The CalDAG-GEFI/Rap1/αIIbβ3 axis minimally contributes to accelerated platelet clearance in mice with constitutive store-operated calcium entry

Abstract Circulating platelets maintain low cytosolic Ca2+ concentrations. At sites of vascular injury, agonist-induced Ca2+ release from platelet intracellular stores triggers influx of extracellular Ca2+, a process known as store-operated Ca2+ entry (SOCE). Stromal interaction molecule 1 (Stim1) senses reduced Ca2+ stores and triggers SOCE. Gain-of-function (GOF) mutations in Stim1, such as described for Stormorken syndrome patients or mutant mice (Stim1 Sax ), are associated with marked thrombocytopenia and increased platelet turnover. We hypothesized that reduced platelet survival in Stim1 Sax/+ mice is due to increased Rap1/integrin signaling and platelet clearance in the spleen, similar to what we recently described for mice expressing a mutant version of the Rap1-GAP, Rasa3 (Rasa3 hlb/hlb ). Stim1 Sax/+ mice were crossed with mice deficient in CalDAG-GEFI, a critical calcium-regulated Rap1-GEF in platelets. In contrast to Rasa3 hlb/hlb x Caldaggef1 −/- mice, only a small increase in the peripheral platelet count, but not platelet lifespan, was observed in Stim1 Sax/+ x Caldaggef1 −/- mice. Similarly, inhibition of αIIbβ3 integrin in vivo only minimally raised the peripheral platelet count in Stim1 Sax/+ mice. Compared to controls, Stim1 Sax/+ mice exhibited increased platelet accumulation in the lung, but not the spleen or liver. These results suggest that CalDAG-GEFI/Rap1/integrin signaling contributes only minimally to accelerated platelet turnover caused by constitutive SOCE. Plain Language Summary What do we know? Platelets are small blood cells which act to prevent blood loss, which circulate in a resting state but are rapidly activated upon exposure to ligands at the site of vascular injury Calcium (Ca2+) is critical for platelet activation, especially for activation of integrins which support platelet–platelet interactions If platelet activation occurs in circulation, platelets can be prematurely cleared from blood and unable to function in hemostasis Disorders of Ca2+ dysregulation such as Stormorken syndrome are associated with reduced platelet counts (thrombocytopenia) and bleeding What did we discover? We used a mouse model expressing a mutation causing higher Ca2+ levels in cells including platelets (Stim1 Sax ), and investigated whether thrombocytopenia is due to stimulation of a specific pathway for integrin activation, mediated by a protein called Rap1 GTPase We crossed Stim1Sax mice with mice lacking an important activator of Rap1, the Ca2+-regulated protein CalDAG-GEFI, and saw no major improvement in thrombocytopenia We also observed more Stim1Sax platelets in the lung but not the liver or spleen, in contrast to mice with activation of platelet integrins in circulation What is the impact? Our results rule out activation of the CalDAG-GEFI/Rap1/integrin pathway as a major cause of thrombocytopenia in Stim1Sax mice Our findings help to narrow down potential causes of thrombocytopenia in disorders such as Stormorken syndrome


Plain Language Summary
What do we know?• Platelets are small blood cells which act to prevent blood loss, which circulate in a resting state but are rapidly activated upon exposure to ligands at the site of vascular injury • Calcium (Ca 2+ ) is critical for platelet activation, especially for activation of integrins which support platelet-platelet interactions • If platelet activation occurs in circulation, platelets can be prematurely cleared from blood and unable to function in hemostasis • Disorders of Ca 2+ dysregulation such as Stormorken syndrome are associated with reduced platelet counts (thrombocytopenia) and bleeding What did we discover?
• We used a mouse model expressing a mutation causing higher Ca 2+ levels in cells including platelets (Stim1 Sax ), and investigated whether thrombocytopenia is due to stimulation of a specific pathway for integrin activation, mediated by a protein called Rap1 GTPase • We crossed Stim1 Sax mice with mice lacking an important activator of Rap1, the Ca 2+ -regulated protein CalDAG-GEFI, and saw no major improvement in thrombocytopenia • We also observed more Stim1 Sax platelets in the lung but not the liver or spleen, in contrast to mice with activation of platelet integrins in circulation

