TIGIT, DNAM-1 and the CD155 Axis: Pathway, Function and Assays

CD155 is read by four different receptors, and only one of them turns the killing programme on. The nectin and nectin-like ligands induced on stressed and transformed cells are simultaneously an activating cue through DNAM-1 (CD226) and an inhibitory cue through TIGIT, CD96 and PVRIG. Which signal wins is decided by receptor affinity, by surface stoichiometry, and by a cis-interaction on the effector membrane that has nothing to do with the ligand at all. That balance sets whether an NK or CD8⁺ T cell degranulates.

One ligand family, four receptors

CD155 (PVR, Necl-5) and CD112 (nectin-2) are scarce on healthy tissue. They are induced by the DNA-damage response through ATM and ATR, by Ras–MAPK signalling and by Toll-like receptor stimulation, which is why they accumulate on transformed, senescent and virally infected cells. Four receptors read that stress report and they disagree: DNAM-1 (CD226) activates, while TIGIT, CD96 and PVRIG inhibit. TIGIT binds CD155 about a hundred-fold more tightly than CD226.

So the axis is governed by stoichiometry rather than by the presence of any single molecule. Chronic antigen exposure and TGF-β raise TIGIT while CD226 is internalised, ubiquitinated and shed, so the CD226:TIGIT ratio collapses even where total CD155 is unchanged. PVRIG adds a second, CD155-independent brake, binding CD112 alone. Reporting CD155 alone therefore tells you little; quantifying CD226, TIGIT, CD96 and CD112 together makes the readout interpretable.

The activating arm: DNAM-1 through Grb2, Vav1 and PI3K

DNAM-1 carries no ITAM. Its short tail has a PKC-dependent serine and a Src-family-phosphorylated tyrosine, and it works by physically associating with LFA-1 — integrin ligation is what brings Fyn into position to phosphorylate the tail. CD226 is therefore a costimulatory amplifier, not a primary trigger: in a T cell the base signal still arrives through the TCR and ZAP-70, and in an NK cell DNAM-1 cooperates with NKG2D and 2B4 rather than replacing them.

Once phosphorylated, the tail recruits Grb2, and Grb2 is the branch point. Its SH3 domains bring in Vav1, the Rho-family exchange factor driving actin clearance at the synapse, microtubule-organising-centre reorientation and lytic-granule delivery. Grb2 also nucleates PI3K, generating PIP₃ and activating AKT, while the SOS arm feeds Ras–MEK–ERK. Vav1 is where signalling turns mechanical: a cell can phosphorylate every upstream kinase correctly and still fail to kill.

The brake: TIGIT, the ITT-like motif and SHIP-1

TIGIT's tail carries a classical ITIM and an immunoglobulin tail tyrosine (ITT-like) motif. Phosphorylation of the ITT-like tyrosine docks Grb2 and β-arrestin 2, which recruit the inositol phosphatase SHIP-1. SHIP-1 hydrolyses PIP₃ back to PI(3,4)P₂, collapsing the PI3K–AKT output, and dampens the adaptor arms feeding NF-κB; the ERK branch is blunted in parallel. Cytokine transcription is usually lost before cytotoxicity is, so the two effector outputs do not disappear on the same timescale.

Grb2 appears on both sides of this map. The adaptor coupling DNAM-1 to Vav1 and PI3K is the same one coupling TIGIT to SHIP-1, and only the tail it docks on decides the outcome — so total Grb2 is not a directional readout. TIGIT also inhibits by a second route, binding CD226 in cis and disrupting the homodimerisation DNAM-1 signalling requires. Removing one brake while CD96 and PVRIG remain explains much about modest single-agent responses.

The output: perforin, granzyme B and NF-κB

Everything upstream converges on two measurable outputs. Vav1-driven cytoskeletal reorganisation delivers granules containing perforin and granzyme B to the synapse, where perforin permeabilises the target membrane and granzyme B enters to cleave BID and caspase-3. In parallel, NF-κB (RelA/p65) translocation downstream of AKT and ERK drives IFN-γ, TNF and the survival genes that keep an effector cell alive through repeated engagements. Granzyme B is the closest available proxy for the balance.

There is a regulatory dimension to the same wiring. TIGIT is high on Tregs, where it marks a more suppressive subset biased towards restraining Th1 and Th17 responses, and TIGIT engagement of CD155 on dendritic cells shifts them towards an IL-10-high, IL-12-low state. The axis therefore suppresses at three points — effector, regulatory and antigen-presenting cell — worth remembering when a blocking experiment gives a larger effect than a single-cell assay predicted.

Why anti-TIGIT is a combination agent

The clinical dependency runs through CD226. Anti-TIGIT works by freeing CD155 for the activating receptor, so if CD226 has been shed or silenced — common in exhausted tumour-infiltrating lymphocytes — there is no accelerator left to press. A TIGIT-high, CD226-low population looks like an ideal target on a TIGIT stain and responds poorly, which makes CD226 the better stratification variable. PD-1 compounds this by recruiting SHP-2 to dephosphorylate CD226 directly.

Lineage matters too. NK cells depend on DNAM-1 to kill MHC-I-low targets and carry TIGIT and CD96 constitutively, so blockade can restore cytotoxicity without antigen-specific priming, whereas CD8⁺ T cells acquire TIGIT progressively with chronic stimulation and respond more slowly. One practical point is easy to miss: an anti-TIGIT with an active IgG1 Fc engages FcγR and depletes TIGIT-high Tregs, so Fc format changes the mechanism of action, not merely the pharmacokinetics.

Key targets and matching reagents

Studying the TIGIT–DNAM-1 axis in vivo

This axis is a receptor–ligand competition, so it suits functional-grade blocking antibodies paired with quantitative readouts. Experiments fall into three groups.

1. Manipulating the ligand and the co-expressed checkpoints

Anti-mouse CD155 (PVR) is the direct handle on the shared ligand: blocking it removes input to CD226, TIGIT and CD96 at once, the cleanest way to show a phenotype is nectin-driven before dissecting individual receptors. Pair it with anti-PD-1 (RMP1-14) to test co-blockade synergy. All are supplied low-endotoxin and azide-free — important here, because LPS itself induces CD155 through TLR signalling and would confound the very measurement you want.

2. Assigning the phenotype to NK cells or T cells

Because both lineages carry the full receptor set, depletion controls are not optional. Anti-NK1.1 removes NK cells and anti-CD3 (145-2C11) addresses the T-cell compartment; running the blocking experiment in each depleted background shows which population the effect requires. This separates DNAM-1-dependent, MHC-I-independent NK killing from TCR-dependent CD8⁺ responses that would otherwise be scored as one result.

3. Quantifying the module and the output

Measure receptors and ligands with TIGIT, DNAM-1, CD96 and CD112 kits, and cover the CD155-independent arm with PVRIG/CD112R reagents. For the proximal module assay ZAP-70, Grb2, Vav1, PI3Kδ, AKT, ERK and the brake, SHIP-1. Close the loop with RelA/p65 for transcription and perforin and granzyme B for cytotoxic output.

Featured products for this pathway

Frequently asked questions

For research use only. Not for use in diagnostic or therapeutic procedures.