Annexin A2 System in Human Biology: Cell Surface and Beyond

Abstract

Annexin A2 (A2) is a multicompartmental, multifunctional protein that orchestrates a growing spectrum of biologic processes. At the endothelial cell surface, A2 and S100A10 (p11) form a heterotetramer, which accelerates tissue plasminogen activator–dependent activation of the fibrinolytic protease, plasmin. In antiphospholipid syndrome, anti-A2 antibodies are associated with clinical thrombosis, whereas overexpression of A2 in acute promyelocytic leukemia promotes hyperfibrinolytic bleeding. A2 is upregulated in hypoxia, and mice deficient in A2 are resistant to oxygen-induced retinal neovascularization, suggesting a role for A2 in human retinal vascular proliferation. In solid malignancies, the (A2•p11)₂ tetramer may promote cancer cell invasion, whereas in multiple myeloma A2 enables malignant plasmacyte growth and predicts prognosis. In the central nervous system, the p11 enables membrane insertion of serotonin receptors that govern mood. In the peripheral nervous system, p11 directs sodium channels to the plasma membrane, enabling pain perception. In cerebral cortex neurons, A2 stabilizes the microtubule-associated tau protein, which, when mutated, is associated with frontotemporal dementia. In inflammatory dendritic cells, A2 maintains late endosomal/lysosomal membrane integrity, thus modulating inflammasome activation and cytokine secretion in a model of aseptic arthritis. Together, these findings suggest an emerging, multifaceted role for A2 in human health and disease.

Cell Biology of Annexin A2

Annexin Family of Proteins

The annexins are an evolutionarily ancient and conserved family of Ca²⁺-regulated phospholipid-binding proteins.¹ Annexin proteins, which have existed for over 500 million years, typically possess two main structural domains. The core domain, usually 30 to 35 kDa in mass, contains four α-helical “annexin” repeats that bind Ca²⁺ and form a convex face that can associate with membrane phospholipid. The more hydrophilic amino-terminal “tail” or “interaction” domains are essentially unique to each family member and allow the annexins to partner with a wide variety of protein ligands. The term “annexinopathy,” first coined in 1999² and expanded in more recent reviews,^3-7 describes the mechanistic role played by the annexins in human disease.

Of the 12 annexins (annexins A1 to A11 and A13) expressed in humans, annexin A2 (A2) is among the most extensively studied, especially with respect to mammalian biology and human disease.⁸ A 36-kDa protein produced by endothelial cells, monocytes, macrophages, dendritic cells, trophoblast cells, epithelial cells, tumor cells, and many others, A2 can exist as a free monomer in the cytoplasm, in association with intracellular membranes, or tethered to the external face of the plasma membrane.^9,10 Human A2 is the product of the ANXA2 gene, composed of 13 exons distributed over 40 kb of genomic DNA on chromosome 15 (15q21).¹¹ Versions of A2 protein among mammalian species are ~98% identical at the amino acid level.

Protein S100A10/p11 and (A2•p11)₂ Complex Formation

Protein S100A10, often designated p11, is a recognized binding partner of A2.^12,13 p11 belongs to the S100 family of proteins by virtue of its solubility in 100% ammonium sulfate at neutral pH and its Ca²⁺-binding helix-loop-helix motifs. p11 endows A2 with increased phospholipid binding affinity. Although most S100 proteins, in response to Ca²⁺, undergo a conformational change that allows them to associate with target proteins, p11 exists in a permanent “calcium-on” state, due to two key amino acid substitutions at positions E65 and D56 within its Ca²⁺-binding domain.^14,15 Within the (A2•p11)₂ heterotetramer, p11 exists as a noncovalently linked homodimer; together the N-terminal HI and C-terminal HIV helices of adjacent p11 molecules form a groove, which is occupied by the α-helical N-terminal 14 amino acids of A2.¹⁵ p11 appears to be stabilized by its interaction with partner proteins such as A2.

Subcellular Localization of Annexin A2

A2 is a multicompartmental protein, which, like other family members, fulfills a spectrum of membrane organizing functions. The available evidence indicates that A2 consolidates membrane microdomains, recruits specialized membrane proteins, regulates membrane fusion events, and participates in membrane repair.⁹ Although heterotrameric (A2•p11)₂ assembles fibrinolytic proteases on the extracellular face of the plasma membrane, monomeric A2 remains soluble in the cytoplasm of cells at resting Ca²⁺ concentration. The subcellular and organellar localization of A2 appears to be governed by a multitude of factors that include Ca²⁺ concentration, pH, membrane phospholipid composition, post-translational modifications, and the availability of other proteins.

