F11R/JAM-A: why do platelets express a molecule which is also present in tight junctions?

Abstract

F11 receptor (F11R)/Junctional Adhesion Molecule –A (JAM-A) is a transmembrane protein which belongs to the immunoglobulin superfamily of cell adhesion molecules. F11R/JAM-A is present in epithelial cells, endothelial cells, leukocytes, and blood platelets. In epithelial and endothelial cells, it takes part in the formation of tight junctions. In these structures, molecules of F11R/JAM-A located on adjacent cells form homodimers and thus take part in stabilization of cellular layer integrity. In leukocytes, F11R/JAM-A was shown to play role in their transmigration through the vascular wall. Paradoxically, the function of F11R/JAM-A in blood platelets, where it was primarily discovered, is much less understood. It has been proven to regulate downstream signaling of αIIbβ3 integrin and to mediate platelet adhesion under static conditions. It was also shown to contribute to transient interactions of platelets with inflamed vascular wall. The review is aimed at summarizing the current state of knowledge of the platelet pool of F11R/JAM-A. The article also presents perspectives of the future research to better understand the role of this protein in hemostasis, thrombosis, and other processes where blood platelets are involved.

Plain Language Summary

The molecule of a complex name F11R/JAM-A is a protein which was primarily discovered on blood platelets. Later, the presence of the same molecule was confirmed on endothelial cells and epithelial cells. From the moment of the discovery, most of the research was focused on the role of this protein in the latter types of cells. It was found to be an important element of so-called tight junctions. These structures are crucial for maintaining of integrity and selective permeability of cellular layers composed of these types of cells. In the following years, the presence of F11R/JAM-A has also been reported on leukocytes. An important role of specific type of leukocytes is their penetration to the sites of inflammation. Interplay of F11R/JAM-A present on endothelium and that on leukocyte is involved in this process. But what about the role of this protein in blood platelets where it was originally discovered? There is limited knowledge regarding this issue. It was found to play a role in the ability of platelets to adhere to a surface under static conditions, but it is not known if the same is true under flow. Is the protein necessary for platelets to aggregate and form thrombus? Genetically engineered mice were created which lack this protein in blood platelets to answer this question. These platelets were abnormally reactive, as it transpired that the protein plays a role of a negative regulator to one of the most important mechanisms, which triggers platelet aggregation. But is this inhibitory function the only task F11R/JAM-A has to fulfil in platelets? Presented review collects all the knowledge regarding this protein in blood platelets and tries to show interesting routes which need exploration.

Keywords:

• Endothelium
• F11R/JAM-A
• inflammation
• platelets
• thrombus
• tight junctions

Introduction

The purpose of this review is to aggregate the current status of knowledge of F11R/JAM-A (F11 receptor/junctional adhesion molecule A) in blood platelets. Historically, the protein was first discovered in platelets as a target of an activating F11 antibody.¹ Shortly after it was also confirmed as a functional part of tight junctions in endothelium and epithelium² and its expression was also found in immune cells³ and smooth muscle cells.⁴ In the following years, a number of studies addressed the function of this protein in a range of physiological processes and an overwhelming number of studies addressed F11R/JAM-A function in these phenomena. At the same time, only a minority of studies were aimed at understanding of F11R/JAM-A function in platelets and in the processes where platelets are involved. As of writing this article, there were 645 original publications mentioning JAM-A. For only 39 of those, platelets are within the main focus of the study. Therefore, only about 6% of papers mention the platelet population of F11R/JAM-A in anything more than a passing acknowledgment of its existence. The presented work is aimed at a review of this relatively limited amount of experimental data.

The search was conducted using keywords and controlled vocabulary within three databases: PMC, Embase, and Cochrane library. The primary keyword and grouping variable for search results were different names of the protein in question (JAM-A/F11R/JAM-1/CD321/Junctional molecule A/Junctional adhesion molecule 1). The search parameters were further filtered by use of additional terms and phrases for related topics such as a junctional molecule, tight junction, platelet function, platelet proteome, platelet transcriptome, platelet omics, atherosclerosis, arteriosclerosis, and atherogenesis. Papers with fewer than five iterations of primary keywords were screened for secondary terms. Aggregated result of less than 10 keywords/terms was a basis for article exclusion unless the article provided further sources in its bibliographical references. Given the relative scarcity of valid sources, no articles were excluded based on species or the date of publication.

