Rhesus Macaque Activating Killer Immunoglobulin-Like Receptors Associate With Fc Receptor Gamma (FCER1G) and Not With DAP12 Adaptor Proteins Resulting in Stabilized Expression and Enabling Signal Transduction

Activating killer cell immunoglobulin-like receptors (KIR) in macaques are thought to be derived by genetic recombination of the region encoding the transmembrane and intracellular part of KIR2DL4 and a KIR3D gene. As a result, all macaque activating KIR possess a positively charged arginine residue in the transmembrane region. As human KIR2DL4 associates with the FCER1G (also called Fc receptor-gamma, FcRγ) adaptor, we hypothesized that in contrast to human and great ape the activating KIRs of macaques associate with FcRγ instead of DAP12. By applying co-immunoprecipitation of transfected as well as primary cells, we demonstrate that rhesus macaque KIR3DS05 indeed associates with FcRγ and not with DAP12. This association with FcRγ results in increased and substantially stabilized surface expression of KIR3DS05. In addition, we demonstrate that binding of specific ligands of KIR3DS05, Mamu-A1*001 and A1*011, resulted in signal transduction in the presence of FcRγ in contrast to DAP12.


INTRODUCTION
Stimulatory NK cell receptors usually do not transmit signals directly but through association with activation motif-containing accessory proteins DAP12 and FcRg (also called g and encoded by the FCER1G gene). This non-covalent interaction is mediated by basic residues (lysine and arginine) in the transmembrane region of the receptor and an acidic residue (aspartic acid) in the adaptor proteins. The prototypical activating KIRs in human interact with DAP12 (1)(2)(3)(4), while KIR2DL4 is the only KIR known so far to associate with the FcRg adaptor (5). Responsible for this different usage of adaptor proteins are positions of the charged amino acid residues in the transmembrane regions (6): a lysine and an aspartic acid residue can be found in prototypical KIRs and DAP12 at position 9, respectively, while KIR2DL4 has an arginine and FcRg an aspartic acid residue at position 4 and 3, respectively. Both adaptors contain a single immunoreceptor tyrosine-based activation motif (ITAM) of the sequence D/ ExxYxxL/I(x [6][7][8] )YxxL/I in their cytoplasmic region, which is responsible for signal transduction upon binding of a specific ligand by the associated receptor. In addition, these adaptor proteins contribute to stabilization of the cell surface expression of interacting stimulatory receptors of different types such as CD16 (7), NKp30 and NKp46 (8), NKp44 (9), KIR (4,5,10), CD94/NKG2C (3,11), ILT1 (12) and many others (13).
KIR genomics in rhesus macaques is more complex as in humans (14)(15)(16)(17)(18)(19)(20)(21) and the genomic diversity and plasticity are focused on lineage II KIR, i.e. to genes encoding three-domain KIR proteins. In contrast to the other KIR genes, human and macaque KIR2DL4 are orthologous (14,22). Interestingly, the activating rhesus macaque KIR genes were formed by recombination between a KIR3D gene and KIR2DL4 (14) and subsequent splice site mutation in intron 8 (23). Thus, the transmembrane region of activating KIR proteins and of KIR2DL4 are very similar. Hence, all activating rhesus macaque KIR proteins possess an arginine residue at position 4 of the transmembrane region instead of a lysine residue that is found in human and great ape activating KIR at position 9. Which of the two adaptor proteins can associate with macaque activating KIR was so far unknown. Here we demonstrate that a prototypical rhesus macaque activating KIR protein indeed interacts with FcRg and not with DAP12.

Rhesus Macaque PBMC Samples
Peripheral blood samples of 8 rhesus macaques kept at the German Primate Center were obtained from the Animal Husbandry Unit of the DPZ for MHC class I genotyping purposes. Peripheral blood mononuclear cells (PBMC) were isolated as described previously (24). Briefly, peripheral blood was diluted with RPMI medium (1:1) and transferred to a leucosep tube filled with ficoll. After 40 min centrifugation at 800 x g, the PBMC layer was removed carefully and the cells were washed with RPMI. Leftovers of PBMCs from genotyping were used for co-immunoprecipitation experiments. As the exact KIR genotype of the animals was not known, we pooled PBMCs of 4 animals (about 5x10 6 cells in total). Two pools of samples were frozen and stored at -140°C until use for co-immunoprecipitation (see below).

