- Top of page
- ROLE OF GALECTIN-1 IN ADULT NSCs AND NEUROGENESIS
- GALECTIN-1 IN THERAPY USING NSCs
Neural stem cells (NSCs) in the adult brain have been a consistent focus of biomedical research largely because of their potential clinical application. To fully exploit this potential, the molecular mechanisms that regulate NSCs must be clarified. Several lines of evidence show that a multifunctional protein, Galectin-1, is expressed and has a functional role in a subset of adult NSCs. Researchers, including our group, have explored the physiological role of Galectin-1 in NSCs and its application in the treatment of animal models of neurological disorders such as brain ischemia and spinal cord injury. Here, we summarize what is currently known regarding the role of Galectin-1 in adult NSCs. Furthermore, we discuss current issues in researching the role of Galectin-1 in adult NSCs under both physiological and pathological conditions. © 2012 Wiley Periodicals, Inc. Develop Neurobiol 72: 1059–1067, 2012
- Top of page
- ROLE OF GALECTIN-1 IN ADULT NSCs AND NEUROGENESIS
- GALECTIN-1 IN THERAPY USING NSCs
Since adult neurogenesis was first described by the scientific community (Altman et al.,1965), a great number of studies have been conducted to elucidate its physiology (Alvarez-Buylla et al.,2004; Kempermann et al.,2004; Ming et al.,2011) and potential clinical application (Alvarez-Buylla et al.,2004; Lindvall et al.,2010; Okano,2002, 2010; Okano and Sawamoto,2008). In addition to the continuous regeneration of skin, blood and other cell types (Weissman,2000), specific types of neurons in particular area of the brain are also in a state of continuous addition and/or replacement during adulthood (Ming et al.,2011). The common feature of these regenerating bodily area is that fully functional, mature cells are continuously regenerated from so called somatic stem cells (Weissman,2000). So far, the clearest example of somatic stem cells occurs in the adult hematopoietic system. It has been shown that the transplantation of only one hematopoietic stem cell (HSC) to a mouse that received a potentially lethal dose of radiation resulted in the repopulation of stem cells (self-renewal) and the regeneration of all blood cell types (multipotent differentiation) (Osawa et al.,1996; Matsuzaki et al.,2004). In the adult mammalian nervous system, cells can be reprogrammed artificially (Kim et al.,2008), but single cells that could regenerate the entire nervous system under physiological conditions are not likely to exist. Instead, there is a class of immature cells, located in specific regions of the brain, which can reproduce themselves (self-renewal) and some, but not all, types of neurons and glial cells (oligopotent differentiation) (Ming et al.,2011). The cells that retain these capabilities throughout an animal's life-span could be defined as adult neural stem cells (NSCs). Recently, a study demonstrated that mouse genetic strategies can be used to asses NSCs at the single-cell level in the adult hippocampus (Bonaguidi et al.,2011). Attempts such as these will clarify the definition and full potential of adult NSCs.
Adult neurogenesis occurs predominately in two different niches in the mammalian brain (Ming et al.,2011). One is in the dentate gyrus (DG) of the hippocampus. The NSCs in the DG mainly produce excitatory granule neurons. These mature adult-born granule neurons serve a function in learning and memory, such as in pattern separation, a cognitive process to distinguish similar inputs as independent representations (Aimone et al.,2011; Sahay et al.,2011). Anatomically, the other niche extends from the lateral wall of the lateral ventricle (LV) into the olfactory bulb (OB). The NSCs reside in the subventricular zone (SVZ, or alternatively, the subependymal zone) of the LV. The pathway between the SVZ and OB is connected by a glial sheath surrounding migrating immature neurons called the rostral migratory stream. NSCs in the SVZ mainly produce granule neurons and periglomelular neurons in the OB (Belvindrah et al.,2009). Adult-born inhibitory neurons in the OB are believed to play a role in odor information processing (Lazarini et al.,2011). These two niches of adult neurogenesis are finely regulated by homeostatic mechanisms consisting of a complex network of molecules (Kempermann,2011).
