Natural killer cells are not classical antigen-presenting cells, but under the right conditions they can behave remarkably like them. The traditional textbook answer names dendritic cells, macrophages, and B cells as the professional antigen-presenting cells (APCs) of the immune system, and NK cells do not appear on that list. Yet research over the past two decades has shown that activated NK cells can display the surface molecules needed to present antigen to T cells, can directly boost T cell proliferation and cytokine production, and can do so in the very lymph node regions where adaptive immune responses are organized. The real answer is more interesting than a simple yes or no.
What Makes a Cell an Antigen-Presenting Cell
To present antigen to helper T cells, a cell generally needs two things on its surface. The first is MHC class II, a molecule that holds fragments of proteins so that CD4+ T cells can inspect them. The second is a set of costimulatory molecules, particularly CD80 and CD86, which deliver a confirming signal that tells the T cell the antigen is worth responding to. Professional APCs like dendritic cells carry both of these as part of their core identity. They produce MHC class II internally, load it with antigen fragments, and shuttle it to the cell surface. NK cells, in their resting state, typically do neither of these things, which is why immunology courses have long classified them strictly as innate effectors whose job is to kill infected or abnormal cells, not to educate T cells.
That classification started to crack when researchers noticed that activated human NK cells could express HLA-DR, the human version of MHC class II, along with costimulatory molecules. The question then became: are these NK cells genuinely presenting antigen, or is something else going on?
How NK Cells Get MHC Class II on Their Surface
There are at least two routes by which NK cells end up displaying MHC class II, and the distinction matters because each route implies something different about what the NK cell is actually doing.
The first route involves a pre-existing subset. When human blood NK cells are cultured with the cytokine IL-2, more than a quarter of them end up expressing HLA-DR. For years, researchers assumed this was because IL-2 flipped on HLA-DR production in cells that previously lacked it. A closer look revealed something different: a small population of NK cells already carries HLA-DR, and IL-2 drives that subset to expand rather than switching on expression in previously negative cells.1PubMed Central. A distinct subset of human NK cells expressing HLA-DR expand in response to IL-2 and can aid immune responses to BCG This means some NK cells may be wired from the start to participate in antigen presentation, rather than being drafted into the role only when inflammation demands it.
The second route is more surprising. Mouse NK cells activated in vivo or in vitro did not turn on MHC class II gene transcription at all, yet they still ended up with MHC class II protein on their surface. The explanation was trogocytosis, a process in which one cell physically strips patches of membrane from another cell during close contact. NK cells were grabbing MHC class II directly off the surface of dendritic cells. When MHC class II-deficient NK cells were transferred into normal mice or cultured with normal dendritic cells, they rapidly became MHC class II-positive despite being genetically unable to produce the molecule themselves.2PubMed Central. Natural killer (NK)-dendritic cell interactions generate MHC class II-dressed NK cells that regulate CD4+ T cells These “dressed” NK cells carried borrowed antigen-presentation machinery without ever making their own.
This distinction between intrinsic expression and borrowed molecules has been a source of real debate. If NK cells are merely wearing someone else’s MHC class II, are they truly presenting antigen, or are they just echoing a signal that originated from a dendritic cell? The answer depends on what downstream effect the display actually has on T cells.
Evidence That NK Cells Can Stimulate T Cells
The functional test of an antigen-presenting cell is whether it can get T cells to respond. On this front, activated human NK cells pass at least a partial version of the exam. When researchers cultured activated NK cells with CD4+ T cells that were receiving a baseline signal through their T cell receptor, the NK cells boosted T cell proliferation and the production of IL-2 and interferon-gamma in a dose-dependent manner. The strongest effect appeared at a ratio of two NK cells per T cell. Unactivated NK cells, by contrast, did not enhance T cell responses beyond what happened spontaneously.3The Journal of Clinical Investigation. Novel APC-like properties of human NK cells directly regulate T cell activation
Blocking the costimulatory molecules CD80 and CD86 on the activated NK cells reduced their T cell-boosting effect by roughly a third for proliferation and by more than half for cytokine secretion.3The Journal of Clinical Investigation. Novel APC-like properties of human NK cells directly regulate T cell activation This is significant because CD80 and CD86 are the same costimulatory molecules that professional APCs use. The fact that blocking them only partially reduced the effect suggests that NK cells also stimulate T cells through additional pathways that professional APCs may not rely on as heavily.
