To select the backbone biomaterial, the main consideration should be based on the fact that the brain tissue is one of the softest tissues in the body. In this regard, many natural biomaterials or synthetic polymers have been examined as backbone structures to mimic lower mechanical properties. Table 2 provides a list of commonly used natural and synthetic biomaterials in neural stem cell studies, each of which has some pros and cons required to be considered [refer to review paper (Doblado et al., 2021)]. Natural biomaterials can be derived from either ECM proteins or the polysaccharides produced in other organisms. They are popular because of their biologically active nature and their excellence in promoting cell adhesion and growth (O’Brien, 2011). Moreover, natural biomaterials are biodegradable and are easily remodeled by cell-mediated proteases such as matrix metalloproteinases, allowing cells to deposit their own ECM and grow (Kamatar et al., 2020; Caliari and Burdick, 2016). Many natural biomaterials have direct analogues in the human body, reducing the risk of cytotoxic reactions (Boni et al., 2018). Although natural biomaterials have been widely used due to their advantages, certain drawbacks can still make their use limited. First, most of the natural-based biomaterials are large and complex proteins making them hard to be produced through a recombinant expression system. Therefore, they are mainly purified from an animal- or human-derived ECM and have the risk of pathogen and immunogen contamination and lot-to-lot variability. A lack of control over their mechanical properties, thermal sensitivity, and induced biological signals due to complex chemical structures make their broad use for a reductionist’s view more challenging. Synthetic polymers have several advantages over natural ones, including the potential to maintain a range of mechanical (e.g., stiffness) and biochemical (e.g., ligand concentration) properties. They can also be designed in different forms, including bulk hydrogels, microgel-based scaffolds, and fiber-aligned polymer scaffolds to provide different types of topographical features for inducing favorable responses. One key drawback, however, is that most synthetic materials lack biological signaling capability. Therefore, they need to be modified with natural bioactive materials to promote cell adhesion, activation, and to ultimately be remodeled by cells (Kamatar et al., 2020). The backbone biomaterial is usually selected based on the application under investigation and choosing between natural and synthetic biomaterials depends on the priorities of each study. Because of the complex nature of the natural tissue, many structural and compositional improvements are still needed to make both natural and synthetic biomaterials better candidates for neural stem cell studies.

Degradability is one of the important factors to be employed if the scaffold is to provide support for cells to grow, as degradation is required to make space for proliferating cells. While natural biopolymers typically have sequences within their backbone that are broken by enzymes such as matrix metalloproteinases within the body, synthetic polymers must be designed with degradability in mind. Researchers have long investigated co-polymerization with a biodegradable polymer, such as PLA or PGA, to add degradability to otherwise non-degradable polymers. More recently, various groups have shown that matrix metalloproteinases can be incorporated into synthetic polymer systems to support cellular degradation (Barros et al., 2019) (Lutolf et al., 2003) (Lévesque and Shoichet, 2007). For applications to neural systems, the addition of matrix metalloproteinases-sensitive cleavage peptides to PEG-heparin-hydrogel system supported the generation of elaborate neuronal network while modeling Alzheimer’s disease (Papadimitriou et al., 2018). The advantage of using naturally occurring proteinase-sensitive materials is that the degradation of these crosslinkers relies on cell demand, so using them allows tuned rates of degradation based on tissue remodeling. In the design process, both the rate of biodegradability and degraded byproducts should be considered since accumulation of degraded components can instigate high astrocyte response, resulting in the glial scar formation and inflammatory response (Bjugstad et al., 2010). Furthermore, the byproducts can induce new signals to stem cells influencing their function and fate decisions (Murphy et al, 2014). In addition, the mechanical properties of the environment where the cells are encapsulated varies dynamically as the material degrades, and an important design consideration is a balance between providing space for cell growth and vascularization and the strength of the scaffold (Peressotti et al., 2021). Therefore, decoupling the cellular infiltration and scaffold’s mechanical properties is one of the main challenges that engineers have faced when using degradable biomaterials.

