Src family kinases and paclitaxel sensitivity

Abstract

Src-family Kinases (SFKs) participate in the regulation of proliferation, differentiation, apoptosis, autophagy, adhesion, migration, invasion and angiogenesis in normal and cancer cells. Abnormal expression of SFKs has been documented in cancers that arise in breast, colon, ovary, melanocyte, gastric mucosa, head and neck, pancreas, lung and brain. Targeting SFKs in cancer cells has been shown to be a promising therapeutic strategy in solid tumors, particularly in ovarian, colon and breast cancers. Paclitaxel is one of most widely used chemotherapeutic agents for the management of ovarian, breast, lung and head and neck cancers. As a microtubule-stabilizing agent, paclitaxel possesses both mitosis-dependent and mitosis-independent activities against cancer cells. A variety of mechanisms such as deregulation of P-glycoprotein, alteration of tubulin isotypes, alteration of microtubule-regulatory proteins, deregulation of apoptotic signaling pathways, mutation of tubulins and overexpression of copper transporters have been implicated in the development of primary or secondary resistance to paclitaxel. By affecting cancer cell survival, proliferation, autophagy, microtubule stability, motility, and/or angiogenesis, SFKs interact with mechanisms that regulate paclitaxel sensitivity. Inhibition of SFKs can potentiate the anti-tumor activity of paclitaxel by enhancing apoptosis, autophagy and microtubule stability. Based on pre-clinical observations, administration of SFK inhibitors in combination with paclitaxel could improve treatment for ovarian, breast, lung and head and neck cancers. Identification and validation of predictive biomarkers could also permit personalization of the therapy.

Introduction

Src-family kinases (SFKs) are a group of membrane-associated non-receptor tyrosine kinases that include c-Src, Fyn, c-Yes, Lyn, Lck, c-Fgr, Blk, Hck and Yrk.^1–4 While Lyn, Lck, c-Fgr, Blk and Hck are expressed primarily in hematopoietic lineages, c-Src, Fyn and c-Yes can be detected in a variety of cell types.^2,4 In addition to these nine members of the Src-family, there are two functionally related kinases (Csk and Fyb) that modulate the activities of SFKs.^5–8 SFKs regulate a number of biological functions such as cell proliferation, survival, differentiation, autophagy, adhesion, motility, invasion and angiogenesis both in normal cells and in cancers.^1–4,9 In this review, we will briefly describe the structure, regulation and signaling induced by SFKs, and then focus on the expression of SFKs in ovarian, breast, lung and head/neck cancers, the mechanisms of action and resistance of paclitaxel, and on how SFKs inhibition sensitizes cancer cells to treatment with paclitaxel.

Structure and Regulation of SFKs

Each of the nine Src-family members exhibits a similar structure composed of four common Src homology domains (SH1–SH4), a unique domain (UD) characteristic of each individual kinase, and a carboxy-terminal tail (CT) containing the major regulatory tyrosine residue (Fig. 1).^2–4 The SH1 domain is the catalytic constituent of SFKs, while the SH4 domain links SFKs to the cell membrane. The SH2 domain mediates phosphotyrosine-dependent protein-protein interactions, while SH3 domain mediates protein-protein interactions by binding to proline-rich motifs.^2–4,10

Figure 1.

Schematic diagram showing the typical structure and integrated signaling and functions of SFKs.

The SH1 domain contains one important tyrosine autophosphorylation site (Y419) whose phosphorylation activates SFKs by presenting a binding pocket for substrates.^2–4 The major regulatory tyrosine residue is Y530 in the CT domain of SFKs that negatively regulates kinase activity. Y530 can be phosphorylated by Csk and Csk homologous kinase, which inactivates SFKs by folding the SH2 and CT domains, thus preventing binding of SFKs to the phosphorylated tyrosine residues of their partner proteins. Y530 can be dephosphorylated by cytoplasmic PTP1B, SHP1/2 and transmembrane enzymes CD45 or PTPα/γ/ε, which reopens up the molecule to an active state.^2–4 In addition, Src phosphorylation on Y213 in the SH2 domain affects its kinase activity.^11,12 Cdc2 is also reported to phosphorylate and activate Src during mitosis.¹³

In addition to tyrosine phosphorylation, SFKs are also regulated through SH2/SH3-directed protein interaction, adaptor proteins and redox regulation.^5–8,10,14 SFKs can be activated by the SH2/SH3 ligands such as a phosphotyrosine residue of the autophosphorylated receptors or a proline motif of β-arrestin.^2–4,10,15 Adaptor protein Cbp/PAG is an E3 ubiquitin ligase that can positively or negatively regulate SFK activity through its ability to interact with SFKs, Csk phosphatases and down stream signaling molecules.^16,17 Other adaptors such as LIME, LAT and NTAL have also been reported to regulate SFK activity in hematopoietic cells.¹⁷ SFKs contain conserved redox-regulated cysteine residues (Cys245 and Cys487) located in the SH1 and SH2 domains and can be oxidized and activated by ROS.¹⁴

