The Inhibitory Receptor GPR56 (Adgrg1) Is Specifically Expressed by Tissue-Resident Memory T Cells in Mice But Dispensable for Their Differentiation and Function In Vivo

Tissue-resident memory T (TRM) cells with potent antiviral and antibacterial functions protect the epithelial and mucosal surfaces of our bodies against infection with pathogens. The strong proinflammatory activities of TRM cells suggest requirement for expression of inhibitory molecules to restrain these memory T cells under steady state conditions. We previously identified the adhesion G protein-coupled receptor GPR56 as an inhibitory receptor of human cytotoxic lymphocytes that regulates their cytotoxic effector functions. Here, we explored the expression pattern, expression regulation, and function of GPR56 on pathogen-specific CD8+ T cells using two infection models. We observed that GPR56 is expressed on TRM cells during acute infection and is upregulated by the TRM cell-inducing cytokine TGF-β and the TRM cell-associated transcription factor Hobit. However, GPR56 appeared dispensable for CD8+ T-cell differentiation and function upon acute infection with LCMV as well as Listeria monocytogenes. Thus, TRM cells specifically acquire the inhibitory receptor GPR56, but the impact of this receptor on TRM cells after acute infection does not appear essential to regulate effector functions of TRM cells.


Introduction
Cytotoxic CD8 + T cells protect the body against pathogenic viruses and intracellular bacteria through the targeted production of cytolytic enzymes and inflammatory cytokines to infected cells and other immune cells, respectively. Cytotoxic CD8 + T cells are activated upon recognition of peptide antigens derived from viruses or bacteria in the context of MHC class I molecules through their T-cell receptors. These CD8 + T cells are also equipped with a comprehensive repertoire of inhibitory receptors that maintain quiescence at steady state but still allow rapid activation upon pathogen encounter [1]. Given that cytotoxic CD8 + T cells are also important for the clearance of tumor cells, current immunotherapies aim to exploit these cells for the benefit of cancer patients. In particular, tuning the balance between inhibition and activation of CD8 + T cells through blockade of immune checkpoints, such as PD-1, appears to be a promising strategy for the treatment of cancer patients [2].
We and others previously identified GPR56 as a surrogate surface marker indicating cytotoxic capacity across human lymphocyte subsets, including CD8 + and CD4 + T cells as well as NK cells [3][4][5][6]. GPR56 is an adhesion G protein-coupled receptor (GPCR) with established roles in brain development, hematopoiesis, male fertility, muscle hypertrophy, and tumor growth and progression [7][8][9]. We obtained evidence that GPR56 negatively

Lymphocytic Choriomeningitis Virus (LCMV) and Listeria Monocytogenes Bacterial Infection
Mice were infected intraperitoneally with 30 plaque-forming units (PFU) of the LCMV strain LCMV-WE or 1 × 10 5 PFU of the LCMV strain LCMV-Armstrong or were infected by oral administration with 2.5 × 10 9 colony-forming unit (CFU) of recombinant Listeria monocytogenes expressing OVA (Listeria-OVA) InlA M (kindly provided by B. Sheridan, Stony Brook University). For rechallenge responses, mice that had been orally infected with 2.5 × 10 9 Listeria-OVA InlA M were reinfected 30 days later with a second dose of 2.5 × 10 10 Listeria-OVA InlA M . At the indicated time points after infection, mice were sacrificed and organs were collected for analysis of CD8 + T-cell responses.

In Vitro CD8 T Cell Stimulation
Murine CD8 + T cells isolated from spleen and small intestine were activated in 96-well plates (Corning, Corning, NY, USA), coated with 10 µg/mL anti-CD3 (clone 145 2C11; BD Bioscience) and 2 µg/mL anti-CD28 (Clone 37.51; BD Bioscience), in the presence of 10 ng/mL IL-2, 10 ng/mL IL-7, and 10 ng/mL IL-15 with or without 10 ng/mL TGF-β (all from PeproTech, London, UK). After 3 days of culture, CD8 + T cells were replated and cultured for 6 additional days with only IL-2, IL-7, and IL-15. For the peptide stimulation assay, GP 33-41 -specific CD8 + T cells were activated in 96-wells plates in the presence of 5 µg/mL of the GP 33-41 peptide KAVYNFATC for 5 hr. The anti-CD107a (clone 1D4B) antibody was added to assess degranulation. Brefeldin A and Monensin (both from eBioscience) were added to enable intracellular capture of IFN-γ and Cytofix Fixation Buffer (BD Bioscience) was used for staining of intracellular cytokines.

Statistical Analysis
All analysis was performed in GraphPad Prism 9 (GraphPad Software, San Diego, CA, USA) using one-way or two-way ANOVA test and Tukey's multiple comparisons test.

