|Home | About | Journals | Submit | Contact Us | Français|
The Abelson (ABL) family of nonreceptor tyrosine kinases, ABL1 and ABL2, transduces diverse extracellular signals to protein networks that control proliferation, survival, migration, and invasion. ABL1 was first identified as an oncogene required for the development of leukemias initiated by retroviruses or chromosome translocations. The demonstration that small molecule ABL kinase inhibitors effectively treat chronic myeloid leukemia opened the door to the era of targeted cancer therapies. Recent reports have uncovered roles for ABL kinases in solid tumors. Enhanced ABL expression and activation in some solid tumors, together with altered cell polarity, invasion or growth induced by activated ABL kinases, suggest that drugs targeting these kinases may have utility in treating selected solid tumors.
Tyrosine kinase signaling networks are required for the regulation of multiple cellular processes including growth, survival, invasion and angiogenesis during tumor initiation and progression. Drug discovery efforts in the past two decades have focused on developing tyrosine kinase inhibitors (TKIs) for cancer therapy. To date, the most successful of these efforts has been the development of ABL TKIs for the treatment of chronic myeloid leukemia (CML), which expresses the fusion oncoprotein BCR-ABL1. ABL1 was first discovered over 30 years ago as the oncogene in the Abelson murine leukemia virus1 and later identified as an oncogene associated with chromosome translocations in human leukemias2. These ABL1 fusion genes encode constitutively activated forms of the ABL1 tyrosine kinase required for cellular transformation. The ABL family of protein kinases comprised of ABL1 and ABL2 (ARG), function to link diverse extracellular stimuli to signaling pathways controlling cell growth, survival, invasion, adhesion, and migration3-6. In recent years, activation of ABL1 and ABL2 has been detected in solid tumors, including breast, colon, lung, and kidney carcinoma, as well as melanoma7-16. Activation of ABL kinases in solid tumors is not linked to chromosome translocation events, but rather is driven by enhanced ABL1 or ABL2 expression and/or by oncogenic tyrosine kinases, chemokine receptors, oxidative stress, and/or inactivation of negative regulatory proteins (cbioportal.org; 3, 16-21). Here we compare the modes of ABL family kinase activation in leukemias and solid tumors. Further, we dissect the role of the ABL kinases in the regulation of cellular processes critical for the progression of solid tumors including cell polarity, invasion, proliferation and survival. Several FDA-approved drugs targeting ABL kinases including imatinib and dasatinib have been tested for the treatment of solid tumors with mixed effectiveness; potentially due to the complexity of signaling pathways activated in solid tumors, the emergence of therapy resistance, and the fact these inhibitors have multiple cellular targets. Thus, these results highlight the necessity both to develop ABL kinase inhibitors with greater specificity and to identify biomarkers for patient stratification.
The two mammalian ABL kinases, ABL1 and ABL2, have both unique and overlapping physiological roles as revealed by genetic inactivation in mice. Disruption of murine Abl1 on a mixed (129/SvEv and C57BL/6J) genetic background resulted in neonatal lethality in about 50% of the mice, a phenotype that was more severe on the C57BL/6J genetic background22, 23. The Abl1 knockout mice exhibit splenic and thymic atrophy, reduced numbers of B and T cells, cardiac abnormalities, and osteoporosis linked to defective osteoblast proliferation and premature senescence22-26. In contrast, Abl2 (Arg) knockout mice are viable and exhibit neuronal defects that include age-related dendrite destabilization and regression27, 28. Functional overlap by Abl1 and Abl2 kinases is supported by the embryonic lethality of Abl1−/− Abl2−/− mice27.
ABL1 and ABL2 function redundantly in some cellular contexts, but also have unique roles, owing to distinct and conserved sequences and structural domains (Fig. 1). ABL1 and ABL2 share amino (N)-terminal regulatory and catalytic domains that are over 90% identical and include the Src homology 3 (SH3), SH2 and SH1 (tyrosine kinase) domains (Fig. 1A). The carboxy (C)-terminus of both ABL kinases contains a conserved filamentous (F) actin-binding domain29, 30 (and reviewed in 6) (Fig. 1A). ABL1 contains a G-actin (globular actin)-binding domain upstream of the F-actin-binding domain, and ABL2 has a second internal F-actin binding domain characterized by an I/LWEQ motif, which is required to promote the formation of F-actin-rich structures at the cell periphery30. The ABL kinases share conserved PXXP motifs that mediate binding to SH3 domain-containing proteins. Consistent with its nuclear and cytoplasmic localization, ABL1 has three nuclear localization signal (NLS) motifs and one nuclear export signal (NES) in its C-terminus. In contrast, ABL2, which lacks the NLS motifs, localizes primarily to the cytoplasm and preferentially accumulates at F-actin-rich sites in the cell periphery, focal adhesions, adherens junctions, invadopodia and phagocytic cups31-34. Multiple isoforms of ABL1 and ABL2 have been detected; alternative splicing of the first exons results in isoforms with distinct N-terminal sequences (Fig. 1A). The 1b isoforms of both ABL kinases contain an N-terminal glycine that is myristoylated, while the 1a variants lack this site and the corresponding modification.
