Influenza Vaccine Induces Intracellular Immune Memory Of Human NK Cells

We often suffer from the flu more than once, but do not always seem to benefit from a memory immune response even after influenza vaccination. Influenza vaccines elicit antigen-specific antibodies and immune memory to protect humans from infec-tion with drift variants. However, what supports or limits vaccine efficacy and dura-tion is unclear. Here, we vaccinated healthy volunteers with annual vaccine formula-tions and investigated the dynamics of T cell, natural killer (NK) cell, and antibody responses upon re-stimulation with heterologous or homologous influenza virus strains. Influenza vaccines induced potential immune memory NK cells during the first6months with increased antigen-specific recall IFN-y responses, which could be blocked by anti-NKp46and might result from surface NKp46receptor internalization. Thus, intracellular NKp46expression may elicit memory NK cell generation and be a determining factor of vaccine efficacy. Moreover, IL-2signals worked cooperatively with the HA-NKp46interaction to enlarge the recall response. These findings provide insights into a novel mechanism underlying vaccine-induced immunity and NK-related diseases, which may help to design future persisting and universal vac-cines.Several key points are made in this study:(1) Antibodies to HA are associated with protection from disease or viral replication after natural influenza infection or vaccination. We show that the protection through the humoral arm of the immune response has limitations. As assessed by the influenza hemagglutination-inhibition (HI) assay, increased HI titers were only detected to A/California/7/2009(H1N1) antigens, which were shared among the seasonal vac-cines, but were not detected against antigens from heterologous viruses not included in the seasonal vaccines (like A/PR8). Seasonal influenza vaccine likely provides pro-tection only to antigens contained in that vaccine. Once the vaccinated host encoun-ters a different influenza virus (as they are subject to rapid mutation), however, the ability of the humoral response to protect the host is decreased by low HA-specific antibody titers to a heterologous influenza virus. (2) We show that peripheral NK cells, but not CD3+T cells, were able to elicit IFN-y upon both homologous and heterologous influenza subtypes re-stimulation, suggest-ing that an advantage of the NK cell memory-like response may be its ability to pro-vide broad immunity against many influenza subtypes.(3) Influenza vaccination down-regulates membrane surface NKp46expression on CD3-CD56dim NK cells1month post-vaccination, surface NKp46+NK cell percent-age was lowest at3months, but gradually normalized over6months. Furthermore, while NKp46+CD56bright NK cell frequency remained relatively stable in all subjects, NKp46+CD56dim NK cell frequency significantly decreased in vaccinated subjects. We also evaluated other memory-associated surface molecules on total peripheral CD3"CD56+NK cells by FACS. Besides NKp46, no differences were observed in ex-pression of CD45RO, CD45RA, NKp44, CD57, NKG2C, CXCR6, CD27and CD62L between the vaccinated and unvaccinated subjects. Upon further analysis, we found that low NKp46expression remarkably accompanied high IFN-y secretion from NK cells in response to A/PR8re-stimulation in most vaccinated subjects.(4) Down-regulated surface NKp46expression might result from receptor internaliza-tion upon NK cell interaction with HA. While surface NKp46-positive (NKp46(S)+) NK cell frequency decreased, intracellular NKp46-positive (NKp46(I)+) NK cells in-creased. Pre-vaccination, CD56dim NK cells exhibited little intracellular NKp46and no IFN-y in response to A/PR8. In contrast, after vaccination, CD56dim NK cells up-regulated IFN-y in response to A/PR8re-stimulation, which remained low in the absence of A/PR8. NKp46(I)+CD56dim NK cells increased within1month, peaked at3months, and dropped to near baseline levels by6months, coinciding with the wave of IFN-y production against A/PR8. Taken together, the dynamic increase in NKp46(I)+CD56dim NK cells positively correlated with increased IFN-y production to A/PR8re-stimulation after vaccination, suggesting the emergence of intracellular NKp46converted and greatly promoted NK cell effector function against viral re-stimulation.(5) Influenza vaccines induced potential immune memory NK cells during the first6months with increased antigen-specific recall IFN-y responses, which could be blocked by anti-NKp46. We also treated the FluaRIX vaccine-stimulated PBMCs or purified NK cell cultures with an IL-2-neutralizing antibody and showed that neutral-izing IL-2did not significantly inhibit IFN-y production from the purified NK cells in response to the recall antigen, while PBMC cultures exhibited higher IFN-y levels than purified NK cells. Taken together, our results suggest that NK memory recall re-sponses are dependent on NKp46and specific to influenza antigens, while IL-2sig-nals work cooperatively with the HA-NKp46interaction to enhance the NK recall response to fluA. NK recall responses are finite, which may restrict long-lasting vac-cine-mediated protection.In this study, we identify a previously undemonstrated role for circulating human NK cell memory in a broad response to influenza vaccination. Our recent results further imply that NK cells may be an important consideration for the design of improved influenza virus vaccines. Influenza vaccine induces memory NK cells with the feature of intracellular NKp46expression. Thus, our study suggests a novel mechanism of memory NK cell generation, which may help to design future persisting and universal vaccines.