Like many annexins, A2 orchestrates a wide range of membrane trafficking events.⁹ A2 promotes Ca²⁺-regulated exocytosis, including Ca²⁺-induced release of Weibel-Palade body proteins,¹⁶ chromaffin granules,¹⁷ and surfactant-containing lamellar bodies.¹⁸ A2 possesses an endosome targeting sequence, and its depletion within cells leads to abnormal morphology of recycling endosomes.¹⁹ A2 appears to be involved in the biogenesis of multivesicular bodies and was the fifth most frequently cited protein among 140 proteins identified in 19 different proteomic studies of exosomes.^20-22 A2 possesses a nuclear export signal within residues 3 to 12, which overlap with the p11 binding site; when A2 is overexpressed, its translocation from nucleus to cytoplasm is subject to inhibition by leptomycin B.²⁰ Because A2 binds to specific messenger RNAs via its fourth core domain repeat, it may serve to escort specific RNAs from the nucleus to specific subcellular locations for localized protein synthesis.²¹ The actin and microfilament interactions of A2 are complex and appear to be connected with many of these functions.²² How these multiple activities may relate to human health and disease, however, is largely unknown.

Annexin A2 and p11 Dynamics at the Cell Surface

The cell surface is a discrete compartment for expression of both A2 and protein p11.^23-25 Here, especially on vascular endothelial cells, the (A2•p11)₂ tetramer serves as an assembly site for plasminogen and tissue plasminogen activator (tPA), an endothelial cell secretory product. Cleavage of the R⁵⁶⁰-V⁵⁶¹ peptide bond of plasminogen gives rise to the active, primary fibrinolytic protease, plasmin.^26-28 This membrane-oriented reaction accelerates the catalytic efficiency of tPA-dependent plasminogen activation by up to 2 log orders of magnitude. In classic fibrinolysis, on the other hand, activation of plasminogen by tPA is even more dramatically accelerated in the presence of fibrin, which serves as a cofactor for its own destruction.²⁶ Thus, the A2-based fibrinolytic system may serve as a protective, surveillance mechanism for fibrin clearance, whereas the more potent classic fibrin-based system may be more important in clearance of already formed fibrin thrombi.

Cell surface expression of the (A2•p11)₂ tetramer is a dynamic process, and translocation of the complex from the cytoplasm to the outer leaflet of the endothelial cell’s plasma membrane appears to be a key regulatory step in fibrinolysis.^23,24 In nonendothelial cells, the cell surface appearance of A2 has been linked to plasma membrane fusion of multivesicular bodies in NIH 3T3 fibroblasts,²⁹ to exocytosis of secretory granules in chromaffin cells,³⁰ to fusion of secretory vesicles with the plasma membrane in porcine intestinal epithelial cells,³¹ and to caspase-1 activation in stimulated macrophages or ultraviolet-irradiated keratinocytes.³² Still, the exact mechanism by which endothelial cells regulate surface (A2•p11)₂ remains unclear.

Endothelial cell translocation of A2 to the cell surface is initiated by several factors, including heat stress, thrombin stimulation, and hypoxia, and can occur within minutes.^33-35 A2 was originally identified as an src kinase substrate,³⁶ and its translocation is driven by phosphorylation at Y²³ by a pp60src-like kinase.³⁴ In the endothelial cell, translocation also requires expression of p11, which is stabilized by A2.³⁴ Upon binding to p11, A2 masks a critical “degron,” or polyubiquitination site on p11, so that in the absence of sufficient A2, p11 is polyubiquitinated and targeted to the proteasome for degradation.³⁷ In AnxA2^−/− mice, which demonstrate low to nondetectable p11 expression in most cells, treatment with bortezomib, a proteasome inhibitor, restored p11 expression in lung tissue, thus verifying that p11 is regulated via polyubiquitination in vivo.

Formation of the (A2•p11)₂ heterotetramer appears to be regulated by protein kinase C.^38,39 Protein kinase C phosphorylates S¹¹ or S²⁵ residues within the N-terminal tail domain of intracellular A2, thereby dissociating the complex and preventing further translocation to the cell surface. Dissociation of p11 from the complex allows it to be polyubiquitinated and destroyed in the proteasome.³⁸ Serine phosphorylation of A2 by protein kinase C appears to be initiated by plasmin, which, once generated at the cell surface, cleaves A2, activates toll-like receptor 4, and signals conventional protein kinase C–mediated phosphorylation of intracellular A2. Through this negative feedback mechanism, plasmin may regulate its own formation.