The catalytic activities of the ABL kinases are tightly regulated by inter- and intramolecular interactions and post-translational modifications (Fig. 1B). This topic has been extensively reviewed and thus will only be briefly discussed here5, 6, 35. Among prominent intramolecular interactions required for autoinhibition are those involving the SH3 and SH2 domains35 (Fig. 1B). The SH3 domain binds to the polyproline-containing linker sequence that connects the SH2 and kinase domains, while the SH2 domain interacts with C-terminal lobe of the kinase domain (SH1) forming a SH3-SH2-SH1 clamp structure35. Autoinhibition is also mediated by the binding of the myristoylated residue in the ABL N-terminus to a hydrophobic pocket within the C-lobe of the kinase domain36. This interaction stabilizes the autoinhibited conformation of the myristoylated 1b isoform of the ABL kinase (Fig. 1B). Inhibitory intramolecular interactions that stabilize the 1a isoform of ABL are poorly understood and may involve binding to lipid residues in trans. In this regard, both ABL1 and ABL2 are inhibited by the lipid phosphatidylinositol 4,5-bisphosphate (PIP2) and experimentally decreasing PIP2 levels in the cell stimulates ABL kinase activity37.
ABL activity can be negatively or positively regulated by intermolecular interactions with distinct binding partners. In general, intermolecular interactions that disrupt autoinhibitory interactions and stabilize the active (open) conformation of the ABL kinases promote increased enzymatic activity. For example, the catalytic efficiency of the ABL kinases can be enhanced by interactions with adaptor proteins such as RIN1, which interacts with both the ABL SH3 and SH2 domains38. Intermolecular interactions that stabilize the inactive conformation of the ABL kinases inhibit enzymatic activity and downstream signaling. Among the proteins reported to negatively regulate ABL kinase activity are PRDX1 (PAG), FUS, and the PSTPIP1 adaptor protein5, 11, 39, 40.
The enzymatic activity of the ABL kinases is also regulated by tyrosine phosphorylation. Among the tyrosine residues phosphorylated on ABL1 are Y412 in the activation loop (corresponds to ABL2 Y439) and Y245 in the SH2-kinase domain linker (corresponds to ABL2 Y272). Phosphorylation of these tyrosines stabilizes the active conformation of the ABL kinases, and can occur by transphosphorylation between ABL proteins or may be mediated by Src family kinases41. The stability of ABL kinases can be modulated by tyrosine phosphorylation. Phosphorylation of ABL1 on Y245 and Y412 decreases protein stability through ubiquitination and subsequent degradation by the 26S proteasome machinery42.
Oncogenic activation of the ABL kinases occurs in Philadelphia-positive (Ph+) human leukemias as a consequence of the t(9;22)(q34;q11) chromosome translocation that generates BCR-ABL1 fusion proteins with constitutive tyrosine kinase activity. Three different BCR-ABL1 proteins have been identified that differ in the amount of BCR sequences retained in the fusion protein, and which are associated with distinct types of leukemia: P210 BCR-ABL1 is the molecular hallmark of CML (reviewed in 43); P185 BCR-ABL1 is present in 20-30% of adult and 3-5% of childhood B-cell acute lymphocytic leukemia (B-ALL)44, 45; and P230 BCR-ABL1 has been detected in neutrophilic-CML46 as well as rare cases of CML47. The preferential association of the various BCR-ABL1 forms with distinct leukemias is not absolute as some cases of Ph+ B-ALL express P210 BCR-ABL143. The role of the BCR-ABL1 fusion proteins in the initiation and progression of leukemia has been extensively covered in previous reviews and will not be discussed here48, 49. Oncogenic activation of ABL1 in the BCR-ABL1 chimeric proteins is dependent on the presence of an N-terminal coiled-coil (CC) oligomerization domain in BCR50, 51. Multiple signaling pathways have been identified that function to mediate the oncogenic activity of BCR-ABL1, and include the RAS, NF-κB, PI3K-AKT, JUN, β-catenin, and STAT3 signaling pathways48, 52. Recent proteomic approaches have identified proximal signaling networks linked to BCR-ABL153. Interestingly, some of the adaptors and signaling enzymes identified as components of the BCR-ABL1 network have been functionally linked to the ABL1 and/or ABL2 kinases5, 54.
