INTEGRIN FUNCTION IN LEUKOCYTE-MEDIATED INFLAMMATION—ACTINOPATHIES IN IMMUNE DISEASES

ABSTRACT

Integrins play a critical role in leukocyte recruitment and activation within inflamed tissues. These heterodimeric cell-surface receptors recognize ligands on vascular endothelium or extracellular matrix to initiate intracellular signals leading to leukocyte adhesion, migration, and activation. The best-described role for integrins is in the leukocyte adhesion cascade, which is the process by which leukocytes exit the blood vasculature and enter the tissues in response to infection or injury. During the adhesion cascade, integrin signaling is required for changes in leukocyte cytoskeletal structure required for firm adhesion to endothelial cells, followed by intravascular crawling and transmigration from the bloodstream into the tissues. During this process, integrin signaling augments leukocytes’ inflammatory and antimicrobial functions. Mutations in the genes encoding integrins or their downstream signaling molecules result in immunodeficiency and altered tissue repair following injury. Many of these mutations occur in proteins involved in the reorganization of the actin cytoskeleton and have become known as actinopathies, the classic example being Wiskott–Aldrich syndrome. We describe a new actinopathy-type mutation in the integrin signaling molecule SKAP2, which is associated with autoimmunity and type 1 diabetes.

INTRODUCTION

The Leukocyte Adhesion Cascade

The leukocyte integrins play a central role in regulating inflammatory cell recruitment to tissue injury or infection sites (1,2). Integrins are the primary cell adhesion receptors that bind to vascular endothelial cell-surface proteins, allowing immune cells to leave the vasculature and enter the tissues. The leukocyte adhesion cascade is the process by which immune cells travel throughout the body and then move into the tissues (3-5). The process is highly regulated and involves multiple steps: leukocyte capture, rolling, activation, firm adhesion, and transmigration. Integrins play a central role in the activation and firm adhesion steps. Leukocyte capture refers to the initial interaction between the immune cell and the vascular endothelium, which slows down the white cell from a free-flowing state in the blood to a rolling state along the endothelium. Leukocyte capture and rolling are mediated primarily by selectins, a family of adhesion molecules on the surface of endothelial cells and leukocytes (6,7). Selectins recognize counter receptors on the leukocyte (or endothelium), forming lower-affinity interactions to start the rolling process along the vascular wall. Rolling is essential for leukocytes to survey the endothelium for signals indicating inflammation. These signals initiate the next step in the cascade, referred to as activation (8). The rolling leukocytes encounter chemokines and other signaling molecules presented on the endothelial surface or released into the local environment. These proinflammatory signals initiate an intracellular signaling response in the leukocyte that activates the integrins’ high-affinity binding state, which causes the cell to arrest (i.e., stop) on the vascular endothelium (9,10). The transition from rolling to firm adhesion also initiates changes in the cytoskeletal structure of the leukocyte, causing the cell to flatten out over the surface of the endothelium. The high-affinity engagement of integrins with their counter receptors on the endothelium is pivotal for stabilizing leukocyte-endothelial interactions. Once firmly adhered, leukocytes spread out and crawl along the endothelial surface (11). This lateral movement helps leukocytes find optimal sites for transmigration out of the vasculature and into the tissue (12). Crawling involves the continued engagement of integrins and their ligands and the reorganization of the leukocyte cytoskeleton. The final step of the leukocyte adhesion cascade is transmigration, also known as diapedesis. Leukocytes pass through the endothelial barrier to enter the underlying tissue. Transmigration can occur through endothelial cell junctions (paracellular route) or directly through endothelial cells (transcellular route) (13). Integrins play a significant role in this process by interacting with endothelial junctional proteins and aiding in the dissolution of these junctions. Strong engagement of integrins by various extracellular matrix proteins in the tissues, in the setting of other inflammatory mediators, also leads to activation of effector functions (especially in neutrophils) (14-16). This will prime the cells for enhanced functional responses (such as phagocytosis or degranulation) that are required for host defense against pathogens but can also contribute to tissue injury if poorly controlled (17). In general, the steps of the leukocyte adhesion cascade are the same when comparing neutrophil recruitment to sites of tissue infection versus lymphocyte entry into lymphoid structures. However, the particular integrins and endothelial counter receptors may differ.

