NBEAL2 is required for neutrophil and NK cell function and pathogen defense

Mutations in the human NBEAL2 gene cause gray platelet syndrome (GPS), a bleeding diathesis characterized by a lack of α granules in platelets. The functions of the NBEAL2 protein have not been explored outside platelet biology, but there are reports of increased frequency of infection and abnormal neutrophil morphology in patients with GPS. We therefore investigated the role of NBEAL2 in immunity by analyzing the phenotype of Nbeal2-deficient mice. We found profound abnormalities in the Nbeal2-deficient immune system, particularly in the function of neutrophils and NK cells. Phenotyping of Nbeal2-deficient neutrophils showed a severe reduction in granule contents across all granule subsets. Despite this, Nbeal2-deficient neutrophils had an enhanced phagocyte respiratory burst relative to Nbeal2-expressing neutrophils. This respiratory burst was associated with increased expression of cytosolic components of the NADPH oxidase complex. Nbeal2-deficient NK cells were also dysfunctional and showed reduced degranulation. These abnormalities were associated with increased susceptibility to both bacterial (Staphylococcus aureus) and viral (murine CMV) infection in vivo. These results define an essential role for NBEAL2 in mammalian immunity.


Proteome analysis
For analysis of subcellular location of differentially expressed proteins (DEPs), the limma generated DEPs were first annotated with known or predicted cellular location using the UniProt database (http://www.uniprot.org/uploadlists/). The DEP list was then grouped based on subcellular location/organelle for both upregulated and downregulated proteins. Proteins with more than one localisation entry were grouped based on the additional 'note' information included with the UniProt subcellular location annotation. For example, a protein described as: "cytoplasmic unless phosphorylated post-activation and then translocated to the nucleus" would be grouped as "cytoplasmic related". For PANTHER Gene Ontology Enrichment, significantly downregulated (positive fold-change) proteins in Nbeal2 -/neutrophils were filtered based on the log-fold change (≥2.5-fold). Unique proteins were then analysed using the PANTHER Overrepresentation Test (release 20160715) via the online analysis tool (http://geneontology.org/page/go-enrichment-analysis). All filtered proteins identified in the MS analysis were used as the reference for the analysis. Each list was tested against the GO Ontology database (release 2015-10-27) for enrichment of "cellular components" and/or "biological processes". For protein set enrichment, the filtered and normalised MS expression set was analysed using the java based GSEA software (http://www.broad.mit.edu/gsea/). Custom protein sets were designed following the GSEA instructions using the dataset referenced earlier in the text. GSEA was run under default settings with the exception of 'Permutation type' (gene_set) and 'Number of permutations' (10,000). For analysis of GTP family enrichment, a GTPase list was downloaded from the qiagen resource website (https://www.qiagen.com/us/resources/resourcedetail?id=1cff0638-82a2-4003-a86a-bebeadc448f8&lang=en). This list was filtered to remove any GTPases not expressed in our proteome data and then further subdivided into GTPase families. These lists were then used for protein set enrichment as described above.
Processing of murine tissues [6][7][8][9][10] week old mice were sacrificed and whole blood, spleen and hind legs were collected. Spleens were weighed before being processed by manual dissociation, red blood cell (RBC) lysis treated, washed in FACS buffer (PBS 0.5% BSA) and enumerated. Bone marrow prepared by first removing excess muscle tissue from both the femurs and tibias of sacrificed mice. Bone marrow cells were then flushed from each bone using a 26G needle (BD Bioscience, UK) and 5ml of cold PBS. Cells were RBC lysis treated, washed and enumerated. Blood was collected in clean 1.5 ml centrifuge tubes and allowed to clot for 1 hour on the bench at room temperature. Coagulated blood was spun at 1500g for 15 minutes and serum was collected and stored at -20 °C until needed.