Inhibition of P210 BCR-ABL1 with small molecule chemical inhibitors in chronic phase CML is perhaps the most successful example of molecular targeted therapy to date55 (see Box 1). The BCR-ABL1 inhibitor imatinib (Gleevec; STI571) is an effective first-line therapy for the majority of patients diagnosed with chronic phase CML, but it is much less effective when used to treat blast crisis CML and Ph+ B-ALL patients43. Inhibitors of the ABL kinases are classified into three main classes based on their mechanism of action: type 1 inhibitors target the active conformation of the kinase domain (dasatinib, bosutinib), whereas type 2 inhibitors stabilize the inactive conformation of the kinase domain preventing its activation (imatinib, nilotinib, ponatinib). The third category of inhibitors includes allosteric inhibitors which do not compete for ATP binding, but rather bind to regulatory domains to inhibit kinase activity (Figs. 1B,C). Among these are GNF-2 and GNF-5 which target the myristoyl-binding pocket in the C-lobe of the ABL kinase domain56. In contrast to the ATP-competitive inhibitors that target multiple kinases, the allosteric inhibitors are highly selective for the ABL kinases (Fig. 1C). These allosteric inhibitors decrease BCR-ABL1 enzymatic activity and inhibit BCR-ABL1-driven leukemogenesis in mice56; thus underscoring the importance of intramolecular autoinhibitory mechanisms regulating kinase activity of the oncogenic fusion protein.
Imatinib (Gleevec, STI571; Novartis), was the first targeted ABL inhibitor developed and approved for the treatment of CML. Treatment of CML and other Ph+ leukemias expressing BCR-ABL1 produced durable remissions in adults and children treated in the chronic phase49, 150. Remarkably, treatment of children with Ph+ ALL with imatinib plus chemotherapy has shown improved early event-free survival151. While imatinib has proven an effective drug for patients with CML, up to one-third of CML patients require alternate therapies132. The most common reason for imatinib resistance is the development of mutant forms of BCR-ABL1, including the T315I gatekeeper mutant, which prevent imatinib from docking in the ATP-binding pocket. Second-generation inhibitors such as nilotinib (Tasigna, AMN107; Novartis), which inhibits ABL kinases at 10- to 20- fold lower concentrations than imatinib, are used to treat imatinib-resistant disease131. Other second-generation inhibitors, including dasatinib (Sprycel, BMS-54825; Bristol-Myers Squibb) and bosutinib (Bosulif, SKI-606; Pfizer), inhibit ABL kinases at even lower concentrations than nilotinib, but like nilotinib and imatinib are ineffective against the BCR-ABL1 T315I mutant131. More recently, ponatinib (Iclusig, AP24534; Ariad Pharmaceuticals) was shown to inhibit both wild type and imatinib-resistant BCR-ABL1 mutants, including the T315I mutant, with similar potency. Similar to dasatinib and bosutinib, ponatinib has broad target specificity152 (Fig. 1C).
ABL1 has been identified as a fusion partner with an increasing number of genes in rare chromosome translocations in T cell acute lymphoblastic leukemia (T-ALL), B-ALL and other leukemias (Fig. 2). In addition to BCR, at least eight other ABL1 fusion partners have been identified to date including NUP214, ETV6 (TEL), EML, ZMIZ1, RCSD1, SFPQ, FOXP1 and SNX257-59 (Fig. 2). The BCR-ABL1 and ETV6-ABL1 fusions are associated with several types of leukemia including CML, B-ALL, T-ALL and acute myeloid leukemia (AML), and these ABL1 translocations may occur in hematopoietic pluripotent or committed stem cells (reviewed in 58, 60.
In contrast, the NUP214-ABL1 and EML1-ABL1 fusions are primarily linked to T-ALL, while the RCSD1-ABL1, SFPQ-ABL1, ZMIZ1-ABL1, FOXP1-ABL1 and SNX2-ABL1 fusions are associated with B-ALL. In general, the fusion partners encode proteins that contain coiled-coil or helix-loop-helix motifs that promote oligomerization of the resulting chimeric ABL kinases. However, the NUP-ABL1 fusion, which is present in about 6% of T-ALL cases, requires localization to the nuclear pore complex rather than oligomerization for enhanced transforming activity61. While the majority of chimeric fusions involve ABL1, the ABL2 gene was identified as a partner of ETV6 in rare cases of T-ALL and AML60 (Fig. 2). In general, the distinct partners fused to the N-terminus of the ABL kinases promote enhanced kinase and transforming activities by disrupting inhibitory intramolecular interactions, providing sequences that facilitate oligomerization, enhancing tyrosine phosphorylation, and/or by recruiting the chimeric kinases to distinct subcellular sites and protein complexes. The effectiveness of the ABL TKIs for the treatment of leukemias associated with the diverse ABL fusion partners other than BCR remains to be established.
