- Email: [email protected]
Complexities of lysophospholipid signalling in glioblastoma
Journal of Clinical Neuroscience 21 (2014) 893–898
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Review
Complexities of lysophospholipid signalling in glioblastoma Wayne Ng a,b,c,f, Alice Pébay a,d, Katharine Drummond a,b,c,f, Antony Burgess a,e, Andrew H. Kaye a,b,c, Andrew Morokoff a,b,c,⇑ a
University of Melbourne, Parkville, VIC, Australia Department of Surgery, Royal Melbourne Hospital, Grattan Street, Parkville, VIC 3050, Australia c Department of Neurosurgery, Royal Melbourne Hospital, Parkville, VIC, Australia d Centre for Eye Research Australia, Royal Victorian Eye and Ear Hospital, East Melbourne, VIC, Australia e Walter and Eliza Hall Institute of Medical Research, Parkville, VIC, Australia f Melbourne Brain Centre at Royal Melbourne Hospital, Parkville, VIC, Australia b
a r t i c l e
i n f o
Article history: Received 12 January 2014 Accepted 22 February 2014
Keywords: Autotaxin Glioblastoma multiforme Lysophosphatidic acid
a b s t r a c t Glioblastoma multiforme (GBM) is the most malignant brain tumour and continues to have a very poor median survival of 12–16 months despite current best therapies. These aggressive tumours always recur after treatment and are defined by their ability to diffusely infiltrate and invade normal brain parenchyma. Autotaxin is overexpressed in GBM, and is a potent chemotactic enzyme that produces lysophosphatidic acid. Lysophospholipid (LPL) signalling is known to increase invasion of solid tumours and is also dysregulated in GBM. The LPL pathway has been shown to interact with known cancer-related signalling pathways, including those for epidermal growth factor and yes-associated protein, which are also dysregulated in GBM. The interactions between these pathways provide insights into the complexities of cancer signalling and suggest potential novel targets for GBM. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Glioblastoma multiforme (GBM) is the most malignant (World Health Organization grade IV) glioma and continues to have a very poor median survival of 12–16 months despite current best therapies (maximal safe resection with concurrent temozolomide chemotherapy and radiation therapy) [1]. These aggressive tumours always recur after treatment and are defined by their ability to diffusely infiltrate and invade normal brain parenchyma. Thus the search for targeted agents inhibiting cell proliferation, survival and invasion has intensified. Research into the epidermal growth factor receptor (EGFR) pathway has led to clinical trials of EGFR and phosphatidylinositol-3-kinase (PI3K) inhibitors that modulate cell survival and proliferation in pre-clinical models. However, results from early EGFR inhibitor trials have not delivered on their promise and PI3K inhibitor trials are ongoing [2–4]. These trials have made it obvious that strategies combining therapies against multiple targets are required to account for the existence of complex pathway interactions and redundancies. For example, matrix metalloproteases (MMP) degrade extracellular matrix components to produce more favourable conditions for cell migration ⇑ Corresponding author. Tel.: +61 3 9035 8586 E-mail address: [email protected] (A. Morokoff). http://dx.doi.org/10.1016/j.jocn.2014.02.013 0967-5868/Ó 2014 Elsevier Ltd. All rights reserved.
and invasion [5]. However, they also cleave and activate growth factors such as epidermal growth factor (EGF) [6]. The function of integrins and their influence on cell morphology and migration has also been enlightening [7,8]. More recently, other promising factors have been identified, including autocrine motility factor receptor, heparin-binding epidermal growth factor, ephrin-B3, netrin 4 and autotaxin (ATX) [9]. The latter three are of interest because their role in cell migration and motility in neural stem cells suggests a similar role in glioma-derived cancer stem cells, that have a putative role in GBM progression [10,11]. ATX in particular is a powerful chemotactic enzyme involved in lysophospholipid (LPL) signalling, and its recent prominence in the literature has highlighted the importance of lipid signalling within complex intracellular pathway interactions. This review focuses on the role that LPL signalling may play in gliomagenesis and its potential as a target in the treatment of this highly malignant disease. 2. Lysophosphatidic acid Lysophosphatidic acid (LPA) and sphingosine-1-phosphate (S1P) are the main membrane-derived lipid signalling molecules. LPA has a 3-carbon glycerol backbone, with an attached single acyl or alkyl chain of varying length which imparts some differences in receptor efficacy [12]. Whilst some LPA production may occur
894
W. Ng et al. / Journal of Clinical Neuroscience 21 (2014) 893–898
intracellularly, much of it is produced extracellularly by secreted enzymes. There are three known pathways: (1) cleavage of LPL (such as lysophosphatidylcholine) by lysophospholipase D, (2) deacylation of phosphatidic acid by phospholipase A1 and A2, and (3) mild oxidation of low-density lipoprotein (non-enzymatic). Lysophospholipase D is now more commonly known as ATX which derives its name from early characterisation of its stimulatory effect on melanoma cell motility [13–15]. It is also known as ectonucleotide pyrophosphatase and phosphodiesterase-2 (ENPP-2) and is part of the family of ENPP enzymes which are traditionally known for their involvement in nucleotide metabolism [13]. LPA production is mainly via ATX catalysis. This is confirmed by the uniformly fatal outcome of homozygous ATX (Atx/) knockout mice and the 50% reduction in circulating plasma LPA levels in embryos with heterozygous ATX (Atx+/) expression [16,17]. The phospholipase A enzymes play an important role in determining the position of the acylation of the phosphoglycerol backbone and so, whilst there is currently no evidence of its upregulation, it may still be an important pathway in pathological states, as it may influence the action of liberated LPA species by playing a role in determining the dominant LPA species produced (acyl or alkyl) [18]. Physiologic LPA is present in small amounts in tissues since it regulates its own production via negative feedback inhibition of its main synthesising enzyme, ATX [19]. Thus, the influence of this signalling pathway is largely autocrinic/paracrinic [12–14]. Further, circulating platelets and erythrocytes secrete LPA which is then bound to albumin to protect it from rapid enzymatic degradation by lysophospholipases, lipid phosphate phosphatases and LPA acyl transferases [12]. Cancer cells can produce increased amounts of ATX/LPA and it has been previously reported that some malignant effusions (such as those of ovarian cancer) have elevated levels of LPA compared to other malignancies [2,20,21]. In addition, circulating plasma levels of LPA can be elevated in malignancy and this may be the result of elevated production (induction of ATX) [14]. 3. LPA receptors Only a brief review of the LPA receptors will be provided here as there are many other detailed reviews available [12,16,22–24]. At the time of writing, the International Union of Basic and Clinical Pharmacology had recognised six definitive G-protein coupledLPA receptors (collectively LPAR) designated LPA1–6 (Table 1). Broadly, the receptors fall into two families: endothelial differentiation gene (Edg) and non-Edg (purinergic) receptors [12]. LPA1 (Edg2), LPA2 (Edg4) and LPA3 (Edg7) are members of the Edg family
