The Interferonopathies and the Self-DNA Problem
In an interferonopathy, nothing is wrong with the sensor. cGAS reports that double-stranded DNA is in the cytosol, exactly as it is designed to \u2014 it does not read sequence and it never could. What has failed is the housekeeping that was supposed to ensure there was nothing there to find. TREX1 clears cytosolic DNA; SAMHD1 restrains the retroelements that generate it; RNase H2 removes ribonucleotides misincorporated during replication; ADAR1 edits endogenous double-stranded RNA so it does not look foreign. Lose any one of them and you get Aicardi\u2013Goutieres syndrome. Push the sensors the other way \u2014 gain-of-function STING or RIG-I \u2014 and you get SAVI or Singleton\u2013Merten. Three of the four protective genes are DNA-repair or nucleotide-metabolism enzymes, which is the real lesson of this map: genome maintenance is an immunological function. Click any protein for the matching Assay Genie In Vivo antibody or ELISA kit.
These are diseases of failed cleaning, not of faulty sensing. The distinction matters because it changes where a therapy has to act. TREX1 loss causes Aicardi\u2013Goutieres syndrome type 1, familial chilblain lupus and retinal vasculopathy; SAMHD1 loss causes type 5; RNase H2 mutations are the commonest cause of all; and ADAR1 loss causes type 6. In every one of them the sensor is behaving correctly on material that should not be there. Three of those four genes are not immune genes by any conventional classification \u2014 RNase H2 is a DNA-repair enzyme, SAMHD1 a dNTP triphosphatase \u2014 and their clinical presentation is nonetheless an interferon-driven inflammatory disease of the brain and skin. Genome maintenance is an immunological function, and that is the most portable idea on this map.
The sensors fail in the opposite direction, and the phenotypes rhyme. Gain-of-function STING causes STING-associated vasculopathy with onset in infancy; gain-of-function RIG-I causes Singleton\u2013Merten syndrome; gain-of-function MDA5 causes Aicardi\u2013Goutieres type 7. Common variants in the same MDA5 gene alter risk of type 1 diabetes and lupus, so a single locus carries both a monogenic interferonopathy and a polygenic autoimmune signal \u2014 which is as direct an argument as exists that the common autoimmune diseases sit on the same axis as the rare ones, differing in degree rather than in kind.
The third way to get an interferonopathy is to fail to stop. USP18 binds IFNAR2 and shuts signalling down after the first wave, and its deficiency causes pseudo-TORCH syndrome. ISG15 is the counter-intuitive one: it is an interferon-stimulated gene, so the naive expectation is that losing it would reduce the response \u2014 but ISG15 stabilises USP18, so ISG15 deficiency causes interferonopathy by removing the brake. Anyone reading an ISG15 result as a simple activity marker will get the direction wrong. This third mechanism also explains why the IFNAR2 node is the most frustrating gap in the reagent set: it is the chain USP18 acts on and the one anifrolumab blocks, and it is ELISA only here.
The In Vivo tie-in, and where this map is thin. 16 of the 36 nodes carry a functional-grade antibody, and none of them is upstream of the interferon. The genes that cause these diseases cannot be blocked: TREX1, SAMHD1, RNase H2, cGAS, STING, RIG-I and MAVS are research grade; ADAR1, MDA5, USP18, TBK1, IRF3, TLR9, IFN-\u03b1, IFNAR2, ISG15, MX1 and OAS1 are ELISA only. What is available is the intervention that matters clinically: IFNAR1 is functional grade, and blocking it removes the entire type I response \u2014 all thirteen alpha subtypes and beta together \u2014 in a single reagent, which is the closest available model of what anifrolumab does. IFN-\u03b2 itself is neutralisable, so the axis can be interrupted at the cytokine or at the receptor and the two answers compared. IFN-\u03b3 (clone XMG1.2) with IFN-\u03b3R1 lets type II be told apart from type I experimentally rather than assumed distinct, which matters more here than anywhere: an "interferon signature" assembled from genes both induce is reporting two systems as one. Around them sit class I in both species, PD-L1 (10F.9G2), CD11c, CD19, CD40 (FGK4.5, ultra-low endotoxin \u2014 LPS drives interferon directly and would be scored as CD40 agonism), CD3 (145-2C11), CD4 (GK1.5), CD8 (Ly-2), IL-1\u03b2, TNF and IL-10. The design that follows is to interrupt at IFN-\u03b2 or IFNAR1 and read the signature out through MX1 and OAS1 \u2014 which is also how the clinical literature confirms an interferonopathy in the first place. One reciprocal relationship worth carrying into any experiment here: type I interferon suppresses inflammasome-driven IL-1\u03b2, so blocking one arm can worsen the other, and TNF blockade runs the same relationship in reverse \u2014 it induces anti-nuclear antibodies and occasionally lupus-like disease, apparently through a rise in type I interferon. For research use only; not for use in diagnostic or therapeutic procedures.
Every protein node links to a product — ELISA kit, In Vivo antibody or research antibody.