Limits...
Hunger neurons drive feeding through a sustained, positive reinforcement signal

View Article: PubMed Central - PubMed

ABSTRACT

The neural mechanisms underlying hunger are poorly understood. AgRP neurons are activated by energy deficit and promote voracious food consumption, suggesting these cells may supply the fundamental hunger drive that motivates feeding. However recent in vivo recording experiments revealed that AgRP neurons are inhibited within seconds by the sensory detection of food, raising the question of how these cells can promote feeding at all. Here we resolve this paradox by showing that brief optogenetic stimulation of AgRP neurons before food availability promotes intense appetitive and consummatory behaviors that persist for tens of minutes in the absence of continued AgRP neuron activation. We show that these sustained behavioral responses are mediated by a long-lasting potentiation of the rewarding properties of food and that AgRP neuron activity is positively reinforcing. These findings reveal that hunger neurons drive feeding by transmitting a positive valence signal that triggers a stable transition between behavioral states.

Doi:: http://dx.doi.org/10.7554/eLife.18640.001

No MeSH data available.


Related in: MedlinePlus

Prestimulation of AgRP neurons conditions appetite and flavor preference.(A) Schematic of conditioned appetite assay. Test pellets and home cage chow are similar in energy density but different in shape, size, and texture. Test pellets were either included in home cage or not, as indicated. (B–C) Average 60 min food intake of conditioned appetitive experiments. (B) Food intake of AgRP-ChR2 mice without access to test pellets in homecage (red n = 6) and with access to test pellets in homecage (black n = 5), and WT mice without access to test pellets in homecage (grey n = 3) through consecutive trials. Blue boxes indicate trials with 60 min prestimulation (trials 3,5,7), whereas in white trials animals were subjected to mock stimulation (trials 1,2,4,6,8). (C) Comparison among pre-conditioning, post-conditioning and extinction trials of AgRP-ChR2 mice with (black n = 5) or without (red n = 6) access to test pellets in homecage. Trial 1 and 2 are considered pre-conditioning, trial 4 and 6 are considered post-conditioning and trial 8 is considered extinction. (D) Conditioned flavor preference experiment. (E) Change of preference to conditioned flavor before and after 4 repeats of prestimulation conditioning assay (n = 8). Asterisks on top of brackets indicate significance levels for comparisons with the respective protocols, using one-way-ANOVA with Holm-Sidak’s correction for multiple comparisons (****p≤0.0001, ***0.0001<p≤0.001, **0.001<p≤0.01, *0.01<p≤0.05, ns p>0.05).DOI:http://dx.doi.org/10.7554/eLife.18640.008
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC5016090&req=5

fig5: Prestimulation of AgRP neurons conditions appetite and flavor preference.(A) Schematic of conditioned appetite assay. Test pellets and home cage chow are similar in energy density but different in shape, size, and texture. Test pellets were either included in home cage or not, as indicated. (B–C) Average 60 min food intake of conditioned appetitive experiments. (B) Food intake of AgRP-ChR2 mice without access to test pellets in homecage (red n = 6) and with access to test pellets in homecage (black n = 5), and WT mice without access to test pellets in homecage (grey n = 3) through consecutive trials. Blue boxes indicate trials with 60 min prestimulation (trials 3,5,7), whereas in white trials animals were subjected to mock stimulation (trials 1,2,4,6,8). (C) Comparison among pre-conditioning, post-conditioning and extinction trials of AgRP-ChR2 mice with (black n = 5) or without (red n = 6) access to test pellets in homecage. Trial 1 and 2 are considered pre-conditioning, trial 4 and 6 are considered post-conditioning and trial 8 is considered extinction. (D) Conditioned flavor preference experiment. (E) Change of preference to conditioned flavor before and after 4 repeats of prestimulation conditioning assay (n = 8). Asterisks on top of brackets indicate significance levels for comparisons with the respective protocols, using one-way-ANOVA with Holm-Sidak’s correction for multiple comparisons (****p≤0.0001, ***0.0001<p≤0.001, **0.001<p≤0.01, *0.01<p≤0.05, ns p>0.05).DOI:http://dx.doi.org/10.7554/eLife.18640.008

Mentions: To test this, we investigated whether AgRP neuron prestimulation could condition appetite for specific foods. AgRP-ChR2 mice were acclimated to a feeding chamber that delivered pellets that had a similar energy density to their home cage chow, but had a distinct size, shape, and texture (Figure 5A; see Methods for additional information). We then tested mice in this chamber for pellet consumption during a one hour test period on eight consecutive days (Figure 5B). The trial was designed so that, on days 3, 5, and 7, the test period was immediately preceded by one hour of AgRP neuron prestimulation (Figure 5B, blue), whereas on the intervening days (days 1, 2, 4, 6, and 8), there was mock stimulation. Of note, mice had ad libitum access to chow in their home cage, and all animals were laser naïve at the beginning of the trial, meaning that day 3 was the first time they were exposed to photostimulation.10.7554/eLife.18640.008Figure 5.Prestimulation of AgRP neurons conditions appetite and flavor preference.