and are the best characterised to our knowledge with LPA1 being the dominant LPA receptor in the central nervous system (CNS) [13,23,25]. LPA1 is coupled to the G-proteins, Gi/o, Gq and G12/13, which allows it to signal via multiple pathways, including major cancer-related pathways such as mitogen-activated protein kinase (MAPK), Akt/PKB, and small GTPases such as Rho/ROCK [12,26]. LPA1-induced PI3K signalling (which activates Akt/PKB) via the p110b/c subunits (of PI3K) has also been reported [27–30]. Whilst LPA2 is expressed in embryonic brain it has little to no expression in the adult CNS [31]. LPA3 is expressed in the brain, but unlike LPA1 and LPA2 it is coupled to the G-proteins, Gi/o and Gq, but not G12/13 and so is less responsive to LPL than LPA1 [27]. As LPA3 does not signal via G12/13, it is not involved in changes in cell morphology, and therefore unlikely to be involved in cell migration [27]. The non-Edg or purinergic family of LPA receptors are genetically distinct from the original Edg family of LPA receptors [23]. Importantly, this results in the non-Edg family having an increased affinity for alkyl-LPA species, as opposed to the Edg family having increased affinity for the acyl variants [22,23]. Current members of the non-Edg family include LPA4 (P2Y9), LPA5 (GPR92) and LPA6 (P2Y5). LPA4 signals via Gs, Gq and G12/13-proteins and probably plays a role in cell motility and migration [23]. Mouse embryonic fibroblasts (MEF) derived from Lpa4 (null) knockout mice have been reported to exhibit hypersensitivity to LPA induced motility. These MEF had increased levels of phosphorylated Akt (pAkt) when stimulated by LPA. The elevated pAkt levels and associated motile response were attenuated when Lpa4 was reintroduced into the cells, suggesting that LPA4 has an action which might suppress signalling activity of the LPA1 receptor [32]. LPA5 (GPR92) and LPA6 (P2Y5) were both discovered subsequent to LPA4 and to our knowledge there is no reported role for LPA5 in tumourigenesis. LPA6 however, may cause morphological changes in vascular endothelial cells and therefore may play a minor or indirect role in gliomagenesis via effects on vascular development [22,23,27]. LPL have a wide range of physiological and pathological effects owing to the myriad G-proteins they are coupled to. As a result, they modulate numerous normal physiologic processes, including cell proliferation and apoptosis, cell differentiation, cell adhesion and migration, cell morphology (including neurite retraction and synaptic cleft shape modulation), normal CNS development and autoimmunity [12,13,16,18,25,33–35]. Dysregulation of these events is strongly implicated in tumourigenesis. 4. LPA signalling in cancer Since Stracke et al. discovered the promotile effects of ATX on melanoma cells in 1992, the LPA pathway has been investigated
Table 1 LPA receptors and their relevance to cancer research LPA receptor
Coupled G-proteins
Relevance to cancer research
Reference
Edg
LPA1
Gi/o, Gq, G12/13
Kishi et al. [2]
LPA2
Gi/o, Gq, G12/13
LPA3
Gi/o, Gq
– – – – – – –
LPA4
Gs, Gq, G12/13
LPA5
Gq, G12/13 Increase in cAMP not shown to be Gs-mediated G12/13, (possible role for Gs, Gi)
Non-Edg
LPA6
Dominant LPA receptor expressed in adult CNS Increased expression in cancers including GBM Linked to cell proliferation, survival, migration/invasion Not present in adult CNS Not markedly elevated in GBM Expressed in adult CNS Not markedly elevated in GBM
Kishi et al. [2] Kishi et al. [2], Choi et al. [16]
– Activation of LPA4 has been shown to be functionally antagonistic (anti-migration, anti-proliferation) to LPA1 – Activation has been shown to reduce cell migration
Kato [59], Yanagida and Ishii [23], Lee et al. [32]
– Possible similar function to LPA4 – May regulate vascular permeability
Yanagida and Ishii [23]
Jongsma et al. [54]
cAMP = cyclic adenosine monophosphate, CNS = central nervous system, Edg = endothelial differentiation gene, GBM = glioblastoma multiforme, LPA = lysophosphatidic acid.
W. Ng et al. / Journal of Clinical Neuroscience 21 (2014) 893–898
for its role in tumour invasion and metastasis [36]. However, its original role in nucleotide metabolism could not be directly reconciled with this biological effect. It was the subsequent discovery by Umezu-Goto et al. [14] and Tokumura et al. [15] that ATX had additional lysophospholipase D activity that triggered further interest in LPA. Recent elucidation of the complex interactions of LPA with known cancer signalling pathways will be discussed below (Fig. 1). 4.1. LPA interacts with cell survival pathways EGFR signalling is commonly upregulated in GBM cells either via over-expression, EGFR gene amplification, or via the EGFR type III mutation, which produces a constitutively active form of EGFR [37,38]. EGFR activates downstream PI3K signalling which regulates cell survival via abrogation of apoptotic cell signals and this is dysregulated in glioma. However, clinical trials with EGFR monoclonal antibodies [3] were less promising than anticipated, possibly because of intracellular pathway cross-communication and redundancy [28,39]. It has been reported that LPA can intracellularly transactivate EGFR via the Gi-protein and that this is independent of normal ligand activation of EGFR by EGF [28–30,40]. Further, it is possible that transactivation of p110b/c subunits of PI3K via Gi-protein coupled LPAR is more important than transactivation of p110a [41]. This LPA-Gi-protein coupled signalling link has been reported in many cancer cell lines including those from prostate, breast, colon and GBM [28,39]. Recently, Schleicher et al. reported that LPA signalling confers radioresistance to the murine glioma cell line GL-261 [42]. Inhibition of LPA signalling with a combination ATX/LPAR inhibitor, a-bromomethylene phosphonate lysophosphatidic acid (BrP-LPA), resulted in impaired survival signals in this cell line when exposed to 3 Gy of irradiation. Knockdown of LPA1 and LPA3 (but not LPA2) signalling using siRNA resulted in reduced pAkt and also correlated with reduced survival in these cells. These results point to the hypothesis that LPA1 and LPA3 play an important role in radioresistance, via interaction with the PI3K/Akt pathway.
895
4.2. LPA signalling in cell proliferation In addition to enhancing cell survival, transactivation of EGFR by LPA also upregulates cell proliferation via MAPK dependent mitogenic signalling. Whilst the transactivation of PI3K signalling can be EGF independent, the transactivation of MAPK appears to be dependent on agonistic activation of EGFR. Further, MAPK transactivation may be linked to activation of PI3K [43]. Its transactivation is also cell type dependent and has also been confirmed to be G-protein dependent [29,30,39,43]. Overall, transactivation of MAPK by LPA appears to be more regulated (than PI3K) and in fact may also require (or be augmented by) concurrent MMP activity, which is also upregulated in gliomas [5,6].
4.3. LPL and ATX signalling influences cancer cell migration and invasion Alterations in cell morphology are tightly regulated and play an important role in terms of chemotactic effect [44]. It is likely that ATX and LPA play their roles in glioma migration and invasion via effects on cell morphology. Overexpression of ATX also correlates with increased invasiveness of breast cancer cells compared to normal breast cells [45]. LPAR overexpression has also been reported in ovarian cancers and this has been postulated to create an autocrine/paracrine loop promoting proliferation (MAPK signalling) and suppressing apoptosis (PI3K signalling) [6,39,43]. Treatment of animals with metastatic ovarian cancer with Ki16425 (selective LPA1/3 receptor antagonist) was demonstrated to have an in vivo anti-tumour effect without obvious adverse health effects on the treated nu/nu mice [46]. Also, use of a pan-LPAR antagonist to treat hepatic metastases in an orthotopic breast cancer model showed significant reduction in tumour deposit size [19]. LPA4 has been shown to promote invadopodia formation in HT1080 fibrosarcoma cells, with Lpa4 shRNA abolishing this effect [47]. Invadopodia are involved in invasion and metastasis and are enriched with MMP, which are known for their key role in invasion and metastasis by providing a favourable milieu for cells to invade
Fig. 1. Overview of intracellular interactions between lysophosphatidic acid (LPA) and epidermal growth factor receptor (EGFR) signalling pathways. Mutations leading to overexpression or elevated activity of autotaxin, LPA1–3, EGFR and phosphoinositol-3-kinase are often found in glioblastoma multiforme. ATX = autotaxin, EGF = epidermal growth factor, EGFR = epidermal growth factor receptor, ERK = extracellular signal-regulated kinases, LPA = lysophosphatidic acid, LPC = lysophosphatidylcholine, MAPK = mitogen-activated phosphokinase, mTOR = mammalian target of rapamycin, PIP2 = phosphatidylinositol 4,5-bisphosphate, PIP3 = phosphatidylinositol (3,4,5)trisphosphate, PI3K = phosphoinositol-3-kinase, PTEN = phosphatase-tensin homologue, ROCK = Rho-associated kinase.
