Mind the GAP: RASA2 and RASA3 GTPase Activating Proteins as gatekeepers of T cell activation and adhesion

Abstract

T cell receptor and co-receptor stimulation integrate multiple intracellular signals to initiate T cell proliferation, migration, gene expression, and metabolism. Among these molecules are the small GTPases RAS and RAP1, which induce MAPK pathways and cellular adhesion to activate downstream effector functions. While many studies have helped elucidate signaling intermediates mediating T cell activation, the molecules and pathways that keep naïve T cells in check are less appreciated. Several recent studies provide evidence that RASA2 and RASA3, which are GAP1-family GTPase-activating proteins (GAPs) that inactivate RAS and RAP1, respectively, are crucial molecules that limit T cell activation and adhesion. In this review, we describe recent data on the roles of RASA2 and RASA3 as gatekeepers of T cell activation and migration.

Balancing T lymphocyte activation

A key feature of T lymphocytes is their ability to become activated by foreign or dangerous antigens, but not by self. Multiple mechanisms participate in this process: during T cell development, T cells with high affinity for self-antigens are purged, yielding a peripheral population of T cells with low affinity for self-antigens yet high affinity for foreign antigens [1]. In mature T cells, both cell-intrinsic (expression of inhibitory receptors, activation of phosphatases, ubiquitination of signaling molecules, and induction of apoptosis) and extrinsic mechanisms (inhibition by regulatory T cells) help limit inappropriate activation [1]. However, many of these inhibitory mechanisms also restrict T cell activation during prolonged antigen exposure in cancer and chronic infection, leading to a state known as “T cell exhaustion”. Understanding the gatekeepers that limit T cell activation is, therefore, essential for understanding the execution of effective immune responses in humans. While the roles of inhibitory receptors and phosphatases have received wide attention, several recent Clustered Regularly Interspaced Short Palindromic Repeats-Cas9 (CRISPR-Cas9) (See glossary) screens have revealed that the RASA2 and RASA3 GTPase activating proteins (GAPs), which inactivate RAS and RAP1, respectively, are key players that maintain T cells in a repressed or less active state [2–4]. The biological functions of these GAPs, how they are regulated to permit T cell activation, and new insights into their emergent roles in different T cell populations from the CRISPR-Cas9 screens, are the subjects of this review.

T cell activation

When a naïve T cell encounters an antigen presented in the context of MHC, engagement of the T cell receptor (TCR) in conjunction with co-stimulatory molecules activates a series of tyrosine kinases, leading to activation of phospholipase C-γ1 (PLCγ1) and subsequent Ca²⁺ mobilization and DAG generation, which result in changes in transcription, translation, and metabolism; these are essential for T cell activation and expression of effector cytokines (Box 1, Figure I) [1]. Mutations affecting components of these signaling pathways cause a wide range of primary immunodeficiencies [5].

Box 1: TCR signaling pathways.

TCR engagement initially activates a series of tyrosine kinases, starting with LCK which phosphorylates the intracellular tails of the TCR/CD3 complex, leading to recruitment, phosphorylation and activation of ZAP-70 [1]. Together these kinases phosphorylate downstream molecules that form a complex nucleated by the adaptors LAT and SLP-76, leading to recruitment of other adaptor molecules (such as GADS, GRB2, ADAP, and SKAP55) and enzymes (ITK and PLC-γ1) leading to PLC-γ1 activation [1]. PLC-γ1 catalyzes the cleavage of the membrane phospholipid PI(4,5)P₂ to: 1) IP₃, which stimulates Ca²⁺ mobilization and the subsequent activation of the NFAT transcription factors, and 2) Diacylglycerol (DAG), which helps recruit proteins containing C1 domains. Key DAG-binding proteins include Protein Kinase Cθ, an essential component of activation of NFκb and JNK (not shown), and RASGRP1, a key activator of the RAS-RAF-MEK-MAPK cascade, which contributes to activation of AP1 transcription factors and many other downstream effectors involved in cytokine production, proliferation, differentiation, and metabolic remodeling [1]. Other downstream effectors of RAS are not shown. RAS is also activated by SOS1 and 2, two homologous GEFs that are recruited to LAT via the adaptor molecule GRB2 and is inhibited by RASA2 and other RAS-GAPs. Phosphoinositide 3 kinase (p85α and p110δ) also contributes to these pathways through the generation of PI(3,4,5)P₃ (PIP₃) that helps recruit proteins with Pleckstrin homology (PH) domains [1]. Together, these events lead to a wide range of downstream changes in transcription and translation that are essential for T cell activation and expression of cytokines and other effector molecules.

Figure I.

Components of TCR signaling. TCR: T cell receptor; CD3: invariant chains of the TCR including ζ chain; LCK: Lymphocyte Cell-Specific Protein-Tyrosine Kinase; ZAP70: ζ chain of T cell receptor associated protein kinase 70; LAT: Linker for Activation of T cells; SLP76: SH2 Domain-Containing Leukocyte Protein Of 76 KDa; GADS: also known as GRAP2, GRB2 Related Adaptor Protein 2; GRB2: Growth Factor Receptor-Bound Protein 2; ITK: Inducible T cell kinase (a tyrosine kinase); PLCγ1: phospholipase C gamma 1; SOS1: Son of Sevenless Guanine Exchange Factor; RASGRP1: Ras Guanyl Releasing Protein 1; RAS: Ras GTPase; RASA2: RAS P21 protein activator 2; RAF: RAF1 serine-threonine kinase; MEK: also known as MAP2K1, Mitogen Activated Protein Kinase Kinase 1; ERK: Extracellular signal-related kinase; p110δ: catalytic subunit of Phosphoinositide 3 kinase δ (PI3Kδ); p85α: regulatory subunit of PI3Kδ. This figure was created with Biorender.