Expression Constructs
The rhesus macaque KIR3DS05 with a C-terminal AcGFP tag was described before (25). We cloned C-terminal myc-tagged versions of either rhesus macaque DAP12 or FcRg in multiple cloning site A of the pIRES expression plasmid (Clontech) and AcGFP-tagged KIR3DS05 in multiple cloning site B. Multiple cloning sites A and B in this vector are separated by an internal ribosomal entry site (IRES). Such bicistronic mRNA expression ensures simultaneous expression of the activating KIR3DS05 and either the DAP12 or the FcRg adaptor protein in transfected cells.

Transfection and Transfected Cell Lines
Plasmid DNA was transfected at 1:3 ratio using Lipofectamine 2000 (Thermo Fisher Scientific) into overnight exponentially grown HEK-293 or HeLa cells in 6-well plates. Transfected cells were cultured in the Medium HEK293 (DMEM, 10% inactivated FBS and 0.1% Gentamycin) and incubated for˜60 -65 hours at 37°C with 5% CO 2 . Stably transfected HEK-293 cells were obtained by selection in neomycin-containing medium for at least 14 d. KIR3DS05 AcGFP , KIR3DS05 AcGFP +FcRg myc , or KIR3DS05 AcGFP +DAP12 myc in pIRES plasmid DNA were also transfected in Jurkat cells using the electroporation-based transfection system Nucleofector II according to the supplier's (Lonza) information for Jurkat cells and these cells were cultured in Medium Jurkat (RPMI, 10% inactivated FBS and 0.1% Gentamycin) at 37°C with 5% CO 2 .
For intracellular staining of myc-tagged adaptor proteins, cells were fixed with fixation buffer (Biolegend), permeabilized with intracellular staining permeabilization wash (Biolegend), and blocked with blocking buffer (1x PBS and 2% BSA). Rabbitanti-myc-tag antibody (1:150) (Cell signaling technology) was then applied for 40 min at RT. Finally, 1 µl Brilliant Violet 421 conjugated donkey anti-rabbit IgG secondary antibody (Biolegend) was added and incubated for another 30 min at RT with staining solution A .
Live/dead staining of cells was performed in 100 ml ice cold Zombie aqua fixable viability kit (Biolegend) for 10 min at RT (dark) and dead cells were excluded from analysis. Cells were analyzed in a BD LSR II flow cytometer (BD Biosciences) and data were analyzed using FlowJo 10.7.
Cells were sorted using the SH800 cell sorter (Sony) and subsequently cultured in appropriate culture medium.

Analysis of Stabilization of KIR3DS05 Expression by Adaptor Proteins
After antibiotic selection, stably expressing KIR3DS05 AcGFP , KIR3DS05 AcGFP +FcRg myc and KIR3DS05 AcGFP +DAP12 myc HEK-293 cells were suspended in sorting buffer (1x PBS, 2% FCS and 2 mM EDTA). Only AcGFP-expressing cells were then gated and equal number of cells (around 80 -90% AcGFPpositive) were sorted and cultured in Medium HEK . We then reanalyzed the cells using antibody 1C7 and used identical frequency of 80 -90% KIR3DS05-positive cells as starting point in all experiments. After different passages with intervals of two to three days, an aliquot of cells was removed after each passage to measure KIR3DS05 expression by flow cytometry.

Stimulation of KIR3DS05 and Adaptor Protein-Expressing Cells
Stimulation of KIR3DS05 was performed using HEK-293 cells expressing AcGFP-tagged MHC class I ligands Mamu-A1*001 or A1*011 and as control the non-interacting Mamu-B*030 (25). Jurkat cells expressing KIR3DS05 with and without adaptor proteins (see above expression constructs) were incubated with these Mamu class I-expressing HEK-293 cells (overnight exponentially grown) in 12-well plates for 14 hours in Medium Jurkat . As readout of stimulation via KIR3DS05 and associated adaptor protein, we measured expression of CD69 by flow cytometry. The non-adherent Jurkat cells were carefully removed from the adherent HEK-293 cells, centrifuged (5 min, 300 x g) and resuspended in 100 ml staining solution A for staining with 3 ml APC anti-human CD69 antibody (Biolegend) for 40 min at RT. Jurkat cells were differentiated from accidentally transferred HEK-293 cells in flow cytometry based on forward and side scatter characteristics.