Integrin Signaling Pathways

Integrins are heterodimeric cell surface receptors consisting of an α and β chain (Figure 1). In neutrophils, the predominant integrins are α_Lβ₂ (LFA-1), α_Mβ₂ (Mac-1), α_Xβ₂ (CR4), and α_Dβ_2, which are expressed to a varying degree based on the cellular activation state (14,18,19). In resting cells, flowing freely in the vasculature, these chains form a bent conformation incapable of binding ligands. Upon sensing inflammatory stimuli such as a chemokine or chemoattractant, an intracellular signaling cascade is initiated, which causes a conformational change in the integrin to form a fully open state that has higher affinity and increased avidity for ligands on the vascular endothelium (such as ICAM-1, 2, or 3) or within the extracellular matrix (such as fibrinogen, fibronectin, or collagen) (20). This signaling cascade is called the “inside-out” pathway of integrin activation, which is required to fully arrest the leukocyte along the endothelium. Several key intracellular signaling molecules are involved in inside-out signaling, including PKC, the GTPase Rap1, RIAM, and others (varying between leukocyte types) (21-24). These molecules interact with the cytoskeletal proteins talin-1 and kindlin-3 and associate with the cytoplasmic tails of the β2 integrins to mediate the conformational change, leading to the formation of the open, high-affinity integrin (25).

Fig. 1.

Integrin signaling induces actin cytoskeletal rearrangement. In resting leukocytes, integrins are in an inactive or closed conformation, which blocks the ligand binding domains. External stimuli, such as chemokine binding to GPCR receptors, lead to a signaling pathway involving RAP1 and other molecules, which results in the unfolding of the integrin into an active or open conformation. This pathway is the “inside-out” signaling response. Full integrin activation to the high-affinity state is achieved by association of additional molecules, such as talin and kindlin. The high-affinity integrin then recognizes endothelial adhesion ligands (such as ICAMs or extracellular matrix proteins like fibrinogen), which then initiate a second signaling response referred to as the “outside-in” pathway. This pathway involves tyrosine kinases and actin-associated molecules, such as WASP and SKAP2, that link the integrin to the actin cytoskeleton. This allows for changes in cell shape that are needed for full adhesion, intraluminal migration, and transmigration out of the vasculature into the tissues.

In the fully open conformation, the integrin binds to its cognate ligand. Following the engagement of high-affinity integrins with their ligands, a second signaling pathway is initiated in the leukocyte, referred to as “outside-in” signaling (26). This signaling response leads to changes in the actin cytoskeletal structures of the leukocytes, allowing for changes in cell shape and motility required for cells to exit the vasculature and enter the tissues. This signaling cascade can also modulate gene expression, enhancing effector functions (in neutrophils, this would include degranulation or activation of superoxide production) or cell proliferation (mainly in lymphocytes) (27,28). Several key signaling molecules are involved in the outside-in pathway. This includes Src-family kinases, the tyrosine kinases FAK and Syk, which signal to downstream molecules such as vinculin, cortactin, and paxillin (29,30). These actin-associated molecules work with the Wiskott–Aldrich protein (WASp) to reorganize the actin cytoskeletal and help physically link the integrins to the actin cytoskeleton to provide the mechanical stability required for cell adhesion and migration (31). The Rho family of GTPases, including RhoA, Rac1, and Cdc42, also plays significant roles in cytoskeletal reorganization and cell motility during outside-in signaling (32).