Historical background of F11R/JAM-A discovery

In 1990, Kornecki et al. showed that certain antibodies which had the capability of activating human blood platelets were binding to a protein target which was not known at the time. The protein was named F11 receptor (F11R) after the name of the clone of these antibodies (F11).¹ Platelet activation induced by binding of the antibodies to F11R was primarily shown to be dependent on the interaction of Fc fragment of F11 antibodies with FcγRII receptor present on blood platelets⁵ (Figure 1). In 1998, a similar protein was isolated during studies on tight junctions in epithelial and endothelial cell monolayers.² The protein was recognized as an immunoglobulin and named after its function as Junctional Adhesion Molecule (JAM). In the wake of full sequencing of the protein, the homology of JAM-A and F11R was established.^6,7 As an effect of this convergent discovery, the two names of the protein exist simultaneously in scientific publications. This status quo gained legitimacy by the decision of the Human Genome Nomenclature committee which has approved F11R as a symbol of the gene number AF207907, and the number BC021876 for the murine molecule, while JAM-A was mentioned as one of the alias symbols among others such as PAM-1, JCAM, JAM-1, JAMA, and CD321.

Figure 1. Activation of blood platelets by F11 antibodies binding to F11R/JAM-A. F11 bind to F11R/JAM-A and Fc fragment of the antibodies is recognized by FcγRII receptor which triggers platelet activation.

Basics of structure and function of F11R/JAM-A

Structure of F11R/JAM-A

JAM-A is a transmembrane protein belonging to the immunoglobulin superfamily (IgSF).^2,8 Several JAMs have been recognized based on their structure,^9,10 however, other members of the JAM family remain outside the scope of this paper. The other proteins from this family are JAM-B which populates mainly tight junctions and was not detected in blood platelets and JAM-C which is expressed on blood platelets and plays a role of a counterreceptor for the leukocyte integrin Mac-1.¹¹ JAM-A has a molecular weight of 29 kDa^1,5 (core protein) which can reach between 32 kDa¹² and 43 kDa¹³ as a result of post-translational modifications (N-glycosylation) or due to differences in mRNA splicing.¹⁴ The protein consists of an extracellular part and transmembrane domain which ends in a short cytoplasmic tail. The extracellular part contains two Ig-like domains, a membrane distal D1 domain, and membrane proximal D2 domain (Figure 2a). In human F11R/JAM-A, the distal domain is of V-type and the proximal is of I-type.¹⁵ In turn, murine protein consists of membrane distal V-type and a proximal C-type domain.^16,17 The extracellular chain contains in its structure a motif which takes part in the homophilic-dimerization of two adjacent F11R/JAM-A molecules known as cis-dimerization.^18,19 A distal Ig domain also contains a motif responsible for the trans-homo-dimerization i.e. dimerization of two F11R/JAM-A molecules located on adjacent cells. The trans-dimerization motif lies on the opposite side to the cis dimerization interface, ensuring non-exclusive access of the binding sites.^20,21 A short linker chain between Ig domains forces a 125°^19,22 bend in its conformation.

Figure 2. Structure and molecular interactions of the extracellular component F11R/JAM-A a) structure of F11R/JAM-A, b) cis-homophilic dimerization – two molecules of F11R/JAM-A located in the same cell interact with each other via D1 domain, c) trans-homophilic interactions – two, or more F11R/JAM-A dimers located on adjacent cells interact with each other via D1 domain, the site of interaction is different from the one responsible for cis-homophilic dimerization, d) heterophilic interaction with lymphocyte function-associated antigen (LFA-1) - LFA-1 is the only identified to date heterophilic ligand for F11R/JAM-A, the interaction occurs via D2 domain.

JAM-A has a single-pass transmembrane domain which ends in a short cytoplasmic tail with a PDZ binding motif^23,24 which is characteristic for transmembrane receptors and signal transduction and is usually associated with internal proteins like ZO-1 or the actin cytoskeleton.²⁵

Interactions of F11R/JAM-A with other proteins

Homo- and hetero-dimerization

As mentioned above, F11R/JAM-A molecules are capable of forming homodimers. This homodimerization occurs in cis-configuration when the dimer is formed by the molecules located in the same cell (Figure 2b). The molecules can also interact in the trans-configuration when they are located on the membrane of two adjacent cells²⁰ (Figure 2c). It is not clear whether cis-homodimerization is required for trans-homophilic interactions. Some data suggest that monomeric F11R/JAM-A can also interact with its counterpart on an adjacent cell.²⁰ The sites responsible for cis- and trans-homophilic interactions lay in distinct parts of D1 domain of the molecule.²⁰ Both types of interactions are believed to be associated with the function of the protein which will be described further.

