The Abelson tyrosine kinases were initially identified as drivers of leukemia in mice and humans. The Abl family kinases Abl1 and Abl2 regulate diverse cellular processes during development and normal homeostasis, and their functions are subverted during inflammation, cancer and other pathologies. Abl kinases can be activated by multiple stimuli leading to cytoskeletal reorganization required for cell morphogenesis, motility, adhesion and polarity. Depending on the cellular context, Abl kinases regulate cell survival and proliferation. Emerging data support important roles for Abl kinases in pathologies linked to inflammation. Among these are neurodegenerative diseases and inflammatory pathologies. Unexpectedly, Abl kinases have also been identified as important players in mammalian host cells during microbial pathogenesis. Thus, the use of Abl kinase inhibitors might prove to be effective in the treatment of pathologies beyond leukemia and solid tumors. In this Cell Science at a Glance article and in the accompanying poster, we highlight the emerging roles of Abl kinases in the regulation of cellular processes in normal cells and diverse pathologies ranging from cancer to microbial pathogenesis.
Abl kinase was first identified from studies of the Abelson murine lymphosarcoma virus (A-MuLV) that induced transformation of murine fibroblasts and lymphoid cells in vitro and lymphomas in vivo (Abelson and Rabstein, 1970; Goff et al., 1980). Subsequent studies demonstrated that chromosomal translocation of human ABL1 to the breakpoint cluster region (BCR) gene sequences results in production of the BCR–ABL1 fusion protein in patients with Philadelphia (Ph) chromosome-positive human leukemia (Wong and Witte, 2004). The BCR–ABL1 chimeric protein exhibits elevated tyrosine kinase and transforming activities, and has been identified in distinct human leukemias, including chronic myelogenous leukemia (CML) and acute lymphocytic leukemia (ALL). Sequencing studies identified ABL2 (also known as Abl-related gene or Arg) as a paralog of ABL1 (Kruh et al., 1990). Oncogenic forms of both ABL1 and ABL2 have been identified in some forms of T-cell acute lymphoblastic leukemia (T-ALL) and acute myeloid leukemia (AML) as a result of chromosomal translocation of the ETV6 gene to either ABL1 or ABL2 (reviewed in De Braekeleer et al., 2012).
The development of small-molecule ATP-competitive inhibitors that target BCR–ABL1 for the treatment of CML was a major breakthrough that opened the door to the era of targeted therapies (Eide and O'Hare, 2015). Treatment with imatinib or other U.S. Food and Drug Administration (FDA)-approved ATP-competitive inhibitors has been shown to induce durable remissions and progression-free survival in the majority of chronic-phase Ph-positive CML patients, but these inhibitors are not as effective for the treatment of blast crisis CML and Ph-positive ALL patients.
Whereas persistent activation of the Abl kinases is linked to the emergence of human and murine leukemias, the activities of the endogenous Abl kinases are highly regulated by diverse stimuli that range from growth factors, chemokines, DNA damage, oxidative stress and adhesion receptors to microbial pathogens (reviewed in Colicelli, 2010; Greuber et al., 2013). Once activated, the Abl kinases regulate signaling pathways implicated in cytoskeletal reorganization that are important for cellular protrusions, cell migration, morphogenesis, adhesion, endocytosis and phagocytosis (reviewed in Bradley and Koleske, 2009; Colicelli, 2010). Abl kinases can also regulate cell survival and proliferation pathways depending on the cellular context (Greuber et al., 2013). Emerging data support a role for abnormally activated Abl kinases in diverse pathologies, including several solid tumors, inflammatory disorders and neurodegenerative diseases. Moreover, accumulating reports have revealed that Abl family kinase function is subverted by numerous microbial pathogens to achieve entry, motility, release and/or survival in mammalian host cells (reviewed in Wessler and Backert, 2011). Thus, targeting the Abl kinases with small-molecule inhibitors, which were initially developed to treat CML patients, might be employed to treat distinct pathologies with hyper-active Abl kinases.
In this Cell Science at a Glance article and accompanying poster, we highlight the emerging roles of Abl kinases in the regulation of cellular processes in normal cells and diverse pathologies. Because of space limitations, we focus on the cytoplasmic-initiated functions of the mammalian Abl kinases and will not cover the nuclear functions of Abl1 that are induced in response to DNA damage (reviewed in Wang, 2014). We highlight cellular networks targeted by the Abl kinases in response to diverse stimuli, revealed primarily by studies that rely on genetic inactivation and/or depletion of the mammalian ABL1 and ABL2 kinases rather than use of non-selective kinase inhibitors.