CD8 + T RM Cells Specifically Upregulate Adgrg1/GPR56
To establish the expression profile of adhesion GPCRs in pathogen-specific T cells, we analyzed RNA sequencing data of memory CD8 + T cells that arose after herpes simplex virus (HSV) or lymphocytic choriomeningitis virus (LCMV) infection in mice. We observed that T cells only expressed 4 out of a panel of 31 adhesion GPCRs, including Adgre5/CD97 and the gene cluster Adgrg1/GPR56, Adgrg3/GPR97, and Adgrg5/GPR114. Adgre5 and Adgrg5 expression was upregulated in naïve CD8 + T cells and was retained in all analyzed memory CD8 + T-cell lineages ( Figure 1A). In contrast, Adgrg1 and Adgrg3 tran-scripts were not expressed in naïve CD8 + T cells but were upregulated in virus-specific CD8 + T cells after HSV or LCMV infection ( Figure 1A). Upregulation of Adgrg3 was most pronounced in HSV-specific T RM cells in skin, whereas expression of Adgrg1 was strongly upregulated in HSV-specific T RM cells of the skin and in LCMV-specific T RM cells of the liver and the small intestine ( Figure 1A). We did not find expression of Adgrg1 in circulating memory CD8 + T cells, including T CM -and T EM -cell populations in the spleen and the liver ( Figure 1A), indicating that this adhesion GPCR was specifically induced in pathogenspecific T RM cells. The quantitative RT-PCR analysis validated the specific upregulation of Adgrg1 in virus-specific T RM cells after acute infection with LCMV ( Figure 1B). The T RM cell-specific upregulation of Adgrg1 appeared to occur in the memory phase, given that we were unable to detect expression of Adgrg1 in effector populations, including terminal effector cells (TECs) and memory precursor effector cells (MPECs) at day 8 after infection with LCMV ( Figure 1B). Interestingly, T RM cells also specifically express high levels of the transcription factor Hobit and the cytolytic enzyme granzyme B, in contrast to circulating T CM and T EM cells [25,28]. Therefore, Adgrg1/GPR56 expression appears to correlate with the expression of Zfp683/Hobit and the presence of granzyme B protein in murine as well as human CD8 + T cells.

CD8 + T RM Cell-Inducing Factors Regulate Adgrg1/GPR56 Expression
We previously showed that Hobit and Blimp-1 strongly collaborate in the transcriptional regulation of T RM cells [28] and that Hobit is the driving transcription factor for the expression of granzyme B in T RM cells [25]. The analysis of Adgrg1 expression in T RM cells of Zfp683-deficient and Prdm1-deficient mice showed strong reduction of Adgrg1 expression in mouse T RM cells lacking Hobit and, to a lesser extent, in those lacking Blimp-1 (Figure 2A,B), in support of a role for Hobit in the transcriptional regulation of GPR56. The expression of Adgrg1 was further reduced in Zfp683 and Prdm1 double-deficient T RM cells ( Figure 2B), indicating coregulation of Hobit and Blimp-1 in the maintenance of Adgrg1 expression in T RM cells. It is important to note that the reduced numbers of T RM cells in Hobit and Blimp-1 double-deficient mice may also impact Adgrg1 expression. However, the partial defects in GPR56 expression in single Hobit and Blimp-1 deficient mice that have normal numbers of T RM in small intestine suggest a direct impact of these transcription factors on the regulation of GPR56 expression.  Figure S1. One-way ANOVA with Tukey's multiple comparisons test; ** p < 0.01, *** p < 0.005.

CD8 + TRM Cell-Inducing Factors Regulate Adgrg1/GPR56 Expression
We previously showed that Hobit and Blimp-1 strongly collaborate in the transcriptional regulation of TRM cells [28] and that Hobit is the driving transcription factor for the expression of granzyme B in TRM cells [25]. The analysis of Adgrg1 expression in TRM cells of Zfp683-deficient and Prdm1-deficient mice showed strong reduction of Adgrg1 expression in mouse TRM cells lacking Hobit and, to a lesser extent, in those lacking Blimp-1 (Figure 2A,B), in support of a role for Hobit in the transcriptional regulation of GPR56. The expression of Adgrg1 was further reduced in Zfp683 and Prdm1 double-deficient TRM cells ( Figure 2B), indicating coregulation of Hobit and Blimp-1 in the maintenance of Adgrg1 expression in TRM cells. It is important to note that the reduced numbers of TRM cells in Hobit and Blimp-1 double-deficient mice may also impact Adgrg1 expression. However, the partial defects in GPR56 expression in single Hobit and Blimp-1 deficient mice that have normal numbers of TRM in small intestine suggest a direct impact of these transcription factors on the regulation of GPR56 expression.   Naïve CD8 + T cells were sorted from uninfected mice, and virusspecific memory CD8 + T-cell populations were sorted from indicated tissues of mice, infected with herpes simplex virus (HSV) or lymphocytic choriomeningitis virus (LCMV), at day 40+ postinfection (memory phase). A panel of 31 adhesion GPCRs was analyzed, of which 19 adhesion GPCRs were present, and 12 adhesion GPCRs were absent in the RNA sequencing data. Data are derived from GSE70813. (B) Adgrg1 transcripts were determined by quantitative RT-PCR in indicated CD8 + T-cell populations from liver, spleen, and small intestinal intraepithelial lymphocytes (SI-IELs). LCMV-specific CD8 + T cells were collected at different time points of infection. Terminal effector cells (TECs) and memory precursor cells (MPECs) were isolated at day 8 (effector phase), and memory CD8 + T cells were obtained at day 30+ postinfection (memory phase). The sorting strategy is shown in Supplementary Figure S1. One-way ANOVA with Tukey's multiple comparisons test; ** p < 0.01, *** p < 0.005.

