Insulin Bidirectionally Alters NAc Glutamatergic Transmission: Interactions between Insulin Receptor Activation, Endogenous Opioids, and Glutamate Release

Abstract

Human fMRI studies show that insulin influences brain activity in regions that mediate reward and motivation, including the nucleus accumbens (NAc). Insulin receptors are expressed by NAc medium spiny neurons (MSNs), and studies of cultured cortical and hippocampal neurons suggest that insulin influences excitatory transmission via presynaptic and postsynaptic mechanisms. However, nothing is known about how insulin influences excitatory transmission in the NAc. Furthermore, insulin dysregulation accompanying obesity is linked to cognitive decline, depression, anxiety, and altered motivation that rely on NAc excitatory transmission. Using whole-cell patch-clamp and biochemical approaches, we determined how insulin affects NAc glutamatergic transmission in nonobese and obese male rats and the underlying mechanisms. We find that there are concentration-dependent, bidirectional effects of insulin on excitatory transmission, with insulin receptor activation increasing and IGF receptor activation decreasing NAc excitatory transmission. Increases in excitatory transmission were mediated by activation of postsynaptic insulin receptors located on MSNs. However, this effect was due to an increase in presynaptic glutamate release. This suggested feedback from MSNs to presynaptic terminals. In additional experiments, we found that insulin-induced increases in presynaptic glutamate release are mediated by opioid receptor-dependent disinhibition. Furthermore, obesity resulted in a loss of insulin receptor-mediated increases in excitatory transmission and a reduction in NAc insulin receptor surface expression, while preserving reductions in transmission mediated by IGF receptors. These results provide the first insights into how insulin influences excitatory transmission in the adult brain, and evidence for a previously unidentified form of opioid receptor-dependent disinhibition of NAc glutamatergic transmission.SIGNIFICANCE STATEMENT Data here provide the first insights into how insulin influences excitatory transmission in the adult brain, and identify previously unknown interactions between insulin receptor activation, opioids, and glutamatergic transmission. These data contribute to our fundamental understanding of insulin's influence on brain motivational systems and have implications for the use of insulin as a cognitive enhancer and for targeting of insulin receptors and IGF receptors to alter motivation.

Keywords: glutamate; insulin receptor; motivation; nucleus accumbens; opioid; striatal plasticity.

Figure 1.

Insulin receptor activation increases, whereas IGFR activation decreases, excitatory transmission onto MSNs in the NAc core. A, Average eEPSC amplitude during baseline (BL), bath application of insulin (black bar), and following washout. B, Summary of average maximum change from BL following insulin (1-500 nm). Effects of insulin on excitatory transmission are concentration-dependent and bidirectional. Right, Recording location within the NAc core. C, Average eEPSC amplitude in the presence of the IGFR antagonist PPP, before and after 30 nm insulin, with and without the membrane-permeable insulin receptor inhibitor HNMPA-(AM)3. D, Average eEPSC amplitude before and after insulin (30 and 100 nm) with the membrane-impermeable insulin receptor inhibitor HNMPA included in the patch pipette. E, Average eEPSC amplitude before and after 100 nm insulin administered in the presence of PPP. F, Average eEPSC amplitude in the presence of PPP, before and after 100 nm insulin followed by the addition of HNMPA-(AM)3 to the bath. Data in all figures are shown as the mean ± SEM. Statistical differences were determined by within-subject, two-way repeated-measures ANOVA comparing BL and treatment conditions. **p < 0.01, main effect of treatment (for full statistical information, see Results).

Figure 2.

Verification of single-cell RT-PCR method. A, Example of single-cell RT-PCR for a D2-MSN (left) and a D1-MSN (right) after whole-cell recordings in adult rat nucleus accumbens. β-actin was used as a positive control. B, Serial dilution of RNA from striatal tissue showing the sensitivity of pDYN primers (149 bp). C, Serial dilution of RNA from striatal tissue showing the sensitivity of pENK primers (220 bp). D, A second round of amplification is sufficient to allow for proenkephalin detection in samples containing 1 pg/µl of RNA.

Figure 3.

