The special AT-rich sequence binding protein 2 (Satb2) was first identified as a cleft palate gene (FitzPatrick et al., 2003). As a transcriptional regulator, Satb2 is shown to be essential for the osteoblast differentiation and craniofacial skeletal development in Satb2-null mice (Britanova et al., 2006; Dobreva et al., 2006). Meanwhile, Satb2 expression in the central nervous system (CNS) caught more attention than that to its role in other systems (Britanova et al., 2005; Szemes et al., 2006). By analyzing the Satb2 knockout mice, two groups have found that Satb2 determines the identity of the neurons in the upper layers of the cerebral cortex (Alcamo et al., 2008; Britanova et al., 2008; Leone et al., 2008; Kwan et al., 2012). In addition, Satb2 is also shown to control the dendritic arborization of cortical upper-layer neurons in early postnatal development (Zhang et al., 2012).
Accumulating evidence focused on the expression and function of Satb2 in the cerebral cortex shows that Satb2 can be regarded as a marker for superficial neurons in the murine or human cerebral cortex (Chen et al., 2008; Hisaoka et al., 2010; Ip et al., 2011; Saito et al., 2011; Baranek et al., 2012; Lickiss et al., 2012; Zgraggen et al., 2012; Zhang et al., 2012). However, the distribution and role of Satb2 beyond the cerebral cortex in other brain regions are largely unknown, in spite of several reports showing Satb2 expression in the hypothalamus, pontine region and retina (Kurrasch et al., 2007; Maeda et al., 2009; Kay et al., 2011).
To gain insights to the function of Satb2 in the adult mouse brain, we performed a comprehensive immunohistochemical study on the expression of Satb2 in the CNS of adult mice as well as the characterization of the neurochemical features of Satb2-expressing neurons in some specific brain regions. We found that Satb2 is abundantly distributed in the neocortex, while relatively a small number of Satb2-expressing neurons are found in the subcortical region and brainstem. Double immunostaining in the neocortex showed that Satb2 is exclusively expressed in excitatory neurons, other than in inhibitory neurons or astrocytes. Most of Satb2-expressing neurons in the hypothalamic periventricular region are dopaminergic, and those in the dorsal part of dorsal raphe nucleus are serotonergic. These results suggest that Satb2 may be involved in the physiological function of certain neuronal subtypes.
MATERIALS AND METHODS
A total number of 14 adult C57BL/6J mice (2–4-months old) were used in this experiment. Mice were deeply anesthetized with sodium pentobarbital (100 mg/kg body weight) and perfused with phosphate-buffered saline (PBS; pH 7.4), followed by 4% paraformaldehyde (PFA; Sigma-Aldrich) in 0.1 M phosphate buffer (PB; pH 7.4). The mouse brain and spinal cord were post-fixed in 4% PFA for 24 hr at 4°C and cryoprotected with 30% sucrose in PBS for another 24 hr at room temperature. Coronal sections of 40-μm thickness were prepared using a cryostat (CM1950; Leica) and then subjected to immunohistochemistry or immunofluorescent staining immediately. Glutamic acid decarboxylase 67 (GAD67)-green fluorescent protein (GFP) knock-in mice (Tamamaki et al., 2003) were used to study Satb2 expression in γ-aminobutyric acid (GABA)-ergic neurons. Animal care procedures and experimental protocols were approved by the Animal Studies Committee at the Tongji University School of Medicine (Shanghai, China). All efforts were made to minimize animal suffering and the number of animals used.
After being pretreated with 3% H2O2 in methanol at room temperature for 15 min and then with citrate buffer (pH 6.0) for 8–9 min at 94°C, brain sections were incubated overnight at 4°C with mouse anti-Satb2 antibody (1:300; Santa Cruz Biotechnology) diluted in PBS containing 0.3% Triton X-100 and 1% bovine serum albumin (BSA). After several washes, the sections were incubated with biotinylated donkey anti-mouse antibody (1:500; Vector Laboratories) for 3 hr at room temperature, followed by incubation with the Vectastain Elite ABC reagent (1:200; Vector Laboratories) at room temperature for 1 hr. After washed in PBS, the sections were placed in a diaminobenzidine tetrahydrochloride (0.5 mg/mL in PBS containing 0.003% hydrogen peroxide and 0.3% nickel ammonium sulfate) reaction solution for visualization. Negative control was performed by incubation of sections in PBS containing 0.3% Triton X-100 and 1% BSA without anti-Satb2 antibody and no immunoreactivity was observed.