Co-Immunoprecipitation (Co-IP)
About 1 -2 × 10 6 sorted HEK-293 cells stably expressing KIR3DS05 AcGFP , KIR3DS05 AcGFP +FcRg myc or KIR3DS05 AcGFP + DAP12 myc were grown individually in T25 tissue culture flask in Medium HEK293 at 37°C with 5% CO 2 . The culture medium was discarded and the cells were washed with ice cold 1× DPBS. After discarding the DPBS, cells were lysed in 400 ml ice cold lysis buffer (10 mM Tris-HCl pH 7.5, 150 mM NaCl, 1% N-Dodecyl b-Dmaltoside (Thermo Fisher Scientific), 0.4 mM EDTA and 1 tablet protease-inhibitor-cocktail complete mini (Roche Diagnostic)) for 20 min at 4°C with gentle shaking. After centrifugation (15 min, 16000 × g), the collected lysate was pre-cleared using 40 µl Protein G sepharose beads (GE healthcare) and the protein concentration of the lysate was determined in a Qubit 4 fluorometer using Qubit protein assay kits (Thermo Fisher Scientific). Subsequently, 4.4 mg of monoclonal antibody 1C7 was added to˜40 mg pre-cleared lysate and incubated overnight at 4°C with rotation. Protein G beads were added and incubated for another 5 hours at 4°C. After washing the beads, bound proteins were released by incubation in 30 ml 5x SDS-PAGE reducing protein loading buffer (Bosterbio) at 95°C for 5 min and centrifuged for 5 min, 10000 × g at RT. The supernatant was collected and either used directly for SDS gel electrophoresis or was stored at -20°C.
About 5 × 10 6 rhesus macaque pooled PBMCs were lysed as described above and pre-cleared lysates were subjected to Co-IP with 1C7 antibody in reducing buffer. The collected samples were ready to use for SDS-page or stored at -20°C.

Confocal Laser Microscopy
Preparation and staining of transiently transfected HEK-293 cells were the same as described above for flow cytometry. Cells were stained with 1C7 antibody and as secondary antibody 1 ml of PE goat anti-mouse IgG (Biolegend) to detect cell surface-expressed KIR3DS05 and were subsequently fixed, permeabilized, blocked and stained intracellularly with 0.5 mg anti-myc antibody (Abcam) and 0.25 mg APC conjugated goat anti-rabbit IgG (Abcam) to detect myc-tagged adaptors. Finally, a drop of Fluoromount-G mounting medium with DAPI (Thermo Fisher Scientific) was applied on a microscope glass slide (Carl Roth), and the stained cells were transferred onto the mounting medium and incubated for 5 min at RT, before fixing the cells with a cover slip (24 × 24 mm). The images were captured with the Plan-Apochromat 63x/1.40 oil objective in confocal laser microscope LSM 800 (Carl Zeiss) fitted with ZEN 2.3 software (Carl Zeiss) and mean fluorescence intensity (MFI) value of KIR3DS05 (1C7) and FcRg/DAP12 (anti-myc) in 11 individual stained cells were measured using the same software.

Statistical Analysis
Differences between two groups were analyzed by applying a Student's t-test (parametric, unpaired, two-tailed, 95% confidence level) using GraphPad Prism 9. Differences between groups with p values >0.05 were regarded as not statistically significant.

