ABSTRACT
The serotonergic system plays important roles in multiple functions of the nervous system and its malfunctioning leads to neurological and psychiatric disorders. Here, we show that the cell adhesion molecule close homolog of L1 (CHL1), which has been linked to mental disorders, binds to a peptide stretch in the third intracellular loop of the serotonin 2c (5-HT2c) receptor through its intracellular domain. Moreover, we provide evidence that CHL1 deficiency in mice leads to 5-HT2c-receptor-related reduction in locomotor activity and reactivity to novelty, and that CHL1 regulates signaling pathways triggered by constitutively active isoforms of the 5-HT2c receptor. Furthermore, we found that the 5-HT2c receptor and CHL1 colocalize in striatal and hippocampal GABAergic neurons, and that 5-HT2c receptor phosphorylation and its association with phosphatase and tensin homolog (PTEN) and β-arrestin 2 is regulated by CHL1. Our results demonstrate that CHL1 regulates signal transduction pathways through constitutively active 5-HT2c receptor isoforms, thereby altering 5-HT2c receptor functions and implicating CHL1 as a new modulator of the serotonergic system.
INTRODUCTION
Cell adhesion molecules, such as close homolog of L1 (CHL1) (Holm et al., 1996; Hillenbrand et al., 1999), have been implicated in numerous functions in the developing and adult nervous system, including neural cell migration and survival, neuritogenesis, axon guidance and synaptogenesis, as well as synaptic plasticity (for a review, see Maness and Schachner, 2007). In humans, mutations of CHL1 are associated with mental retardation, epilepsy, schizophrenia and autism spectrum disorders characterized by deficits in behavior, including cognition and social interactions (Angeloni et al., 1999a,b; Sakurai et al., 2002; Frints et al., 2003; Chen et al., 2005; Chu and Liu, 2010; Tam et al., 2010; Cuoco et al., 2011; Salyakina et al., 2011; Shoukier et al., 2013). Ablation of CHL1 in mice leads to impairments in synaptic transmission, long-term potentiation, working memory, gating of sensorimotor information and the stress response, as well as alterations in social and exploratory behavior (Montag-Sallaz et al., 2002; Frints et al., 2003; Pratte et al., 2003; Demyanenko et al., 2004, 2011; Irintchev et al., 2004; Leshchyns'ka et al., 2006; Nikonenko et al., 2006; Morellini et al., 2007; Wright et al., 2007; Kolata et al., 2008; Pratte and Jamon, 2009).
Being interested in molecules interacting with the intracellular domain of CHL1, with the expectation to find bidirectionally acting modifiers of individual functions, we identified the serotonin 2c (5-HT2c) receptor as a CHL1-binding partner. The seven-transmembrane-spanning G-protein-coupled 5-HT2c receptor triggers different signal transduction pathways and is involved in regulating different types of behavior (Werry et al., 2005, 2008; Berg et al., 2008). Adenosine-to-inosine RNA editing of this receptor alters its signaling characteristics, leading to different degrees of constitutive activity of the differentially edited receptors (Iwamoto et al., 2009; O'Neil and Emeson, 2012). The non-edited isoform and the partially edited isoforms exhibit pronounced or intermediate constitutive activity, whereas the fully edited isoform is not constitutively active and requires binding of serotonin for activation (Burns et al., 1997; Herrick-Davis et al., 1999; Niswender et al., 1999). Reduced reactivity to a novel environment and sensitivity to locomotor stimulants, as well as alterations in sleep, feeding and reward-associated behaviors are observed in 5-HT2c-receptor-deficient mice, whereas mice expressing the non-edited or fully-edited isoforms show anxiety and depression-like symptoms, respectively (for a review and references, see Heisler et al., 2007; Iwamoto et al., 2009; Fletcher et al., 2009; Mombereau et al., 2010; Pennanen et al., 2013). In humans, the 5-HT2c receptor has been associated with obesity as well as neurological and psychiatric disorders, such as schizophrenia, major depression, addiction, epilepsy, anxiety, sleep disorders, motor dysfunction and bipolar disorder (Berg et al., 2008; Lee et al., 2010 ).
In the present study, we show that the interaction of CHL1 with the 5-HT2c receptor regulates the functions of the receptor, 5-HT2c-receptor-mediated signaling and behavior in mice, suggesting that some interdependent functions of the two molecules are linked through their molecular associations.