Function-Associated Signaling of SFKs

SFKs function by associating with a number of membrane-bound receptors and intracellular proteins and are the key mediators of various signal transduction pathways as shown in Figure 1. SFKs integrate signaling from multiple transmembrane-associated receptors and molecules such as RTKs (VEGFR, PDGFR, EGFR, IGF-1R, c-Met or HER2), GPCRs (adrenergic, purinergic or serotonin receptors), integrins, TCR/Ig, cadherin or caveolins, and from multiple cytosolic receptors and molecules such as ER and NOS.^1,18–24 SFKs then activate multiple target proteins including FAK, RhoA, cortactin, p130CAS, paxillin, tensin-3, STAT3, Ras-MAPK, PI3-kinase, p27^Kip1 and Bcl-2.^{1,18–21,25,26} These actions collectively modulate cell migration, invasion, adhesion, survival, differentiation, proliferation, autophagy and angiogenesis. The effects of SFKs on signaling molecules and pathways of cell motility, survival, autophagy and proliferation will modulate drug sensitivity such as paclitaxel, which will be discussed below in detail.

Aberrant Expression of SFKs in Human Cancers

c-Src, the founding member of the SFKs, was originally identified as a retroviral oncoprotein that induces transformation of avian cells and formation of sarcomas in chickens.^2,4,27,28 Increased expression of SFKs, particular c-Src, protein has been detected in various solid tumors including colon, breast, ovarian, lung, head and neck, skin, liver, gastric, prostate, pancreatic and brain cancers.^{1,18–21,29,30}

Colon cancer.

Colon cancer is most often associated with elevated c-Src activity.^29,31 Accumulated data have convincingly demonstrated that c-Src overrexpression and activation not only occurs in the late stages of colon cancer metastasis, but also appears in the early stages of colon cancer progression.²⁹ c-Src activation has been shown to be an independent biomarker of poor clinical prognosis in all stages of colon cancer.²⁹ In addition to c-Src, elevated expression of another SFK member, Yes, have been also indicated in colon cancer.²⁹

Breast cancer.

Evidence from two transgenic mouse models clearly revealed the critical role of c-Src in the formation and progression of breast cancer. Lack of functional c-Src almost completely abolished the formation of mammary tumors in mice expressing the polyomavirus middle T oncogene.³² In another transgenic mice model that expressed a mouse mammary tumor virus-Neu fusion gene, the Neu-induced mammary tumors displayed 6–8 fold-higher c-Src kinase activity than the adjacent non-tumor epithelium.³³ Several clinical studies have demonstrated that c-Src expression and kinase activity are elevated in breast cancer and ductal carcinoma in situ relative to normal breast tissues.^34–38 c-Src activity is increased by HER2/Neu overexpression in human mammary epithelial cells through posttranslational mechanisms.^39,40 Src inhibitors have been tested in clinical trials and appear as a promising therapeutic option in the treatment of breast cancer.^41,42

Ovarian cancer.

A number of studies confirm the important roles of SFKs in ovarian cancer development and progression. c-Src expression in human ovarian cancer cell lines was found to be higher than that of colon cancer cells.³¹ Our own study found that both c-Src and Yes are overexpressed in ovarian cancer cell lines and that c-Src was overexpressed and activated in 15 clinical specimens of metastatic ovarian cancer.⁴³ Knockdown of c-Src activity with an antisense c-Src reduced VEGF mRNA expression, decreased anchorage-independent growth in vitro and diminished growth of SKOv3 xenografts in vivo,⁴⁴ suggesting that c-Src can regulate ovarian cancer growth. Expression and activity of uPA, a protease that contributes to the metastatic potential of many cancers, required c-Src signaling to promote ovarian tumor invasion and progression.^45,46 LPA, a known inducer of ovarian cancer proliferation and motility, relies on SFK family member Fyn and c-Src to induce junction dispersal and migration of ovarian cancer cells.^47,48 c-Src can inhibit TGFβ1-mediated upregulation of anti-angiogenic factors Smad2/3 and PAI-1.⁴⁹ SFKs (c-Src and Yes) not only positively regulate VEGF production, but also can destabilize endothelial barrier function and promote tumor cell extravasation in vivo.⁵⁰ Src activation is required for stress hormone (e.g., norepinephrine)-induced IL-6 expression, which may contribute to ovarian tumor progression.⁵¹ A novel nuclear receptor co-regulator, PELP1, is overexpressed in 60% of ovarian cancers and depends upon Src and PI3K-AKT signaling pathways to exert its effects.⁵² Src is also shown to partially mediate endothelin-1/β-arrestins-induced ovarian tumorigenesis.⁵³ Finally, c-Src expression correlates with cortactin-regulated cytoskeleton function and hyaluronic acid-dependent ovarian cell tumor migration.⁵⁴

Lung cancer.