Introduction
Calcium (Ca 2+ ) signaling plays a multitude of roles in cellular function, mainly by modifying the activity of Ca 2+ -binding proteins [1,2].In platelets, maintaining low intracellular Ca 2+ concentrations ([Ca 2+ ] i ) is critical for preventing untimely platelet activation in circulation, but upon vascular injury a rapid increase in [Ca 2+ ] i is required for platelet adhesion and aggregation [3].Platelets, like many cell types, utilize store-operated calcium entry (SOCE) as a source of robust Ca 2+ influx following agonist stimulation.Receptor stimulation, such as activation of protease activated receptors (PARs) by thrombin or glycoprotein VI (GPVI) by collagen, leads to phospholipase C activation and generation of IP 3 which triggers release of Ca 2+ stores from the platelet dense tubular system (DTS).Stim1 (stromal interaction molecule 1) is a single pass transmembrane protein residing in the DTS membrane which senses Ca 2+ store concentrations ([Ca 2+ ] s ) through an N-terminal EF hand domain [4].Upon store release and reduced [Ca 2+ ] s , Ca 2+ dissociation from the EF hand causes a conformational change in Stim1, which subsequently clusters proximal to the plasma membrane (PM) and interacts with the plasma membrane Ca 2+ channel Orai1 to induce Ca 2+ influx [5].
An important Ca 2+ -binding protein in platelets is Calciumand DAG-regulated Guanine Nucleotide Exchange Factor 1 (CalDAG-GEFI) which activates the small GTPase Rap1, a central regulator of αIIbβ3 integrin activation in platelets [6].CalDAG-GEFI is a cytosolic protein with dual EF hands which bind Ca 2+ with high affinity (K d ~ 80 nM) [7].Ca 2+ binding induces the release of autoinhibition in CalDAG-GEFI [8].CalDAG-GEFI is counteracted by the GTPase activating protein (GAP), Rasa3, which is active in circulating platelets to maintain Rap1 in its inactive GDP-bound state [1].P2Y12 receptor signaling inactivates Rasa3, allowing for greater and sustained Rap1-GTP binding, and both activation of CalDAG-GEFI and inactivation of Rasa3 are required for the full extent of αIIbβ3 activation required for hemostasis [9].Importantly, loss of Rasa3 promotes activation of platelet integrins in circulation, leading to rapid platelet clearance in spleen and liver [10].Mice with a homozygous point mutation in Rasa3 (Rasa3 hlb/hlb ) have approximately 5% platelets remaining in circulation, while mice with global or platelet-specific knockout are virtually platelet-free, which results in fetal hemorrhage and lethality [1].Both the thrombocytopenia and embryonic lethality associated with mutations in or loss of Rasa3 were corrected by concomitant deletion of CalDAG-GEFI, revealing the critical role of Rap1 signaling in platelet homeostasis.
Gain-of-function (GOF) mutations in Stim1 or Orai1 cause thrombocytopenia in humans [11].Stormorken syndrome, caused by Stim1 GOF, is an autosomal dominant disease associated with thrombocytopenia and increased risk of bleeding [12].The cause of thrombocytopenia in Stormorken syndrome is not fully understood.A mouse model generated with a mutation commonly observed in Stormorken syndrome, R304W [13], reproduces much of the clinical phenotype including muscle pathologies and thrombocytopenia [14].Additionally, a mouse model derived from a mutagenesis screen was described with a heterozygous point mutation in the EF hand of Stim1 (D84 G, "Stim1 Sax/+ "), leading to constitutive Ca 2+ entry and severe thrombocytopenia [15].Homozygosity for this mutation in mice causes embryonic hemorrhage and lethality.Interestingly, some of these phenotypes (thrombocytopenia, splenomegaly, embryonic bleeding) are reminiscent of mice lacking Rasa3 [1,10].Therefore, we hypothesized that increased platelet [Ca 2+ ] i due to Stim1 GOF causes platelet pre-activation and clearance by triggering CalDAG-GEFImediated Rap1/integrin signaling.

Methods
Mice: Stim1 Sax/+ mice were previously described [15] and crossed with Caldaggef1 -/-("Cdg1 -/-") mice [16] to obtain Stim1 Sax/+ x Cdg1 +/+, ± or -/-mice.Littermate controls were wild-type for Stim1 with mixed Cdg1 status for platelet counts and lifespan, and wild-type for Cdg1 for tissue immunofluorescence.Stim1 fl/fl x Pf4-Cre + mice were previously described [17].All mice were on a C57BL/6J background, and both male and female mice between 8 and 16 weeks were used for experiments.All experimental protocols were approved by University of North Carolina IACUC.
Platelet count and lifespan: Platelet counts and platelet lifespan in circulation were determined as previously described [10].To determine the impact of in vivo αIIbβ3 blockade on platelet counts, mice received two intravenous (IV) injections of anti-αIIbβ3 F(ab') 2 antibody (75 μg) 24 h apart [10].
Statistics: Data shown as mean ± SD.Platelet count data was analyzed using one-way ANOVA with Bonferroni's test for multiple comparisons.Platelet lifespan data was analyzed using two-way ANOVA with Bonferroni's test for multiple comparisons.Tissue immunofluorescence was analyzed by Student's t-test for each organ.Data were analyzed using GraphPad Prism (version 9.4).P < .05 was considered statistically significant.

