Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2025 Jul 24;55(7):e70009. doi: 10.1002/eji.70009

LAG3 Marks Activated but Hyporesponsive NK Cells

Valeria Vasilyeva 1,2,3, Olivia Makinson 1,2,3, Cynthia Chan 1,2,3, Maria Park 1,2,3, Colin O'Dwyer 1,2,3, Ayad Ali 4, Abrar Ul Haq Khan 2,3, Christiano Tanese de Souza 1, Mohamed S Hasim 1,2,3, Sara Asif 1,2,3, Reem Kurdieh 1,2,3, John Abou‐Hamad 1,5, Edward Yakubovich 1,5, Jonathan Hodgins 1,2,3, Paul Al Haddad 1,2,3, Giuseppe Pietropaolo 6,7, Julija Mazej 7, Hobin Seo 8,9,10,11,12, Qiutong Huang 8,9,10,11,12, Sarah Nersesian 1,5, Damien Chay 13, Nicolas Jacquelot 8,9,10,11,12, David Cook 1,5, Seung‐Hwan Lee 2,3, Giuseppe Sciumè 7, Stephen Waggoner 4,14, Michele Ardolino 1,2,3,✉, Marie Marotel 1,2,3
PMCID: PMC12288813  PMID: 40705324

ABSTRACT

NK cells are critical for immunosurveillance, yet become dysfunctional when chronically stimulated by virally infected or cancerous cells. This phenomenon is similar to T cell exhaustion but less characterized, limiting therapeutic interventions. As shown for T cells, NK cells often display an increased expression of immune checkpoint proteins (ICP) following chronic stimulation, and ICP blockade therapies are currently being explored for several cancer types, with remarkable patient benefits. Nevertheless, the nature of ICP expression in NK cells is still poorly documented. In this study, we aimed to identify the conditions that lead to and the phenotype of immune checkpoint LAG3‐expressing NK cells. Using various experimental models, we found that LAG3 is expressed by murine NK cells upon activation in different contexts, including in response to cancer and acute viral infections. LAG3 marks a subset of immature, proliferating, and activated cells, which, despite activation, have a reduced capacity to respond to a broad range of stimuli. Further characterization also revealed that LAG3+ NK cells exhibit a transcriptional signature similar to that of exhausted CD8+ T cells. Taken together, our results support the use of LAG3 as a marker of dysfunctional NK cells across diverse chronic and acute inflammatory conditions.

NK cells quickly upregulate LAG3 upon activation. LAG3+ NK cells present hallmarks of immune exhaustion. The role of LAG3 in imposing NK cell dysfunction remains to be elucidated.

graphic file with name EJI-55-e70009-g003.jpg

1. Introduction

While natural killer (NK) cells play a key role in immunity against cancer and viral infections, loss of NK cell functional activity has often been observed in pathological conditions in both mice and humans [1‐12]. Little is known about NK cell dysfunction, precluding the development of therapeutic interventions. Mechanistic investigations have also been challenging due to the lack of reliable surface markers that identify bona fide dysfunctional NK cells. This contrasts with the progress made in describing T cell exhaustion, whereby several molecular flags, including immune checkpoint proteins (ICP), have been used to paint a detailed molecular picture [13]. To this end, we and others have investigated the validity of using ICPs as markers of dysfunctional NK cells, but obtained mixed results; for example, PD‐1+ NK cells exhibit a more activated and functional phenotype than their PD‐1‐negative counterparts [14, 15].

Recently, we turned our attention to LAG3, whose expression on exhausted T cells has been well‐characterized [16]. Although LAG3 expression on NK cells was first described in the 1990s, its function remains unclear. LAG3‐deficient NK cells in mice exhibit reduced cytotoxic function in vitro [17], but blockade of LAG3 on human NK cells did not affect their ability to kill tumor cells [18]. Combined blockade of LAG3 and PD‐1 in cancer patients resulted in enhanced cytotoxic potential of NK cells [20]. Moreover, LAG3 was proposed to inhibit cytokine production in human NK cells [21]. These studies have revived the debate on the contribution of LAG3 to NK cell biology.

Here, we set out to investigate the expression and function of LAG3 in NK cells. Using multiple models, we defined common phenotypic and functional characteristics of LAG3+ NK cells, including (i) a more immature and activated phenotype; (ii) unresponsiveness to stimulation; and (iii) a transcriptional profile resembling that of exhausted CD8+ T cells.

2. Results

2.1. LAG3 is Expressed on NK Cells in Several Inflammatory Contexts

To determine which stimuli trigger LAG3 expression on NK cells, we isolated murine splenic NK cells and stimulated them for 48 h with IL‐2, IL‐15, IL‐12/18, or type I IFNs. NK cells stimulated with these cytokines robustly upregulated LAG3 expression, while PD‐1 was not induced, consistent with the notion that murine NK cells fail to express PD‐1 [15, 22]. Epigenetic analysis of the Lag3 locus revealed histone marks for promoter accessibility, such as H3K4me3, and for active enhancers, including H3K27ac and H3K4me1, that were enriched after cytokine treatment compared with those observed in untreated NK cells (Figure 1B). Moreover, the recruitment of the acetyltransferase p300 at the LAG3 locus corroborates the increased enhancer activity induced by stimulation with IL‐2+IL‐12 (Figure 1B).

FIGURE 1.

FIGURE 1

Open in a new tab

LAG3 is expressed on NK cells in several inflammatory contexts. (A) Expression of LAG3 and PD‐1 in cytokine‐treated mouse NK cells (n = 4). A representative flow plot is shown. (B) Genomic snapshots depicting chromatin accessibility (ATAC‐seq), histone marks (H3K4me3, H3K27ac, H3K4me1), and p300 distribution in resting and activated NK cells Lag3 locus. (C, D) LAG3 expression on splenic mouse NK or T cells upon Poly (I:C) treatment. Unpaired nonparametric (Mann–Whitney) test. (E) Organ‐wide analysis of LAG3 expression in NK cells upon poly (I:C) treatment (n = 3). (F–I) LAG3 expression in NK cells gated using CD11b/CD27 (F, n = 4); KLRG1 (G, n = 6); DNAM‐1 (H, n = 6; or CD62L/CD27 (I, n = 5). Nonparametric paired Friedman test (p‐value <0.0001 except at D1) for (F) and (I). Nonparametric Wilcoxon test for (G) and (H).

Next, we analyzed whether inflammation induced LAG3 expression on NK cells in vivo. We treated C57BL/6 mice with poly(I:C), a Toll‐like receptor‐3 (TLR3) agonist often used to induce NK cell activation. NK cells rapidly upregulated LAG3 expression, which peaked at day 2 after stimulation (Figure 1C,D). However, this upregulation was transient, and LAG3 levels returned to baseline by 4 days poststimulation (Figure 1C). By contrast, T cells failed to upregulate LAG3 at any time point analyzed (Figure 1C). Poly (I:C) treatment induced LAG3 upregulation also in NK cells from lymph nodes, liver, and lungs (Figure 1E). On the other hand, NK cells in the Peyer's patches and the small intestine, which abundantly expressed LAG3 at baseline, did not exhibit further LAG3 upregulation.

Next, we sought to define common phenotypic features associated with LAG3 expression. Murine CD49b+ NK cells undergo three maturation stages determined by the expression of CD27 and CD11b: immature NK cells (R1: CD27highCD11blow), CD27 and CD11b double‐positive (R2), and CD27lowCD11bhigh mature NK cells (R3) [23]. While all these NK subsets expressed LAG3 upon poly(I:C) treatment, the highest frequency of LAG3+ NK cells was found in the R1 subset, where a distinct LAG3+ population was observed not only at baseline but also after stimulation (Figure 1F).

