Key takeaways

• CD8+ T cells recognise short peptides presented on MHC class I, which every nucleated cell expresses — so any infected or transformed cell can be inspected and killed.
• CD8 is a co-receptor, not the recognition unit: it binds the invariant alpha-3 domain of MHC class I and stabilises the TCR-peptide-MHC interaction.
• Priming of naive CD8+ T cells requires dendritic cells, and frequently cross-presentation — routing external antigen onto MHC class I.
• Killing proceeds by two routes: granule exocytosis, where perforin admits granzyme B to activate caspases, and Fas-FasL death receptor signalling.
• Killing is directional — granules are released into an immunological synapse, so neighbouring cells are spared and one T cell can kill serially.
• After clearance, a small memory population persists as central, effector and tissue-resident subsets with different locations and behaviours.
• Under chronic antigen, CD8+ T cells enter a distinct exhaustion programme marked by PD-1, LAG-3 and TIM-3 — the state checkpoint blockade targets.

Contents

• What CD8+ T cells are
• Antigen recognition and the role of CD8
• Priming and activation
• Mechanisms of cytotoxicity
• Cytokine secretion
• CD8+ T cells in viral infection
• Memory subsets
• T cell exhaustion
• Cancer immunity and immune evasion
• Immunotherapy approaches
• References
• Choosing reagents
• Frequently asked questions

Reagents for CD8+ T cell work

Identification needs CD3 with CD8; function is measured through the effector molecules released. PD-1 covers the exhaustion axis.

CD8a

Purified Anti-Human CD8a Antibody [OKT-8]

PurifiedHuman

Identifies cytotoxic T cells; paired with CD3 to exclude CD8-expressing NK subsets.

View product →

CD3

FITC Anti-Human CD3 Antibody [OKT-3]

FITC conjugateHuman

The pan-T lineage gate — required alongside CD8 for an unambiguous CTL identification.

View product →

Granzyme B

Human Granzyme B ELISA Kit

Sandwich ELISAHuman

The principal effector protease; a direct measure of cytotoxic degranulation.

View product →

Perforin

Human Perforin ELISA Kit

Sandwich ELISAHuman

Forms the pores through which granzymes reach the target cell cytosol.

View product →

IFN-γ

Human IFN-gamma ELISA Kit

Sandwich ELISAHuman

The signature CD8 cytokine, and the basis of interferon-gamma release assays.

View product →

PD-1

Human PD-1 (PDCD1) ELISA Kit

Sandwich ELISAHuman

The canonical exhaustion and checkpoint marker targeted by anti-PD-1 therapy.

View product →

What CD8+ T cells are

CD8+ T cells, also called cytotoxic T lymphocytes, are the arm of adaptive immunity that kills host cells which have become infected or transformed. They develop in the thymus, circulate as naive cells, and on encountering their specific antigen differentiate into effectors capable of destroying target cells directly.

Their counterpart, the CD4+ helper T cell, coordinates other cells rather than killing. The division reflects which MHC class each reads — CD8 with class I, CD4 with class II — and that in turn determines which compartment each polices.

Antigen recognition and the role of CD8

A point often glossed over: CD8 itself does not recognise the antigen. The T cell receptor does. CD8 is a co-receptor that binds the invariant alpha-3 domain of the MHC class I molecule, stabilising the interaction and recruiting the kinase Lck to initiate signalling. Without CD8, the TCR-peptide-MHC interaction is generally too weak to trigger a response.

What the TCR reads is a peptide of roughly 8 to 10 amino acids sitting in the MHC class I groove. Those peptides derive from proteins degraded by the proteasome in the cytosol, transported into the endoplasmic reticulum by TAP and loaded onto class I molecules.

The consequence is that MHC class I displays a continuous sample of what a cell is making internally. Because every nucleated cell expresses class I, any cell producing viral or mutated protein becomes visible — which is precisely the surveillance function CD8+ T cells provide.

Priming and activation

Naive CD8+ T cells cannot be activated by an infected tissue cell alone. They require a professional antigen-presenting cell, in practice a dendritic cell, and three signals:

A dendritic cell that has not itself been infected can still prime a response through cross-presentation — taking up material from dying infected or tumour cells and routing it onto MHC class I rather than class II. The cDC1 subset specialises in this, which is why cDC1 abundance in tumours correlates with response to checkpoint blockade.

CD4+ helper T cells frequently license the dendritic cell beforehand, via CD40, which is why CD8 responses are often weaker when CD4 help is absent.

Mechanisms of cytotoxicity

Two pathways, used together, and both requiring direct contact with the target.