896
W. Ng et al. / Journal of Clinical Neuroscience 21 (2014) 893–898
through and also cleave and activate growth factors such as EGF [5,6]. In glioma, LPA has been shown to induce astrocyte motility by dramatically altering astrocyte shape [24,33,48]. ATX overexpression has also been shown to enhance the invasion of U87 and U251 malignant glioma cells (autocrine effect) through oligodendrocyte monolayers in vitro, whilst simultaneously reducing the adhesiveness of oligodendrocytes (paracrine effect). Depletion or inactivation of ATX ameliorates these effects. These findings may provide clues to the underlying pathogenesis of in vivo white matter tract invasion of glioma cells [9]. Overexpression of ATX has been shown to be proportionately higher at the leading invasive edges of tumours [9]. There are currently no definitive reports to address whether the overexpression of ATX and upregulation of LPAR is specific to certain cell types within the tumour. But microglia may be recruited to the invasive edge of the tumour by various chemokines (including LPA) and subsequently facilitate invasion by producing ATX, LPA and EGF [49]. The dominant LPA receptors involved in mediating cell migration and motility appear to be LPA1 and LPA4. LPA1 receptors have been linked to tumour cell motility in vitro and have been reported to be overexpressed in GBM cell lines (SNB-78, SNB-75, SF-268, SF539 and SF-295) in combination with overexpression of ATX. This receptor/ligand dual overexpression may contribute to autocrinic stimulation of GBM cell motility and this can be abolished when the cells are treated with the LPA1-3 receptor antagonist, Ki16425 [2]. Early work on the cellular mechanisms of LPA induced motility revealed a link to Rho signalling through either Gi (pertussis toxin sensitive) or G12/13-protein [13,50,51]. The subsequent Rhodependent cytoskeletal rearrangement was reported to induce cell rounding in astrocytes and this may contribute to producing discohesive cells [51]. However, there have been some contradictory reports with regards to Rho-related LPA signalling. For example, LPA induced glioma cell migration can be inhibited by blocking Rho activation. Contrary to this, stimulation of Rho activity has also been demonstrated to cause immobilisation of glioma cells [52]. Recall the different evolutionary pathways of the Edg and non-Edg families of LPAR and there may exist functional antagonism between them to provide an explanation for these inconsistencies. Lee et al. reported that LPA4 is likely a functional antagonist of LPA1, leading to a reduction in cell motility in MEF cells as well as DLD1 colorectal cancer cell lines and B103 neuronal cell lines [32]. It is possible that if LPA4 antagonises LPA1, that LPA4
somehow modulates activation of LPA1 or its signalling. Interestingly, activation of LPA5 has recently been reported to inhibit B16 melanoma cell migration [53]. This further supports the notion that the non-Edg (purinergic; LPA4–6) family of LPA receptors may modulate the effects of the Edg (LPA1–3) family. In addition to the reported Rho-dependent mechanisms, LPAmediated cell migration is also thought to occur via MAPK and PI3K signalling. LPA1 has been shown to activate MAPK and PI3K/ Akt signalling to promote cell migration [16,54]. Also, it has been confirmed that ATX induced motility in melanoma cells is mediated via the p110c subunit (of PI3K) and that PI3K inhibitor can inhibit this motile response in a dose-dependent manner [41]. Further, Kim et al. reported that simultaneous knockout of Akt1 and Akt2 abolished LPA induced motility in MEF [54]. Only reexpression of Akt1 following the double knockout restored the motile response to LPA. They also established that exposure of MEF to the pan-PI3K inhibitor (LY294002) or the LPAR antagonist, Ki16425, also completely abolished the motile response. This would make it likely that the downstream activation of Akt1 is via PI3K activation. However, it is not completely clear whether the subsequent phosphorylation of Akt is via transactivation of EGFR, or via interaction of LPAR associated Gi-protein with PI3K. 4.4. The potential role of LPA in angiogenesis LPA signalling seems to have a minor physiological role in angiogenesis, as S1P normally maintains this process. However, in pathological states, the role of LPA in angiogenesis may become more important [12]. Lpa(/) null mice have been shown to have 1 a small but significant incidence (2.5%) of frontal intracerebral haemorrhages which increases substantially in Lpa(/) /Lpa(/) 1 2 double-null mice to 26%. This may support a role for LPA in vessel maturation [12]. Furthermore, Atx(/) null murine embryos have a profound absence of vessel maturation that causes uniform embryonic mortality by day 10 [16]. This points to a potential role for LPA signalling in neoangiogenesis. In vivo murine models investigating ATX have shown that ATX plays a role in angiogenesis similar to that of vascular endothelial growth factor [55]. Other in vivo studies investigating knockouts of Lpa1–4 have not had the profound effect of Atx(/) knockout on vascular development, but instead, G12/13-protein knockouts have been found to have effects which parallel those of ATX. Therefore, it is possible that the LPAR may play redundant roles in neoangiogenesis, or the effect ATX has on vascular development may be mediated through other pathways such as vascular endothelial growth factor [23].
Table 2 Evidence for the potential therapeutic effect of lysophosphatidic acid and autotaxin modulation Compound
Mechanism
Animal
Cell line (s)
Cancer
Effect
Reference
Ki16425
Selective LPA1/3 antagonist
Balb-c nu/nu mice
Selective ATX inhibitor
NA
Ovarian Breast Melanoma
NSC 48300 bithionol
Small molecule ATX inhibitors Dual pan-LPA antagonist/ ATX inhibitor
NA
A2058
Melanoma
Balb-c nu/nu mice
MDA-MB-231
Breast
MDA-MB-231 A-549 HCT-116 Mouse GL-261
Breast Lung Colon GBM
In vivo – regression of bone metastases In vitro – reduction in LPA production and subsequent cell migration and invasion In vitro – reduced cell migration and invasion In vitro – reduced cell migration and invasion In vivo – tumour regression In vivo – tumour growth delay
Boucharaba et al. [46]
Cyclic phosphatidic acid analogs
CHOb3wt MDA-BO2 A2058
BrP-LPA
BrP-LPA
Dual pan-LPA antagonist/ ATX inhibitor
Balb-c nu/nu mice
BrP-LPA
Dual pan-LPA antagonist/ ATX inhibitor
Balb-c nu/nu mice
In vitro – radioprotection of glioma cell lines In vivo – tumour growth delay
ATX = autotaxin, GBM = glioblastoma multiforme, BrP = a-bromomethylene phosphonate, LPA = lysophosphatidic acid, NA = not applicable.
Baker et al. [58]
Saunders et al. [59] Zhang et al. [57]
Xu et al. [60]
Schleicher et al. [42]
W. Ng et al. / Journal of Clinical Neuroscience 21 (2014) 893–898
Recent evidence also suggests that LPA may bind intracellular targets such as peroxisome proliferator-activated receptor-c (PPAR-c) to cause neointimal formation [27]. It is unclear as to whether LPA migrates across the cell membrane but the effects of LPA induced PPAR-c activity are significantly diminished in Lpa(/) knockout mice. Regardless, specific unsaturated acyl LPA 1 interact with PPAR-c to induce neointimal formation and are likely produced by phosphorylation of monoacylglycerols (mitochondrial acylglycerol kinase) or by conversion of phosphatidic acid to LPA by cytosolic phospholipase A2 [18,27,56]. The role of LPA and ATX in angiogenesis no doubt complements their putative roles in tumour growth and invasion/metastasis. 5. Discussion Decades of glioma research are now seeing the beginnings of molecular and genetic profiling of gliomas. Additionally, gliomaderived cancer stem cells have been identified, providing a new framework to investigate this disease’s resistance to treatment and subsequent recurrence. Much of the evidence for the role of LPA in cancer is spread across different cell lineages making it difficult to extrapolate some of the knowledge to gliomas. However, the emerging reports regarding LPA and gliomagenesis for the most part echo the findings in other solid tumours. The discovery of ATX’s chemotactic effects has brought the spotlight to this field, resulting in invaluable insights into the complexities of intracellular crosscommunication and pathway redundancies. It has also lead to preclinical trials demonstrating the potential of this pathway as a target in malignancies (Table 2). Recently, the potential benefits of targeting ATX-LPA signalling were shown in murine breast cancer models using a combination ATX inhibitor and LPAR antagonist. This treatment reduced in vitro migration of breast cancer cells and caused tumour regression in vivo [57]. This strategy is based on the theory that the Edg (LPA1–3) receptors may be the main procarcinogenic mediators of cancer in dysregulated LPA signalling. An alternative strategy to targeting LPA signalling might look to produce agonists to the non-Edg (LPA4/5) receptor family, as there is mounting evidence that they might mitigate the pro-carcinogenic effects of the Edg receptors. This strategy may have benefits of being more selective. However, the relative underexpression or absence of these receptors in some tumours may limit its value. LPA1 has been shown to be the dominant LPAR expressed in gliomas but some glioma cell lines have also been shown to express moderate levels of LPA4 [2]. Therefore, at least in some gliomas, there may be merit to developing LPA4 agonists as an anti-cancer therapy. Also, an investigation into expression patterns of LPAR in different cell subpopulations (glioma-derived cancer stem cells, progenitor cells, and recruited host cells such as microglia and reactive astrocytes and oligodendroglia) within the glioma microenvironment is critical to ascertain the true value of LPA modulation as a treatment strategy. Conflicts of Interest/Disclosures The authors declare that they have no financial or other conflicts of interest in relation to this research and its publication. Acknowledgements This study was funded by the Neurosurgical Society of Australasia/LifeHealthCare scholarship (2012) and the Brain Foundation of Australia Brain Tumours Award (2012) and the Victorian State Government’s Department of Innovation, Industry and Regional Development’s Operational Infrastructure Support Program. AP is supported by a National Health and Medical Research Council Career Development Fellowship.