INBORN ERRORS IN INTEGRIN FUNCTION

The Leukocyte Adhesion Deficiency Syndromes

Inborn errors (i.e., genetic mutations) affecting leukocyte integrin function lead to immune dysfunction, manifesting in immunodeficiency, autoimmunity, or dysregulated inflammation (33,34). These mutations are generally divided into two types: those affecting integrin receptors and those affecting integrin intracellular signaling molecules. One of the most well-known disorders associated with integrin malfunction is leukocyte adhesion deficiency (LAD) syndrome. There are three types of LAD (33,35,36). LAD1 results from mutations affecting the β₂ integrin itself, resulting in loss of cell surface expression (which is easily recognized by flow cytometry of peripheral blood leukocytes—looking for reduced CD18 expression). LADII is caused by mutations in fucose transporters required for proper glycosylation of selectin receptors. This results in a failure to initiate the early capture and rolling steps of the adhesion cascade. LADIII is caused by mutations affecting kindlin-3, leading to poor integrin activation with reduced integrin affinity and avidity. The LAD syndromes are all loss-of-function mutations that present with immunodeficiency (frequent bacterial or fungal infections). Kindlin-3 deficiency also impairs platelet function, leading to increased bleeding (37). Another classic integrin receptor defect is Glanzmann thrombasthenia, which involves mutations in the genes encoding the glycoprotein (GP) IIb/IIIa complex, also known as integrin αIIbβ3. Mutations in either gene result in defective or absent GP IIb/IIIa expression, leading to impaired platelet aggregation, prolonged bleeding times, and an increased risk of severe bleeding episodes (38).

The Actinopathies

Several genetic mutations that result in altered integrin function have been found in the proteins involved in integrin signaling pathways. These mutations have a common pathophysiology in that they all affect (in one way or another) the reorganization of the actin cytoskeleton during leukocyte adhesion and migration. These mutations are part of a broader class of genetic inborn errors called actinopathies, which can affect immune function, tissue/organ development, and even neurologic function (39). The number of disease-associated genetic mutations classified as actinopathies has grown exponentially over the last five years as exome and genome sequencing have become widely applied (40). The most well-studied actinopathy mutation affecting primarily leukocyte integrin function is Wiskott–Aldrich syndrome. Loss-of-function mutations in WASp lead to defects in actin branching following integrin adhesion, affecting all leukocyte types. This results in immune deficiency (and platelet dysfunction) characterized by recurrent sinopulmonary infections, eczema, bleeding diathesis, and, in some cases, even autoimmunity. A similar phenotype is seen in patients with coronin 1A deficiency, which is particularly important in lymphocyte integrin signaling. Coronin 1A functions with the Arp2/3 complex to regulate actin polymerization; loss of coronin 1A results in impaired T cell development and selection, which in its most severe form can result in T-B+NK+ severe combined immunodeficiency (41,42). Another loss-of-function mutation affecting leukocyte integrin function is defects in the gene encoding the WDR1 protein, which functions in the elongation of actin filaments (41,43). WDR1 mutation results in poor neutrophil adhesion, migration, and activation (often called the lazy leukocyte syndrome), which presents as frequent oral and genital infections (along with frequent bleeding due to platelet dysfunction).

Not all actinopathies result from loss-of-function mutations. Gain-of-function mutations (which affect only one gene allele and are hence heterozygous in affected individuals) can also produce dysregulation of actin cytoskeletal organization, producing hyperactive leukocyte adhesive phenotypes and altering immune cell activation. A classic example is X-linked neutropenia, which results from gain-of-function mutation in WASp (44). The dysregulated actin polymerization dramatically affects neutrophil development from myeloid progenitors; the few cells that do appear in the blood have a hyperactive phenotype. As a result, individuals with X-linked neutropenia present early in life with multiple bacterial infections. Gain-of-function mutations in the RAC2 GTPase, which is involved in outside-in integrin signaling, have also been described (45). These mutations produce hyperactive actin polymerization and branching at the leading edges of migration. In their most severe form, this leads to impaired myeloid and lymphoid cell development, producing severe combined immunodeficiency. The few neutrophils in the blood produce excessive reactive oxygen species following integrin-mediated adhesion. The many types of inborn genetic errors that result in immune defects due to altered actin cytoskeletal regulation are summarized in Papa, et al (46).