Mechanical properties of a biomaterial are key factors that need to be intentionally designed into scaffolds to mimic neural stem cell niche. Failure of tissue integration due to mechanical mismatch between a biomaterial and its surrounding tissue has been well described in other systems such as bone or cardiovascular tissue (Soliman et al., 2022; Boccafoschi et al., 2017), and when selecting materials for use, it is an important design constraint here due to the brain’s soft nature. Biomechanical measurement of brain tissue is challenging due to its softness, leading to inconsistent findings by comparing different reports; the properties of the brain vary with age, sex and disease (Budday et al., 2017). Recently, the human mechanical properties could be measured non-invasively using magnetic resonance elastography (Hiscox et al., 2016). Hiscox et al. (2016) reviewed forty-one studies using magnetic resonance elastography to measure the brain shear stiffness of healthy and patient participants with an age range between 16 and 94 years old. They reported the mechanical properties of human brain as a whole with a range of shear stiffness from .62 kPa to 2.99 kPa with a mean shear stiffness of 2.07 kPa ±.42 kPa at actuation frequency of 50 Hz in healthy participants and showed significant changes in shear stiffness due to different types of neurological disorders [for full review, see (Hiscox et al., 2016)]. Therefore, mechanical properties of a biomaterial used for brain tissue regeneration should then match the native brain tissue in healthy young—and not pathological—conditions to match the desired response of the niche appropriately. In this respect, Saha et al. (2008) demonstrated that soft substrates promote neuronal differentiation while stiffer ones supported glial differentiation for the first time. Following that report, Leipzig and Shoichet (2009) reported a consistent result that soft substrates with E < 1 kPa enhanced neuronal differentiation; medium-stiff substrates with 1 kPa < E < 3.5 kPa favored astrocyte differentiation, and stiff substrates with E > 7 kPa promoted differentiation toward the oligodendrocytes. Furthermore, using PEG-based hydrogels decorated with RGD, Stukel et al. found that softer scaffolds (.1–.8 kPa) provided a better platform for neural stem cell differentiation and neurite extension compared to stiffer ones (4.2–7.9 kPa) (Stukel and Willits, 2018). In addition, Segel et al. (2019) found that aged oligodendrocyte progenitor cells (OPCs) could be rejuvenated when cultured on young-mimicked stiffness scaffolds. These types of studies highlight the importance of scaffold mechanical properties in lineage decisions. Therefore, both investigation into changes in vivo under pathological conditions and designing biomaterials with tunable properties are top priorities in harnessing the therapeutic potential of neural stem cells (Zhu M et al., 2019; Zhu Y et al., 2019).

Like mechanical properties, cells sense the physical environment, e.g., topography and architecture, and are therefore important when studying designing scaffolds for neural stem cells. Flat, 2D cultures have long been studied in vitro; however, they are not representative of in vivo conditions which are inherently 3D. A 2D system could affect the cell response since the distribution of cues that cells receive is different in 2D culture than 3D culture. Therefore, the ideal in vitro model is 3D model in which the cells are encapsulated, as it better mimics the natural niche of neural stem cells. However, both 2D and 3D cultures can have different spatial and physical features from the nanoscale to microscale in a variety of shapes including plain, fibrous, or spherical structures with rough or smooth characteristics.

One method of introducing topography into this system is through fiber formation. Fibers can be formed by various techniques and used for neural tissue engineering [for full review, see (Uyar Tamer, 2017)]. However, the diameter (Christopherson et al., 2009) and alignment (Amores de Sousa et al., 2020) of fibers has influenced the neural stem cell response. For example, higher neuronal differentiation occurs on larger fiber diameters compared to smaller fiber diameters with laminin-coated polyethersulfone fibers (Christopherson et al., 2009). Without the presence of any growth factors, electrospun blended-polycaprolactone fiber mats induced higher levels of glial or neuronal differentiation of the neural stem cells depending upon the hydrophilicity of the mats (Fernández et al., 2019), so care should be taken to fully evaluate the material properties of samples. The next step for fibers is to be able to functionalize with molecules that instruct neural stem cell response. For example, fibers modified with adhesive RGD ligands were designed to guide neural stem cell distribution and promote higher cell elongation leading to neuronal differentiation (Amores de Sousa et al., 2020). In addition, spatial control over peptide or protein functionalization was achieved to increase and direct cell migration, regulate neural differentiation, and gradient formation, allowing for increased control over how bioactive cues are presented (Silantyeva et al., 2019; Cavanaugh et al., 2019; Motta et al., 2019; Silantyeva et al., 2018). Manufacturing systems that regulate both the topographical nature of the fibers at scale, as well as the bioactive moiety placement, are not broadly available and need more engineering development to have a wider impact.