To confirm the higher expression of LAG3 on immature NK cells, we analyzed its co‐expression with KLRG1, an inhibitory receptor that mainly marks mature and functional NK cells [24]. KLRG1− NK cells expressed higher levels of LAG3 than the more mature KLRG1+ counterparts (Figure 1G). We then stratified NK cells based on the expression of DNAM‐1, an activating receptor which marks an alternative functional maturation program [25], but the two subsets expressed similar levels of LAG3 (Figure 1H). Finally, the frequencies of LAG3+ cells were higher in a type one innate lymphoid cell‐like (ILC‐1)‐like NK cell subset (characterized by high expression of CD27 but low expression of CD62L) compared with conventional NK cells that expressed CD62L and lower levels of CD27 [26] (Figure 1I). Collectively, these data show that NK cells upregulate LAG3 expression upon activation and that LAG3 expression is higher in less mature NK cells.

2.2. LAG3+ NK Cells Are More Activated and Metabolically Active

To reveal functional differences between LAG3+ and LAG3− NK cells, we next investigated their transcriptional profile by RNA sequencing. Splenic LAG3+ and LAG3− NK cells were FACS sorted 2 days after poly (I:C) injections (Figure 2A, sorting gating strategy shown in Figure S1A). At this time point, LAG3 was expressed on ∼50% of NK cells (Figure 1C; Figure S1B). To maximize the biological representation within each replicate, while also facilitating higher RNA yield, we pooled NK cells sorted from 10 spleens in each replicate. The principal component analysis revealed a clear separation between LAG3− and LAG3+ cells (Figure 2B).

FIGURE 2.

FIGURE 2

Open in a new tab

LAG3+ NK cells have a distinct transcriptomic profile. (A) 30 C57BL/6 mice were injected with Poly (I:C). Two days postinjection, LAG3‐CD11b+ and LAG3+CD11b+ splenic NK cells were sorted for RNA sequencing. (B) Principal component analysis. (C) Hallmarks pathways GSEA of differentially expressed genes. Selected terms are shown among the most significant ones.

Examination of the differentially expressed genes identified 581 upregulated genes in LAG3+ NK cells relative to their LAG3− counterparts, while 611 were downregulated (Log2 (Fold Change) >0.5 and FDR <0.05) (Figure 2C). We performed a gene set enrichment analysis (GSEA) using the hallmark gene sets to highlight differentially active pathways in NK cells that expressed or did not express LAG3 (Figure 2C). Most of the terms associating with genes enriched in LAG3+ NK cells referred to proliferation: “E2F targets”, “G2M checkpoints”, “Myc target V1 and V2”, and “Mitotic spindle”. One of the other enriched terms, “mTORC1 signalling” is also associated with NK cell growth and proliferation [27]. The downregulated genes in LAG3+ NK cells revealed an enrichment in the terms “apoptosis”, “p53 pathway” and cytokine signaling (Figure 2C). Altogether, these data show that LAG3+ NK cells present a signature of activated and proliferating cells.

2.3. LAG3+ NK Cells Display Higher Metabolic Activity and Proliferative Capacity

To validate the transcriptomic results, we first determined if activated NK cells presented higher expression of LAG3. Sca‐1 and CD69 are commonly used as markers of NK cell activation [28, 29]; hence, we co‐stained these two molecules with LAG3 on NK cells (Figure 3A). Both CD69 and Sca‐1 were rapidly upregulated on splenic NK cells upon poly(I:C) injection (Figure 3A, day 1). While CD69 expression rapidly declined at day 2, Sca‐1 was more stably expressed (Figure 3A). In accordance with our hypothesis, LAG3 induction on NK cells mirrored that of Sca‐1, suggesting that while the cytokine response caused by poly(I:C) treatment kept NK cells activated, LAG3 was also expressed (Figure 3A). Furthermore, Sca‐1+ NK cells presented a more robust upregulation of LAG3 expression (Figure 3B), thus confirming that LAG3 is more highly expressed by activated NK cells.

FIGURE 3.

FIGURE 3

Open in a new tab

LAG3 expression couples with more pronounced metabolic activity and proliferative capacity of NK cells. (A) Expression of Sca1+, LAG3+, and CD69+ on splenic NK cells upon poly (I:C) treatment (n = 4 for each day). (B) Co‐expression of LAG3 and Sca1 on splenic NK cells two days after poly (I:C) treatment (n = 5). (C‐G) Mice were injected with poly (I:C), and NK cells were gated as LAG3+ vs. LAG3−. Different parameters were analyzed (n = 4–6). (H) Splenic NK cells were stained with CTV and cultured with IL‐15 for 3 days. The percentage of divided NK cells for each individual was calculated using the proliferation modeling tool on the FlowJo software (n = 6, 3 experiments). (I) Splenic NK cells were stained with CTV and left in culture for 3 days with IL‐15. LAG3 expression is depicted in each of the CTV peaks. All the data were analyzed using a paired nonparametric Wilcoxon test. p‐values are indicated on each graph.

A second hallmark pathway found upregulated in LAG3+ NK cells was mTORC1 signaling (Figure 2C) that is required for NK cell functionality and reactivity [30, 31]. Hence, we measured the basal phosphorylation level of two proteins that act downstream mTORC1/2 (S6 and Akt [32]) in LAG3+ versus LAG3− NK cells. LAG3+ NK cells exhibited higher phosphorylation of both S6 and Akt (Figure 3C,D), indicating a stronger activity of the mTOR pathway. As mTOR plays a central role in NK cell metabolism [30], we examined the metabolic activity of LAG3+ NK cells by using two molecular probes, Mitotracker and CellRox, to, respectively, evaluate the global mitochondrial mass and the oxidative stress in NK cells. LAG3+ NK cells showed a significant increase in these markers (Figure 3E). We then analyzed the expression of other commonly used metabolic markers, such as the heavy chain of the system L amino acid transporter (CD98) and the transferrin receptor (CD71), whose expression is controlled by mTOR [33]. LAG3+ NK cells expressed higher levels of CD98 and CD71, supporting their higher metabolic demand (Figure 3F). Forward scatter (FSC) and Side scatter (SSC) measurements were consistently greater for LAG3+ NK cells compared with LAG3‐negative counterparts (Figure 3G), consistent with increased cell size and granularity that are additional indicators of metabolic activity in NK cells [33]. Taken together, all these results suggest that LAG3+ NK cells have a higher metabolic activity and demand.

Finally, the transcriptomic analysis indicated that LAG3+ NK cells have higher proliferative capacities (Figure 2C). To further support this observation, we analyzed the capacity of LAG3+ or LAG3− NK cells to proliferate in response to cytokines. Splenic NK cells labeled with Cell Trace Violet (CTV) were treated with IL‐15 for 72 h before assessing cell division via dilution of CTV intensity by flow cytometry. A greater proportion of LAG3+ NK cells than LAG3− NK cells diluted CTV signal (Figure 3H), indicative of increased proliferation of NK cells expressing LAG3. A link between LAG3 and proliferative capacity was also corroborated when LAG3 expression was assessed in NK cells at different “proliferation peaks” after 3 days of culture with IL‐15, with higher expression of LAG3 observed in proliferating cells (Figure 3I).

These findings validate the transcriptomic analysis and confirm that LAG3+ NK cells are more activated and have higher metabolic activity and proliferative capacity.

