Interleukin-10 regulates progenitor differentiation and modulates neurogenesis in adult brain

The subventricular zone (SVZ) of the lateral ventricle (LV) is the main neurogenic area in the adult murine brain. The neural stem cells (NSCs, Type B cells) present in the SVZ are astrocytes characterized by long cell cycles and GFAP expression. NSCs give rise to rapid transit amplifying dividing cells (TAPs) or Type C cells (GFAP^–, EGFR⁺, Ki67⁺ and Mash1⁺). Most cells that arise in the SVZ generate neuroblasts (Type A cells: PSA^–, NCAM⁺, DCX⁺, TUBB3⁺), which migrate over long distances through a rostral migratory pathway (RMS) to reach the olfactory bulb where they finally become mainly GABAergic interneurons (Doetsch and Alvarez-Buylla, 1996; Doetsch et al., 1997; Lois and Alvarez-Buylla, 1994; Merkle et al., 2004). Also, under physiological conditions, oligodendrocytes (Olig2⁺, Ng2⁺) and a small number of glutamatergic neurons (Tbr2⁺, vGlut2⁺) are generated in the SVZ (Brill et al., 2009; Gonzalez-Perez et al., 2009; Marshall et al., 2005; Menn et al., 2006; Parras et al., 2004). Nestin is a cell marker for NSCs, but it remains expressed in TAPs (Mash1⁺) and immature committed neuroblasts (TUBB3⁺) (Doetsch et al., 1997). The adult SVZ is an extensive area where proliferation and neurogenesis occur along the entire rostrocaudal and dorsoventral axes (Mirzadeh et al., 2008). Experiments with viral or genetic lineage tracing indicate that the SVZ is arranged as a mosaic. The position of NSCs within the SVZ determines the types of differentiated progeny generated. In particular, deep granule interneurons and a subpopulation of periglomerular cells arise from the ventral SVZ, whereas superficial granule interneurons, distinct periglomerular cells, and glutamatergic cells are derived from the dorsal SVZ (Brill et al., 2009; Ihrie et al., 2011; Kohwi et al., 2007; Young et al., 2007; Merkle et al., 2007). Progenitors generated along the entire ventral and dorsal regions of the SVZ move from the walls of the LV by means of a series of complex migratory movements to reach more anterior dorsal SVZ regions and form complex interconnected chains to incorporate into the RMS and reach the OB (Kakita and Goldman, 1999; Geraerts et al., 2006; Sawamoto et al., 2006; Nam et al., 2007). Postnatal neurogenesis is regulated by physiological and pathological conditions, and the integration of extracellular signaling activation and the presence of some intrinsic mechanisms lead to cell fate determination and lineage differentiation (Guillemot, 2007; Lim et al., 2009). The identity and lineage of stem cells has been elucidated; however, the regulatory mechanism underlying stem cell renewal, survival and differentiation in vivo are not fully understood (Aguirre et al., 2010; Androutsellis-Theotokis et al., 2006; Andreu-Agulló et al., 2009; Ramírez-Castillejo et al., 2006; Fernando et al., 2011; Khodosevich et al., 2010; Mobley et al., 2009; Rabaneda et al., 2008; Romero-Grimaldi et al., 2012; Teixeira et al., 2012). NSC self-renewal is modulated by several factors, such as VEGF and PEDF, which are released by ependymal and endothelial cells, and the Notch mechanism plays a crucial regulatory function in this process (Aguirre et al., 2010; Andreu-Agullo et al., 2009; Androutsellis-Theotokis et al., 2006; Marqués-Torrejón et al., 2013; Ramírez-Castillejo et al., 2006). Although these events have been extensively studied in the ventral part of the LV, the regulatory mechanisms acting on progenitor cell proliferation and differentiation in the dorsal SVZ before these cells enter the RMS to regenerate the OB are unclear.

IL-10 is a general anti-inflammatory molecule that is present at low levels in circulating blood. It exerts key functions for the maintenance of the pro- and anti-inflammatory balance in the body to prevent inflammatory and autoimmune pathologies (Ouyang et al., 2011; Pestka et al., 2004; Saraiva and O'Garra, 2010). Here we demonstrate a new physiological role for this cytokine. We found that IL-10 activity targeted the whole heterogeneous population of Nestin⁺ progenitors, generated along the entire SVZ, when these cells were present in the dorsal SVZ region to later enter the RMS. When Nestin⁺ progenitors were present in dorsal SVZ, IL-10 upregulated the expression of undifferentiated neural progenitor cell markers, the number of progenitors in the active cell cycle and the inhibitory phosphorylation of CDC2. Subsequently, neuronal differentiation was impaired and the number of postmitotic neuroblasts reduced.

