Dependence of innate lymphoid cell 1 development on NKp46

NKp46, a natural killer (NK) cell–activating receptor, is involved in NK cell cytotoxicity against virus-infected cells or tumor cells. However, the role of NKp46 in other NKp46+ non-NK innate lymphoid cell (ILC) populations has not yet been characterized. Here, an NKp46 deficiency model of natural cytotoxicity receptor 1 (Ncr1)gfp/gfp and Ncr1gfp/+ mice, i.e., homozygous and heterozygous knockout (KO), was used to explore the role of NKp46 in regulating the development of the NKp46+ ILCs. Surprisingly, our studies demonstrated that homozygous NKp46 deficiency resulted in a nearly complete depletion of the ILC1 subset (ILC1) of group 1 ILCs, and heterozygote KO decreased the number of cells in the ILC1 subset. Moreover, transplantation studies confirmed that ILC1 development depends on NKp46 and that the dependency is cell intrinsic. Interestingly, however, the cell depletion specifically occurred in the ILC1 subset but not in the other ILCs, including ILC2s, ILC3s, and NK cells. Thus, our studies reveal that NKp46 selectively participates in the regulation of ILC1 development.


Author summary
Group 1 innate lymphoid cells (ILCs) comprise two subsets: natural killer (NK) cells and ILC1s. Although NK cells and ILC1s are functionally distinct, a factor that regulates one subset but not the other has not been identified. In the current study, we discovered that NKp46, a marker expressed by both NK cells and ILC1s, is critical for the development of ILC1s but not NK cells. In mice lacking NKp46, and in wild-type (WT) mice depleted of immune cells by irradiation and then transplanted with bone marrow (BM) cells lacking NKp46, ILC1s that express cell surface receptor tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL) were almost completely absent in all organs or tissues examined, including liver, spleen, BM, and small intestine (SI). In contrast, the cell number and signature cytokine expression of all other ILC subsets-namely NK cells, ILC2s, and ILC3s-were not significantly affected. Collectively, our findings provide new evidence supporting an essential role for NKp46 in the development of ILC1s. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111

Introduction
Natural cytotoxicity receptor NKp46, encoded by the Ncr1 gene, is a natural killer (NK) cell-activating receptor that plays roles in regulating the NK cell's clearance of virus and rejection of tumor [1]. Following binding to its putative ligands, the receptor activates intracellular signaling through immune-receptor tyrosine-based activating motifs (ITAMs) [2]. Some non-NK innate lymphoid cell (ILC) populations also express NKp46, including the ILC1 subset (Lin ─ NKp46 + NK1.1 + CD49b ─ CD49a + ) [3] of group 1 ILCs and the ILC3 subset (Lin ─ CD127 + RORγt + ) of group 3 ILCs [4]. However, the role of NKp46 in these non-NK ILCs is still poorly understood. We previously reported that NKp46 defines a subset of NKT cells susceptible to malignant transformation in the presence of interleukin 15 (IL-15) and has a role in the NK cell clearance of herpes simplex virus 1 [5,6]. In the current study, we aimed to unravel the role of NKp46 in regulating the development and function of NKp46 + ILCs, especially ILC1s, using a genetic approach.