CD8 + TRM Cell-Inducing Factors Regulate Adgrg1/GPR56 Expression
We previously showed that Hobit and Blimp-1 strongly collaborate in the transcriptional regulation of TRM cells [28] and that Hobit is the driving transcription factor for the expression of granzyme B in TRM cells [25]. The analysis of Adgrg1 expression in TRM cells of Zfp683-deficient and Prdm1-deficient mice showed strong reduction of Adgrg1 expression in mouse TRM cells lacking Hobit and, to a lesser extent, in those lacking Blimp-1 (Figure 2A,B), in support of a role for Hobit in the transcriptional regulation of GPR56. The expression of Adgrg1 was further reduced in Zfp683 and Prdm1 double-deficient TRM cells ( Figure 2B), indicating coregulation of Hobit and Blimp-1 in the maintenance of Adgrg1 expression in TRM cells. It is important to note that the reduced numbers of TRM cells in Hobit and Blimp-1 double-deficient mice may also impact Adgrg1 expression. However, the partial defects in GPR56 expression in single Hobit and Blimp-1 deficient mice that have normal numbers of TRM in small intestine suggest a direct impact of these transcription factors on the regulation of GPR56 expression.  are shown upon different condition treatments as indicated. D FPKM, fragments per kilobase of transcript per million mapp determined using quantitative RT-PCR in sorted spleen CD8 + TGF-β for 3 and 9 days from two independent experiments (n shown for OTI cells from WT or Tgfbr2-deficient mice in skin Data are derived from GSE178769. One-way or two-way ANO sons test; * p < 0.05, ** p < 0.01, *** p < 0.005. TGF-β has been shown to specifically promote the TRM cells in vivo [29] in the skin and small intestine [30 the residency-specific transcriptional profile of TRM c Itgae/CD103 [31]. To determine the impact of TGF-β on RNA sequencing data of spleen CD8 + T cells cultured in β. We observed that TGF-β not only induced expressio but also strongly upregulated expression of Adgrg1 (Fig (D) Adgrg1 transcripts were determined using quantitative RT-PCR in sorted spleen CD8 + T cells after culture with or without TGF-β for 3 and 9 days from two independent experiments (n = 6). (E) Adgrg1 transcript counts are shown for OTI cells from WT or Tgfbr2-deficient mice in skin of HSV-OVA-infected recipients. Data are derived from GSE178769. One-way or two-way ANOVA with Tukey's multiple comparisons test; * p < 0.05, ** p < 0.01, *** p < 0.005. TGF-β has been shown to specifically promote the development or maintenance of T RM cells in vivo [29] in the skin and small intestine [30]. TGF-β can also establish part of the residency-specific transcriptional profile of T RM cells, including the induction of Itgae/CD103 [31]. To determine the impact of TGF-β on Adgrg1 expression, we analyzed RNA sequencing data of spleen CD8 + T cells cultured in the presence of IL-2 and/or TGF-β. We observed that TGF-β not only induced expression of Itgae under these conditions but also strongly upregulated expression of Adgrg1 ( Figure 2C). The presence of IL-2 did not impact Adgrg1 expression ( Figure 2C). Analysis using the quantitative RT-PCR validated the upregulation of Adgrg1 transcripts in splenic CD8 + T cells after culture with TGF-β ( Figure 2D). Furthermore, the expression of Adgrg1 is substantially reduced in skin T RM cells of mice with a deficiency in the TGF-β receptor II (Tgfbr2 −/− ) ( Figure 2E). Our results suggest that the T RM cell-inducing transcription factor Hobit and the T RM cell-inducing cytokine TGF-β contribute to the expression of Adgrg1 in T RM cells.

Adgrg1/GPR56 Is Dispensable for the Development of CD8 + T RM Cells
To study the involvement of GPR56 in CD8 + T-cell differentiation, we crossed floxed Adgrg1 mice [21] with Lck-Cre mice [24], thereby generating mice with deficient GPR56 expression in T cells. Analysis of the activity of Lck-driven CRE at the Adgrg1 locus in T cells indicated the specific deletion of this gene (Supplementary Figure S2). The impact of GPR56 was analyzed on virus-specific CD8 + T cells arising after an acute viral infection with LCMV ( Figure 3A). Virus-specific CD8 + T cells recognizing the dominant epitope GP 33-41 were detected using tetramers at different locations and time points after infection. We found that GPR56 was not essential for the formation of GP 33-41 + CD8 + T cells in spleen, liver, and small intestine at effector (day 8) and memory time points (day 30+) after LCMV infection ( Figure 3B). Expression of IL-7Rα chain (CD127) in conjunction with KLRG1 has been used to distinguish memory precursor effector cells (MPECs; KLRG1 − CD127 + ) from terminal effector cells (TECs; KLRG1 + CD127 − ) in mice [32,33]. Our analysis showed that these two effector populations appear to develop normally in the absence of GPR56 during the effector phase of an LCMV infection ( Figure 3C). Expression of CD69 and CD62L divides memory CD8 + T cells into CD69 − CD62L + T CM , CD69 − CD62L − T EM , and CD69 + CD62L − T RM cells. In contrast to T RM cells in liver, a large fraction of T RM cells in the intraepithelial compartment of the small intestine expressed CD103 [34]. Analysis of the memory subsets in WT and Adgrg1-deficient mice showed that the formation of T CM -, T EM -, and T RM -cell populations in the spleen and liver and the formation of CD103 − and CD103 + T RM cells in the small intestine was not dependent on GPR56 ( Figure 3D). Thus, GPR56 does not appear to impact the differentiation of virus-specific CD8 + T cells into memory precursors and the development of downstream T CM -, T EM -, and T RM -cell populations.