RESULTS
CHL1 interacts with the 5-HT2c receptor
To find new intracellular interaction partners for CHL1, we performed a phage display screening of a random 12mer peptide library using the intracellular domain of CHL1 (CHL1-ICD) (Fig. 1A), which belongs to the L1 (L1 is also known as L1CAM) family of structurally related proteins. A phage expressing a CHL1-ICD-binding peptide was obtained that showed similarity to a short sequence stretch within the third intracellular loop of the 5-HT2c receptor, comprising the amino acids 292–304 (Fig. 1B). To ascertain that the peptide sequence expressed by the phage binds to CHL1-ICD, a synthetic peptide with the phage sequence was probed by ELISA using increasing concentrations of CHL1-ICD. As a control, we used increasing concentrations of L1-ICD, which shows homology to CHL1-ICD like the other members of the L1 family, namely neurofascin and NrCAM (Fig. 1A). A concentration-dependent and saturable binding of CHL1-ICD to the peptide was observed, whereas L1-ICD did not bind to the peptide (Fig. 1C). To substantiate the interaction between CHL1 and the 5-HT2c receptor, a synthetic peptide comprising amino acids 281–309 of the 5-HT2c receptor and carrying a biotin moiety at its N-terminus was applied in increasing concentrations to substrate-coated CHL1-ICD or to the control substrate L1-ICD. Using horseradish peroxidase (HRP)-conjugated streptavidin, a concentration-dependent and saturable binding of the 5-HT2c-receptor-derived peptide to CHL1-ICD, but not L1-ICD was observed (Fig. 1D).
As a further verification of the interaction between the two molecules, pulldown experiments were performed using mouse brain homogenate and CHL1-ICD. As controls, we used L1-ICD and the intracellular domain of NCAM140 which belongs to the immunoglobulin superfamily like CHL1. Western blot analysis of the precipitates revealed that the mouse anti-5-HT2c-receptor antibody recognized a double band of ∼50 and 55 kDa in the CHL1-ICD precipitate, whereas no immunopositive bands for the 5-HT2c receptor were detectable in the L1-ICD and NCAM-ICD precipitates (Fig. 1E). These experiments show that the 5-HT2c receptor specifically interacts with CHL1-ICD.
To further substantiate the interaction of CHL1 with the 5-HT2c receptor, we performed immunoprecipitation experiments using detergent extracts of brain homogenate from wild-type and CHL1-deficient mice using polyclonal rabbit CHL1 antibody. In a western blot analysis, the mouse anti-5-HT2c-receptor antibody recognized predominantly the 50-kDa band in the CHL1 immunoprecipitates from wild-type brain, but not from CHL1-deficient brain (Fig. 1F), indicating that the 5-HT2c receptor and CHL1 are associated with each other in the intact tissue.
To narrow down the 5-HT2c-receptor-binding site in CHL1-ICD, we performed ELISA with substrate-coated 5-HT2c-receptor-derived peptide and recombinant N- and C-terminal truncations of CHL1-ICD fused to GST (Fig. 2A). CHL1-ICD with C-terminal truncations bound to the 5-HT2c-receptor-derived peptide with the shortest CHL1-ICD construct showing the highest concentration-dependent binding to the peptide (Fig. 2B,C). In contrast, CHL1-ICD with the N-terminal truncation and GST did not bind to the peptide (Fig. 2B,C).
The combined results indicate that CHL1 binds to a sequence stretch in the third intracellular loop of the 5-HT2c receptor through the 29 N-terminal amino acids in its intracellular domain and that the 5-HT2c receptor does not interact with L1 and NCAM, which are structurally related to CHL1.
CHL1 colocalizes with the 5-HT2c receptor in GABAergic neurons in the striatum and hippocampus
The interaction between CHL1 and 5-HT2c receptor was further assayed by immunohistology in the adult mouse brain. Double-labeling with goat antibody against CHL1 and mouse antibody against the 5-HT2c receptor showed a considerable overlap of immunostaining in the striatum (Fig. 3A), which contains a large number of inhibitory GABAergic medium spiny neurons (90–95% of the total neuronal population) and GABAergic interneurons. In the CA3 region of the hippocampus, co-immunostaining of CHL1 and the 5-HT2c receptor was observed in distinct cells (Fig. 4A), which likely represent GABAergic interneurons, as judged by their scarcity, position and cell body size. Co-immunostaining of CHL1 and the 5-HT2c receptor in some cells was also observed in the CA1 region of the hippocampus (data not shown). To test whether CHL1 and the 5-HT2c receptor are expressed in GABAergic interneurons, we performed double immunostaining with antibodies against CHL1 or 5-HT2c receptor and parvalbumin, a marker protein for GABAergic interneurons (Houser, 2007; Tepper et al., 2010). Co-stainings of 5-HT2c receptor or CHL1 with parvalbumin were observed in cells in the putamen (Fig. 3B,C) and stratum pyramidale in the CA3 region of the hippocampus (Fig. 4B,C), indicating that the 5-HT2c receptor and CHL1 are associated with each other in GABAergic interneurons in the hippocampus and striatum.
To investigate whether CHL1 and 5-HT2c receptor directly interact with each other in hippocampal and striatal neurons of intact adult mouse brain, we used the antibodies against CHL1 and 5-HT2c receptor in a proximity ligation assay, which allows the detection of close protein interactions with high sensitivity and specificity by generation and amplification of fluorescent signals from a pair of oligonucleotide-labeled secondary antibodies, when the primary antibodies have bound to their antigens in close proximity (less than 40 nm). Fluorescent signals were observed throughout the striatum (Fig. 5A) and in the CA3 and CA1 regions of the hippocampus (Fig. 5B,C), suggesting a close interaction of CHL1 and 5-HT2c receptor in striatal GABAergic medium spiny neurons and in striatal and hippocampal GABAergic interneurons.