IL-10 signals through two membrane receptor chains, IL-10R1 and IL-10R2 (Pestka et al., 2004). IL-10R2 is widely expressed in the brain. By contrast, IL-10R1 expression is low in normal brain, and the activation of this subunit is responsible for STAT3 docking, which mediates intracellular signaling (Pestka et al., 2004; Gonzalez et al., 2009). Here we focused on the expression of IL-10 and its receptor in the SVZ in rat and mouse brain. A similar expression pattern was detected in both species. Regular RT-PCR showed the presence of IL-10R1 transcripts in postnatal (P) and adult (Ad) SVZ tissue, as well as in primary cultures of postnatal SVZ (supplementary material Fig. S1). IL-10 mRNA was expressed in cell cultures, and low levels of this molecule were detected in brain samples (supplementary material Fig. S1). In order to determine the spatial distribution of IL-10R1 in the whole SVZ, we performed histological studies on a series of rostrocaudal coronal sections from adult brain. IL-10R1 expression was detected from dorsal to ventral sub-regions of the SVZ (Fig. 1A) and in migrating cells (Nestin⁺) in the RMS (Fig. 1B), whereas no staining was detected in the OB (data not shown). A representative drawing of a coronal section in Fig. 1 indicates the ventral and dorsal regions analyzed. Confocal imaging showed that, in the entire SVZ, IL-10R1 is expressed in the cell monolayer that limits the LV, where ependymal cells are located (Fig. 1A, arrows). Double-immunofluorescence analysis corroborated the presence of IL-10R1 in cells positive for S100B, a marker for ependymal cells in dorsal and ventral SVZ regions (Fig. 1C,D). In ventral areas, Nestin⁺ progenitors grouped in patches in a second line of cells next to the open ventricle and did not express IL-10R1 (Fig. 1C, arrow). By contrast, in the dorsal region of the SVZ, IL-10R1 staining was observed in a wide lateral zone next to the open ventricle, and this staining matched the region where Nestin⁺ progenitors were located (Fig. 1D). High-magnification confocal pictures demonstrated the presence of IL-10R1 on the cellular limits of Nestin⁺ cells (Fig. 1D, arrowheads, bottom panels).

Double-labeling confocal images were analyzed on coronal sections containing the subgranular zone (SGZ) of the hippocampus, the second neurogenic niche of the adult brain. Staining revealed that Nestin⁺ (red) cells and DCX⁺ neuroblasts (red) do not express IL10-R1 (green) (supplementary material Fig. S2), thereby suggesting that IL-10 does not act on the adult hippocampal neurogenic niche.

To identify the SVZ cell phenotype expressing IL-10R1, we analyzed the expression of this protein in dissociated cultures of dorsal regions of the whole SVZ at high confocal resolution. IL-10R1 colocalized with the membrane-associated protein Spectrin, thereby indicating that the IL-10 receptor is present in cellular membranes, as expected for an active receptor (Fig. 2A). Immunofluorescence analysis also demonstrated the presence of IL-10R1 (green) in most of the cultured cells (79.5±4.3%). Double-immunofluorescence analysis with cellular markers revealed that nearly all Nestin⁺ progenitors were IL10R1⁺ (96.89±1.6%; P<0.01; Fig. 2A,B) and most TUBB3⁺ committed neuroblasts (red) expressed IL-10R1 (79.3±% 5.0; P<0.01; Fig. 2A,B).

The unexpected IL-10R1 expression pattern described here points to a new unknown biological activity for IL-10 in the adult SVZ neurogenic niche. Nestin is a marker of NSCs but its expression remains in intermediate progenitors, including a sub-population of immature neuroblasts (Doetsch et al., 1997). Nearly all Nestin⁺ progenitors obtained from the whole dorsal SVZ expressed Il-10R1; this observation is compatible with IL-10 signaling in a majority and heterogeneous population of progenitors at different steps of differentiation. This suggests that when these cells are located dorsally and entering the RMS, IL10 signaling is involved in the modulation of adult progenitor neurogenesis.

To explore the effects of IL-10 on NSCs, we performed neurosphere assays and assessed the effects of this molecule on the stem cell population (Ferron et al., 2007). It should be taken into account that a number of TAPs also develop as neurospheres in this culture model (Pastrana et al., 2011).

The number of secondary neurospheres increased in the presence of IL-10 (Fig. 3A). To evaluate whether IL-10 improved cell survival, secondary neurospheres were triturated, and cell viability was evaluated by the Trypan Blue exclusion assay. No differences were observed between treated and untreated neurospheres, thus showing that cell survival was not enhanced by this cytokine (Fig. 3B). Furthermore, analysis of neurosphere sizes manifested a normal distribution in treated and untreated cultures, indicating that IL-10 had no effect on proliferation rates (Fig. 3C) (Ferron et al., 2007). To decipher the impact of IL-10 on self-renewal, secondary neurospheres, either exposed or non-exposed to this cytokine, were passaged in the absence of IL-10. The number and size distribution of newly formed tertiary neurospheres were similar between previously IL-10-treated or untreated secondary neurospheres (Fig. 3D), indicating that this cytokine did not affect NSC self-renewal. Consequently, the higher number of secondary neurospheres observed in the presence of IL-10 could be explained by a direct growth effect induced by this molecule on NSCs or TAPs and suggests that IL-10 acts as a factor promoting maintenance of neural progenitors in an undifferentiated state. When tertiary neurospheres were disaggregated and cultured in differentiating conditions, the presence of IL-10 maintained a higher population of NSCs in culture (Nestin⁺ cells, undifferentiated) and neurogenesis was reduced (TUBB3⁺ cells), whereas glial differentiation was maintained (Fig. 3E,F).