Identification of the Adaptor Protein for Rhesus Macaque Activating KIR
To find out the interacting adaptor proteins of activating KIR proteins in rhesus macaques, we expressed AcGFP-tagged KIR3DS05 alone or together with myc-tagged FcRg or DAP12 adaptor protein in HEK-293 cells. Equal numbers of KIR3DS05 AcGFP expressing cells were sorted and their lysate was used for Co-IP with anti-rhesus macaque KIR antibody 1C7 (26). The cell lysis buffer contained N-Dodecyl b-D-maltoside, which was previously shown to be superior to other detergents in identifying receptor-adaptor complexes in NK cells (9). Western blot analysis with an HRP-conjugated anti-myc antibody was performed to detect which adaptor protein is associated with KIR3DS05. Bands of the expected size of 10 -13 kDa were detected for both FcRg and DAP12 in the respective control samples ( Figure 1A). When KIR3DS05 was immunoprecipitated, we detected only the FcRg protein, but not DAP12 ( Figure 1A). To confirm this finding obtained from transfected cells and tagged adaptor proteins, we immunoprecipitated KIR proteins from two pools of rhesus macaque PBMCs with anti-rhesus pan-KIR3D antibody 1C7 (26) and used polyclonal antibodies against FcRg and DAP12 in western blots. In accord with the transfection experiments, also in primary cells only FcRg and not DAP12 was found in immunoprecipitated PBMC samples ( Figure 1B). These findings clearly demonstrate that stimulatory KIR in rhesus macaques interact with the FcRg adaptor protein and not with DAP12.

FcRg Adaptor Protein Promotes High Cell Surface Expression of Rhesus Macaque Activating KIR
Enhanced cell surface expression was previously demonstrated for human KIR2DS1, KIR2DS2, KIR2DS4 (10), and KIR3DS1 (4) in the presence of DAP12, and for human KIR2DL4 in the presence of FcRg (5). Thus, we transiently transfected AcGFPtagged KIR3DS05 with or without adaptor proteins in HEK-293 and HeLa cells and compared its cell surface expression in flow cytometry. In both cell lines, KIR3DS05 AcGFP revealed about 2 -3 times higher cell surface expression in the presence of FcRg as compared to the presence of DAP12 or without any adaptor (Figures 2A, B). The enhanced expression is seen in both %-positive cells and mean fluorescent intensity. This higher expression is not due to transfection efficiency as reflected by comparable GFP expression in both experimental settings ( Figure 2C), while 89.0% ( ± 4.8%) of these GFP-positive cells express KIR3DS05 in combination with FcRg and only 40.2% (± 2.7%) of the GFP-positive cells express KIR3DS05 in the absence of any adaptor protein or 49.6% ( ± 7.6%) with DAP12 ( Figures 2C, D).
Expression of KIR3DS05 AcGFP with myc-tagged FcRg and DAP12 in transfected HEK-293 cells was also investigated by confocal laser microscopy. Strong signals of KIR3DS05 (green) and co-localization of KIR3DS05 with FcRg myc (red) complexes were detected, whereas weaker signals of KIR3DS05 and no clear signs of co-localization with DAP12 myc were observed ( Figure 3A). The measured MFI from a comparable area among these cells shows stronger KIR3DS05 AcGFP expression when FcRg myc is present as compared to co-expression with DAP12 ( Figure 3B). The positive correlation of KIR3DS05 and myc expression was more evident with FcRg ( Figure 3C). The ratio of % cells expressing KIR3DS05 and adaptor and only the adaptor being 1.4 in the combination of KIR3DS05 with FcRg and 0.5 in combination with DAP12.
As KIR3DS05 AcGFP was expressed from a bicistronic mRNA with either FcRg myc or DAP12 myc , the similar amount of the two adaptor proteins in both settings implies that also comparable amounts of KIR3DS05 protein are present. Thus, the observed differences in KIR3DS05 cell surface expression are due to functional differences in pairing of KIR3DS05 with these two adaptor proteins and not to experimental variation.

FcRg Adaptor Protein Stabilizes Cell Surface Expression of Rhesus Macaque KIR3DS05
As the proper adaptor protein stabilizes expression of activating human KIR proteins, we tested whether the presence of FcRg also results in stabilized KIR3DS05 cell surface expression. Thus, we sorted 80-90% AcGFP-positive HEK-293 cells stably transfected with either KIR3DS05 AcGFP , KIR3DS05 AcGFP +FcRg or KIR3DS05 AcGFP +DAP12 and stained these cells with 1C7 antibody to measure KIR3DS05 cell surface expression in the A B different cell passages every 2 -3 days ( Figure 4A). KIR3DS05 expressed in the absence of any adaptor protein rapidly leads to reduced cell surface expression over time and a similar loss was noticed for KIR3DS05 in the presence of tagged DAP12: only about 3 -7% KIR3DS05-positive cells remained after 5 passages ( Figure  4B). Contrasting this loss is a relatively stable expression of KIR3DS05 in the presence of FcRg and only a slow reduction noticed over time ( Figures 4A, B). The differences between KIR3DS05 AcGFP +FcRg and the other two conditions are statistically significant for all time points, whereas the comparison between KIR3DS05 AcGFP +DAP12 and KIR3DS05 AcGFP is statistically significant only after passage 1 (p=0.0372) but not thereafter ( Figure 4B). Thus, FcRg not only physically interacts with KIR3DS05, its association also stabilizes cell surface expression of this stimulatory KIR protein.