The only reported ligand capable of forming a heterodimer with F11R/JAM-A is the lymphocyte function-associated antigen (LFA-1)²⁶ – leukocyte integrin involved in a plethora of processes such as adhesion or extravasation of lymphocytes. The site of interaction with LFA-1 is located in the proximal D2 domain of JAM-A/F11R (Figure 2d).

Interaction between F11R/JAM-A and other representatives of the JAM family such as JAM-B and JAM-C was reported in Danio rerio.²⁷

It is important to mention that both dimeric and monomeric form of F11R/JAM-A undergoes shedding from the cell membrane and thus occurs also in a soluble form.^1,28 Two sheddases were identified to be responsible for this process: ADAM17 and to a lesser extent ADAM10.^13,29,30 The soluble form consists of both extracellular D1 and D2 domains. Such a soluble form of the protein is able to interact with intact, membrane-bound F11R/JAM-A.³¹

Interactions with integrins and tetraspanins

F11R/JAM-A interacts laterally with integrins and tetraspanins. At least two tetraspanins, CD9 and CD151^21,32,33, act as intermediaries in the formation of ternary complexes of F11R/JAM-A with integrins such as α3β1 or αvβ3 in endothelial or epithelial cells^32–35 or with platelet-specific αIIbβ3.¹² Importantly, in such complexes, F11R/JAM-A is present in a monomeric form^12,32 (Figure 3).

Figure 3. Interaction of F11R/JAM-A with αIIbβ3. Monomeric F11R/JAM-A interacts with αIIbβ3 via tetraspanin CD9.

The architecture of these interactions seems to be complex and depends on the type of cells in question. In endothelial cells, interaction of CD9 with JAM-A requires the PDZ domain-binding motif of JAM-A and therefore is mediated by the intracellular portion of JAM-A.³² In contrast, in MDCKII cells, JAM-A interacts with α3β1 via the extracellular JAM-A domain.³³

Proteins associated with intracellular part of F11R/JAM-A

The intracellular COOH terminus of F11R/JAM-A contains a type II PDZ domain-binding motif. It has been shown to interact with several PDZ domain-containing proteins: AF-6 (afadin) and ZO-1²⁵ in endothelial cells, afadin and PDZ-GEF2 but not with ZO-1,³⁶ with ZO-2, afadin and PDZ-GEF1,³⁷ CASK,³⁸ MUPP1,³⁹ and PAR3⁴⁰ in epithelial cells.

Function of JAM-A/F11R in endothelial cells, epithelial cells, and leukocytes

To date, JAM-A/F11R has been shown to be an important player in a large number of processes such as inflammation, tumor progression, and metastasis, differentiation of T-helper lymphocytes, or entry of viruses to host cells. The presented work is intended to review the current state of knowledge on JAM-A/F11R in blood platelets. This cannot be done, however, without outlining the information on this protein which, as stated above, was to a large extent gathered in other cellular models. To avoid an extensive review of JAM-A/F11R function in other types of cells which a reader can familiarize with in other excellent papers, such as,^16,41 we briefly introduce arbitrarily selected aspects of JAM-A/F11R function in endothelial cells, epithelial cells, and leukocytes, which we considered most relevant for further discussion of JAM-A/F11R function in blood platelets.

Component of tight junctions

F11R/JAM-A is a component of tight junctions. In these structures, extracellular parts of F11R/JAM-A molecules are assembled as cis-homodimers interacting via trans-homodimerization with cis-homodimers on adjacent cells.² The intracellular part of the protein is associated with a complex of proteins which integrates tight junctions with the actin filaments. These trans-homophilic interactions are crucial for maintaining the integrity and regulation of permeability of an endothelial or epithelial layer. There are, however, discrepancies as to whether F11R/JAM-A plays a structural or regulatory role in these phenomena. A profound analysis of JAM-A’s role in the physiology of tight junctions and discussion regarding these discrepancies can be found in excellent reviews by Hartmann et al.⁴² and Steinbacher et al.²¹

An important aspect of F11R/JAM-A physiology in the endothelium is that upon inflammatory activation of endothelial cell, the protein translocates from tight junctions to the apical side of the cell and thus to the luminal surface of the blood vessel.^43,44 This in turn facilitates the adhesion of monocytes and platelets to the inflamed vascular wall and therefore perpetuates the inflammatory process.

