TNFA

In one line: TNF was named in 1975 for a protein that makes tumors die — but it turned out to be the master switch of inflammation, and the drugs that switch it off (for arthritis, not cancer) became the best-selling medicines in history.

---

The gene

TNF (also written TNF-α / TNFA) sits on chromosome 6p21.33, buried inside the MHC class III region — the densest immune-gene neighborhood in the genome, flanked by the lymphotoxin genes. That address is fitting: the protein it encodes, released mostly by macrophages, is a central pro-inflammatory cytokine — a master switch for inflammation. (One housekeeping note: the official gene symbol is TNF; the channel slug "TNFA" is an older alias, and the gene should not be confused with its neighbor LTA, the old "TNF-β.")

The hunt — two names, one molecule

The molecule was discovered twice, under two very different names:

• 1975 — "Tumor Necrosis Factor." Lloyd Old's group at Memorial Sloan-Kettering in New York found that the serum of mice primed with BCG and then given bacterial endotoxin contained a factor that caused solid tumors to blacken and die. They named it for what it could destroy (Carswell, Old et al., PNAS 1975). Even then they suspected the source was a host cell — "probably macrophages."
• ~1985 — "cachectin." Anthony Cerami and Bruce Beutler were chasing a different problem: the wasting (cachexia) that hollows out the body in chronic infection. They purified the culprit, "cachectin" — and then showed it was the very same protein as TNF (Beutler, Cerami et al., Nature 1985). The tumor-killer and the body-waster were one molecule.

Two faces — defender and destroyer

TNF is a textbook double-edged sword:

• In short bursts, it defends you. It is essential to mount an immune response and, in particular, to build and maintain the granulomas that wall off tuberculosis. This isn't theoretical: blocking TNF carries a boxed warning because it lets latent TB — and other serious infections — reactivate.
• When it never shuts off, it destroys you. Chronic, unresolving TNF signaling drives rheumatoid arthritis, Crohn's disease and ulcerative colitis, psoriasis, and ankylosing spondylitis — which is exactly why blocking it treats all of them.

The mechanism

TNF acts as a homotrimer — three identical subunits locked together. The trimer docks onto TNF receptors on a cell's surface and clusters them; that clustering flips an internal master switch, NF-κB, which drives the transcription of inflammation genes. A single, self-limiting burst clears an infection and shuts down. But when the signal jams on, the same machinery floods the tissue and grinds away at joints and gut. The therapeutic idea writes itself: catch TNF before it can dock, and the switch goes dark.

The stakes

Roughly 18 million people live with rheumatoid arthritis worldwide (GBD 2021: ~17.6 million in 2020), and it is one of the leading causes of years-lived-with-disability — swollen, eroding joints, and, for decades, little that could halt the damage.

The blockers — and the biggest business in pharma

The fix was to intercept TNF before it can dock, and the first generation did it with two distinct designs, both acting at the protein level:

1. Monoclonal antibodies that clamp onto the TNF protein and pull it out of circulation — infliximab (Remicade, chimeric, FDA 1998) and adalimumab (Humira, the first fully-human monoclonal antibody, FDA Dec 31 2002).
2. A decoy receptor — etanercept (Enbrel, FDA Nov 1998): two copies of the extracellular, TNF-binding part of TNF receptor 2 fused to an antibody's IgG1 Fc tail. It floats free and soaks up TNF before it reaches a cell. Crucially, it is not an antibody — it's a lure.

This class became the best-selling drug class in the history of medicine. Humira alone was the world's top-selling drug from 2012 to 2020, peaked near $20 billion a year, and has booked over $200 billion in cumulative sales.

The frontier — discover → design

Every one of those drugs had to be discovered: Humira came from years of phage-display screening. The newest approach inverts that. With deep-learning protein design — the field that won the 2024 Nobel Prize in Chemistry (David Baker) — you can hand a computer the structure of TNF and have it design a tiny protein, a "minibinder," to grip it. These designed binders are small (under ~65 amino acids), hyperstable, and bind with antibody-grade (nanomolar-or-better) affinity, which could make them cheaper to manufacture and easier to deliver than the giant antibodies — and they can be made to order rather than found. Researchers have already de-novo-designed minibinders that neutralize human TNFα in the lab (Weng et al., Communications Biology, 2025), with cell-based activity matching or beating an existing anti-TNF. These are research-stage lead molecules — not approved drugs and not yet in patients — but they point past the blockbuster antibodies toward biologics that are designed, not discovered.

A protein named for killing cancer became the most valuable target in medicine — by teaching us when to switch it off.

Sources

Full claim-by-claim evidence is in references.md. Primary anchors:

• Gene/locus & function: NCBI Gene 7124 · OMIM 191160
• Named "Tumor Necrosis Factor" (1975): Carswell, Old et al., PNAS 72:3666-3670 — PMID 1103152
• Cachectin = TNF (1985): Beutler, Cerami et al., Nature 316:552-554 — PMID 2993897
• Mechanism (trimer → NF-κB): Regulation of NF-κB by TNF-family cytokines, PMC4156877 · Reactome: TNFR1-induced NF-κB
• Defender face / infection risk: TNF Inhibitors, StatPearls NBK482425 · Anti-TNF & infection, PMC3615849
• Disease burden: GBD 2021 rheumatoid arthritis, Lancet Rheumatol (PMC10546867)
• Drugs: Adalimumab/Humira FDA history · Etanercept = TNFR-Fc fusion (review) · Mechanisms of anti-TNF agents, PMID 27567553
• The business: Pharmaceutical Accountability Foundation: Humira · AJMC: the first $20B drug

Accuracy note: the episode is careful with the things people routinely get backwards — (1) TNF was named to kill tumors, but the blockbuster drugs work by switching it off, and they treat autoimmune disease, not cancer; (2) "Tumor Necrosis Factor" and "cachectin" are the same molecule; (3) TNF is a needed defender in bursts (blocking it reactivates TB) and a destroyer when chronic; and (4) the two drug designs are distinct — a monoclonal antibody (infliximab/Humira) versus a decoy receptor (etanercept), which is not an antibody.