Hunger neurons drive feeding through a sustained, positive reinforcement signal
Prestimulation of AgRP neurons conditions appetite and flavor preference.(A) Schematic of conditioned appetite assay. Test pellets and home cage chow are similar in energy density but different in shape, size, and texture. Test pellets were either included in home cage or not, as indicated. (B–C) Average 60 min food intake of conditioned appetitive experiments. (B) Food intake of AgRP-ChR2 mice without access to test pellets in homecage (red n = 6) and with access to test pellets in homecage (black n = 5), and WT mice without access to test pellets in homecage (grey n = 3) through consecutive trials. Blue boxes indicate trials with 60 min prestimulation (trials 3,5,7), whereas in white trials animals were subjected to mock stimulation (trials 1,2,4,6,8). (C) Comparison among pre-conditioning, post-conditioning and extinction trials of AgRP-ChR2 mice with (black n = 5) or without (red n = 6) access to test pellets in homecage. Trial 1 and 2 are considered pre-conditioning, trial 4 and 6 are considered post-conditioning and trial 8 is considered extinction. (D) Conditioned flavor preference experiment. (E) Change of preference to conditioned flavor before and after 4 repeats of prestimulation conditioning assay (n = 8). Asterisks on top of brackets indicate significance levels for comparisons with the respective protocols, using one-way-ANOVA with Holm-Sidak’s correction for multiple comparisons (****p≤0.0001, ***0.0001<p≤0.001, **0.001<p≤0.01, *0.01<p≤0.05, ns p>0.05).DOI:http://dx.doi.org/10.7554/eLife.18640.008
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC5016090&req=5

fig5: Prestimulation of AgRP neurons conditions appetite and flavor preference.(A) Schematic of conditioned appetite assay. Test pellets and home cage chow are similar in energy density but different in shape, size, and texture. Test pellets were either included in home cage or not, as indicated. (B–C) Average 60 min food intake of conditioned appetitive experiments. (B) Food intake of AgRP-ChR2 mice without access to test pellets in homecage (red n = 6) and with access to test pellets in homecage (black n = 5), and WT mice without access to test pellets in homecage (grey n = 3) through consecutive trials. Blue boxes indicate trials with 60 min prestimulation (trials 3,5,7), whereas in white trials animals were subjected to mock stimulation (trials 1,2,4,6,8). (C) Comparison among pre-conditioning, post-conditioning and extinction trials of AgRP-ChR2 mice with (black n = 5) or without (red n = 6) access to test pellets in homecage. Trial 1 and 2 are considered pre-conditioning, trial 4 and 6 are considered post-conditioning and trial 8 is considered extinction. (D) Conditioned flavor preference experiment. (E) Change of preference to conditioned flavor before and after 4 repeats of prestimulation conditioning assay (n = 8). Asterisks on top of brackets indicate significance levels for comparisons with the respective protocols, using one-way-ANOVA with Holm-Sidak’s correction for multiple comparisons (****p≤0.0001, ***0.0001<p≤0.001, **0.001<p≤0.01, *0.01<p≤0.05, ns p>0.05).DOI:http://dx.doi.org/10.7554/eLife.18640.008
Mentions: To test this, we investigated whether AgRP neuron prestimulation could condition appetite for specific foods. AgRP-ChR2 mice were acclimated to a feeding chamber that delivered pellets that had a similar energy density to their home cage chow, but had a distinct size, shape, and texture (Figure 5A; see Methods for additional information). We then tested mice in this chamber for pellet consumption during a one hour test period on eight consecutive days (Figure 5B). The trial was designed so that, on days 3, 5, and 7, the test period was immediately preceded by one hour of AgRP neuron prestimulation (Figure 5B, blue), whereas on the intervening days (days 1, 2, 4, 6, and 8), there was mock stimulation. Of note, mice had ad libitum access to chow in their home cage, and all animals were laser naïve at the beginning of the trial, meaning that day 3 was the first time they were exposed to photostimulation.10.7554/eLife.18640.008Figure 5.Prestimulation of AgRP neurons conditions appetite and flavor preference.

View Article: PubMed Central - PubMed

ABSTRACT

The neural mechanisms underlying hunger are poorly understood. AgRP neurons are activated by energy deficit and promote voracious food consumption, suggesting these cells may supply the fundamental hunger drive that motivates feeding. However recent in vivo recording experiments revealed that AgRP neurons are inhibited within seconds by the sensory detection of food, raising the question of how these cells can promote feeding at all. Here we resolve this paradox by showing that brief optogenetic stimulation of AgRP neurons before food availability promotes intense appetitive and consummatory behaviors that persist for tens of minutes in the absence of continued AgRP neuron activation. We show that these sustained behavioral responses are mediated by a long-lasting potentiation of the rewarding properties of food and that AgRP neuron activity is positively reinforcing. These findings reveal that hunger neurons drive feeding by transmitting a positive valence signal that triggers a stable transition between behavioral states.

Doi:: http://dx.doi.org/10.7554/eLife.18640.001

No MeSH data available.


Related in: MedlinePlus