Inside the contact

A 500 nm window on the place where a T cell meets a tumor cell. Drag to turn it. Then step through what actually has to happen — because being next to each other is necessary and about 5% sufficient.

window 500 × 500 nm · gap 14 nm

vertical scale exaggerated 5×

engager density —

drag to rotate · scroll to zoom

CD70 — anchoring antigen CD4 — gating antigen CD3 on the T cell CD45 — 28 nm, excluded engager

Where our model lives in that sequence

stage	measured	in our framework
0 · proximity	two cells within reach	ours geometry
1 · conjugation	median 15 min, 95% transient	ours adhesion first-passage
2 · close contact	~14 nm, CD45 excluded	ours as an exclusion
3 · bound engager density	N₂₀ = 40.16/µm²	ours the whole binding model
4 · receptor triggering	occupancy vs. serial engagement	edge read, not derived
5 · granule polarization	MTOC reorients to the contact	not ours no cytoskeleton
6 · perforin event	first at 6–16 min	not ours no term
7 · damage accumulation	≥3 hits, 57 min half-life	not ours no term
8 · death	1.8 ± 1.5 h after binding	ours the endpoint we fit
9 · detachment, serial kill	~1 per 2 h	not ours no term

Our model covers stages 0 to 3 and the endpoint, and nothing in between. That is stated in the paper. What is new is seeing how many stages the assumption spans — and that five of them have measured numbers of their own.

What that does to our own threshold

N₂₀ was inferred by assuming killing rises monotonically with bound density. But a 6 h assay holds about 3.3 complete kill cycles and 6.3 damage half-lives, each target needs roughly three hits inside a 57 min window, and only ~5% of contacts deliver one. So the 20% endpoint is a population statistic, not one synapse crossing a line.

Stated properly, N₂₀ is the bound-engager density at which the hit rate across the population is high enough that 20% of targets accumulate enough coincident damage in 6 h. That is a threshold on a rate — and our model contains no contact frequency and no contact duration. It changes no ratio. It changes what the absolute means.

Lateral distances true to scale; vertical exaggerated 5× so the molecules are legible — at true scale a 14 nm gap across a 500 nm window is a hairline. Receptor densities, rate constants and N₂₀ from the primary tables of the two published affinity series; contact, hit and killing statistics from the intravital and in vitro cytotoxicity literature. Nothing here was measured by us.

Observations at the SynapseThe Engager ProgramLuminary Studies