When tissue from CHL1-deficient mice was used as negative control for proximity ligation assay with antibodies against CHL1 and the 5-HT2c receptor, no fluorescent signals were detectable, for example, in the hippocampal CA3 region (Fig. 5D), showing the specificity of the close interaction between CHL1 and 5-HT2c receptor.
Given that PTEN binds to the 5-HT2c receptor (Ji et al., 2006), a proximity ligation assay was performed with antibodies against PTEN and the 5-HT2c receptor to verify this interaction with a positive control. Fluorescent signals were found in wild-type and CHL1-deficient brain tissues with higher signal levels in the CHL1-deficient than wild-type brain regions in the hippocampal CA3 region (Fig. 5E,F). This result not only confirms the close interaction between PTEN and the 5-HT2c receptor, but also indicates that the level of this close interaction is enhanced in the absence of CHL1.
Reduced reactivity of CHL1-deficient mice to novelty results from abnormal signaling through constitutively active 5-HT2c receptor isoforms
Based on the finding that CHL1 and 5-HT2c receptor interact with each other predominantly in inhibitory GABAergic neurons and the findings that CHL1 and the 5-HT2c receptor are associated with mental disorders in humans and with abnormal novelty-seeking exploratory behavior in mice, we assumed that binding of CHL1 to the 5-HT2c receptor is physiologically relevant in the intact adult mouse brain to control behavior. Thus, we investigated whether the altered exploratory behavior of CHL1-deficient mice resulted from alterations in responsiveness or activity of serotonin-reactive and/or constitutively active 5-HT2c receptor isoforms. For this, CHL1-deficient and wild-type mice were monitored for locomotor activity in an open field test, as a read-out of exploratory behavior, immediately after intraperitoneal injection of the serotonin agonist Ro60-0175, which activates serotonin-activatable receptor isoforms, the antagonist (i.e. inverse agonist) SB-206553, which inactivates constitutively active receptor isoforms, or of vehicle solution. CHL1-deficient females moved less than wild-type females when vehicle solution was applied (Fig. 6A), indicating that CHL1-deficient females show a pronounced and prolonged reduction of reactivity to the novel environment in the open field. In contrast to CHL1-deficient females, CHL1-deficient males showed reduced locomotor activity only within the first 5 min in the open field after application of vehicle solution (Fig. 6B), indicating that CHL1-deficient male mice show a less pronounced and short-term reduction of reactivity to novelty in the open field. The distance moved by wild-type and CHL1-deficient females within 30 min and males within 5 min in the open field was not altered by Ro60-0175 treatment (Fig. 6C,D). Wild-type females and males did not respond to SB-206553, whereas the distance moved by CHL1-deficient females within 30 min and males within 5 min increased after SB-206553 treatment and reached values observed for wild-type females and males treated with SB-206553 or vehicle (Fig. 6C,D). Given that rearing is a typical novelty-induced behavior in males and that CHL1-deficient males show drastically reduced rearing within the first few minutes in the open field (Morellini et al., 2007), we analyzed whether SB-206553 treatment also affected the rearing behavior of CHL1-deficient males in the open field. CHL1-deficient males showed less rearing activity within the first 5 min in the open field after injection of vehicle solution when compared to wild-type males, but this reduced rearing activity was not observed after application of SB-206553 (Fig. 6E).
These results indicate that the behavioral changes caused by CHL1-deficiency are due to a dysregulation of SB-206553-sensitive constitutively active 5-HT2c receptor isoforms and that Ro60-0175-reactive serotonin-activatable 5-HT2c receptor isoforms are not affected by the lack of CHL1. Moreover, the results suggest that an altered 5-HT2c-receptor-mediated signaling in CHL1-deficient mice is associated with reduced reactivity to novel stimuli and/or abnormalities in exploratory behavior.
Lack of CHL1 alters phosphorylation of the 5-HT2c receptor and interaction of the receptor with PTEN and β-arrestin 2
Given that activity of the 5-HT2c receptor and downstream signaling pathways triggered by activated 5-HT2c receptor are regulated by phosphorylation and dephosphorylation of the 5-HT2c receptor, we tested whether ablation of CHL1 affects the association of the receptor with PTEN, which dephosphorylates the 5-HT2c receptor (Ji et al., 2006).
