Preserved effector functions of human ORAI1 and STIM1 deficient neutrophils

To the editor,

The endoplasmic reticulum (ER) calcium sensor STIM1 and the calcium channel ORAI1 are the core components of store-operated calcium entry (SOCE), an essential pathway of lymphocyte activation. In SOCE, engagement of immunostimulatory receptors triggers Ca²⁺ release from the ER through inositol 1,4,5 trisphosphate (InsP₃) formation. The Ca²⁺-sensing ER proteins STIM 1 and 2 subsequently induce opening of Ca²⁺ release-activated Ca²⁺ (CRAC) channels in the plasma membrane, which are composed of ORAI proteins (ORAI1-3).

Inherited mutations in ORAI1 or STIM1 genes, which abolish SOCE, cause a combined immunodeficiency syndrome accompanied by various non-hematopoietic manifestations¹. Whereas the T cell defect is well established as a key mechanism mediating the immunophenotype of SOCE-deficient patients, the contribution of dysfunctional innate immunity to the broad susceptibility to infections in affected patients is unclear. In particular, data on human polymorphonuclear neutrophils (PMN) derived from SOCE-deficient patients are lacking². Studies in mice and in human neutrophil-like cell lines suggest a substantial role for SOCE in antimicrobial PMN functions, including phagocytosis, degranulation and superoxide (ROS) production^{3, 4}. In contrast, other reports implicate non-SOCE mechanisms like receptor-operated calcium entry (ROCE) - which is directly mediated by cell surface receptors and independent of ER stores - as important mediators of PMN activation⁵. A more detailed understanding of SOCE in PMN function is of obvious importance, first, for understanding the susceptibility to the broad spectrum of microorganisms in in SOCE-deficiency and second, since SOCE inhibition in PMN has been proposed as an anti-inflammatory treatment strategy.

Here, we analyzed Ca²⁺ signals and effector functions in PMN from two patients with loss-of-function mutations in ORAI1 and STIM1 genes. Surprisingly, and in contrast to T cells, Ca²⁺ influx was only mildly reduced in ORAI1- and STIM1-deficient PMN, and detailed studies did not reveal overt defects in PMN function. Accordingly, in contrast to mice, STIM1 and ORAI1 do not appear to be critical for various effector functions of human PMN.

Material & Methods

Human PMN were isolated from EDTA blood by Biocoll density gradient centrifugation. Intracellular Ca²⁺ mobilization was measured by flow cytometry after labeling with Indo-1 (Invitrogen) and stimulation with formylated methionyl-leucylphenylalanine (fMLP, 100 µM), heat-fixed streptococci (10⁸ CFU/ml), ionomycin (2 µM) or thapsigargin (1 µM). The functional PMN assays (adhesion, chemotaxis, IL-8 production, ROS formation) were described earlier⁶. Phagocytosis of fluorophore-stained heat-fixed streptococci was measured by flow cytometry after PMN fixation with paraformaldehyde and trypan blue quenching. More details are provided in the online repository.

Informed consent was obtained from the patients’ families in accordance with the Declaration of Helsinki and Institutional Review Board approval from the University of Freiburg Ethics Committee.

Results and discussion

First, we compared the expression levels of ORAI and STIM homologues between PMN, B and T lymphocytes using publicly available RNA sequencing data from healthy individuals⁷. ORAI1 and STIM1 were similarly expressed in PMN and lymphocytes. Moreover, mRNA levels of the different ORAI and STIM homologues are comparable across leukocyte subsets (Supplementary Figure 1).

To investigate the functional role of ORAI1 and STIM1 in primary human PMN, we analyzed Ca²⁺ influx and effector functions in PMN isolated from a previously published patient with a homozygous missense mutation in ORAI1 (ORAI1 p.R91W)⁸. The patient underwent hematopoietic stem cell transplantation (HSCT) without conditioning from a matched related donor at the age of four months. Subset specific mixed chimerism was found 6 months after HSCT. This study was performed when the patient was 19 years old (Supplementary Figure 2). At this point, short tandem repeat analysis showed complete autologous reconstitution of his granulocyte compartment (Supplementary Table 1, sensitivity ≥ 95%), and his PMN counts were within normal limits (data not shown). As expected, T cell receptor engagement of CD4⁺ and CD8⁺ T cells showed a complete inhibition of SOCE (Supplementary Figure S3). We analyzed Ca²⁺ influx in PMN after agonist stimulation with streptococci and fMLP. Unexpectedly, both signals induced a strong Ca²⁺ influx in ORAI1-deficient PMN, which was only mildly reduced compared to control PMN (Figure 1A). Next, we stimulated PMN with the sarco/endoplasmic reticulum Ca²⁺-ATPase inhibitor thapsigargin and with ionomycin, both of which deplete ER Ca²⁺ stores and thereby trigger SOCE bypassing immunoreceptors. Again, only a modest reduction of SOCE could be observed in ORAI1 p.R91W PMN after both stimuli. Quantification of calcium flux as an area under the curve (AUC) revealed a reduction in the ORAI1-deficient patient in response to fMLP and ionomycin, but not in response to streptococci or thapsigargin (Figure 1A).

