Deazaflavin

Deazaflavin (TND1128) is a synthetic 5-deazaflavin — a flavin scaffold where the N5 nitrogen has been replaced by carbon. That single atom swap changes the redox chemistry dramatically: instead of the one-electron radical cycling that classical flavins do, TND1128 runs a two-electron hydride transfer like NAD(P)H. It behaves less like a vitamin cofactor and more like a synthetic NAD(P)H mimetic that can auto-cycle in and out of mitochondrial redox pools without being consumed the way NMN and NR are. The downstream phenotype is a polarized inner mitochondrial membrane, more on-demand ATP, and structural reinforcement of neurons — at roughly 1/100th the mass dose of NMN.

Auto-Redox Cycling and the NAD(P)H-Mimetic Premise#

The defining feature of TND1128 is that it can shuttle hydride equivalents reversibly. In a working mitochondrion, NADH delivers electrons to Complex I; FADH₂ delivers to Complex II; the electron transport chain pumps protons and ΔΨm builds. TND1128 inserts into that loop as an additional redox carrier that is not stoichiometrically depleted — it cycles. The practical consequence is that the same milligram of compound continues to support electron flow over a long functional window, which is why a single subcutaneous dose in mice produces effects measurable nearly a day later rather than dissipating within hours.

This is the mechanistic divergence from NMN/NR. Those compounds work by raising the size of the NAD⁺ pool; TND1128 works by directly participating in the redox shuttle. Same destination, different vehicle.

Mitochondrial Membrane Potential and On-Demand ATP#

The cleanest functional readout from the published work is ΔΨm polarization and increased ATP synthesis in brain tissue.

"A single subcutaneous administration of TND1128 in mice led to a significant increase in ΔΨm and ATP generation, which peaked at 22 ± 2 hours post-dosing." — Takahashi N. et al. J Pharmacol Sci, 2024

The 22-hour peak is important. It tells the community two things: first, this is not an acute stimulant — there is no minute-to-minute "feel" the way caffeine or methylene blue can produce. Second, once-daily dosing is sufficient because the functional effect window outlasts the plasma half-life. For physique-focused and longevity-focused users, the relevant downstream is mitochondrial supply-side support — the cell's capacity to meet demand from heavy training, GH-axis peptide cycles, or simply age-related ETC drift.

Calcium Buffering and Excitotoxic Protection#

A second mechanistic axis is Ca²⁺ overload buffering. When neurons depolarize hard (KCl challenge, ischemia, excitotoxic glutamate flood), cytosolic Ca²⁺ spikes and mitochondria absorb the excess — at high enough loads, this triggers permeability transition and cell death.

"TND1128 buffered both cytosolic and mitochondrial Ca²⁺ overload after depolarization, while β-NMN at 100-fold higher dose only buffered cytosolic Ca²⁺." — Takahashi N. et al. J Pharmacol Sci, 2023

The mitochondrial-compartment effect is what NMN cannot reproduce. For users running heavy stimulants, recreational drug history, post-concussion recovery, or simply chasing neuroprotection in an aggressive PED context, this is the most directly translatable mechanism on the page.

Structural Neuroplasticity — Dendrites, Axons, Synapses#

Beyond the bioenergetic effects, TND1128 produces a structural growth signal in hippocampal neurons that NMN does not.

"TND1128 significantly promoted dendritic growth, axonal arborization, and excitatory synapse formation in cultured hippocampal neurons, surpassing the effect of β-NMN despite matching NAD⁺ concentrations." — Katsurabayashi S. et al. Biochem Biophys Res Commun, 2021

Matching NAD⁺ concentrations is the key control. It rules out the "more NAD⁺ = more synapses" interpretation and forces the conclusion that TND1128 is doing something downstream of bulk NAD⁺ levels — most likely supporting sirtuin- and PGC-1α-driven mitochondrial biogenesis at the dendrite, where local ATP demand for synapse formation is highest. This is the mechanistic foundation for the nootropic / cognitive use case and explains why the cognition score on this profile (6) outpaces the recovery score (5).

Oxidative-Stress Cytoprotection#

The final mechanism worth flagging is direct antioxidant / cytoprotective activity against ROS.

"TND1128 exerted robust cytoprotective activity against H₂O₂-induced cell death and promoted cell proliferation in non-malignant lines." — Kubota K. et al. bioRxiv, 2024

H₂O₂ is the standard in vitro proxy for ROS stress — the same kind of redox load generated by heavy training volume, AAS-driven elevations in baseline oxidative stress, and the chronic low-grade inflammation that defines aging tissue. TND1128 preserves cell viability under that load and supports proliferation in healthy (non-malignant) cells. Practically, this is the rationale for stacking it during heavy training blocks and during cycles where oxidative load is known to be elevated.