In contrast to the well-established role of BCR-ABL1 in CML, less is known about the role of the ABL kinases in solid tumors. Accumulating data support a role for ABL kinases in solid tumors characterized by enhanced expression and/or activation of these kinases. The Cancer Genome Atlas (TCGA) and other studies have shown that ABL1 and ABL2 are amplified and/or overexpressed in invasive breast carcinoma, ovarian serous cystadenocarcinoma, lung adenocarcinoma, and lung squamous cell carcinoma (17-20) (Table 1). Alterations in ABL1 or ABL2 are also found in skin cutaneous melanoma, bladder urothelial carcinoma, colorectal adenocarcinoma, rhabdoid tumors, as well as renal medullary and clear cell renal carcinoma (cbioportal.org; 20, 21, 62). Recent data generated from large-scale sequencing projects revealed that alterations (amplification, mRNA increase, or somatic mutations) are more common in ABL2 than ABL1, with observed ABL2 alterations in breast invasive carcinoma (18%), uterine endometrioid carcinoma (19%), ovarian serous cystadenocarcinoma (15%), lung adenocarcinoma (16%) and lung squamous cell carcinoma (15%) (cbioportal.org; 17) (Table 1). Prevalent among these alterations is ABL2 amplification, which has been detected in invasive breast carcinomas (12%)17 and non-small cell lung adenocarcinomas (9%) (cbioportal.org; 18, 19). These findings are consistent with earlier reports of elevated ABL2 expression in advanced high-grade colorectal, pancreatic, renal and gastric tumors, implicating an association between ABL2 activity and tumor progression21, 63-65. The prevalence of ABL2 amplification and enhanced gene expression are comparable to those reported for several kinases previously implicated in breast invasive carcinoma (TCGA; 17). Somatic mutations of ABL1 and ABL2 in solid tumors are rare, but have been reported in selected solid tumors, including lung cancer and uterine corpus endometroid carcinoma (66-69, cbioportal.org). The role of these mutations in regulating ABL oncogenic activity remains to be determined.
In addition to genomic amplification and changes in ABL mRNA expression in tumors, increased ABL kinase activity has been reported in several solid tumors. Quantitative proteomic analysis in cancer cells expressing the ERBB receptors showed that ABL1 was rapidly activated in response to epidermal growth factor (EGF) stimulation7. This is consistent with previous findings that showed ABL kinases are activated downstream of the receptor tyrosine kinases (RTKs) platelet-derived growth factor receptor (PDGFR) and EGF receptor (EGFR, also known as ERBB1) in fibroblasts70. Subsequent studies demonstrated that ABL kinases are tyrosine phosphorylated and activated in breast, lung, colorectal, gastric, and prostate cancer cells as well as in melanoma8-16. Analysis of global tyrosine phosphorylation profiles of lung cancer cells showed higher levels of ABL tyrosine phosphorylation, but the identity of the upstream signals promoting tyrosine phosphorylation of the ABL kinases, and their role in the initiation and progression of lung tumors is largely unknown10-13.
The mechanisms regulating ABL activation in solid tumors differs from the constitutive kinase activation observed in BCR-ABL1+ disease (Table 2). In solid tumor cells, ABL activation has often been linked to stimulation by hyperactive RTKs and chemokine receptors8, 33. Indeed, activation of ABL kinases in breast cancer cells has been shown to occur downstream of the EGFR, Her2 (ERBB2), and insulin-like growth factor receptor (IGFR) as well as the chemokine receptor, CXCR48, 33, 71-73. Activation of ABL1 in GTL-16 gastric carcinoma cells and HepG2 hepatocellular carcinoma cells was induced by ligand-dependent activation of MET15, while ABL1 activation in HT-29 colon cancer cells was stimulated by PDGF9. Enhanced ABL1 kinase activity in human anaplastic thyroid carcinoma cells was induced by the RET-PTC3 fusion protein, a constitutively active form of the RET RTK74. Thus, the catalytic activity of the ABL kinases can be up-regulated by ligand-dependent and ligand-independent activation of RTKs. In addition to activation by upstream kinases, ABL kinases can also be activated through the inactivation of negative regulatory networks including interactions with inhibitory proteins. In this regard, activation of ABL1 in two lung adenocarcinoma cell lines correlated with loss of the tumor suppressor protein FUS11.
Activation of ABL kinases in solid tumor types has been linked to alterations in cell growth and survival. Several studies have reported inhibitory effects of imatinib, nilotinib and related ABL TKIs (Box 1) on cancer cell proliferation and survival in vitro and in xenograft studies, but the cellular responses to these compounds cannot be solely attributed to inhibition of the ABL kinases as these compounds have numerous kinase and some non-kinase targets75, 76 (Fig. 1C). Depletion of either ABL1 or ABL2 with specific siRNAs in breast cancer cells that express elevated levels of one or both proteins showed that both kinases were required for anchorage-independent growth72. Both ABL kinases were also required for maximal proliferation of melanoma cell lines16. In contrast, growth and survival under nutrient-deprivation required ABL2 but not ABL1 expression in some breast cancer cells (MDA-MB-435s and BT-549)72. Additionally, depletion of ABL2 alone in non-small cell lung cancer lines decreased cell growth77, and ABL1 alone was required for survival and anchorage-independent growth of gastric and hepatocellular carcinoma cells15. Among pathways mediating ABL1-dependent proliferative pathways are the Rac, p38, and ERK5 signaling pathways (reviewed in 78). Thus, cancer cells expressing high levels of ABL1 and/or ABL2 can become reliant on one or both kinases for cell growth and viability.
