Instructor Notes — Chapter 22

Neuropeptide Y, Substance P, and the Peptides That Regulate Stress, Pain, and Appetite in the Brain Part 4 · advanced · prerequisites: Chapters 2, 20, 21

What this chapter is for

Chapter 22 is the book's stress test for its own central discipline. Everywhere else, "mechanism is not evidence" is applied to marketing — to people who have every incentive to overstate and no obligation to be rigorous. Here it is applied to the pharmaceutical industry at its most careful, and it still holds.

If your students take one thing from Part IV, it should be this: the NK1 program had a real mechanism, strong preclinical data, good compounds, and confirmed target engagement in humans — and the clinical hypothesis was wrong anyway. Everything else in the chapter exists to give that fact company, contrast, and consequences.

Plan roughly seven to nine hours of student time. In a course, this is comfortably two sessions: one on §§22.1–22.5 (the failure) and one on §§22.6–22.9 plus the case studies (the contrast and the payoff). Do not compress it to one. The comparison in §22.9 does not land unless students have sat with the NK1 failure long enough to feel it.

Sequencing

Session 1 — the mechanism and the failure (§§22.1–22.5). Open with §22.1's co-transmission idea, because it does real work later: substance P's selectivity for sustained noxious input — the property that made the analgesic hypothesis so attractive — is a direct consequence of dense-core vesicle release requiring high-frequency firing. Students who skip §22.1 read §22.4 as a list of facts instead of an argument.

Then run NPY (§22.2–22.3) as a warm-up failure. It is lower-stakes than NK1 and has the same shape: correct mechanism, engaged target, no drug. Students who see the pattern once in a low-stakes case recognize it faster in the high-stakes one.

Give §22.5 the time it needs. The single most important beat is the PET occupancy data. Until students internalize that the drugs reached and occupied the receptor, they will keep reaching for "underdosed" or "didn't get into the brain," which are the comfortable explanations and are both ruled out.

Session 2 — the contrast and the payoff (§§22.6–22.9 + case studies). Orexin (§22.6) is the pivot. It looks like a success story and is really an asymmetry story: the antagonist worked, the replacement did not, and the difference is delivery. That sets up §22.8.

Then CGRP (§22.7) and the barrier (§22.8) and the comparison (§22.9). §22.9 is the chapter and should get at least a third of the session.

The three places students reliably go wrong

1. "The NK1 failure means substance P isn't really a pain peptide." This is the most common misreading and it is worth heading off explicitly before students encounter §22.5. The chapter's exercise L (about NPY) is a deliberate rehearsal for exactly this error at lower stakes; assign it before the NK1 material if you can. The correction: a molecule's physiological role and the therapeutic utility of blocking it are separate questions, and evidence about the second does not overturn evidence about the first.

2. "Mechanism is useless, then." The overcorrection. §22.9 refuses it in writing and you should refuse it out loud. CGRP's success came directly out of decades of mechanistic work. The lesson is about what mechanism supports — a hypothesis worth testing — not about whether mechanism is worth doing.

Watch for the student who arrives at cynicism and mistakes it for rigor. It is a comfortable position and it is not the position this book teaches.

3. "CGRP drugs are peptide drugs." Almost every student will say this at least once, because almost every popular article says it. The -mab suffix from Chapter 1 §1.8 is the fastest correction available and it is worth making students do it themselves: put the eight drug names on a board and ask them to sort. Nobody who sorts them forgets afterward.

Assessment notes

The quiz is 22 items and can be given closed-book; several items have distractors that are true but irrelevant (item 3's option D is the clearest), which reliably separates students who read carefully.

Of the exercises, six are worth prioritizing if you cannot assign all thirty-four:

• T (rewrite the 1995 grant application, then annotate each sentence) — the single best item in the set. Students discover the mechanism/inference ratio themselves rather than being told it.
• W (what PET occupancy eliminates and what it leaves) — tests precision about negative results.
• Y (strong ❌ versus weak ❌) — tests the epistemics.
• CC (narcolepsy/type 1 diabetes parallel, and where it breaks) — tests cross-chapter integration.
• NN (why the divergent fates are surprising, then resolve it) — tests §22.9 directly.
• OO (run the four-item checklist on a real current claim) — the transfer item. Expect wide variance and mark on honesty about the "could not determine" category, not on the conclusion.

Both case studies carry six discussion questions each and are designed for seminar use rather than written submission, though 22.1 Q1 and 22.2 Q3 make good short essays.

Materials and preparation

No laboratory, computation, or software is required. Nothing in this chapter involves protocols, dosing, or product evaluation, and student questions that head that direction should be redirected to Chapter 39.

If you can obtain two papers for the class, get the 1998 positive NK1 depression trial and the 2006 negative replication (see the chapter's Further Reading). Reading them back to back is the most efficient hour available in this material, and the "If you only do one thing" recommendation is built around it.

Sensitivity note: some students will have personal experience with migraine, narcolepsy, chronic pain, or chemotherapy. The chapter treats all four clinically and without sentimentality, which is the right register, but be aware that the discussion of chronic pain trial heterogeneity in §22.9 can read as dismissive to someone living with the condition. It is not — the argument is about measurement, not about legitimacy — and saying so once, plainly, is usually enough.