Structure, modular domains and enzymatic regulation of the Abl kinases
Abl kinases have been identified in all metazoan and premetazoan unicellular organisms, such as the choanoflagellate Monosiga brevicollis (Colicelli, 2010). The structural domains of mammalian Abl kinases are shown in a panel within the poster. Vertebrate metazoans express ABL1 (c-Abl) and ABL2 proteins that share a highly conserved SH3–SH2–SH1 cassette, containing a tyrosine kinase Src homology 1 (SH1) domain and regulatory SH3 and SH2 domains. This cassette is preceeded by an N-teminal cap region. Both kinases also have F-actin-binding domains within their large C-terminal sequences. The DNA-binding domain, nuclear localization signals and nuclear export signal in the ABL1 C-terminus are consistent with its nucleo-cytoplasmic localization, whereas the microtubule- and actin-binding domains in the ABL2 C-terminus are consistent with its localization at the cell periphery (Miller et al., 2004). Conserved proline-rich sequences in the C-terminal domain of the Abl kinases mediate protein–protein interactions (Colicelli, 2010). Alternative splicing of the first exons produces various isoforms of ABL1 and ABL2 that harbor distinct N-terminal sequences. The 1b isoforms of both Abl kinases contain an N-terminal glycine that is myristoylated. Additional ABL2 isoforms have been reported but little is known regarding their function (Bianchi et al., 2008).
The activity of the Abl kinases is tightly regulated by intra-molecular and inter-molecular interactions that result in alternative conformations of the kinase domain corresponding to auto-inhibited and active states (Nagar et al., 2006,, 2003). These include the binding of the SH3 domain to the polyproline-containing sequence that connects the SH2 and SH1 domains, as well as interactions of the SH2 domain with the SH1 domain, leading to the formation of a SH3–SH–SH1 clamp structure (Hantschel and Superti-Furga, 2004). The auto-inhibited conformation is further stabilized by an N-terminal cap and binding of the N-terminal myristoylated residue to a hydrophobic pocket within the kinase domain (see poster). Interestingly, the M. brevicollis ABL2 lacks the N-terminal myristoylation and cap sequences, and was reported to be constitutively active when expressed in mammalian cells (Aleem et al., 2015).
The ATP-competitive inhibitors of the Abl kinases can be sub-classified into type-I inhibitors that target the active conformation of the kinase domain (dasatinib, bosutinib) and type-II inhibitors targeting the inactive conformation of the kinase domain (imatinib, nilotinib, ponatinib). A third class includes allosteric inhibitors that do not target the ATP-binding pocket but, instead, bind to regulatory domains to inhibit kinase activity. Among the allosteric inhibitors are GNF2 and GNF5, which bind to the myristoyl-binding pocket in the C-lobe of the Abl kinase domain (Zhang et al., 2010). In contrast to ATP-competitive inhibitors that target multiple kinases, the allosteric inhibitors are highly selective for the Abl kinases. Crystal structures of Abl kinases in complexes with several inhibitors have permitted visualization of their inactive and active conformations. The crystal structure of the ABL1 SH2–SH1 domains bound to dasatinib revealed an extended conformation that included a contact interface between the SH2 domain and the N-lobe of the kinase domain (Lorenz et al., 2015) (see poster). Interestingly, a crystal structure of the isolated ABL2 kinase domain with imatinib revealed the presence of two imatinib molecules, with one bound to the ATP-binding site and the other to the myristate binding site (Salah et al., 2011). These and other studies have shown that binding of inhibitors to the SH1 domain can capture multiple conformational states of the Abl kinases.
The enzymatic activity of Abl kinases can also be regulated by inter-molecular interactions with distinct binding partners that negatively or positively regulate their activity, and by autophosphorylation or phosphorylation by other kinases, such as Src and receptor tyrosine kinases (RTKs) (Colicelli, 2010). Phosphorylation of amino acid (aa) residue Y412 in ABL1 (which corresponds to aa residue Y439 in ABL2) in the activation loop of the kinase domain enhances catalytic activity, whereas phosphorylation of aa residue Y245 in the SH2-kinase domain linker (which corresponds to (aa) residue Y272 in ABL2) further enhances catalytic activity by disrupting inhibitory intra-molecular interactions. Upon activation, Abl kinases phosphorylate and bind to multiple targets to regulate diverse processes, which we describe below.