Accumulating data show that a growing number of signaling networks (such as the transforming growth factor β (TGFβ), WNT, and Hippo pathways) exhibit context-dependent activities in solid tumors and can promote tumorigenesis or tumor suppression under different circumstances79-81. While ABL kinases promote cell proliferation and survival in leukemias and some solid tumor cell lines, three reports have shown that ABL kinases suppress the growth of breast cancer xenografts82-84. Stimulation of the EphB4 receptor with ephrin-B2 in breast cancer cells decreased cell growth through an ABL-CRK pathway, and imatinib blocked the inhibitory effect of ephrin-B2 on the growth of these breast cancer xenografts82. Further, expression of a constitutively active form of ABL1 in the 4T1 murine mammary tumor cell line enhanced the growth suppressive effects of TGF-β in 3-dimensional (3D) cultures, and inhibited growth of 4T1 tumor xenografts83. A recent report showed that knockdown of ABL2 in the MDA-MB-231 breast cancer cell line enhanced the growth of MDA-MB-231 xenograft tumors due to increased cell proliferation 84. ABL1 and ABL2 have overlapping functions during cell proliferation in some cell types, but the consequences of depleting both kinases on the growth of the MDA-MB-231 xenograft tumors was not reported in this study84. In contrast to the growth inhibitory role of ABL kinases in the 4T1 and MDA-MB-231 xenografts, depletion of both ABL1 and ABL2 in the MCF7 breast cancer cell line markedly impaired the growth of MCF7 xenograft tumors, supporting a tumor-promoting function for ABL kinases in this cell line85. A role for ABL kinases in the growth of mammary tumors may depend on the cellular context, the repertoire of oncogenes and tumor suppressors expressed in the tumor, redundancy in the function of ABL1 and ABL2 kinases, and upstream signals that regulate tumor suppressive versus promoting pathways in distinct cell lines.
ABL kinases can be activated in response to distinct types of cellular stress including those mediated by reactive oxygen species (ROS)86-89. Elevated levels of ROS are a feature characteristic of many solid tumors90, 91 and are also an inevitable by-product of cellular metabolism. High levels of ROS can induce cell death, whereas lower chronic levels of ROS may induce inflammation, promote tumorigenesis, and other disease states92. Oxidative stress results when the level of oxidants within the cells supersedes the antioxidant defense and may be induced by exogenous as well as endogenous sources93. Due to increased cellular metabolism, cancer cells are vulnerable to oxidative stress as the antioxidant defense system may be operating at its maximum threshold94. ABL kinase activation and signaling appear to be differentially affected by ROS levels. ABL2 was reported to promote cell survival in response to low levels of oxidative stress, but high levels of oxidative stress correlated with ABL2 degradation and apoptosis95, 96. In this regard, increased tyrosine phosphorylation of ABL2 on Y439 induced by high ROS levels promoted ABL2 ubiquitination and protein degradation95. In contrast, phosphorylation on Y261 enhanced ABL2 protein stability in response to low ROS levels95.
ABL kinases can mediate anti-proliferative and apoptotic responses induced by ROS and DNA damage86, 97, 98. Whether ABL1 functions to promote or inhibit cell growth may depend on the status of p53 (TP53) or the related p73 (TP73)99-103. ABL1 was shown to enhance cell proliferation of fibroblasts lacking p53 under non-stress conditions99, 104. However, in response to cellular stresses, including DNA damage, ABL1 induced cell growth arrest and/or apoptosis mediated by p53 or p73104. Interestingly, ABL1 can differentially modulate p53 activity and cellular responses through alteration of Mdm2, a regulator and transcriptional target of p53. Mechanistically, ABL1 protects p53 from the inhibitory effect of Mdm2, by promoting Mdm2 tyrosine phosphorylation, which restricts its interactions with p53, allowing for the accumulation of p53 and the promotion of apoptosis under stress conditions104. Recently, ABL1 was shown to be required for cell survival and activation of a p38 MAP kinase-p53-Mdm2 pathway in response to ligand-activated Met kinase in hepatocellular carcinoma cells15. Thus, ABL kinases can differentially disrupt the p53-Mdm2 regulatory loop to suppress or promote cell growth and survival depending on the upstream signals leading to ABL activation and the cellular context.
In addition to oxidative stress, growth factor deprivation was reported to activate ABL1 in mammary epithelial cells by inducing binding to the Mig6 (ERRFI1) adaptor and triggering p73-dependent apoptosis89. Silencing ABL1 in primary mammary epithelial cells (pMECs) in 3D cultures impaired lumen formation, which may be due, in part, to decreased cell death89. However, in comparison to wild-type pMECs, the ABL1-depleted pMECs failed to reconstitute the mammary gland after transplantation into the cleared mouse fat pads, suggesting impaired cell proliferation in the absence of ABL189. It is likely that activation of ABL kinases by oxidative stress and/or reduced nutrient availability prevalent in many solid tumors, may modulate tumor cell survival, invasion, and metastasis90, 91.