Adgrg1/GPR56 Does Not Regulate Cytokine Production and Release of CD8 + T RM Cells
Virus-specific memory CD8 + T cells that develop after LCMV infection can effectively respond with the production and release of proinflammatory cytokines, such as IFNγ. To examine whether GPR56 is involved in the regulation of IFN-γ production upon reactivation, we briefly cultured spleen and liver cells of LCMV-infected mice in the presence of the LCMV peptide GP 33-41 ( Figure 4A,B). Taking advantage of the T RM cellassociated molecule CD69 to distinguish circulating memory T cells from T RM cells, we observed that restimulated T RM cells from liver produced more IFN-γ than circulating memory T cells from the spleen and liver ( Figure 4A,B). Moreover, T RM cells upregulated expression of CD107a, indicating the ability to degranulate and release cytokines, to a higher extent than circulating memory T cells ( Figure 4A,B). However, IFN-γ production and degranulation was independent of GPR56 ( Figure 4A,B). Taken together, we did not find evidence that GPR56 has an effect on the production and release of the proinflammatory cytokine IFN-γ by CD8 + T cells.

Adgrg1/GPR56 Does Not Regulate Cytokine Production and Release of CD8 + TRM Cells
Virus-specific memory CD8 + T cells that develop after LCMV infection can effectively respond with the production and release of proinflammatory cytokines, such as IFN-γ. To from the spleen and liver ( Figure 4A,B). Moreover, TRM cells upregulated expressio CD107a, indicating the ability to degranulate and release cytokines, to a higher extent t circulating memory T cells ( Figure 4A,B). However, IFN-γ production and degranula was independent of GPR56 ( Figure 4A,B). Taken together, we did not find evidence GPR56 has an effect on the production and release of the proinflammatory cytokine I γ by CD8 + T cells.

Adgrg1/GPR56 Does Not Regulate Cytotoxic Function of CD8 + TRM Cells
We exploited the Listeria-ovalbumin (OVA) infection model to validate our find on the role of GPR56 in TRM cells in the LCMV infection model. Listeria monocytogenes i intracellular bacterium that, upon oral administration, establishes acute infection in small intestine and after systemic spread in other organs such as the liver. Similar a the LCMV infection model, TRM-cell populations develop in liver and small intestine a Listeria-OVA infection [35]. We used tetramers recognizing OVA-specific CD8 + T cel determine the effect of GPR56 on the differentiation of OVA-specific T cells after prim infection with Listeria-OVA ( Figure 5A). Corroborating our results in the LCMV infec model, GPR56 did not appear to impact the differentiation of pathogen-specific CD cells into TCM, TEM, and TRM cells in the spleen and liver or into CD103 − and CD103 + cell populations in the intraepithelial and lamina propria compartment of the small in tine after Listeria-OVA infection ( Figure 5B). Effector CD8 + T cells upregulate protein pression of the cytolytic enzyme granzyme B, which is retained in TRM cells after patho clearance, but not in circulating memory CD8 + T cells [25]. Consistent with these findi we observed that granzyme B protein expression is elevated in TRM cells in liver and s intestine compared to TCM and TEM cells in liver and spleen ( Figure 5C). However, c parison between WT and Adgrgr1-deficient CD8 + T cells did not reveal differences in expression of granzyme B ( Figure 5C). Thus, GPR56 does not appear to essentially tribute to TRM-cell differentiation in the Listeria-OVA infection model and the regula of granzyme B expression in Listeria-OVA specific TRM cells. TGF-β has been shown to specifically promote the development or mainten TRM cells in vivo [29] in the skin and small intestine [30]. TGF-β can also establish the residency-specific transcriptional profile of TRM cells, including the induc Itgae/CD103 [31]. To determine the impact of TGF-β on Adgrg1 expression, we an RNA sequencing data of spleen CD8 + T cells cultured in the presence of IL-2 and/o β. We observed that TGF-β not only induced expression of Itgae under these con but also strongly upregulated expression of Adgrg1 ( Figure 2C). The presence of I not impact Adgrg1 expression ( Figure 2C). Analysis using the quantitative RT-PC dated the upregulation of Adgrg1 transcripts in splenic CD8 + T cells after cultur TGF-β ( Figure 2D). Furthermore, the expression of Adgrg1 is substantially reduced TRM cells of mice with a deficiency in the TGF-β receptor II (Tgfbr2 −/− ) ( Figure 2E results suggest that the TRM cell-inducing transcription factor Hobit and the TRM ducing cytokine TGF-β contribute to the expression of Adgrg1 in TRM cells.

Adgrg1/GPR56 Is Dispensable for the Development of CD8 + TRM Cells
To study the involvement of GPR56 in CD8 + T-cell differentiation, we crossed Adgrg1 mice [21] with Lck-Cre mice [24], thereby generating mice with deficient expression in T cells. Analysis of the activity of Lck-driven CRE at the Adgrg1 loc cells indicated the specific deletion of this gene (Supplementary Figure S2). The im GPR56 was analyzed on virus-specific CD8 + T cells arising after an acute viral in with LCMV ( Figure 3A). Virus-specific CD8 + T cells recognizing the dominant e GP33-41 were detected using tetramers at different locations and time points after in We found that GPR56 was not essential for the formation of GP33-41 + CD8 + T cells in liver, and small intestine at effector (day 8) and memory time points (day 30+) after infection ( Figure 3B). Expression of IL-7Rα chain (CD127) in conjunction with KLR TGF-β has been shown to specifically promote the development or maint TRM cells in vivo [29] in the skin and small intestine [30]. TGF-β can also establi the residency-specific transcriptional profile of TRM cells, including the ind Itgae/CD103 [31]. To determine the impact of TGF-β on Adgrg1 expression, we RNA sequencing data of spleen CD8 + T cells cultured in the presence of IL-2 an β. We observed that TGF-β not only induced expression of Itgae under these c but also strongly upregulated expression of Adgrg1 ( Figure 2C). The presence o not impact Adgrg1 expression ( Figure 2C). Analysis using the quantitative RTdated the upregulation of Adgrg1 transcripts in splenic CD8 + T cells after cul TGF-β ( Figure 2D). Furthermore, the expression of Adgrg1 is substantially reduc TRM cells of mice with a deficiency in the TGF-β receptor II (Tgfbr2 −/− ) ( Figure  results suggest that the TRM cell-inducing transcription factor Hobit and the T ducing cytokine TGF-β contribute to the expression of Adgrg1 in TRM cells.