Much effort has been made to identify the molecules that regulate the proliferation of adult NSCs. Early successes in identifying molecules that promote proliferation of HSCs (or hematopoietic progenitor cells) for use in clinical settings (e.g., Granulocyte Colony-Stimulating Factor after chemotherapy) (Dahlberg et al., 2011; Heike et al., 2002) have been one of the primary motivations in the search for molecules that stimulate adult neurogenesis. In our exploration for such molecules, we identified a unique protein, Galectin-1, which promotes proliferation of NSCs in the adult SVZ (Sakaguchi et al.,2006). Galectin-1 is distinct from other molecules in terms of its functional mechanisms (Leffler et al.,2004). Although full classification of its receptors is not yet complete, Galectin-1 apparently binds a variety of different molecules (more than 10 identified so far) in biological contexts that are not necessarily related (Elola et al.,2005). This property of Galectin-1 stems from its ability to bind to carbohydrate moieties, and is also the reason behind the seemingly conflicting experimental results concerning the function of Galectin-1 across different tissues or biological contexts (Leffler et al.,2004; Scott et al.,2004). In this review, we provide an overview of what is known about the function of Galectin-1 in adult NSCs, the molecular mechanisms of this function, and the potential clinical application of Galectin-1.
ROLE OF GALECTIN-1 IN ADULT NSCs AND NEUROGENESIS
- Top of page
- ROLE OF GALECTIN-1 IN ADULT NSCs AND NEUROGENESIS
- GALECTIN-1 IN THERAPY USING NSCs
Details of the two adult neurogenic niches have been described elsewhere (Kempermann et al.,1999; Ming et al.,2011; Temple et al.,1999). Here, we briefly review these two niches to allow better understanding of the expression and function of Galectin-1 in adult neurogenesis.
Adult Neurogenesis in the DG
In the adult hippocampus, NSCs exist in the inner-most layer of the DG granule cell layer (called the subgranular layer (SGL)). The SGL of the adult mouse brain is easily observed by standard light microscopy, and consists of a two or three cell-layer-thick zone lining the inner part of the granule cell layer. Immunohistochemistry and electron microscopy analysis revealed that several cell types reside in the SGL (Ehninger et al.,2008; Filippov et al.,2003). A type 1 cell is characterized by a triangular-shaped soma and radial glia-like processes that extend into the granule cell layer. These radial glia-like type 1 cells, which appear to be NSCs (Bonaguidi et al.,2011; Ehninger et al.,2008; Seri et al.,2001), slowly proliferate and produce progenitor cells. These progenitors undergo several sequential stages of maturation before becoming fully mature granule neurons or glial cells in the DG. One recent study showed that adult-born granule neurons constitute up to 10 % of the inner granule cell layer of the adult mouse DG (Imayoshi et al.,2008).
Galectin-1 Expression in Adult-Born DG Neurons
In the adult mouse DG, Galectin-1 expression is confined to glial fibrillary acidic protein (GFAP)-immunopositive cells in the SGL and granule cell layer, as well as some NeuN-positive mature neurons in the hilus (Sakaguchi et al.,2006; Imaizumi et al.,2011). GFAP-positive cells in the SGL consist of type 1 and 2a neural precursor cells and mature astrocytes (Steiner et al.,2006; Ehninger et al.,2008). Type 2a cells have a nonradial cell shape, a small cell soma, and an irregularly shaped nucleus (Filippov et al.,2003; Ehninger et al.,2008). Morphological analysis combined with multiple immunoshistochemical analyses revealed that although most Galectin-1-positive cells are type 2a cells, 7% of Galectin-1-positive cells match the phenotype of type 1 cells (i.e., morphology and marker expression) (Imaizumi et al.,2011), suggesting that Galectin-1 is expressed in a subset of adult DG NSCs.
Function of Galectin-1 in Adult DG Neurogenesis
5-bromo-2′-deoxyuridine (BrdU) pulse-chase analysis has been widely used to quantify the number of progenitor cells in both the animal and human adult DG (Eriksson et al.,1998). BrdU is a thymidine analog that is incorporated into the DNA of cells when they proliferate (i.e., when DNA is copied). Therefore, BrdU can be used to tag cells that are born during the period of its administration. Since most cells in the intact adult brain do not proliferate, with only minor exceptions (e.g., endothelial cells of blood vessels or microglia), BrdU administration and its subsequent immunohistochemical detection can be used to visualize newly-generated progenitors in the adult brain. BrdU pulse-chase analysis is most powerful when it is combined with multicolor immunohistochemical assays using cell type-specific molecular markers (Imaizumi et al.,2011).