An important caveat: in these experiments, the NK cells were enhancing T cell responses that had already been initiated by T cell receptor engagement. They were functioning as accessory cells or costimulators rather than as the primary trigger of a new immune response. This is a meaningful distinction. A dendritic cell can pick up a pathogen, process it, and present its fragments to a naive T cell that has never seen that pathogen before, launching an entirely new adaptive immune response from scratch. There is no strong evidence that NK cells can do that. What they appear to do is amplify and shape responses that are already underway.
Where NK Cells and T Cells Actually Meet
For NK cells to influence T cell behavior, the two cell types need to be in the same physical space. The immune system is not a well-mixed soup; its cells are organized into specific tissues with distinct neighborhoods. T cell activation mostly happens in lymph nodes, particularly in zones called parafollicular areas or T cell zones. If NK cells never entered those areas, their ability to present antigen would be biologically irrelevant regardless of what surface molecules they carry.
It turns out that a specific subset of NK cells, the CD56-bright population, does reside in human lymph nodes. Histological examination of normal lymph nodes found distinct CD56-positive and CD94-positive lymphocytes sitting within the parafollicular T cell areas, exactly where T cells are activated by dendritic cells.4Blood. CD56bright natural killer cells are present in human lymph nodes and are activated by T cell–derived IL-2: a potential new link between adaptive and innate immunity These lymph node NK cells are activated by IL-2 produced by nearby T cells, creating a feedback loop: T cells activate NK cells, and the NK cells in turn may amplify the T cell response.
This finding matters because it places NK cells in the right anatomical context for APC-like behavior to have real consequences. A CD56-bright NK cell expressing HLA-DR and costimulatory molecules, sitting in the same lymph node region as antigen-specific T cells, is positioned to meaningfully influence the adaptive immune response.
Trogocytosis and Its Darker Side
The membrane-stealing process that allows NK cells to borrow MHC class II from dendritic cells is not limited to that one interaction. Trogocytosis is a general phenomenon in which immune cells trade surface molecules during close contact, and it can cut both ways. In the context of cancer, tumor cells that express the immunosuppressive molecule HLA-G can transfer it to NK cells during cell-to-cell contact. NK cells that acquire HLA-G stop proliferating, lose their ability to kill, and actually start behaving as suppressor cells that inhibit other NK cells and protect tumors from being destroyed.5PubMed Central. Trogocytosis-based generation of suppressive NK cells
This flips the script on NK cell antigen presentation in an unsettling way. If NK cells can acquire MHC class II from dendritic cells and use it to boost T cell responses, they can also acquire suppressive molecules from tumors and use those to shut immune responses down. The same flexibility that makes NK cells interesting as potential APCs also makes them vulnerable to being co-opted by the very targets they are supposed to eliminate. For cancer immunotherapy, this is a practical concern: therapies that activate NK cells need to account for the possibility that those cells may pick up tumor-derived molecules that convert them from killers into protectors of the tumor.
NK Cells as Immune Regulators During Chronic Infection
The relationship between NK cells and T cells is not always cooperative. During chronic viral infections, NK cells can actively suppress T cell activity, and in some contexts this suppression turns out to be protective rather than harmful. In mice chronically infected with murine cytomegalovirus, NK cells expressing the death-inducing ligand TRAIL specifically eliminated activated CD4+ T cells in the salivary gland. Removing this NK cell activity allowed T cells to accumulate unchecked, which cleared the virus more quickly but came at a cost: the mice developed a condition resembling Sjögren’s syndrome, an autoimmune disease characterized by immune cells invading the glands that produce saliva and tears, along with the production of autoantibodies.6Immunity. TRAIL+ NK cells control CD4+ T cell responses during chronic viral infection to limit autoimmunity
This finding reframes NK cell interactions with T cells. Rather than simply helping or hindering, NK cells appear to act as regulators that calibrate the strength of T cell responses. Too little NK cell oversight and the adaptive immune system can overshoot, attacking the body’s own tissues. This regulatory function is conceptually different from antigen presentation, but it uses some of the same machinery, particularly the close physical contact between NK cells and T cells, and it depends on NK cells being able to distinguish activated T cells from resting ones.
What This Means for Transplant Immunology
Organ transplantation is one area where the APC-like and regulatory properties of NK cells both come into sharp focus. After a transplant, the recipient’s immune system recognizes the donor organ as foreign and mounts a response that can destroy it. NK cells participate in this process, but their contribution is not straightforward. Depending on which NK cell subsets are activated and what cytokines they release, NK cells can push the immune response toward rejection by promoting inflammatory T cell responses or toward tolerance by favoring regulatory T cell populations.7PubMed Central. Natural killer cells in rejection and tolerance of solid organ allografts
This dual capacity is one reason transplant immunologists have not been able to neatly categorize NK cells as either harmful or helpful. An NK cell subset that presents donor antigen to helper T cells and costimulates them could accelerate rejection. A different subset that kills activated T cells via TRAIL could promote tolerance. The same cell type can tip the balance in either direction depending on the local environment, the timing relative to the transplant, and the signals it receives from surrounding cells.