The neurosphere assay was performed in both rat and mouse samples, and comparable outcomes were observed in both species. Together, these results suggest that IL-10 acts on neural SVZ progenitors and plays a role in the regulation of neuronal differentiation.

To evaluate the direct effect of IL-10 on SVZ progenitors, we used a more direct culture than the neurosphere assay to better represent the heterogeneous population present in the whole dorsal SVZ in situ. For this purpose, we performed monolayer primary cultures from dissociated cells from whole dorsal postnatal SVZ from rats; unfortunately, mouse cultures did not attach properly in these conditions. Cells were cultured in medium conditions similar to those used for the neurosphere differentiation assay with the absence of mitogenic factors. Microglial (IBA1⁺) cells were not detected in monolayer culture conditions. These cultures consisted of a heterogeneous cell population in which proliferation and differentiation occurred (Fig. 4A,D). As observed in the neurosphere culture, IL-10 did not affect cell viability or proliferation (PI and BrdU incorporation, respectively); however, the number of cells in active cell cycle (Ki67⁺) increased (Fig. 4A), thus pointing to an effect of this cytokine on cell cycle activity. Moreover, IL-10 upregulated neural markers such as Nestin, Sox1, Sox2, Musashi and Mash1, as well as activated Notch (NICD) (Fig. 4B), thereby suggesting that this cytokine causes a more pronounced undifferentiated progenitor state in SVZ cells. By contrast, the expression of Numb, a protein involved in neuronal differentiation (Imai et al., 2001; Kuo et al., 2006; Li et al., 2003; Nishimoto et al., 2010; Petersen et al., 2004) was reduced (Fig. 4B). Actin was unaltered and used as a loading control (Fig. 4B). Phenotypic analysis of monolayer cultures by immunofluorescence showed that, in the presence of IL-10, undifferentiated neural cell types accumulated, whereas the number of cells with a neuronal phenotype decreased. The number of Nestin⁺ cells markedly increased, and these cells underwent considerable morphological changes (Fig. 4C,D). They had more pronounced long processes that resembled the radial processes of radial glia, its developmentally originary cell (Fig. 4C, arrow; Merkle et al., 2004). Moreover, IL-10 increased the number of cells positive for Mash1, which is a marker for the TAP population (42.4±4.7%, Fig. 4D). By contrast, this cytokine decreased the number of neuroblasts (TUBB3⁺ cells; 53.03±3.8% in controls versus 33.7±1.7% in IL-10-treated animals) and induced a higher proportion of TUBB3⁺ cells that co-expressed Nestin (66.6±1.7% in controls versus 81.1±1.9% in treated cells) and were in the active cell cycle (11.3±1.5% in controls versus 18.4±0.9% in IL-10-treated animals). The numbers of glial progenitors (either Olig2⁺ or NG2⁺ cells) were unchanged after IL-10 treatment (Fig. 4D; and data not shown). Together, our results indicate that IL-10 does not interfere with the cell proliferation rate but does induce the accumulation of undifferentiated progenitors, which reduces neurogenesis. This finding suggests that IL-10 participates in the G2–M cell cycle transition.

To decipher alterations in this transition, we analyzed the presence of mitotic cells by staining for the phosphorylated form of histone 3 (PH3) as an M-phase-specific marker. The percentage of PH3⁺ cells was reduced by the presence of IL-10 at a range of post-treatment times (Fig. 4E, increase versus initial time). Furthermore, to determine whether IL-10 slowed down cell cycle progression through G2–M, a series of BrdU pulses (time 0), followed by various chase times, were conducted. Between 8 and 10 hours of chase, BrdU⁺ cells duplicated, indicating that this is the period required to transit to G2–M in order to leave the cell cycle (data not shown). To assess the acute effects of IL-10 on the S to M transition, BrdU labeling was combined with staining for PH3. At 3 and 6 hours, IL-10-treated cultures showed a lower percentage of double-labeled cells (BrdU⁺, Ph3⁺) (Fig. 4F). Moreover, at 6 hours, a higher number of cells left the cell cycle (BrdU⁺, Ki67⁻) in the control cultures (Fig. 4G,H). These findings support the notion that IL-10 delays the G2–M transition and subsequently the exit of the SVZ dorsal progenitors from the active cell cycle.

The continuation of mitosis universally involves early activation of M-phase-promoting factor by dephosphorylation of the inhibitory Thr14- and Thr15-phosphorylated sites of CDC2 protein kinase (Abrieu et al., 1997). We analyzed whether this key mechanism that controls G2–M cell cycle progression is regulated by IL-10. This cytokine induced a robust increase in CDC2 phosphorylation without affecting the total levels of expression of this cell cycle regulator in cultures (Fig. 4I). The increase in pCDC2 was due to an increase in the number of pCDC2_Tyr15+ cells, which were mainly Nestin⁺ (Fig. 4J). Moreover, IL-10 increased the percentage of TUBB3⁺ neuroblasts showing pCDC2_Tyr15 (13.7±0.2 in controls versus 21.2±0.3% in IL-10-treated animals, Fig. 4K). These findings suggest that IL-10 directly promotes a neural profile of SVZ cells while also impairing cell cycle exit and accumulating progenitors in G2–M transition, thus preventing these cells from dropping out of the cell cycle to become fully differentiated neurons.