Adhesion of leukocytes to vascular wall and their transmigration

Leukocytes interact via their LFA-1 with endothelial F11R/JAM-A. Such interactions facilitate anchoring and migration of T cells in static²⁶ as well as flow conditions.⁴⁵ A similar effect was shown for monocytes⁴⁶ and neutrophils, but the extent was lesser, due to redundancy provided by an alternate pathway employing Mac-1/ICAM-1 binding.

The affinity studies suggest that JAM easily adopts the dimeric state and preferentially forms hetero- rather than homodimers.^{10,15,20,23,47} This is likely the mechanism used in the diapedesis, facilitating disruption of tight junctions. Once the LFA-1-induced disruption of the barrier⁴⁸ occurs, the uncoupled JAMs can disrupt deeper JAMs, therefore propagating the opening of the paracellular path, “unzipping” tight junctions in the direction of diapedesis.⁴⁹

F11R/JAM-A in blood platelets

Expression, dimerization, and phosphorylation

F11R/JAM-A is present in all phases of platelet life, starting from megakaryocytes,⁵⁰ through to matured platelets where it can be found on the cell surface and inside the alpha granules,⁵¹ ending on platelet-derived microparticles.

Evaluation of the number of copies of F11R/JAM-A on the surface of human blood platelets revealed approx. 8,000 copies,¹ while the number of total copies based on proteomic analysis of the whole platelet was estimated at approx. 13,000 copies.^52,53 The difference between a total number of copies and that on the surface suggests that some amount can be stored internally and presented on the surface upon activation. The latter is in agreement with the presence of the protein in alpha granules⁵¹ and with recent findings that platelet activation results in an increase of F11R/JAM-A expression revealed by flow cytometry.³¹ Total number of copies of F11R/JAM-A places it among other abundant and crucial proteins such as integrin β3, αIIb, or glycoprotein Ibβ (expressed in tens of thousands of copies) and GPVI which (expressed in approximately 10,000 copies) and hints at significant role of F11R/JAM-A in the functioning of blood platelets.

In non-activated human platelets, F11R/JAM-A is predominantly present as a monomer and is associated with tetraspanin CD9.¹² In blood platelets, CD9 is associated with αIIbβ3 integrin and was shown to negatively regulate its function³⁵ and to be involved in microparticle release from activated platelets.⁵⁴ When platelets were activated with F11 antibodies, a shift from monomeric to dimeric form of F11R/JAM-A was observed and this process was associated with a decrease in CD9 binding to F11R/JAM-A and an increase in binding of αIIbβ3. The latter observation is in contrast to what was observed in thrombin or fibrinogen-activated platelets where F11R/JAM-A dissociated from the complex with integrin αIIbβ3.⁵⁵ Similarly, in endothelial and epithelial cells, their activation resulted in dissociation of F11R/JAM-A from the complexes with integrins rather than association.^32,34 Assuming that activation with F11 antibodies is to some extent different than that triggered by “classical” agonists, it can be summarized that similarly to epithelial and endothelial cells F11R/JAM-A in platelets dissociates from the complex with integrins upon platelet activation. There is no evidence, however, whether in platelets activated with “classical” agonists dissociated F11R/JAM-A forms dimers as it does in F11-activated platelets. The function of F11R/JAM-A is regulated by its phosphorylation. Upon platelet activation with thrombin, the protein is phosphorylated at Ser284 by PKC, to a lesser extent when activated with collagen, and no phosphorylation on Ser284 was observed in ADP-activated platelets.^5,56 In resting blood platelets, F11R/JAM-A is phosphorylated on a tyrosine residue (human Y280 or murine Y281). Platelet activation and dissociation of F11R/JAM-A from integrin αIIbβ3 results in dephosphorylation in this site.⁵⁵ This dephosphorylation was shown to be dependent on protein phosphatase non-receptor type 1 (PTPN1) constitutively associated with F11R/JAM-A.⁵⁷ Tyrosine phosphorylation has also been shown to occur as a result of CLEC-2 cascade activation in platelets.⁵⁸ Analysis of phosphoproteomics of platelets revealed that F11R/JAM-A is in a group of proteins that respond with an early and sustained modification upon activation of platelets with ADP.⁵⁹