SKAP2 IN INTEGRIN SIGNALING

SKAP2 Gain-of-Function Mutation as a New Actinopathy-Associated Disease

We recently described the discovery of a single point mutation in the protein Src-kinase associated protein 2 (SKAP2), which functions in the outside-in integrin signaling pathway, in a young patient with autoimmunity and type 1 diabetes (47). SKAP2 is an intracellular adapter protein that links integrins to the actin cytoskeleton through association with WASp and talin-1 (48). SKAP2 is found primarily in myeloid leukocytes. SKAP2-deficient mice manifest defects in leukocyte recruitment to inflammatory sites, increasing susceptibility to bacterial and fungal infections (49-51). Neutrophils from SKAP2-deficient mice have impaired β2 integrin-mediated adhesion, resulting in poor superoxide production when plated on extracellular matrix-coated surfaces. SKAP2 is regulated by conformational changes. The protein is held in an auto-inhibited state in resting cells through an intramolecular association of the dimerization and pleckstrin homology domains. Following integrin activation, SKAP2 becomes tyrosine phosphorylated (by Src-family kinases that initiate the outside-in integrin pathway), which breaks the intramolecular association, allowing for the unfolding of the protein and association with WASp and talin-1 (52). The expression of an engineered SKAP2, with a single substitution mutation in the pleckstrin domain that blocks the formation of auto-inhibited conformation, leads to hyperactive actin polymerization and increased adhesion in macrophages (53).

Through a collaboration with the University of Chicago Monogenic Diabetes Registry Study (https://monogenicdiabetes.uchicago.edu), which collects exome sequencing data in individuals with type 1 diabetes and associated autoimmune diseases, we identified a young woman with a single point mutation in pleckstrin homology domain of SKAP2. Polymorphisms in the SKAP2 gene have been widely described in various genetic studies of individuals with type 1 diabetes. In perhaps the most extensive genetic study of type 1 diabetic patients to date (61,427 participants), the SKAP2 locus was identified as having the second largest number of single nucleotide polymorphisms, just behind IL-27, estimated to be present in nearly 30% of type 1 diabetic patients (54). The mechanism by which these SKAP2 variants contribute to the genetic risk for type 1 diabetes is unclear.

Our patient is heterozygous for a single amino acid substitution in SKAP2 (G153R). Besides type 1 diabetes, she has several additional autoimmune syndromes, including Hashimoto’s thyroiditis, Raynaud’s syndrome, and bouts of autoimmune hemolytic anemia. Examination of her peripheral blood monocytes showed an increased association of SKAP2 and WASP with the plasma membrane in resting cells (similar to that seen in chemokine-stimulated normal cells). Plating of these cells on integrin ligand-coated tissue culture plates resulted in increased downstream integrin signaling events. See Rutsch, et al (47) for the immunofluorescent photomicrographs. We also observed that the patient’s neutrophils exhibited increased production of reactive oxygen species when plated on integrin ligand-coated plates (Figure 2), which is the opposite of what is seen with murine neutrophils derived from SKAP2-deficient mice. These data suggest that G153R mutation results in hyperactive integrin outside-in signaling, leading to dysregulation of actin cytoskeletal structures. We propose that the gain-of-function mutation in SKAP2 is another example of an actinopathy-type inborn error of integrin signaling. Our ongoing work involves modeling the mutation in mice by generating knock-in animals carrying the murine G153R substitution mutation (made through CRISPR mutagenesis methods). We are currently studying the integrin signaling defects in these animals.

Fig. 2.

Increased neutrophil superoxide production in SKAP2 G153R mutant individual. Peripheral blood neutrophils from the SKAP2 patient or parent control were plated on (A) poly-RGD peptide-coated wells, (B) pRGD in the presence of TNFα, or (C) pRGD + LPS, and ROS was monitored by luminol reduction as described (55). Note Y-axis scale differences in samples stimulated with TNFα or LPS, which activates integrin signaling (n=3 per time point, error bars = SEM, ****p <0.0001 by Kruskal–Wallis test).