Adgrg1/GPR56 Is Dispensable for the Development of CD8 + TRM Cells
To study the involvement of GPR56 in CD8 + T-cell differentiation, we cross Adgrg1 mice [21] with Lck-Cre mice [24], thereby generating mice with deficie expression in T cells. Analysis of the activity of Lck-driven CRE at the Adgrg1 l cells indicated the specific deletion of this gene (Supplementary Figure S2). The GPR56 was analyzed on virus-specific CD8 + T cells arising after an acute viral with LCMV ( Figure 3A). Virus-specific CD8 + T cells recognizing the dominan GP33-41 were detected using tetramers at different locations and time points after We found that GPR56 was not essential for the formation of GP33-41 + CD8 + T cells liver, and small intestine at effector (day 8) and memory time points (day 30+) aft infection ( Figure 3B). Expression of IL-7Rα chain (CD127) in conjunction with K ) spleen (A) and liver (B) upon in vitro restimulation with GP 33-41 peptide. Two-way ANOVA with Tukey's multiple comparisons test; n.s., no significant difference.

Adgrg1/GPR56 Does Not Regulate Cytotoxic Function of CD8 + T RM Cells
We exploited the Listeria-ovalbumin (OVA) infection model to validate our findings on the role of GPR56 in T RM cells in the LCMV infection model. Listeria monocytogenes is an intracellular bacterium that, upon oral administration, establishes acute infection in the small intestine and after systemic spread in other organs such as the liver. Similar as in the LCMV infection model, T RM -cell populations develop in liver and small intestine after Listeria-OVA infection [35]. We used tetramers recognizing OVA-specific CD8 + T cells to determine the effect of GPR56 on the differentiation of OVA-specific T cells after primary infection with Listeria-OVA ( Figure 5A). Corroborating our results in the LCMV infection model, GPR56 did not appear to impact the differentiation of pathogen-specific CD8 + T cells into T CM , T EM , and T RM cells in the spleen and liver or into CD103 − and CD103 + T RM -cell populations in the intraepithelial and lamina propria compartment of the small intestine after Listeria-OVA infection ( Figure 5B). Effector CD8 + T cells upregulate protein expression of the cytolytic enzyme granzyme B, which is retained in T RM cells after pathogen clearance, but not in circulating memory CD8 + T cells [25]. Consistent with these findings, we observed that granzyme B protein expression is elevated in T RM cells in liver and small intestine compared to T CM and T EM cells in liver and spleen ( Figure 5C). However, comparison between WT and Adgrgr1-deficient CD8 + T cells did not reveal differences in the expression of granzyme B ( Figure 5C). Thus, GPR56 does not appear to essentially contribute to T RM -cell differentiation in the Listeria-OVA infection model and the regulation of granzyme B expression in Listeria-OVA specific T RM cells.

Secondary CD8 + T-Cell Responses Are Not Regulated by Adgrg1/GPR56
T RM cells have the ability to expand and differentiate into effector cells that combat re-encountered pathogens upon reinfection. In fact, T RM cells substantially contribute to the establishment of secondary CD8 + T-cell responses [23]. We therefore employed the Listeria-OVA infection model to study the role of GPR56 in T-cell responses in the context of prime boost infection ( Figure 6A). We observed that secondary infection of Listeria-OVA resulted in a substantial increase in the percentage and number of pathogen-specific CD8 + T cells in the spleen, liver, and intraepithelial and lamina propria compartment of the small intestine ( Figure 6B-E). This increase in the number of pathogen-specific CD8 + T cells was notable as early as day 8 after reinfection ( Figure 6B-E). However, GPR56 did not appear to have an effect on the magnitude of the secondary CD8 + T cell response in spleen, liver, and small intestine ( Figure 6B-E). We further analyzed the secondary response of pathogen-specific CD8 + T cells using expression analysis of the proliferation-associated marker Ki-67. In line with their re-expansion, we detected increased expression of Ki-67 in pathogen-specific CD8 + T cells at day 8 after reinfection with Listeria-OVA ( Figure 6B-E). However, the expression of Ki-67 was not different between pathogen-specific CD8 + T cells of WT and Adgrg1-deficient mice ( Figure 6B-E). In sum, GPR56 expression did not appear to have an effect on the proliferative responses of pathogen-specific CD8 + T cells upon secondary infection with Listeria-OVA. Thus, we conclude that GPR56, despite its highly specific expression on T RM cells, does not essentially contribute to the regulation of the differentiation of T RM cells after primary infection and their reactivation and re-expansion after secondary infection.