Because PTEN and CHL1 bind directly to the amino acids 284–298 (Ji et al., 2006) and 281–309 (present study) of the 5-HT2c-receptor, respectively, we first investigated whether PTEN interferes with the binding of CHL1-ICD to the 5-HT2c receptor and performed pulldown experiments using a detergent brain extract from wild-type mice and CHL1-ICD. In the presence of a 5-fold molar excess of PTEN–GST or GST, as the negative control, levels of the 5-HT2c receptor in the CHL1-ICD precipitate were reduced by ∼10% (12.2±5.8%, mean±s.d., n=3) (Fig. 7A), indicating that PTEN does not compete with CHL1 for 5-HT2c receptor binding. However, in the presence of the tat/281-309 peptide containing the 5-HT2c receptor sequence stretch 281–309, levels of the 5-HT2c receptor in the CHL1-ICD precipitate were reduced by more than 40% (41.4±3.4%, mean±s.d., n=3) compared to the levels in the absence of the peptide (Fig. 7A). This result indicates that the peptide which contains the CHL1-binding site of the 5-HT2c receptor interferes with the interaction between the 5-HT2c receptor and CHL-ICD.
To substantiate the notion that the receptor-derived peptide tat/281-309 interferes with the interaction between CHL1 and the 5-HT2c receptor and to investigate whether PTEN binds to the receptor in the absence of CHL1, we performed immunoprecipitation experiments with antibodies against CHL1, PTEN and 5-HT2c receptor using brain extracts from wild-type and CHL1-deficient mice in the absence or presence of the tat/281-309 peptide. Similar amounts of PTEN were found in the PTEN immunoprecipitates of both genotypes, whereas approximately three times (2.9±0.4, mean±s.d., n=3) higher amounts of HT2c receptor were observed in the PTEN immunoprecipitate from CHL1-deficient brain extracts than in the immunoprecipitates from wild-type brain extracts (Fig. 7B), indicating that more PTEN associates with the receptor in the absence of CHL1. No CHL1 was detectable in the PTEN immunoprecipitates from extracts of either genotype (Fig. 7B), implying that CHL1 does not directly interact with PTEN. In the presence of the tat/281-309 peptide, the levels of the 5-HT2c receptor in the PTEN immunoprecipitates from wild-type mice were approximately two times (2.2±0.2, mean±s.d., n=3) higher than the levels in the immunoprecipitates obtained in the absence of the peptide, whereas the amounts of 5-HT2c receptor in the PTEN immunoprecipitates from extracts of CHL1-deficient mice were only slightly reduced in the presence of the tat/281-309 peptide (12.8±5.4% less, mean±s.d., n=3) (Fig. 7B). This result indicates that the receptor-derived peptide disturbs the interaction between the 5-HT2c receptor and CHL1, but does not interfere with the interaction between the receptor and PTEN in the absence of CHL1.
In the CHL1 immunoprecipitates from wild-type mice, similar levels of CHL1 were found in the absence and presence of the tat/281-309 peptide, whereas no CHL1 was detectable in the immunoprecipitates from CHL1-deficient mice (Fig. 7C). In the presence of the peptide, reduced levels (by 60.6±6.3%, mean±s.d., n=3) of the 5-HT2c receptor were found in the CHL1 immunoprecipitates from wild-type mice, confirming that the peptide interferes with the interaction between CHL1 and the receptor. PTEN was not detectable in the CHL1 immunoprecipitates from wild-type mice (Fig. 7C), supporting the view that a direct interaction between CHL1 and PTEN can be excluded.
After immunoprecipitation with the rabbit anti-5-HT2c-receptor antibody, reduced levels (by 31.5±5.5%, mean±s.d., n=3) of the 50-kDa receptor form were detectable by western blot analysis with the goat antibody against the receptor in the presence of the tat/281-309 peptide relative to the levels in its absence (Fig. 7D). Relative to the receptor levels in the immunoprecipitates, the PTEN levels were ∼1.5-fold (1.6±0.1, mean±s.d., n=3) higher in immunoprecipitates from CHL1-deficient extracts and from wild-type extracts in the presence of the peptide, when compared to the levels of PTEN in immunoprecipitates from wild-type extracts in the absence of peptide (Fig. 7D). These results show that association of PTEN with the receptor is enhanced in the absence of CHL1 and that the peptide interferes with the interaction between CHL1 and the 5-HT2c receptor, thereby allowing PTEN to associate with the receptor in the absence of CHL1.
It has been reported that phosphorylation of the 5-HT2c receptor promotes binding of β-arrestin 2 to the 5-HT2c receptor and that an enhanced interaction of β-arrestin 2 with SB-206553-sensitive constitutively active 5-HT2c receptor isoforms modulates the activity of these isoforms (Marion et al., 2004). Given that we observed that the behavioral changes of CHL1-deficient mice are due to dysregulation of SB-206553-sensitive constitutively active 5-HT2c receptor isoforms, we analyzed whether binding of β-arrestin 2 to the 5-HT2c receptor was altered in the absence of CHL1. Relative to the β-arrestin 2 levels in the immunoprecipitates from wild-type brains without peptide, the levels in immunoprecipitates from CHL1-deficient brains were reduced by ∼70% (69.5±4.5, mean±s.d., n=3), whereas the levels in the immunoprecipitates from wild-type brains were decreased by ∼35% (36.8±3.5, mean±s.d., n=3) in the presence of the tat/281-309 peptide (Fig. 7D). This result indicates that the association of β-arrestin 2 with the 5-HT2c receptor is reduced in the absence of CHL1, suggesting that CHL1 regulates the phosphorylation of the receptor and the consecutive association of β-arrestin 2 with the receptor.