2.4. LAG3+ NK Cells Are Hyporesponsive and Exhibit Signatures of Functional Exhaustion

In T cells, LAG3 and other immune checkpoint receptors are upregulated upon activation and are associated with an exhaustion program mainly driven by the transcription factor TOX [34]. We thus conducted a GSEA using a published dataset [34] to examine whether LAG3+ NK cells resembled exhausted T cells. The transcriptional signature of the “TOX driven exhaustion” within intratumoral CD8+ T cells was strongly enriched in LAG3+ NK cells (Figure 4A). In addition to TOX, the transcriptional factors TCF‐1 and EGR2, also involved in CD8+ T cell exhaustion, were more expressed in LAG3+ NK cells (Figure 4B).

FIGURE 4.

FIGURE 4

Open in a new tab

LAG3+ NK cells exhibit an exhaustion signature and are hyporesponsive. (A) GSEA plot comparing LAG3+ and dysfunctional T cells. (B) Mice were injected with Poly (I:C) and sacrificed on Day 2. The MFI for TOX, TCF1, and Egr2 was determined by flow cytometry on LAG3+ vs. LAG3− splenic NK cells (n = 3–7). (C, D) Mice were injected intraperitoneally with Poly (I:C) and sacrificed at Day 2. Splenocytes were then stimulated for 4 h with plate‐bound NKp46 antibody (n = 30). All the data were analyzed using a paired nonparametric Wilcoxon test. (E) GSEA plot comparing LAG3+ and anergic NK cells. (F) Intracellular staining for IFN‐γ and the proportion of NK cells expressing CD107a was determined by immunostaining in LAG3− and LAG3+ splenic NK cells from MHC II‐deficient mice (n = 5).

Our transcriptomic and phenotypic analyses indicate that LAG3+ NK cells are activated yet present a transcriptional program similar to exhausted T cells. To more directly assess the functionality of LAG3+ NK cells, we analyzed their capacity to produce IFN‐γ and degranulate (surface‐exposed CD107a) in response to ex vivo stimulations. Upon stimulation with NKp46 antibodies, LAG3+ NK cells produced less IFN‐γ and exhibited decreased degranulation compared with LAG3− NK cells (Figure 4C,D). Of note, at the basal level (i.e., nonstimulated condition), LAG3+ NK cells displayed greater surface CD107a (Figure 4D), suggesting they may be more activated in vivo than LAG3‐ NK cells. The lower reactivity displayed by LAG3+ NK cells was also apparent in response to stimulation with agonist antibodies specific for NKR‐P1C or the cytokines IL‐12 and IL‐18 (Figure S2A,B). Because LAG3 expression is higher on immature NK cells, which are less responsive, we stratified functional responses by maturation stage. LAG3+ NK cells exhibited reduced responsiveness compared with LAG3− NK cells regardless of their maturation stage (Figure S2C). Taken together, these results indicate that NK cells expressing LAG3 are hyporesponsive despite their activation profile.

In addition, we observed an overlap of LAG3+ NK cells’ transcriptional profile and the signature of anergic human NK cells lacking self‐MHC I inhibitory receptors as described by Sabag et al. [35] (Figure 4E). Although the enrichment was less pronounced than the “TOX driven exhaustion” signature, the GSEA analysis revealed that key genes that are upregulated in anergic NK cells were also enriched in LAG3+ NK cells, further confirming their hyporesponsive state.

LAG3 has several ligands, MHC II being the most studied one. To determine if MHC II binding to LAG3 was required to induce dysfunction in NK cells, we took advantage of MHC II‐deficient mice. We treated MHC II‐deficient mice, or control littermate, with poly I:C and assessed the functionality of LAG3+ versus LAG3− splenic NK cells. We found that the dysfunctional phenotype was conserved in the absence of MHC II expression (Figure 4F), ruling out that LAG3‐MHC II interactions imposed NK cell dysfunction in LAG3+ NK cells.

2.5. LAG3 Expression Is a Shared Feature of NK Cell Activation

We next sought to ascertain whether LAG3 is a marker of NK cells with reduced functions in the context of viral infections and cancer. First, we took advantage of the murine cytomegalovirus model (MCMV), which has been extensively used to study the role of NK cells in viral infections [36, 37]. C57BL/6 mice were infected with MCMV (15,000 PFU) and sacrificed at D2, D4, D7, and D10. Unlike T cells, LAG3 expression was enhanced on splenic NK cells from mice infected with MCMV, in particular 2 and 4 days postinfection (Figure 5A). This was also true for liver NK cells (Figure S3A), while hepatic T cells expressed LAG3 with a delayed kinetics (Figure S3A). A higher percentage of NK cells expressed LAG3 within the activated Sca1+ population in both the spleen (Figure 5B) and the liver (Figure S3B). In contrast, similar LAG3 levels were expressed by Ly49H+ and Ly49H− NK cells (Figure S3C,D), suggesting that the inflammatory response, and not antigen recognition, was the primary driver of LAG3 expression.

FIGURE 5.

FIGURE 5

Open in a new tab

LAG3 expression is a shared feature of NK cell activation. (A, B) Mice were infected with MCMV, and expression of LAG3 on splenic NK and T cells was determined by flow cytometry (n = 3). (B) LAG3 expression was determined among Sca1+ or Sca1− splenic NK cells. (C) LAG3 expression on intratumoral NK in mice injected with AT3, 4T1, MCA205, B16‐BL6, C1498, RMA, or RMA‐S, or on intratumoral NK cells in the spleen or tumors from spontaneous cancer models (Eμ‐Myc models or NDL and PyMT). (D) Splenocytes from Eμ‐myc mice or intra‐tumoral NK cells from ectopic tumor model (RMA‐S) were stimulated for 4 h with plate‐bound NKp46 antibody. Intracellular staining for IFN‐γ was performed. Representative flow‐cytometry plot is shown (n = 3–5). (E) LAG3 expression among CD27+CD11b− (R1), CD27+CD11b+ (R2), and CD27‐CD11b+ (R3) in NK cells from Eμ‐myc mice (n = 5). Paired nonparametric Friedman test. (F) LAG3 expression among Sca1− or Sca1+ NK cells (from the spleen or Bone Marrow) from Eμ‐myc mice (n = 5). Paired nonparametric Wilcoxon test. (G) Splenic NK cells from Eμ‐myc mice were stained for the phospho‐epitope pS6 Ser235/236 or pAkt S473 (n = 13). Paired nonparametric Wilcoxon test.

We next investigated if NK cells upregulated LAG3 in models of cancer. LAG3 was abundantly expressed on tumor‐infiltrating NK cells in various ectopic or spontaneous murine tumor models, although with some degree of heterogeneity (Figure 5C). To further characterize LAG3+ intratumoral NK cells, we focused on the Eμ‐Myc tumor model, where mice spontaneously develop a B cell lymphoma in several organs (bone marrow, lymph nodes, spleen, and thymus), allowing us to recover a higher number of NK cells from tumors for subsequent analyses. LAG3+ NK cells were highly hyporesponsive in the Eμ‐Myc model (Figure 5D, left), an observation also corroborated in RMA‐S ectopic tumors (Figure 5D, right). Importantly, the profile of tumor infiltrating NK cells expressing LAG3 resembled those induced by poly (I:C) in terms of enrichment of an immature (CD27+ CD11b−, KLRG1−, DNAM1−) and activated (Sca1+) phenotype (Figures 5E,F and 2G,H) with higher mTOR signaling (pS6 and pAkt) activity (Figure 5G). Altogether, these results show that, in tumor models, LAG3 is more expressed by immature NK cells that are activated yet hyporesponsive and suggest that LAG3 could be used as a marker of dysfunctional NK cells.

