Friend retrovirus infection induces the development of memory-like natural killer cells
Traditionally, NK cells belong to the innate immune system and eliminate virus-infected cells through their germline-encoded receptors. However, NK cells were recently reported to possess memory-like functions that were predominantly provided by hepatic NK cells. Memory properties were mainly documented in contact hypersensitivity models or during cytomegalovirus infections. However, the precise role and the physiologic importance of memory-like NK cells during retroviral infections are still under investigation. Here, we show that Friend retrovirus (FV) infection of mice induced a population of phenotypically memory-like NK cells at 28 days post infection. Upon secondary antigen encounter, these NK cells showed an increased production of the pro-inflammatory cytokines IFNγ and TNFα as well as the death ligand FasL in comparison to naïve NK cells. Furthermore, we found an augmented elimination of antigen-matched but not antigen-mismatched target cells by these memory-like NK cells. In adoptive cell transfer experiments, equal antiviral activities of splenic and hepatic memory-like NK cells during the late phase of acute FV infection were found. Our results strongly imply the existence and antiviral activity of spleen and liver memory-like NK cells in FV infection, which efficiently respond upon secondary exposure to retroviral antigens.
KeywordsNatural killer cells Immunological memory Friend virus Retrovirus Memory NK cells Antigen specificity
Natural killer (NK) cells classically belong to the innate immune system and fight cancers as well as intracellular pathogens. They sense altered cells with their germline-encoded cell surface receptors mainly through changes in the major histocompatibility complex (MHC) class I expression or via the upregulation of ligands for their activating receptors. Activated NK cells eliminate target cells by the release of cytotoxic granules that contain granzymes and perforin. Granules are exocytosed into the immunologic synapse, ultimately inducing apoptosis . NK cells also express death receptor ligands on the cell surface, e.g. Tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL) and FasL, that result in elimination of abnormal cells upon binding with its death receptors . In addition, NK cells produce a broad range of pro-inflammatory cytokines such as interferon (IFN) γ, Tumor necrosis factor (TNF) α and Granulocyte–macrophage colony-stimulating factor (GM-CSF), which can interfere with tumor cell proliferation or virus replication .
Although NK cells have traditionally been considered as innate lymphocytes with no memory formation, O´Leary and colleagues described for the first time memory-like features of NK cells in a contact hypersensitivity (CHS) model . Immunological memory is classically defined as the ability to rapidly respond and provide protection against previously encountered pathogens. The process of memory formation includes the longevity of memory cells, which is usually initiated during the acute infection, and recall responses upon a secondary challenge with the identical pathogen. Classical immunological memory that is usually attributed to T and B cells is divided into three phases: (1) expansion of immune cells; (2) contraction phase, in which the majority of effector cells undergo apoptosis; (3) memory phase with long-living, self-renewing subsets of immune cells . Similar to NK cell memory to haptens, skin sensitization with virus-like particles containing antigens from HIV-1, influenza or vesicular stomatitis virus resulted in memory NK cell responses, although NK cells lack antigen-specific rearranged receptors . Hence, for the memory formation of these distinct immune cells antigen-specificity is required that should differ between NK cells and T cells. Currently, two concepts of antigen specific recognition of virus-infected target cells by NK cells are being postulated. (1) For the expansion and differentiation of adaptive NK cells, viral proteins can be recognized by their activating receptors, e.g. the MCMV-encoded protein m157 that directly interacts with the activating NK cell receptor Ly49H in the mouse . In human cytomegalovirus (CMV) infection, the activating receptor CD94-NKG2C+ recognize the non-classical MHC I molecule HLA-E as cognate ligand . HLA-E surface molecules require peptide loading for stabilization, which are provided by the host or by HCMV-encoded peptides (e.g. UL40) that mimic peptide sequences of host [9, 10]. Hammer and colleagues demonstrated that a single substitution in a UL40-derived peptide resulted in altered NK cell binding to HLA-E and distinct UL-40 peptides mediated different degrees of NK cell proliferation . (2) The second concept is based on the role of inhibitory receptors on NK cells, which usually recognize MHC I molecules . However, on the cell surface MHC I molecules are always associated with a bound peptide within the binding groove. Virus-induced changes in the peptide reservoir presented by MHC I molecules can alter the strength of NK cell receptor binding to the inhibitory MHC I, hence, influencing NK cell inhibition. This can ultimately lead to NK cell activation and target cell lysis if a viral peptide in the MHC I binding groove strongly interferes with MHC I recognition . Interestingly, single amino acid changes in the presented peptide can result in abolished or enhanced MHC I binding and subsequently triggers or dampens the NK cell degranulation, e.g. in HIV-1 infection [14, 15].