Figure 1. Robust Ca²⁺ influx and preserved effector functions in activated human ORAI1-deficient (ORAI1 p.R91W) PMN.

(A) Ca²⁺ influx in PMN from a patient with ORAI1 p.R91W loss-of-function mutation (orange) and a healthy donor (control, black) after stimulation with fMLP, streptococci, ionomycin and thapsigargin (TG). Each line graph represents average intracellular Ca²⁺ levels (F409/F519) from independent blood samples and repeat experiments (indicated by n). The bar graphs represent (top) the fold increase of average peak Ca2+ levels normalized to baseline Ca2+ levels (+/-standard errors) and (bottom) the area under the curve (AUC) of the Ca2+ response in arbitrary units (a.u.) after addition of the different stimuli.. (B) IL-8 ELISA of patient and control PMN after stimulation for 5 hours as indicated. (C) Measurement of reactive oxygen species (ROS) in patient and control PMN after 90 minutes of stimulation with the indicated stimuli. (D) Static PMN adhesion after stimulation in a 96 well chamber as indicated for 30 min. (E) Migration of calcein-A-labeled PMN towards the indicated chemoattractants fMLP (100 nM), PAF (100 nM) and C5a (1 nM) over 55 min in a modified Boyden chamber. Depicted is the slope of PMN migration. (F) Phagocytosis of ORAI1-deficient neutrophils after coincubation with fluorophore-labeled streptococci as determined by flow cytometry. C5a, complement factor 5a; LPS, lipopolysaccharide; PAF, platelet-activating factor; PMN, polymorphonuclear neutrophils; PMA, phorbol-12-myristate-13-acetate; TNFa, TNFalpha.

Finally, FcγRIII (CD16) crosslinking also induced Ca²⁺ influx in ORAI1 p.R91W PMN that was only moderately reduced compared to control PMN (Supplementary Figure S4). Since cytosolic Ca²⁺ concentrations are considered to be essential for antibacterial PMN functions, we asked whether ORAI1 deficiency impaired the PMN response to various microbial and non-microbial ligands. IL-8 production was unaffected, both in response to soluble ligands for TLRs 2, 4, 5 and 6, and to whole E.coli and streptococci (Figure 1B). Furthermore, ORAI1 p.R91W PMN were indistinguishable from control PMN with respect to formation of ROS, static cell adhesion, chemotaxis and phagocytosis of fluorophore-labeled streptococci (Figure 1C-F).

To confirm our findings, we analyzed SOCE and inflammatory functions in PMN from a STIM1-deficient patient. In this patient, a missense mutation in STIM1 (STIM1 p.R429C) was shown to abolish SOCE in T cells by interfering with STIM1 binding to ORAI1⁹. While the patient’s PMN showed robust peak Ca²⁺ influx in response to fMLP and streptococci similar to control cells, the Ca²⁺ response appeared less sustained (Figure 2A). STIM1-deficient PMN, however, showed normal production of IL-8 and ROS, as well as unaltered chemotaxis and static adhesion as compared to control PMN (Figure 2B-E).

Figure 2. STIM1-deficient (STIM1 p.R429C) PMN show normal Ca²⁺ influx after pattern recognition receptor activation and have unaltered antibacterial effector functions.

(A) Ca²⁺ influx in STIM1-deficient PMN after stimulation with fMLP or streptococci as described in Figure 1. (B) IL-8 ELISA in STIM1-deficient PMN after stimulation with streptococci, E.coli or Curdlan (1mg/ml) for 5 hours as compared to control PMN. (C) ROS generation by STIM1-deficient PMN after stimulation with streptococci (10⁸ CFU/ml), zymosan or PMA measured by lucigenin. (D) PMN adhesion of STIM1-deficient PMN versus control PMN measured by fluorescence after stimulation with LPS, Pam₃CSK₄, fMLP, PAF or TNFa. (E) Chemotaxis of STIM1-deficient and control PMN towards fMLP, PAF or C5a. Unless specifically indicated, the experimental procedures were identical to the assays performed with ORAI1-deficient PMN as described in Figure 1.

Our findings in primary human ORAI1- and STIM1-deficient PMN are unexpected since they contrast those in STIM1-deficient mice and knockdown of ORAI1 or STIM1 in HL-60 tumor cells. In both models, ORAI1- and STIM1-mediated SOCE was identified as an important mechanism of Ca²⁺ influx and antibacterial PMN functions^{3, 4}. By contrast, our data indicate that loss-of-function mutations in STIM1 and ORAI1, which abolish SOCE in T cells, have only a small effect on SOCE in human PMN. Our data suggest that ORAI1/STIM1-independent mechanisms, for instance SOCE mediated by STIM2, ORAI2 or ORAI3, ROCE⁵ or other yet to be identified channels, contribute to Ca²⁺ influx in human PMN activation. Importantly, ORAI1 and STIM1 deficiency did not cause an overt defect in any of the PMN effector functions we analyzed. At least in our patients, PMN dysfunction does not seem to be a dominant contributor to their immunodeficiency.