MED02 Research Brief — The Endocannabinoid System and Cannabinoid Pharmacology: What Is Established, Emerging, and Overstated?

DRAFT v0.1 — NOT FROZEN

Drafted 2026-10-15 by Claude at Danny’s request, from the claim list the Corpus Cannabis Body Lab needs verified (ROADMAP §3 + §4; §11 candidate “Endocannabinoid System: What is established, emerging, and overstated?”).

Before freeze:

• ChatGPT and Gemini editorial passes on this brief.
• The §M12 archive overlap check.
• Danny’s decisions in §10.

Researchers receive the frozen version only, as a git .md attachment.

Governing protocol: PROTOCOL-MEDICAL v0.5 BASELINE (git 7beaff6) · base PROTOCOL v1.4

Independence rule: ChatGPT and Gemini receive this brief with substantively identical instructions. Neither sees the other’s pass until both are filed (§M1 firewall).

Not medical advice. This round describes a research literature about how the body’s cannabinoid system works and how cannabinoids are handled by the body. It does not evaluate any treatment for any reader. Doses appear only as what a study administered.

0. How MED02 is organized

• Structure (proposed; Danny decides, §10).

  • Two parts, run in sequence. Each part has its own two independent passes, firewall, verification and cross-audit, and reaches PART LOCK before the next part’s passes start.
  • One MED02-wide evidence search cutoff, set at the first freeze.
  • One shared ledger (MED02-C001 onward) and one final synthesis. Each part also gets a short interim synthesis.

• A third lane is not sent to the researchers. Facts about deposited receptor structures (database records and what each structure paper reports about its own structure) are checked by the verifier directly against the record. They are ledgered with origin verifier-direct.

• Already ledgered, do not re-research. Treat these as given context:

  • MED01-C007–C011: discovery of the binding site, CB1, CB2, anandamide and 2-AG.
  • C044: retrograde messenger role; FAAH-knockout mice.
  • C045: CBD as a CB1 negative allosteric modulator in cell systems.
  • C046: “clinical endocannabinoid deficiency” as a hypothesis.
  • C047: entourage/synergy as a hypothesis.
  • C064: adverse outcomes of ECS-targeting drugs.
  • C072: cannabis withdrawal time course.

  You may cite later evidence that changes any of these, and must flag it as such.

1. Research question

P1: What does current evidence establish about the human endocannabinoid system? Where are its receptors, how are its signalling molecules made, used and removed, and which physiological roles are demonstrated, in which species and by which methods? Which common descriptions of the system go beyond that evidence?

P2: What does current evidence establish about how THC, CBD and other named cannabinoids act at cannabinoid and other receptors, at which concentrations? How are THC and CBD absorbed, distributed, metabolized and eliminated by route of administration, and how do those concentrations relate to the concentrations used in laboratory studies?

2. Scope

In scope:

• Human evidence first; animal, tissue and cell evidence where human evidence is absent, always labelled.
• Receptor distribution and function.
• Endocannabinoid biosynthesis, degradation and signalling.
• Receptor pharmacology of named cannabinoids.
• Human pharmacokinetics and metabolism by route.
• Receptor adaptation (tolerance) in humans.
• Cannabinoid–drug interactions at the level of metabolism.
• Terpenes only where a direct, tested action at a cannabinoid receptor is reported.

Out of scope, with handoffs:

• Whether any cannabinoid treats any condition: Phase B indication rounds.
• General terpene pharmacology and “entourage” effects: ROADMAP §5 round; C047 stands.
• Plant chemistry and cultivation: ROADMAP §6.
• Long-term health outcomes of use: ROADMAP §7.
• Regulatory status.
• The MED01-ledgered items listed in §0.

Scope guard. This is a map of established mechanism and pharmacokinetics, not a textbook. For each question, characterize the evidence body (which methods, species and designs, and how consistent they are). Fully extract only the studies your answer materially relies on.