A second major outcome of T cell activation is an increase in cell adhesion [6], secondary to a change in the conformation of integrins [7–9]. These heterodimeric cell surface adhesion receptors bind to ligands on other cells or the extracellular matrix, allowing T cells to migrate and interact with their environment. The importance of integrin activation in leukocyte function is highlighted by several primary immunodeficiencies, Leukocyte Adhesion Deficiencies, with mutations affecting integrin expression or activation [10]. Among the major integrins on T cells are LFA-1 (αLβ2), which binds to ICAM-1 on other immune cells such as dendritic cells, B cells, and endothelial cells to facilitate transendothelial migration into lymph nodes and T cell activation [7,11], and α4β7, which mediates migration to the gut [12]. In naïve resting T cells, LFA-1 exists in a folded conformation that binds ICAM-1 with low affinity [9]. Stimulation of T cells, either through the TCR or through chemokines, leads to a conformational change of LFA-1 and other integrins that increases their affinity towards their ligands via a process known as inside-out signaling (Box 2, Figure II) [7–9].

Box 2: TCR-mediated LFA-1 activation.

Activation of integrins via TCR signaling.

In naïve resting T cells, the integrin LFA-1 exists in a folded, closed conformation that binds ICAM-1 with low affinity. Stimulation of T cells, either through the TCR or through chemokines [7], leads to activation of the RAP1 GTPase via the CRKL/C3G complex, which is recruited to ZAP-70 [26], and RasGRP2 (CALDAG-GEF1), which is activated by Ca²⁺ binding [74]. Signaling pathways involved in the activation of RAP1 include the recruitment and activation of the adaptor molecule ADAP, which binds to the SLP-76 complex and recruits SKAP-55 and RIAM [75]. The RAP1-GAP RASA3 inhibits RAP1 activation. Active RAP1 interacts with RAPL and RIAM to recruit KINDLIN3 and the cytoskeletal molecule TALIN1, respectively. TALIN1 and KINDLIN3 bind the intracellular tail of integrins, resulting in a conformational change of LFA-1 (and other integrins) that increases integrin affinity towards their respective ligands [6]: this process is known as “inside-out signaling” [7]. In turn, ICAM-1-binding results in increased “outside-in” signals from integrins involving RAP1, KINDLIN3, TALIN1 and other proteins such as FAK and Paxillin (not shown) that link integrins to the cytoskeleton [6]. In conjunction with clustering and increased expression of LFA-1, integrin-mediated adhesion is increased upon T cell activation by the TCR, stabilizing interactions in the immunological synapse between T cells and antigen-presenting cells. Similar mechanisms occur in T cells upon chemokine signaling to initiate migration and transendothelial migration into lymph nodes, as well as in other hematopoietic cells, including myeloid cells and platelets.

Figure II.

Components of TCR-mediated activation of LFA-1. ADAP: Adhesion and degranulation promoting adapter protein; SKAP55: SRC kinase-associated phosphoprotein of 55 kDa; CRKL: CRK-like proto-oncogene; C3G: a GEF for RAP1, also known as RAPGEF1; RASGRP2: also known as CALDAG-GEF1; RASA3: RAS P21 protein activator 3; RAP1: RAS-related protein 1: RAPL: Regulator of adhesion and cell polarization enriched in lymphoid tissues, also known as RAS Association Domain Family Member 5 (RASSF5) or NORE1; KINDLIN3: a member of the KINDLIN family of integrin interacting proteins; TALIN1: Talin1 cytoskeletal protein; RIAM: RAP1-GTP Interacting Protein. This figure was created with Biorender.

TCR signaling leads to downstream activation of a wide range of effectors, including the RAS and RAP1 small GTPases [13]. These structurally homologous members of the RAS family of proteins transduce signals downstream of multiple receptors, including tyrosine kinase growth factor, antigen, G protein-coupled, and cytokine receptors. Activation of RAS initiates the RAS-RAF-MEK-ERK cascade of Serine/Threonine and Threonine/Tyrosine dual kinases which activate transcription factors, including AP1 family members, and other effectors, promoting gene expression, proliferation, and cell survival (Box 1, Figure I) [14]. The importance of RAS is highlighted by the number of cancers, including pancreatic, colon, and melanoma [15,16], as well as developmental disorders, known as Rasopathies, that result from mutations affecting RAS signaling pathways [17] (Clinician’s corner).

BOX: CLINICIAN’S CORNER.

• RAS is an essential component of cellular activation and proliferation that contributes to multiple diseases, including cancer and Rasopathies, a series of genetic developmental disorders including Noonan Syndrome and Neurofibromatosis type 1 (NF1), defined by altered activation of RAS signaling pathways.

• T cell adhesion, integrin activation and migration, which are activated by the RAP1 GTPase, are also implicated in human disease, including thrombotic cardiovascular events, multiple sclerosis, ulcerative colitis and Crohn’s disease, for which integrin antagonists have been developed [83]. Conversely, defects in integrin adhesion resulting from mutations affecting the β2 chain of integrins and signaling proteins upstream of integrin activation such as KINDLIN3, are associated with a group of severe primary immunodeficiencies, the Leukocyte Adhesion Deficiencies (LAD).

• The implication of RASA2 and 3 as gatekeepers of T cell activation and adhesion, that target RAS and RAP1, respectively, raises the question of whether these molecules may provide therapeutic targets for manipulating T cell dysfunction in disease.

• Targeting RASA2 increases T cell sensitivity and improves T cell immunotherapy against low affinity targets (or low antigen-expressing targets) suggesting that inhibiting or targeting RASA2 may help improve immunotherapy against cancer, including reinvigoration of T cells during exhaustion and in adoptive cell therapies. Nonetheless, as RASA2 is mutated in several tumor types, inhibiting RASA2 in vivo may be challenging. However, such therapeutics mays be of particular interest for adoptive cell immunotherapy of solid cancers, where RASA2-deficiency has shown improved adoptive T cell responses in mouse models.

• Activation of RASA2 or 3 might also be useful to dampen T cells activation in autoimmunity. However, mice with Rasa3-deficient T cells fail to develop EAE, suggesting that inhibiting RASA3 might be a therapeutic approach for certain types of autoimmunity. Nonetheless, the broad expression of Rasa3 and its involvement in hematopoiesis and platelet function highlight the challenges of broadly targeting integrin activity.

Originally identified as an inhibitor of RAS [18] that contributes to T cell anergy [19], activated RAP1 is now recognized to recruit KINDLIN3 and the cytoskeletal molecule TALIN1, which bind the intracellular tail of integrins [7,20,21]. This results in a conformational change of LFA-1 (and other integrins) that increases integrin affinity towards its ligands (Box 2, Figure II) [9]. In turn, binding of ICAM-1 leads to “outside-in” signaling involving RAP1, KINDLIN3, TALIN1, as well as Focal Adhesion Kinase (FAK), CasL and Paxillin, which induce cytoskeletal changes contributing to cell adhesion, morphological changes, and migration [7,11,22].