The effects of IL-10 were also assayed in SVZ explants. Postnatal SVZ explants contained a population of Nestin⁺ cells restricted to zones near the core of the explants, whereas TUBB3 and DCX neuronal markers were present in migrating precursors, which also expressed IL-10R1 (Fig. 5A). When SVZ dorsal explants were exposed to IL-10, the area of migrating cells moving away from the core of the explants was severely reduced (Fig. 5B). Furthermore, analysis of explants of dorsal SVZ from IL-10-deficient mice (IL-10-KO) (Kühn et al., 1993) revealed that in the endogenous absence of this cytokine, the area of migrating cells was significantly increased when compared with WT explants (Fig. 5C). These observations could be explained by an impairment of neuronal differentiation; however, IL-10-mediated effects on cell migration cannot be excluded. To rule out this possibility, we performed a migration assay where SVZ explants were confronted by a source of IL-10 (IL-10 soaked beads). No chemorepulsive or chemoattractive effects were induced by IL-10 on SVZ progenitors (Fig. 5D).

All together, these results indicate that the action of IL-10 on SVZ Nestin⁺ progenitors regulates neurogenesis by causing a delay in the G2–M cell cycle progression and promoting an undifferentiated neural stage in dorsal SVZ progenitors.

We developed an in vivo gain-of-function model to decipher the physiological role of IL-10. The LV (right hemisphere) of adult mouse brains was infused with this cytokine at very low doses over 7 days (IL-10-gain). BrdU was administered to the animals once daily to monitor newly born cells. Histological studies were carried out in the left hemisphere, contralateral to that receiving the cannula implantation. The expression levels of astroglial (supplementary material Fig. S3A) and microglial (supplementary material Fig. S3B) markers such as GFAP and Iba-1 were similar between control and IL-10-gain animals, showing that IL-10 did not induce a glial reaction, as an indication that neuroinflammation was not altered.

Histological exploration of the ventral SVZ sub-region showed that both proliferating (PCNA⁺ cells, 339±40.2 versus 292±14.2) and cycling (Ki67⁺, 189±32.3 versus 217±50.6) cells were observed in the ventral neurogenic niche and indicated that IL-10 did not affect these specific progenitors, as expected because Nestin⁺ ventral progenitors are negative for IL-10-R1.

In-depth histological analyses of the dorsal SVZ indicated no changes in the total number of exiting cells (TO-PRO⁺) or cell death (see Materials and Methods) in the IL-10-gain in vivo model. Likewise, the numbers of newly generated (BrdU⁺) and PCNA⁺ cells (164±17.8 versus 160±5.4) were similar between control and IL-10-treated animals, thus supporting the notion that proliferation was not altered (Fig. 6A,B). However, the number of cells in the active cycle (Ki67⁺ cells) increased (Fig. 6A,B). Moreover, the expression of neural undifferentiating markers, such as Nestin, Musashi, NICD and Mash1, were increased in the IL-10 gain-of-function model (Fig. 6C). By contrast, NUMB and the neuronal marker DCX were severely reduced (Fig. 6C). The loading control, actin, was unaltered (Fig. 6C). Studies on sections at the cellular level revealed a sharp increase in the number of Nestin⁺ and Mash1⁺ cells in the dorsal SVZ after IL-10 treatment (Fig. 6D,E). By contrast, the number of committed neuroblasts (either DCX⁺ or PSA-NCAM⁺ cells) was significantly reduced (Fig. 6D,E). The percentage of newly generated progenitors (Nestin⁺, BrdU⁺) and neuroblasts (DCX⁺, BrdU⁺) decreased (Fig. 6F); however, a higher proportion of neuroblasts were actively cycling (DCX⁺, Ki67⁺, Fig. 6F). These observations showed that IL-10 does not affect the number of newly born cells. IL-10 intervenes in the existing pool of Nestin⁺ progenitors located dorsally, in which it promotes the prevalence of an undifferentiated transition state that leads to the impairment of neuronal differentiation. In this regard, the numbers of Nestin⁺ progenitors and DCX⁺ neuroblasts recovered in the dorsal SVZ of IL-10-gain animals two weeks after cytokine treatment was stopped (supplementary material Fig. S4A,B).

To further support the relevance of IL-10 as a regulator of the adult SVZ niche, we analyzed endogenous neurogenesis in IL-10-KO mice. Expression analysis of brain samples from these animals showed a reduction in the expression of Nestin and neural markers Musashi and NICD, in the dorsal SVZ (Fig. 6G,H). By contrast, DCX expression was strongly increased (Fig. 6G,H).