2.6. LAG3 is Not Required for NK Cell Development nor Acute Killing of RMA‐S

The strong association between an activated, hyporesponsive phenotype and expression of LAG3 in NK cells prompted us to employ a genetic tool to mechanistically link LAG3 expression with NK cell dysfunction. To this end, we obtained Lag3fl/fl mice generated on a mixed C57BL/6 and 129 background [38] that were extensively backcrossed to fix critical components of the C57BL/6 NK receptor complex (NKC) located on chromosome 6. Of note, Lag3 is located close to the NKC on the same chromosome. We then crossed Lag3fl/fl mice with Ncr1‐iCre mice to generate a line lacking LAG3 expression selectively in NK cells (termed LAG3‐KO) that can be compared with Cre‐negative LAG3‐WT littermates (Figure 6A). To validate LAG3 deletion in this model, we injected mice of both genotypes with poly(I:C) and analyzed the expression of LAG3 on NK cells at Day 2. Poly(I:C) injection induced LAG3 expression exclusively on NK cells from LAG3 WT mice but not in LAG3 KO mice (Figure 6B), with levels comparable to those observed in C57BL/6 mice treated with poly(I:C) (Figure 1D). We then assessed potential defects in the NK cell compartment due to the absence of LAG3. LAG3‐KO mice presented a normal number of NK cells (NK1.1+NKp46+) in the spleen and bone marrow (Figure 6C) and no apparent defect in phenotypic maturation of NK cells (Figure 6D).

FIGURE 6.

FIGURE 6

Open in a new tab

LAG3 marks a subset of hyporesponsive NK cells but may not be the driver of the dysfunctional phenotype. (A) Schematic representation of the NCR1‐iCre LAG3‐flox model. (B) NCR1 Cre−/− LAG3 fl/fl (LAG3‐WT) and NCR1 Cre± LAG3 fl/fl (LAG3‐KO) were injected with Poly (I:C) and sacrificed at Day 2. LAG3 expression on splenic NK cells was determined by flow cytometry (n = 5). (C, D) NCR1 Cre−/− LAG3 fl/fl (LAG3‐WT) and NCR1 Cre± LAG3 fl/fl (LAG3‐KO) were sacrificed at 8–9 weeks old (n = 5 for each group). (C) The percentage of NK cells in the spleen and bone marrow was determined by flow cytometry. (D) The percentage of NK cells in each maturation subset in the spleen and bone marrow was determined by flow cytometry. (E) NCR1 Cre−/− LAG3 fl/fl (LAG3‐WT, n = 9) and NCR1 Cre± LAG3 fl/fl (LAG3‐KO, n = 9) were injected with Poly (I:C). 24 h later, they were injected with a 50:50 mixture of RMA/RMA‐S cells. 24 h later, the percentage of remaining RMA‐S and RMA cells in the peritoneal wash was determined by flow cytometry. Unpaired nonparametric (Mann–Whitney) test. (F, G) C57BL/6, NCR1 Cre−/− LAG3 fl/fl (LAG3‐WT) and NCR1 Cre± LAG3 fl/fl (LAG3‐KO) were injected with Poly (I:C) and sacrificed at Day 2 (n = 2 per group). LAG3 expression on splenic NK cells was determined by flow cytometry (F). Representative flow plots are depicted. (G) qPCR of Lag3 mRNA was performed on these NK cells.

To assess the role of LAG3 in NK cell antitumor activity, we performed in vivo killing assays in groups of mice treated with poly(I:C) to induce LAG3 expression prior to injection of a 50:50 mixture of RMA (NK cell resistant) and RMA‐S (NK cell sensitive) tumor cells. After 24 h, we analyzed the percentage of remaining RMA‐S cells relative to RMA cells as an indicator of NK cell cytotoxicity. In this setting, LAG3 KO mice cleared RMA‐S similar to littermate controls, indicating that LAG3 is dispensable for the acute killing of MHC‐deficient tumor cells (Figure 6E).

Unfortunately, an unexpected shift in LAG3 expression occurred after these initial studies in the LAG3 mouse colonies, which calls for caution in the use of this model. While LAG3 was properly upregulated in NK cells in Cre‐negative control mice when the colony was first established and evaluated (November 2020; Figure 6A–E), this response waned in the control mice over time. In subsequent experiments, we failed to observe an induction of LAG3 expression in Cre‐negative LAG3 WT littermate control NK cells after treatment of mice with poly(I:C), in contrast to continued expression of LAG3 on NK cells in poly(I:C)‐treated C57BL/6 controls (Figure 6F). Lack of LAG3 induction at the protein level was associated with a corresponding failure to upregulate Lag3 expression at the mRNA level (Figure 6G). Therefore, our littermate Ncr1‐Cre−/− Lag3fl/fl colony represents a poor control, and follow‐up experiments could not be performed.

3. Discussion

Immune cell dysfunction is a complex and multifaceted phenomenon that remains not fully understood. Therefore, it is important to define precise markers to identify subsets of exhausted leucocytes. Here, we focused on understanding if LAG3 could be used as a marker of dysfunctional NK cells.

LAG3 was initially found to be expressed by activated T and NK cells [39, 40], and it is a well‐established marker of CD8+ T cell exhaustion. Blocking LAG3 restores T cell function in cancer [41, 42, 43], highlighting its potential in not only driving but also maintaining T cell exhaustion. In our hands, LAG3 was expressed on NK cells upon activation, whereas a lower expression was observed in T cells. Since NK cells react quickly to stimulation, the rapid expression of LAG3 fits with their fast‐acting nature [44]. Other immune checkpoint receptors like PD‐1, TIGIT, and TIM3 were only modestly upregulated or absent on NK cells in inflammatory conditions and in cancer compared with LAG3. LAG3 expression was also observed in NK cells under chronic conditions, suggesting that this protein may be important in NK cells facing stimulation, whose fate may be akin to exhausted CD8+ T cells [12].

The role of LAG3 in NK cells remains elusive. We found that LAG3+ NK cells have a similar transcriptional signature to exhausted CD8+ T cells [34] coupled with an activated but hyporesponsive phenotype. In particular, LAG3+ NK cells resembled T cells undergoing early stages of exhaustion (precursors to exhausted CD8+ T cells or Tpex) [45]. LAG3+ NK cells share similarities with Tpex cells, including increased expression of TCF‐1 [46], elevated proliferation, greater mitochondrial mass, and activation of the mTOR pathway. This could reflect a unique dysfunctional phenotype in NK cells, distinct from canonically exhausted CD8+ T cells, due to NK cells having shorter lifespans, a lack of antigen specificity, and a rapid immune response activation [47, 48, 49]. However, studies have shown that T cell function and proliferation in response to stimulation are uncoupled [50], and overstimulation may trigger two separate exhaustion programs [46]. Like T cells [51], we proposed that NK cells expressing LAG3 might be attempting to balance effector function and prevent overstimulation by becoming hyporesponsive to simultaneously provide suboptimal control of harmful cells, but at the same time limiting the amount of tissue damage they may be causing [51]. Other immune checkpoints like CTLA‐4, TIGIT, and TIM3 also show context‐dependent expression on NK cells, but it remains unclear whether they simply also mark dysfunction or drive it. A direct assessment of the role of LAG3 in NK cells using a conditional genetic knock‐out model was complicated by an unexpected loss of LAG3 expression in the control littermates. This loss of gene expression could be due to unpredictable genetic changes near LAG3 on chromosome 6 arising from attempts to backcross to C57BL/6 background based on nearby NKC locus or to the known “leakiness” of the Ncr1 promoter during embryonic development which would lead to deletion of the LAG3 gene on NK cells in the control mice. However, another group found that LAG3‐deficient NK cells have increased proinflammatory functions and numbers, partially corroborating our findings [52]. On the other hand, culturing NK cells with LAG3 antibodies resulted in reduced glycolysis and mTOR signaling, but it is unclear if the antibody used in the assays acted as an agonist (as speculated by the authors) or an antagonist of LAG3 [53]. Still, another study suggested that immune checkpoints, including LAG3, are expressed on dysfunctional NK cells but are not the primary cause of functional impairment, indicating that other mechanisms contribute to NK cell dysfunction [54], and leaving the question of whether LAG3 marks or drives NK cell dysfunction still open.

