Sarcosine

Sarcosine (N-methylglycine) is a GlyT1 inhibitor + NMDA glycine-site partial agonist with A-tier evidence as a schizophrenia adjunct (Tsai 2004, Lane 2008, Singh & Singh 2011 meta-analysis) — when added to non-clozapine antipsychotics, it produces clinically meaningful improvement in negative symptoms and cognitive measures. For healthy adults including Dylan: SKIP. The clinical evidence is entirely in NMDA-hypofunction populations; healthy adults have an approximately saturated NMDA glycine site (same logic that limits oral glycine — see glycine.md) and no published RCT shows cognitive benefit in healthy young adults at any dose. Critical interaction note: sarcosine antagonizes clozapine specifically (avoid the combo); other antipsychotics fine. Dylan = no indication. This is a "fix what's broken" molecule, not a "push above baseline" tool.

Sarcosine is N-methylglycine — glycine with one hydrogen on the amino nitrogen replaced by a methyl group. It is endogenously produced as an intermediate in glycine ↔ dimethylglycine ↔ choline metabolism (via the glycine N-methyltransferase / sarcosine dehydrogenase axis), and is found at low concentrations in plasma and brain. Pharmacologically two mechanisms matter:

1. GlyT1 (glycine transporter 1) inhibition — the primary mechanism

GlyT1 (SLC6A9) is a Na⁺/Cl⁻-dependent transporter expressed predominantly on astrocytes surrounding glutamatergic synapses (with smaller neuronal expression). It is the dominant mechanism for clearing synaptic glycine after release, and its activity sets the steady-state synaptic glycine concentration that supplies the NMDA receptor glycine co-agonist site on the GluN1 subunit. Sarcosine is a competitive substrate-inhibitor of GlyT1 — it occupies the transporter, blocks reuptake of synaptic glycine, and raises the synaptic glycine pool that's available to the NMDA receptor. This is mechanistically distinct from (and more pharmacologically targeted than) oral glycine substrate-loading, because:

• Oral glycine has poor BBB penetration and is rapidly cleared from synaptic compartments by GlyT1 itself.
• Sarcosine acts directly at the synapse to block clearance, producing a more sustained synaptic-glycine elevation per dose.
• Sarcosine also crosses the BBB more efficiently than glycine (smaller hydrophilic drag from the methyl group; favorable PEPT1/amino-acid transporter handling).

The result: at typical 1-2 g/day clinical doses, sarcosine produces a measurable synaptic-glycine increase in target brain regions (frontal cortex, hippocampus, striatum) that correlates with NMDA-receptor function recovery in NMDA-hypofunction states. Pharmaceutical GlyT1 inhibitors (bitopertin/RG1678 from Roche, others) were developed on the same mechanistic premise but mostly failed Phase III trials in schizophrenia for reasons that remain debated (dose, patient stratification, endpoint selection) — sarcosine itself, despite being a "natural" molecule, has more positive trial data than the synthetic GlyT1 inhibitors developed at orders-of-magnitude greater cost.

2. NMDA receptor glycine-site partial agonism

Sarcosine has weak intrinsic agonist activity at the NMDA receptor glycine-binding site (GluN1 subunit) in addition to its GlyT1 inhibition. The agonist activity is modest — full agonists at this site include glycine and D-serine — but it contributes to the net positive modulation of NMDA receptor function. The dual mechanism (raise endogenous glycine via GlyT1 block + directly activate the site) is part of why sarcosine has slightly more reliable clinical signal than pure GlyT1 inhibitors in the schizophrenia adjunct literature.

3. Why this matters less in healthy adults

The NMDA glycine site is approximately saturated under normal physiological conditions in healthy adults (synaptic glycine + D-serine are sufficient — same logic detailed in glycine.md). In schizophrenia, NMDA receptor function is hypoactive (the leading mechanistic hypothesis for the disorder's cognitive and negative-symptom dimension; consistent with PCP/ketamine inducing schizophrenia-like states via NMDA antagonism). Sarcosine raises synaptic glycine and partially agonizes the glycine site — this is therapeutic only if the system is hypoactive. In healthy adults with a saturated glycine site, additional GlyT1 inhibition produces little measurable cognitive effect, because the rate-limiting step in NMDA signaling is glutamate availability and depolarization, not glycine co-agonist supply.

This is the central reason healthy-adult sarcosine evidence is thin: the mechanism inherently doesn't push a healthy system above baseline. It's a pharmacological "fix what's broken" tool, not a cognitive enhancer.

4. Other mechanistic notes

• Sarcosine is metabolized by sarcosine dehydrogenase (SARDH) in mitochondria back to glycine (oxidative N-demethylation, releasing CH₂O which feeds into one-carbon metabolism via 5,10-methylene-tetrahydrofolate). Half-life is short (~1-2 hours plasma), justifying split dosing in clinical protocols.
• Sarcosine is a normal urinary metabolite. It briefly attracted attention as a putative prostate cancer biomarker (Sreekumar et al. 2009 in Nature) — that finding has been substantially walked back in subsequent studies and is not relevant to oral supplementation safety, but it's the reason "sarcosine" sometimes shows up in oncology literature.
• Sarcosine is not a methyl donor in the SAMe sense — although it sits adjacent to one-carbon metabolism, supplementation does not meaningfully alter SAMe/homocysteine status at typical 1-2 g doses.