Another common method to introduce topography is by fabricating granular platforms. Granular scaffolds offer many favorable properties, making them appealing in neural stem cell studies. These materials offer the ability to piece together different structures like Lego^® bricks to regulate the properties of the scaffold. From a topographical view, the granular structure is thought to better mimic the brain structure (George et al., 2020). One such granular construct is the microgel-based scaffold, which uses micro-sized hydrogels as the basic brick to then link together to form a larger structure (Figure 7) (Zhou et al., 2016). The microgels are built and modified chemically and mechanically prior to forming the desired structure and then packed with/without encapsulated cells for their final applications (Scott et al., 2010b; Scott et al., 2011). They can be formed either in an array, film, or 3D structure to provide topographical cues to cells (Riegert et al., 2021). Wilson et al. (2022) showed that engineered microgels can serve to promote the neural stem cell’s neurogenesis or help them to keep their multipotency depending on how they are post-modified with the laminin-derived peptides. Because of the granular nature of the system, the environment of microgel-based biomaterials can either be homogenous, providing encapsulated cells similar signals throughout the scaffold, or can offer heterogeneous microenvironments to provide spatial regulation.

In addition to providing topographical cues, microgel-based scaffolds have been shown to have adjustable properties in all important features talked above to develop an appropriate scaffold for neural stem cell studies. These granular hydrogels can be easily fluidized while compressed because of shear thinning phenomena (Riley et al., 2019). Injectability is one of the aspects that engineers are working to improve. Although solid-like hydrogels are ideal for many aspects of brain tissue regeneration ex vivo, injectable and self-gelling systems are critical design parameters for use in vivo. In this regard, hyaluronic acid-based microgels were injected to the cavity post-stroke and found to reduce inflammation, astrocyte infiltration, and microglia activation. These changes made the environment suitable for triggering neural stem/progenitor cell migration toward the stroke cavity (Nih et al., 2017). They also studied the neural progenitor cell fate in these types of scaffolds designed with adhesive and non-adhesive laminin peptides; the first one led to the cell’s spread, migration and differentiation and the later one led to neurosphere formation, maintaining their level of stemness (Wilson et al., 2022). The microgel-based scaffolds also could be used as an alternative to solve the decoupled degradability and mechanical properties (Scott, Marquardt and Willits, 2010a; Scott et al., 2011; Coronel et al., 2022). Xin et al. (2018) demonstrated that microgel scaffolds, even without degradability potential, work better for cellular spreading comparing conventional bulk hydrogels, even with a degradable crosslinker peptide. The porous structure provides equal soluble nutrient accessibility and vascularization that make these microgels suitable for mimicking the native niche. Furthermore, the mechanical properties of microgel-based scaffolds are a result of the continuous phase and are usually determined by the polymer type, density and level of crosslinking (Daly et al., 2020). Combining these properties, plus the post-modification capability of the microgels, granular scaffolds are now considered as one of the most promising platforms as they meet most of the requirements to recapitulate the native tissue patterns. With the current technologies, we can only generate simple tissue patterns that replicate early tissue development; therefore, engineers need to expand the design technologies to make the scaffolds better meet all of the requirements for mimicking the natural tissue in vitro.

While designing scaffolds for neural stem cell studies, another important aspect is to make these scaffolds bioactive and enable them to support cell adhesion, growth and differentiation. The following section will delve more into the bioactive components that are currently common.