Abl kinases regulate cytoskeletal dynamics required for cellular protrusions, cell adhesion, polarity and migration
The most evolutionarily conserved and well-characterized function of the Abl kinases is their ability to regulate cytoskeletal dynamics (Bradley and Koleske, 2009; Greuber et al., 2013). Abl kinases are transiently activated by growth factors that stimulate RTKs and lead to actin cytoskeleton reorganization that is required for the formation of lamellipodial protrusions, membrane dorsal ruffles, cell migration, invasion, cell scattering and/or tubulogenesis (Boyle et al., 2007; Li et al., 2015; Plattner et al., 2003,, 1999). Abl-dependent regulation of cytoskeletal dynamics is mediated by the phosphorylation of target proteins, such as cortactin that is required for platelet-derived growth factor (PDGF)-induced dorsal membrane ruffles (Boyle et al., 2007), as well as actin nucleating-promoting factors, such as N-WASP and WAVE proteins required for membrane protrusions (Burton et al., 2005; Miller et al., 2004; Sossey-Alaoui et al., 2007; Stuart et al., 2006). Abl kinases also transduce signals from cell surface receptors, such as integrins, cadherins, chemokine receptors and the T cell receptor (TCR), to reorganize the cytoskeleton. Integrin engagement induced by plating fibroblasts on fibronectin promotes activation of the Abl kinases required for the formation of membrane protrusions and remodeling of focal adhesions and F-actin stress fibers (Bradley and Koleske, 2009). ABL2 is activated by direct binding to the β1-integrin cytoplasmic tail, and the phosphorylated ABL2, in turn, phosphorylates the β1 tail (Simpson et al., 2015). Integrin-dependent spindle orientation was shown to require ABL1-mediated phosphorylation of the NuMA microtubule-binding protein, a modification that is required for NuMA localization to the cortex and correct alignment of the spindles (Matsumura et al., 2012). ABL1-knockout mice exhibit spindle misorientation in basal skin cells, a phenotype that has been previously linked to β1-integrin-mediated cell adhesion to the extracellular matrix (Lechler and Fuchs, 2005).
Remodeling of the cytoskeleton by Abl kinases downstream of diverse stimuli is mediated, in part, by the small GTPases, Rac1, RhoA and Rap1 (see poster). Abl kinases activate Rac1 downstream of growth-factor-stimulated RTKs leading to formation of membrane ruffles (Sini et al., 2004). Abl kinases also activate Rac1 in T cells in response to chemokine stimulation which is required for T cell polarization and F-actin polymerization (Gu et al., 2012). Abl kinases activated downstream of chemokine receptors promote Rap1 activation leading to T cell migration and polarization (Gu et al., 2012). In contrast, activation of ABL2 in epithelial cells grown on collagen inhibits β1-integrin signaling through suppression of Rap1 activity, while also inhibiting Rac1 and disrupting the Par3–Par6 polarity complex (Li and Pendergast, 2011). These examples illustrate that regulation of Rac1 and Rap1 by Abl kinases is cell-context- and stimulus-dependent. In response to integrin engagement, ABL2 inhibits RhoA in fibroblasts and dendritic spines by phosphorylating p190RhoGAP (also known as ARHGAP35) that, in turn, inhibits RhoA (Bradley and Koleske, 2009; Warren et al., 2012). However, activation of Abl kinase downstream of the ligand-stimulated RTK MET promotes activation of RhoA (Li et al., 2015). These findings indicate that regulation of RhoA activity downstream of Abl kinases is also cell context dependent.
Abl kinases are also activated in response to E-cadherin and N-cadherin (Bays et al., 2014; Zandy et al., 2007). Active Abl kinases are required for both the maintenance and remodeling of cadherin-dependent adherens junctions (see poster). Simultaneous inactivation of both ABL1 and ABL2 disrupts adherens junctions in fibroblasts and epithelial cells (Li and Pendergast, 2011; Zandy et al., 2007). Recently, Abl kinases were shown to phosphorylate the actin-binding protein vinculin at epithelial cell–cell junctions, and this modification was required to promote E-cadherin-mediated adhesion and force transmission (Bays et al., 2014). Growth-factor-induced turnover of cadherin-based epithelial and endothelial cell–cell junctions also requires Abl kinases. Inactivation of both Abl kinases impairs hepatocyte growth factor (HGF)-mediated dissolution of E-cadherin-mediated cell–cell junctions (Li et al., 2015). Similarly, inhibition of Abl kinases prevents dissolution of VE-cadherin-dependent intercellular junctions in endothelial cells that have been treated with vascular endothelial growth factor (VEGF), thrombin or histamine (Chislock and Pendergast, 2013). Abl kinases link signaling from the HGF receptor MET to activation of RhoA, which leads to increased phosphorylation of myosin light chain 2 (MLC2-P) and actomyosin contractility, thereby disrupting epithelial cell–cell junctions (Li et al., 2015). However, inhibition of the Abl kinases in endothelial cells treated with VEGF or thrombin suppresses MLC2-P levels through a Rho-independent pathway (Chislock et al., 2013). Importantly, treatment of wild-type mice with Abl-specific allosteric inhibitors, or genetic inactivation of Abl1 in Abl2 heterozygous mice, blocks VEGF-induced vascular permeability (Chislock et al., 2013). The phenotypes of mice with global or tissue-specific knockouts of the Abl kinases support a role for these kinases in the regulation of cytoskeletal dynamics required for cell–cell adhesion and other cellular processes (see Box 1).