Infections with cytomegaloviruses (CMV) or simian immunodeficiency virus (SIV) and also haptens were shown to induce an antigen-specific NK cell memory [7, 16]. Effective memory NK cells can be found in spleens and livers of SIV- and MCMV-infected animals, although hapten-induced memory NK cells as well as memory NK cells from influenza-infected mice were only present in livers [4, 7, 16, 17]. Especially hepatic NK cells have been shown to possess characteristics of adaptive immune cells such as clonal expansion, longevity and robust recall responses including the production of IFNγ and degranulation . For the induction of adaptive NK cells also pro-inflammatory signals are essential. Especially IL-12 but also IFNα are important for the induction of memory NK cells [18, 19]. The cytokines IL-12, IL-15 and IL-18 induce long-lived NK cells, which produce IFNγ after restimulation, whereas they do not show increased cytotoxicity .
While there are several studies on memory-like NK cells in CHS and CMV models, not much is known about adaptive NK cells in retroviral infections. To address this issue, we explored the functionality and phenotypical characteristics of memory-like NK cells in FV-infected mice. Here we demonstrate the existence of memory-like NK cells during FV infection. We revealed the occurrence of adaptive NK cells in FV-infected bone marrow, spleens and lymph nodes as well as in the liver. Their killing of target cells was antigen-specific and correlated with the production of pro-inflammatory cytokines as well as the expression of death receptor ligands.
Phenotyping of memory-like NK cells from FV-infected mice
Taken together, only the enhanced expression of KLRG-1 and CXCR6 on NK cells after FV infection suggested the induction of a memory-like NK cell population.
NK cells in FV-infected animals are highly cytotoxic after re-challenge with FV-induced tumor cells
Here, we demonstrate the pro-inflammatory cytokine production and the cytotoxicity of memory-like NK cells during infection of mice with FV.
NK cells from infected livers as well as from other FV-infected organs exhibit augmented antiviral activities in an adoptive transfer experiment
FV-sensitized NK cells eliminate FV antigen-expressing target cells
These data demonstrate that NK cells from FV-infected mice eliminate target cells in an antigen-specific manner and thus develop a memory-like phenotype during infection.
Infections with viruses cause numerous health issues. Currently, 37 million people worldwide are living with the human immunodeficiency virus (HIV) that can cause a deadly immunodeficiency syndrome . The antiretroviral therapy (ART) currently used for the treatment of HIV patients suppresses viral replication; however, a cure is still impossible. Therefore, further basic research is necessary to find a way to cure or functionally cure HIV infections and prevent AIDS-associated disease. Natural killer (NK) cells are one of the cytotoxic cell populations that are important for the control of acute retroviral infections [29, 30]. In this study, we analyzed the NK cell phenotype and functionality during the late phase of acute FV infection in the mouse. Our findings demonstrate the development and antigen-specific activity of memory-like NK cells during retroviral infection.
In the past years, memory-like NK cells have been studied using several mouse models [31, 32]. Memory-like NK cells can be phenotypically characterized analog to memory CD8+ T cells with the markers KLRG-1, CD27, CD62L, Ly-6c but also using CD11b, Thy-1 and CXCR6 [4, 6, 21, 23]. During FV infection, a higher percentage of NK cells expressed KLRG-1 and CXCR6, but not the other memory markers that have been described. CXCR6 is upregulated on hepatic NK cells, according to Paust and colleagues who described CXCR6 as a required chemokine receptor for memory-like liver NK cells in the CHS and virus-like particle (VLP) context .
The localization and the organ-associated function of adaptive NK cells can be different in different models. In models of CHS or influenza virus infection adaptive NK cells reside in the liver whereas an efficient memory response was also detectable within the spleen of MCMV-infected or SIV-infected animals. In FV infection, spleen-derived memory-like NK cells slightly better eliminated target cells upon second antigen encounter compared to hepatic NK cells. FV replicates efficiently in the spleen but very little virus is present in livers of FV-infected animals (data not shown). Thus, our findings indicate that the memory function of adaptive NK cells can develop independently of organ distribution but depend on the amount of viral antigens within these organs.