F11R/JAM-A in platelet activation, regulation of αIIbβ3 integrin signaling

From the moment of its discovery, F11R/JAM-A has been associated with the capacity to regulate blood platelet function. The original study has shown that F11 antibodies, which bound to the F11R protein, activated blood platelets by interacting with FcγRII.⁵ As it turned out, there were also FcγRII -independent effects of F11 antibodies. Adhesion of platelets and spreading on immobilized F11 antibodies was not inhibited by blocking of FcγRII, suggesting that activation elicited by binding of F11 antibodies to F11R/JAM-A was independent of this receptor.⁷ Furthermore, the presence of F11 caused platelet aggregation in response to normally subthreshold concentrations of ADP, thrombin, or collagen. This effect was sustained even when FcγRII was blocked by specific antibodies¹² which confirmed that there was a component of platelet activation mediated directly by F11 antibodies binding to F11R/JAM-A and independent of crosslinking of the antibodies to FcγRII. The exact nature of this FcγRII-independent activation has not been elucidated.

New light has been shed on F11R/JAM-A’s role in the regulation of platelet activity by experiments performed on F11R/JAM-A knock-out mice. The association of F11R/JAM-A with αIIbβ3 integrin turned out to have a regulatory function and a lack of this protein in the murine model of F11R/JAM-A deficiency resulted in platelet hyperreactivity.⁶⁰ In vivo thrombosis in these mice was enhanced and their platelets showed higher reactivity toward agonists than that presented in wild-type animals. Further studies provided an explanation of the molecular mechanism of this phenomenon. It has been known before that downstream outside-in signaling of integrins is mediated via c-Src kinases associated with integrins. These kinases are maintained in an inactive state by phosphorylation mediated by Csk kinases.⁶¹ F11R/JAM-A turned out to associate in resting platelets with Csk and keep it in the proximity of αIIbβ3 associated c-Src thus limiting αIIbβ3-mediated outside-in signaling.⁵⁵ Csk binding to F11R/JAM-A was dependent on tyrosine 280 phosphorylation on F11R/JAM-A. Dephosphorylation of this tyrosine upon platelet activation caused dissociation of Csk from F11R/JAM-A; therefore, releasing its inhibitory effect on c-Src. In F11R/JAM-A-deficient platelets, Csk is not present in the proximity of αIIbβ3 associated c-Src and, as an effect, outside-in signaling from αIIbβ3 is enhanced and these platelets are more prone to activation by a lower level of stimulus than their wild-type counterparts (Figure 4). Importantly, authors have shown that in F11R/JAM-A-deficient platelets Csk is basically not co-immunoprecipitated with αIIbβ3 suggesting that F11R/JAM-A is the main protein responsible for ensuring Csk-dependent inhibitory activity on outside-in signaling from αIIbβ3.

Figure 4. F11R/JAM-A regulates outside-in signaling of αIIbβ3. In resting platelets Csk kinase is associated with F11R/JAM-A and remains in proximity to αIIbβ3-associated c-Src kinase. Phosphorylation of c-Src mediated by Csk limits outside-in signaling of αIIbβ3. Dissociation of F11R/JAM-A from the proximity of αIIbβ3 upon platelet activation, or F11R/JAM-A deficiency results in cessation of Csk inhibitory effect on c-Src. This in turn enhances outside-in signaling from the integrin.

Estimates of F11R/JAM-A number of copies in human blood platelets mentioned above show that F11R/JAM-A molecules are outnumbered by αIIbβ3 by 3 to 4 times and therefore not all copies of αIIbβ3 could be negatively regulated by F11R/JAM-A at the same time in 1:1 manner. Alternatively, it could be speculated that Csk associated with F11R/JAM-A maintains such a pool of c-Src molecule in a phosphorylated, inactive state which is sufficient to effectively limit αIIbβ3-dependent outside-in signaling.