Analysis of mice that lack one or both of the murine Abl kinases has shown that ABL1 and ABL2 exhibit overlapping as well as unique functions. Abl1-knockout mice are either viable or exhibit perinatal lethality depending on the strain, and display phenotypes distinct from those detected in the viable Abl2 (Arg) global knockout mice (Gourley et al., 2009; Li et al., 2000; Moresco et al., 2005; Schwartzberg et al., 1991; Tybulewicz et al., 1991). Abl1-knockout mice exhibit splenic and thymic atrophy, reduced numbers of B- and T-cells, cardiac abnormalities, and osteoporosis that has been linked to defective osteoblast proliferation and premature senescence (Kua et al., 2012; Li et al., 2000; Moresco et al., 2005; Schwartzberg et al., 1991; Tybulewicz et al., 1991). In contrast, Abl2-knockout mice are viable and exhibit neuronal defects that include age-related dendrite destabilization and regression (Gourley et al., 2009; Koleske et al., 1998; Moresco et al., 2005). Functional overlap between ABL1 and ABL2 is supported by the embryonic lethality of Abl1−/− Abl2−/− mice by embryonic day 11 (Koleske et al., 1998). Analysis of conditional knockout mice with tissue-specific deletion of the Abl kinases has shown unique and overlapping roles for these kinases in immune and endothelial cells. For example, Abl kinases have redundant roles in mature T cells because deletion of both ABL1 and ABL2 was required to inhibit TCR-induced proliferation and cytokine production, as well as chemokine-induced T cell migration (Gu et al., 2007). In contrast, only ABL1 is required for thymocyte differentiation (Trampont et al., 2015). Mice with endothelial deletion of Abl1 in Abl2-null mice exhibit late-stage embryonic and perinatal lethality that is partly due to increased apoptosis (Chislock et al., 2013). The unique phenotypes that are induced by loss of ABL1 and ABL2 might be owing to their differential expression in various tissues, and/or might reflect distinct cellular functions of ABL1 and ABL2 that are mediated by their unique domains, differential subcellular localization and/or association with distinct protein complexes.
Abl-dependent regulation of membrane and organelle trafficking
Plasma membrane plasticity is regulated by forces that are generated by the actin and microtubule cytoskeletons. Abl kinases contain both actin- and microtubule-binding domains, and phosphorylate several proteins that regulate the actin and microtubule cytoskeletons (reviewed in Colicelli, 2010). Thus, Abl kinases have the potential to directly and/or indirectly regulate the actin and microtubule cytoskeletons, thereby promoting plasma membrane remodeling and receptor endocytosis. Activated ABL1 was shown to inhibit ligand-induced endocytosis of epidermal growth factor receptor (EGFR), in part, through phosphorylation of the EGFR, which results in decreased binding to the Cbl ubiquitin ligase, a protein that promotes ubiquitylation and lysosomal degradation of the receptor (Tanos and Pendergast, 2006). Additionally, active ABL1 disrupts the interaction of Cbl with Abi1, a binding partner and substrate of the Abl kinases, which is also a component of the WAVE protein complex that promotes Arp2/3-dependent actin polymerization (Tanos and Pendergast, 2007). EGFR endocytosis is also regulated by RAB5 (also known as RABGEF1), a GTPase that promotes early endosome fusion and subsequent degradation of EGFR in the lysosome (Chen et al., 2009). The activity of RAB5 is regulated by the Ras and Rab interactor 1 (RIN1) that binds and activates the Abl kinases, and is also a guanine nucleotide exchange factor (GEF) for RAB5 (Balaji et al., 2012; Hu et al., 2005). Recent data suggest that binding of the RIN1 SH2 domain to EGFR phosphorylated at tyrosine residues balances RIN1–RAB5-mediated EGFR endocytosis and lysosomal degradation with RIN1–ABL-dependent EGFR stabilization by blocking macropinocytosis of the receptor (Balaji et al., 2012) (see poster). Although overexpression of active ABL1 inhibits EGFR internalization, the activity of endogenous ABL1 has been reported to be required for antigen-induced endocytosis of the B-cell receptor (Jacob et al., 2009). ABL1 inactivation impaired capping of B-cell receptors and delayed their internalization, which correlated with decreased Rac activation (Jacob et al., 2009). A role for Abl kinases in clustering of receptors at the plasma membrane has also been shown at the postsynaptic membrane of the neuromuscular junction (Finn et al., 2003). Here, Abl kinase activity is required for agrin-induced clustering of acetylcholine receptors (AChRs) on the postsynaptic membrane of the neuromuscular junction, which may be mediated through Abl-dependent regulation of cytoskeletal dynamics.