ILC1s are absent in NKp46-deficient mice
An NKp46 knockout (KO) mouse model-in which Ncr1, the gene encoding NKp46, was replaced with green fluorescent protein (gfp) (Ncr1 gfp/gfp ) [5,7]-was used in this study. The development of ILC populations was assessed in different organs and tissues comparing wild-type (WT) (Ncr1 +/+ ), heterozygous (Ncr1 gfp/+ ), and KO (Ncr1 gfp/gfp ) mice. The gating strategy for ILC1 flow cytometric analysis is shown in Fig 1A. A clear NK1.1 + NKp46 ─ population (more than 50%) was observed within the NK1.1 + population in the liver (S1A Fig), consistent with previous studies [8,9]. The intensity of CD49a surface expression was higher in the NK1.1 + NKp46 + CD49b ─ CD49a + ILC1 population than that in the NK1.1 + NKp46 ─ CD49b ─ CD49a + population, and the latter population displayed a phenotype of CD3 ─/dim (S1 and S2 Figs). Dadi and colleagues recently defined an ILC1-like population, named T cell receptor (TCR) lineage type 1 innate-like T cells (ILTC1), with a phenotype of NK1.1 + NKp46 ─ CD49a + [10]. Moreover, NKp46 is considered a reliable marker for ILC1s and NK cells [9][10][11]. Thus, in the current study, the ILC1 subset was defined as NK1.1 + NKp46 + (or GFP + for KO mice)CD49b ─ CD49a + (Fig 1A), which does not include the NK1.1 + NKp46 ─ CD49b ─ CD49a + population. It was surprising that, when compared to Ncr1 +/+ littermates, proportionally the ILC1 subset (CD49b ─ CD49a + ) was nearly completely absent in the liver of Ncr1 gfp/gfp mice and was significantly decreased in the liver of Ncr1 gfp/+ mice ( , while the percentage of the NK1.1 + NKp46 ─ (or GFP ─ for KO mice)CD49b ─ CD49a + population seems to be no different among the Ncr1 gfp/gfp , Ncr1 +/gfp , and Ncr1 +/+ groups (S2 Fig, bottom panel). Furthermore, quantification of the ILC1 subset showed that the absolute cell quantity was also drastically reduced in the liver of the Ncr1 gfp/gfp mice and moderately reduced in the liver of Ncr1 gfp/+ mice compared to their Ncr1 +/+ littermates (Fig 1B, lower panel). We also used markers including CD62L, Eomesodermin (Eomes), and T-box expressed in T cells (T-bet) to confirm our observation regarding dependency on NKp46 for ILC1 development by comparing Ncr1 gfp/gfp mice and/or Ncr1 +/gfp mice to Ncr1 +/+ mice (S3 and S4 Figs). The development of ILC populations was also assessed in other organs and tissues using the liver as our point of reference and comparing results in WT (Ncr1 +/+ ) mice to Ncr1 gfp/gfp mice (Fig 1C and 1D). Cell quantification by flow cytometry indicated that ILC1s (CD49b ─ CD49a + ) were nearly completely absent in the bone marrow (BM), spleen, and small intestine (SI) of Ncr1 gfp/gfp mice ( Fig 1E). In contrast, the absolute number of NK cells, which are the main population that expresses NKp46 in tested organs or tissues, did not significantly change in Ncr1 gfp/gfp mice compared to their Ncr1 +/+ littermates, consistent with a previous report [7] (Fig 1E, lower panel). decision to publish, or preparation of the manuscript. NCI https://projectreporter.nih.gov/ project_info_details.cfm?aid=9388328&map=y (grant number CA185301). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. NCI https://projectreporter.nih.gov/ project_info_details.cfm?aid=9186831&map=y (grant number CA210087). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. NCI https://projectreporter.nih.gov/ project_info_details.cfm?aid=9243984&map=y (grant number CA068458). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests:
The authors have declared that no competing interests exist.

Ncr1 gfp/gfp mice lack tumor necrosis factor (TNF)-related apoptosisinducing ligand (TRAIL) + ILC1s
TNF-related apoptosis-inducing ligand (TRAIL) is a functional protein, selectively expressed on the ILC1 subset, that plays an essential role in mediating the cytotoxicity of this population against target cells through triggering the death receptor-transduced signaling pathway [12]. The expression of TRAIL and lack of CD49b expression can also be used to distinguish ILC1s from NK cells [13] because resting NK cells in WT mice do not express TRAIL, while ILC1s do (Fig 2A, left). Consistent with the data shown in Fig 1 regarding the near-complete lack of CD49a + CD49b ─ ILC1s, the TRAIL + CD49b ─ population of NKp46 + (or GFP + )NK1.1 + type I ILCs was barely detectable in the liver (Fig 2A and 2B) and other organs (Fig 2C) of Ncr1 gfp/gfp mice compared to their Ncr1 +/+ littermate controls.