The same proliferative signal is also why active malignancy sits on the contraindication list — the compound has no oncology safety dataset, and a redox-cycling cytoprotectant is not something to administer alongside an active tumor until that data exists. That limit is real, not a hedge. For everyone outside that population, the mechanism stack — auto-redox cycling, ΔΨm polarization, Ca²⁺ buffering, synaptic growth, and ROS protection — is one of the cleaner mitochondrial pharmacologies currently available to the research-peptide community.

Common (most users)#

Deazaflavin's adverse-event profile in the published literature is genuinely sparse — the mouse work to date shows no acute toxicity at doses up to 10 mg/kg i.p., and the cytoprotective study demonstrated proliferative effects on healthy cells rather than damage (Kubota et al., 2024). What follows is drawn from that preclinical record plus early-adopter community reports.

• Transient insomnia or "wired" feeling with late-day administration — the on-demand ATP pharmacology (Takahashi et al., 2024) drives this. Mitigation: morning or early-afternoon dosing only; the 22-hour functional window means a single AM dose covers the day.
• Mild injection-site reactions (redness, transient bump, mild soreness) on subcutaneous protocols. Mitigation: dilute the reconstituted solution further with bacteriostatic water (a 10mg vial in 3–4mL rather than 2mL), rotate sites, and let the solution come to room temperature before injection.
• Mild headache or pressure sensation in the first few days, particularly when stacked with NMN/NR. Mitigation: hydration, magnesium, and starting at the low end (0.5–1mg SubQ) before titrating.
• Vivid dreams — reported anecdotally in the same way as with several mito-active compounds. Not documented in the literature, generally self-resolving within the first 1–2 weeks.

Uncommon (dose-dependent or individual)#

• Jitteriness, mild tachycardia, or stimulant-like restlessness at the upper end of the dose range (3–5mg SubQ). Mitigation: drop back to 1–2mg, move dosing earlier, and reduce or cycle off concurrent methylene blue, caffeine, or other mito-stimulants.
• Fatigue or "flat" feeling 24–48h after a high dose — paradoxical but consistent with the delayed ΔΨm peak at ~22h (Takahashi et al., 2024). Mitigation: every-other-day dosing rather than daily, or a 5-on/2-off pattern.
• GI discomfort on oral capsule protocols — typically the nano-absorption matrix rather than the molecule itself. Mitigation: dose with food.
• Sleep architecture shifts in users who layer TND1128 onto an already-heavy nootropic/mito stack. If HRV or resting heart rate trend upward across a block, drop the stack down to TND1128 alone for a week to isolate the driver.

Rare but serious#

The peer-reviewed safety record is too thin to confidently enumerate serious effects — none have been documented in the published mouse work, and there is no human pharmacovigilance dataset. The two mechanism-derived concerns worth flagging:

• Theoretical pro-proliferative signal in undiagnosed malignancy. The H₂O₂ cytoprotection study demonstrated proliferative effects on non-malignant cells (Kubota et al., 2024). No tumor-bearing safety study has been published. Warning signs that warrant discontinuation: unexplained lumps, persistent night sweats, unexplained weight loss, or any new symptom that would prompt a cancer workup anyway.
• Excessive mitochondrial drive when stacked with multiple ETC-active agents (methylene blue, urolithin A, PQQ, CoQ10, NMN all at once). Warning signs: persistent tachycardia, elevated resting heart rate, declining HRV across a block. Drop everything except TND1128 and reassess.

Hard contraindications#

• Active malignancy or recent cancer history. The proliferative cell-viability signal combined with the absence of any tumor-bearing safety data makes this the one population where TND1128 is not appropriate until the literature catches up. This is a real limit, not a hedge.
• Pregnancy and lactation. No data exist. Avoid entirely.
• Concurrent DNP (2,4-dinitrophenol) use. Mechanistically incoherent — DNP uncouples the proton gradient while TND1128 polarizes ΔΨm and supports ATP synthesis. DNP's safety margin is already razor-thin; layering an opposing mitochondrial agent on top is reckless.

Gender and PCT considerations#

Deazaflavin is non-hormonal. There is no documented effect on the HPG axis, androgen receptors, the estrogen pathway, or SHBG — no PCT is required and no ancillary support is needed. The same dose ranges apply across the full subject pool, with no virilization or feminization risk on either side. Female protocols use identical SubQ and oral ranges to male protocols. The only gender-specific limit is the pregnancy/lactation contraindication above.