Insulin (30 nm) increases mEPSC frequency and the probability of glutamate release, but not mEPSC amplitude. A, Average mEPSC frequency before (BL) and after bath application of insulin (30 nm). B, Average mEPSC amplitude before and after insulin (30 nm). C, Cumulative probability distributions of mEPSC frequency before and after insulin (30 nm). D, Cumulative probability distributions and histograms of mEPSC amplitude before and after insulin (30 nm). E, Representative mEPSC traces before and after insulin (30 nm), with and without the membrane-impermeable insulin receptor inhibitor HNMPA included in the patch pipette. Top traces, arrows indicate regions in which the time scale was expanded in the lower traces. F, Average mEPSC frequency before and after bath application of insulin (30 nm) with membrane-impermeable HNMPA included in the patch pipette. G, Cumulative probability distributions of mEPSC frequency before and after insulin (30 nm) with HNMPA included in the patch pipette. H, Verification of paired-pulse facilitation in MSNs. Average PPR across increasing interpulse intervals (50-400 ms). Inset, Representative traces at a 50 ms interpulse interval. As expected, the probability of glutamate release is relatively low in NAc MSNs, and facilitation occurs at interpulse intervals ≤100 ms. I, Average PPR before and after insulin (30 nm). Inset, Representative traces before (black) and after insulin (gray) (50 ms interstimulus interval). J, Average mEPSC amplitude before and after insulin (30 nm) with membrane-impermeable HNMPA included in the patch pipette. K, Cumulative probability distributions and histograms of mEPSC amplitude before and after insulin (30 nm) with HNMPA included in the patch pipette. Statistical differences determined by two-tailed paired t-tests, *p < 0.05, ***p < 0.007.

Figure 4.

Insulin (100 nm) reduces mEPSC frequency and the probability of glutamate release without altering mEPSC amplitude. A, Average mEPSC frequency before (BL) and after insulin (100 nm). B, Average mEPSC amplitude before and after insulin (100 nm). C, Cumulative probability distributions of mEPSC frequency before and after insulin (100 nm). D, Cumulative probability distributions and histograms of mEPSC amplitude before and after insulin (100 nm). E, Representative mEPSC traces before and after insulin (100 nm). Left traces, arrows indicate regions in which the time scale was expanded in traces shown at the right. F, Average PPR before and after insulin (blue bar, 100 nm) and following insulin washout. Inset, representative traces before (black) and after insulin (blue). G, Average change in the PPR following insulin (100 nm). Statistical differences were determined by two-tailed paired t tests (A,G; **p < 0.01) and two-way repeated-measures ANOVA comparing BL and treatment conditions (F; *main effect of treatment, p < 0.01).

Figure 5.

Insulin-induced increases in glutamate release are GABA-B receptor- and opioid-receptor dependent. A, Average PPR at BL, and after insulin (30 nm) in the absence and presence of the GABA_B receptor antagonist phaclofen (20 μm). Phaclofen prevents insulin-induced decreases in PPR. B, Average mEPSC frequency before and after insulin (30 nm) in the presence of phaclofen (20 μm). Phaclofen prevents insulin-induced increases in mEPSC frequency. C, Proposed mechanism by which activation of insulin receptors on MSNs enhances glutamate release. We propose that activation of insulin receptors on MSNs results in the release of endogenous opioids (1) that reduces GABAergic transmission (2), thereby causing disinhibition of presynaptic glutamate release (3). D, Average eEPSC amplitude in the presence of the opioid receptor antagonist (–)-naloxone (1 μm) before and after bath application of insulin (30 nm). (–)-Naloxone prevents insulin-induced increases in eEPSC amplitude. E, Average PPR in the presence of (–)-naloxone (1 μm) before and after bath application of insulin (30 nm). (–)-Naloxone prevents insulin-induced decreases in PPR. F, Average PPR before and after bath application of (–)-naloxone (1 μm) alone confirms there is no effect of (–)-naloxone alone. G, Average PPR at BL, and after insulin (30 nm) in the absence or presence of (–)-naloxone (1 μm). (–)-Naloxone does not reverse insulin-induced decreases in PPR. H, Average PPR in the presence of (+)-naloxone (1 μm) before and after bath application of insulin (30 nm). (+)-Naloxone, which does not inhibit opioid receptors, does prevent insulin-induced decreases in PPR. Example traces are shown within each panel. Statistical differences were determined by two-way repeated-measures ANOVA comparing BL and treatment conditions (D), two-tailed paired t tests (C,E,F,H), and one-way ANOVA followed by Sidak's multiple comparisons post-test (B,G). *p < 0.05.

Figure 6.

High-fat diet-induced obesity results in a loss of insulin-induced increases in excitatory transmission and a reduction in NAc IRβ surface expression. Average concentration of fasted plasma insulin (A) and fat mass (B) in chow and high-fat diet groups. C, Average eEPSC amplitude following bath application of increasing concentrations of insulin (blue bars) and following insulin washout in MSNs from chow (circles) and high-fat groups (squares). Right, Representative traces for each group before (black) and after (blue) each insulin concentration. D, Average NAc IRβ surface expression in high-fat and chow-fed groups. Left, immunoblot for TH in the bound (B) and unbound (UB) fractions. Consistent with its intracellular localization, TH protein levels were nearly absent in the bound (surface) fraction. E, Total NAc IRβ expression in high-fat and chow-fed groups. Representative blot images are shown below each graph. Statistical differences were determined by two-tailed unpaired t tests (A,B: **p < 0.001), two-way repeated-measures ANOVA comparing BL and treatment conditions (C: *chow group, main effect of treatment; ^#high-fat group, main effect of treatment, p < 0.05), and two-tailed unpaired t test (D: p = 0.046).