After senolysis, residual NAD ranks precursor above CD38i
STATUS_LABEL: NEGATIVE
Claim
After documented senescent-cell / SASP clearance lowers CD38 load, a residual NAD-precursor pulse (NMN / NR class) raises tissue NAD more than a matched CD38-inhibitor pulse at the same post-clearance timepoint — pre-registered residual ΔNAD_precursor − ΔNAD_CD38i ≥ 20% of the larger arm (or precursor wins on a pre-registered superiority test).
Why it matters
Longevity stacks treat CD38 drugs and NMN/NR as interchangeable post-senolysis NAD boosters. L5 claimed the opposite residual ranking (CD38i > precursor) on the same missing H2H panel; it did not prove that ranking or this flip. Soft-claims: discussion — head-to-head residual ranking UNKNOWN; L5 and L6 are mutually exclusive predictions on one assay.
Mechanism sketch
Subtract-before-add: clear SASP-driven CD38+ macrophage load first, then pulse leftover levers. CD38 gates age NAD decline and NMN degradation; senescent/SASP ablation lowers CD38 and partially restores NAD; ecto-CD38 block raises NAD via an NMN-dependent route; 78c-class CD38i improves aged-mouse healthspan as monotherapy (not a post-senolysis factorial). Flip motivation: after CD38↓ the residual CD38i target shrinks; because ecto-block is NMN-dependent, residual precursor can regain the larger ΔNAD once CD38 load is already low — opposite ranking to L5.
Prior art
- PMID 27304511 / DOI 10.1016/j.cmet.2016.05.006 / PMC4911708 — age NAD↓ requires CD38; CD38 is major in vivo NMN-degrading activity (Camacho-Pereira 2016).
- PMID 33199924 / DOI 10.1038/s42255-020-00305-3 / PMC7908681 — SASP induces CD38-high macrophages that reduce tissue NAD (Covarrubias 2020).
- PMID 33199925 / DOI 10.1038/s42255-020-00298-z / PMC8752031 — senescent/SASP ablation ↓CD38 and partially restores NAD; ecto-CD38 block raises NAD NMN-dependently (Chini 2020).
- PMID 35263032 / DOI 10.1111/acel.13589 / PMC9009115 — CD38i 78c improves median survival/healthspan in aged mice; monotherapy, not post-senolysis residual ranking (Tarragó 2022).
Baseline claimed
After senolysis, residual CD38 inhibition always beats residual NAD precursor (L5 seed), or stacking both is always better than ranking which residual lever wins.
Baseline measured
CD38 gates age NAD / NMN fate LITERATURE (PMID 27304511). SASP→CD38+ macrophages→tissue NAD↓ LITERATURE (PMID 33199924). Clearance ↓CD38 + partial NAD rebound; ecto-block NMN-dependent LITERATURE (PMID 33199925). 78c aged-mouse benefit LITERATURE monotherapy (PMID 35263032). Post-clearance H2H residual ΔNAD precursor vs CD38i UNKNOWN (PubMed soft-complete 0). Residual ranking margin ≥20% PREDICTION (flip of L5). L5 does not already prove this flip — same missing assay; rankings mutually exclusive.
Actionable limitations
- Do not treat L5 as proving L6 — L5 left H2H UNKNOWN and only listed this flip as a competing PREDICTION; one 2-arm residual pulse decides both cards.
- Incomplete clearance: residual CD38+ macrophages can still favor CD38i even after partial senolysis — flip is not automatic from “CD38 fell.”
- 78c monotherapy ≠ residual after subtract — aged-mouse survival does not rank post-senolysis residual arms.
Method
propose-assay
Prediction variables
| name | kind | assay | threshold |
|---|---|---|---|
| CD38_gates_age_NAD_and_NMN | LITERATURE | Aging NAD metabolome vs CD38; NMN degradation (Camacho-Pereira 2016) | age NAD↓ requires CD38; major in vivo NMN-degrading activity (PMID 27304511) |
| SASP_CD38_macrophage_NAD_drain | LITERATURE | Senescent/SASP-conditioned macrophage CD38 and tissue NAD (Covarrubias 2020) | SASP induces CD38-high macrophages that reduce tissue NAD (PMID 33199924) |
| clearance_CD38_drop_NAD_rebound_ecto_NMN_dependent | LITERATURE | Senescent/SASP ablation → CD38↓ + partial NAD↑; ecto-CD38 block NAD boost (Chini 2020) | CD38↓ with partial NAD recovery; ecto-block raises NAD via NMN-dependent route (PMID 33199925) |
| CD38i_78c_aged_mouse_monotherapy | LITERATURE | Chronological-aging 78c survival/healthspan (Tarragó 2022) | ~10% median survival ↑ class effect as monotherapy — not senolytic→then-78c factorial (PMID 35263032) |
| post_clearance_H2H_residual_NAD_ranking | UNKNOWN | Same post-clearance window — residual precursor vs residual CD38i; tissue NAD + CD38 activity | dedicated residual ranking assay not found (soft-complete 0) — UNKNOWN |
| residual_precursor_beats_CD38i_margin | PREDICTION | Pre-registered aged tissue or SASP-conditioned system; senolytic/SASP-ablation run-in → wash → 2-arm residual pulse | ΔNAD_precursor − ΔNAD_CD38i ≥ 20% of the larger arm (or pre-registered superiority); α/N pre-specified — do not invent observed Δ |
| competing_residual_CD38i_wins_if_clearance_incomplete | PREDICTION | Same design; falsifier if residual CD38i ≥ residual precursor when post-clearance CD38 activity remains high | opposite ranking (L5 direction) allowed as pre-registered competing outcome when residual CD38 load stays elevated |
Ablate
- remove post-clearance residual 2-arm contrast (precursor vs CD38i at matched timepoint): Without the H2H residual pulse, solo CD38i or precursor LITERATURE cannot rank which lever wins after subtract.
- remove documented clearance run-in with CD38↓ / partial NAD rebound verification: Without subtract-first confirmation, the card collapses into generic precursor-vs-CD38i comparison (not L6 flip).
- remove L5≠proof + incomplete-clearance competing falsifier: Treating L5 as already deciding residual ranking, or ignoring residual CD38+ macrophages, would invent a flip or overclaim CD38↓ as automatic precursor win.
Refute if
On the pre-registered post-clearance panel, residual CD38i ΔNAD ≥ residual precursor ΔNAD (or the ≥20% precursor-superiority margin fails) when CD38↓ is verified — especially under incomplete clearance where residual CD38+ load remains high and L5 ranking can still win.
Ask a human
Lock tissue (WAT/liver), clearance verification (CD38 activity + SC/SASP marker), and identical post-clearance sampling before ranking arms; do not treat L5 as already deciding the flip; no dosing.
Cheap IP
kill_cost_tier=2 · time_to_refute=6w · ip_class=subtract-before-add · vault_hold=no
Execution
execution_status=NOT_RUN · fto_status=UNKNOWN · related_prior=L5 (not proof of flip; mutually exclusive on same H2H)