In addition, we evaluated NSC self-renewal in vivo in these models by assessing the presence of long cells retaining BrdU (LCR) (Ferron et al., 2007). Alteration of IL-10 expression either by gain-of-function or knockout did not induce differences in the number of LCR⁺ cells in the adult SVZ (supplementary material Fig. S5A,B). This finding indicates that IL-10 does not influence self-renewal in vivo and that its effects are related to its contribution to the maintenance of an undifferentiated pool of progenitors.

To precisely define whether IL-10 affects the phenotypic diversity of original progenitors in vivo, we administered a single injection of BrdU followed by 4 days of IL-10 treatment. Four days is the required time to detect newly formed neuroblasts in SVZ regeneration experiments (Doetsch et al., 1999). This approach allowed us to confirm that under physiological conditions the SVZ is mainly a neurogenic region because 64.1±2.3% of newly generated cells became committed neuroblasts (BrdU⁺, DCX⁺) whereas only 6±3.3% of BrdU⁺ cells expressed Olig2 as a marker of glial progenitors (Fig. 7A). The number of BrdU⁺ cells was similar in IL-10 gain animals and the control group, but IL-10 selectively reduced the proportion of committed neuronal cells to 50±2.6%, whereas it did not affect the number of BrdU and Olig2 double-labeled cells (4.5±1.3%) (Fig. 7A). In conclusion, IL-10 reduced neurogenesis by decreasing the pool of committed neuronal progenitors without inducing a switch to a glial phenotype. To further support the relevance IL-10 as a regulator of adult neurogenesis, we analyzed endogenous neurogenesis in IL-10-KO mice. In IL-10-deficient animals the number of new neurons (BrdU⁺, DCX⁺) was enhanced (84.9±1.4% versus 70.5±4.1% of WT) after four days of BrdU administration (Fig. 7B). No differences in the number of newly generated glial cells were observed between the two genotypes (BrdU⁺, Olig2⁺, 6.9±1.1% versus 6.1±0.6%) (Fig. 7B).

Finally, to study whether the effects of IL-10 on the SVZ influences final neurogenesis, we analyzed the OB (the final target area of newly generated neurons) of IL-10 gain- and loss-of-function animal models. The experimental design consisted of BrdU administration at days 0 and 1 followed by 19 monitoring days. This waiting time is required to allow newly formed cells to become fully differentiated neurons (NeuN⁺) and reach the OB, the target area (Doetsch and Alvarez-Buylla, 1996). Immunoanalysis of the OB of the IL-10 gain animals showed a reduced number of BrdU⁺ cells and also a decrease in the total number of newly generated neurons (Brdu and NeuN double-positive cells) (Fig. 7C,D). By contrast, IL-10 deficiency caused a significant increase in the number of newly formed cells (BrdU⁺) and neurons (BrdU⁺, NeuN⁺) that incorporated into the OB (Fig. 7E,F). Moreover, western blot analysis of OB samples revealed that the presence of GAD 65/67, as a marker of GABAergic cells, the extensively main neuronal phenotype of SVZ-derived cells, was increased in samples from IL-10-KO mice (Fig. 7G).

Together, these results indicate that IL-10 activity participates in the control of neurogenesis in the adult SVZ. IL-10 maintains SVZ neural progenitors in an undifferentiated condition, thereby confining neuronal differentiation; consequently, this cytokine regulates the final production of new neurons that will later regenerate the OB.

The SVZ niche is a complex system composed of progenitors, the lateral ventricle limited by the ciliated ependymal cells and the vascular system. The niche determines important properties of the stem cell biology including self-renewal, proliferation, differentiation and fate (Suh et al., 2009). Two types of progenitors are present in the SVZ niche: type B and type C cells. Type B cells are exposed to the ventricle and contact with blood vessels. Type C cells do not contact the ventricular space but contact blood vessels (Doetsch et al., 1996; Lois et al., 1994; Merkle et al., 2004). Here we propose a new biological role for IL-10, by which it modulates adult neurogenesis by targeting the Nestin⁺ progenitors located in the dorsal SVZ.

The adult SVZ is an extensive area where proliferation and neurogenesis occur along the entire rostrocaudal and dorsoventral axes (Mirzadeh et al., 2008). Several studies based on stem cell targeting (Merkle et al., 200743), cell grafting or transcription-factor-based fate mapping show that the NSCs present in various SVZ regions generate distinct types of neurons. In particular, deep granule interneurons and a subpopulation of periglomerular cells arise from the ventral SVZ, whereas superficial granule interneurons, distinct periglomerular cells and glutamatergic cells are derived from the dorsal SVZ (Kelsch et al., 2007; Kohwi et al., 2007; Merkle et al., 2007; Young et al., 2007; Brill et al., 2009). The molecular mechanisms controlling such positional specification are little known. The specific presence of Shh signaling in the ventral SVZ is instructive in cell fate specification and the generation of specific neuronal subtypes (Ihrie et al., 2011). Our data indicate the presence of molecular patterning in the adult SVZ that distinguishes, in a general manner, the dorsal versus ventral region. Progenitors generated along the entire ventral and dorsal regions of the SVZ move from the walls of the LV to reach more anterior dorsal SVZ regions (Kakita and Goldman, 1999; Geraerts et al., 2006; Sawamoto et al., 2006; Nam et al., 2007). IL-10R1 was not expressed in Nestin⁺ progenitors when these cells were in the ventral SVZ whereas it was expressed in nearly all Nestin⁺ progenitors located in the dorsal SVZ and ready to enter the RMS to reach their final target, the OB. This open dorsal IL-10R1 expression confers IL-10 the capacity to control final neurogenesis by acting on a wide population of SVZ Nestin⁺ progenitors, independently of their origin. Our data demonstrate a general action of IL-10 in the regulation of SVZ neuronal differentiation. This is one of the first molecular mechanisms described to be active in the dorsal region of the adult SVZ, which controls neurogenesis. In the future, it would be interesting to address the relevance of IL-10 on interneuron diversity specification and/or the determination of the glutamatergic neuronal phenotype.