The immunological synapse

Killing is not indiscriminate. On recognising its target, the T cell forms a tightly organised contact zone — the immunological synapse — and reorients its microtubule-organising centre so that granules are released only into that space. Two things follow: bystander cells are spared despite lethal mediators being present, and the T cell can detach and kill again, serially, rather than being consumed by a single encounter.

Apoptosis rather than necrosis is the outcome, which matters because it contains viral contents rather than spilling them, and because granzyme B activates caspase-independent nucleases that degrade viral genomes.

Cytokine secretion

Beyond killing, CD8+ T cells shape the surrounding response:

• Interferon gamma activates macrophages, upregulates MHC class I on nearby cells — making them easier to inspect — and has direct antiviral effects.
• TNF-alpha contributes to target cell apoptosis and to local inflammation.
• Chemokines including MIP-1 alpha and beta recruit further effector cells to the site.

Interferon-gamma release is also the practical readout for antigen-specific CD8 responses — the basis of ELISpot and of the interferon-gamma release assays used in tuberculosis testing.

CD8+ T cells in viral infection

Because viruses replicate inside cells, antibody alone cannot clear an established infection — infected cells must be destroyed, and that is the CD8 compartment’s job.

That several unrelated viruses independently evolved ways to downregulate MHC class I is strong evidence of how much selective pressure CD8+ T cells apply. It also creates an opening for NK cells, which kill cells that have lost class I — the two systems cover each other’s blind spots.

Memory subsets

Most effector cells die once antigen is cleared, but a small fraction persists. Memory is not one population:

Tissue-resident memory is the more recent addition and has changed thinking about vaccination: protection at a barrier surface may depend on establishing local T_RM rather than on circulating memory alone.

T cell exhaustion

Under persistent antigen — chronic infection or a growing tumour — CD8+ T cells do not simply tire. They enter a distinct, transcriptionally and epigenetically defined differentiation state called exhaustion, and it is progressive.

• Inhibitory receptors accumulate — PD-1 first, then LAG-3, TIM-3, TIGIT and CTLA-4. The number expressed tracks the severity.
• Function is lost in sequence: IL-2 production first, then TNF-alpha, then interferon gamma and cytotoxicity last.
• Metabolism shifts, with mitochondrial dysfunction limiting the energy available for effector function.
• The state is epigenetically imprinted, which is why checkpoint blockade reinvigorates exhausted cells only partially and often transiently — the underlying programme is not fully erased.

Exhaustion is best understood as an adaptation that limits immunopathology during unresolvable infection, rather than as failure. That framing explains why it is so entrenched, and why reversing it is hard.

Cancer immunity and immune evasion

CD8+ T cells recognise tumours through neoantigens — peptides from mutated proteins that the immune system has never been tolerised to. Tumour mutational burden correlates with checkpoint inhibitor response largely because it determines how many such targets exist.

Tumours evade this in several ways:

• Loss of MHC class I or of beta-2 microglobulin, removing the display surface entirely. This is a recognised mechanism of acquired resistance to checkpoint blockade.
• PD-L1 expression, engaging PD-1 to shut down infiltrating T cells — often induced by the interferon gamma those T cells themselves produce.
• An immunosuppressive microenvironment of regulatory T cells, myeloid-derived suppressor cells, TGF-beta and adenosine.
• Physical exclusion, where T cells remain in surrounding stroma and never contact tumour cells — the “immune-excluded” phenotype.

Immunotherapy approaches

The distinction between CAR-T and TCR-engineered cells is worth holding onto: CAR-T sees only surface antigen but is unaffected by MHC downregulation, while a TCR can reach intracellular targets but fails if class I presentation is lost.

References

• Joshi, N.S. and Kaech, S.M., 2008. Effector CD8 T cell development: A balancing act between memory cell potential and terminal differentiation. The Journal of Immunology, 180(3), pp.130-134.
• Gallimore, A., and Godkin, A., 2013. Evasion of cytotoxic T lymphocytes by tumor cells. Cancer Immunology, Immunotherapy, 62(9), pp.1391-1401.
• Schumacher, T.N. and Schreiber, R.D., 2015. Neoantigens in cancer immunotherapy. Science, 348(6230), pp.69-74.
• Lim, A., et al., 2016. Checkpoint inhibitors in immunotherapy. Nature Reviews Immunology, 16(9), pp.512-524.
• Wherry, E.J. and Kurachi, M., 2015. Molecular and cellular insights into T cell exhaustion. Nature Reviews Immunology, 15(8), pp.486-499.
• Restifo, N.P., et al., 2012. Adoptive immunotherapy for cancer: harnessing the T cell response. Nature Reviews Immunology, 12(4), pp.269-281.
• Klebanoff, C.A., et al., 2016. The plasticity of T cell function during cancer immunotherapy. Nature Reviews Immunology, 16(11), pp.713-723.

Frequently asked questions