Consistent with a role for Abl kinases in membrane trafficking events, Abl kinases were recently shown to control the inward trafficking of caveolin-1, in part, through regulation of actin stress fibers by diaphanous-1 (DIAPH1), a member of the formin family of actin regulatory proteins (Echarri et al., 2012) (see poster). Further, Abl kinases have been shown to regulate phagocytosis of antibody-coated pathogens in macrophages (Greuber and Pendergast, 2012; Wetzel et al., 2012) (see poster). Bone marrow-derived macrophages from mice that lack both Abl1 and Abl2 display impaired phagocytosis, and treatment of macrophages with pharmacological inhibitors of the Abl kinases impairs FcγR-mediated phagocytosis in part by decreased Abl-mediated phosphorylation of the tyrosine-protein kinase Syk (Greuber and Pendergast, 2012). Inhibition of the Abl kinases also decreases phagocytosis of opsonized polystyrene beads and reduces the uptake of Leishmania amazonensis by macrophages (Wetzel et al., 2012). Several pathogens rely on the activity of the endogenous Abl kinases to enter into host cells (see Box 2). Although loss of ABL1 reduces complement-mediated phagocytosis, ABL2 activity is required for immunoglobulin-mediated phagocytosis (Wetzel et al., 2012). Further, Abl kinases are required for human immunodeficiency virus type 1 (HIV-1) infection by regulating virus-induced lipid mixing or hemifusion of the plasma membrane, which is required for pore formation and virus entry (Harmon et al., 2010).
Accumulating reports have shown that the activity of the Abl kinases is subverted by an ever increasing number of bacterial and viral pathogens (see poster, Table) (Wessler and Backert, 2011). The first report of a role for Abl kinases in microbial pathogenesis was the demonstration that Abl kinase activity is required for internalization of Shigella flexneri, a bacterium that enters the non-phagocytic cells of the colonic mucosa (Burton et al., 2003). Abl kinases are catalytically activated during Shigella internalization, and inactivation of the Abl kinases impairs Shigella uptake and phosphorylation of the Crk adaptor protein required for activation of the Rac1 and Cdc42 GTPases (Burton et al., 2003). Phosphorylation of N-WASP by Abl kinases is required for Shigella-induced formation of actin comet tails, bacteria intracellular movement and cell-to-cell spread (Burton et al., 2005) (see poster). Subsequent reports showed that Abl kinases are also required for infection by several bacteria (e.g. Helicobacter pylori, EPEC, Salmonella enterica, Anaplasma phagocytophilum, Pseudomonas aeruginosa, Chlamydia trachomatis, Mycobacterium tuberculosis) (Bauer et al., 2009; Bruns et al., 2012; Elwell et al., 2008; Jayaswal et al., 2010; Lin et al., 2007; Ly and Casanova, 2009; Muller, 2012; Napier et al., 2011; Pielage et al., 2008; Poppe et al., 2007; Swimm et al., 2004; Tammer et al., 2007) and viruses [vaccinia virus, coxsackievirus, enterovirus 71 (EV71), HIV, hepatitis C virus (HCV) and Ebola virus] (Chen et al., 2007; Coyne and Bergelson, 2006; Garcia et al., 2012; Harmon et al., 2010; Newsome et al., 2006; Reeves et al., 2005; Yamauchi et al., 2015). Microbial pathogens exploit the ability of Abl kinases to regulate cytoskeletal processes in order to achieve efficient internalization, intracellular motility, pedestal formation, cell-to-cell spread, membrane modeling and release (egress). Microbial pathogens subvert the function of mammalian Abl kinases by two means: (1) through the deregulation of host cell factors targeted by the Abl kinases, such as actin regulatory proteins (N-WASP, cortactin, Crk, Vav) and GTPases (Rac, Cdc42) and (2) through Abl-mediated phosphorylation of microbial factors required for pathogen entry, motility, and/or release, such as H. pylori CagA (Muller, 2012; Poppe et al., 2007), EPEC Tir (Swimm et al., 2004), vaccinia virus A36R (Newsome et al., 2006; Reeves et al., 2005), Ebola virus VP40 (Garcia et al., 2012) and HCV NS5A (Yamauchi et al., 2015). Mammalian Abl kinases have also been shown to regulate assembly and release of viral particles in the case of HCV (Yamauchi et al., 2015), as well as EV71-induced apoptosis of neuronal cells by targeting cyclin-dependent kinase 5 (CDK5) of the host cell (Chen et al., 2007). These unexpected findings suggest that pharmacological inhibitors of the Abl kinases that have been developed for the treatment of leukemia can be used for treating infections that are induced by diverse microbial pathogens.