Stimulatory Signals of Activating KIR Are Transmitted via FcRg
After demonstrating that FcRg is the proper adaptor protein for activating KIR, we investigated whether signal transduction can be demonstrated in the presence of FcRg as compared to DAP12. KIR3DS05 was particularly suitable as its ligand specificity was determined by us before (25). For this, we transfected KIR3DS05 AcGFP, KIR3DS05 AcGFP +FcRg myc or KIR3DS05 AcGFP +DAP12 myc in Jurkat cells. Also in Jurkat cells, which lack endogenous expression of both FcRg and DAP12, the expression of KIR3DS05 was higher in the presence of FcRg (31.8%) as compared to DAP12 (17.5%) or without adaptor protein (15.7%; Figure 5A). These transfected cells were then stimulated for 14 h with HEK-293 cells expressing KIR3DS05 ligands Mamu-A1*001 AcGFP and Mamu-A1*011 AcGFP as well as the non-interacting Mamu-B*030 AcGFP as control. The Mamu class I protein expression of about 85% positive cells and the MFI according to AcGFP expression was comparable between the three cell lines ( Figure 5B). As readout of stimulation, we measured CD69 expression on Jurkat cells by flow cytometry. KIR3DS05 AcGFP alone (0.6-0.9%) or in combination with DAP12 myc (2.8-3.6%; Figure 5C). In the presence of the noninteracting Mamu-B*030, no CD69 induction was seen in all three experimental settings as anticipated ( Figure 5C). Altogether these results demonstrate that in rhesus macaques the FcRg protein is the adaptor that interacts with activating KIR such as KIR3DS05. The presence of FcRg is necessary not only for the stable expression of KIR3DS05 on the cell surface, but also for signal transduction. Figure 6 summarizes our findings and the functional consequences of differential presence of the two adaptor proteins for expression and function of activating KIR in rhesus macaques.