To test whether phosphorylation of the 5-HT2c receptor is altered in the absence of CHL1, we used wild-type and CHL1-deficient brain extracts for immunoprecipitation with the rabbit antibody against the 5-HT2c receptor and subjected the immunoprecipitates to western blot analysis with a antibody against phosphorylated serine (phospho-serine) residues followed by a goat antibody against the 5-HT2c receptor. Similar amounts of the 5-HT2c receptor were immunoprecipitated from brains of both genotypes (Fig. 7E), whereas 5-fold (4.7±0.7, mean±s.d., n=3) lower levels of phosphorylated 5-HT2c receptor were observed in the immunoprecipitates from CHL1-deficient extracts when compared to immunoprecipitates from wild-type brains (Fig. 7E). These results indicate that CHL1 regulates the phosphorylation of the 5-HT2c receptor and suggest that increased association of PTEN with the 5-HT2c receptor in the absence of CHL1 maintains the receptor in a dephosphorylated state.
CHL1 regulates signaling through constitutively active 5-HT2c receptor isoforms
The results from behavioral experiments suggest that CHL1 regulates the activity of constitutively active SB-206553-sensitive 5-HT2c receptor isoforms. To confirm the interaction of CHL1 with these isoforms, we performed immunoprecipitation experiments with transfected COS-7 cells that either expressed the non-edited constitutively active INI isoform alone or co-expressed the INI isoform with CHL1. Pronounced levels of the INI isoform were co-immunoprecipitated by a CHL1 antibody from cells co-expressing CHL1, whereas low levels were non-specifically immunoprecipitated from CHL1-negative cells (Fig. 8A), indicating that CHL1 associates with the constitutively active 5-HT2c receptor isoforms.
Next, the transfected cells were taken to analyze whether CHL1 regulates signaling of the constitutively active 5-HT2c receptor isoforms. Given that the non-edited INI isoform can activate phospholipase D (PLD) (McGrew et al., 2004) and because SB-206553 inhibits Erk (Erk1/2, also known as MAPK3 and MAPK1) phosphorylation triggered by the INI isoform (Labasque et al., 2010), we asked whether CHL1 affects PLD activity and Erk phosphorylation by regulating the constitutively active INI isoform. Transfected COS-7 cells expressing the INI isoform alone showed similar levels of PLD activity and Erk phosphorylation as in cells co-expressing the INI isoform and CHL1 (Fig. 8B,C). Treatment with SB-206553 reduced PLD activity and phosphorylated Erk (phospho-Erk) levels in cells expressing the INI isoform alone, whereas SB-206553 had no effect on PLD activity and Erk phosphorylation in cells co-expressing CHL1 and the INI isoform (Fig. 8B,C). These results indicate that binding of SB-206553 to the INI isoform triggers signaling pathways leading to reduction in phospho-Erk levels and PLD activity when CHL1 is absent, indicating that CHL1 controls signal transduction pathways mediated by the constitutively active INI isoform.
DISCUSSION
The 5-HT2c serotonin receptor is essential for normal nervous system functions. Several human mental disorders have been linked to abnormalities of the 5-HT2c receptor and disturbance of 5-HT2c-receptor-regulated physiological processes. Because many different 5-HT2c receptor isoforms exist, receptor-triggered signal transduction pathways and cellular responses are complex. Different 5-HT2c receptor isoforms are generated by adenosine-to-inosine RNA editing: the non-edited isoform is constitutively active, the partially edited isoforms exhibit intermediate constitutive activity and the fully edited isoform becomes active only in the presence of its ligand serotonin. Thus, elucidation of the molecular mechanisms underlying regulation of 5-HT2c-receptor-mediated signaling in the context of interaction partners as possible modulators of the 5-HT2c receptor isoform activities is of interest. Such knowledge might contribute to a better understanding of 5-HT2c receptor functions in the normal brain as well as in the pathogenesis of 5-HT2c-receptor-related mental disorders, such as addiction, obesity, epilepsy, anxiety, sleep disorders, schizophrenia, major depression and bipolar disorder (Berg et al., 2008; Lee et al., 2010). Here, we identified CHL1 as a new binding partner of the 5-HT2c receptor in vitro and in vivo. In humans, CHL1 has been linked to neurological and psychiatric disorders, such as mental retardation, schizophrenia, epilepsy and autism, but so far has not been associated with addiction, obesity, anxiety or sleep disorders (Angeloni et al., 1999a,b; Sakurai et al., 2002; Frints et al., 2003; Chen et al., 2005; Chu and Liu, 2010; Tam et al., 2010; Cuoco et al., 2011; Salyakina et al., 2011; Shoukier et al., 2013). CHL1-deficient mice showed reduced reactivity to novelty or external stimuli, but no anxiety-related behavior (Montag-Sallaz et al., 2002; Pratte et al., 2003; Morellini et al., 2007; Pratte and Jamon, 2009). Mice expressing only the non-edited or the fully-edited isoform of the 5-HT2c receptor display anxiety- or depression-like symptoms and 5-HT2c-receptor-deficient mice display reduced reactivity to a novel environment or external stimuli, as well as alterations in sleep, feeding and reward-associated behaviors (for review and references, see Heisler et al., 2007; Fletcher et al., 2009; Iwamoto et al., 2009; Mombereau et al., 2010; Pennanen et al., 2013). Alterations in sleep, feeding and reward-associated behaviors have so far not been reported for CHL1-deficient mice.