Abl kinases have been shown to regulate organelle trafficking by promoting fusion of autophagosomes to lysosomes, and regulating lysosome motility and localization in fibroblasts and alveolar carcinoma cells (Yogalingam and Pendergast, 2008). Abl kinases also promote endosome maturation and trafficking of proteins to the lysosome for degradation (Fiore et al., 2014; Yogalingam and Pendergast, 2008). Trafficking of early endosomes and lysosomes along microtubules facilitates organelle motility and fusion events. It is, currently, unknown whether the phenotypic consequences of ABL1 and ABL2 inactivation on endosome and lysosome trafficking and function is mediated through Abl-dependent regulation of the microtubule and actin cytoskeletons. Regardless, it is clear that Abl kinases regulate membrane trafficking processes that are required for receptor clustering, internalization, particle uptake and organelle fusion.
Abl-dependent regulation of cell proliferation and survival pathways in normal cells
Abl kinase-mediated signal transduction from diverse cell surface receptors also regulates cell proliferation and survival pathways during development and cellular homeostasis (Sirvent et al., 2008) (see poster). Cytoplasmic Abl kinases promote mitogenic signaling downstream of the PDGF receptor (PDGFR). Mouse embryo fibroblasts deficient for Abl1 exhibit a delay in DNA synthesis following stimulation with PDGF (Furstoss et al., 2002; Plattner et al., 1999). Activation of Abl kinases downstream of the PDGFR is mediated, in part, by Src kinases, leading to induction of Myc and increased DNA synthesis (Furstoss et al., 2002). PDGFR–Src–Abl signaling also increases mitogenesis through Rac activation, which, in turn, activates the Jun N-terminal kinase (JNK) and NADPH oxidase (Nox) pathways in fibroblasts (Boureux et al., 2005). Analysis of osteoblasts derived from Abl1-null mice showed that Abl1 is required for osteoblast proliferation, and its loss results in premature senescence (Kua et al., 2012). Abl1 also mediates cell proliferation downstream of the EphB2 receptor in the intestinal epithelium, as shown by decreased proliferation in the small intestine and colon of Abl1-mutant mice (Genander et al., 2009). Abl1 is required for EphB signaling to cyclin D1, which promotes cell proliferation in the intestinal epithelium (see poster).
Abl kinases regulate cell proliferation downstream of the antigen receptors in T and B cells. Abl1-deficient mice have decreased pro-B, pre-B and peritoneal B-1 cells (Liberatore and Goff, 2009; Schwartzberg et al., 1991; Tybulewicz et al., 1991). Abl1-deficient B cells exhibit decreased proliferation in response to antibody against immunoglobulin M (IgM) and have defective B-cell-receptor-induced activation and signaling (Liberatore and Goff, 2009; Zipfel et al., 2000). Abl kinases are activated downstream of the T cell receptor (TCR), and are required for tyrosine phosphorylation of downstream kinases (ZAP70) and adaptor proteins (LAT and Shc) that are required for maximal T cell proliferation and IL-2 production (Gu et al., 2007; Zipfel et al., 2004). Interestingly, murine Abl1 has recently been shown to function downstream of ShcA in immature thymocytes to regulate their proliferation, differentiation and migration (Trampont et al., 2015).