It was previously described that memory-like NK cells produce higher amounts of IFNγ and increase the expression of CD107a after secondary antigen encounter, indicating a direct cytotoxic function of NK cells after stimulation . After the challenge of FV-experienced mice with FV-induced FBL-3 tumor cells (48 h), we also found an increased percentage of IFNγ and TNFα producing NK cells as well as a higher expression of FasL. We also detected an increased expression of granzyme B very early after challenge with FBL-3 cells (6 h, data not shown). These findings confirm the rapid secretion of pro-inflammatory cytokines as well as the direct cytotoxicity of memory-like NK cells towards secondary target cells.
NK cells recognize MHC I molecules with their inhibitory receptors, but it was also reported that the nature of the peptide, which is presented in the MHC binding groove, is of substantial importance for this recognition and may affect NK cell functions . It was shown that amino acid changes in peptide sequences can abolish or trigger the binding of NK cells to target cells [33, 34, 35]. In the current study we show that FV antigen-experienced NK cells potently eliminate FV antigen-matched target cells. Beside studies in murine models, a study in a macaque SIV and SHIV infection model revealed that the recognition of target cells by memory NK cells is antigen-specific . Moreover, NK cells from spleens and livers of macaques were both functional and able to lyse matched target cells similar to NK cells from FV model.
Taken together, we demonstrated the existence of memory-like NK cells during FV infection and their antigen-specific elimination of target cells upon secondary exposure to FV antigen. Overall, the FV mouse model is very powerful to study adaptive NK cells in order to get deeper insights into the molecular pathways of their induction and to find solutions for their clinical use. This might be important for cancer immunotherapy, the treatment of viral diseases or the improvement of vaccination strategies.
Materials and methods
Mice and infection
Inbred C57BL/6 mice were purchased from Harlan Envigo Laboratories and maintained under pathogen-free conditions. Experiments were performed using female C57BL/6 mice of the age of at least 7 weeks. Infection of mice was done with the FV complex containing B-tropic Friend murine leukemia helper virus and polycythemia-inducing spleen focus-forming virus. The FV stock was prepared as a 15% spleen cell homogenate from susceptible BALB/c mice infected 14 days previously with 3000 SFFU of FV. Mice were injected intravenously with 0.1 ml phosphate-buffered saline (PBS) containing 40,000 SFFU of FV, if not indicated differently. The virus stock was free of lactate dehydrogenase-elevating virus. Mice were sacrificed at 28 dpi by cervical dislocation.
Single cell suspensions were prepared from spleens and livers. Erythrocytes were lysed in liver samples. After gradient centrifugation, hepatocytes were washed and used for stainings. Cells were stained with the following antibodies: CD3 (17A2), CD11b (M1/70), CD27 (LG.3A10), CD62L (MEL-14), CD69 (H1.2F3), CXCR-6 (SA051D1), FasL (MFL3), GzmB (GB11), IFNγ (XMG1.2), KI-67 (B56), KLRG-1 (2F1), Ly-6c (HK1.4), NK1.1 (PK136) or TNFα (MP6-XT22). Zombie UV Fixable Viability Kit (BioLegend) was used for the exclusion of dead cells. If stimulation was necessary, cells were stimulated with ionomycin (500 ng/ml), phorbol myristate acetate (PMA; 25 ng/ml), monensin (1X), and brefeldin A (2 µg/ml) diluted in Iscove’s modified Dulbecco’s medium (IMDM) buffer at 37 °C for 3 h. For intracellular stainings, cells were fixed with the Cytofix/Cytoperm Fixation/Permeabilization kit (BD) or with Foxp3/Transcription Factor staining buffer set (eBioscience). Cells were measured at LSR II (BD).
Infectious center assay
Infectious centers (IC) were detected by tenfold serial dilution of single-cell suspensions onto Mus dunni cells. Co-cultures were incubated for 3 days and fixed with 96% ethanol. Fixed cells were washed twice with PBS plus bovine serum albumin (BSA) and stained with F-MuLV envelope-specific monoclonal antibody 720. After a second wash with PBS + BSA cells were incubated with peroxidase-conjugated goat anti-mouse antibody. For the detection of foci, assay was developed with aminoethylcarbazol.