RAS family GTPases (including RAS and RAP1) are active in their GTP-bound state, allowing binding to and activation of downstream effectors [23]. Hydrolysis of GTP to GDP prevents these interactions and inactivates RAS proteins (Figure 1). To regulate RAS and RAP1, Guanine exchange factors (GEFs) are well-known to induce the exchange of GDP with GTP, resulting in activation of the proteins. In parallel, GTPase-activating proteins (GAPs) are known to inactivate the RAS family of proteins by inducing intrinsic catalytic GTPase activity. Some GEFs and GAPs are promiscuous and regulate multiple RAS-family proteins, whereas others are more selective, targeting only a subset [24,25]. Consequently, the activity of RAS-family members is regulated through a multitude of mechanisms.

Figure 1: RAS/RAP1 circuits in T cells;

TCR stimulation leads to activation of both RAS and RAP1. The RAS-GEFs (SOS1 and RASGRP1) are activated downstream of TCR. This leads to an accumulation of RAS-GTP which in turn activates T cell survival and proliferation and regulates differentiation. Simultaneously, TCR signaling leads to activation of RAP1-GEFs (C3G, RASGRP2), which results in RAP1-GTP accumulation and activation of integrins such as LFA-1 (inside-out signaling) and actin remodeling. The mechanisms by which the RAS and RAP1-GAPs are regulated are less clear [35], but RASA2 and RASA3 appear to be major GAPs for RAS and RAP1, respectively. GAP1 proteins are in red, and the key GAP1 members are in bold. This figure was created with Biorender.

In T cells, RAS activity is activated by the RAS-GEF, SOS1, which is recruited to the LAT-SLP-76 signalosome, and RASGRP1, which is recruited by Diacylglycerol (DAG); both GEFs are activated downstream of TCR signaling (Box 1, Figure I). RAP1, by contrast, is activated by the RAP1-GEFs, C3G and RASGRP2 (CALDAG-GEF1) (Box 2, Figure II). C3G is recruited to ZAP-70 via CRKL, which coimmunoprecipiates with ZAP-70 upon TCR stimulation [26]. RASGRP2 is activated primarily by Ca²⁺ [27] and plays an important role in RAP1 activation in both platelets and T cells, as evidenced by defective RAP1 activation in immune cell populations from Rasgrp2-deficient mice [28,29].

RAS/RAP1-GAPs in T cells

Whereas the roles of RAS and RAP1-GEFs in T cells have been studied extensively, the GAPs responsible for switching off RAS and RAP1 activity are perhaps less well appreciated. There are multiple GAP families, defined by conserved protein domain structures, including GAP domains that have activity for RAS, RAP1 or both GTPases (dual RAS/RAP1-GAPs, Box 3. Figure III). The GAP1 family, which includes RASAL1 (previously known as RASAL), RASA2 (GAP1^m), RASA3 (GAP1^IP4BP), and RASA4 (CAPRI), contains two conserved Ca²⁺-binding C2 domains and a Pleckstrin homology (PH) domain/BTK domain that can bind phospholipids or other proteins. The PH domains of RASA2 and RASA3 have high affinity for PI(3,4,5)P₃, a product of Phosphoinositide-3-kinase (PI3K), as shown by high-resolution quantitative mass spectometry of proteins that bound agarose beads coupled to different PIP isoforms [30]. The RASA3 PH domain was also previously found to bind Inositol(1,3,4,5)P₄ (IP₄) and PI(4,5)P₂ [31,32].

Box 3: RAS and RAP1 GAPs.

Members of the RAS and RAP1 families of GAPs.

A) The GAP1 family of RAS/RAP1-GAPs consists of RASAL1, RASA2, RASA3 and RASA4. B) RAS-specific GAPs are classified into multi-protein families that have homologous RAS-GAP-like motifs [76,77]. RAS-specific GAPs elicit their functions by supplying a catalytic arginine residue into the active site of RAS (RASA2 aa: R511), inducing the degradation of GTP to GDP [78]. The IQGAP proteins have limited or no GAP activity. C) RAP1-specific GAPs have 2 primary families. These RAP1-specific GAPs include RAP1GAP I and II [79] and the SPA-1 family of proteins [80]. RAP1-specific GAP domain functionality relies on a catalytic asparagine (RAP1GAPI aa: N290), through a distinct mechanism from RAS-GAPs, to stimulate GTP hydrolysis in RAP1 [81].

Several RAS-GAP proteins have dual RAS/RAP1-GAP activity, including the SynGAP family and members of the GAP1 family, RASAL1, RASA3, and RASA4 [25]. The ability of the GAP1 family to elicit dual RAS/RAP1-GAP activity is thought to rely on an arginine (RASA3 aa: R371) similar to RAS-specific GAPs [25,54].

Figure III.

Domains of GAP proteins: C2: Ca²⁺-binding domain; PH: Pleckstrin Homology domain; BTK: Bruton’s tyrosine kinase domain; SH1: SRC Homology 1 domain; SH2: SRC Homology 2 domain; EC: Extracellular domain; TM: transmembrane domain; JM: Juxtamembrane domain; CH2: calponin-homology 2 domain; WW: poly proline protein–protein domain; IQ: IQ calmodulin-binding motif; PDZ: PDZ domain-like motifs. This figure was created with Biorender.

Early work, including biochemical studies in human peripheral blood T cells and the Jurkat cell line, as well as studies of T cells from Nf1^+/− and Rasal3^−/− mice indicated a role for RAS-GAPs RASA1, NF1, and the SynGAP, RASAL3, in RAS activation, and T cell development, survival, and function [33–35]. However, data from recent CRISPR-Cas9 screens suggest that RASA2 is also a key regulator of T cell RAS activity in T cells (see below) [2].