Recently, Abl kinases have been shown to regulate angiopoietin 1 (Angpt1)-mediated endothelial cell survival by activation of the angiopoietin-1 receptor Tie2 (also known as TEK) that is activated by binding Angpt1 (Chislock et al., 2013). Inactivation of both ABL1 and ABL2 in endothelial cells markedly decreased Angpt1 signaling, mediated in part by inhibition of the pro-survival Akt pathway (see poster). Inactivation of the Abl kinases also decreased gene expression of Tie2 (Chislock et al., 2013). Interestingly, expression of exogenous Tie2 only partially rescued Angpt1-mediated survival, suggesting that the Abl kinases modulate Angpt1–Tie2 signaling through additional mechanisms that go further that regulating Tie2 levels. ABL1 may also promote vascular development through neuropilin-1, a receptor for VEGFA (Raimondi, 2014; Raimondi et al., 2014). Endothelial deletion of Abl1 in Abl2-null mice resulted in focal loss of vasculature due to apoptosis, leading to localized tissue necrosis, with late-stage embryonic and perinatal lethality (Chislock et al., 2013). The phenotypes of endothelial Abl knockout mice consistently point to a role for these kinases in the regulation of cell survival pathways (Box 1).
Role for Abl kinases in solid tumors
Whereas a role for Abl kinases in the initiation and progression of human and mouse leukemia is understood well, a role for ABL1 and ABL2 in solid tumors is only beginning to be appreciated (reviewed in Greuber et al., 2013). Abl activation in solid tumors is not linked to chromosome translocation events, such as those found in human leukemias, but, rather, is driven by enhanced activation of Abl in response to stimulation through oncogenic tyrosine kinases, chemokine receptors, oxidative stress and metabolic stress. A number of reports have suggested a role for Abl kinases in the regulation of tumor cell migration, growth and/or survival following the use of ATP-competitive inhibitors, such as imatinib, nilotinib and dasatinib. However, these and other inhibitors target a broad spectrum of kinases (Greuber et al., 2013) and, therefore, we here only highlight studies that employ mouse models with a genetic inactivation (knockout or knockdown) of the Abl kinases.
Activation of Abl kinases in some solid tumors has been linked to alterations in the growth of primary tumors. Inactivation of Abl1 in the Apcmin/+ mouse model of intestinal adenoma impaired EphB2-mediated promotion of tumors and extended the life span of Apcmin/+ mice (Kundu et al., 2015). Notably, the ABL1 kinase was found to be hyperactive in human hereditary leiomyomatosis and in renal cell carcinoma (HLRCC) that is characterized by deficiency in the enzyme fumarate hydratase leading to the accumulation of high levels of fumarate (Sourbier et al., 2014). High fumarate levels activate ABL1, which, in turn, promotes aerobic glycolysis through the activation of the mTOR-HIF1α pathway and, so, compensates for high levels of oxidative stress through enhanced nuclear localization of the antioxidant response transcription factor NRF2 (also known as NFE2L2; see poster).
Abl kinases also regulate motility, invasion and metastasis of solid tumors; knockdown of ABL2 alone decreases cancer cell invasion and intravasation following implantation of MDA-MB-231 breast cancer cells in the mammary fat pad (Gil-Henn et al., 2013). Moreover, Abl kinases have been shown to promote invasion and metastasis of melanoma cells (Fiore et al., 2014). ABL2 localizes to invadopodia and was shown to regulate the maturation of invadopodia by linking activation of the EGFR and Src kinases to tyrosine phosphorylation of cortactin (Mader et al., 2011). Furthermore, Abl kinases promote the expression and/or localization of other proteins that regulate cell invasion, such as the type 1-matrix metalloproteinase MT1-MMP (also known as MMP14) (Chevalier et al., 2015; Mader et al., 2011; Smith-Pearson et al., 2010). A recent report demonstrated a so-far-unknown role for Abl kinases in promoting the invasion and metastasis of colorectal cancer cells, by linking the activation of NOTCH to the phosphorylation of TRIO, which enhances its Rho-GEF activity, leading to a corresponding increase of Rho–GTP levels (Sonoshita et al., 2015) (see poster).