The equilibrium between pro- and anti-inflammatory signals in the healthy body is required for the regular physiology of most tissues, including the brain. Here we show that an alteration of IL-10 expression (either gain- or loss-of-function) induces an abnormal ratio of neuronal differentiation in the adult neurogenic niche. As a final consequence, a significant impairment in the number of new neurons regenerating the OB is induced. The interaction of progenitors with the ventricular region and vessels makes these cells susceptible to regulation by local and systemic factors. A decline in neurogenesis and cognitive impairments observed during aging are associated with changes in blood factors, such as chemokines (Villeda et al., 2011). The CSF and blood of healthy animals contain low concentrations of IL-10, and after CNS inflammation or injury circulating levels of this cytokine increase. Furthermore, neurons and glial cells are also able to produce IL-10 (Lacki et al., 1995; Vila et al., 2003; Yasuda et al., 2008; Bromander et al., 2012). Our data indicate that IL-10 levels are determinants of the normal regulation of adult SVZ neurogenesis (summarized in Fig. 8). IL-10 levels regularly circulate in the body and are upregulated in response to disease, lesion or infection. Therefore, IL-10 could be a sensor of body physiology, continuously providing inputs to the SVZ niche in order to regulate final neurogenesis (summarized in Fig. 8).

Adult neurogenic niches increase proliferation and cell differentiation in response to cerebral aggression and neurodegeneration, but the maintenance of newly generated neurons is compromised (Arvidsson et al., 2002; Baron et al., 2008; Jin et al., 2004; Kokaia et al., 2006). The acute activation of local or systemic innate pro-inflammatory cascades is detrimental for progenitor proliferation. Much research effort has been dedicated to studying the cellular effects of these pro-inflammatory cytokines in an inflammatory context (Ekdahl et al., 2003; Monje et al., 2003). Following these first observations, other pro-inflammatory molecules and their receptors have been associated with the negative regulation of progenitor proliferation (Baron et al., 2008; Gómez-Nicola et al., 2011; Iosif et al., 2006; Koo et al., 2008; Li et al., 2010; Wachs et al., 2006). In addition, some pro-inflammatory signals have been linked to neuronal differentiation (Baron et al., 2008; Li et al., 2010). The molecular and cellular mechanisms that mediate the activity of these cytokines, as well as the relevance of these molecules in SVZ biology, are unclear. Furthermore, little is known about the effects of anti-inflammatory cytokines on SVZ biology in health or in diseased states. In animal models of neurodegeneration, the administration of high doses of IL-10 as an anti-inflammatory agent improves functional parameters and progenitor proliferation (Kiyota et al., 2012; Yang et al., 2009). On the basis of our findings, we propose that increases in systemic IL-10 levels cause a downregulation of SVZ neurogenesis by directly delaying the normal activity of the neurogenic niche (summarized in Fig. 8).

Here we report for the first time a novel endogenous action of IL-10 on adult neural progenitors in the healthy brain. IL-10R1 is present in the dorsal SVZ niche, and its expression is restricted to Nestin⁺ progenitors when these cells are located in that area before migrating to the OB. IL-10 directly acts on neural cells from the dorsal SVZ. It promotes a cellular non-differentiating stage and impedes neuronal differentiation among the entire pool progenitors. In vivo animal models of IL-10 gain- and loss-of-function showed that alterations in the levels of this cytokine lead to defects in neurogenesis in the healthy adult mouse brain.

Future studies will be crucial to identify the specific intracellular signaling pathways activated by IL-10 in adult brain progenitors and the effects of this cytokine on endogenous neurogenesis in a brain injury context involving the induction of a complex inflammatory process.

IL-10 was purchased from Prepotech (Rocky Hill, CT) and R&D (Minneapolis, MA), and U0126 was from Merck-Millipore (Darmstadt, Germany). Rodents were obtained from Charles River (Lille, France) and IL-10-KO mice (Kuhn et al., 1993) were originally purchased from the Jackson Laboratory. All animals were male and age-matched. Mice for knockout studies were obtained by crossing heterozygote animals. Cerebral volume was measured in vivo by MRI in WT and IL-10-KO mice. Mice were scanned in a 7.0 T BioSpec 70/30 horizontal animal scanner (Bruker BioSpin, Ettlingen, Germany). The analysis of T2 maps using AMIRA software or ImageJ showed no differences in forebrain volume between the two genotypes. Animals were kept and treated in accordance with the European Community Council Directive on animal welfare. Every effort was made to minimize animal suffering.

