The Neurosphere Culture System

After the initial reports of free-floating cultures of neural stem cells termed neurospheres (1,2), a wide array of studies using this promising culture system emerged. In theory, this was a nearperfect system for large-scale production of neural cells for use in cell replacement therapies and to assay for and characterize neural stem cells. More than a decade later, after rigorous scrutiny and ample experimental testing of the neurosphere culture system, it has become apparent that the culture system suffers from several disadvantages, and its usefulness is limited for several applications. Nevertheless, the bulk of high-quality research produced over the last decade has also shown that under the right circumstances and for the appropriate purposes, neurospheres hold up to their initial promise. This article discusses the pros and cons of the neurosphere culture system regarding its three major applications: as an assay for neural stem cells, as a model system for neurogenesis and neural development, and for expansion of neural stem cells for transplantation purposes. Index Entries: Neurosphere; in vitro; neural stem cells; regional specification; transplantation; clonal analysis; neuron; glia. Received May 23, 2006; Accepted July 4, 2006. *Author to whom correspondence and reprint requests should be [email protected]. B27 supplemented medium containing fibroblast growth factor (bFGF) and/or epidermal growth factor (EGF) (2,4–7). One of the major problems with the neurosphere culture system is that it appears sensitive to the culturing method used. Variations in cell density alter the microenvironment, which in turn may affect both the proliferation capacity (6) and the positional cues that the cells are exposed to (8). Different constituents or concentrations of factors in the media (9–11), method and frequency of passaging (10), whether the neurosphere is dissociated before differentiation (10), and the number of passages after isolation (12) also lead to differences in both the composition of cell types as well as the properties of the cells within each neurosphere. Because of the sensitivity for such variation in culture method, it is often hard to consolidate data from different groups (8,13) or even to interpret the results within the same study (14) to make comprehensive conclusions about the properties of the cells within the neurospheres. Another difficulty with the system comes from the inherent properties of suspension cultures: the neurospheres represent “little black boxes” that you cannot peer into, making it hard to carefully monitor the properties of individual cells during the culture period. The neurogenic capacity of the neurosphereexpanded cells declines after extended numbers of passages (15). It is not clear whether this is because progenitors with capacity to form neurons are outnumbered by glial progenitors and thus progressively lost or whether the intrinsic properties of the cells within the neurospheres change their properties over time. Furthermore, the neurospheres are heterogeneous in nature, and only a small percentage of cells within each sphere holds the neurosphere-forming capacity (2); even fewer fulfill the criteria of being neural stem cells (16). Each neurosphere contains cells at various stages of differentiation (Fig. 1), including stem cells as well as proliferating neural progenitor cells and postmitotic neurons and glia (17,18). This heterogeneity increases with sphere size because more differentiated cell types arise after longer time in culture. To monitor the presence of neural stem and progenitor cells in the neurospheres, we generated neurospheres from mice that expressed green fluorescent protein (GFP) from the Sox1 locus (19), enabling identification of neural stem and progenitor cells by their GFP expression (20). We found that even small neurospheres contain a mix of GFP-positive and -negative cells shortly after passage (M.P., unpublished observation). Therefore, even at a very early stage, the neurospheres appear heterogeneous. This heterogeneity makes it difficult to study any specific event or cell in isolation, and studies using neurospheres should be seen and interpreted as studies on a mixed population of precursor cells and not as studies of neural stem cells. The Neural Stem Cell Assay The neurosphere culture system was the first in vitro system to unequivocally demonstrate the presence of cells in the adult brain with characteristics of true neural stem cells (1,2) and remains an extremely useful tool to analyze proliferation, self-renewal capacity, and multipotency of neural stem and progenitor cells. Testing for neurosphere-forming capacity over serial clonal passaging (Fig. 2A) followed by in vitro differentiation to show multipotency of individual spheres (Fig. 2B,C) is widely used and, when performed properly, provides the best functional assay for neural stem cells available (4,6,21,22). The system is particularly attractive because clonal analysis can be performed with ease compared to adherent culture systems where, in the past, complicated labeling methods had to be employed to determine clonal relation between a cell and its progeny (23). Recently, a new adherent monolayer culture system has been reported that allows for serial clonal expansion of neural stem cells (24), and this culture provides a relatively uncomplicated alternative assay for neural stem cells. 154 Jensen and Parmar Molecular Neurobiology Volume 34, 2006 Neurospheres as a Physiologically Relevant Model System In addition to its function as a neural stem cell assay, the neurosphere culture system is a valuable in vitro model system to study neurogenesis and neural development. Most studies on the genetic and molecular control of regional specification of neural precursors have been performed in vivo. Complimentary in vitro approaches are useful to determine the degree of intrinsic specification present in neural precursors at various developmental time-points, as well as to study the full potential of the cells when removed from extrinsic cues provided by their normal environment. The neurospheres are a good system for such studies because they are maintained under defined serum-free conditions where the environmental cues are limited to those of surrounding cells. Additionally, it is easy to manipulate the extrinsic cues the cells are exposed to during their development by changing the environment during either the expansion or differentiation phase. This can be done by simply adding precise and variable amounts of factors of interest to the media (25–27) or by culturing the neurospheres together with other types of cells, such as primary cells from the developing CNS (26,28,29). Neurosphere Cultures 155 Molecular Neurobiology Volume 34, 2006 Fig. 1. Neurospheres are heterogeneous in nature with RC2 (B,C) and nestin-positive (D,F) progenitor cells generally located toward the outside of the sphere and GFAP-positive (A,C) cells in the center of the spheres. Few βIII-tubulin-positive neurons (E,F) can also be found evenly distributed within some spheres. (Figure modified from ref. 17.)