897
References [1] Stupp R, Hegi ME, Mason WP, et al. Effects of radiotherapy with concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the EORTCNCIC trial. Lancet Oncol 2009;10:459–66. [2] Kishi Y, Okudaira S, Tanaka M, et al. Autotaxin is overexpressed in glioblastoma multiforme and contributes to cell motility of glioblastoma by converting lysophosphatidylcholine to lysophosphatidic acid. J Biol Chem 2006;281: 17492–500. [3] Camara-Quintana JQ, Nitta RT, Li G. Pathology: commonly monitored glioblastoma markers: EFGR, EGFRvIII, PTEN, and MGMT. Neurosurg Clin N Am 2012;23:237–46. [4] Wen PY, Lee EQ, Reardon DA, et al. Current clinical development of PI3K pathway inhibitors in glioblastoma. Neuro Oncol 2012;14:819–29. [5] Pullen NA, Anand M, Cooper PS, et al. Matrix metalloproteinase-1 expression enhances tumorigenicity as well as tumor-related angiogenesis and is inversely associated with TIMP-4 expression in a model of glioblastoma. J Neurooncol 2012;106:461–71. [6] Gschwind A, Prenzel N, Ullrich A. Lysophosphatidic acid-induced squamous cell carcinoma cell proliferation and motility involves epidermal growth factor receptor signal transactivation. Cancer Res 2002;62:6329–36. [7] Tchaicha JH, Reyes SB, Shin J, et al. Glioblastoma angiogenesis and tumor cell invasiveness are differentially regulated by b8 integrin. Cancer Res 2011;71: 6371–81. [8] D’Abaco GM, Kaye AH. Integrins: molecular determinants of glioma invasion. J Clin Neurosci 2007;14:1041–8. [9] Hoelzinger DB, Nakada M, Demuth T, et al. Autotaxin: a secreted autocrine/ paracrine factor that promotes glioma invasion. J Neurooncol 2007;86: 297–309. [10] Singh SK, Clarke ID, Terasaki M, et al. Identification of a cancer stem cell in human brain tumors. Cancer Res 2003;63:5821–8. [11] Hemmati HD, Nakano I, Lazareff JA, et al. Cancerous stem cells can arise from pediatric brain tumors. Proc Natl Acad Sci USA 2003;100:15178–83. [12] Ishii I, Fukushima N, Ye X, et al. Lysophospholipid receptors: signaling and biology. Annu Rev Biochem 2004;73:321–54. [13] Moolenaar WH. Lysophospholipids in the limelight: autotaxin takes center stage. J Cell Biol 2002;158:197–9. [14] Umezu-Goto M, Kishi Y, Taira A, et al. Autotaxin has lysophospholipase D activity leading to tumor cell growth and motility by lysophosphatidic acid production. J Cell Biol 2002;158:227–33. [15] Tokumura A, Majima E, Kariya Y, et al. Identification of human plasma lysophospholipase D, a lysophosphatidic acid-producing enzyme, as autotaxin, a multifunctional phosphodiesterase. J Biol Chem 2002;277:39436–42. [16] Choi JW, Lee CW, Chun J. Biological roles of lysophospholipid receptors revealed by genetic null mice: an update. Biochim Biophys Acta 2008;1781: 531–9. [17] Tanaka M, Okudaira S, Kishi Y, et al. Autotaxin stabilizes blood vessels and is required for embryonic vasculature by producing lysophosphatidic acid. J Biol Chem 2006;281:25822–30. [18] Aoki J, Inoue A, Okudaira S. Two pathways for lysophosphatidic acid production. Biochim Biophys Acta 2008;1781:513–8. [19] Prestwich GD, Gajewiak J, Zhang H, et al. Phosphatase-resistant analogues of lysophosphatidic acid: agonists promote healing, antagonists and autotaxin inhibitors treat cancer. Biochim Biophys Acta 2008;1781:588–94. [20] Baker DL, Morrison P, Miller B, et al. Plasma lysophosphatidic acid concentration and ovarian cancer. JAMA 2002;287:3081–2. [21] Westermann AM, Havik E, Postma FR, et al. Malignant effusions contain lysophosphatidic acid (LPA)-like activity. Ann Oncol 1998;9:437–42. [22] Chun J, Hla T, Lynch KR, et al. International Union of Basic and Clinical Pharmacology. LXXVIII. Lysophospholipid receptor nomenclature. Pharmacol Rev 2010;62:579–87. [23] Yanagida K, Ishii S. Non-Edg family LPA receptors: the cutting edge of LPA research. J Biochem 2011;150:223–32. [24] Frisca F, Sabbadini RA, Goldshmit Y, et al. Biological effects of lysophosphatidic acid in the nervous system. Int Rev Cell Mol Biol 2012;296:273–322. [25] Anliker B. Lysophospholipid G protein-coupled receptors. J Biol Chem 2004;279:20555–8. [26] Keller JN, Steiner MR, Holtsberg FW, et al. Lysophosphatidic acid-induced proliferation-related signals in astrocytes. J Neurochem 1997;69:1073–84. [27] Meyer zu Heringdorf D, Jakobs KH. Lysophospholipid receptors: signalling, pharmacology and regulation by lysophospholipid metabolism. Biochim Biophys Acta 2007;1768:923–40. [28] Prenzel N, Zwick E, Daub H, et al. EGF receptor transactivation by G-proteincoupled receptors requires metalloproteinase cleavage of proHB-EGF. Nature 1999;402:884–8. [29] Daub H, Weiss FU, Wallasch C, et al. Role of transactivation of the EGF receptor in signalling by G-protein-coupled receptors. Nature 1996;379:557–60. [30] Daub H, Wallasch C, Lankenau A, et al. Signal characteristics of G proteintransactivated EGF receptor. EMBO J 1997;16:7032–44. [31] Goldshmit Y, Munro K, Leong SY, et al. LPA receptor expression in the central nervous system in health and following injury. Cell Tissue Res 2010;341:23–32. [32] Lee Z, Cheng CT, Zhang H, et al. Role of LPA4/p2y9/GPR23 in negative regulation of cell motility. Mol Biol Cell 2008;19:5435–45.