Laminin is the most commonly studied ECM protein in the SVZ niche serving as an adhesive site. Therefore, adding naturally-derived, recombinant, or synthetic versions of laminin to biomaterial scaffolds can mimic the natural tissue by providing links for laminin receptor- or integrin-mediated cell adhesion. Laminin is not often used as a 3D scaffold by itself, but it has been used a lot with other natural ECM components (Willerth and Sakiyama-Elbert, 2019). Laminin 111 is the most studied type of laminin in neural stem cell studies since it can be easily isolated and purified from Engelbreth-Holm-Swarm (EHS) mouse sarcoma (Tate et al., 2009; Addington et al., 2015; Distler et al., 2021). There are also recombinant versions of laminin 111 protein, but their use is not still that common in neural stem cell cultures. Although natural-derived laminin 111 has been widely used for primary studies due to price and availability, laminin 111 is not appropriate for mimicking many disease models since it is only expressed in the CNS during embryogenesis, and it disappears progressively from basement membranes during development. Instead, laminin 511 and laminin 521 become ubiquitous during adulthood (Barros et al., 2020). Therefore, laminins 511 and 521 should be the preferred laminins to be utilized in engineered scaffolds for studying adulthood disease models. However, apart from mouse-derived laminin 111, other laminins, including laminin 511 and 521, are almost impossible to be isolated in their native form (Rodin et al., 2010). Although recombinant full-length laminins 511 and 521 are commercially available, they are large and difficult to be produced recombinantly. To overcome this challenge, the E8 fragment of laminin 511 and 521 is a functionally shorter form of laminins that supports integrin-mediated cell adhesion, enabling engineers to study the effect of niche laminin components. In this regard, Silva et al. (2017) showed that laminin 511 could support neural stem/progenitor cell adhesion and migration better than laminin 111 and laminin 511-based differentiation methods are used to produce clinical quality dopaminergic neurons to be used for Parkinson disease treatment (Yap et al., 2019). Although laminin 511 has been shown to support neurons in 2D (Zhang et al., 2017; Rodin et al., 2020; Zekonyte et al., 2016; Doi et al., 2020), 3D scaffolds using laminin 511 and 521 are not commonly examined for neural stem cell studies. Therefore, we have a gap in knowledge about neural stem cell-laminin 511 and 521 interaction in a mimicked niche that needs to be filled by designing new advanced scaffolds based on these proteins.

In addition to recombinant proteins, chemical peptide synthesis is available to generate binding sites and promote cell adhesion. Short laminin peptides are stable, easily synthesized, and target-specific (Li et al., 2014). Some of the laminin peptides that are used to interact with neural stem cells are listed in Table 3 [see review paper (Mukherjee et al., 2020), for more information]. These peptides are derived from different laminin chains, but there are few of them derived from laminin α₅ chain of laminin 511 and laminin 521 that are the laminin of interest in the SVZ niche (AGQWHRVSVRWG and TWSQKALHHRVP). Huettner, Dargaville and Forget (2018) reports that 89% of biomaterials are functionalized with RGD, 6% with IKVAV and 4% with YIGSR, highlighting the need to extend studies on desirable peptides (derived from laminin α₅, laminin β₁, laminin β₂, and laminin γ₁) and discover more bioactive peptides to mimic the neural stem cell niche better.

Despite its low abundance in adult SVZ, collagen is another ECM protein that can serve as a regulator for neural stem cell’s fate decisions. Since collagen I and collagen IV are the collagen types available in fractones and basement membranes, they could be studied to mimic the neural stem cell niche. In contrast with laminin, collagen is used both as a 3D structure backbone and as the added bioactive component. However, like laminin, there are natural, recombinant and synthetic versions of both collagen I and collagen IV. The whole-length animal-derived collagen I is the most common type of collagen to be used for in vitro models. However, Vagaska et al. (2020) found that collagen I alone is not sufficient for human neural stem cells to adhere and survive. They suggested that human neural stem cells do not express collagen I binding subunits, so other factors present in basement membranes including laminin, collagen IV, nidogen and perlecan are possibly required for human neural stem cells adhesion, survival and proliferation (Vagaska et al., 2020). In this regard, Bergström et al. (2014) confirmed that mRNA pattern encoding collagen I-binding integrins differs from postnatal to adulthood, especially integrin α1, α2 and integrin α11 and that is why adult human-derived neural stem cells cannot adhere to the collagen I-treated substrates while postnatal ones can. In studying the vascular basement membrane protein assembly for the neurogenic niche, Kim et al. (2021) illustrated that progenitor cells were less likely to adhere to collagen IV-overlaid laminin substrates versus laminin-overlaid collagen IV, suggesting the importance of laminin recognition within the niche. In pathological conditions, investigating laminin and collagen IV organization could be something interesting to study; however, there are still very limited studies on the effect of collagen IV on neural stem cell behavior. Although collagen does not seem to provide adherence sites for neural stem cells, understanding its effect on neural stem cell fate in concert with other matrix proteins or properties would give valuable information about its regulating role. Que et al. (2018) developed a recombinant collagen III scaffold decorated with integrin-binding peptide sequence from collagen I (GFOGER [Gly-Phe-Hyp-Gly-Glu-Arg]) and from laminin α₁ (IKVAV [Ile-Lys-Val-Ala-Val]). They found that α₁β₁ integrin-mediated adhesion of fetal-derived neural stem/progenitor cells to the GFOGER peptide supported cell proliferation and differentiation toward both glial and neuronal lineages, but it was not clear if it works with adult-derived neural stem cells whose collagen binding receptors are not expressed through adulthood. Examining materials and proteins via high-throughput studies may provide further insight into the effects of interactions of the ECM with aging neural stem cells.