Using this method, our group analyzed the function of Galectin-1 in adult DG neurogenesis using galectin-1 knock-out mice on a C57BL6/J background (C57gal1ko) (Imaizumi et al.,2011). In these mice the total number of putative type 1 NSCs and their progeny was increased compared to their litter mate control mice (C57gal1 heterozygote). More detailed analysis using proliferation and survival markers revealed that loss of galectin-1 resulted in an increase in the proliferation of type 1 and/or 2a cells. These results suggest that Galectin-1 down-regulates early types of progenitor cells in the adult DG. Currently, there is no information available regarding down-stream regulators of Galectin-1 in the DG progenitor cells.
Influence of Genetic Background on the Function of Galectin-1 in Adult DG Neurogenesis
Genetic background has been found to influence adult neurogenesis and hippocampal-dependent behaviors in mice (Kempermann et al.,2002). To date, there have been two independent studies analyzing adult DG neurogenesis in galectin-1 knock-out mice using different strains (Kajitani et al.,2009; Imaizumi et al.,2011). One week after BrdU administration, Kajitani et al. (2009) observed fewer numbers of BrdU-positive cells in galectin-1 knock-out mice (129gal1ko) compared with wild type mice in a 129 background. However, using a similar BrdU administration protocol, Imaizumi et al. (2011) observed greater numbers of BrdU-positive cells in galectin-1 knock-out mice (C57gal1ko) compared with littermate control mice (galectin-1 heterozygotes) in a C57BL6/J background. Since the C57gal1ko mice were derived from back-crosses of the same 129gal1ko mice used by Kajitani et al. (2009), it is suggested that the Galectin-1 loss-of-function phenotype could differ depending on genetic background (Imaizumi et al.,2011).
Why is the Galectin-1 loss-of-function phenotype in adult DG neurogenesis different between the two genetic backgrounds? There are four major factors that can affect the function of Galectin-1 in different biological contexts (Scott et al.,2004). First, there are three distinct functional domains in the structure of Galectin-1 (Leffler,2001)—those that induce cell death, bind to glycans harboring lactosamine sequences (the carbohydrate recognition domain (CRD)), or promote dimerization of Galectin-1 molecules. In different biological contexts, the function of Galectin-1 could depend on the availability of receptor(s) through which any of the domains can function. Second, the counter-receptors to Galectin-1 that harbor lactosamine sequences are variable among cells (Elola et al.,2005). Therefore, even when the carbohydrate epitope to which Galectin-1 binds is the same, the signaling pathway of the molecules that harbor the epitope can be different (Laderach et al.,2010). Third, Galectin-1 can function as a monomer or a dimer (Nishioka et al.,2002). Although the dimer Galectin-1 can bind to two different (or two of the same) lactosamine-rich carbohydrates by divalent CRDs, the monomer cannot. Fourth, oxidative conditions in the extracellular space induce formation of intramolecular Cys-Cys bonds, which change the structure of Galectin-1 and result in the complete loss of Galectin-1's carbohydrate binding activity (Inagaki et al.,2000; Scott et al.,2004). Differences in genetic background could affect any of these factors and thereby alter the function of Galecitin-1 in adult neurogenesis. Similar phenomena in other biological contexts have been previously reported (Adams et al.,1996; Scott et al.,2004; Vas et al.,2005). By considering the above possibilities, future analyses will clarify the mechanisms that result in phenotypic differences between the two backgrounds of galectin-1 knock-out mice.
Function of Galectin-1 in Hippocampal-Dependent Memory
The hippocampus plays a critical role in certain types of memory (Kim et al.,1992). Using transgenic mice, Arruda-Carvalho et al. specifically ablated adult-born neurons before or after mice learned one of three separate hippocampal-dependent memory tasks: contextual fear conditioning (CFC) and two different versions of the Morris water maze (MWM). The authors that postlearning deletion of adult-born neurons deteriorated hippocampal-dependent memory, but prelearning deletion of the same number of adult-born neurons had no effect. This phenotypic contrast between pre- and postlearning deletion of adult-born neurons suggests that adult-born neurons are incorporated into memory circuits (Arruda-Carvalho et al.,2011).