For RAP1-GAPs, Signal-Induced Proliferation-Associated Protein 1 (SPA-1, encoded by Sipa1) appears to be important in T cells; T cells from Sipa1^−/− mice exhibit elevated TCR-induced RAP1 activation and are poorly responsive to secondary recall challenges relative to wildtype mice [36]. However, a recent CRISPR-Cas9 screen from our laboratories, as well as other studies of RASA3-deficient T cells suggest that RASA3 is also a major regulator of RAP1 activity and LFA-1-mediated ICAM-1-binding in mouse T cells (discussed below) [4,37]. Thus, genetic screens have revealed RASA2 and RASA3 as important components of T cell signaling (Figure 1).

RASA2 as a gatekeeper of T cell activation

RASA2 was identified as an inhibitor of RAS signaling in multiple human tumors, including melanoma, where inactivating RASA2 mutations have been found [38]. RASA2 is also one of the genes mutated in Noonan syndrome, a Rasopathy associated with developmental defects and elevated RAS activity [39]. RASA2 is expressed in both CD4⁺ and CD8⁺ T cells [2], yet, until recently, its role in T cells remained unexplored.

One group of researchers performed a whole-genome CRISPR-Cas9 screen for genes that controlled TCR-induced proliferation in human primary CD8⁺ T cells from peripheral blood. Specifically, using dilution of Carboxyfluorescein succinimidyl ester (CFSE) as a sensitive measure of cell division (see Figure 2A), they identified RASA2 as a negative regulator of T cell proliferation [3]. CRISPR-Cas9-mediated deletion of RASA2 resulted in a 2-fold increase in proliferation of T cells, as well as increased expression of the activation markers, CD69 and CD154, in response to stimulation with anti-TCR antibodies in vitro. Furthermore, RASA2 targeting in TCR-transduced primary human T cells increased antigen-specific killing of cancer cells in culture as measured by increased annexin staining and decreased cell numbers [2,3].

Figure 2: CRISPR-Cas9 screens identifying RASA2 and RASA3 as regulators of T cell proliferation and adhesion;

A) CRISPR-Cas9 screen for cell proliferation [2,3]. Single guide RNA (sgRNA) lentiviral infection was followed by Cas9 protein electroporation (SLICE) to perform a CRISPR-Cas9 screen using dilution of Carboxyfluorescein succinimidyl ester (CFSE) as a sensitive readout for cell division. RASA2 was identified as a negative regulator of T cell proliferation that increased cell proliferation under multiple immunosuppressive conditions, including adenosine, Cyclosporin or Tacrolimus, TGFβ, and T_reg cells [2,3]. Cells were sorted by flow cytometry into CFSE^lo (highly proliferating) and CFSE^hi (low proliferation) gates and guide sequences were sequenced to determine enrichment in each population. B) CRISPR-Cas9 screen for TCR-induced LFA-1-mediated ICAM-1-binding [4]. A CRISPR-Cas9 screen was performed in primary mouse T cells by transducing a retroviral library of sgRNAs directed against PIP₃-binding proteins and controls into Cas9-expressing cells and using a flow cytometry-based assay for TCR-induced binding of fluorescently-labeled multimerized ICAM-1. High and low ICAM-1-binders were sorted and gRNAs were sequenced to determine enrichment in each population [4]. Both screens relied on highly sensitive flow cytometric assays for readouts, as opposed to cell growth. Soluble complexed ICAM-1: scICAM-1, anti-CD3: aCD3. This figure was created with Biorender.

To further identify negative regulators of T cell proliferation, the authors of this study performed a series of parallel whole-genome CRISPR-Cas9 screens for genes that rescued T cell survival and proliferation in the presence of immunosuppressive agents, including adenosine, Cyclosporine (NFAT inhibitor), Tacrolimus (mTORC1 inhibitor), TGFβ, or regulatory T cells (T_reg), chosen to replicate a suppressive tumor microenvironment [2] (Figure 2A). Only two genes, RASA2 and TMEM222, increased T cell proliferation in the presence of any of these inhibitory agents [2]. In contrast, targeting the gene encoding the RAS-GEF, RASGRP1, decreased proliferation, highlighting the importance of RAS in this process. Limited or no effects were observed by targeting other RAS-GAPs, suggesting that RASA2 was a critical RAS-specific GAP, inhibition of which permitted increased proliferation following TCR stimulation under conditions of immunosuppression. Similarly, RASA2 was one of multiple genes identified in a CRISPR-Cas9 screen in human tumor lines for genes that allowed cell growth in the presence of the SHP2 inhibitor, SHP099, which blocks RAS activation, again implicating RASA2 as a major inhibitor that restrains RAS activation [40]. RASA2 was also identified as an inhibitor of RAS activation in an earlier siRNA screen for genes affecting neurite outgrowth and cell proliferation in human PC12 pheochromocytoma cells in response to Nerve Growth Factor (NGF) [41].

CRISPR-Cas9-mediated targeting of RASA2 in human T cells led to an increase in GTP-bound RAS, as well as phospho-ERK and phospho-S6, two downstream effectors of RAS [2]. Moreover, RASA2-deficient T cells showed increased antigen sensitivity, as evidenced by induction of phospho-ERK and proliferation at lower peptide and anti-CD3/CD28 concentrations than seen with control T cells. Furthermore, CD19-targeting Chimeric Antigen Receptor (CAR)-T cells lacking RASA2 killed target NALM6 human cancer cells with low surface CD19 to a greater extent than WT CAR-T cells [2]. Similar results were seen in an in vivo model injecting NALM6 cancer cells and CAR-T cells into immunodeficient mice [2]. RNAseq analyses of RASA2-deficient CAR-T cells co-cultured with target cancer cells showed increased expression of genes associated with cell cycle and metabolism compared to RASA2-sufficient cells. An inverse correlation between oxidative phosphorylation signatures and RASA2 expression was noted across multiple public gene expression databases [2].