3. Key questions

Part 1 — Endocannabinoid system physiology

A. CB1 in the human nervous system

• A1. Where is CB1 densest and sparsest in the adult human brain and spinal cord? What methods established this (post-mortem receptor binding, mRNA, protein staining, PET), and how well do the methods agree?
• A2. What is known about CB1 in the brainstem areas that control breathing and heart rate? What evidence connects that distribution to the acute safety profile of cannabinoids, and is that connection tested or inferred?
• A3. How does CB1 density or distribution change across development and age in humans?
• A4. Which brain functions have human evidence (not only animal evidence) of involving CB1 signalling?

B. CB2

• B1. Which human tissues and blood-cell types express CB2, at mRNA and at protein level, and how do levels compare with CB1?
• B2. What is the current evidence for CB2 in the central nervous system, on microglia and on neurons? Which methodological problems (for example antibody specificity) affect that evidence?
• B3. What is known in humans about whether selective CB2 activation produces CB1-like psychoactive effects?

C. Peripheral organs

• C1. Which peripheral organs and tissues express CB1 and/or CB2, by which method and in which species? Include the lung, heart, adrenal gland, reproductive organs, skeletal muscle, pancreas, kidney, skin, bone and eye.
• C2. What is established about cannabinoid receptors in the gastrointestinal tract and its nerve network?
• C3. What roles of the endocannabinoid system in appetite, liver and fat metabolism are demonstrated, and in which species? How does receptor expression change in conditions such as fatty liver?
• C4. What is established about cannabinoid receptors in the heart and blood vessels, and about the receptor dependence of cannabinoids’ acute cardiovascular effects in humans?

D. Making, signalling and removing endocannabinoids

• D1. How are 2-AG and anandamide made? Which enzymes and pathways are established, which are proposed, and what do knockout studies (which species) show?
• D2. Which enzymes break down 2-AG and anandamide? Where are they located in the cell and at the synapse? What changes when each is inhibited, including effects on other related lipids?
• D3. What forms of retrograde endocannabinoid signalling are described, over what timescales, at which synapse types? Where are receptors located relative to the synapse, and how much of this is shown in human tissue?
• D4. How strongly do 2-AG and anandamide activate CB1 and CB2 (efficacy), and which other targets do they act on?
• D5. What levels of endocannabinoids are measured in human blood, cerebrospinal fluid and tissue, and how reliable are those measurements?

E. How the system is described

• E1. Which precise physiological statements about the endocannabinoid system’s role in maintaining balance (“homeostasis”) are supported? Which common generalizations about it are broader than the evidence?

Part 2 — Cannabinoid pharmacology and pharmacokinetics

F. THC pharmacokinetics by route

• F1. For smoked, vaporized, oral (including the effect of food and formulation) and oromucosal/sublingual THC: what bioavailability, time to peak and peak concentrations are reported, in which populations, at which doses? What drives the variability?
• F2. How is THC metabolized? Which enzymes are involved, and how do the amount and timing of 11-OH-THC (and other metabolites) differ by route? What are the known sources of variation between people, including genetic ones?
• F3. How are THC and its metabolites distributed, stored and eliminated? What half-lives are reported, by which methods, and do they differ between occasional and frequent users?
• F4. What is known about conversion of acidic cannabinoids to THC/CBD by heat, and about losses during combustion and vaporization?
• F5. How does the time course of measurable effects (subjective, cardiovascular, performance) relate to blood concentrations, by route?
• F6. For oromucosal/sublingual cannabinoid products: what evidence shows how much is absorbed through the mouth versus swallowed, and what are their pharmacokinetics?

G. Receptor pharmacology and adaptation

• G1. How tightly does THC bind CB1 and CB2, and how strongly does it activate them, relative to the endocannabinoids? How does its action depend on receptor density and on the body’s own endocannabinoid activity? What do human receptor-blockade studies show about which of its effects run through CB1?
• G2. Which non-cannabinoid-receptor targets does THC act on, at what concentrations?
• G3. Which mechanisms are proposed for CBD at CB1, CB2 and other targets, at which concentrations, in which systems? What does CBD do to drug-metabolizing enzymes, and what clinical drug-interaction evidence exists?
• G4. For THCV, CBN, CBG, CBC, CBDV, THCA and CBDA: which receptor actions are reported, at what concentrations, in which systems? How many phytocannabinoids have any pharmacological characterization?
• G5. Which terpenes found in cannabis have a directly tested action at a cannabinoid receptor, and what evidence supports it?
• G6. How do synthetic cannabinoid receptor agonists compare with THC in potency and efficacy, and what harms evidence is linked to those differences?
• G7. What do human studies (including imaging) show about cannabinoid receptor adaptation with repeated use, and about its time course of recovery?
• G8. What blood and brain concentrations of THC and CBD are reached in humans by route and dose (including medical oral CBD doses)? How do these compare with the concentrations at which laboratory effects in G1–G4 were observed?

