Physical and Functional Interaction between 5-HT6 Receptor and Nova-1

Abstract

5-HT₆ receptor (5-HT₆R) is implicated in cognitive dysfunction, mood disorder, psychosis, and eating disorders. However, despite its significant role in regulating the brain functions, regulation of 5-HT₆R at the molecular level is poorly understood. Here, using yeast two-hybrid assay, we found that human 5-HT₆R directly binds to neuro-oncological ventral antigen 1 (Nova-1), a brain-enriched splicing regulator. The interaction between 5-HT₆R and Nova-1 was confirmed using GST pull-down and co-immunoprecipitation assays in cell lines and rat brain. The splicing activity of Nova-1 was decreased upon overexpression of 5-HT₆R, which was examined by detecting the spliced intermediates of gonadotropin-releasing hormone (GnRH), a known pre-mRNA target of Nova-1, using RT-PCR. In addition, overexpression of 5-HT₆R induced the translocation of Nova-1 from the nucleus to cytoplasm, resulting in the reduced splicing activity of Nova-1. In contrast, overexpression of Nova-1 reduced the activity and the total protein levels of 5-HT₆R. Taken together, these results indicate that when the expression levels of 5-HT₆R or Nova-1 protein are not properly regulated, it may also deteriorate the function of the other.

Keywords: 5-HT6 receptor; Neuro-oncological ventral antigen 1; Neurological diseases; RNA binding proteins; Serotonin.

Fig. 1. Nova-1 interacts with 5-HT₆R in vitro and in vivo. (A) Schematic diagrams showing the structure of 5-HT₆R, the CT (carboxyl-terminus) of 5-HT₆R as bait and complete Nova-1 protein as prey. (B) When the AH109 strain with 5-HT₆R CT and the Y187 yeast strain with Nova-1 were mated, a blue color was detected. (C) GST pull-down assay showed that Flag-Nova-1 specifically interacted with GST-fused CT of 5-HT₆R (GST-6RCT). (D) Full-length Myc-tagged 5-HT₆R (Myc-5-HT₆R) signal was detected in the immunoprecipitation with anti-Flag. (E) Flag-Nova-1 signal was detected in the immunoprecipitation with anti-Myc. (F) Western blot with anti-5-HT₆R or anti-Nova-1 antibody showed the expression patterns of 5-HT₆R and Nova-1 in the rat brain. (G) GST-6RCT interacted with endogenous Nova-1 in the rat brain. (H) Co-immunoprecipitation assay confirmed the interaction between endogenous 5-HT₆R and Nova-1 in the rat brain. (I) When GST pull-down assays were performed between Nova-1 and CTs of several serotonin receptors (GST-4RCT, GST-6RCT and GST-7BRCT), Nova-1 only interacted with 6RCT but not 4RCT and 7BRCT. GAL4, galactose-responsive transcription factor GAL4; BD, binding domain; AD, activation domain; C, Cortex; Hi, Hippocampus; S, Striatum; B, Brainstem; Hy, Hypothalamus.

Fig. 2. The KH3 domain of Nova-1 is responsible for 5-HT₆R binding. (A) Schematic diagram of HA- and venus C-terminal (VC)-tagged full length and truncated Nova-1 constructs (top). GST-tagged CT of 5-HT₆R (GST-6RCT) (bottom). (B) The expression of HA-VC-tagged full length or truncated Nova-1 constructs was identified by immunoblotting with anti-HA antibody. (C) GST pull-down assay showed that GST-6RCT interacted with all three different Nova-1 constructs. (D) Specific interaction between GST-6RCT and three different Nova-1 constructs was validated by co-immunoprecipitation. (E) Co-immunoprecipitation assay showed that endogenous 5-HT₆R interacted with all three different Nova-1 constructs. (F) 5-HT₆R CT interacted with GST-KH3 but not with GST-Nova-1 (239–419 aa).