Double Immunofluorescent Staining
For double immunofluorescent staining with antibodies of Satb2 and other specific cortical markers, brain sections pretreated with citrate buffer (pH 6.0) were incubated at 4°C overnight with mouse anti-Satb2 antibody (1:300, Santa Cruz) and one of the following antibodies: goat anti-Brn2 (1:300; Santa Cruz), rabbit anti-Cux1 (1:300; Santa Cruz), goat anti-Sox5 (1:300; Santa Cruz). After several washes with PBS, sections were incubated with Alexa Fluor 488-conjugated donkey anti-mouse IgG (1:500; Invitrogen) and biotinylated donkey anti-goat or rabbit IgG (1:500; Vector Laboratories) antibodies for 3 hr and then Cy3-conjugated streptavidin (1:1,000; Jackson ImmunoResearch Laboratories) for 1 hr at room temperature.
For double immunofluorescent labeling outside of the cerebral cortex with two primary antibodies originating from different species, Satb2 was detected by tyramide signal amplification system (TSA Plus Biotin Kit, PerkinElmer). In brief, brain sections pretreated with 3% H2O2 and citrate buffer (pH 6.0) were then incubated with both mouse anti-Satb2 antibody (1:5,000; Santa Cruz Biotechnology) and one of the following antibodies: rabbit anti-glial fibrillary acidic protein (GFAP; 1:500; Dako), rabbit anti-GFP (1:2,000; Invitrogen), rabbit anti-S100calcium binding protein β (S100β; 1:600; Sigma-Aldrich), rabbit anti-Orexin-A (1:300; Millipore), rabbit anti-Oxytocin (1:200; Millipore) and rabbit anti-tryptophan hydroxylase 2 (Tph2; 1:5,000; a gift from Dr.Klaus-Peter Lesch) (Gutknecht et al., 2009, 2008) antibodies. Signals were developed by incubating sections with a mix of Alexa Fluor 488-conjugated donkey anti-rabbitIgG (1:500; Invitrogen) and horseradish peroxidase (HRP)-labeled goat anti-mouse IgG (1:100; KangChen) antibodies for 3 hr at room temperature, followed by TSA Biotin Amplification Reagent (1:50; PerkinElmer) for 10 min at room temperature, and then 1 hr incubation with Cy3-conjugated streptavidin (1:1,000; Jackson ImmunoResearch Laboratories).
For double immunofluorescent labeling outside of the cerebral cortex with two primary antibodies originating from the same species, the above procedure is slightly modified to avoid cross-reaction (Shindler and Roth, 1996; Brouns et al., 2002). Briefly, brain sections pretreated with 3% H2O2 and citrate buffer (pH 6.0) were first incubated overnight at 4°C with mouse anti-Satb2 antibody (1:5,000; Santa Cruz), a dilution at which no immunoreactivity for Satb2 could be detected by the conventional fluorescent immunostaining procedure. Sections were then incubated with HRP-labeledgoat anti-mouse IgG (1:100; KangChen) for 3 hr and TSA Biotin Amplification Reagent (1:50; PerkinElmer) for 10 min at room temperature. Signals were visualized with Cy3-conjugated streptavidin (1:1,000; Jackson ImmunoResearch Laboratories). After being washed with PBS, sections were incubated with the second primary antibody, mouse anti-NeuN (1:1,000; Millipore) or mouse anti-tyrosine hydroxylase (TH) (1:20,000; Sigma-Aldrich) antibody overnight at 4°C and Alexa Fluor 488-conjugated donkey anti-mouse IgG (1:500; Invitrogen) for 3 hr at room temperature.
