Innate Immunity
Natural Killer Cell
Innate lymphocytes that kill virus-infected and tumor cells through missing-self detection
NK cells are innate lymphocytes that kill virus-infected and tumor cells without prior sensitization. Inhibitory KIRs read MHC class I; when self is missing, perforin and granzyme fire within 30 minutes.
- Frequency5-15% of circulating lymphocytes
- Kill timeTarget apoptosis within 15-30 min
- Inhibitory receptorsKIRs reading MHC class I (HLA-A,B,C)
- Activating receptorsNKG2D, NKp46, NKp44, NKp30, CD16
- Lytic granulesPerforin pore + granzyme B caspase activator
- Therapeutic formatCAR-NK — 73% response in CD19 lymphoma
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
How NK cells decide who lives
An NK cell does not have an antigen-specific receptor. It scans every cell it touches in a fraction of a second, integrating signals from a dozen surface receptors into a single binary verdict: kill or move on. The arithmetic looks like this:
- Inhibitory signals from KIR2DL and KIR3DL family receptors recognizing self HLA-A, B, C — these recruit SHP-1 phosphatase via ITIM motifs and shut down activating kinases.
- Activating signals from NKG2D (binding MICA/B and ULBP — stress-induced ligands upregulated in DNA-damage and viral infection), NKp46, NKp30, NKp44, and CD16 (for IgG-coated targets).
- Cytokine priming by IL-2, IL-12, IL-15, and IL-18 — produced by macrophages and dendritic cells in inflamed tissue — boosts NK activation thresholds and granule content.
The default verdict is "do not kill." Healthy cells license themselves through MHC I. Only when the inhibitory signal disappears, or when stress ligands accumulate, or when an antibody coats the target, does the NK cell tip into kill mode. This logic is precisely the inverse of T cells, which need a positive antigen signal to fire.
Missing self in practice
Why does this design exist? Because viruses evolved a beautiful exploit. CD8 T cells kill any cell presenting viral peptide on MHC I. So herpesviruses (HSV, CMV, EBV, KSHV) and HIV encode proteins that pull MHC I off the surface — US2, US11, ICP47, Nef — and become invisible to T cells. The immune system's answer: NK cells that read MHC I as a signal of safety, and kill any cell that lacks it. Selective NK deficiency presents with disseminated herpesvirus disease; the loophole is closed exactly where it would otherwise be wide open.
Tumors face the same selection pressure. Cells that downregulate MHC I escape CD8 T cells and are visible to NK cells; cells that retain MHC I escape NK cells and are visible to T cells. The combined coverage means very few tumors get away with either strategy alone. This duality drives therapeutic interest in NK-T cell combination therapy.
Worked clinical example: post-transplant CMV
A 47-year-old has received an allogeneic stem cell transplant for AML. Day 35 post-transplant, screening PCR detects CMV viremia at 1,200 copies/mL. The transplant immunologist knows that NK cell reconstitution is the first to recover — within 2-4 weeks — while T cell reconstitution takes months. The donor was selected for KIR-B haplotype because B haplotypes have a higher density of activating KIRs (KIR2DS1, 2DS4), and patients receiving KIR-B donor grafts have significantly lower CMV reactivation and relapse rates. Letermovir is started, and the recovering NK pool — with high CD16 and NKG2C expression — controls the virus through ADCC against CMV-specific IgG that the donor B cells begin producing by day 60. The patient never develops symptomatic CMV disease. Every step of that course relies on NK biology that was unknown 30 years ago.
Why NK cells matter clinically
- Herpesvirus control. NK cells are the dedicated countermeasure to viral MHC I downregulation; their loss permits disseminated HSV/CMV/EBV/VZV disease.
- Tumor surveillance. Low NK activity (assessed historically by 51Cr release on K562 cells) predicts higher cancer incidence in 11-year follow-up studies.
- ADCC and antibody drugs. CD16 polymorphism FcγRIIIa-V158F predicts rituximab and trastuzumab response — V/V homozygotes outperform F/F.
- Stem cell transplant. KIR ligand mismatch between donor and recipient drives the graft-versus-leukemia effect without graft-versus-host.
- Pregnancy biology. Uterine NK cells regulate trophoblast invasion; KIR-HLA-C mismatch is implicated in preeclampsia risk.
- Emerging cell therapy. CAR-NK and allogeneic NK products promise off-the-shelf cancer immunotherapy with lower cytokine release toxicity than CAR-T.
NK cell vs CD8 cytotoxic T cell
| NK cell | CD8 T cell | |
|---|---|---|
| Recognition | Missing self + stress ligands | Specific peptide-MHC I |
| Receptors | KIR, NKG2, NKp, CD16 (germline) | TCR (V(D)J rearranged, ~10¹⁵ specificities) |
| Priming required | No — kills on first encounter | Yes — antigen presentation by DC over days |
| Time to kill target | 15-30 minutes | 30-60 minutes after primed |
| Kills MHC-low targets | Yes (preferred targets) | No (cannot see them) |
| Memory | Limited; some CMV-induced memory-like | Robust antigen-specific memory |
Common misconceptions
- NK cells are non-specific. They have specificity — just at the level of receptor-ligand integration, not antigen-MHC.