Non-small-cell lung cancer (NSCLC) cell lines are found to have even higher c-Src activity than that of colon cancers.³¹ Increased expression of c-Src is observed in about 60% of patients with lung cancers, including neuroendocrine tumors (small-cell cancer and atypical carcinoid) and non-small-cell tumors such as adenocarcinoma, bronchoalveolar and squamous-cell lung cancers.⁵⁵ c-Src expression and its kinase activity are markedly activated in NSCLC samples, especially in adenocarcinomas, and associated with the Y530 dephosphorylation.⁵⁶ c-Src activation in NSCLC has been confirmed in a second clinical study.⁵⁷ c-Src inhibition in lung cancer cell lines with small molecule inhibitors or small interfering RNAs is associated with impressive antitumor effects, further supporting the role of c-Src in lung cancer.^58–64

Head and neck cancer.

Head and neck squamous cell carcinomas (HNSCC) are characterized by EGFR overexpression.⁶⁵ c-Src is found as the key mediator of EGFR-promoted cell growth and invasion in HNSCC.⁶⁶ c-Src is overexpressed in cancer samples of patients with HNSCC.⁶⁷ Similar to other cancer types described above, c-Src inhibition in HNSCC with small molecule inhibitors or small interference RNAs exhibits impressive antitumor effects^58,65,68,69

Mechanisms of activation of SFKs in human cancers.

As described above, elevated expression of wild-type SFKs is associated most frequently with the development and progression of human cancers. SFK genes are rarely mutated in human malignancy.²³ Although exact mechanisms of SFK overexpression remain complex and elusive, increased expression of SFKs in human cancers by post-translational regulation is likely associated with their deregulated upstream regulators and signaling (such as aberrant expression of EGFR, HER2, c-Met).^1,18–23 In addition, there are other mechanisms by which SFKs are activated in human cancers. As indicated above, SHP-1 can dephosphphorylate Y530 of SFKs and activate these kinases. One study has found that SHP-1 (PTPN6) was not only overexpressed in ovarian epithelial carcinoma cell lines, but also in 10 of 11 invasive ovarian cancer tissue specimens (91%) examined.⁷⁰ Histone deacetylase (HDAC) inhibitors can repress c-Src transcription in cancer cell lines,³⁰ indicating that SFKs can be transcriptionally regulated in human cancers. In some cancer types, it is likely that multiple mechanisms such as deregulation of upstream regulators, associated kinases (including Csk and Fyb) or phosphatases, SH2/SH3 ligand interactions, expression of adaptor proteins and redox regulation may collectively activate SFKs.

Multiple Mechanisms of Paclitaxel Action

Over nearly two decades since the FDA first approved paclitaxel for treatment of ovarian cancer in 1992, paclitaxel alone or in combination with other chemotherapeutic agents has become a standard therapy for patients with ovarian cancer, breast cancer, NSCLC and Kaposi sarcoma.^71,72 Understanding the mechanisms by which paclitaxel inhibits and/or kills cancer cells and by which resistance to paclitaxel develops could lead to improve clinical applications of current paclitaxel-based therapy and possibly to develop novel microtubule-targeted drugs.

About 30 years ago, treatment of human HeLa and mouse fibroblast cells with cytotoxic paclitaxel was found to stabilize cytoplasmic microtubules and block cells in the M phase of the cell cycle.^73,74 Paclitaxel targets microtubules by binding to β-tubulin, which affects cell proliferation (mitosis), cell shape, intracellular trafficking of macromolecules and organelles, and motility.^71,75 Paclitaxel-induced M phase arrest is ultimately followed by induction of apoptosis and/or mitotic catastrophe.⁷⁶

Although the anticancer activity of paclitaxel is primarily attributed to this anti-mitotic activity, the effects of paclitaxel on non-mitotic events such as intracellular trafficking and motility have recently been emphasized and thought to play important roles in paclitaxel's anticancer activity.⁷⁷ A number of regulatory and structural proteins such as p53, BRCA1, Rb, androgen receptor and survivin have been shown to depend on microtubule trafficking for normal function in both dividing and interphase cells.⁷⁷ Suppression of microtubule-associated intracellular trafficking can lead to cell death.⁷⁷ Paclitaxel is also reported to inhibit sodium channel-dependent but cell cycle-independent cell motility in breast cancer cells.⁷⁸