Imaging and Quantitative Analysis
Images were obtained by an epifluorescent microscope (Eclipse 80i; Nikon) equipped with a Coolpix digital camera (DS-Ri1; Nikon) or a confocal microscope (TCS SP5; Leica). Four mice were used for the quantitative analysis of Satb2/Brn2, Satb2/Cux1, Satb2/Sox5-double labeled cells in the neocortex, Satb2/TH-double labeled neurons in the hypothalamus and Satb2/Tph2 in the rostral and dorsal part of the dorsal raphe nucleus (Bregma −4.72 mm). Six sets of consecutive 40-μm thick coronal sections were collected from each brain. One set of sections was processed for the double immunostaining and then used for cell counts. All statistical data are presented as mean±standard errors of the mean (s.e.m.).
Using immunohistochemistry, we found that Satb2 was abundantly expressed in the all six layers of the neocortex (Fig. 1), which is consistent with the previous reports (Britanova et al., 2005; Alcamo et al., 2008; Britanova et al., 2008; Nielsen et al., 2010; Balamotis et al., 2012; Zhang et al., 2012). To investigate the colocalization of Satb2 with layer-specific genes, we performed double immunostaining for Satb2 with Brn2 (Sugitani et al., 2002), Cux1 (Nieto et al., 2004) and Sox5 (Lai et al., 2008). We found that Brn2 and Cux1 were highly expressed in the superficial layers (Fig. 1A,B), so the colocalizations of these two proteins with Satb2 were evaluated in the layers II–III separately (Fig. 1A′-A″′,B′-B′″). Cell counting data showed that 90.1±1.1% of Brn2-expressing cells was Satb2-positive; 98.0±0.9% of Satb2-immunoreactive cells was expressing Brn2 (322.5±17.5 cells were counted from each brain; arrows in Fig. 1A′-A″′). Similar results were obtained from Satb2/Cux1 double immunostaining: 99.0±0.4% of Cux1-expressing cells was Satb2-positive and 97.2±0.3% of Satb2-immunoreactive cells was expressing Cux1 (321.5±23.6 cells were counted from each brain; arrows in Fig. 1B′-B″′). Strong expression of Sox5 was observed in the deep layers of neocortex (Fig. 1C), and the colocalization of Sox5 with Satb2 was examined in the layer V (Fig. 1C′-C″′). Cell counting data showed that about 79.3±2.0% of Sox5-expressing cells was positive for Satb2, whereas approximately 68.7±1.3% of Satb2-positive cells expressed Sox5 (299±19.9 cells were counted from each brain; arrows in Fig. 1C′-C″′).
To further examine whether Satb2 is expressed in neuronal and/or glial cells in the neocortex, we performed double immunostaining for Satb2 with NeuN (a pan-neuronal marker), S100β or GFAP (both are astrocytic markers). We found that all Satb2-expressing cells were NeuN immunoreactive (Fig. 2A-A″, arrows), and no Satb2-positive cells were expressing S100β or GFAP (Fig. 2B-B″,C-C″). Therefore, Satb2 is expressed exclusively in neurons, similar to the distribution of Satb1 in a previous report (Huang et al., 2011). To explore whether Satb2 is expressed in excitatory or inhibitory neurons, double immunostaining for Satb2 and GFP was performed in brain sections from GAD67-GFP knock-in mice (Tamamaki et al., 2003), in which GFP expression is driven by the promoter of GABA-synthesizing enzyme GAD67. No colocalization of Satb2 and GFP was observed (Fig. 2D-D″), indicating that Satb2-expressing cells are not GABAergic in adult mouse neocortex.