Disruption of cell polarity is an early event in the evolution of epithelial tumors105. A role for ABL kinases in the regulation of epithelial apical-basal polarity was first described in Drosophila106-108. Drosophila Abl (D-Abl) also regulates planar cell polarity during axis elongation by modulating the tyrosine phosphorylation and localization dynamics of β-catenin109. Recently, D-Abl was shown to be required for planar cell polarity in the Drosophila eye and wing by regulation of Frizzled/Dishevelled signaling110. Furthermore, expression of activated ABL2, and to a lesser extent ABL1, induces a striking inversion of apical-basal polarity in epithelial cells grown in a 3D cell culture system in the presence of collagen111 (Fig. 3A). Inversion of apical-basal polarity by active ABL2 is mediated by at least two pathways, which include the disruption of β1-integrin localization and signaling downstream of Rap1, and impaired Rac1-mediated laminin assembly111 (Fig. 3B). Thus, ABL kinases may contribute to loss of epithelial polarity at early stages of tumor development.
Spindle orientation is important for morphogenesis, asymmetric cell division and stem cell self-renewal, processes that are deregulated during tumorigenesis112-115. Notably, a recent report showed that ABL1 regulates spindle orientation in epithelial cells in culture and the mouse epidermis116. Loss of ABL1 induces spindle misorientation through regulation of the LGN (LeuGly-Asn repeat-enriched protein; GPSM2) and nuclear mitotic apparatus (NuMA) polarity proteins. Whether the ABL1-induced changes in spindle orientation contributes to tumor progression remains to be determined. Interestingly, β1-integrin regulates apical-basal epithelial cell polarity and controls spindle orientation in embryonic mouse skin, and the localization of LGN and NuMA is disrupted in the epidermis of β1-integrin knockout mice117. Because expression of activated ABL2 disrupts β1-integrin localization and signaling in epithelial cells111, an intriguing possibility is that activated ABL kinases might disrupt spindle orientation indirectly through β1-integrin as well as through directly targeting the LGN and NuMA polarity proteins.
The progression of solid tumors involves dramatic changes in cell morphology, adhesion, motility and polarity that are dependent on cytoskeletal rearrangements. Epithelial tumors can activate dormant developmental pathways that promote conversion from organized epithelial tissue into isolated cells with mesenchymal morphology through the Epithelial-to-Mesenchymal Transition (EMT) program118, 119. EMT is characterized by loss of cell polarity, disassembly of cell-cell junctions, acquisition of mesenchymal cell morphology and gene expression profiles, together with enhanced cell motility and invasion. Notably, these cellular processes are linked to signaling networks regulated by ABL family kinases. ABL1 was required for PDGF-induced EMT in colon cancer cells; this involved ABL1 phosphorylation of the p68 helicase (DDX5), which promoted β-catenin nuclear translocation9. Consequently, depletion and pharmacological inhibition of ABL kinases in colon and cervical cancer cells promoted epithelial phenotypes in cells that had been induced to undergo EMT9, 120. Recently, ABL1 was shown to be required for claudin-1-induced EMT in human liver cells and hepatocellular carcinoma cell lines by regulating the activation of RAS-RAF-ERK pathway121. The role of ABL1 in EMT may be context-dependent, however, as pharmacological inhibition or silencing of ABL kinases in normal epithelial cells results in dissolution of cell-cell junctions without inducing expression of genes required for the EMT program32, 111. Interestingly, alternative splicing promoted by EMT in breast cancer cells results in the formation of a spliced ABL2 isoform predominantly expressed in mesenchymal cells122. Roles for this ABL2 isoform specifically and the ABL kinases in general during EMT in solid tumors are poorly defined and require further study. The EMT program is linked to acquisition of therapeutic resistance in solid tumors118, and ABL signaling has recently been shown to be up-regulated in chemoresistant breast tumors123. Thus, an intriguing possibility to be evaluated is whether ABL kinases are important for induction and/or maintenance of EMT pathways which contribute to acquisition of therapeutic resistance in diverse solid tumors.