In other experiments, the authors found that RASA2 was downregulated in human T cells upon TCR stimulation and in mouse T cells in response to acute bacterial infection. In contrast, RASA2 was upregulated in cells that were repeatedly exposed to antigen in culture and in “exhausted” T cells found in chronic infection with Listeria sp in mice, and in human tumors[2]. Targeting RASA2 rescued many of the dysfunctional phenotypes in cultured T cells after repeated exposure to antigen, particularly metabolic dysfunction, as measured by metabolic assays for oxidative phosphorylation and glycolysis (Seahorse); this led to increased killing of target cells both in vitro and in vivo relative to controls [2]. Similar findings were seen using CAR-T cells in a number of in vivo preclinical tumor models, including an intraperitoneal osteosarcoma model for solid tumors, which are particularly challenging to target in T cell immunotherapies [42]. Nonetheless, RASA2-deficient T cells did not show increased phospho-ERK prior to TCR simulation [2]. The authors concluded that RASA2 serves as an “intrinsic signaling checkpoint” that may act to limit T cell activation, particularly under conditions of T cell exhaustion and chronic antigen exposure. Thus, RASA2 might represent a new putative therapeutic target to help increase the cytotoxic and/or helper activity of T cells against tumors, as well as for cancer immunotherapy with CAR-T cells [2], although further studies remain to be conducted to test the outcomes of RASA2 targeting.

RASA3, a dual RAS/RAP1-GAP that regulates T cell adhesion

In parallel work, we recently uncovered RASA3 as an important GAP regulating RAP1 in a CRISPR-Cas9 screen of PI3K effectors that alter integrin adhesiveness in mouse T cells [4] (Figure 2B).

PI3K is a lipid kinase that phosphorylates PI(4,5)P₂ to generate the phospholipid PI(3,4,5)P₃, which recruits proteins containing PH and other PI(3,4,5)P₃-binding domains to the plasma membrane [43]. This generally leads to activation of effectors such as AKT kinases which phosphorylate downstream targets that alter transcription, translation, metabolism, and protein stability.

Previous data indicated that pharmacological or shRNA-mediated inhibition of PI3Kδ, the major Class Ia PI3K expressed in lymphoid cells, decreased TCR-induced RAP1-GTP, a measurement of RAP1 activation, and LFA-1-mediated ICAM-1-binding, as determined using a flow cytometry-based assay to detect binding of fluorescently-labeled multimerized ICAM-1 in vitro [4,43,44]; this assay detects increases in both integrin affinity and avidity [45]. However, inhibition or knockdown of AKT did not reduce ICAM-1-binding to the same extent, suggesting that other PI3K effectors were required for RAP1 activation [4,44].

To probe how PI3K affects T cell adhesion, we bioinformatically curated a list of all known proteins containing PH domains, or that have been experimentally shown to bind PIP₃; we subsequently generated a retroviral library of CRISPR-Cas9 gRNAs for a targeted screen of genes affecting ICAM-1-binding in TCR-activated mouse CD8⁺ T cells. Among several hits of interest, targeting (ablating) RASA3, increased ICAM-1-binding and RAP1-GTP [4].

RASA3 (also known as GAP1^IP4BP) is broadly expressed in mouse and human tissues, but until recently, was primarily described in the context of angiogenesis and hematopoiesis [46,47]. Global ablation of Rasa3 in mice is fatal due to bleeding [48], but point mutants that cause protein instability (RASA3^G125V and RASA3^H794L), result in thrombocytopenia and defects in erythropoiesis, platelet function, and megakaryocytosis [47,49,50]. These defects are accompanied by increased GTP-bound RAP1 and activated ɑIIβ3 integrin on platelets [49,50], implicating RASA3 as a negative regulator of integrin activation. Of note, RASA3 was found to translocate to the membrane in a PI3K-dependent manner, but was hypothesized to be inhibited by PI3K signaling; PI3K inhibition prevented platelet spreading and RAP1 activation in wildtype but not Rasa3-deficient platelets [50,51], This finding was intriguing because PI3K usually activates proteins having PH domains.

Consistent with a role for RASA3 in inhibiting RAP1, murine T cells deficient in RASA3 showed increased TCR and chemokine-induced LFA-1-mediated ICAM-1-binding, increased RAP1-GTP, altered migration on platebound ICAM-1, and decreased trans-endothelial migration into lymph nodes (LNs) compared with wildtype T cells, as determined by intravital microscopy of congenically-marked T cells transfered into wildtype recipient mice. Moreover, LN egress of RASA3-deficient T cells was also impaired (seen via microscopy of transferred T cells) [4,37]. Accordingly, mice specificlly deficient in Rasa3 in T cells (Rasa3^fl/flxCD4^Cre mice) exhibited low numbers of T cells in LNs and especially the blood, and showed impaired antibody responses to the T cell-dependent immunogen, NP-OVA in alum, relative to wildtype mice [4].

Nonetheless, T cell numbers in the mesenteric LNs (mLN) and Peyer’s patches (PP) were normal [4,37], perhaps reflecting different requirements for activation of the gut-homing integrin α4β7 [28,52], although this remains to be further investigated. Recent work also highlighted increased RAP1-dependent changes in T cell shape in response to the chemokine CCL21 in mouse T cells deficient in Rasa3 and SPA-1 (Rasa3^−/−Sipa1^−/− mice) [53]. Thus, Rasa3-deficient mouse T cells show defective T cell morphology, migration, and localization [4,37] suggesting that RASA3 is an important RAP1-GAP in T cells.

Rasa3-deficient mouse CD4⁺ T cells also showed increased proliferation in culture under conditions of stimulation with suboptimal concentrations of anti-TCR antibodies [4]. However, in contrast to Rasa2-deficient T cells, ERK phosphorylation, a downstream readout of RAS activation, was not elevated in response to TCR stimulation under the conditions examined. Thus, in T cells, RASA3 might act more specifically on RAP1 than RAS, although this remains to be formally proven (Figure 1). Indeed, although RASA3 has been described as a dual RAS/RAP1-GAP, it has high RAP1-GAP activity compared to RASA2 [50,54]. Whether the increased sensitivity to TCR stimulation in RASA3-deficient T cells is solely due to increased integrin-ligand binding or other signaling features of RASA3 (described below) remains to be further investigated. Conversely, Rasa2-deficient mouse T cells did not show increased TCR-induced ICAM-1-binding in our hands [4], consistent with earlier findings defining RASA2 as a RAS-specific GAP [25]. Thus, in T cells, RASA2 and RASA3 may have distinct roles in keeping T cells in check.