Recently a novel mechanism of platelet activity regulation by F11R/JAM-A was proposed where binding of soluble F11R/JAM-A to the protein located in the platelet membrane triggers a cascade of platelet activation.³¹ The authors were prompted to study the effects of soluble F11R/JAM-A on blood platelets by an observation that serum levels of the protein were increased in patients with acute coronary syndrome and patients with higher serum concentrations of soluble F11R/JAM-A were at higher risk of recurrent myocardial infarction. Authors reported that the presence of soluble F11R/JAM-A increased the kinetics of platelet spreading on fibrinogen and enhanced platelet response to agonists such as ADP or TRAP, measured both by assessing integrin activation and degranulation, with the use of flow cytometry and by evaluating platelet aggregation. Thrombosis in vitro and in vivo in a murine model was also increased in presence of soluble F11R/JAM-A. These effects were associated with enhanced phosphorylation of p-Src kinase which bears some similarity to the effects of F11R/JAM-A knock-out observed in mice. Authors suggest that soluble F11R/JAM-A by interacting with F11R/JAM-A located in the platelet membrane may emulate trans-homophilic interaction of these two molecules in adjacent cells as it occurs in epithelial or endothelial cells and thus propagate the outside-in signaling. The authors noticed that the physiological relevance of this finding might be disputed as the concentrations of soluble F11R/JAM-A which caused the described effect in experimental conditions were several orders of magnitude higher than that measured in human serum samples. They argued, however, that the concentration of soluble F11R/JAM-A in the thromboinflammatory milieu may be in fact higher than that assessed in the plasma due to intense in situ shedding of the protein from inflamed endothelium.

F11R/JAM-A role in platelet adhesion and consequences for atherosclerosis

Considering that trans-homophilic interactions of F11R/JAM-A molecules contribute to the formation of tight junctions, it could be assumed that such interactions can translate to adhesive properties. Engagement of molecules located on blood platelet and their counterparts presented on activated endothelium or on another platelet could therefore be involved in platelet adhesion to an inflamed vascular wall or platelet recruitment to the growing thrombus.

Such a concept has been positively verified in static conditions as human blood platelets were shown to adhere to immobilized F11R/JAM-A.^31,43 Similarly, platelet adhesion to activated endothelial cells in vitro in static conditions was partially blocked by a soluble form of F11R/JAM-A, peptides which constituted fragments of F11R/JAM-A⁴³ or by silencing of F11R/JAM-A gene.⁶² This indicated that in static conditions, F11R/JAM-A-dependent interactions can contribute to platelet adhesion.

Whether these interactions could ensure platelet adhesion under flow conditions is less evident. Deposition of platelet-derived microparticles on activated endothelium under flow conditions was shown to be partially mediated by F11R/JAM-A as it was reduced in the presence of soluble F11R/JAM-A.⁶³ In line with this, lower deposition of platelet-derived chemokine RANTES was observed when F11R/JAM-A-deficient mouse platelets were perfused over activated endothelial cells expressing F11R/JAM-A when compared to deposition of F11R/JAM-A expressing platelets perfused over the same cells.⁶⁴ In ApoE^−/− mice treated with the peptide which specifically blocked F11R/JAM-A trans-homophilic dimerization⁶⁵ and platelet adhesion to inflamed endothelium under static conditions, platelets interacted less efficiently with vascular wall.⁶⁶

Contrary to these observations, soluble F11R/JAM-A has been shown to increase platelet adhesion to inflamed endothelium in vitro under flow conditions which was attributed by the authors to the activating effect of soluble F11R/JAM-A on platelets.³¹

Another result which undermines the notion that F11R/JAM-A enhances the ability of platelets to adhere to the vascular wall was obtained in a study using hyperlipidemic mice with platelet-specific F11R/JAM-A deficiency. These mice were characterized by a higher release of platelet-derived microparticles and higher adhesion of platelets to activated endothelium than mice expressing F11R/JAM-A in blood platelets.⁵⁷ The mice lacking F11R/JAM-A in blood platelets also more rapidly developed atherosclerotic lesions. These findings were further confirmed by wire injury of the vascular wall in the same animal model. The site of vascular injury was more effectively covered with platelets in mice with F11R/JAM-A-deficient platelets than in mice whose platelets expressed F11R/JAM-A.⁶⁷ Interestingly, interactions observed in vitro under flow conditions of monocytes with adherent platelets were more pronounced in the case of F11R/JAM-A-deficient platelets than with F11R/JAM-A expressing platelets. These effects were explained by higher reactivity of F11R/JAM-A deficient platelets in comparison with WT platelets in a mechanism explained in the previous section.