898
W. Ng et al. / Journal of Clinical Neuroscience 21 (2014) 893–898
[33] Tabuchi S, Kume K, Aihara M, et al. Expression of lysophosphatidic acid receptor in rat astrocytes: mitogenic effect and expression of neurotrophic genes. Neurochem Res 2000;25:573–82. [34] Savaskan NE, Rocha L, Kotter MR, et al. Autotaxin (NPP-2) in the brain: cell type-specific expression and regulation during development and after neurotrauma. Cell Mol Life Sci 2007;64:230–43. [35] Steiner MR, Holtsberg FW, Keller JN, et al. Lysophosphatidic acid induction of neuronal apoptosis and necrosis. Ann N Y Acad Sci 2000;905:132–41. [36] Stracke ML, Krutzsch HC, Unsworth EJ, et al. Identification, purification, and partial sequence analysis of autotaxin, a novel motility-stimulating protein. J Biol Chem 1992;267:2524–9. [37] Omidfar K, Shirvani Z. Single domain antibodies: a new concept for epidermal growth factor receptor and EGFRvIII targeting. DNA Cell Biol 2012;31: 1015–26. [38] Gan HK, Kaye AH, Luwor RB. The EGFRvIII variant in glioblastoma multiforme. J Clin Neurosci 2009;16:748–54. [39] Gschwind A, Zwick E, Prenzel N, et al. Cell communication networks: epidermal growth factor receptor transactivation as the paradigm for interreceptor signal transmission. Oncogene 2001;20:1594–600. [40] Hernández M, Barrero MJ, Crespo MS, et al. Lysophosphatidic acid inhibits Ca2+ signaling in response to epidermal growth factor receptor stimulation in human astrocytoma cells by a mechanism involving phospholipase C(gamma) and a G(alphai) protein. J Neurochem 2000;75:1575–82. [41] Lee HY, Bae GU, Jung ID, et al. Autotaxin promotes motility via G proteincoupled phosphoinositide 3-kinase gamma in human melanoma cells. FEBS Lett 2002;515:137–40. [42] Schleicher SM, Thotala DK, Linkous AG, et al. Autotaxin and LPA receptors represent potential molecular targets for the radiosensitization of murine glioma through effects on tumor vasculature. PLoS One 2011;6:e22182. [43] Shah BH, Neithardt A, Chu DB, et al. Role of EGF receptor transactivation in phosphoinositide 3-kinase-dependent activation of MAP kinase by GPCRs. J Cell Physiol 2006;206:47–57. [44] Ren XD, Kiosses WB, Schwartz MA. Regulation of the small GTP-binding protein Rho by cell adhesion and the cytoskeleton. EMBO J 1999;18: 578–85. [45] Yang SY, Lee J, Park CG, et al. Expression of autotaxin (NPP-2) is closely linked to invasiveness of breast cancer cells. Clin Exp Metastasis 2002;19:603–8. [46] Boucharaba A, Serre CM, Guglielmi J, et al. The type 1 lysophosphatidic acid receptor is a target for therapy in bone metastases. Proc Nat Acad Sci USA 2006;103:9643–8.
[47] Harper K, Arsenault D, Boulay-Jean S, et al. Autotaxin promotes cancer invasion via the lysophosphatidic acid receptor 4: participation of the cyclic AMP/EPAC/Rac1 signaling pathway in invadopodia formation. Cancer Res 2010;70:4634–43. [48] Frisca F, Crombie DE, Dottori M, et al. Rho/ROCK pathway is essential to the expansion, differentiation, and morphological rearrangements of human neural stem/progenitor cells induced by lysophosphatidic acid. J Lipid Res 2013;54:1192–206. [49] Hoelzinger DB, Demuth T, Berens ME. Autocrine factors that sustain glioma invasion and paracrine biology in the brain microenvironment. JNCI J Natl Cancer Inst 2007;99:1583–93. [50] Clair T, Lee HY, Liotta LA, et al. Autotaxin is an exoenzyme possessing 50 -nucleotide phosphodiesterase/ATP pyrophosphatase and ATPase activities. J Biol Chem 1997;272:996–1001. [51] Seasholtz TM, Radeff-Huang J, Sagi SA, et al. Rho-mediated cytoskeletal rearrangement in response to LPA is functionally antagonized by Rac1 and PIP2. J Neurochem 2004;91:501–12. [52] Khalil BD, El-Sibai M. Rho GTPases in primary brain tumor malignancy and invasion. J Neurooncol 2012. [53] Jongsma M, Matas-Rico E, Rzadkowski A, et al. LPA is a chemorepellent for B16 melanoma cells: action through the cAMP-elevating LPA5 receptor. PLoS One 2011;6:e29260. [54] Kim EK, Yun SJ, Do KH, et al. Lysophosphatidic acid induces cell migration through the selective activation of Akt1. Exp Mol Med 2008;40:445–52. [55] Nam SW, Clair T, Kim YS, et al. Autotaxin (NPP-2), a metastasis-enhancing motogen, is an angiogenic factor. Cancer Res 2001;61:6938–44. [56] Tigyi G. Aiming drug discovery at lysophosphatidic acid targets. Br J Pharmacol 2010;161:241–70. [57] Zhang H, Xu X, Gajewiak J, et al. Dual activity lysophosphatidic acid receptor pan-antagonist/autotaxin inhibitor reduces breast cancer cell migration in vitro and causes tumor regression in vivo. Cancer Res 2009;69:5441–9. [58] Baker DL, Fujiwara Y, Pigg KR, et al. Carba analogs of cyclic phosphatidic acid are selective inhibitors of autotaxin and cancer cell invasion and metastasis. J Biol Chem 2006;281:22786–93. [59] Saunders LP, Ouellette A, Bandle R, et al. Identification of small-molecule inhibitors of autotaxin that inhibit melanoma cell migration and invasion. Mol Cancer Ther 2008;7:3352–62. [60] Xu X, Yang G, Zhang H, et al. Evaluating dual activity LPA receptor panantagonist/autotaxin inhibitors as anti-cancer agents in vivo using engineered human tumors. Prostaglandins Other Lipid Mediat 2009;89:140–6.
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Review
Complexities of lysophospholipid signalling in glioblastoma Wayne Ng a,b,c,f, Alice Pébay a,d, Katharine Drummond a,b,c,f, Antony Burgess a,e, Andrew H. Kaye a,b,c, Andrew Morokoff a,b,c,⇑ a
University of Melbourne, Parkville, VIC, Australia Department of Surgery, Royal Melbourne Hospital, Grattan Street, Parkville, VIC 3050, Australia c Department of Neurosurgery, Royal Melbourne Hospital, Parkville, VIC, Australia d Centre for Eye Research Australia, Royal Victorian Eye and Ear Hospital, East Melbourne, VIC, Australia e Walter and Eliza Hall Institute of Medical Research, Parkville, VIC, Australia f Melbourne Brain Centre at Royal Melbourne Hospital, Parkville, VIC, Australia b
a r t i c l e
i n f o
Article history: Received 12 January 2014 Accepted 22 February 2014
Keywords: Autotaxin Glioblastoma multiforme Lysophosphatidic acid
a b s t r a c t Glioblastoma multiforme (GBM) is the most malignant brain tumour and continues to have a very poor median survival of 12–16 months despite current best therapies. These aggressive tumours always recur after treatment and are defined by their ability to diffusely infiltrate and invade normal brain parenchyma. Autotaxin is overexpressed in GBM, and is a potent chemotactic enzyme that produces lysophosphatidic acid. Lysophospholipid (LPL) signalling is known to increase invasion of solid tumours and is also dysregulated in GBM. The LPL pathway has been shown to interact with known cancer-related signalling pathways, including those for epidermal growth factor and yes-associated protein, which are also dysregulated in GBM. The interactions between these pathways provide insights into the complexities of cancer signalling and suggest potential novel targets for GBM. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Glioblastoma multiforme (GBM) is the most malignant (World Health Organization grade IV) glioma and continues to have a very poor median survival of 12–16 months despite current best therapies (maximal safe resection with concurrent temozolomide chemotherapy and radiation therapy) [1]. These aggressive tumours always recur after treatment and are defined by their ability to diffusely infiltrate and invade normal brain parenchyma. Thus the search for targeted agents inhibiting cell proliferation, survival and invasion has intensified. Research into the epidermal growth factor receptor (EGFR) pathway has led to clinical trials of EGFR and phosphatidylinositol-3-kinase (PI3K) inhibitors that modulate cell survival and proliferation in pre-clinical models. However, results from early EGFR inhibitor trials have not delivered on their promise and PI3K inhibitor trials are ongoing [2–4]. These trials have made it obvious that strategies combining therapies against multiple targets are required to account for the existence of complex pathway interactions and redundancies. For example, matrix metalloproteases (MMP) degrade extracellular matrix components to produce more favourable conditions for cell migration ⇑ Corresponding author. Tel.: +61 3 9035 8586 E-mail address: [email protected] (A. Morokoff). http://dx.doi.org/10.1016/j.jocn.2014.02.013 0967-5868/Ó 2014 Elsevier Ltd. All rights reserved.