NKp46 deficiency has no substantial adverse effect on ILC2 and ILC3 subsets
Due to our observation that NKp46 deficiency restricted the development of ILC1s, we next set out to test whether this effect also occurred in other ILC subsets. However, using the gating strategy as described in Fig 3A, we did not observe a significant difference in the quantities or frequencies of Lin ─ CD127 + Gata3 + ILC2s and Lin ─ CD127 + RORγt + ILC3s when Ncr1 gfp/gfp mice were compared to their Ncr1 +/+ littermate controls (Fig 3B and 3C). There was an insufficient quantity of ILC1 cells from Ncr1 gfp/gfp mice to study how this NKp46 deficiency affects ILC1 function(s); however, we did observe that the interferon γ (IFN-γ) production by NK cells in response to co-stimulation with IL-12 and IL-18 was unaltered between cells isolated from Ncr1 gfp/gfp mice versus those from Ncr1 +/+ littermate controls (Fig 3D). Likewise, IL-22 production by ILC3 cells isolated from Ncr1 gfp/gfp mice versus those from Ncr1 +/+ littermate controls was not significantly different (Fig 3E). Consistent with our results, Satoh-Takayama and colleagues previously demonstrated that NKp46 is not required for IL-22-mediated intestinal innate immune cells in the gut to defend against Citrobacter rodentium [14]. Together, these results suggest that NKp46 does not control homeostasis or signature ILC cytokine production of ILC2s, NK cells, or ILC3s, but does selectively participate in the regulation of ILC1 development.

NKp46 controls ILC1 development in a cell-intrinsic manner
Although NK cells and ILC1s are closely related, ILC1s do not develop through Lin ─ CD122 + NK1.1 ─ DX5 ─ NK cell precursors (NKPs) but can develop through Lin ─ c-kit low α4β7 + CD127 + CD25 ─ Flt3 ─ common helper innate lymphoid precursors (CHILPs). Both NKPs and CHILPs are derived from a common lymphoid progenitor (CLP) [15]. antibody. We found that there was a nearly complete absence of ILC1s in the liver, spleen, and BM of WT recipients when Ncr1 gfp/gfp BM cells were used as donor cells. However, the ILC1 population was present in significantly larger quantities when Ncr1 +/+ BM cells were used as donor cells (Fig 4B and 4C and S6 Fig). In contrast, the quantity of ILC2 cells and ILC3 cells in the SI was not affected in recipient mice engrafted with Ncr1 gfp/gfp or Ncr1 +/+ BM donor cells (Fig 4D and 4E). The moderate increase in the proportion of NK cells could be occurring due to the complete lack of ILC1s among NKp46 + NK1.1 + type I ILCs (Fig 4B top right; Fig 4C bottom right). Collectively, these results further confirm that ILC1 development depends on NKp46, and this dependency is cell-intrinsic (Fig 1).
In conclusion, our findings provide novel evidence that NKp46 plays a critical role in ILC1 development. Previous studies in this area focused on transcriptional control of ILC development. It is known that T-bet and Eomes regulate NK cell development, Gata3 controls ILC2 development, and RORγt defines the ILC3 lineage [15]. Several transcription factors-such as nuclear factor interleukin 3 regulated (Nfil3), runt related transcription factor 3 (Runx3), and T-bet-have been found to control ILC1 development; however, these factors do not play a selective role in determining ILC1 development [15]. That is, these transcription factors have some overlapping roles in several types of ILCs and thus individually cannot determine the fate of ILC1 development. For example, T-bet has been shown to play a role not only in ILC1 development but also in NK cell and ILC3 development [15]. Here, we identified a receptor that selectively determines the developmental fate of ILC1s. Our study also supports the notion that ILC1s and NK cells belong to different lineages, although both belong to group 1 ILCs, and both can produce IFN-γ when activated (e.g., by cytokines). Our study also shows that Ncr1 gfp/gfp mice may serve as a useful animal model for investigating the physiological or pathological functions of ILC1s, given their near-complete absence in various organs, while the development and function of other ILCs are kept nearly intact, with the exception of functions related to NKp46's role in NK cells [7,16].

Ethics statement
All animal experiments were performed according to the protocol (# 2012A00000090), which has been approved by The Ohio State University Institutional Animal Care and Use Committee (IACUC). No human subjects were involved in this study.

Enzyme-linked immunosorbent assay (ELISA)
1 × 10 5 NK cells were seeded into a 96-well plate and cultured with or without IL-12 plus IL-18 for 24 h. Supernatants were then harvested to detect IFN-γ production, which was assessed by the Mouse IFN-gamma Uncoated ELISA Kit (Catalog #88-7314-88, Invitrogen)

Statistics
Student's t test or paired t test was used to analyze two independent or paired groups, respectively. A p value less than 0.05 was considered statistically significant.