Added to laminin and collagen, HSPGs are considered as influential factors present in SVZ niche that can interact with other ECM proteins and contribute to neural stem cell regulation [for further information, see review by (Yu et al.t, 2017)]. HSPGs have been widely incorporated to biomaterials by taking advantage of high affinity between heparin and growth factor for tissue repairs (Sakiyama-Elbert, 2014). For example, Sakiyama-Elbert group found that delivery systems incorporated with the heparin and nerve growth factors help nerve regeneration (Wood et al., 2010). Perlecan and agrin are HSPGs present in both fractones and basement membranes that serve as mediators between laminin, collagen IV, growth factors and cell receptors. Therefore, in culture, HSPGs are used in combination with either laminin, collagen or growth factors. The combination of heparan sulfate with collagen as a scaffold for neural stem cells was used to investigate its effect on nerve regeneration and restoration after traumatic brain injury (Zhang and Wang, 2021). They found better neuronal recovery in the groups loaded with neural stem cells and suggested that the interaction between neural stem cells, collagen, heparan sulfate and FGF acted as anti-fibrosis and anti-inflammation inhibiting astrocyte activation and promoting neuronal differentiation (Zhang and Wang, 2021). A key role of heparan sulfate was in the stabilization of FGF2-FGF receptor complex to regulate neural stem cell proliferation and differentiation through activation of phosphorylation of extracellular regulated kinase 1 and 2 (pERK1/2) signaling (Latchoumane et al., 2022). Heparan sulfate offers great potential in biomaterials to mimic interactions with both ECM and growth factors within the neurogenic niche and improve neural stem cell response. Future work to expand our knowledge about the interplay of how these interactions affect neural stem cell fate will be critical to better understanding engineering design considerations to mimic the niche.

The final factor to discuss that adds bioactivity to these material scaffolds are cells from the SVZ, in addition to neural stem cells. In this regard, the most studied cells are endothelial cells, given the vascular changes that occur with age and pathology in the SVZ niche. Sosa et al. (2007) found that endothelial cells influence neural progenitor cells differently while contacting them directly or indirectly. Through indirect contact, an endothelial cell’s secreted factors (e.g., VEGF) helped support neural progenitor cell maintenance and proliferation, while direct contact led to neuronal differentiation (Sosa et al., 2007; Sun et al., 2010). Winkelman et al. (2021) reviewed the literature on neural stem cell and endothelial cell interactions and discussed that Ephrin-B2 and Jagged1 molecules on endothelial cells bind to Eph and Notch receptors located on neural stem cells to inhibit their proliferation. Another example of providing niche residing cells to neural stem cells is the co-culture of microglia and neural stem cells positively affecting neural stem cell differentiation toward dopaminergic neurons, which is the potential therapeutic for Parkinson’s disease treatment (Schmidt et al., 2021); co-culture with indirect contact provided higher yield of TH⁺ neurons possibly because of secreted factors (TNFa, IL-1b, and IGF1). However, they mentioned that different microglia cell lines may result in differentiation toward different neuronal subtypes like β-tubulin III⁺ neurons that should be considered (Schmidt et al., 2021) as future design considerations.