Why NK Cells Are Not Simply Reclassified as APCs
Given all of this evidence, it might seem like the immunology field should just add NK cells to the list of professional APCs and update the textbooks. There are several reasons that has not happened, and they go beyond mere conservatism.
Professional APCs like dendritic cells have dedicated molecular machinery for processing antigen. They take up proteins from the environment, chop them into fragments inside specialized compartments, load those fragments onto newly synthesized MHC class II molecules, and transport the loaded complexes to the cell surface. This entire processing pipeline is constitutive in dendritic cells; it is what they do as a primary function. NK cells lack much of this machinery. When NK cells display MHC class II, they are in many cases wearing pre-loaded complexes borrowed from dendritic cells via trogocytosis rather than processing antigen independently.2PubMed Central. Natural killer (NK)-dendritic cell interactions generate MHC class II-dressed NK cells that regulate CD4+ T cells They are displaying a finished product without having done the manufacturing.
Additionally, the functional evidence shows that NK cells costimulate T cells whose receptors have already been engaged rather than initiating responses in naive T cells. Professional APCs are defined in part by their ability to prime naive T cells, converting them from ignorant bystanders into activated effectors for the first time. NK cells appear to operate downstream of that event, tuning the intensity and character of an existing response rather than sparking a new one.
There is also the practical question of consistency. Not all NK cells display APC-like behavior. It appears to be limited to specific subsets, particularly the HLA-DR-positive minority among circulating NK cells and the CD56-bright population found in lymph nodes.1PubMed Central. A distinct subset of human NK cells expressing HLA-DR expand in response to IL-2 and can aid immune responses to BCG Calling all NK cells APCs based on the behavior of a subset would be like calling all T cells cytotoxic because one subset kills infected cells.
Implications for Immunotherapy
The APC-like properties of NK cells are not just an academic curiosity. They have concrete implications for how therapies that harness NK cells might work. NK cell-based immunotherapies, which are being developed for cancer and other conditions, typically involve activating and expanding NK cells outside the body and then infusing them into patients. If those expanded NK cells carry HLA-DR and costimulatory molecules, they could potentially boost anti-tumor T cell responses in addition to directly killing tumor cells. That would be a two-for-one effect: innate killing plus adaptive immune amplification.
On the other hand, the trogocytosis problem means that infused NK cells could also pick up immunosuppressive molecules from the tumor microenvironment and become counterproductive. The conditions under which NK cells acquire helpful versus harmful surface molecules are still being worked out, and this is an active area of research. Therapies that combine NK cell infusions with agents that block suppressive molecules on tumor cells might help ensure that the NK cells’ APC-like properties work in the patient’s favor rather than against them.
There is also the transplant angle. If clinicians could selectively promote the tolerogenic NK cell subsets that favor regulatory T cells while suppressing the inflammatory subsets, it might be possible to reduce the need for long-term immunosuppressive drugs after organ transplantation.7PubMed Central. Natural killer cells in rejection and tolerance of solid organ allografts That remains a goal rather than a clinical reality, but the recognition that NK cells can modulate adaptive immunity through APC-like functions is part of what makes it conceivable.
The Lymph Node Feedback Loop
One of the more elegant aspects of NK cell biology is the reciprocal relationship between NK cells and the adaptive immune system in lymph nodes. CD56-bright NK cells residing in the parafollicular T cell zone are activated by IL-2 that nearby T cells produce during an ongoing immune response.4Blood. CD56bright natural killer cells are present in human lymph nodes and are activated by T cell–derived IL-2: a potential new link between adaptive and innate immunity Once activated, those NK cells upregulate costimulatory molecules and can in turn amplify the T cell response. The T cells then produce more IL-2, which further activates the NK cells, creating a positive feedback loop.
This loop could serve as a signal amplifier for weak immune responses. If an infection generates only a modest initial T cell response, the small amount of IL-2 produced could still activate local NK cells, which then provide the costimulation needed to push T cell proliferation over a functional threshold. In that sense, NK cells in lymph nodes may serve as a bridge between the innate and adaptive arms of immunity, ensuring that adaptive responses reach the strength needed to clear a pathogen even when the initial signal is modest. The fact that this depends on specific anatomical positioning, with NK cells sitting in exactly the right lymph node compartment, suggests it evolved as a deliberate feature of immune architecture rather than an accidental byproduct of cellular flexibility.