NK in vitro degranulation assay
For the in vitro degranulation of NK cells a previously described protocol was used (4). In brief, splenocytes prepared as before were counted and resuspended at a concentration of 2 x 10 6 per ml in RPMI containing anti-LAMP-1-PE or isotype antibody conjugated to PE at a dilution of 1:200. 1 x 10 6 splenocytes were plated into wells of a 24-well plate and stimulated with 100 /ml PMA and 150ng/ml ionomycin (both Sigma) for indicated time points. Post-stimulation cells were retrieved from wells, washed and processed for extracellular staining as described earlier.

T cell Cytotoxcity assay
Single cell suspensions of splenocytes from Nbeal2 +/+ and Nbeal2 -/mice were made, before being enumerated. Splenocytes were plated in 6-well plates pre-coated with anti-CD3e (1µg/ml) and anti-CD28 (2µg/ml) at a density of 1.7 × 10 6 /ml (8ml total) in T cell media (RPMI (Sigma) plus 10% fetal calf serum (Sigma), 10U/ml Penicillin Streptomycin (Thermo), 1mM Sodium Pyruvate (Sigma), 2mM L-glutamine (Sigma), 50µM 2-Mercaptoethanol and IL-2 (20ng/ml). 48 hours post-stimulation cells were washed, resuspend in 20ml T cell media and 5ml plated per well of a 6-well plate. On day 6 cells were used for cytotoxicity assay. The target cell line P815 (ATCC, UK) was cultured in DMEM (Sigma) with 10% fetal calf serum and split one day prior to assay being carried out. On the day of the assay P815 were washed, and counted, and loaded with anti-mouse CD3e (clone 145-2c11, eBioscience). Cytotoxic T cells were counted, washed, resuspended media plus 1% fetal calf serum before being plated in a U-bottomed 96-well plate in triplicate per condition/genotype. A starting effector:target ratio of 10 was used and diluted serially. 10 4 P815 cells were added to each plate as targets. No target controls were included for each dilution, as well as media alone, targets alone, targets with lysis buffer and media with lysis buffer. The lysis buffer was part of the lactate dehydrogenase (LDH) release kit (Promega, UK), which was used to quantify target cell lysis following the manufacturers instructions.

NK Cytotoxicity assay
Single cell suspensions of splenocytes from Nbeal2 +/+ and Nbeal2 -/mice were made, before being enumerated. NK cells were enriched using the NK cell isolation kit II (Miltenyi, UK) according to the manufacturers protocol. Isolated NK cells were culture at 1 × 10 6 /ml in T cells media with high dose IL-2 (1000U/ml) human IL-2 (Peprotech, UK) for 4 days. After four days cells were washed and used in an LDH release assay (Promega) following the manufacturers protocol. YAC-1 cells (a kind gift from Professor Gillian Griffiths, University of Cambridge, Cambridge, UK) were used as target cells in the assay. YAC-1s were maintained in RPMI, 10% fetal calf serum and 10U/ml Penicillin Streptomycin. 10 4 targets were used at a starting effector:target ratio of 5:1.

Infection studies
Infection of mice was carried out as part of the pathogen-screening pipeline in the mouse genetics project (5). Age and sex matched Nbeal2 +/+ or Nbeal2 -/mice were infected i.v. with 0.2 ml Salmonella enterica serovar Typhimurium M525 (phoN::tetC) containing 5 x 10 5 CFU of bacteria in sterile PBS. Mice were monitored as outlined in the mouse genetics project and results were made available through the International Mouse Phenotyping Consortium (http://www.mousephenotype.org/). For infection with virulent Salmonella enterica serovar Typhimurium SL1344, mice were orally infected with 10 9 CFU of bacteria. For infection with Staphylococcus aureus, mice were infected i.v. with 4 x 10 7 CFU of bacteria as established previously (6). MCMV was cultured as described before (3), and 3 x 10 4 PFU of virus was used to infect mice i.p. For neutrophil depletion studies, mice were administered with 300µg InVivoPlus anti-mouse Ly6G (Clone: 1A8) or InVivoPlus Rat IgG2a Isotype control (both BioXCell, USA) i.p. on days -1 and +1 post infection. Mice were sacrificed on day 3 post-infection. NK cells were depleted as described previously (7). In brief, mice were administered with 250µg InVivoMAb anti-mouse NK1.1 (Clone: PK136) or InVivoMAb Mouse IgG2a Isotype control (Clone: C1.18.4) i.p on days -2, 0 and +2 post-infection. Mice were sacrificed as before on day 4 post-infection. For all infections mice were monitored daily for changes in weight and mice were culled if they lost more than 20% of their body weight. Otherwise, mice were sacrificed at time points indicated in the main text and spleen, liver, kidneys, blood, caecal contents were harvested/collected. Bacterial CFUs for each tissue was quantified using serial dilution and plating onto agar plates. MCMV titres were calculated as described previously (7).