Secondary CD8 + T-Cell Responses Are Not Regulated by Adgrg1/GPR56
TRM cells have the ability to expand and differentiate into effector cells that combat re-encountered pathogens upon reinfection. In fact, TRM cells substantially contribute to the establishment of secondary CD8 + T-cell responses [23]. We therefore employed the Listeria-OVA infection model to study the role of GPR56 in T-cell responses in the context of prime boost infection ( Figure 6A). We observed that secondary infection of Listeria-OVA resulted in a substantial increase in the percentage and number of pathogen-specific CD8 + T cells in the spleen, liver, and intraepithelial and lamina propria compartment of the small intestine ( Figure 6B-E). This increase in the number of pathogen-specific CD8 + T cells was notable as early as day 8 after reinfection ( Figure 6B-E). However, GPR56 did not appear to have an effect on the magnitude of the secondary CD8 + T cell response in spleen, liver, and small intestine ( Figure 6B-E). We further analyzed the secondary response of pathogen-specific CD8 + T cells using expression analysis of the proliferationassociated marker Ki-67. In line with their re-expansion, we detected increased expression of Ki-67 in pathogen-specific CD8 + T cells at day 8 after reinfection with Listeria-OVA (Figure 6B-E). However, the expression of Ki-67 was not different between pathogen-specific CD8 + T cells of WT and Adgrg1-deficient mice ( Figure 6B-E). In sum, GPR56 expression did not appear to have an effect on the proliferative responses of pathogen-specific CD8 + T cells upon secondary infection with Listeria-OVA. Thus, we conclude that GPR56, despite its highly specific expression on TRM cells, does not essentially contribute to the regulation of the differentiation of TRM cells after primary infection and their reactivation and re-expansion after secondary infection.

Discussion
In this report, we have studied the role of the inhibitory receptor GPR56 on pathogenspecific CD8 + T cells. We found that Adgrg1 was specifically upregulated in CD8 + TRM cells in the memory phase after infection with HSV or LCMV. Moreover, we established evidence that implicates the TRM cell-associated transcriptional regulators Hobit and Blimp-1 and the TRM cell-inducing cytokine TGF-β and in the upregulation of Adgrg1 in CD8 + TRM cells. However, under the setting of acute infection with LCMV and Listeria monocytogenes, we did not observe an essential inhibitory role of GPR56 in the regulation of proliferative, cytokine, or cytotoxic responses of CD8 + TRM cells. We conclude that GPR56 is specifically upregulated on CD8 + TRM cells but does not provide an essential contribution in the regulation of the proinflammatory activity of TRM cells after acute infection.
Of note, the balance of GPR56 and GPR97 in hematopoietic stem cells (HSCs) is crucial for the development and differentiation of HSCs [36]. Therefore, it is possible that compensatory effects of other adhesion GPCRs such as GPR97 negate the impact of Ad-grg1-deficiency on TRM cells. In line with potential redundant functions between adhesion GPCRs in TRM, we found pronounced expression of Adgrg3 in skin TRM and ubiquitous expression of Adgrg1 throughout the analyzed TRM-cell populations. The study of the combined role of these two adhesion GPCRs in TRM cells is an important future research direction.
TRM cells have been established as a separate lineage of memory CD8 + T cells with a unique transcriptional profile [33]. Here, we have identified GPR56 as a TRM cell-associated receptor. Adgrg1 was expressed in TRM cells in skin, liver, and small intestine, but not in circulating TCM-and TEM-cell populations in spleen and liver. TRM-cell populations in different tissues are exposed to distinct environmental conditions and therefore may have tissue-specific expression profiles. The expression of Adgrg1 in TRM cells in skin, liver, and liver of wild type (WT, ) and Zfp683-deficie tion. The gating strategy is shown in Supplem mined using quantitative RT-PCR in LCMV-s thelial lymphocytes (SI-IELs) of WT, Zfp683 −/C Zfp683 −/CRE × Prdm1 flox/flox (DKO) mice at day 3 strategy is similar, as shown in Supplementar are shown upon different condition treatmen FPKM, fragments per kilobase of transcript p determined using quantitative RT-PCR in sor TGF-β for 3 and 9 days from two independen shown for OTI cells from WT or Tgfbr2-defici Data are derived from GSE178769. One-way sons test; * p < 0.05, ** p < 0.01, *** p < 0.005. TGF-β has been shown to specifical TRM cells in vivo [29] in the skin and sma the residency-specific transcriptional p Itgae/CD103 [31]. To determine the impa RNA sequencing data of spleen CD8 + T c β. We observed that TGF-β not only ind but also strongly upregulated expression not impact Adgrg1 expression ( Figure 2C dated the upregulation of Adgrg1 transc TGF-β ( Figure 2D). Furthermore, the exp TRM cells of mice with a deficiency in th results suggest that the TRM cell-inducin ducing cytokine TGF-β contribute to the

Adgrg1/GPR56 Is Dispensable for the D
To study the involvement of GPR56 Adgrg1 mice [21] with Lck-Cre mice [24] expression in T cells. Analysis of the act cells indicated the specific deletion of thi GPR56 was analyzed on virus-specific C with LCMV ( Figure 3A). Virus-specific ) and Adgrg1-deficient (G56KO, as been shown to specifically promote the development or maintenance of ivo [29] in the skin and small intestine [30]. TGF-β can also establish part of -specific transcriptional profile of TRM cells, including the induction of 31]. To determine the impact of TGF-β on Adgrg1 expression, we analyzed ing data of spleen CD8 + T cells cultured in the presence of IL-2 and/or TGFed that TGF-β not only induced expression of Itgae under these conditions gly upregulated expression of Adgrg1 ( Figure 2C). The presence of IL-2 did dgrg1 expression ( Figure 2C). Analysis using the quantitative RT-PCR valiregulation of Adgrg1 transcripts in splenic CD8 + T cells after culture with e 2D). Furthermore, the expression of Adgrg1 is substantially reduced in skin ice with a deficiency in the TGF-β receptor II (Tgfbr2 −/− ) ( Figure 2E). Our st that the TRM cell-inducing transcription factor Hobit and the TRM cell-inne TGF-β contribute to the expression of Adgrg1 in TRM cells.