In vivo cytotoxicity assay
NK cell transfer
NK cells were isolated from spleen, bone marrow and lymph nodes (mix) or livers of mice according to the manufacturer’s protocol (MojoSort Mouse NK cell isolation kit, BioLegend). Purity of NK cell isolation was checked at LSR II (> 85%). The fraction of < 15% NK1.1-negative cells contained almost no B cells, T cells and granulocytes but DCs and macrophages. 5 × 105 hepatic cells or 1 × 106 mixed NK cells were transferred intravenously into naïve mice that were infected at the same day with 20,000 SFFU of FV. At 3 dpi, spleens were removed and viral loads were detected.
In vitro antigen-specific NK cell killing assay
Based on the previously described protocol, the generation of bone marrow-derived DCs was done with some modifications . In brief, cells were isolated from murine tibias and femurs and placed on petri dish plates containing 10 ml of DC media (RPMI supplemented with 10% FCS, 2 mM l-glutamine, 50 nM 2-mercapotoehanol, 1 mM sodium pyruvate, 0.5% penicillin/streptomycin, 5 ng/mL GM-CSF, and 1 ng/mL IL-4) and were incubated at 37 °C in a humidified 5% CO2 atmosphere. After 24 h, 10 ml of DC media was added. 15 ml of media was changed at day 3. At day 7, non-adherent cells were used for experiments. DCs were loaded either with FV GagL peptide (2 µg/ml), with the class I peptide epitope of ovalbumin (OT-I, 2 µg/ml) or without any peptide (unloaded) for at least 1 h at 37 °C. Unloaded DCs were stained with carboxyfluorescein succinimidyl ester (CFSE, 2.5 µM, BioLegend) whereas loaded cells were labeled with Tag-it Violet Proliferation and Cell Tracking Dye (2.5 µM, BioLegend). Loaded and unloaded cells were mixed in a 1:1 ratio. NK cells were isolated from spleens and livers according to the manufacturer’s instructions (MojoSort Mouse NK cell isolation kit, BioLegend). Purity of NK cell isolation was checked at LSR II (> 85%). Mixed DCs were co-cultured with NK cells in a 10:1 effector to target ratio for 18 h at 37 °C. Cells were immediately measured at LSR II. For calculation, the ratio of violet-labeled cells (loaded) versus CFSE-labeled cells (unloaded) was calculated and normalized to the ratio of mixed DCs without NKs.
Statistical analyses were done with GraphPad Prism version 6. Statistical differences between two groups were analyzed by an unpaired t test (parametric) and differences between multiple groups were examined by Kruskal–Wallis test (non-parametric) or Ordinary one-way ANOVA (parametric).
ELS supervised, designed and performed the experiments, analyzed the data, participated in the statistical analysis. TN performed several experiments and participated in the analysis of data. SS performed several experiments. ELS and UD conceived the study and wrote the paper. All authors read and approved the final manuscript.
This work was supported by a Grant from the University of Duisburg-Essen (BIOME-PEP).
The authors declare that they have no competing interests.
Availability of data
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.
Consent for publication
Ethics approval and consent to participate
Animal experiments were performed under strict consent with the German regulations of the Society for Laboratory Animal Science (GV-SOLAS) and the European Health Law of the Federation of Laboratory Animal Science Associations (FELASA). North Rhine-Westphalia State Agency for Nature, Environment and Consumer Protection (LANUV) approved all experiments and protocols.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 25.Zelinskyy G, Dietze KK, Husecken YP, Schimmer S, Nair S, Werner T, Gibbert K, Kershaw O, Gruber AD, Sparwasser T, Dittmer U. The regulatory T-cell response during acute retroviral infection is locally defined and controls the magnitude and duration of the virus-specific cytotoxic T-cell response. Blood. 2009;114:3199–207.CrossRefGoogle Scholar
- 28.UNAIDS. http://www.unaids.org/en/resources/fact-sheet. Accessed 28 April 2018.
- 35.Holzemer A, Thobakgale CF, Jimenez Cruz CA, Garcia-Beltran WF, Carlson JM, van Teijlingen NH, Mann JK, Jaggernath M, Kang SG, Korner C, et al. Selection of an HLA-C*03:04-restricted HIV-1 p24 gag sequence variant is associated with viral escape from KIR2DL3 + natural killer cells: data from an observational cohort in South Africa. PLoS Med. 2015;12:e1001900 (discussion e1001900).CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.