Expression of both Rasa3 and Rasa2 was markedly repressed upon TCR stimulation, as seen in mouse and human T cells stimulated in culture, and in multiple datasets of T cell gene and protein expresssion [2,4,55]. Accordingly, the increase in ICAM-1-binding seen upon TCR stimulation was particularly large in naïve Rasa3-deficient mouse T cells, suggesting that Rasa3 is a major factor that maintains resting T cells in a less adhesive state. Of note, studies using ICAM-1-deficient antigen-presenting dendritic cells suggest that naïve T cells expand independently of LFA-1-ICAM-1 interactions [56]; whether this is due to the effects of RASA3 in suppressing RAP1 activity and LFA1-mediated adhesion in naïve T cells, remains an intriguing question. Thus, the dynamic regulating expression of these GAPs may be part of what keeps naïve and other T cell populations, such as exhausted cells, in a more refractory state compared with acutely activated cells – a finding that might have therapeutic relevance for regulating T cell activation during T cell exhaustion and in scenarios of autoimmunity.

Furthermore, although most proteins with PH and other PIP₃ binding domains are activated when recruited to the plasma membrane by PIP₃, RASA3 appears to be inactivated by PI3K and by its PH domain; inhibition of PI3K prevented RAP1 activation but not in RASA3-deficient T cells [4], suggesting that PI3K acts to inhibit RASA3 function. Moreover, overexpression of a PH domain mutant of RASA3 (which cannot bind PIP₃) led to a greater decrease in ICAM-1-binding than overexpression of wildtype RASA3, suggesting that the PH domain inhibits RASA3 activity [4,51]. RASA3 was originally identified as an IP₄-binding protein [31]: however, its PH domain also has high affinity for PIP₃ [30]. PH domains also interact with proteins; for example, the PH-BTK domains of BTK and RASA2 interact with Gα12 [57]. Thus, it will be of interest to determine whether interactions with other lipids, proteins, or other mediators are important for RASA3 function in T cells and whether RASA2 and RASA3 are controlled by similar mechanisms (see outstanding questions box). Indeed, data suggest that RASA2 is activated by PI3K in PC12 cells; inhibiting PI3K increased RAS and ERK activation in response to NGF, but not in the absence of RASA2 [41]. Whether this results from differing binding preferences of the PH domains of RASA3 and RASA2 [32] remains to be evaluated. Early work also pointed to a role of cAMP in preventing TCR-induced integrin activation [6] — whether cAMP contributes to the regulation of GAPs is not clear, but is another intriguing question. Finally, both activating and inactivating mutants of PI3Kδ cause immunodeficiencies associated with immune dysregulation, including altered distribution of lymphocytes [58]. Whether altered RASA3 activation contributes to these phenotypes is unknown but may represent a fruitful area of future investigation.

Outstanding questions.

• Given the large number of GAPs expressed in T cells, how can we define physiologically relevant upstream regulators and targets for their GAP activity? Are RASA2 and RASA3 affected by other upstream pathways such as from growth factor, chemokine, or cytokine receptors? To what extent do the GAPs have functions beyond inactivating small GTPases? Understanding the redundancy or lack thereof may help identify targets for further study and potential therapeutic modulation of immune cell function.

• How are the GAP1 proteins regulated by other proteins and second-messengers such as IP₄ and PIP₃? Most proteins, such as AKT and BTK are activated when their PH domains bind PIP₃, whereas RASAS3 is inhibited by its PH domain and PI3K. Does the interaction between the RASA3 PH domain and PIP₃ remove RASA3 from a relevant substrate (e.g. RAP1), or does PIP3-binding inhibit the catalytic activity directly or interfere with other interacting proteins or lipids, such as IP₄? Are RASA2 and other GAP1 family members similarly restrained or activated by their PH domains? Understanding these questions may provide insight as to how to therapeutically manipulate these pathways.

• Expression of human and mouse RASA2 and mouse Rasa3 is rapidly repressed upon stimulation of T cells through the TCR. What is the mechanistic basis for this repression? In what contexts are these GAPs re-expressed and how is this regulated? Knowledge of the dynamic regulation of these GAPs by different signaling pathways may help shed light on the context-dependent functions of RASA2 and RASA3 in repressing T cell activation and adhesion.

• Does altered integrin activation contribute to the Th17 defect in RASA3-deficient murine CD4⁺ T cells? Are these and other phenotypes controlled by overlapping or distinct RASA3-mediated signaling complexes? And does integrin activation affect other effector T cell populations, both in lymphoid organs and other tissues? This is key to understanding how and the extent to which integrins might influence differentiation and function of other T cell subsets. Similarly, are there specific T cell subpopulations that RASA2 controls?

• Both activating and inactivating mutations affecting Pik3cd lead to immunodeficiency and immune dysregulation in humans and mice. Does RASA3 contribute to these phenotypes and could this provide insight into understanding phenotypes in these disorders?

• Can the GAP1 family members be effectively targeted by small-molecule inhibitors, which could have broad implications for the therapeutic targeting of small GTPases and their downstream pathways?

• Both RASA2 and RASA3 have emerged as important components of T cell activation from CRISPR and RNAi screens. Are such screens particularly useful for identifying inhibitors of cellular activation?

RASA3 in T_H17 cells

CD4⁺ T cells can differentiate into distinct cytokine producing populations that help different arms of the immune system but can also induce pathology. T_H17 cells produce IL17, and are important for clearing fungal and bacterial infections, but can also drive autoimmunity [59]. Naïve CD4⁺ T cells can be induced to express IL17 in culture in the presence of TGFβ and IL-6 [60,61], yet these cells are not pathogenic [62]. Inclusion of IL-23 induces IL-17-producing cells that are more inflammatory than those generated in the absence of IL-23; these are highly efficient at inducing autoimmunity, as in the case of the Experimental Autoimmune Encephalitis (EAE) mouse model [62].

In other investigations into RASA3 biology, researchers found that mouse Rasa3 was highly expressed in pathogenic T_H17 (pT_H17) cells [63]. Furthermore, Rasa3-deficient CD4⁺ T cells produced less IL-17 in the presence of IL-23 (pT_H17-inducing conditions) than wildtype CD4⁺ T cells cultured under the same conditions [63]. Rasa3^fl/flxCD4^Cre mice showed decreased incidence and severity in a model of EAE, with reduced IL17-expressing CD4⁺ T cells in the CNS compared to heterozygous controls. Instead, Rasa3-deficient T cells produced increased amounts of the less inflammatory cytokine IL-4, both in vitro and in vivo in EAE, compared with wildtype T cells. The increase in IL-4 correlated with increased expression of IRF4, a transcription factor involved in T cell expansion, metabolism, and cytokine expression. Blocking anti-IL4 antibodies or knockdown of IRF4 by shRNA increased IL-17 production and the pathogenicity of Rasa3-deficient cells in EAE, suggesting that RASA3 can restrict IL-17 expression via IL-4 and IRF4.