The Chicago Monogenic Diabetes Registry Study has resulted in the identification of several families with mutations in integrin signaling molecules (including another family with a single point mutation in the SKAP2 dimerization domain, which should produce the same integrin hyperactivation phenotype that we observed with our index patient). The affected gene products function in both the inside-out and outside-in integrin signaling pathways. This suggests that dysregulation of leukocyte integrin signaling should be considered a potential genetic risk factor for type 1 diabetes, similar to the known involvement of altered integrin signaling in other autoimmune and inflammatory diseases. Therapeutic modulation of leukocyte integrin signaling may offer alternative approaches to limiting disease progression in subsets of individuals with type 1 diabetes.

CONCLUSION

The extravasation of leukocytes from the peripheral blood into tissues is a highly ordered process that depends on the regulated intracellular signaling pathways from cell adhesion receptors (selectins and integrins). Dysregulation of these pathways, either through excessive inflammatory signaling or inborn genetic mutations, leads to a growing list of immune disorders. We suggest that the list should include type 1 diabetes. In the era of personalized molecular medicine, it is clear that a more in-depth evaluation of the immune status of all these patients is warranted to direct appropriate therapy at underlying disease drivers. From a scientific point of view, we clearly need a better understanding of the immune mechanisms that lead to any particular disease in an individual with altered integrin signaling or function. For example, why do gain-of-function mutations in SKAP2 associate with type 1 diabetes but the related gain-of-function mutations in WASp cause neutropenia? We simply don’t know. Obviously, alteration in integrin signaling will have different effects on different leukocyte subsets (e.g., SKAP2 deficiency has little effect on T lymphocytes because it is not expressed in these cells, whereas WASp is). Dissecting the molecular mechanisms by which specific alterations in leukocyte adhesion signaling lead to particular immune disorders is a clear priority for investigators in this field.

DISCUSSION

Tweardy, Houston: Well, thank you, for that excellent talk. You nicely summarized that the macrophages are the cells that lead to this increased inflammatory process that leads to type-1 diabetes. Are there any other contributing cells that have the SKAP to gain a function?

Lowell, San Francisco: That’s a good question. Remember, this is a germ-line mutation, so every cell in the body has this mutation. We know for sure that macrophages are hyperactivated, but we really focused on the dendritic cells in the mouse model. This is the advantage of building the mouse model because we can take it apart and look at it. We know that the dendritic cells in mice have enhanced adhesion to T cells through integrin-mediated pathways. This results in increased antigen presentation and increased selection for self-reactive T cells. The neutrophils probably also contribute; they contribute to the general inflammatory state that occurs. Once diabetes gets going in a mouse, it becomes systemic. These mice have other autoimmune syndromes, sialadenitis, pulmonary inflammation, and all that—usually later if they survive from diabetes—so it’s probably a contribution of many factors. Remember, though, that SKAP2 is primarily expressed in the innate immune system. This is another example of mutations in the innate compartment driving autoimmune disease which is dependent upon the adaptive compartment. When you talk to many immunologists about autoimmune disease, they think it’s a disease of lymphocytes. Well, this is a disease of macrophages, dendritic cells, and maybe even neutrophils. To really understand what cells individually contribute, we have to learn how to make these mutations in each specific lineage alone. That technology is evolving; pretty soon, we’ll be seeing that though.

Tweardy, Houston: Thanks for that answer. This may be a bit moot based on your answer, but one of the questions is in the ill-defined area of M-1 versus M-2 macrophages. Does this link in any way to the phenotype of macrophages in the patient or mice?

Lowell, San Francisco: Yes, the macrophages are all type M-1. This is a proinflammatory IFNγ-driven disease. We can see that by single-cell RNA-seq, which we’ve done up the wazoo, and it’s an M-1 disease.