Tumor progression and metastasis require the induction of an invasive program that promotes dissemination into surrounding tissues by primary tumor cells, followed by intravasation, migration, extravasation, and formation of metastases at distant sites124. ABL kinases engage the actin polymerization machinery in response to growth factor and chemokine stimulation as well as adhesion cues, and promote formation of membrane protrusions, morphological changes, altered cell adhesion, migration and invasion of diverse cell types6, 37, 54, 70. ABL kinases are required for cell motility and invasion induced by IGF-1, EGF, serum and chemokines in breast, hepatocellular carcinoma, and melanoma cells8, 16, 33, 73, 84, 125. One or both ABL kinases are required for invasion in several cancer cell lines, but ABL2 is preferentially required for invasion of the MDA-MB-231 breast cancer cell line84. ABL kinases can regulate invasion by directly phosphorylating proteins required for invasive behavior, and also by regulating the expression of genes linked to invasion and cell motility (Fig. 4). ABL2 localizes to invadopodia, where it is required for both matrix degradation and invasion in breast cancer cells33, 73. Knockdown of ABL2 decreases matrix degradation and invasion by MDA-MB-231 breast cancer cells, and ABL2 regulates the maturation of invadopodia by linking activation of the EGFR and Src kinases to tyrosine phosphorylation of cortactin, which is required for Arp2/3-dependent actin polymerization73. In addition to cortactin, other actin regulatory proteins are targeted by ABL kinases at invadopodia (Fig. 4). Among these are the Abl interactor 1 (Abi1) adaptor protein, a component of the WAVE (WASF) protein complex, which is required for invadopodia formation, ECM degradation and induction of matrix metalloproteinase (MMP)-9 expression in MDA-MB-231 breast cancer cells126. ABL kinases also regulate the activity of MMPs during invadopodia maturation (Fig. 4). Active ABL2 interacts with and promotes phosphorylation of the membrane type 1-matrix metalloproteinase (MT1-MMP, also known as MMP14) and is required for its localization and function at invadopodia33. Both ABL1 and ABL2 kinases are required for melanoma cell invasion by regulating MMP expression through STAT3-dependent and independent pathways16 (Fig. 4). Other potential targets of ABL kinases at invadopodia include the N-WASP (WASL) actin regulatory protein, which is required for Arp2/3-mediated actin polymerization at invadopodia, and is phosphorylated by ABL kinases127-129. Recently, HEF1 (NEDD9) was identified as a direct substrate and binding partner of the ABL kinases in T cells54. The HEF1 protein also regulates cancer cell invasion, and ABL-mediated phosphorylation of HEF1 may be required for promoting invasive behavior as observed during chemokine-induced T cell migration130. Consistent with its ability to positively regulate invadopodia formation and function, knockdown of ABL2 decreased cancer cell invasion and intravasation following implantation of MDA-MB-231 cells in the mammary fat pad84.
To date, there are five approved ABL inhibitors for the treatment of BCR-ABL1+ CML (Box 1). The pharmacology and clinical data of these drugs has been recently reviewed131. While imatinib has proven remarkably successful for the treatment of CML, particularly those in chronic phase, up to one-third of patients will require alternative therapies132. The effectiveness of imatinib decreases during the acute and blast stages of CML. Development of imatinib resistance may result from the development of mutations in BCR-ABL1 that alter drug binding, including the T315I gatekeeper mutation. Other ATP-competitive inhibitors, such as nilotinib, dasatinib, and bosutinib are also ineffective against the T315I mutant form of BCR-ABL1131. Newer allosteric inhibitors have been tested in preclinical models of resistance; among these are the GNF-2 and GNF-5 allosteric inhibitors that target the myristate pocket of ABL kinases (Fig. 1C). Combining GNF-5 with imatinib or nilotinib prolonged survival of mice with imatinib-resistant BCR-ABL1 T315I-induced leukemia56. Another type of allosteric inhibitor, an engineered ABL1 SH2-binding fibronectin type III monobody, blocked the intramolecular interaction between the SH2 and kinase domains of BCR-ABL1, inhibited kinase activity in vitro and in vivo, impaired leukemogenesis in mice, and increased the sensitivity of imatinib-resistant BCR-ABL1 mutants to TKI therapies133. Thus, the combined use of allosteric and ATP competitive inhibitors may represent an effective strategy to overcome resistance to either type of compound alone.
In contrast to the success of TKI therapies for the treatment of BCR-ABL1-induced leukemias, use of the ATP-competitive inhibitors imatinib, nilotinib and dasatinib has not achieved similar success for the treatment of solid tumors (reviewed in 134, 135). The variable clinical responses to these TKIs may reflect the heterogeneous nature of solid tumors, which often have acquired mutations in multiple tumor promoting pathways. In addition, recent work has shown that signaling networks can be rewired in response to treatment with single kinase inhibitors136, 137, which may underlie poor responses to treatment. Another mechanism that underlies the poor response to TKI therapy in some solid tumors and chemoresistant leukemias is the targeting of alternative pathways resulting in paradoxical activation of proliferative pathways. Treatment with imatinib, nilotinib, and dasatinib can activate BRAF/ CRAF (RAF1) complexes in cells expressing active RAS leading to enhanced activation of the MEK-ERK pathway in diverse cancer cell lines including melanoma, lung, colorectal and pancreatic carcinoma, as well as BCR-ABL1 TKI-resistant leukemic cells76. The paradoxical MEK-ERK activation by these TKIs is reminiscent of the increased MEK-ERK pathway activation induced by vemurafenib, a B-RAF inhibitor, in cancer cells expressing active RAS, which was reported to promote cell proliferation in soft agar and accelerate cutaneous squamous-cell tumors in mouse models and patients138. At the molecular level, whereas imatinib, nilotinib, and dasatinib promote the formation of B-RAF/C-RAF complexes in cells with activated RAS, these TKIs are weak RAF inhibitors in vitro76. Therefore, the indiscriminate use of imatinib, nilotinib, dasatinib and other ABL-targeted ATP-competitive TKIs is not recommended; rather, screening tumors for ABL amplification, enhanced expression and hyperactivation, as well as co-expression of other mutations associated with poor response to specific TKIs, is crucial for identifying those tumors that may benefit from therapies with selective ABL TKIs.