Annexin A2 at the Cell Surface

Thrombosis and Vascular Occlusion

Increasing evidence shows that the A2 system functions at the cell surface to regulate hemostasis in vivo (►Fig. 1A). This concept evolved from findings that the (A2•p11)₂ complex binds both tPA and plasminogen^40,41 and accelerates plasmin generation. Specific binding sites reported for plasminogen include lysine (K) 308 of A2²⁴ or the C-terminal K of p11,⁴² or both.⁴³ tPA binding to A2 appears to require a linear sequence within the tail domain (LCKLSL).⁴⁴ The Ca²⁺-dependent sequence KGLGT…D in core domain 2 facilitates membrane phospholipid interaction on the endothelial cell.⁴⁵

Fig. 1.

Some examples of cell surface and intracellular A2 system functions that relate to human health and disease. (A) Thrombosis and hemostasis. 1. In the vasculature, the endothelial cell surface A2 tetramer [(A2•p11)₂] tetramer binds plasminogen (Plg) and tissue plasminogen activator (tPA), accelerating the catalytic generation of plasmin (PN). Plasmin hydrolyzes insoluble fibrin (FN) into soluble fibrin degradation products (FDPs), thus maintaining blood vessel patency. In the absence of A2, fibrinolytic surveillance system is impaired, and fibrin accumulates. 2. In antiphospholipid syndrome and related autoimmune disorders, anti-A2 antibodies lead to fibrin deposition, either by inhibiting cell surface fibrinolysis or by activating endothelial cells and inducing a loss of cell surface thromboresistance. 3. Overexpression of (A2•p11)₂, as occurs in acute promyelocytic leukemia (APL), may lead to excessive plasmin generation, depletion of plasmin inhibitors, and induction of a hemorrhagic state (arrow). (B) Angiogenesis. Endothelial cell surface (A2•p11)₂, which is upregulated in hypoxia, promotes neovascularization of underperfused tissues, most likely through direct profibrinolytic activity. A2-generated plasmin may also activate downstream matrix metalloproteinases (MMPs), which further remodel the extracellular matrix (ECM). (C) Neurodegenerative disease. Within neurons, A2, presumably in complex with p11, enables the microtubule (MT)-associated protein, tau, to associate with the plasma membrane within axonal growth cones. In familial frontotemporal dementia, the mutated R406W tau fails to associate with plasma membrane-associated A2, leading to axonal instability, and is thought to lead neurodegeneration. (D) Aseptic arthritis. When dendritic cells are exposed to nonbiodegradable wear debris particles (WDP) derived from joint prostheses, the particles are endocytosed and traffic to lysosomes (L). The ensuing late endosomal/ lysosomal membrane damage is thought to be normally repaired through a (A2•p11)₂-dependent mechanism. In the absence of A2, escape of lysosomal cathepsin B (CatB) and hydrogen ions (H⁺) leads to rapid activation of the NLRP3 inflammasome, with activation of caspase-1, and cleavage of pro-interluekin-1 (pro-IL-1β) into its mature, secretable form. Release of IL-1β initiates a potent inflammatory response that exacerbates osteolysis and joint damage.

The AnxA2^−/− mouse has been informative in investigating the role of the A2 system in vascular homeostasis in vivo. Although A2-deficient mice display uncompromised development, fertility, and lifespan, fibrin accumulation is a key phenotypic hallmark. Fibrin can be observed in both intra- and extravascular locations within the lungs, spleen, small intestine, liver, and kidney.⁴⁶ In addition, injury to the carotid artery through application of FeCl₃ at the adventitial surface led to a significant decline in carotid blood flow and increased frequency of complete thrombotic occlusion in the AnxA2^−/− versus AnxA2^+/+ mouse. Cultured microvascular endothelial cells isolated from neonatal AnxA2^−/− mice, moreover, failed to support tPA-dependent plasmin generation. S100A10^−/− mice also display increased vascular fibrin, reduced clearance of thrombi, and impaired neovascularization of Matrigel implants,⁴⁷ reflecting either the loss of S100A10 (as a binding site for Plg), loss of p11 “chaperone” function in directing A2 to the plasma membrane surface, or both.