In the basal forebrain, a large number of Satb2-positive neurons were distributed in the nucleus of the horizontal limb of the diagonal band (HDB; Fig. 3A,C), while a small number of them were observed in the bed nucleus of the stria terminalis (BST; Fig. 3A,B). In the diencephalon, Satb2 expression was found in the hypothalamus but not in the dorsal thalamus. Satb2-expressing neurons were present in the lateral hypothalamus (LH; Fig. 3D–F), arcuate nucleus (ArC; Figs. 3G and 5A), and the paraventricular nucleus (PVN; Figs. 3D,E and 5C,D). In the lateral hypothalamic area, it is well known that a few neurons secret orexin, which is involved in feeding behavior (Sakurai et al., 1998). However, we only found Orexin-positive fibers in Satb2-expressing LH areas (Fig. 4A-A″′). Distribution of Satb2 in the PVN promoted us to examine if these neurons expressed oxytocin, but no colocalization of Satb2 and oxytocin was observed (Fig. 4B-B″′).
In addition to the hypothalamic regions mentioned above, Satb2 was also expressed in the periventricular regions (Figs. 3G and 5A). Dopaminergic neurons are also distributed in the periventricular region of hypothalamus along the anterior–posterior axis (Dahlstroem and Fuxe, 1964; Paxinos and Franklin, 2001). Overlapping expression of both Satb2 and TH was observed, butcoexpression of Satb2 and TH was only found in the A12 region, which is located dorsal to the ArC (Fig. 5A–B″). Cell counting data showed that approximately 60.5±6.2% of TH-positive neurons was immunostained with Satb2 antibody (157.8±6.9 cells were counted from each brain). Meanwhile, about 85.1±6.8% of Satb2-positive neurons expressed TH. However, in the A14 which is located in the rostral part of PVN, no neurons expressing both Satb2 and TH was observed (Fig. 5C–D″). The largest population of dopaminergic neurons is located in the ventral midbrain, in which no detectable Satb2-immunosignal was found in the ventral tegmental area (VTA) or compact part of substantia nigra (SNc).
In the midbrain, intense Satb2 expression was found in the ventral tegmental nucleus (VTg) and the laterodorsal tegmental nucleus (LDTg) (Fig. 6A,E), while weak expression of Satb2 was found in the rostral part of dorsal raphe nucleus (Figs. 6A,B and 7A,B). Satb2 expression was mainly localized in the dorsal and rostral parts of the dorsal raphe nucleus (DRD), which is centered around the level of Bregma −4.72 mm (Fig. 6A,B). A large population of serotonergic neurons are located in the dorsal raphe nucleus, which are termed as B6 and B7 cell groups (Dahlstroem and Fuxe, 1964). To examine whether Satb2 is expressed in these serotonergic neurons or not, we performed double immunostaining for Satb2 and tryptophan hydroxylase 2 (Tph2), which is the limited enzyme controlling the serotonin biosynthesis in the brain (Zhang et al., 2004). We found that most of Satb2-positive neurons also showed Tph2 immunoreactivity (Fig. 7A-A″,B-B″, arrows). Cell counting data showed that about 60.3±1.1% of Tph2-expressing neurons was Satb2-positive in the DRD (152.8±26.4 cells were counted from each brain). Meanwhile, there was approximately 88.5±4.0% of Satb2-positive neurons colabeled with Tph2. On the contrary, no Satb2-immunoreactive signals were found in other raphe nuclei like the ventral part of dorsal raphe nucleus (DRV) (Fig. 6A-A″,C-C″).
In the hindbrain, Satb2 expression was restricted in a few cells of some nuclei. Ventrally to the dorsal raphe nucleus, Satb2 was expressed in the rostral periolivary region (RPO; Fig. 6C,D). Besides, Satb2 was expressed strongly in the parabrachial region including both medial and lateral nuclei (MPB and LPB; Fig. 6E). The immunoreactive signal for Satb2 in this region was as strong as in the neocortex, consistent with a previous report (Maeda et al., 2009). There were no signals of Satb2-immunoreactivity in the cerebellum or other regions in the medulla oblongata or spinal cord (Fig. 6E and data not shown).