Notably, in contrast to the ABL-targeted ATP-competitive TKIs, the ABL allosteric inhibitor GNF-2 does not inhibit BRAF activity in vitro, fails to induce formation of BRAF/CRAF dimers, and thus, treatment with this compound does not elicit paradoxical activation of the MEK-ERK pathway76. Development of additional selective ABL TKIs and further mechanistic understanding of the signaling networks disrupted in response to distinct ABL TKIs, may better inform the selection of therapies that can be used in combination with ABL inhibitors. Thus, tumors with ABL2 genomic amplification and/or robust kinase activity (as detected by high levels of pCRKL) might be predicted to be good candidates for TKIs that are selective for ABL kinases (and do not target B-RAF or other kinases that may induce paradoxical activation of MEK-ERK and other tumor-promoting pathways). Additionally, developing an ABL signature (genomic, transcriptional, or phospho-proteomic) might be critical to identify subset of patients to rationally test the effects of these inhibitors.
One interesting possibility for the future of ABL TKIs in cancer therapy may lie in overcoming resistance to chemotherapies. Enhanced expression of ABL1 signaling networks was recently detected in endocrine-resistant breast cancer123. ABL1 and PDGFR were identified as the top adaptive pathways characteristic of estrogen receptor (ER)-positive breast cancers that acquired resistance to aromatase inhibitor (AI) therapy. Activation of the PDGFR and ABL1 pathways was linked to long-term estrogen deprivation in MCF7 breast cancer cells, and was associated with decreased anti-proliferative response to AI treatment in primary ER-positive breast carcinomas123, 139. Moreover, treatment of these cells with nilotinib (which inhibits both the PDGFR and ABL1), suppressed ER-mediated transcription by destabilizing the ER. This finding is consistent with reports that ABL1 regulates ERα-mediated transcription by inducing phosphorylation and stabilization of ERα140 and that silencing ABL1 in ER-positive breast cancer cells sensitized these cells to tamoxifen treatment141. Treatment of breast cancer cells with imatinib has been shown to sensitize these cells to selected drugs such as 5-fluorouracil, cisplatin and vinorelbine142, 143. Furthermore, imatinib treatment or ABL1 depletion was shown to restore sensitivity to the EGFR and ErbB2 inhibitor lapatinib in breast cancer cells that had acquired lapatinib-resistance144. Future studies are warranted to define how ABL1 inhibition sensitizes cancer cells to chemotherapies and whether such an approach may improve the effectiveness of chemotherapy in vivo.
While the ABL1 kinase was first linked to the induction of human leukemia almost three decades ago2, a role for ABL family kinases in the progression of solid tumors has only recently been recognized. Over the last decade the use of imatinib, dasatinib and nilotinib to treat patients with BCR-ABL1-positive leukemias provided the best example for the successful use of TKI-targeted therapy. However, it is clear that the use of these drugs is inadequate for the treatment of solid tumors as monotherapies due to the complexity of mutations even in early-stage tumors, and the potential for inappropriate activation (rather than inhibition) of proliferative pathways by some TKIs with multiple protein targets. Ongoing preclinical and future clinical studies are expected to evaluate whether treatment of patients with tumors with ABL amplification, enhanced expression and hyper-activation could benefit from therapies with ABL-selective TKIs. The identification of new allosteric inhibitors with greater specificity against ABL family kinases may allow for evaluation of the effectiveness of targeting these kinases in the treatment of solid tumors with hyperactive ABL kinases, and whether anti-ABL therapies display greater effectiveness at specific tumor stages. Recent findings suggest that ABL1 and ABL2 may regulate overlapping as well as distinct cellular processes in various cell types, and may differentially contribute to tumor progression. Future studies will be required to evaluate unique roles for the ABL kinases not only in selected solid tumors, but also in cells in the tumor microenvironment. ABL kinases regulate diverse signaling pathways in fibroblasts, macrophages, and T cells34, 54, 70, 100, 145, 146. These cell types are known to contribute to tumor progression and metastasis in various cancers147-149, and a role for ABL kinases in tumor-promoting pathways in distinct cell types in the microenvironment should be an area of active enquiry. Exciting research over the next few years is expected to evaluate the contribution of ABL kinases to distinct tumor types and establish whether these kinases prove to be useful targets for the treatment of multiple solid tumor types.
The authors regret that due to space limitations, they could not directly cite the work of many investigators. We thank Ran Li for confocal images. The Pendergast laboratory is supported by grants from the National Cancer Institute, including NCI grants R01CA155160 and R01CA070940 to A.M.P. We also acknowledge the support by training grants from the Department of Defense Breast Cancer Research Program Fellowship W81XWH-10-1-0345 and the Pharmaceutical Research and Manufacturers of America to E.K.G., and ACS Spin-Odyssey Postdoctoral grant 11984-7-PF-10-228-01-CSM to P.S.S.P.