and invasion [5]. However, they also cleave and activate growth factors such as epidermal growth factor (EGF) [6]. The function of integrins and their influence on cell morphology and migration has also been enlightening [7,8]. More recently, other promising factors have been identified, including autocrine motility factor receptor, heparin-binding epidermal growth factor, ephrin-B3, netrin 4 and autotaxin (ATX) [9]. The latter three are of interest because their role in cell migration and motility in neural stem cells suggests a similar role in glioma-derived cancer stem cells, that have a putative role in GBM progression [10,11]. ATX in particular is a powerful chemotactic enzyme involved in lysophospholipid (LPL) signalling, and its recent prominence in the literature has highlighted the importance of lipid signalling within complex intracellular pathway interactions. This review focuses on the role that LPL signalling may play in gliomagenesis and its potential as a target in the treatment of this highly malignant disease. 2. Lysophosphatidic acid Lysophosphatidic acid (LPA) and sphingosine-1-phosphate (S1P) are the main membrane-derived lipid signalling molecules. LPA has a 3-carbon glycerol backbone, with an attached single acyl or alkyl chain of varying length which imparts some differences in receptor efficacy [12]. Whilst some LPA production may occur
894
W. Ng et al. / Journal of Clinical Neuroscience 21 (2014) 893–898
intracellularly, much of it is produced extracellularly by secreted enzymes. There are three known pathways: (1) cleavage of LPL (such as lysophosphatidylcholine) by lysophospholipase D, (2) deacylation of phosphatidic acid by phospholipase A1 and A2, and (3) mild oxidation of low-density lipoprotein (non-enzymatic). Lysophospholipase D is now more commonly known as ATX which derives its name from early characterisation of its stimulatory effect on melanoma cell motility [13–15]. It is also known as ectonucleotide pyrophosphatase and phosphodiesterase-2 (ENPP-2) and is part of the family of ENPP enzymes which are traditionally known for their involvement in nucleotide metabolism [13]. LPA production is mainly via ATX catalysis. This is confirmed by the uniformly fatal outcome of homozygous ATX (Atx/) knockout mice and the 50% reduction in circulating plasma LPA levels in embryos with heterozygous ATX (Atx+/) expression [16,17]. The phospholipase A enzymes play an important role in determining the position of the acylation of the phosphoglycerol backbone and so, whilst there is currently no evidence of its upregulation, it may still be an important pathway in pathological states, as it may influence the action of liberated LPA species by playing a role in determining the dominant LPA species produced (acyl or alkyl) [18]. Physiologic LPA is present in small amounts in tissues since it regulates its own production via negative feedback inhibition of its main synthesising enzyme, ATX [19]. Thus, the influence of this signalling pathway is largely autocrinic/paracrinic [12–14]. Further, circulating platelets and erythrocytes secrete LPA which is then bound to albumin to protect it from rapid enzymatic degradation by lysophospholipases, lipid phosphate phosphatases and LPA acyl transferases [12]. Cancer cells can produce increased amounts of ATX/LPA and it has been previously reported that some malignant effusions (such as those of ovarian cancer) have elevated levels of LPA compared to other malignancies [2,20,21]. In addition, circulating plasma levels of LPA can be elevated in malignancy and this may be the result of elevated production (induction of ATX) [14]. 3. LPA receptors Only a brief review of the LPA receptors will be provided here as there are many other detailed reviews available [12,16,22–24]. At the time of writing, the International Union of Basic and Clinical Pharmacology had recognised six definitive G-protein coupledLPA receptors (collectively LPAR) designated LPA1–6 (Table 1). Broadly, the receptors fall into two families: endothelial differentiation gene (Edg) and non-Edg (purinergic) receptors [12]. LPA1 (Edg2), LPA2 (Edg4) and LPA3 (Edg7) are members of the Edg family
and are the best characterised to our knowledge with LPA1 being the dominant LPA receptor in the central nervous system (CNS) [13,23,25]. LPA1 is coupled to the G-proteins, Gi/o, Gq and G12/13, which allows it to signal via multiple pathways, including major cancer-related pathways such as mitogen-activated protein kinase (MAPK), Akt/PKB, and small GTPases such as Rho/ROCK [12,26]. LPA1-induced PI3K signalling (which activates Akt/PKB) via the p110b/c subunits (of PI3K) has also been reported [27–30]. Whilst LPA2 is expressed in embryonic brain it has little to no expression in the adult CNS [31]. LPA3 is expressed in the brain, but unlike LPA1 and LPA2 it is coupled to the G-proteins, Gi/o and Gq, but not G12/13 and so is less responsive to LPL than LPA1 [27]. As LPA3 does not signal via G12/13, it is not involved in changes in cell morphology, and therefore unlikely to be involved in cell migration [27]. The non-Edg or purinergic family of LPA receptors are genetically distinct from the original Edg family of LPA receptors [23]. Importantly, this results in the non-Edg family having an increased affinity for alkyl-LPA species, as opposed to the Edg family having increased affinity for the acyl variants [22,23]. Current members of the non-Edg family include LPA4 (P2Y9), LPA5 (GPR92) and LPA6 (P2Y5). LPA4 signals via Gs, Gq and G12/13-proteins and probably plays a role in cell motility and migration [23]. Mouse embryonic fibroblasts (MEF) derived from Lpa4 (null) knockout mice have been reported to exhibit hypersensitivity to LPA induced motility. These MEF had increased levels of phosphorylated Akt (pAkt) when stimulated by LPA. The elevated pAkt levels and associated motile response were attenuated when Lpa4 was reintroduced into the cells, suggesting that LPA4 has an action which might suppress signalling activity of the LPA1 receptor [32]. LPA5 (GPR92) and LPA6 (P2Y5) were both discovered subsequent to LPA4 and to our knowledge there is no reported role for LPA5 in tumourigenesis. LPA6 however, may cause morphological changes in vascular endothelial cells and therefore may play a minor or indirect role in gliomagenesis via effects on vascular development [22,23,27]. LPL have a wide range of physiological and pathological effects owing to the myriad G-proteins they are coupled to. As a result, they modulate numerous normal physiologic processes, including cell proliferation and apoptosis, cell differentiation, cell adhesion and migration, cell morphology (including neurite retraction and synaptic cleft shape modulation), normal CNS development and autoimmunity [12,13,16,18,25,33–35]. Dysregulation of these events is strongly implicated in tumourigenesis. 4. LPA signalling in cancer Since Stracke et al. discovered the promotile effects of ATX on melanoma cells in 1992, the LPA pathway has been investigated
Table 1 LPA receptors and their relevance to cancer research LPA receptor
Coupled G-proteins
Relevance to cancer research
Reference
Edg
LPA1
Gi/o, Gq, G12/13
Kishi et al. [2]
LPA2
Gi/o, Gq, G12/13
LPA3
Gi/o, Gq
– – – – – – –
LPA4
Gs, Gq, G12/13
LPA5
Gq, G12/13 Increase in cAMP not shown to be Gs-mediated G12/13, (possible role for Gs, Gi)
Non-Edg
LPA6
Dominant LPA receptor expressed in adult CNS Increased expression in cancers including GBM Linked to cell proliferation, survival, migration/invasion Not present in adult CNS Not markedly elevated in GBM Expressed in adult CNS Not markedly elevated in GBM
Kishi et al. [2] Kishi et al. [2], Choi et al. [16]
– Activation of LPA4 has been shown to be functionally antagonistic (anti-migration, anti-proliferation) to LPA1 – Activation has been shown to reduce cell migration
Kato [59], Yanagida and Ishii [23], Lee et al. [32]
– Possible similar function to LPA4 – May regulate vascular permeability
Yanagida and Ishii [23]
Jongsma et al. [54]
cAMP = cyclic adenosine monophosphate, CNS = central nervous system, Edg = endothelial differentiation gene, GBM = glioblastoma multiforme, LPA = lysophosphatidic acid.
