Portal de Periódicos da CAPES

Endothelium-Neutrophil Communication via B1-Kinin Receptor–Bearing Microvesicles in Vasculitis

2017; American Society of Nephrology; Volume: 28; Issue: 8 Linguagem: Inglês

10.1681/asn.2017030300

1533-3450

Pierre‐Louis Tharaux, Neeraj Dhaun,

Complement system in diseases

The kinin system is activated during inflammation. Plasma prekallikrein is cleaved on the extracellular surface of endothelial cells or neutrophils generating plasma kallikrein. This, in turn, cleaves high molecular weight kininogen (HK) releasing bradykinin and Lys-bradykinin, with subsequent cleavage of the C-terminal arginine residue by carboxypeptidase that yields the fragments des-Arg9-bradykinin and Lys-des-Arg9-bradykinin. Kinins have multiple effects. Kinin activation of the B1 receptor (B1R) regulates local hemodynamics and inflammation and induces capillary leakage. This has been shown to promote inflammation partly through increased endothelial permeability and neutrophil motility. The C1 inhibitor (C1-INH) is at the crossroads of both the circulating complement and bradykinin-forming cascades. It is the key inhibitor of activated complement factors C1r and C1s, and it also inhibits the conversion of HK to bradykinin via prekallikrein. Furthermore, because activated factor XII is generated, it converts prekallikrein to kallikrein, which in turn, cleaves HK to release bradykinin.1 Activated factor XII is further cleaved by kallikrein or plasmin to yield factor XII fragment. Factor XII fragment can activate the C1r subcomponent of C1, particularly when C1-INH is absent. Microvesicles are released from the plasma membrane of many cell types and have a diameter of 100–1000 nm. They can be distinguished from exosomes by the presence of phosphatidylserine on their outer plasma membrane. Microvesicles can be taken up by other cells and are able to transport membrane proteins and cytosolic cargo between cells. In this issue of the Journal of the American Society of Nephrology, Mossberg et al.2 analyzed plasma from patients with active small vessel vasculitis, comparing it with that of healthy controls. They show the presence of B1R on endothelial cell–derived microvesicles (EMVs) that are shed during active vasculitis. They then studied the contribution of EMV B1R to the inflammatory process, specifically neutrophil chemotaxis. They show that plasma from patients with active small vessel vasculitis is chemoattractant to neutrophils isolated from healthy volunteers. Interestingly, the effect of patients’ plasma on neutrophil migration was reduced after EMV depletion as well as using pharmacologic blockade of B1R. Although circulating EMVs also contain IL-8, a recognized neutrophil chemokine, proof of principle for a sufficient effect of EMV B1R to attract neutrophils is shown using microvesicles derived from B1R-transfected human embryonic kidney cells. These display chemoattractant activity for neutrophils, an effect that is abrogated by B1R antagonism. The authors go on to show that plasma from patients with active small vessel vasculitis is able to stimulate cultured endothelial cells to release B1R-expressing EMVs. C1 inhibition reduced the amount of EMVs released, specifically those that were B1R positive.2 There are shared elements between the plasma bradykinin-forming and complement cascades, including transactivation, common control mechanisms, and shared binding proteins.3 Inactivation of C1-INH in inflamed tissues by proteolytic enzymes (e.g., elastase) released from activated neutrophils may potentiate both the complement- and the kinin-producing cascades as well as the proinflammatory release of EMVs. The roles of activated complement factors C1r and C1s on endothelial cell phenotype and EMV release remain unclear, and further studies will be needed to better understand the mechanisms for such a feedforward amplification system. There are a few limitations to this study. First, the authors defined EMVs by the expression of CD105+ and/or CD144+. Importantly, CD105 (endoglin) may also be expressed by monocytes. Second, Annexin V labeling might have been useful to distinguish exosomes from microvesicles. Nevertheless, CD144 is exclusive to endothelial cells. As described in previous work by the authors, the microvesicles analyzed here ranged from 8 to 451 nm in size,4 suggesting that they were most likely derived from plasma membrane. The presence of EMVs in the plasma of patients with inflammatory diseases is well recognized and has been described in atherosclerosis, rheumatoid arthritis, systemic sclerosis, SLE, and systemic vasculitides.5,6 Given these studies, it is perhaps unsurprising that Mossberg et al.2 found increased numbers of B1R-bearing EMVs in the plasma of patients with small vessel vasculitis. Also, although the authors show that plasma from patients with active small vessel vasculitis is chemoattractant to neutrophils isolated from healthy volunteers, it remains unclear what the effect would be on neutrophils from those with active vasculitis and whether plasma components other than EMVs might contribute to neutrophil chemotaxis. This study raises a number of novel pathophysiologic questions. The precise nature of the effect of EMV B1R on neutrophil function remains unclear. Previous in vitro data from the same group using both endothelial and epithelial cell lines suggest that microvesicles are able to transfer B1R from donor to recipient cells. Interestingly, using flow cytometric analysis of plasma from patients with ANCA vasculitis, Kaplan