Another outstanding question is the effect of LAG3 engagement on NK cells, and whether different ligands would play a different role in this putatively inhibitory axis. MHC class II‐LAG3 interactions suppress T cell responses in anticancer immunity [55]. Both MHC class II and FGL‐1 engagement triggered similar conformational changes in LAG3, leading to the same downstream signaling [56]. We have cursorily explored the effect of LAG3 engagement with MHC II molecules on NK cells using an MHC class II‐deficient mouse model. Even in the absence of MHC II, LAG3+ NK cells remained hyporesponsive to secondary stimulations, suggesting that LAG3‐MHC II engagement is not required to induce NK dysfunction.

In conclusion, in our study, we have identified LAG3 as a marker of NK cell dysfunction, offering a valuable tool for identifying dysfunctional NK cells in future studies and leading to a better understanding of the mechanisms underlying this phenomenon in different pathological conditions and immunotherapies. Surprisingly, we found that immune cells exhibiting reduced functionality, such as LAG3+ NK cells, are paradoxically more metabolically active, likely due to their overactivation. It will be important to follow up on these data, particularly in light of ongoing efforts to use LAG3 blocking reagents in the clinic [57].

4. Material and Methods

4.1. Animals

Mice were housed at the University of Ottawa Animal Care and Veterinary Services or the University of Calgary. C57BL/6 mice, NCR1‐GFP, MHC II KO, and Eμ‐Myc mice were purchased from Jackson Laboratories. Both male and female 8–15‐week‐old animals were used. Eμ‐Myc mice were sacrificed when they experienced cancer symptoms, including enlarged masses on the neck or lethargy, around 11–15 weeks. NCR1‐Cre LAG3‐flox mice were obtained from Dr. Stephen Waggoner, who initially obtained LAG3fl/fl mice from Dr. Dario Vignali (University of Pittsburgh) [38]. NDL mice were a kind gift from Dr. Luc Sabourin (OHRI). PyMT mice were provided by Dr. Pamela Ohashi (University of Toronto), who obtained them from Dr. William Muller (McGill) [58].

4.2. In Vivo Procedures

C57BL/6 mice were intraperitoneally injected with 200 µg of polyinosinic:polycytidylic acid (poly I:C) (Invivogen).

For s.c. and orthotopic injections, tumor cells resuspended in 100 µL RPMI were injected in the left flank or the mammary fat pad. For the in vivo killing assay, a mixture composed of 5 million RMA cells and 5 million RMA‐S cells was injected intraperitoneally. Twenty‐four hours later, we performed peritoneal washes and analyzed using flow cytometry the RMA‐S/RMA ratio as an indicator of NK cell cytotoxicity.

4.3. Tissue Processing

Collected tumors were cut into small pieces and dissociated using the GentleMacs (Miltenyi) in media supplemented with DNase I and collagenase IV for 45 min at 37°C. Cells were then homogenized using Fisherbrand Sterile Cell Strainers. Red blood cell lysis was performed using ammonium chloride‐potassium (ACK) lysing buffer.

Spleens were homogenized using Fisherbrand Sterile Cell Strainers and resuspended in PBS. Red blood cell lysis was performed using ACK lysing buffer.

4.4. NK Cell Isolation

NK cells were isolated from splenocytes using the Mouse NK cell Isolation Kit (StemCell Technologies EasySep) according to the manufacturer's instructions.

4.5. In Vitro Stimulation Assay

Isolated NK cells were cultured for 48 h at 200,000 cells/mL in the presence of different cytokines: 1000 IU/mL IFN‐α or IFN‐β (both from PBL Assay Sciences), 20 ng/mL of IL‐12 (Peprotech) + 100 ng/mL of IL‐18 (Peprotech), 100 ng/mL of IL‐15 (Peprotech), or 1000 IU/mL IL‐2 (NIH Biological Resources Branch Preclinical Repository).

4.6. Flow Cytometry Analysis

Cells were incubated with BD Mouse Fc Block Purified Rat Anti‐Mouse CD16/CD32 and then immunostained for 30 min at 4°C with the appropriate monoclonal antibodies (Table S1). For phosphoflow assays, cells were then incubated for 1 h in 5% Gibco RPMI at 37°C to preserve LAG3 extracellular staining. Intracellular staining of cytokines was performed with Cytofix/Cytoperm (BD Biosciences). Intracellular staining of transcription factors or cytotoxic molecules was performed using Foxp3 fixation/permeabilization buffer (eBioscience). Intracellular staining of phosphorylated proteins was performed with Lyse/Fix and Perm III buffers (BD Biosciences). Flow cytometric analysis was performed on LSR Fortessa 5L (BD). Fluorescence Minus One or IgG controls were used to set the gates, and data were analyzed with FlowJo 10.9.0 software.

4.7. Ex Vivo Functionality Assay

Isolated splenocytes were collected and stimulated using plate‐bound antibodies against NKp46, NKR‐P1C, and NKG2D or with Recombinant Murine IL‐12 (20 ng/mL) and Recombinant Murine IL‐18 (100 ng/mL) for 4 h at 37°C in the presence of the CD107a antibody and Golgi Stop and Plug (BD biosciences).

4.8. In Vitro Proliferation Assay

Isolated NK cells were stained with 0.6 µM Invitrogen CTV according to the manufacturer's instructions and then incubated with 100 ng/mL IL‐15 for 72 h at 37°C. The percentage of divided NK cells for each individual was calculated using the proliferation modeling tool on the FlowJo 10.9.0 software.

4.9. RNA Sequencing

NK cells were sorted as CD3‐CD19‐NK1.1+NKp46+DX5+LAG3+ or CD3‐CD19‐NK1.1+NKp46+DX5+LAG3− (Figure S1) using a Beckman Coulter MoFlo XDP.

RNA from sorted NK cells was isolated using the Sigma GenElute RNA Miniprep Kit. mRNA sequencing libraries were prepared according to the manufacturer's instructions for the TruSeq Stranded mRNA Library Prep kit (Illumina). Library preparation and sequencing were performed by DNA Link (Seoul, Republic of Korea). The quality of the libraries was verified by Tapestation (Agilent Technologies). Sequencing was performed using an Illumina NovaSeq 6000 system following the provided protocols for 2 × 100 sequencing. Transcript quantification for each sample was performed using Kallisto (v0.45.0) with the GRCm38 transcriptome reference and the ‐b 50 bootstrap option. The R package DESeq2 (v1.44.0) was then used to construct general linear models of each gene across experimental conditions. Wald's test was used to test for differential expression between groups, and the resultant p‐values were adjusted using the Benjamini–Hochberg false discovery rate method.

4.10. Epigenetic Analyses of LAG3 Locus

ATAC‐Seq/ChIP‐Seq were previously generated and mapped onto mouse genome build mm9, as described [59]. A genomic snapshot of the Lag3 locus was generated using IGV software (version 2.8.2) [60].