W. Ng et al. / Journal of Clinical Neuroscience 21 (2014) 893–898
for its role in tumour invasion and metastasis [36]. However, its original role in nucleotide metabolism could not be directly reconciled with this biological effect. It was the subsequent discovery by Umezu-Goto et al. [14] and Tokumura et al. [15] that ATX had additional lysophospholipase D activity that triggered further interest in LPA. Recent elucidation of the complex interactions of LPA with known cancer signalling pathways will be discussed below (Fig. 1). 4.1. LPA interacts with cell survival pathways EGFR signalling is commonly upregulated in GBM cells either via over-expression, EGFR gene amplification, or via the EGFR type III mutation, which produces a constitutively active form of EGFR [37,38]. EGFR activates downstream PI3K signalling which regulates cell survival via abrogation of apoptotic cell signals and this is dysregulated in glioma. However, clinical trials with EGFR monoclonal antibodies [3] were less promising than anticipated, possibly because of intracellular pathway cross-communication and redundancy [28,39]. It has been reported that LPA can intracellularly transactivate EGFR via the Gi-protein and that this is independent of normal ligand activation of EGFR by EGF [28–30,40]. Further, it is possible that transactivation of p110b/c subunits of PI3K via Gi-protein coupled LPAR is more important than transactivation of p110a [41]. This LPA-Gi-protein coupled signalling link has been reported in many cancer cell lines including those from prostate, breast, colon and GBM [28,39]. Recently, Schleicher et al. reported that LPA signalling confers radioresistance to the murine glioma cell line GL-261 [42]. Inhibition of LPA signalling with a combination ATX/LPAR inhibitor, a-bromomethylene phosphonate lysophosphatidic acid (BrP-LPA), resulted in impaired survival signals in this cell line when exposed to 3 Gy of irradiation. Knockdown of LPA1 and LPA3 (but not LPA2) signalling using siRNA resulted in reduced pAkt and also correlated with reduced survival in these cells. These results point to the hypothesis that LPA1 and LPA3 play an important role in radioresistance, via interaction with the PI3K/Akt pathway.
895
4.2. LPA signalling in cell proliferation In addition to enhancing cell survival, transactivation of EGFR by LPA also upregulates cell proliferation via MAPK dependent mitogenic signalling. Whilst the transactivation of PI3K signalling can be EGF independent, the transactivation of MAPK appears to be dependent on agonistic activation of EGFR. Further, MAPK transactivation may be linked to activation of PI3K [43]. Its transactivation is also cell type dependent and has also been confirmed to be G-protein dependent [29,30,39,43]. Overall, transactivation of MAPK by LPA appears to be more regulated (than PI3K) and in fact may also require (or be augmented by) concurrent MMP activity, which is also upregulated in gliomas [5,6].
4.3. LPL and ATX signalling influences cancer cell migration and invasion Alterations in cell morphology are tightly regulated and play an important role in terms of chemotactic effect [44]. It is likely that ATX and LPA play their roles in glioma migration and invasion via effects on cell morphology. Overexpression of ATX also correlates with increased invasiveness of breast cancer cells compared to normal breast cells [45]. LPAR overexpression has also been reported in ovarian cancers and this has been postulated to create an autocrine/paracrine loop promoting proliferation (MAPK signalling) and suppressing apoptosis (PI3K signalling) [6,39,43]. Treatment of animals with metastatic ovarian cancer with Ki16425 (selective LPA1/3 receptor antagonist) was demonstrated to have an in vivo anti-tumour effect without obvious adverse health effects on the treated nu/nu mice [46]. Also, use of a pan-LPAR antagonist to treat hepatic metastases in an orthotopic breast cancer model showed significant reduction in tumour deposit size [19]. LPA4 has been shown to promote invadopodia formation in HT1080 fibrosarcoma cells, with Lpa4 shRNA abolishing this effect [47]. Invadopodia are involved in invasion and metastasis and are enriched with MMP, which are known for their key role in invasion and metastasis by providing a favourable milieu for cells to invade
Fig. 1. Overview of intracellular interactions between lysophosphatidic acid (LPA) and epidermal growth factor receptor (EGFR) signalling pathways. Mutations leading to overexpression or elevated activity of autotaxin, LPA1–3, EGFR and phosphoinositol-3-kinase are often found in glioblastoma multiforme. ATX = autotaxin, EGF = epidermal growth factor, EGFR = epidermal growth factor receptor, ERK = extracellular signal-regulated kinases, LPA = lysophosphatidic acid, LPC = lysophosphatidylcholine, MAPK = mitogen-activated phosphokinase, mTOR = mammalian target of rapamycin, PIP2 = phosphatidylinositol 4,5-bisphosphate, PIP3 = phosphatidylinositol (3,4,5)trisphosphate, PI3K = phosphoinositol-3-kinase, PTEN = phosphatase-tensin homologue, ROCK = Rho-associated kinase.
896
W. Ng et al. / Journal of Clinical Neuroscience 21 (2014) 893–898
through and also cleave and activate growth factors such as EGF [5,6]. In glioma, LPA has been shown to induce astrocyte motility by dramatically altering astrocyte shape [24,33,48]. ATX overexpression has also been shown to enhance the invasion of U87 and U251 malignant glioma cells (autocrine effect) through oligodendrocyte monolayers in vitro, whilst simultaneously reducing the adhesiveness of oligodendrocytes (paracrine effect). Depletion or inactivation of ATX ameliorates these effects. These findings may provide clues to the underlying pathogenesis of in vivo white matter tract invasion of glioma cells [9]. Overexpression of ATX has been shown to be proportionately higher at the leading invasive edges of tumours [9]. There are currently no definitive reports to address whether the overexpression of ATX and upregulation of LPAR is specific to certain cell types within the tumour. But microglia may be recruited to the invasive edge of the tumour by various chemokines (including LPA) and subsequently facilitate invasion by producing ATX, LPA and EGF [49]. The dominant LPA receptors involved in mediating cell migration and motility appear to be LPA1 and LPA4. LPA1 receptors have been linked to tumour cell motility in vitro and have been reported to be overexpressed in GBM cell lines (SNB-78, SNB-75, SF-268, SF539 and SF-295) in combination with overexpression of ATX. This receptor/ligand dual overexpression may contribute to autocrinic stimulation of GBM cell motility and this can be abolished when the cells are treated with the LPA1-3 receptor antagonist, Ki16425 [2]. Early work on the cellular mechanisms of LPA induced motility revealed a link to Rho signalling through either Gi (pertussis toxin sensitive) or G12/13-protein [13,50,51]. The subsequent Rhodependent cytoskeletal rearrangement was reported to induce cell rounding in astrocytes and this may contribute to producing discohesive cells [51]. However, there have been some contradictory reports with regards to Rho-related LPA signalling. For example, LPA induced glioma cell migration can be inhibited by blocking Rho activation. Contrary to this, stimulation of Rho activity has also been demonstrated to cause immobilisation of glioma cells [52]. Recall the different evolutionary pathways of the Edg and non-Edg families of LPAR and there may exist functional antagonism between them to provide an explanation for these inconsistencies. Lee et al. reported that LPA4 is likely a functional antagonist of LPA1, leading to a reduction in cell motility in MEF cells as well as DLD1 colorectal cancer cell lines and B103 neuronal cell lines [32]. It is possible that if LPA4 antagonises LPA1, that LPA4
somehow modulates activation of LPA1 or its signalling. Interestingly, activation of LPA5 has recently been reported to inhibit B16 melanoma cell migration [53]. This further supports the notion that the non-Edg (purinergic; LPA4–6) family of LPA receptors may modulate the effects of the Edg (LPA1–3) family. In addition to the reported Rho-dependent mechanisms, LPAmediated cell migration is also thought to occur via MAPK and PI3K signalling. LPA1 has been shown to activate MAPK and PI3K/ Akt signalling to promote cell migration [16,54]. Also, it has been confirmed that ATX induced motility in melanoma cells is mediated via the p110c subunit (of PI3K) and that PI3K inhibitor can inhibit this motile response in a dose-dependent manner [41]. Further, Kim et al. reported that simultaneous knockout of Akt1 and Akt2 abolished LPA induced motility in MEF [54]. Only reexpression of Akt1 following the double knockout restored the motile response to LPA. They also established that exposure of MEF to the pan-PI3K inhibitor (LY294002) or the LPAR antagonist, Ki16425, also completely abolished the motile response. This would make it likely that the downstream activation of Akt1 is via PI3K activation. However, it is not completely clear whether the subsequent phosphorylation of Akt is via transactivation of EGFR, or via interaction of LPAR associated Gi-protein with PI3K. 4.4. The potential role of LPA in angiogenesis LPA signalling seems to have a minor physiological role in angiogenesis, as S1P normally maintains this process. However, in pathological states, the role of LPA in angiogenesis may become more important [12]. Lpa(/) null mice have been shown to have 1 a small but significant incidence (2.5%) of frontal intracerebral haemorrhages which increases substantially in Lpa(/) /Lpa(/) 1 2 double-null mice to 26%. This may support a role for LPA in vessel maturation [12]. Furthermore, Atx(/) null murine embryos have a profound absence of vessel maturation that causes uniform embryonic mortality by day 10 [16]. This points to a potential role for LPA signalling in neoangiogenesis. In vivo murine models investigating ATX have shown that ATX plays a role in angiogenesis similar to that of vascular endothelial growth factor [55]. Other in vivo studies investigating knockouts of Lpa1–4 have not had the profound effect of Atx(/) knockout on vascular development, but instead, G12/13-protein knockouts have been found to have effects which parallel those of ATX. Therefore, it is possible that the LPAR may play redundant roles in neoangiogenesis, or the effect ATX has on vascular development may be mediated through other pathways such as vascular endothelial growth factor [23].