We studied IL-10 and its receptors in the postnatal and adult SVZ niche of both rat and mouse brains. Cell cultures were performed from postnatal brains of rats and mice (P7–P9) because cell viability, amount of material and adherence was better compared with adult cultures, and this reduced the number of animals needed per experiment. Fresh rat and mice coronal sections were obtained from the anterior forebrain using a McIlwain tissue chopper (300–500 µm). The SVZ was gently microdissected from each coronal section with the aid of sharpened metallic needles. Since dissociated mice brain cells did not attach properly in differentiating conditions, cultures were prepared from rat brain.

Dorsal SVZ was digested by trypsin and cultured in the presence of 10 ng/ml bFGF and 20 ng/ml of EGF in DMEM/F12 medium plus supplements (Life Technologies). Self-renewal studies were performed as described previously in different passages (Ferron et al., 2007). Differences between neurosphere cultures prepared from postnatal and adult brains have been described (Gritti et al., 2009). Differentiation studies were performed by mechanical dissociation of neurospheres from passage 2 in DMEM/F12 medium with B27 in the presence or absence of IL-10 (50 ng/ml) for 7 days.

Freshly isolated SVZ was mildly trypsinized and plated in poly-L-lysine-coated plates at medium density (40,000 cells/well on 48-well plates) and cultured in DMEM/F12 supplemented with B27 (Life Technologies, Paisley, UK) in the presence or absence of IL-10 (50 ng/ml) for several days. For viability studies, living cultures were stained with 10 µg propidium iodide (PI) to specifically label dead cells and all nuclei were stained with Hoechst 33342. A concentration-response curve was performed in order to determine the working concentration. Cell death was represented as percentage of the total number of cells. Proliferative studies were analyzed after a 4 hour pulse of bromo-deoxy-uridine (BrdU, 10 µM; Sigma-Aldrich, St Louis, MO) for each time studied. Pulse–chase experiments were performed 12 hours after platting dissociated primary cells; a pulse of 90 minutes of BrdU was applied to the cells and cells were chased at different time points (between 90 minutes and 12 hours) in order to study the S–G2–M transition. Dissociated cultures were fixed at each time point for 10 minutes in 4% paraformaldehyde (PFA). Cells were permeabilized with 0.1% Triton X-100 and stained as below. Nuclei were routinely counterstained with Hoechst 33342.

Isolated postnatal dorsal SVZ was cut in explants and embedded in Matrigel (BD-Bioscience), IL-10 was added at (50 ng/ml). Culture conditions and cell migration were described previously (Pozas and Ibáñez, 2005).

Western blotting was carried out after primary cultures and tissue samples were lysed and sonicated in RIPA buffer with proteases and phosphatase inhibitors. Protein was quantified by BCA (Pierce, Waltham, MA) and run in SDS-PAGE gels and transferred to PVDF membranes (Merck-Millipore). 5 µg total protein from culture lysates and 20–40 µg from tissue samples were loaded on polyacrylamide gels. Membranes were then incubated with pCDC2_Tyr15 (Cell Signaling Technology; 1∶1000), DCX (Cell Signaling Technology; 1∶1000; Santa Cruz Biotechnology, Santa Cruz, CA; 1∶500), Sox1 (Merck-Millipore; 1∶500), Sox2 (Merck-Millipore; 1∶500), Nestin (1∶500; either from Hybridoma Bank, Merck-Millipore or BD Bioscience San Jose, CA), Musashi (Millipore; 1∶500), Mash1 (BD Bioscience; 1∶500), total STAT-1 (BD Transduction; 1∶1000), total STAT-3 (BD Bioscience; 1∶1000), total CDC2 (Cell Signaling; 1∶1000), Notch-ICD (Abcam, Cambridge, UK; 1∶500), Numb (Abcam; 1∶500); GFAP (DAKO, Glostrup, Denmark; 1∶50,000), GAD 65/67 (Assay Designs, Lausen, Switzerland; 1∶1000), Spectrin (Millipore; 1∶1000) and Tubulin (Sigma; 1∶50,000) or Actin (Sigma; 1∶50,000) as loading controls. Blots were then processed by secondary antibodies conjugated to horseradish peroxidase and developed using luminol on X-ray films. Films were scanned and analyzed by Quantity One (Bio-Rad, Hercules, CA).

Total RNA from different cell culture samples and brain from postnatal ages of both mice and rats was isolated with SNAP columns (Life Technologies) or Trizol. Single-stranded cDNA was synthesized using AMV First-Strand cDNA Synthesis KIT (Life Technologies) and subjected to PCR. GenBank cDNA sequences were used to design gene specific primers in Primer3 Output and their specificity was determined by BLAST searching. Primer sequences were: rat IL-10 (ID, NM_12854) forward 5′-CCTGCTCTTACTGGCTGGAG-3′ and reverse 5′-TGTCCAGCTGGTCCTTCTTT-3′; mouse IL-10 (ID, NM_010548) forward 5′-AGTGAGAAGCTGAAGACCCTCAGG-3′ and reverse 5′-TTCATGGCCTTGTAGACACCTTGGT-3′, rat IL-10R1 (ID, NM_057193) forward 5′-TGAAACTCGCTTCACAGTGG-3′ and reverse 5′-TAAATACGGTGGTGCGTGAA-3′; mouse IL-10R1 (ID, L12120) forward 5′-AAAGGCTCAGGCATTGTCCT-3′ and reverse 5′-TCACGCTATCCACTGTCAGA-3′.