In this study, we investigated the expression pattern of Satb2 in the adult mouse brain using immunohistochemistry. Given several reports exploring Satb2 expression in the cerebral cortex (Britanova et al., 2005; Alcamo et al., 2008; Britanova et al., 2008; Gyorgy et al., 2008; Balamotis et al., 2012; Baranek et al., 2012; Zhang et al., 2012), we focused our investigation mainly on subcortical regions and the brainstem, in which the expression of Satb2 is largely unclear. With the help of double immunostaining, we found in the cerebral cortex that Satb2 is exclusively expressed in the excitatory neurons, other than inhibitory neurons or astrocytes (Fig. 2), which is distinct from the expression of Satb1, the closest homolog of Satb2, in both excitatory and inhibitory neurons of the cerebral cortex (Balamotis et al., 2012). This discrepancy may also reflect the different roles of these two related genes in the development of cerebral cortex: Satb2 is essential for the identity determination and dendritic arborization of cortical neurons (Alcamo et al., 2008; Britanova et al., 2008; Zhang et al., 2012), while Satb1 is important for the synapse formation (spine density) of these neurons (Balamotis et al., 2012).
In the hypothalamus, the dopaminergic neurons are identified as A11, A12, A13, and A14 cell groups along the posterio-anterior axis (Björklund and Dunnett, 2007). These diencephalon-originated dopaminergic neurons often project to adjacent regions, which is quite different from the major ascending mesencephalic dopaminergic neurons (Moore and Lookingland, 2000).We found that the majority of Satb2-positive cells are co-immunostained with TH in the A12 group (Fig. 5), while no colocalization was found in other groups, which indicates that only Satb2-expressing cells in the A12 group are dopaminergic. Although the development and function of the hypothalamic dopaminergic neurons are poorly understood so far, it is known that A12 group provides the tuberoinfundibular and the tuberohypophysial projections involved in neuroendocrine regulation (Smidt and Burbach, 2007). Given the critical role of Satb2 in cortical development, it is possible that Satb2 may be involved in the specification of cell fate and maintainance of normal hormone-releasing regulation in the A12 cell group. Besides, it is well known that leptin signaling pathway regulates the mammalian food intake and body weight through the ArC as well as other nuclei of the hypothalamus (Jovanovic and Yeo, 2010), and it has been reported that the expression of Satb2 mRNA is dramatically reduced in response to leptin defficiency in neonatal other than adult mice (Kurrasch et al., 2007). It is also interesting to explore the role of Satb2 expression in the hypothalamus in autonomic functions, such as food intake.
In addition, Satb2 is specifically expressed in the DRD (Fig. 6A,B), which contains the majority of serotonergic neurons in the brainstem. We found that most of Satb2-labeled cells in the DRD also expressed Tph2 (Fig. 7), indicating that these Satb2-positive cells are serotonergic. The dorsal raphe nucleus has unique anatomical and neurochemical properties consistent with a role in the etiology and pathophysiology of anxiety and affective disorders (Lowry et al., 2008). It is very worth examing the role of Satb2 in the DRD-related psychiatric disorders, especially anxiety. Although Satb2 expression was not observed in the VTA and SNc, which contains the midbrain dopaminergic neurons projecting widely to the brain, the majority of TH-positive neurons are expressing Satb1 in these two brain regions (Huang et al., 2011). Conversely, Satb1-positive neurons in the dorsal raphe nucleus are not serotonergic (Huang et al., 2011), wheareas the majority of Satb2-positive neurons was found to be colocalized with Tph2 in the dorsal part of this nucleus. The differential distribution of these two family members suggests their separate, not redundant, roles in the brain.
In summary, Satb2 is expressed widely in the adult mouse brain. Specifically, Satb2-positive cells are dopaminergic in the A12 cell group of hypothalamus and are serotonergic in the dorsal part of the dorsal raphe nucleus. Our findings not only implicate the functions of Satb2 in the development of certain subpopulations of neurons but also provide a morphological basis for studying the role of Satb2 in the adult brain and in mental diseases.
The authors thank Ms. Jia-Yin Chen and Qiong Zhang for technical support and Dr. K.P. Lesch (University of Würzburg) for providing the anti-Tph2 antibody.