Tweardy, Houston: All right, thanks.

Anderson, Chicago: Thank you for that talk. My name is Mark Anderson from the University of Chicago, and I’m not an endocrinologist or even an immunologist.

Lowell, San Francisco: You probably know Lou Philipson, right?

Anderson, Chicago: I know Lou, and I actually know Mark Anderson from the University of California, San Francisco (UCSF) because we sometimes get each other’s mail, but not for a while. Given the ubiquity of integrins and of actin, I just wanted to expand on the previous question, which is every cell type, except maybe red blood cells, should be affected by things that go wrong in that collaboration. Are they always inflammatory? Does it affect the neuro system and cognition? Does it affect the heart? What are the other things you might see?

Lowell, San Francisco: This is leukocyte specific for sure. There are many different integrin signaling molecules in different cell types. For example, cardiac cells also use integrins, especially during development, to guide cell migration. However, they have different integrins on the cell surface. Most integrin-related mutations found in people seem to affect the immune system and platelets, even though these mutations are present in every cell in the body. In mice, engineered mutations that delete integrins involved in cardiac or neuro development are usually lethal, so we would not see them in humans. Mainly, the mutations to affect integrin signaling molecules expressed only in immune cells tend to show up in people. For example, SKAP2 is really only expressed in myeloid leukocytes; so SKAP2 mutant mice or people don’t have a neurologic or cardiac phenotype. WASp, on the other hand, is expressed in platelets, so those patients have a large coagulopathy and other developmental problems.

Anderson, Chicago: Thank you.

Hochberg, Baltimore: There actually is a surgeon at New York University (NYU) with the same name as me, and we do get confused sometimes. My question relates to your index case. You noted that she had Raynaud’s phenomenon so I’m wondering if that’s a red herring. Was there any evidence of vascular abnormalities in your mice model and/or science with the systemic autoimmune rheumatic disease of which Raynaud’s phenomenon is associated?

Lowell, San Francisco: It’s hard to know. I haven’t seen good mouse models of Raynaud’s. SKAP2 mutant animals get systemic autoimmunity much worse than the wild types do. We can take serum from them early in life and do antigen autoantibody arrays and they light up everything. We’re doing a lot of PhIP-Seq on them right now as well. This is another lesson for all of us—type 1 diabetes is just an autoimmune disease, and most patients with type 1 diabetes have other autoimmune sequelae. When we managed type 1 diabetes when I was in med school, we thought of it as a unique single type of disease in a patient. However, it’s really just a manifestation of another type of autoimmunity. We are going to test these animals and other autoimmune models, and we expect that they will be accelerated. Somebody should have asked me whether or not SKAP2 mutations are present in other human autoimmune diseases. It turns out that SKAP2 mutations are found in a variety of other autoimmune processes, which emphasizes that some of the same pathogenic mechanisms that lead to or promote genetic risk for type-1 diabetes also provide risk for other autoimmune diseases.

Schnapp, Madison: Thank you. In your index patient or your mouse model, this seems to be specific for the β2 integrins. Are you looking at treatment with blocking antibodies?

Lowell, San Francisco: Oh, great question. That came up relatively early when we figured out that she had a gain-of-function mutation in the integrin outside-in pathway. We realized that we could treat this patient with, for example, Ibrutinib. Ibrutinib blocks one of the kinases in the outside-in pathway, not Src-kinases but related BTK kinases, and has been used obviously in cancer immunotherapy therapy. We actually have looked into it and not a single company would sponsor an N-of-1 trial in our patient with Ibrutinib, so she’s progressed. Her diabetes will never be cured; her pancreatic β cells are gone, but she’s actually developed a lot of sclerosis and other forms of autoimmune sequelae, and it’s kind of progressing. Other integrin blockers are available, and there are other drugs that you can potentially use. That’s a great question for all the other patients with type 1 diabetes or other autoimmune diseases in which we’re identifying mutations that affect leukocyte integrin pathways: are these actionable?