NF1 Is a Direct G Protein Effector Essential for Opioid Signaling to Ras in the Striatum

Abstract

It is well recognized that G-protein-coupled receptors (GPCRs) can activate Ras-regulated kinase pathways to produce lasting changes in neuronal function. Mechanisms by which GPCRs transduce these signals and their relevance to brain disorders are not well understood. Here, we identify a major Ras regulator, neurofibromin 1 (NF1), as a direct effector of GPCR signaling via Gβγ subunits in the striatum. We find that binding of Gβγ to NF1 inhibits its ability to inactivate Ras. Deletion of NF1 in striatal neurons prevents the opioid-receptor-induced activation of Ras and eliminates its coupling to Akt-mTOR-signaling pathway. By acting in the striatal medium spiny neurons of the direct pathway, NF1 regulates opioid-induced changes in Ras activity, thereby sensitizing mice to psychomotor and rewarding effects of morphine. These results delineate a novel mechanism of GPCR signaling to Ras pathways and establish a critical role of NF1 in opioid addiction.

Keywords: G proteins; GPCR signaling; addiction; neurofibromatosis; opioids; striatum.

Figure 1. NF1 is a novel G protein effector downstream from MORs activation

(A) Schematic of the assay design to study membrane recruitment of NF1 in response to MOR activation. Association of NanoLuc-tagged NF1 with membrane targeted Venus is expected to result in BRET signal. (B) Real time kinetics of BRET signal change upon NF1 plasma membrane recruitment. Arrow indicates time of morphine application. (C) Quantification of maximal amplitude of BRET signals from 3 independent experiments. **p<0.01, One-Way ANOVA post hoc Tukey’s test. (D) Schematic representation of NF1 domain composition and truncation mutants used in this study. (E) Co-immunoprecipitation of endogenous NF1 with over-expressed Gβ1γ2 complex in transfected HEK293T cells. (F) Co-immunoprecipitation of Sec14/PH module with Gβ1γ2 complex in transfected HEK293T cells. (G) Mutual stabilization of Sec14/PH module and Gβ1γ2 upon co-expression in HEK293T cells. (H) Direct interaction between purified recombinant Sec14/PH and Gβ1γ2 proteins studied in pull-down assays. (I) Upper, Schematic diagram of BRET strategy to study NF1-Gβγ interaction in living cells. Lower, Analysis of Gβγ binding to NF1 constructs by titrating ratios of interacting partners. (J) Quantification of maximal BRET ratios generated by various NF1 constructs from three independent experiments. **p<0.01, One-Way ANOVA post hoc Tukey’s test. (K) Upper: Schematic representation of the BRET assay monitoring interaction between Sec14/PH and Gβγ upon MORs activation and inactivation. Lower: Real-time kinetics of BRET signal in response to changes in MOR activity. (L) Quantification of the dose-response relationship of Gβγ-NF1 interaction from 3 independent experiments. Gβγ scavenger GRK3ct blunted the BRET response between Sec14/PH module and Gβγ. Data are represented as mean ± SEM. See also Figure S1.

Figure 2. Gβγ inhibits GTPase accelerating protein (GAP) activity of NF1

(A) Left, Diagram of single turnover principle of Ras GTP hydrolysis assay. Right, Effects of recombinant Gβγ on NF1 fragment (GRD/Sec14/PH)-stimulated GTPase activity of N-Ras. (B) Quantification of the effect of Gβγ on GTPase activity of NF1. (C) Left, Diagram of multiple turnover of Ras GTP hydrolysis assay principle. The assay involves both GAP (NF1) and GEF (sos1). Right, Both RasGEF and RasGAP are necessary for Ras GTP hydrolysis in the multiple turnover assay. (D) Gβγ inhibits Ras GTP hydrolysis rate in dose-dependent manner with IC₅₀ for Gβγ of 0.187 ± 0.044 μM. Data are represented as mean ± SEM.

Figure 3. NF1 is essential for Ras activation by MORs in cultured striatal medium spiny neurons (MSNs)

(A) Schematic of Ras FLIM-FRET sensor. Increase in binding between mGFP-RasGTP and mRFP-Raf-RBD upon Ras activation is measured as a decrease in GFP fluorescence lifetime and quantified as change in binding fraction. (B) Representative fluorescence lifetime image of striatal MSNs before and 10 minutes after 10 μM morphine application. Warmer colors indicate decrease in lifetime, which corresponds to higher Ras activity. (C) Average time course of Ras activation (change in binding fraction) of MSNs in response to 10 μM morphine application (black arrow). Average includes neurons, which responded and those that showed no response to morphine. (D) Morphine induced Ras activation of individual MSNs. (E) Schematic of conditional Nf1 ablation strategy in cultured MSNs. Cultured MSNs from neonatal Nf1^flx/flx mice were infected with AAV-Cre to induce Nf1 elimination. (F) Representative time course of Ras activation (change in binding fraction) in response to morphine application (black arrow) in WT (AAV-Control) and Nf1 cKO (AAV-Cre) MSNs. (G) Quantification of Ras response to morphine in control and Nf1 cKO (Cre) cultures. Unpaired t test *p=0.02, n=26, 15. (H) Schematic of strategy to prevent Gβγ binding to NF1 by expressing the minimal binding module of NF1 (NF1-DN). (I) Representative time-course of Ras activation (change in binding fraction) in response to morphine application (black arrow) in control and NF1-DN expressing MSNs. (J) Quantification of Ras response to morphine in control and NF1-DN expressing cultures. Unpaired t test **p=0.007, n=19, 18. (K) Schematic outlining strategy for expression of Ras sensor in D1R or D2R expressing MSNs. (L) Representative time course of Ras activation (change in binding fraction) in response to morphine application (black arrow) in D1R and D2R expressing MSNs. (M) Quantification of Ras response to morphine in D1R and D2R expressing MSNs. Unpaired t test **p=0.007, n=13,15. Data are represented as mean ± SEM. See also Figure S2.