- NK cells lack memory. CMV induces NKG2C+ memory-like NK in humans, persisting and expanding on re-exposure.
- Missing self alone triggers killing. Inhibition removed plus an activating signal — both are required.
- NK cells are interchangeable with CD8 T cells. They cover different vulnerabilities; deficiency in either causes distinct infection patterns.
- NK cells are minor players. 5-15% of lymphocytes; primary defense in the first 48 hours of viral infection before adaptive immunity engages.
- Perforin alone kills. Perforin makes the pore; granzyme B does the molecular work, activating caspases inside the target.
Frequently asked questions
What is the missing-self hypothesis?
Klas Kärre proposed in 1986 that NK cells survey cells for the presence of MHC class I. Healthy nucleated cells display MHC I as a license — inhibitory KIRs (killer immunoglobulin-like receptors) on the NK cell bind it and shut down killing through ITIM-recruited phosphatases SHP-1 and SHP-2. When a virus (HSV, CMV, HIV) or tumor downregulates MHC I to escape CD8 T cells, the inhibitory signal is lost, activating receptors win, and the NK cell fires. The system is the immune system's solution to T-cell evasion.
How exactly does an NK cell kill?
After receptor balance tips toward kill, the NK cell forms an immunological synapse with the target. Cytotoxic granules — lytic vesicles preloaded with perforin and granzymes A and B — polarize toward the synapse and dock with the membrane via SNARE proteins. Perforin polymerizes into a calcium-dependent transmembrane pore (~10-20 nm), allowing granzyme B to enter the target cytosol. Granzyme B cleaves caspase-3 and BID, triggering apoptosis with DNA fragmentation and membrane blebbing — death within 15-30 minutes. The NK cell then disengages and can serially kill several targets.
How do NK cells differ from CD8 T cells?
Both use perforin and granzymes to kill, but specificity differs profoundly. CD8 T cells recognize one peptide-MHC complex through a rearranged TCR — exquisitely specific, requires prior priming, and takes days. NK cells have germline-encoded receptors integrating signals from dozens of ligands — no priming needed, kill within hours of meeting target. NK cells handle the cells T cells cannot see (MHC-low). The two are complementary, not redundant.
What is ADCC?
Antibody-dependent cellular cytotoxicity. NK cells express CD16 (FcγRIIIa), a low-affinity receptor for the Fc region of IgG. When IgG coats a target (virus-infected cell, tumor cell, or therapeutic antibody bound to its target), CD16 cross-linking activates the NK cell to release perforin and granzyme just as in missing-self killing. ADCC is a major mechanism of action for rituximab (anti-CD20 in lymphoma), trastuzumab (anti-HER2 in breast cancer), and other therapeutic antibodies — and explains why CD16-V158F polymorphism predicts response.
What happens when NK cells are defective?
Selective NK cell deficiency is rare but severe. Patients (e.g., GATA2, MCM4, or IRF8 mutations) suffer recurrent disseminated herpesvirus infections — primary varicella, CMV, EBV — because herpesviruses are masters of MHC I downregulation, and NK cells are the dedicated countermeasure. Chediak-Higashi syndrome combines NK granule defects with albinism. Conversely, NK cells contribute to graft-versus-host disease in stem cell transplant and to preeclampsia (uterine NK regulate trophoblast invasion).
Are NK cells used in cancer therapy?
Yes, and the field is growing fast. Tumor cells often downregulate MHC I, making them NK targets in principle. Strategies: (1) allogeneic NK cell infusion — donor NK kill recipient tumor without graft-versus-host risk; (2) CAR-NK cells — chimeric antigen receptor engineered NK from cord blood or iPSCs, with the safety advantage of short persistence vs CAR-T; (3) NK cell engagers (BiKEs, TriKEs) — bispecific molecules that bridge NK CD16 to tumor antigen; (4) checkpoint inhibitors targeting NKG2A and TIGIT releasing the NK kill brake. CAR-NK against CD19 lymphomas has shown 73% response rates with minimal cytokine release syndrome.
Where do NK cells live and how long?
Develop from common lymphoid progenitor in bone marrow (and partly in lymph nodes and liver). Mature NK cells circulate in blood, comprising 5-15% of lymphocytes (~250 cells/μL). Tissue-resident NK populations live in liver, uterus (decidual NK), skin, and lung — distinct phenotypes and roles. Circulating NK turn over with a half-life of about two weeks; some long-lived memory-like NK populations have been described, especially after CMV exposure, blurring the once-strict innate/adaptive divide.