4. Evidence standards (PROTOCOL-MEDICAL §M3–M8, binding)

1. Every source gets a source type and a study design (§M3): primary | sr-ma | narrative-review | guideline | official × rct | interventional-nr | observational | case | animal | in-vitro | mechanistic. For a review, state the designs it rests on.

2. Name the evidence system for every mechanistic or expression finding. Species · tissue or cell type (native or transfected cell line, named) · method:

   • receptor binding / autoradiography
   • mRNA
   • protein (antibody named, specificity controls stated if reported)
   • PET tracer
   • functional assay, named (e.g. GTPγS, cAMP, β-arrestin, electrophysiology)

3. Concentrations travel with mechanism claims. Give the potency/affinity measure exactly as reported, with units and assay: Ki, Kd, EC50, IC50, KB. A concentration-dependent effect is never reported without its concentration. Do not compare a laboratory concentration to human exposure unless a human pharmacokinetic source supplies the exposure (G8).

4. Pharmacokinetic fields. For every PK study you rely on:

   • population: n, occasional vs frequent users, sex where reported
   • product or formulation, named; dose; route
   • fed/fasted; sampling schedule
   • analytic method and limit of quantification
   • parameters exactly as reported (with variability measures)
   • registration if a trial

5. No drift (§M7, D1–D14).

   • Animal ≠ human; cell ≠ animal.
   • mRNA ≠ protein ≠ functional receptor.
   • “Implicated in” ≠ “demonstrated to regulate”.
   • Binding ≠ activation.
   • A concentration seen in a dish is not a concentration reached in a person.
   • One product or route ≠ “cannabis”.

6. Established vs emerging vs contested. For every key-question answer, say which parts are:

   • established: replicated across methods or labs
   • emerging: few studies, or one method
   • contested: conflicting findings, or known method problems

   Say what each label rests on.

7. Superseded findings. Where later work revised an earlier finding (a receptor once thought absent from a tissue, or a half-life revised by better methods), report both with dates and say what changed.

8. Harms get the same rigor where the question touches them (C4, G6, G7, G3 interactions).

9. Publication integrity. Check and flag retractions, corrections and expressions of concern.

10. Identifiers must resolve, never from memory. Open every DOI/PMID live and copy the resolved title. Never reconstruct an identifier.

11. Quote by copy, not recall.

12. No advice. No efficacy verdicts. No statement about what a cannabinoid does for a patient.

4a. Search requirements (§M12b)

Minimum sources:

• PubMed/MEDLINE
• Cochrane Library
• ClinicalTrials.gov and WHO ICTRP (pharmacokinetic and receptor-occupancy trials)
• IUPHAR/BPS Guide to Pharmacology: cannabinoid receptor and ligand entries, with their cited primary sources
• Clinical-pharmacology sections of official labels or product monographs for approved cannabinoid medicines (dronabinol, nabilone, nabiximols, purified CBD), from the issuing regulator
• At least one recent systematic or narrative review per key-question block, if any exists

Core search concepts:

• endocannabinoid system
• CB1 / CNR1, CB2 / CNR2: distribution, expression, autoradiography, PET
• anandamide, 2-arachidonoylglycerol (2-AG)
• biosynthesis enzymes: DAGL, NAPE-PLD
• degradation enzymes: FAAH, MAGL, ABHD6, ABHD12
• retrograde signalling, DSI, DSE
• named cannabinoids: Δ9-THC, 11-OH-THC, CBD, THCV, CBN, CBG, CBC, CBDV, THCA, CBDA
• pharmacokinetics, bioavailability, route: inhalation, vaporization, oral, oromucosal, sublingual
• metabolism (CYP2C9, CYP3A4, CYP2C19), drug interaction
• tolerance, receptor downregulation, imaging
• synthetic cannabinoid receptor agonists
• terpene AND cannabinoid receptor

Search log (required at the end of the pass):

• tools and databases used
• exact queries (ESearch URLs for PubMed)
• dates run
• counts
• what you could not access
• anything you relied on from training knowledge rather than a live search

This is not a systematic review and must not be described as one.