The authors thof this study performed an unbiased mass spectrometry analysis of proteins co-immunoprecipitating with RASA3 and found both IRF4 and the E3 ubiquitin ligase, CBL-B as RASA3 binding partners [63]. Furthermore, interactions between CBL-B and IRF4 required RASA3; in the absence of RASA3, IRF4 expression was increased, which promoted IL-4 expression (whereas intermediate amounts of IRF4 are required for IL-17 expression). In addition, transfection of CBL-B and IRF4 into heterologous (HEK293T) cells increased ubiquitination and degradation of IRF4 in the presence of co-transfected RASA3 [63]. Thus, RASA3 appears to play a key role in pT_H17 cells by restricting IRF4 and IL-4 production, both of which limit IL-17 expression, However, whether RASA3 affects T cell metabolism and proliferation, (two other readouts of IRF4 activity), is unclear.

Collectively, the above results suggest that RASA3 is involved in multiple cell-signaling pathways. Whether these complexes intersect to regulate both cytokine production/T_H17 pathogenicity and integrin adhesion and how these processes interconnect are open questions. Of note, T cells from Cbl-b-deficient mice show increased TCR-mediated RAP1 activation and ICAM-1-binding relative to wildtype; these phenotypes correlate with increased interactions between Crk-L and C3G, as determined by co-immunoprecipitation, but whether RASA3 is involved remains unknown [64]. It is also of interest that a recent CRISPR-Cas9 screen for genes that alter T_H2 cell differentiation revealed αVβ3 integrin as being essential for the expression of IL-4 and T_H2 cell generation [65]; this raised the question of whether distinct types of integrins can affect different cytokine-producing T cell subpopulations. It should be noted that in the study assessing the role of RASA3 in mouse T_H17 cells, the authors did not evaluate cell adhesion, nor did they observe lymphopenia[63] —perhaps because they used heterozygous Rasa3^fl/wtxCD4^Cre mice controls, which may have lower T cell numbers than wildtype mice. However, a recent study found that Rasa3-deficient T cells transferred into wildtype recipient mice, were retained in the lung parenchyma due to increased LFA-1 binding, further supporting the importance of RASA3 in T cell migration [37].

Memory CD8⁺ T cells also express higher amounts of RASA3 compared to effector cells [55], perhaps reflecting their dependency on cytokines, rather than on TCR stimulation [66]. Conversely, some effector populations, such as follicular T helper (T_FH) cells, exhibit low RASA3 expression [67], perhaps secondary to multiple repeated rounds of TCR stimulation during B cell selection in the germinal center. Studies using blocking anti-LFA-1 antibodies or ICAM-1-deficient B cells transferred into recipient mice indicate that T_FH cells require LFA-1-mediated interactions with B cells for their development and function [68,69]; downregulation of RASA3 may therefore be particularly important for T_FH cells. How RASA3 affects different T cell populations and whether its effects on T cell adhesion and migration contribute to distinct functions of different effector and memory T cells are intriguing open questions. Moreover, whether RASA2 is similarly regulated and how it affects distinct T cell populations remain important questions (Outstanding Questions box). Nonetheless, these genetic studies highlight the importance of these GAP1 proteins in T cell biology.

Other GAP1 Members

Although we focused on RASA2 and RASA3, other GAP1 members are also expressed in T cells. Although RASAL1 was not a prominent hit in the CRISPR-Cas9 screens in human primary T cells [2], RASAL1 has been identified as a negative regulator of TCR signaling that co-immunoprecipitates with ZAP-70 in the TCR signaling complex; indeed, Rasal1 knockdown in mouse T cells increased T cell activation and proliferation in mice following antigen injection, as well as enhanced tumor clearance using mouse B16 melanoma and EL4 lymphoma tumor models [70], Thus, RASAL1 and RASA2 may play similar roles in keeping RAS-signaling in check downstream of TCR and both may be potential immunotherapeutic targets during adoptive cell therapy, although this remains conjectural.

Similarly, another previous study found that knockout of RASA4 in mice resulted in greater RAP1-GTP in T cells following CD28 stimulation compared to wildtype T cells [71]; the authors proposed that CD28 inhibited LFA-1 activity. While this contrasts with other studies that suggest that engagement of CD28, like the TCR, activates LFA-1 in a partially PI3K-dependent manner [72,73], the net effect of CD28 on LFA-1 activity might depend in part on the balance between TCR and CD28 signals in different contexts. Nonetheless, these data suggest that multiple GAP1 family members serve as gatekeepers that maintain T cells in a less active state, thereby fine tuning T cell activation.

Concluding remarks

Understanding mechanisms by which T cell activation and function are tightly controlled is essential both for understanding the requirements for productive immune responses and for uncovering potential therapeutic targets for T cell-mediated diseases. Such targets may help reinvigorate exhausted T cells in cancer and chronic infections, while also providing insight into how to limit T cell activation in autoimmunity. It is now clear that RASA2 and RASA3 play key roles in keeping T cells in a repressed or less active state, preventing inappropriate activation (Figure 3, Key figure). Thus, these proteins provide potential therapeutic targets for manipulating T cell activity.

Key Figure, Figure 3: RASA2 and RASA3 restrict T cell activation and adhesion.

Left: RASA2-deficient human T cells show increased antigen-sensitivity, proliferation and metabolic changes associated with elevated RAS activation relative to RASA2-sufficient T cells [2]. RASA2-deficiency also permits T cell proliferation under inhibitory conditions, such as those found in a tumor environment, also increasing tumor killing in several human T cell models, both in vitro and in vivo in NSG mice [2,3]. Right: RASA3-deficient mouse T cells show increased LFA-1-mediated ICAM-1-binding, altered migration and homing, morphological changes, and impaired responses to T-dependent immunization, associated with elevated RAP1 activation, relative to RASA3-sufficient mouse T cells [4]. RASA3 deficiency in mouse T cells also prevents pathogenic T_H17 cell differentiation and development of mouse models of EAE [63]. Protein structures are predicted by alpha-fold [82]. This figure was created with Biorender.