The effect of paclitaxel on tumor angiogenesis and vascular structures contributes to the anticancer activity of the drug.^77,79,80 Paclitaxel inhibits the proliferation and motility of endothelial cells and hampers formation of new blood vessels in vivo.^77,79,80

Multiple Mechanisms of Paclitaxel Resistance

Primary paclitaxel resistance refers to the intrinsic resistance of cancer cells to an initial paclitaxel treatment, whereas secondary paclitaxel resistance is acquired only after an initial response to paclitaxel treatment.⁸¹ Both forms of resistance limit the use of paclitaxel to treat different cancers. Many mechanisms described below have been proposed and tested to account for primary and acquired paclitaxel resistance.^71,81,82

Deregulation of P-glycoprotein.

Intracellular levels of paclitaxel are determined by net cellular uptake and export of the drug. As a lipophilic molecule, paclitaxel is readily taken up by cancer cells and is exported by several trans-membrane ABC transporters. Among the 15 ABC transporters (ABCA2, ABCA3, ABCB1, ABCB4, ABCB5, ABCB11, ABCC1–6, ABCC11–12 and ABCG2) that have been linked to drug resistance,⁸³ only P-glycoprotein (Pgp or ABCB1), ABCC10 (MRP7) and ABCB11 (BSEP/PFIC2/SPGP) have been implicated in paclitaxel resistance in cancers to date.^71,83 Pgp overexpression in breast and ovarian cancers has been shown to be a poor prognostic factor that correlates inversely with response to paclitaxel.^84,85 Higher expression of ABCC10 in 17 NSCLC cell lines correlated more precisely with paclitaxel resistance than did levels of Pgp.⁸⁶ Increased ABCC10 expression has also been found in salivary gland adenocarcinoma cell lines that had developed docetaxel resistance.⁸⁷ Although the role of many other ABC transporters have not been evaluated extensively in paclitaxel resistance, multidrug resistance protein 1 (MRP1 or ABCC1) and breast cancer resistance protein (BCRP/MXR or ABCG2) seem to play a minimal role in resistance to taxanes.^88,89 Aside from overexpression of the transporters, approximately 50 SNPs in the Pgp gene have been identified which may affect function of the transport protein and export of paclitaxel, contributing to drug resistance.⁸³

Alteration of tubulin isotypes.

Increased expression of several tubulin isotypes, including class III β-tubulin, α-tubulin and class IVa-tubulins, has been associated with paclitaxel resistance.⁹⁰ Class III β-tubulin dimers depolymerize more readily than the constitutively expressed Class I and IVb β-tubulin isotypes in vitro.⁹¹ Overexpression of class III β-tubulin is associated with paclitaxel resistance in several cancer types including ovarian cancer.^{71,81,82,90,92} Importantly, class III β-tubulin knockdown with small interfering RNA (siRNA) or antisense silencing can sensitize cancer cells to paclitatexel^93–95 through suppression of microtubule dynamics at low drug concentrations and a mitosis-independent mechanism of apoptosis at higher drug concentrations.⁹⁴

Alteration of microtubule-regulatory proteins.

A number of proteins such as MAP4, MAP2, Tau, stathmin, fragile histidine triad protein (FHIT) and survivin, can directly interact with microtubule and regulate their polymerization and stability.^90,96 There are evidence suggesting that enhanced microtubule assembly and stability generally increases paclitaxel sensitivity, whereas decreased microtubule stability favors paclitaxel resistance. For example, Increased expression of MAP4 and E-MAP-115 (epithelial microtubule-associated protein of 115 kDa), microtubule stabilizers, has been demonstrated in vitro to increase sensitivity to paclitaxel treatment.^97,98 Stathmin promotes microtubule depolymerization and is frequently overexpressed in cancers and negatively regulates paclitaxel sensitivity.^99,100 Knockdown of stathmin using siRNA increases sensitivity to paclitaxel treatment.^99,100 Loss of TGFBI (transforming growth factor β induced), an extracellular matrix (ECM) protein, is reported to induce specific resistance to paclitaxel via integrin-dependent de-stabilization of microtubules in ovarian cancer cells.¹⁰¹