Here, we show that the 29 N-terminal amino acids in the intracellular domain of CHL1 mediate binding to a sequence stretch in the third intracellular domain of the 5-HT2c receptor and provide evidence that the interaction of CHL1 with the receptor effectively reduces the binding of PTEN to the receptor and enhances the interaction of the receptor with β-arrestin 2, which is involved in regulating the activity of constitutively active 5-HT2c receptors (Marion et al., 2004). Given that PTEN dephosphorylates the 5-HT2c receptor (Ji et al., 2006), an enhanced interaction between the 5-HT2c receptor and PTEN in the absence of CHL1 leads to prolonged dephosphorylation of 5-HT2c receptor in CHL1-deficient mice, and sustained dephosphorylation of the 5-HT2c receptor at the plasma membrane by PTEN might affect the activity of the 5-HT2c receptor and the signal transduction pathways downstream of the 5-HT2c receptor. Furthermore, reduced interaction of the dephosphorylated receptor with β-arrestin 2 might affect receptor activity and receptor-dependent downstream signaling. The results from behavioral analysis indicate that CHL1 predominantly controls the activity of SB-206553-sensitive constitutively active isoforms of the 5-HT2c receptor and regulates the signaling through these isoforms, whereas CHL1 does not appear to regulate expression of the serotonin-activatable 5-HT2c receptor isoforms. Application of SB-206553, which specifically reduces the activity of the constitutively active 5-HT2c receptors isoforms, increases locomotor activity of CHL1-deficient, but not wild-type mice, and reverted the phenotype of CHL1-deficient mice to that of wild-type mice. This finding strongly suggests that the functions of constitutively active 5-HT2c receptor isoforms are abnormal in the absence of CHL1 and that hypolocomotion of CHL1-deficient mice might be due to inhibition of locomotion mediated by the constitutively active 5-HT2c receptor isoforms. Indeed, association of CHL1 with the constitutively active 5-HT2c receptor isoforms regulates the activities of these receptor isoforms and controls downstream signal transduction pathways; Erk phosphorylation and PLD activity is reduced after SB-206553 treatment of cells expressing the constitutively active INI isoform but not CHL1, whereas cells co-expressing the INI isoform and CHL1 did not respond to SB-206553.
On the basis of these observations, we propose that ablation of CHL1 reduces the activity of constitutively active 5-HT2c receptor isoforms resulting in an enhancement of tonic inhibition of locomotor activity. It is likely that absence of CHL1 leads to an enhancement of inhibitory transmission as seen in the CA1 region of juvenile CHL1-deficient mice (Nikonenko et al., 2006). Of note, CHL1 and 5-HT2c receptor are co-expressed in GABAergic interneurons (Serrats et al., 2005; Liu et al., 2007; Ango et al., 2008; Jakovcevski et al., 2009), and they colocalize in GABAergic neurons of the putamen and hippocampal CA1 and CA3. Enhanced activity of constitutively active 5-HT2c receptor isoforms in GABAergic neurons (e.g. medium spiny neurons and/or interneurons) in the absence of CHL1 might result in increased inhibition of neurons innervated by these GABAergic neurons, resulting in reduced locomotor activity. Binding of SB-206553 to dysregulated constitutively active 5-HT2c receptor isoforms in CHL1-deficient GABAergic interneurons might trigger signaling pathways which, for example, reduce PLD activity and Erk phosphorylation, abolishing enhanced tonic inhibition of locomotion by GABAergic neurons. Further investigations of the role of CHL1 in serotonergic and GABAergic transmission and/or 5-HT2c-receptor-mediated in signaling pathways in different brain areas should provide novel insights into the functions of the serotonergic system under normal and pathological conditions with hope for therapy (see, for instance, Giorgetti and Tecott, 2004; Meltzer et al., 2012).
MATERIALS AND METHODS
Animals
CHL1-deficient mice (Montag-Sallaz et al., 2002), which have been back-crossed onto the C57BL/6J background for more than eight generations, and their age-matched wild-type littermates, as well as C57BL/6J mice were bred and maintained at the Universitätsklinikum Hamburg-Eppendorf. Animals were housed at 25°C on a 12 h-light–12-h-dark cycle with ad libitum access to food and water. Adult 10- to 12-week-old mice of either sex were used. All animal experiments were approved by the local authorities of the State of Hamburg and conform to the guidelines set by the European Union.