Figure 4. Striatal NF1 is essential for morphine-induced Ras and its downstream signaling cascade activation

(A) Generation of striatal specific Nf1 knockout mice (Nf1 cKO) by crossing conditional NF1^flx/flx line with MSNs specific Rgs9-Cre line. (B) Representative images of Nissl-stained coronal brain sections from control and Nf1 cKO mice. Stm, striatum; NAc, nucleus accumbens; aca, anterior part of anterior commissure (Scale bar, 1 mm). We found average volumes of both dorsal striatum (16.4±0.8 mm³ in Nf1 cKO vs 17.2±0.3 mm³ in control littermates, n=4 mice) and nucleus accumbens (1.4±0.1 mm³ in Nf1 cKO vs 1.3±0.1 mm³ in control littermates, n=4 mice) to be unaffected by NF1 deletion. (C) Impact of NF1 elimination on baseline Ras activation and signaling to downstream kinase pathways in striatum. (D) Quantification of Western blot data with activities normalized to samples from control mice. Data are represented as mean ± SEM. **p<0.01, Student’s t-test. (E) Schematic representation of Ras-ERK and Ras-Akt signaling. Ras is activated in response to both GPCRs and receptor tyrosine kinases (RTKs), which in turn is able to activate both Raf/MEK/ERK and PI3K/Akt pathways. Activated Akt can further modulate GSK3 and mTOR activity. High level of ERK activity can result in decreased activity of Akt through crosstalk. (F) Morphine-induced activation of Ras and its downstream pathway in striatum studied by Western blotting. Mice were injected with 20 mg/kg morphine (s.c.) or saline and their NAc regions were dissected 60 minutes after for active Ras pull down as well as Western blot analysis. (G) Quantification of Western blot data with activities normalized to vehicle-treated control of the same genotype. *p<0.05, **p<0.01, n=3 mice per each treatment, Student’s t-test. See also Figure S3.

Figure 5. Striatal NF1 regulates psychomotor responses to morphine

(A) The effect of morphine administration on locomotor activity in control (left) and Nf1 cKO (right) mice in the open field task. Mice received saline or increasing doses of morphine (2, 5, 10, 20 mg/kg, s.c.) and were immediately placed in the activity-recording chamber. (B) Quantification of cumulative distance traveled in open–field chamber during 20 to 80 min following vehicle or morphine treatment. NF1 ablation in striatum results in blunted morphine-induced psychomotor activation. *p<0.05, **p<0.01, Two-Way ANOVA post hoc Tukey’s test (n=8 mice per each group). See also Figure S4.

Figure 6. Disruption of MOR signaling via NF1 in the striatum decreases sensitivity to morphine reward

(A) Effect of striatal specific deletion of NF1 (Nf1 cKO mice) on mouse performance in conditioned place preference (CPP) task. Mice were administered either saline (0 mg/kg morphine) or various concentrations of morphine as indicated. Place preference scores are calculated as the time difference during post-conditioning between drug-paired side versus saline-paired side. * p < 0.05 in comparison between genotypes Two-Way ANOVA post hoc Tukey’s test (n=6–10 each group). (B) Effect of disrupting Gβγ-NF1 interaction in the Nucleus Accumbens (NAc) on morphine CPP. Upper, Schematics of the conditional dominant negative construct (AAV-NF1-DN^FLEX) and its viral delivery into NAc of Rgs9-Cre mice. Lower, Performance of mice in the CPP task. Mice were administered either saline or various concentrations of morphine as indicated. AAV- mediated expression of a dominant negative NF1 construct that prevents Gβγ-NF1 interaction decreases sensitivity of mice to rewarding effects of morphine. The same AAV-NF1-DN^FLEX virus was injected into either Rgs9-Cre(+) mice or their control littermates Rgs9-Cre(−). CPP experiments were conducted as in A. ** p < 0.01 in comparison between genotypes Two-Way ANOVA post hoc Tukey’s test (n=5–6 each group). (C) Animal performance in morphine self-administration task. Lever presses per session are plotted. Active (0.3mg/kg/infusion) vs. Inactive levers showed significant (p<0.01 and p<0.0001) difference with both genotypes and within criteria. Genotype vs. Session (n=5 per group) Interaction p= 0.005, Genotype p<0.0001, Session p<0.0001. Data expressed as mean ± SEM *p<0.05, **p<0.01, ***p<0.001 genotype comparison per session, two-way ANOVA followed by Bonferroni test. (D) Dose-response dependence of drug taking in the self-administration task. Genotype vs. Dose (n=5 per group) Interaction p<0.0001, Genotype p<0.0001, Dose p<0.0001. Data is expressed by mean ± SEM ***p<0.001 genotype comparison, ^###p<0.001 compared to Saline for both genotypes, two-way ANOVA with Bonferroni test.

Figure 7. Proposed model for the role of NF1 in mediating MOR signaling to Ras

NF1 is a novel Gβγ effector downstream from MORs activation and is essential for MORs-induced Ras pathway activation in MSNs.