Survivin, a member of the inhibitors of apoptosis protein (IAP) family, plays dual roles in both mitosis and apoptosis.^102,103 Paclitaxel treatment rapidly induces survivin expression that is independent of cell cycle arrest, which serves as a mechanism for cancer cells to evade paclitaxel-induced apoptosis.¹⁰⁴ High levels of survivin protein expression has been found in more than 70% of advanced ovarian carcinomas and are associated with clinical resistance to a paclitaxel-based regimens, but not to non-paclitaxel-based regimens.¹⁰⁵ While survivin can stabilize microtubules during mitosis, survivin's ability to inhibit apoptosis appears to have a greater impact on interactions with paclitaxel. Downregulating survivin with either siRNA, anti-sense oligonucleotides, dominant negative mutants or ribozymes leads to greater paclitaxel-induced apoptosis and to increased paclitaxel sensitivity.^103,104,106 Enhancement of apoptosis by downregulating survivin appears to be regulated through inhibition of Smac and activation of caspases.¹⁰² On the other hand, increased survivin can induce drug resistance to microtubule de-stabilizers (i.e., vinca alkaloids and combretastatin A-4) due to its effect on microtubule stabilization, rather than by inhibiting apoptosis.¹⁰⁷

Tumor suppressor FHIT can bind to microtubules and exert its activity though inhibition of survivin and PI3K-AKT pathways.^108,109 Increased expression of FHIT can enhance paclitaxel-induced apoptosis.¹¹⁰ Lastly, Tau stabilizes microtubules its expression was initially found to correlate inversely with the response to paclitaxel treatment.¹¹¹ However, a recent National Surgical Adjuvant Breast and Bowel Project (NSABP)-B28 clinical trial found that there was no significant interaction between Tau expression and paclitaxel benefit.¹¹² Instead, this study revealed that higher Tau protein expression is associated with better disease-free survival (DFS) and overall survival (OS).¹¹²

Deregulation of apoptotic signaling pathways.

Acquired resistance to cell death is one of the hallmarks of all cancers.¹¹³ Several important apoptotic regulators such as p53, survivin and Bcl-2 have been shown to physically interact with microtubules and are thought to play roles in microtubule-associated function.^77,96 As discussed earlier, deregulation of some of these apoptotic regulators in cancers can alter microtubule assembly, microtubule stability or intracellular trafficking, leading to paclitaxel resistance. Although p53 mutation is thought to play only a marginal role in paclitaxel resistance,¹¹⁴ alteration of p53 expression and structure can affect paclitaxel sensitivity by regulating microtubule composition and dynamics.^77,96 Increased expression of the anti-apoptotic Bcl-2 protein confers paclitaxel resistance.¹¹⁵ Downregulation of the pro-apoptotic Bcl-2 antagonist killer 1 (Bak1) by miRNA125b results in paclitaxel resistance.¹¹⁶

Mutations of tubulins.

Mutations in class I β- or α-tubulin isotypes (for example, αSer379, βPhe270, βLeu215) have been documented in cancer cell lines that have been selected for resistance to paclitaxel in cell culture models.^117–121 Tubulin mutations may change microtubule assembly, dynamics and stability, leading to drug resistance.⁹⁰ However, very few, if any, tubulin mutations have been detected in clinical specimens.^90,122–124 Further studies are needed to determine whether tubulin mutations occur in patient-derived samples and if they do, whether these tubulin mutations are acquired during/after paclitaxel treatment in patients whose cancer's develop paclitaxel resistance. A recent study demonstrated in vitro that several β-tubulin mutations (A185T, A248V, R306C) that were identified in clinical specimens did confer paclitaxel resistance.¹²⁵

Overexpression of copper transporters.

Two mammalian copper efflux transporters ATP7A and ATP7B regulate copper metabolism, and have been shown in vitro and in vivo to mediate drug resistance to cis-platin.¹²⁶ ATP7A overexpression has recently been found to mediate drug resistance to paclitaxel as well.^127,128 Increased expression of ATP7B has been shown in ovarian cancer cell lines and colorectal cancer patients and associated with poor prognosis.^129,130

Substoichiometric levels of paclitaxel.

Given the known pharmacokinetics and pharmacodynamics of the taxanes, the concentration of the drug within cancer cells is almost always substoichiometric. Ovarian cancers may fail to respond to initial treatment with paclitaxel, not from primary resistance, but rather from a failure to achieve adequate levels of paclitaxel within cancer cells at a maximally tolerated dose of the drug (Ahmed AA, Shibani N, Bast RC Jr, Goldsmith J. Lack of response to taxane chemotherapy: substoichiometry or resistance? Submitted for publication). An optimal level of paclitaxel would occupy each of the binding pockets found on each molecule of β-tubulin. Under these circumstances, new approaches for paclitaxel therapy might attempt to enhance selectively the sensitivity of cancers to paclitaxel at first treatment, rather than attempt to overcome the many mechanisms of primary or acquired drug resistance. Against this hypothesis are older observations that low levels of paclitaxel can arrest living tumor cells at mitosis and induce apoptosis without a change in microtubule mass and microtubule bundling.^131–133

SFK Regulation of Paclitaxel Sensitivity and Resistance

Overexpression of SFKs in cancers, especially in ovarian cancer regulates a variety of cell functions (Fig. 1) that can impact on paclitaxel sensitivity and resistance. SFK inhibition can sensitize cancers cells to paclitaxel-based treatments by modulating cell survival and proliferation, autophagy, microtubule stability, motility, and/or tumor angiogenesis.^134–139

Modulation of apoptosis.