Interestingly, mice with diet-induced hyperhomocysteinemia, like AnxA2^−/− mice, also display fibrin accumulation in a wide range of tissues.⁴⁸ The thiol-containing amino acid, homocysteine, forms as a metabolic intermediate in vivo when methionine is converted to cysteine.⁴⁹ Hyperhomocysteinemia, which occurs in several inborn errors of metabolism or as a consequence of folate or B vitamin deficiency, has been implicated in thrombotic and atherosclerotic vascular disease,⁵⁰ although therapies that lower plasma homocysteine may not fully reverse the risk of recurrent cardiovascular disease.⁵¹ Pretreatment of endothelial cells with homocysteine largely inhibited both tPA binding and tPA-dependent plasmin formation at the cell surface.⁵² Incubation of purified A2 with homocysteine also blocked its ability to bind tPA, and A2 extracted from hyperhomocysteinemic mice failed to support tPA binding or tPA-dependent plasmin generation.⁴⁴ Hyperhomocysteinemia has emerged as an independent risk factor for cerebral small vessel disease, which leads to lacunar infarction and approximately one fourth of all ischemic strokes.^53,54 The cell surface A2 pathway may be central to understanding this process.

Animal models of thrombosis have illustrated the theoretical clinical utility of recombinant A2 in ischemic stroke. Compared with those treated with tPA alone, rats treated with tPA plus supplemental rA2 at 2 or 4 hours after the initiation of focal embolic stroke developed significantly smaller infarcts and showed improved cerebral blood flow.^55,56 In a follow-up study, labeled embolized thrombi were imaged by magnetic resonance and found, in tPA plus A2 treated rats, to be significantly smaller than those in animals treated with tPA alone.⁵⁷ Similarly, when carotid artery thrombosis was induced by adventitial application of FeCl₃ in the rat, administration of recombinant full-length, but not tail-less, A2 was associated with improved cerebral blood flow and reduced thrombus size in comparison with untreated control animals.⁵⁸ Because rA2 appears to improve the lytic efficacy of tPA,⁵⁹ these findings could reduce the small, but significant, risk of cerebral hemorrhage associated with the use of tPA for thromboembolic stroke in humans by lowering the required intensity of tPA treatment.^60-62

Beyond these rodent models, there is strong evidence that the A2 system has important implications for human vasoocclusive disease. In antiphospholipid syndrome (APS), a well-known cause of acquired thrombophilia, A2 has been identified as one of many autoantibody targets.^63,64 High titer autoantibodies to A2 were detected specifically in patients with APS and severe thrombosis and/or pregnancy morbidity.^65,66 Anti-A2 antibodies isolated from APS patients impaired endothelial surface tPA-dependent plasmin generation and incited them to express elevated levels of procoagulant tissue factor.⁶⁵ Among a cohort of 40 consecutive patients with cerebral venous thrombosis studied 2 to 6 months after the index thrombotic event, moreover, 12.5% were found to have high titer anti-A2 antibodies compared with 2.1% in healthy subjects.⁶⁷ Although this finding has been confirmed in APS and extended to other autoimmune disorders, such as rheumatoid arthritis, the nature of the target epitopes remains undefined.^68-70 In lupus nephritis, finally, autoantibodies directed against A2 have been found to react with renal mesangial cells.⁷¹ Together, these data suggest that cell surface A2 represents a prominent autoantibody target and that high titer anti-A2 in the vasculature is associated with a thrombotic diathesis.⁷²

In children with sickle cell disease, stroke is a potentially devastating complication that occurs in 6 to 8% of patients.⁷³ Two studies linked single nucleotide polymorphisms in the ANXA2 gene with increased risk of stroke (odds ratio > 2.7),^74,75 whereas additional ANXA2 single nucleotide polymorphisms have been associated with elevated risk of avascular necrosis of bone (osteonecrosis)⁷⁶ and, possibly, pulmonary hypertension.⁷⁷ These data suggest A2 may be a significant modifier gene, which, when modified, reduces blood fluidity and exacerbates the natural history of sickle hemoglobinopathy.