Results
Stim1 Sax/+ mice had platelet counts approximately 30-40% of littermate controls and significantly reduced platelet lifespan in circulation (Figure 1a,b).To investigate a potential contribution of increased Rap1 signaling to accelerated platelet turnover, we crossed Stim1 Sax/+ mice with Caldaggef1 -/- (Cdg1 -/-) mice to obtain Stim1 Sax/+ mice with 2, 1 or 0 copies of CalDAG-GEFI.We did not observe any surviving homozygous Stim1 Sax/Sax pups, regardless of CalDAG-GEFI status.Homozygous deletion of CalDAG-GEFI led to a small but significant improvement in the peripheral platelet count (Figure 1a) but not platelet lifespan (Figure 1b) compared to Stim1 Sax/+ x Cdg1 +/+ mice.CalDAG-GEFI deficiency was previously shown to impair Ca 2+ -dependent platelet functions to a greater extent than Stim1 deficiency [18].To determine the requirement of Stim1 and SOCE for Rap1 activation, we performed a Rap1-GTP pulldown assay in Stim1 fl/fl x Pf4-Cre + or Cre-platelets.Rap1-GTP levels following activation were partially reduced in Stim1-deficient platelets, both in the absence and presence of a P2Y12 inhibitor (Figure 1c), suggesting that Stim1-mediated Ca 2+ influx contributes to but is not absolutely required for Rap1 activation.To determine whether αIIbβ3 mediates clearance of Stim1 Sax/+ platelets, mice were injected with a blocking F(ab') 2 antibody against αIIbβ3, which increased circulating platelet counts in Rasa3 hlb/hlb mice by more than 5-fold [10].Blockade of αIIbβ3 in vivo also increased platelet counts in Stim1 Sax/+ (Figure 1d) but to a much lesser extent than in Rasa3 hlb/hlb mice.To investigate potential sites of increased platelet clearance, we immunostained spleen, liver and lung tissue sections from Stim1 Sax/+ mice and littermate controls and determined the extent of platelet accumulation in these organs.We observed no difference in the clearance of Stim1 Sax/+ platelets in the spleen (Figure 2a,d) or liver (Figure 2b,d), but did observe greater platelet accumulation in lung (Figure 2c,d) compared to controls.Taken together, these findings strongly suggest that the mechanism underlying increased platelet clearance and thrombocytopenia in Stim1 GOF mice is almost entirely distinct from that of unrestricted Rap1/integrin signaling in Rasa3 mutant mice.

Discussion
Gain-of-function mutations in Stim1 or Orai1 cause multi-systemic syndromes affecting muscle, immune cells and platelets [19].Stormorken syndrome is caused by pathogenic Stim1 variants and is associated with thrombocytopenia and/or platelet dysfunction [20], but the cause of thrombocytopenia in Stormorken patients remains unknown.Stim1 Sax/+ mice have severe thrombocytopenia with increased platelet turnover, megakaryocyte hyperplasia but normal proplatelet formation in vitro, enlarged spleens, and homozygosity for the Sax mutation causes embryonic bleeding and lethality [15].In our recent studies, we observed a very similar phenotype in mice expressing a hypomorphic mutant Rasa3, the critical Rap1 GAP in platelets [1,10].Deletion of CalDAG-GEFI, a Rap GEF with very high affinity for Ca 2+ , markedly improved the peripheral platelet count and rescued survival of Rasa3 hlb/hlb mice.Additionally, pharmacological inhibition of SOCE normalized platelet counts and lifespan in a different Stim1 GOF mouse model [21] demonstrating that Ca 2+ entry is responsible for enhanced platelet clearance.Based on these studies, we hypothesized that increased SOCE may lead to platelet clearance and thrombocytopenia due to the activation of CalDAG-GEFI/Rap1/integrin signaling.The data presented here is not consistent with this hypothesis as platelet survival and peripheral counts in Stim1 Sax/+ mice were not markedly improved by CalDAG-GEFI deletion or αIIbβ3 inhibition.Furthermore, deletion of CalDAG-GEFI did not improve the viability of homozygous Stim1 Sax/Sax mice.It is important to note that the [Ca 2+ ] i in isolated Stim1 Sax/+ platelets was measured at a concentration similar to the K d for calcium of CalDAG-GEFI (~80 nM) [15].It seems likely that [Ca 2+ ] i in cleared Stim1 Sax/+ platelets is higher than in circulating cells, i.e. at levels that should lead to significant CalDAG-GEFI activation.Our results suggest that increased SOCE does not cause enough CalDAG-GEFI activation, however, to outcompete the negative feedback provided by Rasa3 or that it triggers both CalDAG-GEFI/Rap1/integrin-dependent and -independent clearance of platelets.Protein kinase C (PKC) also regulates Rap1 activation, although mainly through secretion of ADP and inactivation of Rasa3 [1], but we cannot rule out a direct activation of Rap1 through Ca 2+ -sensitive PKCs.Most likely, the low platelet counts in Stim1 Sax/+ x Caldaggef1 −/-mice or integrin inhibitor-treated mice would be explained by another calciumdependent clearance mechanism.For example, high and sustained [Ca 2+ ] i is a known prerequisite for platelet phosphatidylserine (PS) surface exposure, an "eat-me" signal for phagocytic cells [22].Our findings in Stim1 knockout platelets support this idea as they demonstrated a critical role for SOCE in platelet PS exposure but a lesser contribution to Rap1 activation and integrin-mediated adhesion [18].PS externalization is regulated by TMEM16F in procoagulant platelets [23] and by TMEM16F-independent apoptotic signaling [24], potentially involving Xk-related protein 8 (Xkr8) [25].The physiological relevance of PS as a clearance signal for platelets in general and in the case of constitutive SOCE, however, still needs to be confirmed.Desialylation of surface receptors can also serve as a signal for clearance by liver macrophages [26].Increased numbers of platelets, however, were observed in the lung but not the liver of Stim1 Sax/+ mice.The fact that Stim1 Sax/+ platelets were enriched in the lung is interesting, as platelet recruitment to the lung can be observed in severe pulmonary disease [27] Thus, future studies on the mechanism(s) underlying increased platelet turnover in pathologies with constitutive SOCE could be of high clinical relevance.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This work was supported by the National Institutes of Health under Grant R35 HL144976 (to WB); American Society of Hematology under an ASH Scholar Award (to RHL); and AABB Foundation under an Early-Career Scientific Research Grant (to RHL).