Fig. 3. The 5-HT₆R overexpression attenuates the splicing activity of Nova-1 centered on the excision of GnRH intron A. (A) The structure of the GnRH gene and positions of the primers (top). A standard curve was constructed to analyze the quantitative ratio of 1234 and 1A234 cDNAs by competitive PCR using 10 pg of the 1A234 and serial dilutions of the 1234 gene (bottom). (B) Nova-1 increased the rate of intron A excision of GnRH gene in dose-dependent manner when GnRH minigene and Nova-1 were transiently transfected into NIH3T3 cells (F_{(3, 8)}=29.79, p=0.0001; Dunnett's post hoc test: ^***p<0.001). (C) 5-HT₆R decreased the splicing activity of Nova-1 in dose-dependent manner (F_(4,10)=18.92, p=0.0001; Dunnett's post hoc test: ^**p<0.01, ^***p<0.001). (D) C-terminal truncated 5-HT₆R (5-HT₆RΔCT) did not affect the splicing activity of Nova-1 (F_{(3, 8)}=0.4302, p=0.7371). (E) 5-HT₆R dose-dependently decreased the splicing activity of Nova-1 in GT1-1 neuronal cell line in which GnRH and Nova-1 are endogenously expressed (F_{(4, 10)}=12.69, p=0.0006; Dunnett's post hoc test: ^**p<0.01). (F) The splicing activity of Nova-1was not affected by 5-HT₆RΔCT in GT1-1 neuronal cell line (F_{(4. 5)}=0.194, p=0.9314). Data are expressed as means±S.E.M. (n=3 per group).

Fig. 4. The subcellular localization of Nova-1 is shifted from nucleus to cytoplasm upon 5-HT₆R overexpression. (A) Overexpression of 5-HT₆R in GT1-1 cells did not affect the expression levels of Nova-1 (F_{(4, 5)}=0.8457, p=0.5520). (B) Translocation of Nova-1 (red) to cytoplasm was detected in GT1-1 when co-expressed with 5-HT₆R (green). (C) Translocation of Nova-1 to cytoplasm was confirmed in HEK293 cells. (D) Nova-1 was transported from nucleus to the cytoplasm fraction when co-expressed with 5-HT₆R (F_{(3, 9)}=18.31, p=0.0004; Dunnett's post hoc test: ^**p<0.01). β-Tubulin and histone H3 were used as cytoplasmic and nuclear markers, respectively. Data are expressed as means±S.E.M. (n=3 per group).

Fig. 5. Nova-1 overexpression decreases functional activity of 5-HT₆R. (A) Overexpression of Flag-Nova-1 (open circle) suppressed the Ca²⁺ responses induced by 5-HT (10 µM) compared to control (closed circle). F is the fluorescent intensity, and F0 is the initial fluorescent intensity at 480 nm. Inset, The average of ratio (F/F0) measured at the indicated time (#) was significantly reduced by Nova-1 (t=9.834, ^***p<0.001; n=10 per group). (B) Expression of Nova-1 reduced 5-HT-induced Ca²⁺ responses (effect of treatment, F_{(1, 96)}=131, p<0.0001; effect of 5-HT concentration, F_{(11, 96)}=106.8, p<0.0001; interaction between treatment and 5-HT concentration, F_{(11, 96)}=7.91, p<0.0001; n=12 per group). (C) Nova-1 did not affect 5-HT₄R (t=1.237, p=0.2224; n=24 per group) and 5-HT_7BR-mediated Ca²⁺ responses (t=0.9698, p=0.3386; n=19 per group). (D) Expression of Nova-1 decreased 5-HT-induced ERK phosphorylation levels in a dose-dependent manner (F_{(5, 12)}=6.523, p=0.0037; Dunnett's post hoc tests, ^*p<0.05, ^**p<0.001; n=3 per group). Data are expressed as means±S.E.M..

Fig. 6. Nova-1 overexpression decreases the expression levels of 5-HT₆R. (A, B) 5-HT₆R levels were reduced upon overexpression of Nova-1 in HEK293/6R stable cells (F_{(4, 5)}=3.959, p=0.0818; Dunnett's post hoc test: ^*p<0.05; A) as well as in GT1-1 cells (F_{(4, 5)}=12.62, p=0.008; Dunnett's post hoc test: ^*p<0.05, ^**p<0.01; B). (C) Reduced 5-HT₆R expression caused by Nova-1 was rescued by MG132, a proteasome inhibitor (effect of MG132, F_{(1. 12)}=0.9567, p=0.0281; effect of Nova-1, F_{(2, 12)}=0.9567, p=0.4116; interaction between MG132 and Nova-1, F_{(2, 12)}=10.85, p=0.0020; Student-Newman-Keuls post hoc test: ^*p<0.01, ^**p<0.01). Data are expressed as means±S.E.M. (n=3 per group).