Circulating membrane aminophospholipids contribute to thrombotic risk in rheumatoid arthritis

Abstract

Patients with rheumatoid arthritis (RA) are at elevated risk of thrombotic events, yet the underlying mechanisms remain unknown. The contribution of the procoagulant membrane surface provided by aminophospholipids (aPLs) in driving thrombotic risk in RA has not been investigated. Specifically, neither the type of aPL exposed on circulating blood cell membranes in patients is characterized nor is their ability to support thrombin generation is known. Here, lipidomics was used to characterize the external-facing and total levels of aPL molecular species in RA, specifically phosphatidylserine and phosphatidylethanolamine on extracellular vesicles (EVs), platelets, and white blood cells (WBCs). The ability of the cells and EVs to support thrombin generation from patients and healthy controls was compared using an in vitro prothrombinase assay. RA patient plasma had significantly higher levels of thrombin-antithrombin and d-dimers, indicating increased thrombotic activity in vivo. Higher EV and platelet counts were seen in RA, but WBC counts were not elevated. EVs from RA patients supported higher levels of thrombin generation compared with healthy controls, whereas for platelets and WBC, thrombin generation was similar for both groups. EVs from RA patients also showed elevated external-facing phosphatidylserine molecular species, with total aPL also increased. For platelets and WBC, total and external-facing aPL levels were similar. Thrombin-antithrombin (TAT) complexes significantly correlated with EV particle counts, indicating that their circulating numbers are directly related to coagulation in vivo. Overall, our data suggest that elevated plasma EV levels in RA are a major source of procoagulant membranes, contributing to thrombotic risk in RA.

Keywords: aminophospholipids; extracellular vesicles; lipidomics; rheumatoid arthritis; thrombosis.

Graphical abstract

Figure 1

RA patients show higher coagulation markers and EV counts in plasma, and their EV can support increased in vitro thrombin generation. A and B: d-dimer levels are raised in RA patients. d-dimer levels were measured as described in Materials and methods section, using ELISA in plasma from RA patients (n = 21) and HC (n = 16), showing absolute levels (ng/ml) and frequency of raised levels (%) above clinical cutoff of 500 ng/ml shown. C: TAT complexes are increased in plasma from RA patients (22). TAT complexes were measured using ELISA in plasma from RA patients (n = 25) and HC (n = 19). D: EVs from RA patients generate more thrombin. The ability of EVs in 6 ml plasma to support coagulation reactions was assessed using the prothrombinase assay, as described in Materials and methods section, for HC (n = 24) and RA patients (n = 22). E and F RA patients have increased plasma EV counts but no increase in diameter. EV particles were counted, from HC (n = 24) and RA (n = 24) plasma, as described in Materials and methods section. G: Thrombin generation is increased in RA plasma because of increased EV levels. Prothrombinase assay, using 10¹⁰ EV particles, was applied to plasma from HC (n = 24) and RA (n = 21). Data were analyzed using Mann-Whitney test (∗P < 0.05, ∗∗P < 0.001, ∗∗∗P < 0.001). H: EV particle counts correlate with thrombin generation and TAT levels in RA. Spearman correlation was performed comparing EV particle counts with TAT complexes (n = 22) in RA plasma. I: Heatmap for external-facing PE and PS levels in RA and HC plasma. EV particles were isolated followed by biotinylation of externalized lipids, as described in Materials and methods section. Lipids were extracted from plasma from HC (n = 19) and RA patients (n = 21) as described in Materials and methods section and analyzed by LC-MS/MS. Externalized aPLs in EV are shown in a heatmap (log10, ng/ml). J: Total aPLs are increased in EVs from RA. Total (internal + external) individual aPLs in EVs were analyzed using LC-MS/MS and shown in a heatmap (log10, ng/ml). K and L: Externalized PS was barely detected in EVs from RA plasma and undetectable in HC, whereas external-facing PE was similar in both groups. Externalized PS and PE levels were determined (ng/ml) for HC (n = 19) and RA (n = 21), using LC-MS/MS as described in Materials and methods section. M and N: Total PS and PE levels are increased in EVs in RA. Total (external and internal) aPLs were quantified as described in Materials and methods section or HC (n = 19) and RA (n = 22) (ng/ml of plasma). O–Q: Externalized PE is decreased in EVs from RA patients following normalization of EV count. Total aPL concentrations were adjusted to EV counts (10¹⁰ EV particles), for HC (n = 18) and RA (n = 21), along with externalized PE for HC (n = 19) and RA patients (n = 20) (ng/10¹⁰ EV particles). Data were analyzed using the Mann-Whitney test (∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001).

