Am I full: How does dopamine track the value of what we consume?

In Chapter 9 we learned about innate behaviors and the evolutionarily ancient neural circuits that drive their performance. Across the animal kingdom, many behaviors are associated with homeostatic drive. A cold mouse will nest to keep warm while an owl will hunt that same mouse to quell its hunger. Behaviors that maintain homeostasis are necessary for survival and so evolution has hardwired circuits to execute them. Among such innate homeostatic behaviors are eating and drinking.

All animals must maintain adequate fluid and caloric intake to live. This process abides by a sequence of physiological steps. Imagine you skipped lunch. The caloric deficiency sensed in the periphery is relayed to the brain upon which a brain state is generated that is associated with the subjective feeling of hunger. This then culminates in the innate behavior to seek and consume food. Our brain then receives feedback based on the caloric content of what we consumed. For example, although having a large bowl of spinach might fill your stomach, it would not sate you the way a far more calorie dense meal such as a cheeseburger would. Thus, the differential sating value of ingested foods influences our decision on what food to seek. This evinces a critical question: How is information about the hydrating or sating value of ingested water/food evaluated? Recent work by Grove et al. (2022) probed this question.   

The authors identified some dopamine (DA) neurons in the midbrain Ventral Tegmental Area (VTA) in thirsty mice whose activity was associated with drinking water. Interestingly, a large subset of these VTA-DA neurons would increase their rate of firing about 10 minutes after drinking or when water was infused intragastrically. Indeed, the authors demonstrated that the activity of these neurons tracked the osmotic state of the animal. How might caloric information converge onto similar circuits? The authors approached this question by infusing liquid food and water intragastrically and measuring the neural responses of VTA-DA neurons. Remarkably, they found that VTA-DA neurons encoding food versus water are largely non-overlapping.

After characterizing the input and output of VTA-DA neurons that represent systemic hydration, the authors wanted to determine whether the activity of these neurons are associated with assigning the learned hydrating properties of a fluid. (Many animals must learn through ingestive experience the hydration values of specific fluids.) To test this, they devised a behavioral assay where two fluids with different hydrating value (water vs. a salty solution) were infused intragastrically to a thirsty mouse and were associated with two flavors of a separate reward (Figure; left panel). Over time, the mouse would learn to prefer the water that was more hydrating (Figure; right panel). In this way, the authors can assay how the value of a reward is determined and learned. Finally, the authors wanted to test whether VTA-DA neural activity was necessary for this learning. When they inhibited VTA-DA neurons during the learning period (training) they found that mice no longer developed a preference for the more hydrating solution (Figure, right panel), suggesting that these neurons are crucial for post-ingestive evaluation of hydrating value.

Collectively, these results identify populations of dopamine neurons whose activity tracks the sating value of ingested food or water. Furthermore, they demonstrate that the activity of hydration-representing dopamine neurons is essential for post-ingestive learning.

Figure: (left) Schematic of experimental paradigm to assay post-ingestive learning. The paradigm couples the flavor of a reward to the IG infusion of a salty solution or water. (right) Control mice (expressing mCherry in VTA-DA neurons) learn to prefer the flavor associated with water over the flavor associated with salty solution. The VTA-DA neurons of experimental mice express GtACR, a genetically encoded light sensitive channel that hyperpolarizes and therefore inhibits activity. In these mice, when VTA-DA activity is inhibited, they no longer learn to prefer the flavor associated with water. (Adapted from Grove et al., 2022)

Reference

Grove, J. C. R., Gray, L. A., la Santa Medina, N., Sivakumar, N., Ahn, J. S., Corpuz, T. v., Berke, J. D., Kreitzer, A. C., & Knight, Z. A. (2022). Dopamine subsystems that track internal states. Nature, 608(7922), 374–380. Link