As described above, one of the major mechanisms of paclitaxel-induced cancer cell death is the induction of apoptosis following M phase arrest of the cell cycle. Our unpublished data and published data of others show that SFK inhibition with pharmacologic Src kinase inhibitors, siRNA knockdown or expression of a dominant-negative Src can potentiate paclitaxel-induced apoptosis by activating caspase-3.¹³⁴ SFK inhibition can induce cancer cell apoptosis to enhance paclitaxel cytotoxicity through several signaling molecules and pathways. SFK inhibition can downregulates PI3K-AKT,^25,106,136 STAT3/5,¹⁴⁰ Bcl-2,²⁵ Bcl-x_L¹⁴¹ and Mcl-1,¹⁴⁰ and induce apoptosis^140–142 SFK inhibition also downregulates expression of antiapoptotic survivin.¹⁰⁶ SFKs can phosphorylate tumor suppressor FHIT on tyrosine 114 residue and inactivate FHIT function.¹⁰⁸ Since FHIT can inhibit two downstream targets, survivin and AKT, inhibition of SFK is thought to activate FHIT and apoptosis.¹⁰⁸

Modulation of autophagy.

Autophagy can mediate both tumor cell survival and cell death. In the short term, release of amino acids and fatty acids by digestion of cellular organelles through autophagy provides substrates for the generation of ATP that permits cells to survive in a nutrient poor environment. When autophagy is prolonged, however, cells die from nonapoptotic type II programmed cell death.^143,144 Recent data indicate that SFK inhibition can induce significant autophagy.^25,145,146 Knockdown of two key autophagy regulators Beclin-1 and Atg12 expression with their respective siRNAs diminished SFK inhibitor dasatinib-induced autophagy and growth inhibition in ovarian cancer cells.²⁵ Induction of autophagy has recently been shown to enhance paclitaxel activity in cancers. ARHI, a Ras-related imprinted gene, inhibits cancer cell proliferation and motility and induce autophagy in ovarian and breast cancers.¹⁴⁷ ARHI re-expression induces autophagy and enhances the growth inhibitory effect of paclitaxel in vitro and in vivo.¹⁴⁸ Increased expression of the pro-autophagic Beclin-1 can enhance paclitaxel-induced apoptotic and autophagic cell death in cervical cancer cells.¹⁴⁹ In addition, SFK inhibition with dasatinib in combination treatment with temozolomide, an oral alkylating agent, also resulted in a significant increase in autophagic cell death in glioma cells.¹⁴⁵ Although the exact mechanisms by which SFK inhibition modulates autophagy are still not clear, inhibition of Bcl-2 expression²⁵ and AKT-mTOR-p70S6K pathway^25,106,136 by SFK inhibition appear important. Constitutive expression of AKT1 and Bcl-2 inhibited dasatinib-induced autophagy in ovarian cancer cells.²⁵

Modulation of microtubule stabilization.

Enhanced microtubule stability can increase paclitaxel sensitivity in cancer cells. Increased expression of MAP4,⁹⁸ E-MAP-115,⁹⁷ and TGFBI,¹⁰¹ decreased expression of stathmin^99,100 and survivin^103,104,106 have been shown to promote microtubule assembly and stability, and to increase sensitivity to paclitaxel treatment. Overexpression of tubulin binding cofactor C (TBCC), a protein for proper polymerization, sensitized breast cancer cells to paclitaxel treatment in a xenograft model.¹⁵⁰ Higher expression of Tau, another microtubules stabilizing protein, is a favorite prognostic biomarker in breast cancer.¹¹² SFKs, particularly Fyn have been shown to regulate microtubules polymerization and spindle stabilization.^151,152 Inhibition of SFKs with inhibitor PP2 had no effect on MDR1 expression but enhanced microtubule stabilization in paclitaxel-resistant human (CaOV3TaxR) and mouse (ID8TaxR) ovarian cancer cell lines, restoring paclitaxel sensitivity.¹³⁷ Microtubule-associated MAP2 has been shown to bind directly to the SH3 domain of c-Src and Fyn.^153,154 Fyn can phosphorylate MAP2 on tyrosine 67 and phosphorylated MAP2 can bind to the adaptor protein Grb2.¹⁵⁴ Fyn also physically associates with tubulins.^151,152 These data suggest that SFKs may directly and indirectly affect microtubules assembly and stabilization.