Since C57gal1ko mice showed an increase in adult-born neurons in the DG (Imaizumi et al.,2011), these mice were used to examine the function of Galectin-1 in hippocampal-dependent memory. Interestingly, C57gal1ko mice showed impairments in both CFC and MWM tasks compared with their wild-type litter mates (Sakaguchi et al.,2011). Several reasons for the behavioral deficits have been proposed. First, many studies including our own (Sakaguchi et al.,2006) have confirmed that Galectin-1 is not expressed exclusively in the DG but is also expressed in other brain areas (Sakaguchi et al.,2007; Imaizumi et al.,2011). We measured motor and sensory function and levels of anxiety C57gal1ko mice and observed predominately normal behavior, suggesting that CFC and MWM performance deficits were due to hippocampal-dependent deficits (Sakaguchi et al.,2011). Currently, no evidence is available to clarify whether the memory deficits were caused by an increase in DG neurogenesis or a loss of Galectin-1 in hippocampal neurons and astrocytes. In future studies, site-specific knock-down of Galectin-1 only in adult-born neurons may clarify this issue. Secondly, although decreases in adult DG neurogenesis have been correlated with impaired performance in hippocampal-dependent tasks (Dent et al., 2010), in this case, drawing a correlation between increased adult DG neurogenesis and poor behavioral task performance may not be straightforward. Rather, this finding may just suggest that an increase in the number of new DG neurons is not sufficient to enhance memory or restore memory deficits after loss of Galectin-1. Future analyses will clarify whether adult-born neurons normally integrate into memory circuits in C57gal1ko mice.
Adult Neurogenesis in the SVZ
NSCs also reside in the SVZ (Doetsch,2003; Ming et al.,2011). In the SVZ, NSCs (type B cells) slowly proliferate (self-renew) and give rise to transit-amplifying cells (type C cells) (Doetsch et al.,1997). Type C cells are considered to contribute to an increase in the final number of adult-born neurons (Doetsch et al.,1997). Type C cells eventually exit the cell cycle and produce neuroblasts (type A cells). Type A cells proliferate occasionally and migrate a long distance from the LV to the core region of the OB. After receiving some unknown signal in the OB, type A cells change their direction and migrate toward the surface of the OB (Belvindrah et al.,2009). During the course of this radial direction of migration, they undergo final differentiation into mostly inhibitory granule neurons in the OB.
Galectin-1 Expression in Adult SVZ NSCs
Most adult-born cell types in the SVZ can be identified using cell type-specific molecular markers (Sakaguchi et al.,2007; Imayoshi et al.,2011). Historically, type B cells were identified by their morphology observed by electron microscopy (Doetsch et al.,1997), making it difficult to determine their number and analyze their immunohistochemical properties. Alternatively, type B cells can be identified as slowly-dividing GFAP-positive astrocytes (Doetsch et al.,1999; Nakano et al.,2007). Since GFAP is commonly expressed in nontype B “regular” astrocytes, the visualization of type B cells can be accomplished using GFAP marker staining combined with BrdU pulse-chase analysis. In the pulse-chase analysis, BrdU is administrated for a period during which the NSCs proliferate at least once (1-2 weeks depending on experimental design). Since other proliferating cells (i.e., type C and A cells) also uptake BrdU, a time period called a “wash-out period” is introduced after BrdU administration. During this period, type C and A cells, which incorporated BrdU, start to differentiate and migrate out of the SVZ. Therefore, after the wash-out period, only slowly-dividing cells, which include type B cells, can be detected in the SVZ (Doetsch et al.,1999; Morshead et al.,1994; Nakano et al.,2007).
Several studies have confirmed that Galectin-1 is expressed in a subset of GFAP-positive SVZ astrocytes (Ishibashi et al.,2007; Sakaguchi et al.,2006, 2007). BrdU pulse-chase analysis also confirmed that Galectin-1-positive cells include slowly-dividing cells (Sakaguchi, et al.,2006). These two lines of evidence strongly suggest that Galectin-1 is expressed in adult SVZ NSCs.