and Ghebrehiwet4 recently found high levels of leukocyte-derived microvesicles bearing B1R. Although one might hypothesize that a shuttling of B1R (or its ligand bradykinin) occurs between different cellular compartments (in this case, endothelium to leukocyte), these data are insufficient to support this. Where do these data fit more broadly with previous studies? Although B1R is expressed at low levels under physiologic conditions, its expression increases in response to a range of stimuli, including bacterial endotoxin,7 proinflammatory cytokines,8,9 and angiotensin II,7 and in pathologic conditions, such as ischemic cardiomyopathy, diabetes,8 and ischemia-reperfusion injury.10 Interestingly, B1R-null mice show a lower expression of proinflammatory molecules, such as monocyte inflammatory protein 1 and IL-6. Furthermore, in the unilateral ureteric obstruction model, B1R antagonism prevents macrophage infiltration with a reduced level of renal fibrosis.11,12 The potential therapeutic advantages of B1R antagonism are also shown by reduced inflammation in the adriamycin-induced model of FSGS.13 Together, these data support a deleterious role of B1R activation in renal inflammation. More relevant to patients with crescentic GN, the typical histology seen in small vessel vasculitis, a seminal study has reported that treatment with a B1R antagonist reduced renal chemokine expression alongside improved renal histology and function in the nephrotoxic nephritis model.14 This work adds to these data by suggesting a novel role for B1R and EMVs at the communication interface between the inflamed endothelium and activated neutrophils. Thus, the observed alleviation of macrophage accumulation and renal fibrosis by B1R antagonism seen in earlier studies might have been a result of reduced neutrophil recruitment to areas of injury. This study also raises questions about the role of EMV-derived B1R signaling in neutrophil activation and endothelial damage in the setting of small vessel vasculitis. It is tempting to reconsider previous data from this group, which showed that proteinase 3 (PR3) incubated with HK induced breakdown and release of a tridecapeptide termed PR3-kinin, which consisted of bradykinin with two additional amino acids on each terminus. Release of PR3-kinin was inhibited by both anti-PR3 antibodies and α1-antitrypsin.15 Furthermore, PR3-kinin binds to and activates the human B1R. Thus, one may hypothesize that insufficient inhibition of the PR3 protease activity in peripheral tissues may be responsible for accentuated endothelial injury through the unopposed production of PR3-kinin.16 Indeed, genome-wide association studies have revealed that polymorphic variants of genes encoding PR3, the predominant antigenic target of ANCA in patients with granulomatosis with polyangiitis, and its main inhibitor, α1-antitrypsin, are highly associated with granulomatosis with polyangiitis and even more significantly, PR3 ANCA positivity.17 Patients with ANCA vasculitis might exhibit a protease/antiprotease imbalance, which is either genetically determined in the rare patients with deficient α1-antitrypsin phenotypes or more often, acquired through both α1-antitrypsin inactivation in various pathologic conditions and possible inhibition of PR3/α1-antitrypsin complexing by anti-PR3 ANCA. This imbalance may at least contribute to disease aggravation or propagation. Both neutrophils and endothelial cells are a source of active PR3. Priming of HUVECs with TNF-α induced endothelial upregulation of PR3 surface expression.18 Neutrophil- and/or endothelium-derived PR3 may proteolyze HK and liberate PR3-kinin, thereby initiating kallikrein-independent activation of the kinin pathway. Thus, one might speculate that local production of PR3-kinin by the inflamed endothelium may promote release of EMVs and B1R-dependent neutrophil chemoattraction. However, these investigators found no qualitative difference between the chemoattraction elicited by plasma from a limited set of patients with vasculitis, and it was also unaffected by immunoglobulin (including ANCA) depletion, suggesting a more general effect than one mediated specifically by ANCA. Although this study does not answer the question of whether B1R-positive EMVs contribute to the development of GN, it does open the door to future potentially important work in the area. Experiments in animal models using tools to specifically interfere with microvesicles would be required to answer this question. In vivo, many variables could influence the function of microvesicles. As described in these recent studies, high levels of kinins in the plasma of patients with vasculitis may saturate B1R on microvesicles before they are transferred to neutrophils. These data support C1 inhibition as an important modulator of B1R-mediated vascular inflammation. It was known that prolonged incubation of plasma of patients with hereditary angioedema (but not healthy controls) leads to bradykinin formation and conversion of prekallikrein to kallikrein, which is reversed by reconstitution with C1-INH.1 Hereditary angioedema can be treated by C1-INH replacement with C1-INH concentrate or recombinant human C1-INH and/or by the selective and reversible kallikrein inhibitor, ecallantide. This therapeutic strategy might be useful in small vessel vasculitis. This study also moves the field forward by adding novel mechanistic insights to the interesting previous report of a beneficial effect of B1R antagonism in experimental immune complex–mediated GN.14 Disclosures None.