Thrilled to share my PhD work showing dopamine and serotonin form a gas-brake system for reward has won Stanford's Sammy Kuo Paper of the Year Award!

I'm at #sfn25 presenting new work that uncovers a key circuit mechanism underlying this opponency. Stop by poster AA3 Monday morning to learn more!

We think this provides one putative mechanism to explain longstanding observations of serotonin counteracting dopamine's rewarding effects.

Our work also provides a characterization of the mesolimbic serotonin system that we hope will be resource for the field.

7/n

Finally we tested whether this also happens in vivo. We used DAT (bup) and SERT (esc) blockers to upregulate dopamine or serotonin and labeled cfos as a proxy for neural activity.

‼️ Enhancing dopamine activity preferentially activated D1-MSNs while enhancing serotonin activated D2-MSNs

6/n

This suggested that serotonin may differentially modulate D1 and D2 MSN activity in a manner opposite to their regulation by dopamine. To test this, we recorded from genetically identified D1 and D2-MSNs.

⬆️⬇️ Dopamine selectively excited D1-MSNs while serotonin selectively activated D2-MSNs

5/n

Next we used RNAscope to map the their distribution.

📊 expression varied across subregions and cell-types for every gene examined

❗️ Overall though, D1-MSNs expressed a mix of Gi, Gs, and Gq coupled serotonin receptors while D2-MSNs were enriched for receptors in the Gq coupled 5HT2 family

4/n

To answer this question, we first determined which serotonin receptors are expressed on striatal MSNs.

🎯 Mining a scRNA-Seq dataset revealed that >90% of MSNs express serotonin receptors, and that this is primarily limited to a subset of ~8 receptor genes

3/n

📚 It's known that dopamine drives reinforcement via differential modulation of D1 and D2 MSNs, and last year we showed serotonin counteracts dopamine's reinforcing effects.

This inspired us to ask:

❓How does serotonin affect striatal neuron activity?

2/n

Finally we asked whether these effects generalize beyond the NAc pmSh. Repeating key experiments in a canonical reward center (NAc Core) we found

✅ inverse reward responses
✅ dopamine stimulation + serotonin inhibition was more reinforcing than dopamine stim alone‼️

8/n

This led us to hypothesize that artificially recreating dopamine & serotonin reward responses together would drive learning more potently than either manipulation alone

🐭 That's exactly what we found across multiple assays for contextual, Pavlovian, & instrumental learning

7/n

To test whether these signals are necessary for learning, we used optogenetics to block dopamine and/or serotonin reward responses

🤯 mice that lost both reward responses totally failed to learn a cue-reward association, bc they perceived the reward to be less reinforcing

6/n

So what does the activity of these inputs look like while mice perform a simple reward learning task? Across multiple experimental approaches, we consistently found the same thing:

❗️When mice consumed a reward, dopamine ⬆️ serotonin ⬇️

5/n

We could then look for overlap between red & green axons in other brain regions to ask, where do dopamine & serotonin signals converge?

🔥 We identified a subregion of the striatum (NAc posterior medial shell) as a hotspot for converging dopamine & serotonin inputs

4/n

To tackle this question, we developed genetic strategies to access dopamine & serotonin systems simultaneous for the first time.

In one example, we labeled dopamine & serotonin neurons in the same mouse brain w/ different fluorescent proteins

Dopamine 🔴
Serotonin 🟢

3/n

My paper is out!! 🎉

"Opponent control of reinforcement by striatal dopamine and serotonin", @Nature

Here, we show that dopamine and serotonin signals form a gas-brake system for reward in the mammalian brain

THREAD ⬇️

1/n