Histology
Mouse kidneys from S. aureus infected isotype control treated Nbeal2 +/+ and Nbeal2 -/-(described in Infection Studies) were removed and immediately fixed in 10% formalin when the mice were culled. Kidney's were processed, sectioned (2µm thickness), stained (haematoxylin and eosin) and analysed as part of the core histology service at the Metabolic Research Laboratories (Addenbrooke's, Cambridge).

Statistics
Statistical analysis was carried out using GraphPad Prism (Prism version 5.00 for Mac OS X, GraphPad Software, La Jolla California USA, www.graphpad.com). Mann-Whitney U unpaired t-tests (2-tailed), Kruskal-Wallis one-way ANOVAs, 2-way ANOVAs or log-rank tests were used as indicated in the figure legends. Significance threshold of P = 0.05 was used throughout.
(arrowheads) (X40). III shows focal abscess formation (arrowhead) and surrounding area of acute tubular necrosis (double arrowheads) (X200). IV shows an increased magnification of a bacteria-rich inflammatory lesion with monocytic and neutrophilic infiltrate (X400). (B) In vitro phagocytosis of fluorescently labelled S. aureus by MACS purified bone marrow neutrophils from either C57BL/6 control (top panels) or Nbeal2 -/mice (bottom panels). Representative fluorescence microscopy images are shown together with FACS quantification of phagocytosed bacteria. (n = 3-4). (C) Bone marrow neutrophils were stimulated with serum-opsonised S. aureus at a ratio of 20:1 (bacteria:cells) and reactive oxygen species were measured using the chemiluminescent substrate lucigenin. Graph shows the area under the curve. (n = 3) (D) Nbeal2 +/+ and Nbeal2 -/mice were treated with an isotype control antibody or an anti-Ly6G antibody (1A8) that specifically depletes neutrophils (9), on days -1, and +1 post-infection with 4 × 10 7 of S. aureus sh1000. Mice were sacrificed on day 3 post-infection and bacterial counts were determined from blood, spleen, and kidney (n = 10-13). Data shows mean and s.e.m. (C) or median (D). Data in (D) are representative of 2 pooled independent experiments. * P < 0.05, 2-tailed student's t-test (C) or 2-tailed Mann-Whitney U test (D). Figure 7 NK functionality, CD8 cytotoxicity and longer term in vivo mCMV experiments (A) Splenic NK cells (CD3-, B220-, NK1.1+, NKp46+) were stained intracellularly for LAMP-1 as described in Materials and Methods. The geometric mean fluorescence is shown for both Nbeal2 +/+ and Nbeal2 -/-NK cells (n=5-7). (B) Effect of NK depletion on mCMV infection. Mice were treated with an isotype control antibody or an anti-NK1.1 antibody on days -2, 0 and +2 post-infection, and infected as before. On day 4 mice were culled and viral plaque assays were used to determine PFUs in the lung and spleen of infected mice. n = 6. (C) Splenocytes from Nbeal2 +/+ or Nbeal2 -/mice were used in a cytotoxicity assay that measures lactose dehydrogenase release by the P815 target cell line, as described in the Methods. Displayed is the % cytotoxicity at different cell:target ratios. n = 3.