4.11. Statistical Analysis

Univariate statistical analyses were performed on GraphPad Prism. Unpaired or paired statistical tests were used as appropriate and indicated in each figure legend.

Author Contributions

Conceptualization, funding acquisition, and methodology: Marie Marotel and Michele Ardolino. Investigation: Valeria Vasilyeva, Marie Marotel, Olivia Makinson, Cynthia Chan, Maria Park, Colin O'Dwyer, Ayad Ali, Christiano Tanese de Souza, Mohamed S. Hasim, Sara Asif, Reem Kurdieh, John Abou‐Hamad, Edward Yakubovich, Jonathan Hodgins, Paul Al Haddad, Giuseppe Pietropaolo, Julija Mazej, Hobin Seo, Qiutong Huang, Giuseppe Sciumè, and Michele Ardolino. Resources: Giuseppe Sciumè, Stephen Waggoner, David Cook, and Nicolas Jacquelot. Marie Marotel, Michele Ardolino, Stephen Waggoner, Giuseppe Sciumè, Seung ‐ Hwan Lee and David Cook. Visualization: Valeria Vasilyeva, Colin O'Dwyer, Sarah Nersesian, Damien Chay, David Cook, and Marie Marotel. Writing — original draft: Valeria Vasilyeva, Marie Marotel, Sarah Nersesian, Stephen Waggoner, and Michele Ardolino. Writing — review and editing: All authors.

Conflicts of Interest

MA is a Scientific Advisory Board Member for Aakha Therapeutics and was under a contract agreement to perform sponsored research with Actym Therapeutics and Dragonfly Therapeutics. Neither consulting nor sponsored research is related to the present article.

Peer Review

The peer review history for this article is available at https://publons.com/publon/10.1002/eji.70009.

Supporting information

Supporting Figure 1: eji70009‐sup‐0001‐FigureS1.pdf

EJI-55-e70009-s001.pdf (507.2KB, pdf)

Supporting Figure 2: eji70009‐sup‐0002‐FigureS2.pdf

EJI-55-e70009-s005.pdf (69.1KB, pdf)

Supporting Figure 3: eji70009‐sup‐0003‐FigureS3.pdf

EJI-55-e70009-s002.pdf (213.6KB, pdf)

Supporting Figure 4: eji70009‐sup‐0004‐FigureS4.pdf

EJI-55-e70009-s004.pdf (63.3KB, pdf)

Supporting Figure 5: eji70009‐sup‐0005‐TableS1.pdf

EJI-55-e70009-s003.pdf (308.6KB, pdf)

Acknowledgments

We thank members of the Ardolino laboratory and all the authors for critically reading the manuscript. The OHRI and uOttawa flow‐cores for support with flow cytometry, and the ACVS facility at the University of Ottawa for support with animal studies. We are thankful to Dr. Vignali (University of Pittsburgh) for providing the LAG3 fl/fl mice, Dr. Ohashi (University of Toronto) for providing the PyMT mice, and Dr. Sabourin (OHRI) for providing the NDL mice.

Michele Ardolino and Marie Marotel contributed equally to this study.

Funding: M.A. is supported by CIHR and CRS; G.S. is funded by the Italian Association for Cancer Research (AIRC) IG‐28719. Research in the Jacquelot lab is supported by grants from Cancer Council NSW (RG 21‐05 to N.J.), Alberta Cancer Foundation/Arnie Charbonneau Cancer Institute laboratory start up package (to N.J.), Canadian Cancer Society Emerging Scholar Research Grant (grant #708072, to N.J.), Canada Foundation for Innovation ‐ John R. Evans Leaders Fund (#44762 to N.J), the Dr. Robert C. Westbury Fund for Melanoma Research (to N.J.), SSHRC Explore VPR Catalyst Grant (to N.J), Grant 1274078 from the Cancer Research Society and the Canadian Institutes of Health Research – Institute for cancer Research (to N.J), scholarships and fellowships from the University of Calgary Cumming School of Medicine Graduate Scholarship (to H.S.), Alberta Graduate Excellence Scholarships ‐ Master's Research (to H.S.), University of Calgary Eyes High postdoctoral fellowship (to Q.H.), and Canadian Cancer Society Research Training Award – PDF level (CCS award #708378, to Q.H.). M.M. is the recipient of a CAAIF and CIHR postdoctoral fellowship. GP: PNRR‐MAD‐2022‐12375947, Next Generation EU ‐ PNRR M6C2 ‐ Investimento 2.1 Valorizzazione e potenziamento della ricerca biomedica del SSN. JM: Marie Skłodowska‐Curie Actions (MSCA) Innovative Training Networks (ITN): H2020‐MSCA ITN‐2019 (grant agreement No 813343 J.M.).

Data Availability Statement

The RNA‐sequencing dataset is available as GEO: GSE284572.