The higher reactivity of F11R/JAM-A-deficient platelets causes certain difficulties in terms of understanding the role of trans-homophilic interactions of F11R/JAM-A in platelet adhesion with the use of knock-out models. Any potential effect caused by the lack of these interactions can be masked by an enhancement in the activity of other mechanisms in these platelets. The fact that F11R/JAM-A is not indispensable for both aggregation and adhesion does not mean that the protein is devoid of any other function that regulation of αIIbβ3 activity.

As it was discussed above, studies of the role of platelet F11R/JAM-A in the development of atherosclerosis provide conflicting results. One of the aspects that has not been addressed is the level of JAM-A expression on blood platelets among the human population. Taking into account its crucial role in the regulation of outside-in signaling of αIIbβ3 integrin, it would be interesting to assess whether F11R/JAM-A expression in platelets correlates with platelet reactivity, cardiovascular disease, and/or cardiovascular events as it has been shown for the soluble portion of the protein.^31,68

Perspectives

It is still not clarified what is the role of the formation of cis-homodimers in platelets upon activation. It is suggested that cis-dimerization of monomers dissociated from αIIbβ3 keeps them at bay from the integrin and therefore allows the outside-in signaling during platelet activation. Are therefore cis-dimers only a storage form of the protein, or have they some other role to play after platelet activation? Would the inhibition of cis-homodimerization impair this mechanism and limit outside-in signaling thus decreasing platelet reactivity? Animal models expressing in blood platelets F11R/JAM-A with site-specific mutations in the D1 domain, which is responsible for cis-dimerization, would greatly help to answer these questions.

Based on F11R/JAM-A’s involvement in tight junctions formation, an obvious question occurs: can this protein take part in the formation of similar connections between blood platelets during thrombus formation or in any other interaction between platelets and other types of cells? The only evidence published to date in favor of such a role of F11R/JAM-A is an electron microscopy image revealing clusters of F11R/JAM-A on contacts of thrombin-activated platelets.⁵⁶ This raises another question: why do blood platelets contain some of the proteins which play a part in tight junctions formation? According to the platelet proteome identification, approx. 2,400 copies of claudin-5, 1,100 copies of claudin-3, and 3,900 copies of zonula occludens 2 (ZO-2) were detected.^52,53,69 In concert with this, a recent conference communication presented on International Society on Thrombosis and Haemostasis congress in 2022 in London reported enriched ZO-2 in platelet-platelet contacts.⁷⁰ Both proteins are essential components of tight junctions in epithelial and endothelial cells. Claudins are transmembrane proteins directly involved in formation of cell contacts while ZO-2 is a peripheral membrane protein which connects actin cytoskeleton with transmembrane components of tight junctions such as claudins. On the other hand, no occludins, another important transmembrane protein component of tight junctions, were found in platelets’ proteome. These data suggest that blood platelets may form a kind of tight junction analogs. Taking into account the role of tight junctions in regulation of permeability of endothelial and epithelial layers, it is conceivable that similar structures on platelet-platelet contacts may be responsible for the regulation of thrombus porosity.

The involvement of platelets in the regulation of immunological processes is well known and was profoundly reviewed.^71,72 Until today, there are no data regarding platelet F11R/JAM-A involvement in such regulation. There are however certain indications suggesting such a possibility. It has been recently shown that F11R/JAM-A present on dendritic cells is involved in regulation of T cells differentiation.⁷³ The site of contact of these two types of cells where the interaction of the molecules crucial for the process occurs is called immunological synapse. This structure is known to contain LFA-1⁷⁴ which, as mentioned above, is the only known protein interacting with F11R/JAM-A other than F11R/JAM-A itself. Taking this into account, it could be assumed that platelet F11R/JAM-A may be involved in immunomodulatory processes.

It has been shown that F11R/JAM-A in endothelial cells acts as a mechanosensor.⁷⁵ Tension or shear stress imposed on this protein resulted in a series of intracellular signaling events which led to increased cell stiffness. Since a growing thrombus is subjected to a constant shear stress, it is conceivable that platelet F11R/JAM-A located on the luminal thrombus surface as well as the portion of the protein engaged in trans-homophilic interactions between adjacent platelets in deeper layers of thrombus can affect the mechanistic properties of platelets. This in turn could have an impact on thrombus stability.

There is still much to be learned about the many physiological roles of platelet F11R/JAM-A. The innocuous junctional protein might be so much more than its humble name implies. Only time and concerted effort will tell.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the National Science Centre grant OPUS [UMO-2020/37/B/NZ3/00301].