Modulation of tumor angiogenesis.

Activation of angiogenesis in tumors is another hallmark of all solid cancers.¹¹³ Increasing production of pro-angiogenic factors such as VEGF and IL-8 by cancer cells and the tumor microenvironment (e.g., pericytes) (Fig. 2) supports endothelial cell proliferation and migration and is one of most common mechanisms by which angiogenic switch is turned on persistently during tumor progression.¹¹³ Both human and murine endothelial cells express high levels of VEGF receptors (VEGFR1, VEGFR2 and VEGFR3).¹⁵⁵ In addition, pericytes can stabilize and drive maturation of blood vessels through activation of PDGF-PDGFR signaling.^113,156 Targeting breast cancer cells that overexpress HER2 with trastuzumab can significantly inhibit tumor vessel formation by decreasing production of VEGF and IL8.¹⁵⁷ Dual targeting of cancer cells (e.g., with trastuzumab) and endothelial cells (e.g., with VEGF-Trap) produces superior inhibition of tumor angiogenesis in a breast cancer model.¹⁵⁵ Dual targeting of pericytes (e.g., anti-PDGF aptamer AX102) and endothelial cells (bevacizumab) also results in better inhibition of tumor angiogenesis in an ovarian cancer model.¹⁵⁸

Figure 2.

Paracrine interaction and interventions for cancer cells, vascular endothelial cells and pericytes during tumor angiogenesis.

As shown in Figure 2, Inhibition of SFKs can block signaling produced by EGFR and HER2 in cancer cells, by VEGFR in endothelial cells, and by PDGFR in pericytes.^1,18,23 Inhibtion of SFKs results in inhibition of a number of downstream targets such as FAK, RhoA, cortactin, p130CAS, paxillin, tensin-3, STAT3, Ras-MAPK and PI3K-AKT,^{1,18–21,25,26} which produce multiple phenotypic changes including inhibition of cell migration/invasion, survival and angiogenesis. On the other hand, paclitaxel can also inhibit the proliferation and motility of endothelial cells, pericytes and cancer cells.^77,79,80 Therefore, the combination of SFK inhibition and paclitaxel treatment can target simultaneously at cancer cells, endothelial cells and pericytes, which likely results in better inhibition of tumor angiogenesis. Inactivation of pericytes with AX102 in combination with paclitaxel significantly reduces pericyte coverage on endothelial cells and causes significant tumor reduction.¹⁵⁸ Paclitaxel or Ixabepilone has been shown to greatly increase killing of endothelial cells, inhibit tumor angiogenesis and potentiate antitumor activity of bevacizumab and PDGFR/VEGFR inhibitor sunitinib in multiple animal models of breast, lung, colon and kidney cancers.¹⁵⁹

Potential Biomarkers that Predict SFK Inhibition-Enhanced Paclitaxel Activity

As described above, SFK inhibition can increase paclitaxel sensitivity in a number of cancer types in vitro and in vivo. Given the heterogeneity of cancer, however, not every cancer patient would benefit from treatment of a combination of an SFK inhibitor and paclitaxel. Therefore, identification of potential biomarkers that can predict SFK inhibition-enhanced paclitaxel activity is critical for personalizing treatment and maximizing the benefit of this combination.

Expression of biomarkers such as moesin, caveolin-1 and yesassociated protein-1 have been found to predict response to SFK inhibition with dasatinib treatment in the “triple-negative” breast cancer.^160,161 Increased expression of Yes, Lyn, Eph2A, caveolin-1, moesin, annexin-1 and uPA also confer increased sensitivity to dasatinib treatment in ovarian cancer.¹⁴² Urokinase-type plasminogen activator (uPA) and EphA2 are identified as the biomarkers that could predict dasatinib response in prostate cancer.¹⁶² Tyrosine phosphorylation of Y419 c-Src and Y118 paxillin has been indicated as the biomarkers for response to dasatinib in colon cancer.¹⁶³ Among above mentioned biomarkers, cavelolin-1 is a particularly interesting biomarker that was confirmed in multiple cancer types. Caveolin 1 can interact with SFKs on membrance (Fig. 1) and inhibits protein kinase C phosphorylation on c-Src.²⁴ Increased expression of phosphorylated caveolin 1 can sensitize MCF7 breast cancer cells to apoptosis.¹⁶⁴ Overexpression of caveolin 1 inhibits c-Src expression and kinase activity, in vitro migration and in vivo metastatic potential.¹⁶⁵