DISCUSSION
The association of signaling adaptor molecules and activating KIR proteins in Old World monkeys such as rhesus macaques was unknown so far. Here we could unambiguously show by different methods that a prototypical activating KIR protein of rhesus macaques associates with the FcRg and not the DAP12 adaptor protein.
For the simultaneous expression of an activating KIR protein together with either the FcRg or DAP12 adaptor, we used an IREScontaining expression vector. To ensure equal expression of FcRg and DAP12 adaptor proteins, we placed the genes for adaptors upstream and the KIR3DS05-encoding gene downstream of the IRES sequence in the expression vector. The downstream position of a gene is known for lower expression efficiency as compared to the position upstream of the IRES (28). Through this order of the genes in the expression construct, we achieved not only simultaneous expression of adaptor and KIR, but also that the amount of KIR3DS05 is likely less than the adaptor protein and, hence, KIR3DS05 and not the adaptor protein would (if at all) be the limiting factor in the following experiments. After demonstrating the association of KIR3DS05 and FcRg by coimmunoprecipitation in transfected HEK-293 cells as well as in rhesus macaque PBMCs, we analyzed in transfected cells the surface expression of KIR3DS05 in the presence or absence of adaptor proteins in flow cytometry. These experiments showed that the highest percentages of KIR3DS05-positive cells as well as the highest mean fluorescence intensity were achieved when FcRg was present and not DAP12. Furthermore, also confocal microscopy not only indicated that KIR3DS05 co-localizes with FcRg but showed also the highest expression level of KIR3DS05 in this combination. These data indicate that the KIR-interacting FcRg obviously stabilized the cell surface expression of KIR3DS05 and that its absence results in low cell surface expression of KIR3DS05. Indeed, when we followed the heterologous expression of KIR3DS05 with or without adaptors over several cell passages, its expression remained high in the presence of FcRg with just very few loss over time, whereas the expression of KIR3DS05 with DAP12 or without any adaptor protein rapidly decreased over time. Thus, the higher percentage of KIR3DS05-expressing cells that was noticed after transfection is most probably due to the stabilizing capability of FcRg. This also indicates a crucial role of FcRg in the regulation of stimulatory KIR3D proteins in macaque NK cells. Indeed, when FcRg is absent or only the 'wrong' adaptor is present, signal transduction upon interaction with a cognate ligand is abrogated as shown by our coincubation experiments with rhesus macaque MHC class I proteins. It should be mentioned that we noticed a small but statistically significant effect of the presence of DAP12 on the stability of cell surface expression of KIR3DS05 (Figures 2 and 4). A possible explanation for this observation might be that we used transfected cells with a strong promoter and, thus, higher levels of adaptor expression than under normal conditions of endogenous promoter control. Indeed, this effect was only seen shortly after transient transfection ( Figure 2) or after start of a time-course kinetic experiment with stably transfected cells (Figure 4). This ectopic expression of DAP12 in the cell membrane may contribute to slightly stabilizing KIR3DS05 cell surface expression on the short term, but due to lack of molecular interaction, DAP12 is not able to stabilize it over time.
It was previously shown that genes encoding activating KIRs in rhesus macaques were most probably derived from recombination involving KIR2DL4 and inhibitory KIR3D genes (14,23). Thus, our data shown here are not only in accord with human KIR2DL4 associating with FcRg (5), our data also support the finding that activating KIR in macaques were indeed derived from such gene recombination during evolution of Old World monkey KIR genes. Besides Old World monkeys, an arginine residue instead of a lysine in the transmembrane region of activating KIR is also present in New World monkeys (29,30) and in small apes (31). In contrast, all hominid species (great apes and human) display a lysine residue in activating KIR proteins that associate with the DAP12 adaptor. This indicates that in the evolution of primates, a change from FcRg to DAP12 in the usage of the activating KIR-associating adaptor protein along with corresponding adaptations in the transmembrane region of activating KIR were fixed in the lineage leading to extant hominid primates. Hence, the usage of FcRg for activating KIR in platyrrhini (New World monkeys), catarrhini (Old World monkeys) and gibbons is likely the ancestral situation. The reason for the change in the usage of specific adaptors for activating KIR in hominid primates is unknown. The evolution of KIR genes in primates is very dynamic and frequently leads to formation of new genes by recombination (14)(15)(16)(17)(18)(19)(20)(21)(22)(23), most probably driven by infections. Thus, it appears advantageous to   have at hand different opportunities for adaptor proteins in order to cope with different requirements from activating KIR. Indeed, the broad expression of FcRg, DAP12 or DAP10 in different cells of the immune system, their association with diverse receptors and their stabilizing role of receptor expression mirror their importance and their regulatory roles (10,13,32,33). Adaptive NK cells in humans generated as a consequence of infection with cytomegalovirus (CMV) are characterized by loss of FcRg expression through epigenetic silencing (34,35). While such loss of FcRg is thought to result in concomitant loss of stimulatory receptors as shown for NKp46 and NKp30 (8,(35)(36)(37)(38), it also results in increased killing capacity of adaptive NK cells mediated via CD16 (34,35,39). This might be explained by association of CD16 with both FcRg and TCRϛ signaling proteins (7,(40)(41)(42) and the higher number of ITAMs in TCRϛ compared to FcRg. Thus, CMV-associated adaptive human NK cells are regulated by loss of FcRg to sharpen CD16-mediated recognition of antibody-tagged cell targets. FcRg-deficient adaptive NK cells were also reported in CMV-infected rhesus macaques (43). We hypothesize, that loss of FcRg in adaptive NK cells of macaques not only sharpens their CD16-mediated effector function as described previously by others (44), but that concomitant loss of activating KIRs further refines their immune function: moving towards antibody-driven and away from ligand (MHC class I)driven recognition. If this would indeed be the case, then macaque activating KIRs are expected to have a more prominent role in shaping adaptive NK cell functions as their human counterparts.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

AUTHOR CONTRIBUTIONS
LW conceived the project. MZH and LW designed the experiments and wrote the manuscript. MH carried out all experiments. MH and LW analysed the data. All authors contributed to the article and approved the submitted version.