Antibodies and reagents
Polyclonal antibodies against the extracellular domain of CHL1 (Chen et al., 1999; Rolf et al., 2003) were used for western blot analysis and a polyclonal goat antibody against intracellular CHL1 epitopes from R&D Systems (AF-2147) was used for immunoprecipitation, immunostaining and proximity ligation assay. Polyclonal antibodies against 5-HT2c receptor from Santa Cruz Biotechnology (goat N-19, sc-15081; mouse D-12, sc-17797) (Heidelberg, Germany) were used for western blot analysis, immunostaining and proximity ligation assay and a polyclonal rabbit antibody (Pierce, PA1-18069) from Life Technologies was used for immunoprecipitation. PTEN antibody (clone 6H2.1) from Merck Chemicals was used for immunoprecipitations and western blot analysis. Antibodies against β-arrestin 2, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and total Erk were from Santa Cruz Biotechnology, anti-phospho-Erk antibody [phospho-p44/42 (Erk1/2) (T202/Y204) antibody] was from New England Biolabs and monoclonal parvalbumin antibody was from Sigma-Aldrich (p-3088). Cloning and production of recombinant His-tagged or GST-tagged intracellular domains of CHL1 (CHL1-ICD), His-tagged L1 (L1-ICD) and His-tagged NCAM140 (NCAM-ICD) have been described previously (Xiao et al., 2009; Andreyeva et al., 2010). The plasmid pGEX3X-GST-PTEN (Addgene) was used for the expression of recombinant PTEN-tagged with GST. Production of PTEN–GST was as described previously (Xiao et al., 2009). Horseradish peroxidase (HRP)-coupled and fluorescent dye-coupled secondary antibodies were from Jackson Laboratory. The synthetic phage peptide (KKDRRPRSLHPH), the 5-HT2c-receptor-derived peptide (NPNPDQKPRRKKKEKRPRGTMQAINNEKK) and the peptide tat/281-309 containing the HIV tat sequence YGRKKRRQRRR (Schwarze and Dowdy, 2000) and the amino acids 281–309 of the murine 5-HT2c receptor (YGRKKRRQRRRNPNPDQKPRRKKKEKRPRGTMQAINNEKK) were from Schafer-N (Copenhagen, Denmark). Ro60-0175 fumarate (CAS 169675-09-6) and SB-206553 hydrochloride, (CAS 158942-04-2) were from Biozol (Eching, Germany). For the determination of the PLD activity, the Amplex® Red Phospholipase D Assay Kit (A12219) from Molecular Probes (LifeTechnologies) was used.
Phage display
The Ph.D.-12™ Phage Display Peptide Library (New England Biolabs) displaying 108–1010 random 12-mer peptides at the pili of M13-like phage particles in fusion with the N-terminus of the pVIII major coat protein and the intracellular domain of CHL1 (CHL1-ICD) were used for screening of peptides binding to CHL1-ICD. All selection steps were performed according to the Ph.D.-12™ Phage Display Peptide Library Kit instruction manual version 2.0 (New England Biolabs) and have been described previously (Wang et al., 2011).
Western blot analysis, immunoprecipitation, biochemical crosslinking, pulldown assay and ELISA
Western blot analysis, immunoprecipitation, the pulldown assay and ELISA as well as biochemical crosslinking were performed as described in detail previously (Makhina et al., 2009; Xiao et al., 2009). For preparation and immunoprecipitation of phosphorylated proteins PhosSTOP Phosphatase Inhibitor Cocktail Tablets (Roche) were used and TBS buffers were used for detection. For visualization of immunopositive bands by western blot analysis the chemiluminescent substrate with extended duration (Pierce) was used. Band intensities were densitometrically quantified using ImageJ software (http://imagej.nih.gov/ij/).
Tissue preparation, immunohistochemistry and proximity ligation assay
Brains from 3-month-old mice were fixed with 4% formaldehyde in 0.1 M cacodylate buffer, pH 7.3, at room temperature, post-fixed overnight at 4°C in the formaldehyde solution, incubated in 15% sucrose solution in 0.1 M cacodylate buffer, pH 7.3, for 2 days at 4°C, frozen by immersion for 2 min in 2-methyl-butane (isopentane) precooled to −80°C and stored in liquid nitrogen. Serial coronal 25-μm-thick sections were cut in a cryostat (Leica CM3050, Leica Instruments) and collected on SuperFrost Plus glass slides (Carl Roth). Antigen retrieval was performed by incubating the sections in 10 mM sodium citrate solution (pH 9.0) for 30 min at 80°C, followed by blocking of non-specific binding sites with phosphate-buffered saline, pH 7.3 (PBS) containing 0.2% Triton X-100, 0.02% sodium azide, and 5% normal donkey serum for 1 h at room temperature. The sections were incubated with primary antibodies diluted in PBS overnight at 4°C in a humidified chamber. For immunohistochemistry, the sections were washed in PBS (3×15 min at room temperature), incubated at room temperature with donkey Cy3- or Cy2-conjugated secondary antibodies diluted in PBS for 2 h, washed in PBS, and incubated for 10 min at room temperature with bis-benzimide solution (Hoechst dye 33258, 5 μg/ml in PBS, Sigma-Aldrich) to stain cell nuclei. Finally, the sections were washed again and mounted with Fluoromount G (Southern Biotechnology Associates, Birmingham, AL, USA) and microphotographs were taken on a Leica confocal laser scanning microscope at a 1024×1024 digital resolution. For proximity ligation assay, the sections were incubated with a pair of secondary antibodies conjugated with oligonucleotides (Duolink anti-goat PLA probe MINUS and Duolink anti-rabbit PLA probe PLUS) and the Duolink detection reagent RED (Sigma-Aldrich) was used. Coverslips were mounted with Roti-Mount FluorCare DAPI (Carl Roth, Karlsruhe, Germany) and confocal images were taken with an Olympus Fluoview FV1000 confocal laser scanning microscope (Hamburg, Germany) in sequential mode with a 60× objective.