Hyperfibrinolysis and Hemorrhage

Acute promyelocytic leukemia (APL) is commonly associated with life-threatening hemorrhage at the time of presentation.⁷⁸ The presence of the [t(15;17)(q22–24;q12–21)] chromosomal translocation gives rise to the transcriptionally active promyelocytic leukemia-retinoic acid receptor α (PML-RARα) fusion protein. At diagnosis, coagulopathy appears to reflect a combination of disseminated intravascular coagulation and hyperfibrinolysis due to consumption of α₂-antiplasmin and excessive plasmin generation. Blast cells isolated from newly diagnosed APL patients expressed unusually high levels of A2, and all patients had evidence of hyperfibrinolysis, with elevated circulating fibrin degradation products, including D-dimer, and depletion of plasma fibrinogen.⁷⁹ NB4 cells, which carry the t(15;17) translocation and express the PML-RARα fusion protein, displayed steady-state A2 messenger RNA levels that were ~10-fold higher than those found on leukemia cells that lacked the fusion protein. Treatment of NB4 cells with the retinoic acid receptor ligand, all trans-retinoic acid, attenuated A2 expression in a time frame associated with clinical resolution of bleeding.⁷⁹

In a more recent prospective study, 26 APL patients were found at diagnosis to have enhanced fibrinolysis, despite normal tPA levels and increased plasminogen activator inhibitor-1; harvested APL cells expressed threefold higher levels of A2 and supported elevated tPA-dependent plasmin generation.⁸⁰ Elevated expression of p11 in NB4 cells has also been reported and appears to respond, like overexpression of A2, to all trans-retinoic acid treatment.⁸¹ These studies confirm the likely role of the A2 system in hyperfibrinolytic bleeding in APL, especially in the central nervous system where microvascular expression of A2 may be relatively high.⁷⁸ Interestingly, induction of diet-induced hyperhomocysteinemia in a mouse model of APL (hCG-PML-RARα mice) reversed hyperfibrinolysis, thus reinforcing the relevance of A2 to APL in vivo and suggesting a potentially novel the rapeutic approach to APL.⁸²

Retinal Angiogenesis

In at least three different models of postnatal angiogenesis, including Matrigel implant, corneal pocket, and oxygen-induced retinopathy, AnxA2^−/− mice show a diminished capacity to form new blood vessels (►Fig. 1B).⁴⁶ In mice with diet-induced hyperhomocysteinemia, which renders A2 unable to bind tPA effectively, corneal neoangiogenesis is also reduced and can be restored to normal upon intravenous treatment with recombinant A2.⁴⁸ These data suggest that either absence of A2 or its modification by homocysteine deprives the endothelial cell of plasmin-generating capacity, leading most likely to accumulation of fibrin as well as diminished activation of downstream matrix metalloproteases. Unable to adequately remodel extracellular matrix, these cells are unable to mount an appropriate angiogenic response.

In the oxygen-induced retinopathy (OIR) model, newborn mice are maintained with their nursing mothers in a 75% oxygen environment for 5 days and then abruptly transferred to room air; the ensuing relative hypoxia initiates a robust retinal vascular proliferative reaction.⁸³ Interestingly, the in vivo neoangiogenic response is reduced by about 50% in AnxA2^−/− pups, and synthesis of A2 is upregulated by the relative hypoxia induced in this model and also by true hypoxia in vitro.^35,84,85 Further experiments revealed that this reflects the direct action of hypoxia-inducible factor-1, a master regulator of oxygen-sensitive genes, upon the hypoxia responsive element–containing A2 promoter.³⁵ Electrophoretic mobility shift experiments, chromatin immunoprecipitation studies, and luciferase promoter reporter assays all demonstrate a functional interaction between hypoxia-inducible factor-1α and hypoxia-inducible factor-1β and the human A2 promoter. Gene expression profiling of retinal tissue in murine OIR, furthermore, revealed a 4.3-fold increase in A2 expression at P17, 5 days after the mice were returned to room air.⁸⁶ A2 was also found to be upregulated and to promote neovascularization after laser rupture of Bruch’s membrane in a model of choroidal angiogenesis.⁸⁷ In addition, an agent (TM601) that binds directly to A2 appears to suppress both choroidal and retinal neovascularization in the mouse,⁸⁸ and an A2 N-terminal peptide specifically blocked neoangiogenesis in the chick chorioallantoic membrane model and the mouse Matrigel plug assay.⁸⁴ These data suggest that A2 is likely to be upregulated in other tissues in response to hypoxia and may be central to the angiogenic response in many pathologic settings.