Figure 1 .
Figure 1.Minor role for CalDAG-GEFI/Rap1/αIIbβ3 integrin activation in clearance of Stim1 Sax/+ platelets.(a) Platelet counts in Stim1 Sax/+ mice crossed with Cdg1 -/-mice.Platelet counts were determined in whole blood samples by flow cytometry.(b) Platelet lifespan in Stim1 Sax/+ mice crossed with Cdg1 -/-mice.All circulating platelets were labeled at t = 0 by injection of anti-GPIX antibody, and the percentage of GPIX+ platelets remaining at t = 24 and 48 hours was determined by flow cytometry.(c) Rap1-GTP pulldown assay in Stim1-deficient platelets.Washed platelets from Stim1 fl/fl Pf4-Cre + mice or Cre-littermate controls were activated with convulxin (Cvx) or Par4 activating peptide (Par4p) in the presence or absence of the P2Y12 inhibitor 2-MesAMP.Platelets were lysed, and lysate was incubated with RalGDS-RBD beads to pulldown GTP-bound Rap1.Active and total Rap1 levels were determined by SDS-PAGE and western blot.Normalized densitometry values for Rap1-GTP/total Rap1 bands are listed above the blot.Representative blot of n = 2 experiments is shown.(d) Platelet counts in Stim1 Sax/+ mice treated with a αIIbβ3 blocking antibody.Stim1 Sax/+ mice were treated with anti-αIIbβ3 F(ab') 2 (75 μg once per day), and platelet counts were determined at 24 and 48 hours.Statistical significance was determined between Stim1 Sax/+ x Cdg1 +/+ , Stim1 Sax/+ x Cdg1 ± and Stim1 Sax/+ x Cdg1 +/+ mice by one-way (a) or two-way (b) ANOVA with Bonferroni post-hoc test, or by one-way ANOVA with Bonferroni post-hoc test for effect of αIIbβ3 inhibition in Stim1 Sax/+ mice (d).ns=not statistically significant, *p < .05,**p < .01.Data shown as mean ± SD.

Figure 2 .
Figure 2. Clearance of Stim1 Sax/+ platelets in spleen, liver and lung.Stim1 Sax/+ and littermate control mice were sacrificed, perfused, and organs were fixed for cryosectioning.Sections were immunostained for platelets and nuclei, and Z stacks (~10 μm) were collected using scanning confocal microscopy.Representative Z max projections are shown for spleen (a), liver (b) and lung (c) with platelets in red, nuclei in blue and autofluorescence/background in green.Scale bars = 20 µm.(d) Platelet area in Z max projections was determined using Fiji software.Statistical significance was determined by unpaired t-test.ns=not statistically significant, *p < .05.Data shown as mean ± SD.