Drug administration and measurement of locomotor activity
The 5-HT2c-receptor-specific agonist Ro60-0175 or inverse agonist SB-206553 were dissolved in saline and injected intraperitoneally before behavioral testing. All injections (1 mg/kg body weight) were given in a volume of 0.1 ml. Saline injection served as vehicle control. Locomotion was evaluated by the open field test. The open field consisted of a wooden box (50×50×40 cm) laminated with rough, matted, light-gray resin and illuminated by a white bulb (100 Lux). After injection, each mouse was gently introduced into a cylinder of opaque Plexiglas placed at one corner of the box for 5 s. As the cylinder was lifted, the mouse could move freely in the arena for 30 min. Locomotor activity was measured at 5 min intervals and cumulative counts were taken for data analysis with the software EthoVision (Noldus, Wageningen, The Netherlands).
Cloning of expression vectors and transfection of COS-7 cells
For the single expression of the INI isoform and the co-expression of the INI isoform and CHL1, COS-7 cells (ATCC® CRL-1651™; tested for contamination every 3 months) were transfected with INI or INI and CHL1 coding bicistronic mammalian expression vectors using Turbofect™ (Thermo Fisher Scientific). For cloning, the InFusion Cloning Kit (Clontech) and PCR amplifications with Phusion Polymerase (New England Biolabs) were used. The product resulting from PCR using the INI-coding plasmid (Herrick-Davis et al., 1999) and the primers pAG/HT2c fw (5′-GTTGCCTTCGCCCCGATGGTGAACCTTGGCAACG-3′; start codon in bold) and HT2c/IRES rev (5′-TCGCACGATTACCATTTA- CACACTACTAATCCTCTC-3′; stop codon is underlined) were used for cloning of INI-coding cDNA. The INI-coding plasmid was a kind gift from Katharine Herrick-Davis (Albany Medical College, Albany, NY). For cloning of CHL1-coding cDNA, the product from PCR with CHL1-coding vector (Holm et al., 1996) and the primers Chl1 fw (5′-ATGATGGAATTGCCATTATGT-3′) and pCAG/Chl1 rev (5′-TGGCGGCCGGCCGCTTCATGCCCGGAGTGGGAA-3′) were used. For amplification of the IRES element, the pCAG-GFP (Addgene, Cambridge, MA) and the primers IRES fw (5′-ATGGTAATCGTGCGAGAGG-3′) and IRES/CHL1 rev (5′-TGGCAATTCCATCATGGTTGTGGCCATATTATCAT-3′) were used. The products from PCR with pCAG-GFP and the primer pAG fw (5′-AGCGGCCGGCCGCCAGCACAGTGG-3′) and pCAG rev (5′-CGGGGCGAAGGCAACGCAGCGACT-3′) or IRES/pCAG rev (5′-TGGCGGCCGGCCGCTGGTTGTGGCCATATTATCAT-3′) were used for cloning of INI-coding or of INI- and CHL1-coding PCR products. The vector resulting from the fusion of the pCAG PCR product with 5-HT2c receptor-, IRES- and CHL1-coding PCR products was used to co-express the 5-HT2c receptor and CHL1 and the vector resulting from the fusion of the pCAG PCR product with the 5-HT2c-receptor- and IRES-coding PCR products was used for single expression of the receptor.
Acknowledgements
We are grateful to Eva Kronberg for excellent animal care, to Jeanette Reinshagen (Fraunhofer IME ScreeningPort), Ute Bork and Emanuela Szpotowicz for excellent technical assistance, Igor Jakovcevski for technical support and fruitful discussions and Katharine Herrick-Davis for plasmids.
Footnotes
Author contributions
R.K. and M.S. conceived the study; R.K. and G.L. designed experiments; H.C., N.K., J.K., A.K., K.G., G.L. and R.K. performed experiments and analyzed the data together with M.S.; R.K. and M.S. wrote the manuscript.
Funding
M.S. is supported by the Li Kashing Foundation at Shantou University Medical College.
References
Competing interests
The authors declare no competing or financial interests.