Function of Galectin-1 in Adult SVZ NSCs
The physiological function of Galectin-1 in adult SVZ NSCs has been examined by infusion of several variants of Galectin-1 recombinant proteins or function-neutralizing antibody (Gal1ab) into the LV and by analysis of 129gal1ko mice (Sakaguchi et al.,2006). Since Galectin-1 loses its carbohydrate binding ability (lectin ability) by oxidation, a mutant Galectin-1, CS-Galectin-1 recombinant protein (CS-rGal1), which retains lectin ability even under oxidative conditions, was utilized (Inagaki et al.,2000). In addition, because oxidized Galectin-1 can acquire different functions to CS-rGal1 (Sakaguchi et al.,2007), an artificially oxidized recombinant Galectin-1 (Ox-rGal1) was produced for functional comparison (Inagaki et al.,2000). Interestingly, infusion of CS-rGal1, but not Ox-rGal1, into the LV promoted proliferation of adult SVZ NSCs. Importantly, when the native form of rGal1 was infused, the number of adult NSCs was lower than that observed after infusion of CS-rGal1. These results suggest that a portion of rGal-1 changed structure (e.g., oxidation or dissociation as monomers) in the brain tissue after infusion. Furthermore, a decrease in the number of adult SVZ NSCs was observed both after infusion of the neutralizing antibody and in 129gal1ko mice (Sakaguchi et al.,2006). Taken together, these results suggest that Galectin-1 promotes proliferation of SVZ NSCs through its lectin binding ability.
Mechanisms of Galectin-1 Function in Adult SVZ NSCs
There are many molecular players that regulate proliferation, survival, differentiation and other processes of adult SVZ NSCs (Kempermann,2011). Compared to other molecules involved in NSC regulation, the function of Galectin-1 is distinct. First, Galectin-1 can act as a lectin, which binds primarily to carbohydrate structures. Second, Galectin-1 can act in both extracellular and intracellular space. Since Galectin-1 is secreted by a pathway independent of the trans-Golgi pathway, Galectin-1 can also act in intracellular space without its secretion from cell the membrane (Liu et al.,2002). Third, Galectin-1 can lose its property as a lectin and gain different functions as discussed above (Camby et al.,2006). These properties of Galectin-1 are rarely seen in typical protein receptor-binding molecules.
Since infusion of Gal1ab decreases the number of SVZ NSCs (Sakaguchi et al.,2006), we examined the function of Galectin-1 in extracellular space. In the adult SVZ, we found CS-rGal1 bound to type A and E cells and, in rare cases, putative type B cells (i.e., GFAP-positive SVZ astrocytes) but not striatal or cortical neurons adjacent to the SVZ (Sakaguchi et al.,2010). This suggests that Galectin-1 can bind to multiple cell types in the SVZ. Subsequent biochemical analysis using CS-rGal1 affinity columns suggested that Galectin-1 can bind to the β1 subunit of Integrin (β1 Integrin). β1 Integrin is expressed mostly in type A cells and activated NSCs (Kazanis et al.,2011). Considering that Galectin-1 bound to other molecules in the biochemical analysis, and β1 Integrin expression partially covers the area where CS-rGal-1 binds (Sakaguchi et al.,2010), there should be molecules other than β1 Integrin to which Galectin-1 binds in the adult SVZ. Nevertheless, functional analysis using Gal1ab and β1 Integrin neutralizing antibodies, and rGal-1 in vivo and in vitro, suggests that Galectin-1 plays a role up-stream of the β1 Integrin signaling pathway in the adult SVZ (Sakaguchi et al.,2010). In future analyses, it will be necessary to identify other binding partners of Galectin-1 in the adult SVZ to understand the dynamic role of Galectin-1 not only in NSCs but also in other SVZ cell types.
Future Studies of Galectin-1 in the Adult SVZ
Previous biochemical studies showed that CS-rGal1 binds to carbohydrate moieties, which are rich in lactosamine sequences (Hirabayashi et al.,2002). Consistent with the localization of lactosamine harboring molecule LeX-epitope (Capela et al.,2002; Katagihallimath, et al.,2010), the binding of CS-rGal1 is limited to the the SVZ (Sakaguchi et al.,2010). Future studies using emerging techniques to identify the exact sequence of the carbohydrate moieties and their core molecules from tissue samples, such as lectin-affinity capture (Kaji et al.,2003) or lectin array (Kuno et al.,2005), will be essential to understand the regulatory mechanism of Galectin-1 in the SVZ (Sakaguchi et al.,2007).