Specific biomarkers have not yet been identified that predict the ability of SFK inhibitors to potentiate paclitaxel therapy. Dasatinib response biomarkers such as caveolin 1, EphA2 and uPA are reasonable candidates for further evaluation. Indeed, EphA2 inhibition with siRNA in couple with paclitaxel treatment dramatically decreases tumor growth in an orthotopic ovarian cancer mouse model associated with decreased levels of VEGF and phosphorylation of c-Src.¹⁶⁶ Combined siRNA targeting FAK (a critical downstream target of SFKs) and EphA2 have also been shown to significantly decrease tumor cell proliferation and tumor angiogenesis.¹⁶⁷ Our unpublished data indicate that low expression of EphB2 is a potential biomarker to predict the potential of SFK inhibitor dasatinib to increase paclitaxel treatment in ovarian cancer cell lines.

Conclusions

SFKs play important roles in regulating a broad range of cellular functions. Abnormally high expression of SFKs has been documented in a number of different malignancies including ovarian, breast, lung and head and neck cancers, which respond to treatment with paclitaxel. Inhibition of SFKs alone or in combination with paclitaxel treatment is currently being evaluated in clinical trials. SFK inhibition also represents a new approach to decrease paclitaxel resistance that often develops from a variety of mechanisms during treatment with taxanes. Several precedents exist for enhancing sensitivity to paclitaxel treatment by inhibiting SFKs in cell culture and xenograft models. A phase I clinical trial in metastatic breast cancer has recently found that dasatinib plus weekly paclitaxel treatment produced a partial response in four of 13 patients (31%) including two patients previously treated with taxanes.¹⁶⁸ Similar studies might be extended to many other types of cancer that are partially responsive to treatment with taxanes. Mechanisms underlying the interaction of SFK inhibition and enhanced paclitaxel sensitivity include potentiating apoptosis, inducing autophagic death, enhancing microtubule stability and targeting tumor neovasculature. In order to circumvent the development of paclitaxel resistance, inhibition of SFKs with its inhibitors such as dasatinib should also been tested in combination with novel taxanes such as the Epothilones,⁸¹ Larotaxel (RPR 109881A),¹⁶⁹ and nanoparticle albumin-bound (nab)-paclitaxel.¹⁷⁰ In addition, identification of the biomarkers that can predict SFK inhibition-enhanced paclitaxel effect would permit personalization of combined therapy with SFK inhibition and paclitaxel.

Abbreviations

c-Src

avian v-src sarcoma (Schmidt-Ruppin A-2) viral oncogene homolog

Fyn

FYN oncogene related to SRC, FGR, YES

Yes

v-yes-1 yamaguchi sarcoma viral oncogene homolog 1

Blk

B lymphocyte kinase

Fgr

gardner-rasheed feline sarcoma viral oncogene homolog

Hck

hematopoietic cell kinase

Lck

lymphocyte-specific protein tyrosine kinase

Lyn

v-yes-1 yamaguchi sarcoma viral related oncogene homolog

Yrk

YES related kinase

Csk

c-src tyrosine kinase

Fyb

FYN binding protein

FAK

focal adhesion kinase

GPCR

G-protein coupled receptor

STAT

signal transducers and activator of transcription

ER

estrogen receptor

AR

androgen receptor

EGFR

epidermal growth factor receptor

HER2

human epidermal growth factor receptor 2

TCR

T cell receptor

Ig

immunoglobulin

PDGFR

platelet-derived growth factor receptor

IGF-1R

insulin-like growth factor-1 receptor

VEGFR

vascular endothelial growth factor receptor

Cadherin

calcium-dependent adhesion

ERK

extracellular signal-regulated kinases

MAPK

mitogen-activated protein kinases

PTP1B/α/γ/ε

protein-tyrosine phosphatase1B/alpha/lambda/epsilon

SHP1

hematopoietic cell protein-tyrosine phosphatase/protein-tyrosine phosphatase 1C

SHP2

protein-tyrosine phosphatase 2C

MMP

matrix metalloprotease

Csk

C-terminal Src kinase

Cbp

Csk binding protein

PAG

phosphoprotein associated with glycosphingolipid-enriched microdomains

Cbl

Cas-Br-M (murine) oncogene

LIME

Lck-interacting membrane protein

LAT

linker for activation of T cells

NTAL

non-T cell activation linker

ROS

reactive oxygen species

TGFβ1

transforming growth factor-beta1

uPA

urokinase-type plasminogen activator

PAI-1

plasminogen activator inhibitor type-1

Smad2/3

SMAD (the drosophila gene ‘mothers against decapentaplegic’ (Mad) and the C. elegans gene Sma) family member 2/3

PELP1

proline-, glutamic acid- and leucine-rich protein-1

LPA

lysophosphatidic acid