PR56 Is Dispensable for the Development of CD8 + TRM Cells
the involvement of GPR56 in CD8 + T-cell differentiation, we crossed floxed [21] with Lck-Cre mice [24], thereby generating mice with deficient GPR56 T cells. Analysis of the activity of Lck-driven CRE at the Adgrg1 locus in T the specific deletion of this gene (Supplementary Figure S2). The impact of nalyzed on virus-specific CD8 + T cells arising after an acute viral infection

Discussion
In this report, we have studied the role of the inhibitory receptor GPR56 on pathogenspecific CD8 + T cells. We found that Adgrg1 was specifically upregulated in CD8 + T RM cells in the memory phase after infection with HSV or LCMV. Moreover, we established evidence that implicates the T RM cell-associated transcriptional regulators Hobit and Blimp-1 and the T RM cell-inducing cytokine TGF-β and in the upregulation of Adgrg1 in CD8 + T RM cells. However, under the setting of acute infection with LCMV and Listeria monocytogenes, we did not observe an essential inhibitory role of GPR56 in the regulation of proliferative, cytokine, or cytotoxic responses of CD8 + T RM cells. We conclude that GPR56 is specifically upregulated on CD8 + T RM cells but does not provide an essential contribution in the regulation of the proinflammatory activity of T RM cells after acute infection.
Of note, the balance of GPR56 and GPR97 in hematopoietic stem cells (HSCs) is crucial for the development and differentiation of HSCs [36]. Therefore, it is possible that compensatory effects of other adhesion GPCRs such as GPR97 negate the impact of Adgrg1-deficiency on T RM cells. In line with potential redundant functions between adhesion GPCRs in T RM , we found pronounced expression of Adgrg3 in skin T RM and ubiquitous expression of Adgrg1 throughout the analyzed T RM -cell populations. The study of the combined role of these two adhesion GPCRs in T RM cells is an important future research direction.
T RM cells have been established as a separate lineage of memory CD8 + T cells with a unique transcriptional profile [33]. Here, we have identified GPR56 as a T RM cell-associated receptor. Adgrg1 was expressed in T RM cells in skin, liver, and small intestine, but not in circulating T CM -and T EM -cell populations in spleen and liver. T RM -cell populations in different tissues are exposed to distinct environmental conditions and therefore may have tissue-specific expression profiles. The expression of Adgrg1 in T RM cells in skin, liver, and small intestine suggests that GPR56 is upregulated on T RM cells in diverse microenvironments and may be part of the universal gene signature of T RM cells. However, further analysis of T RM cells in other tissues is required to determine whether GPR56 is ubiquitously expressed in T RM -cell populations throughout tissues.
In line with conserved expression of GPR56 between T RM -cell populations of humans and mice, we recently reported GPR56 expression on human brain T RM cells [37]. Although GPR56 has a wider expression pattern in humans than in mice, the retained expression of GPR56 in human T RM -cell populations suggests that our findings in mice have relevance for humans. Importantly, we identified expression of GPR56 protein in human brain T RM cells, indicating that this adhesion GPCR, at least in humans, is not regulated at protein level [37]. Unfortunately, antibodies are not yet available for murine GPR56, preventing us from analyzing protein expression in mice. Therefore, the possibility that lack of protein expression underlies the absence of functional defects in T RM cells of GPR56-deficient mice exists. However, we consider this option unlikely, given that we can detect both GPR56 mRNA and protein in humans.
The expression of Adgrg1 appears to be induced in the T RM -cell lineage at a late time point during T-cell differentiation after acute infection. We did not observe expression of Adgrg1 in naïve T cells or in effector CD8 + T cells including memory precursors that are upstream of T RM cells. We cannot exclude the possibility that a minor subset of these memory precursors upregulates Adgrg1 expression. Memory precursors with exclusive potential to form T RM cells appear to separate early from other effector cells [38][39][40]. These T RM precursor cells may be present in the bloodstream [40] and locally within peripheral tissues [38], where they eventually settle and form T RM cells. With current tools, it remains difficult to unequivocally identify precursor stages of T RM cells; therefore, analysis of the expression of GPR56 by these subsets awaits further characterization of T RM precursor cells. The repertoire of surface markers to identify T RM cells includes CD69, CD103, CD49a, and P2XR7 and other surface molecules [33,41,42]. Our identification of the specific expression of GPR56 on T RM cells suggests that this surface receptor may also be employed as a surrogate marker of T RM cells upon the establishment of antibodies recognizing mouse GPR56. As mentioned above, the expression pattern of GPR56 in humans appears broader than in mice. In humans, GPR56 is also expressed on cytotoxic lymphocytes, including NK cells and effector-type CD8 + and CD4 + T cells in peripheral blood [4] besides T RM cells in peripheral tissues [37]. The specific expression of GPR56 on T RM cells, rather than circulating memory T cells, in mice is consistent with their immediate cytotoxic capacity, given that T RM cells retain cytotoxic molecules, such as granzymes, at protein level in contrast to other memory T cells [25]. Thus, although GPR56 in humans may be differentially expressed compared to GPR56 in mice, upregulation of this adhesion GPCR on the cytotoxic fraction of memory T cells appears conserved between both species.
