SAMHD1 p.A565T — the science¶
The enzyme¶
SAMHD1 (SAM domain and HD domain–containing protein 1) sits at a critical evolutionary intersection of nucleotide metabolism, DNA repair, and innate immunity. It is not a single-job protein; it is a hub with at least seven documented functions:
- dNTPase — hydrolyzes the DNA building blocks (dNTPs) to keep cytoplasmic nucleotide pools low. This is its best-known role and the primary way it restricts retroviruses (including HIV-1) and mobile genetic elements.
- 3′→5′ exonuclease — degrades stray single-stranded DNA before it can trip innate-immune DNA sensors.
- Condensate formation (LLPS) — sequesters cytoplasmic nucleic acids away from the sensors MDA5 and TLR3.
- DNA-repair scaffold — supports homologous recombination and stabilizes stalled replication forks.
- Transcriptional co-repressor — restrains a second wave of type I interferon by occupying an inhibitory site on IRF7.
- Mitochondrial guardian — interacts with VDAC1 to stabilize the mitochondrial membrane and limit leakage of mitochondrial DNA into the cytosol.
- Cell-cycle–gated switch — its dNTPase activity is turned off by phosphorylation at residue T592 during DNA synthesis.
The enzyme is deeply conserved across vertebrates, which tells us this checkpoint matters.
The variant¶
| Field | Value |
|---|---|
| Genomic | NM_015474.3 c.1693G>A |
| Protein | p.Ala565Thr (rs779491090) |
| Population frequency | ≈ 1.1 × 10⁻⁵ (gnomAD v4) — very rare |
| Location | extreme C-terminal edge of the HD phosphohydrolase domain, 27 residues from the T592 regulatory hinge |
Swapping a small alanine for a bulkier, hydroxyl-bearing threonine at position 565 is predicted to distort the C-terminal regulatory lobe of the enzyme — the "latch" that gates its activity.
What the lab work shows (established): In cell-line work from a 2022 doctoral dissertation (LMU München), A565T reduced protein stability roughly 2.4-fold and caused near-complete loss of dNTPase activity. That is cell-line data; it has not yet been confirmed in primary human cells from a heterozygous carrier.
Haploinsufficiency vs. dominant-negative (open question): We currently model the variant as acting through haploinsufficiency — one good copy is not enough for full function. Whether the impaired copy actively poisons the healthy protein (a dominant-negative effect on the four-unit SAMHD1 complex) rather than simply dropping out of the pool is an unresolved structural question, and it matters for how any future gene therapy is designed. Structural modeling to test this is ongoing.
The mechanism: one root cause, four streams¶
The working model is that reduced SAMHD1 function lifts several brakes at once. ~40–60% residual dNTPase activity leaves the cytosolic dNTP pool running high, and that single upstream event branches into four color-coded streams — two converging on the NLRP3 inflammasome, two on a chronic, moderate-amplitude type I interferon tone the body cannot switch off. The full breakdown, with a diagram and citations for each link, lives on the pathway map; the clinically relevant summary is below.
🔵🔴 Blue → red — the interferon / JAK-STAT axis (partly treatable today)¶
Loss of the SAMHD1–VDAC1 interaction (blue stream) opens a route for mitochondrial DNA to escape into the cytosol, where cGAS-STING signaling produces type I interferon. That interferon then drives sustained JAK-STAT signaling (red stream), which upregulates ISG15 and blocks mitophagy — so damaged mitochondria pile up instead of being cleared, feeding back into more mitochondrial damage.
Because this axis runs through JAK-STAT, JAK inhibitors can suppress it — which matches the clinical observation that a JAK inhibitor rapidly clears most inflammatory features, and that they relapse within 24–48 hours of stopping (consistent with a re-priming node that resets on that same timescale).
🟣🟡 Purple + gold — the nucleotide / NLRP3 metabolic axis (JAK-resistant)¶
In parallel, the same excess dNTPs overload mitochondrial nucleotide import (purple stream), stalling replication and releasing oxidized mtDNA — a direct NLRP3 ligand — while also being catabolized to uric acid and MSU crystals (gold stream), a second, independent NLRP3 activator. A 2026 Science study (Liu et al.) showed SAMHD1 loss drives insulin resistance and steatohepatitis at normal body weight — a diet-independent metabolic disease mechanism running through this same NLRP3 route. This axis does not respond to JAK inhibition, which likely explains the residual fatigue and metabolic floor that persists even when the interferon axis is well controlled.
Two mechanistically independent axes → why one drug isn't enough
The clearest clinical prediction of this model is that a JAK inhibitor resolves the interferon-driven features almost completely but leaves a hard floor of fatigue and metabolic dysfunction untouched — because that floor is generated by a parallel, JAK-resistant axis. Treating the whole syndrome therefore requires hitting the upstream nucleotide/NLRP3 axis too. See the pathway map for how each stream is scored and cited individually.
What we don't yet know¶
- Whether A565T acts purely by haploinsufficiency, or partly dominant-negative.
- Confirmation of the purple/gold NLRP3 streams in primary cells from a heterozygous carrier (the strongest data so far is from knockout, not heterozygous, systems) — see the pathway map for the gold stream's evidence caveat specifically.
- Structural predictions of the C-terminal latch have not yet been folded into the formal concept papers and are kept strictly separate from established biochemistry.
Primary references informing this model include Liu et al. (Science, 2026) on the dNTP→NLRP3 metabolic mechanism and Rieser/Walczak et al. (Nat Cell Biol, 2026) on the interferon gain-control node. The research blog tracks new evidence weekly.