Primary antibodies (as described above) were used at the following dilutions: TUBB3 (1∶1000), DCX (1∶1000), Nestin (1∶200), Ki67 (Leica Microsystems, Wetzlar, Germany; 1∶1000), pCDC2 (1∶300); PH3 (1∶500; Cell Signaling Technology), IL-10R1 (1∶150; from Millipore or Santa Cruz Biotechnology), BrdU (AbCam; 1∶400), NG2 (Millipore; 1∶400), GFAP (DAKO; 1∶2000), NeuN (Millipore; 1∶1000), Gad67 (Millipore; 1∶300) and Spectrin (Millipore; 1∶1000). Following the first immunoreaction, samples were incubated with secondary antibodies conjugated with Alexa Fluor dyes (Molecular Probes; 1∶1000). In some cases, staining was amplified by secondary biotinylated antibodies (Vector; 1∶200) followed by Alexa-Streptavidin (Molecular Probes; 1∶500). Staining on cell cultures was visualized with a Leica CTR400-DMI400B inverted microscope. Photographs were taken with a DFC300FX camera from Leica. Quantification was carried out using Leica Application Suite (LAS-Leica) or ImageJ (NIH) software. The number of positive cells for each experimental assay was expressed as percentage of the total number, and/or referred as percentage of the control.

The administration of IL-10 (50 ng/ml) and/or U0126 (25 µM) in the lateral ventricle was carried out by continuous infusion with Alzet osmotic minipumps (model 1007D) and Alzet Brain Infusion Kit 3 (DURECT Corporation, Cupertino, CA). The animal was placed in a stereotaxic frame using a mice adaptor (Stoelting, Wood Dale, IL) and the delivery of saline or IL-10 to the lateral ventricle of the right hemisphere was achieved by inserting the cannula 1.7 mm deep from the brain surface at −0.1 mm posterior, and 0.6 lateral coordinates from Bregma. The cannula was connected to the osmotic minipump placed in a subcutaneous pocket between the scapulae. Infusion of vehicle, IL-10 (50 ng/ml) was carried out at a low flow rate of 0.5 µl/hour and it was prolonged during 4 (4-day-treated animals) and/or 7 days (7-day-treated animals) depending on the experimental group. Contralateral hemisphere (left) was always considered for histological analysis and some ipsilateral hemispheres were used for biochemical studies.

Intraperitoneal (IP) BrdU injections were administered in three independent experimental groups. In the first group, the animals received BrdU (50 mg/kg) daily for 7 days until the day of sacrifice. In the second group, animals received a single dose of BrdU at day 0 and were killed at day 4. The last group of animals received BrdU on the two first days (0 and 1) and were perfused at day 19. These BrdU injections were performed in the IL-10 gain-of-function animal model and in IL-10-KO mice.

Postnatal and adult animals (8 weeks) were always perfused with saline followed by 4% paraformaldehyde and cryoprotected in 30% sucrose. The forebrain was snap-frozen in dry-ice and coronal sections (16 µm) were obtained and were collected in eight consecutive slice series. Slices were frozen and were air-dried at room temperature before use for histological analyses. In all animals, the left (contralateral) hemisphere was analyzed on histological sections to avoid any mechanical damage that might be caused by inserting the delivery cannula in the right ventricle.

Routinely, the evaluation of total number of cells present in the dorsal SVZ was carried out after TO-PRO staining (Invitrogen). Cell death was evaluated by morphological observation and cleaved Caspase 3 immunostaining (Cell Signaling; 1∶150). Cell death was also evaluated by the pattern of cleaved Spectrin by elecrophoresis and western blotting (see above).

The primary antibodies used for immunofluorescence were as described above. Brain sections were scanned and evaluated under a Leica DM5500Q confocal microscope. At least four consecutive sections from the same slice were evaluated for each staining and the number of positive cells for each section was counted after a Z projection by ImageJ or LAS (Leica application suite) software. For the SVZ analysis, images were taken between +1.10 mm and +0.38 mm from Bregma. For OB studies, we analyzed a number of consecutives series starting around +3.56 from Bregma and going rostrally. The total number of cells per section or the percentages of a particular cell population were represented on the histograms.

Analyses of significant differences between group means were performed using the two-tailed Student's t-tests. In each case, n indicates the number of independent cultures or mice used. Statistical significance was considered where P<0.05.

We are grateful to Dr Fernando Aguado for helpful discussions, and Toffa Evans and Amanda Billington for editorial assistance. We are indebted to the Cytomics and Image Units of the Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS) for technical help. Some Nestin antibodies were obtained from the Developmental Studies Hybridoma Bank at the University of Iowa.