The differentiation of T RM cells is instructed by cytokines in their local environment. In particular, TGF-β has been found instrumental in the differentiation of T RM cells in the epithelial compartment of the skin and the small intestine [29,33]. We have identified TGF-β as a potent inducer of Adgrg1 expression in CD8 + T cells in in vitro cultures. Notably, developing T RM cells in skin require TGF-β signaling to acquire Adgrg1 expression. These findings suggest that T RM -cell populations in the skin and small intestine upregulate GPR56 expression in response to TGF-β signaling. Expression of CD103, the αE component of the αEβ7 integrin, also strongly depends on TGF-β signaling [29]. In contrast to CD103, we have observed that GPR56 is also expressed on liver T RM cells outside of CD103-expressing T RM -cell populations in skin and intestine. Therefore, the expression regulation of GPR56 in T RM cells appears more complex and may include environmental cues other than TGF-β. We have previously observed in human NK cells that the expression of GPR56 is downregulated in the presence of the homeostatic cytokine IL-15 and upon activation with inflammatory cytokines, including IL-2 and IL-18 [10]. Therefore, it is unlikely that IL-15, which can contribute to T RM -cell maintenance in mice [43][44][45], and these other cytokines play a role in the upregulation of GPR56 on T RM cells. We have previously also observed that GPR56 was induced by the transcription factor Hobit in human NK cells [10]. Hobit and the related transcription factor Blimp-1 are master regulators of T RM -cell differentiation, which, through suppression of tissue exit molecules, such as S1PR1 and CCR7, can permanently lock these memory T cells into the peripheral tissues [28]. Consistent with these findings, we observed here that Hobit together with Blimp-1 contributed to the transcriptional regulation of GPR56 expression in T RM cells. Hobit and Blimp-1 are broadly expressed in T RM cells throughout tissues including those of the skin, liver, and small intestine [28], suggesting that the transcription factor may drive GPR56 expression on CD103 − and CD103 + T RM -cell populations in these tissues. The expression of Hobit and Blimp-1 appears independent of TGF-β signaling, but currently, it remains unknown how the expression of these transcription factors is upregulated during T RM -cell differentiation. Thus, GPR56 expression may be regulated through separate pathways, involving TGF-β and yet unresolved cytokines that trigger Hobit and Blimp-1 expression.
The strong proinflammatory activities of T RM cells after reinfection may have harmful impact on the surrounding healthy tissues. Pathogen-specific T RM cells express a multitude of inhibitory receptors including PD-1, TIM-3, LAG-3 and CD39 that can restrain T RM cell-driven immune responses [33]. Interestingly, CD103 + T RM cells appear to be under stronger regulation by inhibitory receptors than CD103 − T RM cells [30]. In line with these findings, we also found that the inhibitory receptor GPR56 is responsive to TGF-β suggesting elevated expression of this receptor on CD103 + T RM cells compared to CD103 − T RM cells. We have previously found that GPR56 is an inhibitory receptor, which can suppress the cytotoxic and cytokine responses of NK cells [10]. In this paper, we addressed the suppressive function of GPR56 on T RM cells after infection with LCMV and Listeria monocytogenes. In contrast to its essential role in regulating proinflammatory responses of NK cells, we did not observe that GPR56 substantially contributed to the control of the magnitude of T RM -cell responses or the cytokine production of T RM cells. We also did not find evidence that GPR56 contributed to the regulation of the cytolytic enzyme granzyme B in T RM cells, suggesting that the regulatory role of GPR56 is not essential under these conditions. Evidence is accumulating that expression of inhibitory receptors, such as PD-1, on T RM cells is highly relevant to protect against T RM cell-driven inflammatory responses that otherwise may develop in the intestine, pancreas, and the lungs [18][19][20]. Possibly, the inhibitory impact of GPR56 in the regulation of T RM cells is masked by the presence of other inhibitory receptors on T RM cells. Given that proinflammatory cytokines suppress GPR56 expression [10], unleashing the inhibitory potential of GPR56 may also require an anti-inflammatory environment, such as occurs in a tumor setting. More recently, T RM cells have been detected in tumor tissue of mouse models of melanoma [46], and in patients with melanoma, lung carcinoma, and ovarian carcinoma [47][48][49]. These tumor-residing T RM cells highly express PD-1, and inhibition of the PD-1-driven checkpoint blockade pathway appears to reinvigorate their antitumor activity [49]. Therefore, inhibitory receptors appear relevant on T RM cells to restrain antitumor immune responses. Our findings show that T RM cells also express dedicated inhibitory receptors such as GPR56 that suggest the presence of unique regulatory mechanisms in T RM compared to circulating memory T cells. However, resolving the potential role of GPR56 on tumor-resident T RM cells awaits future research.

Institutional Review Board Statement:
The study was approved by the national Animal Ethics Committee of The Netherlands (protocol code: AVD3010020172205 and date of approval: 18 July 2017).

Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon reasonable request.