Nonetheless, our understanding of the actions of these molecules in T cells is still early and there are many limitations to these studies that will require further study. RASA2 and possibly RASAL1 appear to be promising targets for overcoming inhibition of T cell function in the immunosuppressive environment of a tumor, yet whether these GAPs affect other T cell populations and how to take advantage of them as putative therapeutic targets are open areas (Outstanding Questions box). While RASA3 has been studied extensively in platelets [46,49,50], understanding its roles in TCR signaling and how it intersects with other signaling pathways remains to be further explored. How GAP1 proteins, the receptors they are controlled by, and the signaling pathways they influence affect distinct types of T cell function and metabolism are important future questions. Finally, whether other GAP1 family members are negatively regulated by PI3K in the manner observed for RASA3 remains to be determined (Outstanding questions box). Nonetheless, studies of these GAP1 proteins suggest they are important new players in T cell biology and function.

Highlights.

• The RASA2 and RASA3 GTPase-activating proteins are recently recognized negative regulators of T cell activation and adhesion.

• RASA2 was identified as a novel immunotherapy target in a CRISPR-Cas9 screen for molecules permitting human primary T cell proliferation in the presence of inhibitors. RASA2-depletion increased antigen sensitivity in conventional and CAR-T cells and improved tumor rejection.

• RASA3 was identified in a CRISPR-Cas9 screen as an inhibitor of TCR-induced LFA-1-mediated ICAM-1-binding in mouse T cells. Rasa3 deficiency altered T cell adhesion, migration, and T cell-dependent antibody responses in mice.

• Rasa3 is highly expressed in pathogenic T_H17 cells and is required for the development of Experimental Autoimmune Encephalitis in mice, suggesting T cell-type and tissue-specific functions.

• Both Rasa2 and Rasa3 are rapidly repressed upon TCR stimulation of human and mouse T cells, promoting T cell activation. RASA3 is also inhibited by PI3K activity.

SIGNIFICANCE.

Understanding the mechanisms that keep T cells in check can provide insight into targets for reinvigorating T cell function in cancer and chronic infection or for suppressing T cell function in the context of autoimmunity and transplantation. The RASA2 and RASA3 GAPs have recently emerged as key inhibitors of T cell activation, adhesion, and migration, thereby introducing a potential new set of therapeutic targets for manipulating T cell activity.

Glossary:

Anergy

state of dormancy for T cells where they do not proliferate or respond to activation; classically described as a result of activation without costimulation or in the presence of high coinhibitory receptor signaling.

Avidity

Overall combined adhesiveness of a set of receptors on a cell or surface to its target. For integrins, this is determined by the affinity for ligands and density of integrins and their ligands (a combination of clustering and protein expression).

Carboxyfluorescein succinimidyl ester (CFSE)

membrane-permeable dye that enters cells and is diluted upon division. CFSE is used to evaluate cell proliferation.

Chimeric antigen receptor (CAR) T cells

CAR-expressing T cells that have been genetically modified to express an antibody fragment specific to a defined target, such as CD19 expressed on CD19⁺ leukemia.

Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)-Cas9 (CRISPR-associated endonuclease 9)

genome editing technology used in bacteria as a defense system. CRISPR-Cas9 has been adapted to cut specific DNA sequences determined by short guide RNA sequences and used in mammalian cells to target and mutate genes. The Cas9 endonuclease originating from Streptococcus pyogenes is the most commonly used CRISPR-Cas enzyme used for the binding and cleavage of DNA guided by a gRNA.

Extracellular signal-regulated kinase (ERK)

Mitogen activated protein kinase (MAPK) activated by a series of kinases; dually phosphorylated on threonine and tyrosine residues for its activation. ERK activates a number of downstream pathways including those regulating transcription, translation, and metabolism.

Experimental autoimmune encephalitis (EAE)

mouse model for multiple sclerosis, induced by various types of immunization with myelin antigens and leading to an inflammatory demyelinating central nervous system disorder.

E3 Ubiquitin ligase

enzyme that catalyzes the addition of ubiquitin short protein chains to specific proteins. Ubiquitination can target a protein for degradation or promote protein-protein interactions.

LAT-SLP-76 signalosome

TCR activation leads to downstream phosphorylation of LAT and SLP-76 resulting in nucleation of a large signaling complex comprised of SLP76, GADS, ITK, PLCγ1, and VAV1. This signalosome is critical for T-cell immune responses as it is central in the TCR signaling pathway.

Leukocyte Adhesion Deficiencies

primary immunodeficiencies that cause a loss of functioning leukocyte adhesion resulting in poor chemotaxis and migration. These deficiencies manifest as increased susceptibility to infections.

Mass Spectrometry

analytical method which measures the mass:charge ratio of molecules that can be used to identify unknown compounds in a sample; can be used to identify peptides from proteins interacting and co-immunoprecipitating with a known protein.

Pleckstrin homology (PH) domains

Conserved and found in many signaling proteins; mediate interactions with inositol phospholipids and other proteins to regulate protein localization and interactions.

Rasopathy

One of several developmental and cancer predisposition syndromes involving mutations that affect Ras signaling, including Noonan’s syndrome (genetic syndrome characterized by short stature, specific facial features, heart defects, skeletal abnormalities, and bleeding), and Neurofibrosis Type 1 (NF1, a syndrome that include multiple skin coloration changes and tumors growing along nerves).

T cell Exhaustion

State of T cell dysfunction, seen in cancer and chronic infections; characterized by poor effector function, altered metabolism, and epigenetic changes; caused by prolonged antigen exposure.

Transendothelial migration (TEM)

process whereby leukocytes cross endothelial barriers to migrate from blood to tissues. There are 3 steps to TEM, rolling on selectins on the high endothelial vessels, activation of integrins by chemokines and resulting adhesion to the endothelial cells, and transmigration through the endothelial barrier.

Tyrosine kinase

enzyme that adds a phosphate to a specific tyrosine on itself and/or other proteins. Tyrosine kinases initiate many signal transduction cascades including TCR signaling.