Figure 2

Platelet counts in RA are elevated, whereas PE externalization and thrombin generation are similar. A: Platelet count is increased in RA patients.²² Platelets were counted as described in Materials and methods section, HC (n = 23) and RA (n = 26). Data were analyzed using the Mann-Whitney test (∗∗P < 0.01). B and C: Thrombin generation by platelets is similar between RA and HC. Around 2 × 10⁸ platelets from HC (n = 25) and RA patients (n = 23) were assessed using the prothrombinase assay and expressed either on a per platelet basis or per blood platelet count from HC (n = 23) and RA patients (n = 23). Data were analyzed using Student's t-test. D: Externalized aPL profile is similar for RA and HC platelets. Lipids were extracted from resting platelets or following thrombin activation (0.2 U/ml). Externalized lipids were analyzed as described in Materials and methods section using LC-MS/MS. Platelet externalized aPL species, either resting or thrombin activated, from HC (n = 19) and RA patients (n = 22 and n = 23, respectively), were analyzed using LC-MS/MS and shown in heatmaps (log10, ng/2 × 10⁸ platelets). E and F: Externalized PS and PE levels are similar in resting or activated platelets from RA patients and HC. The sum of externalized PS in resting or activated platelets, along with externalized PE, was calculated, in HC (n = 19 for both) and RA patients (n = 22 and n = 23, respectively) (ng/2 × 10⁸ platelets). Data were analyzed using Kruskal-Wallis test and Dunn's multiple comparisons test (∗∗∗∗P < 0.0001). G: Platelets from RA patients contain less total aPLs. Total individual aPLs species in platelets, either resting or thrombin activated, from HC (n = 13 and n = 14) and RA patients (n = 20 and n = 21, respectively), were analyzed using LC-MS/MS, and are shown in heatmaps (log10, ng/2 × 10⁸ platelets). H and I: Resting platelets from RA patients have less total PE. The sum of total analyzed PS and PE was determined for both resting and thrombin-activated platelets, from HC (n = 13 and n = 14, respectively) and RA patients (n = 20 and n = 21, respectively), (ng/2 × 10⁸ platelets). Data were analyzed using the Kruskal-Wallis test and Dunn's multiple comparisons test.

Figure 3

WBC counts are similar in RA and HC, whereas WBC from RA support similar thrombin generation and contain similar levels of total and external aPL to HC. A: WBC counts were similar between RA patients and HC (22). WBCs were counted in samples from HC (n = 25) and RA patients (n = 26), as described in Materials and methods section. B: Thrombin generation is similar for WBC from RA patients and HC. WBCs were tested for their ability to support prothrombinase activity, as described in Materials and methods section from HC (n = 24) RA (n = 22). Data were analyzed using Student's t-test. C: Externalization of aPLs in resting or activated WBCs from RA patients and HC is similar. Lipids were extracted from resting and ionophore-activated (10 μM) WBC, followed by analysis of externalized lipids as described in Materials and methods section using LC-MS/MS. Externalized individual aPL species in WBC, either resting or activated, from HC (n = 20 and n = 19, respectively) and RA patients (n = 23 and n = 21, respectively), were analyzed using LC-MS/MS, and shown in heatmaps (log10, ng/4 × 10⁶ WBC). D and E: Levels of external-facing PS and PE in WBC are similar for RA patients and HC. The sum of externalized PS and PE in resting and ionophore-activated WBC, in both HC (n = 20 and n = 19, respectively) and RA patients (n = 23 and n = 21, respectively), was calculated (ng 4 × 10⁶ WBs). Data were analyzed using the Kruskal-Wallis test and Dunn's multiple comparisons test (∗∗∗∗P < 0.0001). F: Total amount of aPLs in resting or activated WBC from RA patients and HC is similar. Total amounts of individual aPL species in WBC, either resting or activated, from HC (n = 15 for both) and RA patients (n = 21 for resting and n = 22 for activated), were analyzed using LC-MS/MS, and are shown in heatmaps (log10, ng/4 × 10⁶ WBC). G and H: Total PS and PE levels are similar for resting or activated WBC in RA patients and HC. The sum of total PS, along with total PE, was determined as described in Materials and methods section, in WBC, either resting or following ionophore activation, HC (n = 15 for both) and RA patients (n = 21 for resting and n = 22 for activated) (ng/4 × 10⁶ WBC). Data were analyzed using the Kruskal-Wallis test and Dunn's multiple comparisons test.