From a mechanistic point of view, it appears that the blockade to normal angiogenesis in the AnxA2^−/− mouse in OIR reflects an inability of endothelial cells to traverse an extracellular fibrin barrier. Hypoxic angiogenesis, which is initiated by vascular endothelial growth factor, involves loss of interendothelial cell junctions and leakage of plasma into the extracellular space with fibrin formation. Neovascularization during OIR in the AnxA2^−/− retina can be normalized upon systemic pretreatment of the mice with ancrod, an agent that depletes fibrinogen, thus preventing fibrin formation.³⁵ Similarly, the robust angiogenic response that occurs in the wild-type mouse can be tempered by pretreatment with tranexamic acid, a global fibrinolytic inhibitor.³⁵ These results link fibrinolysis with neoangiogenesis, suggesting new therapeutic approaches for proliferative retinal vascular disorders, such as retinopathy of prematurity or diabetic retinopathy.

Cancer Progression

Glioblastomas are highly aggressive and treatment-refractory malignancies that typically exhibit widespread infiltration of surrounding structures early in the development of the disease.⁸⁹ Gliomas produce proteases, including plasminogen activators and matrix metalloproteinases, both of which are thought to promote malignant cell invasion.^89,90 In proteomic analyses, high expression of A2 has been associated with more aggressive phenotypes,⁹¹ and A2 appears to localize to pseudopodia of invasive glioma cells.⁹² In both rat and mouse models, stable knockdown of A2 significantly inhibited glioblastoma invasion, proliferation, and angiogenesis after stereotactic implantation in the brain; interestingly, these effects were independent of p11 expression.⁹³ High expression of A2 has also been documented in human glioblastoma, where it correlates positively with histologic grade and central nervous system dissemination.^93-96 These data suggest that A2-directed treatment might offer a new therapeutic modality for human glioblastoma.

In a xenograft model of highly invasive breast cancer, similarly, malignant human cells were implanted into nude mice, and both tumor growth and vascular density was blocked by administration of anti-A2 monoclonal antibody.⁹⁷ The tumor cells used in this experiment expressed abundant A2, strongly supported tPA binding and tPA-dependent plasmin generation, and exhibited plasmin- and A2-dependent cellular matrix invasion.^98,99 In another study, growth of Lewis lung and T241 sarcoma tumors implanted into S100A10^−/− mice was reduced compared with wild-type controls.¹⁰⁰ Impaired tumor growth was correlated with diminished macrophage density within the tumors, and clodronate-mediated depletion of macrophages in wild-type mice led to a similar reduction in tumor size, suggesting an effect of p11 on macrophage recruitment.⁴² Finally, soluble (A2•p11)₂ tetramer has been reported to activate monocyte-derived macrophages via toll-like receptor 4, thus modulating cytokine production.¹⁰¹

In multiple myeloma, A2 is a candidate prognostic marker. It is expressed in both normal and malignant plasma cells, but elevated expression in myeloma is associated with reduced event-free and overall survival.¹⁰² High-level expression of A2 in human myeloma cells has been confirmed in transcriptional profiling studies.¹⁰³ Primary multiple myeloma cells harvested from a cohort of patients displayed 10-fold higher cell surface A2 expression than that observed on normal plasma cells; silencing of A2 in related cell lines suppressed expression of pro-angiogenic genes.¹⁰⁴ Mechanistically, A2 and a related A2-binding “receptor” appear to promote myeloma cell malignant plasma cell growth and adhesion to stromal cells within the bone marrow microenvironment.¹⁰⁵ Interestingly, prostate cancer cells also appear to use A2 in homing to the bone marrow and in establishing bone metastases.¹⁰⁶ Together, these studies suggest that A2 may contribute to cancer progression through multiple mechanisms.

Neuropsychiatric Disease

Protein S100A10 appears to orchestrate several aspects of neuropsychiatric function.¹⁰⁷ p11 binds to two serotonin receptors (5-HT_1B and 5-HT₄), both involved in mood regulation.¹⁰⁸ S100A10^−/− mice exhibit increased immobility in the tail suspension test, increased thigmotaxis, and decreased responsiveness to a sucrose reward, all thought to be indicative of depression-like behavior. In wild-type mice, suppression of p11 expression within the nucleus accumbens resulted in depression-like behavior, similar to that observed in globally deficient p11 mice; exogenous administration of p11 within the nucleus accumbens of p11-deleted mice seems to restore normal behavior.¹⁰⁹ In humans suffering from depression, p11 levels in the nucleus accumbens were reduced, as were messenger RNA levels in peripheral blood mononuclear cells, suggesting a novel biomarker for patients at high risk for suicide.¹¹⁰ Interestingly, frequently used antiinflammatory agents that attenuate the antidepressive effects of serotonin reuptake inhibitors may do so by inhibiting the effects of interferon-γ, a known inducer of p11.^111,112 Thus, some forms of human depression may be amenable to augmentation of p11 expression; it is not yet known whether this p11 function is A2 dependent.¹¹³