An increase in the activation of endogenous Abl kinases has been reported in some cancers that are resistant to chemotherapy. In breast cancer cells, doxorubicin induces an atypical activation of NF-κB through ABL1 kinase activity but, treatment with the ABL inhibitor imatinib in combination with this chemotherapy drug reversed intrinsic and acquired resistance, and significantly enhanced apoptosis of breast cancer cells (Esparza-Lopez et al., 2013; Sims et al., 2013). In estrogen receptor (ER)-positive breast cancer cells, ABL is a functional partner of ERs, and inhibition of ABL resulted in sensitization to anti-estrogen therapies with tamoxifen, fulvestrant and aromatase inhibitors (Esparza-Lopez et al., 2013; Weigel et al., 2013; Zhao et al., 2011,, 2010). In addition, inhibition of ABL1 has been reported to sensitize breast cancer cells to the EGFR and ERBB receptor tyrosine kinase inhibitor lapatinib (Lo et al., 2011; Stuhlmiller et al., 2015; Wang et al., 2015), and prostate and renal cancer cells to inhibitors of heat shock protein 90 (Hsp90) (Dunn et al., 2015) (see poster).
Role for Abl kinases in neurodegenerative and inflammatory diseases
Recent data have implicated abnormally activated Abl kinases in multiple neurodegenerative diseases (reviewed in Schlatterer et al., 2011) (Jing et al., 2009; Ko et al., 2010). Pharmacological inhibition of Abl kinases in both Alzheimer's disease (AD) and Parkinson's (PD) disease models facilitates amyloid clearance and reduces neuro-inflammation, two key drivers of neuronal cell death. Common to the pathogenesis of both AD and PD is the E3 ubiquitin ligase parkin, which has a neuroprotective role through the ubiquitylation and clearance of misfolded proteins, as well as autophagic clearance of β-amyloid peptides and phosphorylated Tau protein (Imam et al., 2011; Lonskaya et al., 2013a,b). In response to oxidative stress in neurodegenerative disease, ABL1 translocates to mitochondria where it phosphorylates parkin, resulting in loss of its E3-ligase activity (Imam et al., 2011). Treatment with imatinib decreases parkin phosphorylation, thereby, increasing its neuroprotective function (see poster; reviewed in Gaki and Papavassiliou, 2014). Abl kinases also have a role in neuro-inflammation. Treating Alzheimer's disease mouse models with Abl kinase inhibitors not only cleared β-amyloid peptides and dissolved plaque, but also reduced astrocyte and dendritic cell numbers, modulated the cytokine and chemokine profiles (Lonskaya et al., 2015), and improved cognitive performance (Cancino et al., 2008; Lonskaya et al., 2013a,b).
Abl kinases have multiple targets that may contribute to the pathology of Parkinson's disease, such as synuclein – a protein found in the fibrillar aggregates known as Lewy bodies in Parkinson's disease and Lewy body dementia (Goedert, 2001; Mahul-Mellier et al., 2014). Abl kinases also phosphorylate Cdk5 on tyrosine residue 15, which contributes to cell death through the activation of p53 (Lee et al., 2008; Li, 2005) and reduces dopaminergic signaling through phosphorylation of DARPP-32 (also known as PPP1R1B), an important target for dopamine and protein kinase A in the striatum (Tanabe et al., 2014). Treatment of mice with nilotinib in mouse models of Parkinson's disease decreased the loss of dopaminergic neurons, and reduced behavioral deficits and motor symptoms (Karuppagounder et al., 2014; Tanabe et al., 2014).
Concluding remarks and perspectives
The Abl kinases have emerged as crucial integrators of multiple signaling cues during normal development and homeostasis but accumulating data revealed essential roles for hyperactive Abl kinases in diverse pathologies, including cancer, neurodegenerative diseases and inflammatory conditions. Remarkably, a growing number of microbial pathogens subvert the activity of the endogenous Abl kinases in order to infect host cells. The impact of these findings is that inactivation of the Abl kinases with novel specific inhibitors might be employed to treat diverse pathologies that are characterized by hyper-activation of Abl-mediated pathways. Among unresolved questions are whether ABL1 and ABL2 each have unique roles in the initiation and progression of cancer and other pathologies, and which signaling networks that are targeted by active Abl kinases are deregulated in various diseases.
We acknowledge the many scientists who have contributed to advancing our understanding of the roles of the ABL family kinases in health and disease, and apologize to those whose work could not be cited due to space limitations.
This work is supported by National Institutes of Health (NIH) [grant numbers R01CA195549, CA155160 and AI056266 to A.M.P.] and [grant numbers F30HL126448 and T32 GM007171F30 to A.K.]. Deposited in PMC for release after 12 months.
The authors declare no competing or financial interests.