5. Required output format (per part)

1. Narrative answer to the part’s key questions, in order, with inline citations and the §4 rule 6 labels.
2. Candidate claims CAND-01…, normally 20–30, hard maximum 35. Each carries:
   • a one-sentence claim
   • claim type: distribution | mechanism | pharmacokinetic | pharmacology | adaptation | harms-research | synthesis-conclusion | absence | correction-of-common-claim
   • supporting sources with identifiers and resolved titles
   • source type × study design
   • evidence system (§4 rule 2)
   • concentrations or PK fields where relevant
   • established / emerging / contested
   • your confidence
   • what would change your mind
3. Reference list:
   • identifiers, with verified live (resolved title copied) or from memory — unverified
   • access level (full text | abstract only | secondary report)
4. Self-flagged weak points:
   • uncertain citations
   • single-source claims
   • abstract-only reads
   • concentration or species mismatches you noticed
5. Search log (§4a).
6. State at the top which brief version you worked from.

6. Archive probes

To be added after the §M12 overlap check, if needed. Probes go in a separate appendix delivered after the Stage 1 pass is filed, as neutrally worded questions only, in the same pattern as MED01-P2.

7. Verifier-direct lane (not for the researchers)

The verifier checks structural-biology facts directly against the deposited records and the structure papers. These include entry metadata, what each paper reports about its own structure, and the current absence of deposited structures with specific ligands bound. They are ledgered with origin verifier-direct and are a required target of the secondary audit.

8. What “done” looks like

A reader finishes MED02 knowing:

• where the body’s cannabinoid receptors are, and how we know;
• how its own cannabinoids are made, used and cleared;
• which roles are demonstrated in people versus inferred from animals;
• how THC and CBD reach and leave the body by each route, and at what concentrations;
• which receptor actions of each named cannabinoid happen at concentrations people actually reach.

They can also tell which popular descriptions of the system go further than the evidence. Each part adds roughly 20–35 verified claims (MED02-C001 onward; sources continue the shared S-series). Nothing in the ledger reads as advice.

9. Known limits

• This is the program’s first mechanism-heavy round. Most evidence will be animal or cell-based, so the §M3 rule (mechanism ≠ human clinical effect) will be exercised constantly. Expect the verifier to downgrade wording often.
• The base PROTOCOL’s evidence grading fits these claims better than trial risk-of-bias tools. RoB 2 / JBI apply only to the human PK and imaging studies.
• Receptor-expression findings are method-sensitive (especially antibody-based CB2 work). The rule 2 and rule 6 labels are the main safeguard.

10. Open decisions for Danny (block the freeze)

1. Run order: DECIDED (Danny, 2026-10-15). MED02 runs after MED01-P3 locks, one medical round at a time. The editorial passes and the archive check proceed whenever there is slack.
2. Parts: two parts as proposed, or one combined pass?
3. Search cutoff: set at freeze.
4. §M12 archive overlap check.
   • The Corpus Cannabis Body Lab materially overlaps. Its test list is filed separately (git-only, not given to the researchers).
   • the author and researcher’s previous writing and the 2024 medical deck likely cover ECS basics. Each needs checking before the freeze.
5. Editorial passes: ChatGPT and Gemini on this draft.

Editorial history (pre-freeze)

• v0.1 DRAFT, 2026-10-15 (Claude). Built from the Body Lab’s 151 unique EXTERNAL receipts (26 sources), grouped into 53 test items and 11 verifier-direct lookups (see the internal test list). Key questions are worded neutrally: no stance, no numbers and no sources from the lab, per the §M12 anchoring safeguard.