The sensory neuron tetrodotoxin-resistant sodium channel (Na_v1.8/SNS) is the primary pain perception receptor and is expressed in 85% of neurons emanating from the dorsal route ganglia. p11 promotes translocation of to the Na_v1.8/SNS channel to the plasma membrane by binding to its amino terminus,¹¹⁴ but it is not yet clear whether this interaction requires the participation of A2.¹¹⁵ Deletion of p11 in primary nociceptor sensory neurons results in a severe reduction in coding of painful stimuli in dorsal root sensory neurons.¹¹⁶ These data raise the possibility that p11 targeting could provide useful treatment for refractory pain disorders.

Annexin A2 in the Cytoplasm

Frontotemporal Dementia

Neurodegenerative processes are increasing in frequency as the world’s population ages. In frontotemporal lobar degeneration with ubiquitinated inclusions, A2 was found to be overexpressed in expression profiling of postmortem frontal lobe tissue.¹¹⁷ In familial frontotemporal dementia, another neurodegenerative process, mutations in the microtubule-associated protein, tau, have been identified.¹¹⁸ A study revealed that A2 enables wild-type tau to interact with the neuronal growth cone plasma membrane, thus stabilizing process outgrowth by bridging microtubules to the membrane (►Fig. 1C).¹¹⁹ A2 is, however, unable to interact with the R406W, disease-associated form of tau, possibly leading to its intracellular accumulation and neuronal cell death.

Limb-Girdle Muscular Dystrophy

Limb-girdle muscular dystrophy type 2B, a disorder characterized by progressive weakness and atrophy of proximal muscles and elevated serum creatine kinase levels, reflects mutation of the gene encoding dysferlin, a member of the dystrophin-associated complex of muscle proteins.^120,121 Gene expression profiling has revealed an approximately twofold overexpression of A2 transcript levels in biopsied muscle tissue from these patients compared with healthy control subjects,¹²² and the level of expression of A2, as well as A1, in biopsied muscle appears to correlate with clinical severity.¹²³ Of note, both A2 and A1 have been shown to interact with dysferlin by co-immunoprecipitation, to colocalize with dysferlin in normal muscle biopsies, and to promote sarcolemmal wound healing in an in vivo model.¹²⁴ In response to myotube injury, A2 appears to aggregate intracellular vesicles that fuse to form a patch that breaches the plasma membrane defect.

Aseptic Wear Debris Arthritis

Although several million arthroplastic joint replacements are performed each year, 10 to 15% of these procedures fail due to aseptic osteolysis.¹²⁵ In this process, wear debris particles, generated by frictional forces exerted on artificial joint surfaces, are endocytosed by inflammatory cells, such as macrophages and dendritic cells. Eight to 12 carbon alkanes, derived from wear debris particles, can bind directly to toll-like receptors 1 and 2, thereby activating downstream inflammatory pathways.¹²⁶ In addition, wear debris particles can induce endosomal membrane damage, which is associated with dramatic recruitment of cytoplasmic A2 to the endosomal membrane. In the absence of A2, endo/lysosomal disruption leads to leakage of lysosomal cathepsins and H⁺ ions into the cytosol with subsequent activation of the nucleotide-binding, leucine-rich, pyrin-containing-3 (NLRP3) inflammasome, secretion of interleukin-1β, and an accelerated inflammatory response (►Fig. 1D).¹²⁷

Summary

The A2 system fulfills a range of biologic functions both on the plasma membrane and within multiple intracellular compartments. At the cell surface, the A2 system promotes fibrinolysis, angiogenesis, and cell migration. Within the cell, it aids in organelle membrane organization, fusion, and repair. Although we are beginning to understand the physiologic orchestration of these events, key questions remain about the regulation of the system and its precise role in human health and disease. The next few years in A2 biology should prove exciting and illuminating.