References

  • 1. Mamessier E., Sylvain A., Thibult M., et al., “Human Breast Cancer Cells Enhance Self Tolerance by Promoting Evasion From NK Cell Antitumor Immunity,” Journal of Clinical Investigation 121 (2011): 3609–3622, 10.1172/JCI45816. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Peng Y., Zhu Y., Zhang J., et al., “Comprehensive Analysis of the Percentage of Surface Receptors and Cytotoxic Granules Positive Natural Killer Cells in Patients With Pancreatic Cancer, Gastric Cancer, and Colorectal Cancer,” Journal of translational medicine 11 (2013): 262, 10.1186/1479-5876-11-262. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Ardolino M., Azimi C. S., Iannello A., et al., “Cytokine Therapy Reverses NK Cell Anergy in MHC‐deficient Tumors,” Journal of Clinical Investigation 124 (2014): 4781–4794, 10.1172/JCI74337. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Seo H., Jeon I., Kim B., et al., “IL‐21‐mediated Reversal of NK Cell Exhaustion Facilitates Anti‐tumour Immunity in MHC Class I‐deficient Tumours,” Nature Communications 8 (2017): 15776, 10.1038/ncomms15776. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Ohs I., Ducimetière L., Marinho J., Kulig P., Becher B., and Tugues S., “Restoration of Natural Killer Cell Antimetastatic Activity by IL12 and Checkpoint Blockade,” Cancer Research 77 (2017): 7059–7071, 10.1158/0008-5472.CAN-17-1032. [DOI] [PubMed] [Google Scholar]
  • 6. Sun C., Xu J., Huang Q., et al., “High NKG2A Expression Contributes to NK Cell Exhaustion and Predicts a Poor Prognosis of Patients With Liver Cancer,” Oncoimmunology 6 (2017): e1264562, 10.1080/2162402X.2016.1264562. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Zhang Q., Bi J., Zheng X., et al., “Blockade of the Checkpoint Receptor TIGIT Prevents NK Cell Exhaustion and Elicits Potent Anti‐tumor Immunity,” Nature Immunology 19 (2018): 723–732, 10.1038/s41590-018-0132-0. [DOI] [PubMed] [Google Scholar]
  • 8. Mavilio D., Lombardo G., Benjamin J., et al., “Characterization of CD56‐/CD16+ Natural Killer (NK) Cells: A Highly Dysfunctional NK Subset Expanded in HIV‐infected Viremic Individuals,” PNAS 102 (2005): 2886–2891, 10.1073/pnas.0409872102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Peppa D., Micco L., Javaid A., et al., “Blockade of Immunosuppressive Cytokines Restores NK Cell Antiviral Function in Chronic hepatitis B Virus Infection,” Plos Pathogens 6 (2010): e1001227, 10.1371/journal.ppat.1001227. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Sun C., Fu B., Gao Y., et al., “TGF‐β1 Down‐regulation of NKG2D/DAP10 and 2B4/SAP Expression on human NK Cells Contributes to HBV Persistence,” Plos Pathogens 8 (2012): e1002594, 10.1371/journal.ppat.1002594. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Zhang C., Wang X., Li S., et al., “NKG2A is a NK Cell Exhaustion Checkpoint for HCV Persistence,” Nature Communications 10 (2019): 1507, 10.1038/s41467-019-09212-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Marotel M., Villard M., Drouillard A., et al., “Peripheral Natural Killer Cells in Chronic hepatitis B Patients Display Multiple Molecular Features of T Cell Exhaustion,” Elife 10 (2021): e60095, 10.7554/eLife.60095. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. McLane L. M., Abdel‐Hakeem M. S., and Wherry E. J., “CD8 T Cell Exhaustion During Chronic Viral Infection and Cancer,” Annual Review of Immunology 37 (2019): 457–495, 10.1146/annurev-immunol-041015-055318. [DOI] [PubMed] [Google Scholar]
  • 14. Hsu J., Hodgins J. J., Marathe M., et al., “Contribution of NK Cells to Immunotherapy Mediated by PD‐1/PD‐L1 Blockade,” Journal of Clinical Investigation 128 (2018): 4654–4668, 10.1172/JCI99317. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Hasim M. S., Marotel M., Hodgins J. J., et al., “When Killers Become Thieves: Trogocytosed PD‐1 Inhibits NK Cells in Cancer,” Science Advances 8 (2022): eabj3286, 10.1126/sciadv.abj3286. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Ruffo E., Wu R. C., Bruno T. C., Workman C. J., and Vignali D. A. A., “Lymphocyte‐Activation Gene 3 (LAG3): The next Immune Checkpoint Receptor,” Seminars in Immunology 42 (2019): 101305, 10.1016/j.smim.2019.101305. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Miyazaki T., Dierich A., Benoist C., and Mathis D., “Independent Modes of Natural Killing Distinguished in Mice Lacking Lag3,” Science 272 (1996): 405–408, 10.1126/science.272.5260.405. [DOI] [PubMed] [Google Scholar]
  • 18. Huard B., Tournier M., and Triebel F., “LAG‐3 Does Not Define a Specific Mode of Natural Killing in human,” Immunology Letters 61 (1998): 109–112. [DOI] [PubMed] [Google Scholar]
  • 19. Taborda N. A., Hernández J. C., Lajoie J., et al., “Short Communication: Low Expression of Activation and Inhibitory Molecules on NK Cells and CD4(+) T Cells Is Associated With Viral Control,” Aids Research and Human Retroviruses 31 (2015): 636–640, 10.1089/AID.2014.0325. [DOI] [PubMed] [Google Scholar]
  • 20. Huuhtanen J., Kasanen H., Peltola K., et al., “Single‐cell Characterization of Anti‐LAG‐3 and Anti‐PD‐1 Combination Treatment in Patients With Melanoma,” Journal of Clinical Investigation 133 (2023): e164809, 10.1172/JCI164809. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Narayanan S., et al. LAG3 is a Central Regulator of NK Cell Cytokine Production. bioRxiv, 2020.2001.2031.928200 (2020), 10.1101/2020.01.31.928200. [DOI]
  • 22. Judge S. J., Dunai C., Aguilar E. G., et al., “Minimal PD‐1 Expression in Mouse and Human NK Cells Under Diverse Conditions,” Journal of Clinical Investigation 130 (2020): 3051–3068, 10.1172/JCI133353. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Chiossone L., Chaix J., Fuseri N., Roth C., Vivier E., and Walzer T., “Maturation of Mouse NK Cells Is a 4‐Stage Developmental Program,” Blood 113 (2009): 5488–5496. [DOI] [PubMed] [Google Scholar]
  • 24. Huntington N. D., Tabarias H., Fairfax K., et al., “NK Cell Maturation and Peripheral Homeostasis Is Associated With KLRG1 Up‐regulation,” Journal of Immunology 178 (2007): 4764–4770, 10.4049/jimmunol.178.8.4764. [DOI] [PubMed] [Google Scholar]
  • 25. Martinet L., Ferrari De Andrade L., Guillerey C., et al., “DNAM‐1 Expression Marks an Alternative Program of NK Cell Maturation,” Cell reports 11 (2015): 85–97, 10.1016/j.celrep.2015.03.006. [DOI] [PubMed] [Google Scholar]
  • 26. Flommersfeld S., Böttcher J. P., Ersching J., et al., “Fate Mapping of Single NK Cells Identifies a Type 1 Innate Lymphoid‐Like Lineage That Bridges Innate and Adaptive Recognition of Viral Infection,” Immunity 54 (2021): 2288–2304.e2287, 10.1016/j.immuni.2021.08.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Yang C., Tsaih S., Lemke A., Flister M. J., Thakar M. S., and Malarkannan S., “mTORC1 and mTORC2 Differentially Promote Natural Killer Cell Development,” Elife 7 (2018): e35619, 10.7554/eLife.35619. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Fogel L. A., Sun M. M., Geurs T. L., Carayannopoulos L. N., and French A. R., “Markers of Nonselective and Specific NK Cell Activation,” Journal of Immunology 190 (2013): 6269–6276, 10.4049/jimmunol.1202533. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29. Borrego F., Robertson M. J., Ritz J., Peña J., and Solana R., “CD69 is a Stimulatory Receptor for Natural Killer Cell and Its Cytotoxic Effect is Blocked by CD94 Inhibitory Receptor,” Immunology 97 (1999): 159–165, 10.1046/j.1365-2567.1999.00738.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Donnelly R. P., Loftus R. M., Keating S. E., et al., “mTORC1‐Dependent Metabolic Reprogramming Is a Prerequisite for NK Cell Effector Function,” Journal of Immunology 193 (2014): 4477–4484, 10.4049/jimmunol.1401558. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Marçais A., Marotel M., Degouve S., et al., “High mTOR Activity Is a Hallmark of Reactive Natural Killer Cells and Amplifies Early Signaling Through Activating Receptors,” Elife 6 (2017): e26423, 10.7554/eLife.26423. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Huang K. and Fingar D. C., “Growing Knowledge of the mTOR Signaling Network,” Seminars in Cell & Developmental Biology 36 (2014): 79–90, 10.1016/j.semcdb.2014.09.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Marçais A., Cherfils‐Vicini J., Viant C., et al., “The Metabolic Checkpoint Kinase mTOR Is Essential for IL‐15 Signaling During the Development and Activation of NK Cells,” Nature Immunology 15 (2014): 749–757, 10.1038/ni.2936. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Scott A. C., Dündar F., Zumbo P., et al., “TOX Is a Critical Regulator of Tumour‐Specific T Cell Differentiation,” Nature 571 (2019): 270–274, 10.1038/s41586-019-1324-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Sabag B., Puthenveetil A., Levy M., et al., “Dysfunctional Natural Killer Cells Can be Reprogrammed to Regain Anti‐Tumor Activity,” Embo Journal 43 (2024): 2552–2581, 10.1038/s44318-024-00094-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Arase H., Mocarski E. S., Campbell A. E., Hill A. B., and Lanier L. L., “Direct Recognition of cytomegalovirus by Activating and Inhibitory NK Cell Receptors,” Science 296 (2002): 1323–1326, 10.1126/science.1070884. [DOI] [PubMed] [Google Scholar]
  • 37. Adams N. M., Grassmann S., and Sun J. C., “Clonal Expansion of Innate and Adaptive Lymphocytes,” Nature Reviews Immunology 20 (2020): 694–707, 10.1038/s41577-020-0307-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Zhang Q., Chikina M., Szymczak‐Workman A. L., et al., “LAG3 limits Regulatory T Cell Proliferation and Function in Autoimmune Diabetes,” Science Immunology 2 (2017): eaah4569, 10.1126/sciimmunol.aah4569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Triebel F., Jitsukawa S., Baixeras E., et al., “LAG‐3, a Novel Lymphocyte Activation Gene Closely Related to CD4,” Journal of Experimental Medicine 171 (1990): 1393–1405, 10.1084/jem.171.5.1393. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Baixeras E., Huard B., Miossec C., et al., “Characterization of the Lymphocyte Activation Gene 3‐Encoded Protein. A New Ligand for human Leukocyte Antigen Class II Antigens,” Journal of Experimental Medicine 176 (1992): 327–337, 10.1084/jem.176.2.327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Blackburn S. D., Shin H., Haining W. N., et al., “Coregulation of CD8+ T Cell Exhaustion by Multiple Inhibitory Receptors During Chronic Viral Infection,” Nature Immunology 10 (2009): 29–37, 10.1038/ni.1679. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42. Woo S., Turnis M. E., Goldberg M. V., et al., “Immune Inhibitory Molecules LAG‐3 and PD‐1 Synergistically Regulate T‐Cell Function to Promote Tumoral Immune Escape,” Cancer Research 72 (2012): 917–927, 10.1158/0008-5472.CAN-11-1620. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Wherry E. J., “T Cell Exhaustion,” Nature Immunology 12 (2011): 492–499, 10.1038/ni.2035. [DOI] [PubMed] [Google Scholar]
  • 44. Joncker N. T., Shifrin N., Delebecque F., and Raulet D. H., “Mature Natural Killer Cells Reset Their Responsiveness When Exposed to an Altered MHC Environment,” Journal of Experimental Medicine 207 (2010): 2065–2072, 10.1084/jem.20100570. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45. Gabriel S. S., Tsui C., Chisanga D., et al., “Transforming Growth Factor‐β‐Regulated mTOR Activity Preserves Cellular Metabolism to Maintain Long‐Term T Cell Responses in Chronic Infection,” Immunity 54 (2021): 1698–1714.e1695, 10.1016/j.immuni.2021.06.007. [DOI] [PubMed] [Google Scholar]
  • 46. Philip M. and Schietinger A., “CD8(+) T Cell Differentiation and Dysfunction in Cancer,” Nature Reviews Immunology 22 (2022): 209–223, 10.1038/s41577-021-00574-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47. Vivier E., Tomasello E., Baratin M., Walzer T., and Ugolini S., “Functions of Natural Killer Cells,” Nature Immunology 9 (2008): 503–510, 10.1038/ni1582. [DOI] [PubMed] [Google Scholar]
  • 48. Hodgins J. J., Khan S. T., Park M. M., Auer R. C., and Ardolino M., “Killers 2.0: NK Cell Therapies at the Forefront of Cancer Control,” Journal of Clinical Investigation 129 (2019): 3499–3510, 10.1172/JCI129338. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49. Bi J. and Tian Z., “NK Cell Exhaustion,” Frontiers in Immunology 8 (2017): 760, 10.3389/fimmu.2017.00760. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50. Utzschneider D. T., Charmoy M., Chennupati V., et al., “T Cell Factor 1‐Expressing Memory‐Like CD8(+) T Cells Sustain the Immune Response to Chronic Viral Infections,” Immunity 45 (2016): 415–427, 10.1016/j.immuni.2016.07.021. [DOI] [PubMed] [Google Scholar]
  • 51. Freel S. A., Saunders K. O., and Tomaras G. D., “CD8(+)T‐Cell‐Mediated Control of HIV‐1 and SIV Infection,” Immunologic Research 49 (2011): 135–146, 10.1007/s12026-010-8177-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52. Nicosia M., Fan R., Lee J., et al., “LAG3 regulates Antibody Responses in a Murine Model of Kidney Transplantation,” Journal of Clinical Investigation 135 (2025): e172988, 10.1172/JCI172988. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53. Ge H., Guo N., Liu Y., et al., “The Inhibitory Receptor LAG3 Affects NK Cell IFN‐γ Production Through Glycolysis and the PSAT1/STAT1/IFNG Pathway,” MBio 16 (2025): e0023025, 10.1128/mbio.00230-25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54. Pouxvielh K., Marotel M., Drouillard A., et al., “Tumor‐induced Natural Killer Cell Dysfunction Is a Rapid and Reversible Process Uncoupled From the Expression of Immune Checkpoints,” Science Advances 10 (2024): eadn0164, 10.1126/sciadv.adn0164. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55. Maruhashi T., Sugiura D., Okazaki I., et al., “Binding of LAG‐3 to Stable Peptide‐MHC Class II Limits T Cell Function and Suppresses Autoimmunity and Anti‐cancer Immunity,” Immunity 55 (2022): 912–924.e918, 10.1016/j.immuni.2022.03.013. [DOI] [PubMed] [Google Scholar]
  • 56. Jiang Y., Dai A., Huang Y., et al., “Ligand‐induced Ubiquitination Unleashes LAG3 Immune Checkpoint Function by Hindering Membrane Sequestration of Signaling Motifs,” Cell 188 (2025): 2354–2371.e2318, 10.1016/j.cell.2025.02.014. [DOI] [PubMed] [Google Scholar]
  • 57. Tawbi H. A., Schadendorf D., Lipson E. J., et al., “Relatlimab and Nivolumab versus Nivolumab in Untreated Advanced Melanoma,” New England Journal of Medicine 386 (2022): 24–34, 10.1056/NEJMoa2109970. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58. Guy C. T., Cardiff R. D., and Muller W. J., “Induction of Mammary Tumors by Expression of Polyomavirus Middle T Oncogene: A Transgenic Mouse Model for Metastatic Disease,” Molecular and Cellular Biology 12 (1992): 954–961, 10.1128/mcb.12.3.954-961.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59. Sciumè G., Mikami Y., Jankovic D., et al., “Rapid Enhancer Remodeling and Transcription Factor Repurposing Enable High Magnitude Gene Induction Upon Acute Activation of NK Cells,” Immunity 53 (2020): 745–758 e744, 10.1016/j.immuni.2020.09.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60. Robinson J. T., Thorvaldsdóttir H., Winckler W., et al., “Integrative Genomics Viewer,” Nature Biotechnology 29 (2011): 24–26, 10.1038/nbt.1754. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Figure 1: eji70009‐sup‐0001‐FigureS1.pdf

EJI-55-e70009-s001.pdf (507.2KB, pdf)

Supporting Figure 2: eji70009‐sup‐0002‐FigureS2.pdf

EJI-55-e70009-s005.pdf (69.1KB, pdf)

Supporting Figure 3: eji70009‐sup‐0003‐FigureS3.pdf

EJI-55-e70009-s002.pdf (213.6KB, pdf)

Supporting Figure 4: eji70009‐sup‐0004‐FigureS4.pdf

EJI-55-e70009-s004.pdf (63.3KB, pdf)

Supporting Figure 5: eji70009‐sup‐0005‐TableS1.pdf

EJI-55-e70009-s003.pdf (308.6KB, pdf)

Data Availability Statement

The RNA‐sequencing dataset is available as GEO: GSE284572.

Articles from European Journal of Immunology are provided here courtesy of Wiley

RESOURCES