L-Serine

Defects in the three enzymes that synthesize L-serine — most commonly 3-phosphoglycerate dehydrogenase (3-PGDH) — cause congenital microcephaly, severe psychomotor retardation and intractable seizures, and are unusual among neurometabolic diseases in being treatable ^{2Reference 2de Koning et al. · 2004ReviewSerine-deficiency syndromes — reviewView study →}.

Oral L-serine, often combined with glycine, can arrest otherwise drug-resistant seizures: in one 4½-year-old with 3-PGDH deficiency, seizures responded only to L-serine and glycine supplementation ^{3Reference 3Coşkun et al. · 2009Case reportCoşkun, T., Aydin, H. I., Kiliç, M., et al. (2009). 3-phosphoglycerate dehydrogenase deficiency: a case report of a treatable cause of seizures — case report. The Turkish Journal of Pediatrics. https://pubmed.ncbi.nlm.nih.gov/20196394/View study →}.

This is a genuine repletion benefit — the supplement replaces a metabolite the patient cannot make — but it applies to a rare, genetically defined group, and dosing must be watched: in two siblings, very high-dose serine (1400 mg/kg/day) arrested head growth (reversible on dose reduction) and lowered CSF methionine, isoleucine and ornithine through competition for transport ^{4Reference 4Häusler et al. · 2001Case reportPhenotypic heterogeneity and adverse effects of serine treatment in 3-phosphoglycerate dehydrogenase deficiency: report on two siblings — case reportView study →}.

Gap: benefit is confined to people with a proven biosynthetic defect; no RCT (evidence is case reports/series), and the therapeutic window is dose-sensitive.

HSAN1, caused by SPTLC1/SPTLC2 mutations, produces neurotoxic 1-deoxysphingolipids because the mutant enzyme uses alanine instead of serine; loading the system with L-serine pushes the enzyme back toward its normal substrate. In a randomized, placebo-controlled trial (n=18, 400 mg/kg/day for 1 year with an open-label second year), the L-serine group improved on the Charcot-Marie-Tooth Neuropathy Score relative to placebo (−1.5 units, 95% CI −2.8 to −0.1), with reduced plasma deoxysphingolipids; the authors graded it Class I evidence that high-dose L-serine slows disease progression ^{1Reference 1Fridman et al. · 2019RCTRandomized trial of l-serine in patients with hereditary sensory and autonomic neuropathy type 1 — randomised controlled trialView study →}.

An iPSC model of the same disease shows L-serine-responsive neurite deficits, supporting the mechanism ^{12Reference 12Clark et al. · 2021In vitroAn iPSC model of hereditary sensory neuropathy-1 reveals L-serine-responsive deficits in neuronal ceramide — in vitro/iPSCView study →}.

Gap: a single small trial (16 completers) in a rare disease; “slows progression” is not “reverses,” and the 400 mg/kg/day dose is far above ordinary supplemental use.

GRIN disorders are rare developmental encephalopathies caused by variants in NMDA-receptor subunit genes; L-serine is a precursor of D-serine, an NMDA co-agonist, giving a rationale in loss-of-function variants. A phase 2A, open-label, single-arm trial (NCT04646447; n=24, ages 2–18, 52 weeks) reported improvements in several Vineland adaptive-behavior subdomains for the cohort and a significant Bayley cognitive gain in the more severely affected group ^{8Reference 8Juliá-Palacios et al. · 2024RCTL-serine treatment in patients with GRIN-related encephalopathy: a phase 2A, non-randomized study — clinical trialView study →}.

Gap: no control group and no blinding, so developmental gains cannot be separated from natural change or expectation effects; hypothesis-generating only.

The ALS rationale is the BMAA hypothesis: the cyanobacterial non-protein amino acid β-N-methylamino-L-alanine can be misincorporated into proteins in place of L-serine, causing misfolding and aggregation, and excess L-serine inhibits that misincorporation in vitro ^{9Reference 9Dunlop et al. · 2013In vitroThe non-protein amino acid BMAA is misincorporated into human proteins in place of L-serine causing protein misfolding and aggregation — in vitroView study →}.

A randomized, double-blind phase I trial (n=20; 0.5–15 g twice daily for 6 months) found L-serine generally safe and, against historical placebo controls, it “did not appear to accelerate” functional decline ^{5,6Reference 5Levine et al. · 2017RCTPhase I clinical trial of safety of L-serine for ALS patients — randomised controlled trial (phase I)View study →Reference 6Bradley et al. · 2018Clinical trialStudies of Environmental Risk Factors in Amyotrophic Lateral Sclerosis (ALS) and a Phase I Clinical Trial of L-Serine — clinical trial (phase I)View study →}.

Independent review (ALSUntangled 38) judged the evidence preliminary ^{7Reference 7ALSUntangled 38: L-serine · 2017ReviewALSUntangled 38: L-serine (2017) — review. Amyotrophic Lateral Sclerosis & Frontotemporal Degeneration. https://pubmed.ncbi.nlm.nih.gov/27894192/View study →}.

Gap: this is a safety/tolerability study, not an efficacy trial — comparison was to historical controls, three patients died during the trial, and the underlying BMAA-causation hypothesis remains unproven.

Low serine and accumulation of neurotoxic 1-deoxysphingolipids are implicated in diabetic peripheral neuropathy, and human genetic evidence (a 2026 preprint) links the serine-biosynthesis pathway to the disease ^{11Reference 11Fridman et al. · 2026Human genetic evidence links serine biosynthesis to diabetic peripheral neuropathy — genetic association study (preprint). medRxiv. https://pubmed.ncbi.nlm.nih.gov/42326776/View study →}. In db/db mice, 6 months of oral L-serine (5–20% of diet) suppressed toxic 1-deoxysphingolipids and improved functional/sensory measures, though structural axon degeneration still progressed ^{10Reference 10Xia et al. · 2023AnimalLong-term effects of l-serine supplementation upon a mouse model of diabetic neuropathy — animal studyView study →}.

Gap: entirely preclinical plus genetic association — no completed human efficacy trial, and even in mice the structural protection was incomplete.