Table 2 Evidence for the potential therapeutic effect of lysophosphatidic acid and autotaxin modulation Compound
Mechanism
Animal
Cell line (s)
Cancer
Effect
Reference
Ki16425
Selective LPA1/3 antagonist
Balb-c nu/nu mice
Selective ATX inhibitor
NA
Ovarian Breast Melanoma
NSC 48300 bithionol
Small molecule ATX inhibitors Dual pan-LPA antagonist/ ATX inhibitor
NA
A2058
Melanoma
Balb-c nu/nu mice
MDA-MB-231
Breast
MDA-MB-231 A-549 HCT-116 Mouse GL-261
Breast Lung Colon GBM
In vivo – regression of bone metastases In vitro – reduction in LPA production and subsequent cell migration and invasion In vitro – reduced cell migration and invasion In vitro – reduced cell migration and invasion In vivo – tumour regression In vivo – tumour growth delay
Boucharaba et al. [46]
Cyclic phosphatidic acid analogs
CHOb3wt MDA-BO2 A2058
BrP-LPA
BrP-LPA
Dual pan-LPA antagonist/ ATX inhibitor
Balb-c nu/nu mice
BrP-LPA
Dual pan-LPA antagonist/ ATX inhibitor
Balb-c nu/nu mice
In vitro – radioprotection of glioma cell lines In vivo – tumour growth delay
ATX = autotaxin, GBM = glioblastoma multiforme, BrP = a-bromomethylene phosphonate, LPA = lysophosphatidic acid, NA = not applicable.
Baker et al. [58]
Saunders et al. [59] Zhang et al. [57]
Xu et al. [60]
Schleicher et al. [42]
W. Ng et al. / Journal of Clinical Neuroscience 21 (2014) 893–898
Recent evidence also suggests that LPA may bind intracellular targets such as peroxisome proliferator-activated receptor-c (PPAR-c) to cause neointimal formation [27]. It is unclear as to whether LPA migrates across the cell membrane but the effects of LPA induced PPAR-c activity are significantly diminished in Lpa(/) knockout mice. Regardless, specific unsaturated acyl LPA 1 interact with PPAR-c to induce neointimal formation and are likely produced by phosphorylation of monoacylglycerols (mitochondrial acylglycerol kinase) or by conversion of phosphatidic acid to LPA by cytosolic phospholipase A2 [18,27,56]. The role of LPA and ATX in angiogenesis no doubt complements their putative roles in tumour growth and invasion/metastasis. 5. Discussion Decades of glioma research are now seeing the beginnings of molecular and genetic profiling of gliomas. Additionally, gliomaderived cancer stem cells have been identified, providing a new framework to investigate this disease’s resistance to treatment and subsequent recurrence. Much of the evidence for the role of LPA in cancer is spread across different cell lineages making it difficult to extrapolate some of the knowledge to gliomas. However, the emerging reports regarding LPA and gliomagenesis for the most part echo the findings in other solid tumours. The discovery of ATX’s chemotactic effects has brought the spotlight to this field, resulting in invaluable insights into the complexities of intracellular crosscommunication and pathway redundancies. It has also lead to preclinical trials demonstrating the potential of this pathway as a target in malignancies (Table 2). Recently, the potential benefits of targeting ATX-LPA signalling were shown in murine breast cancer models using a combination ATX inhibitor and LPAR antagonist. This treatment reduced in vitro migration of breast cancer cells and caused tumour regression in vivo [57]. This strategy is based on the theory that the Edg (LPA1–3) receptors may be the main procarcinogenic mediators of cancer in dysregulated LPA signalling. An alternative strategy to targeting LPA signalling might look to produce agonists to the non-Edg (LPA4/5) receptor family, as there is mounting evidence that they might mitigate the pro-carcinogenic effects of the Edg receptors. This strategy may have benefits of being more selective. However, the relative underexpression or absence of these receptors in some tumours may limit its value. LPA1 has been shown to be the dominant LPAR expressed in gliomas but some glioma cell lines have also been shown to express moderate levels of LPA4 [2]. Therefore, at least in some gliomas, there may be merit to developing LPA4 agonists as an anti-cancer therapy. Also, an investigation into expression patterns of LPAR in different cell subpopulations (glioma-derived cancer stem cells, progenitor cells, and recruited host cells such as microglia and reactive astrocytes and oligodendroglia) within the glioma microenvironment is critical to ascertain the true value of LPA modulation as a treatment strategy. Conflicts of Interest/Disclosures The authors declare that they have no financial or other conflicts of interest in relation to this research and its publication. Acknowledgements This study was funded by the Neurosurgical Society of Australasia/LifeHealthCare scholarship (2012) and the Brain Foundation of Australia Brain Tumours Award (2012) and the Victorian State Government’s Department of Innovation, Industry and Regional Development’s Operational Infrastructure Support Program. AP is supported by a National Health and Medical Research Council Career Development Fellowship.
897
References [1] Stupp R, Hegi ME, Mason WP, et al. Effects of radiotherapy with concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the EORTCNCIC trial. Lancet Oncol 2009;10:459–66. [2] Kishi Y, Okudaira S, Tanaka M, et al. Autotaxin is overexpressed in glioblastoma multiforme and contributes to cell motility of glioblastoma by converting lysophosphatidylcholine to lysophosphatidic acid. J Biol Chem 2006;281: 17492–500. [3] Camara-Quintana JQ, Nitta RT, Li G. Pathology: commonly monitored glioblastoma markers: EFGR, EGFRvIII, PTEN, and MGMT. Neurosurg Clin N Am 2012;23:237–46. [4] Wen PY, Lee EQ, Reardon DA, et al. Current clinical development of PI3K pathway inhibitors in glioblastoma. Neuro Oncol 2012;14:819–29. [5] Pullen NA, Anand M, Cooper PS, et al. Matrix metalloproteinase-1 expression enhances tumorigenicity as well as tumor-related angiogenesis and is inversely associated with TIMP-4 expression in a model of glioblastoma. J Neurooncol 2012;106:461–71. [6] Gschwind A, Prenzel N, Ullrich A. Lysophosphatidic acid-induced squamous cell carcinoma cell proliferation and motility involves epidermal growth factor receptor signal transactivation. Cancer Res 2002;62:6329–36. [7] Tchaicha JH, Reyes SB, Shin J, et al. Glioblastoma angiogenesis and tumor cell invasiveness are differentially regulated by b8 integrin. Cancer Res 2011;71: 6371–81. [8] D’Abaco GM, Kaye AH. Integrins: molecular determinants of glioma invasion. J Clin Neurosci 2007;14:1041–8. [9] Hoelzinger DB, Nakada M, Demuth T, et al. Autotaxin: a secreted autocrine/ paracrine factor that promotes glioma invasion. J Neurooncol 2007;86: 297–309. [10] Singh SK, Clarke ID, Terasaki M, et al. Identification of a cancer stem cell in human brain tumors. Cancer Res 2003;63:5821–8. [11] Hemmati HD, Nakano I, Lazareff JA, et al. Cancerous stem cells can arise from pediatric brain tumors. Proc Natl Acad Sci USA 2003;100:15178–83. [12] Ishii I, Fukushima N, Ye X, et al. Lysophospholipid receptors: signaling and biology. Annu Rev Biochem 2004;73:321–54. [13] Moolenaar WH. Lysophospholipids in the limelight: autotaxin takes center stage. J Cell Biol 2002;158:197–9. [14] Umezu-Goto M, Kishi Y, Taira A, et al. Autotaxin has lysophospholipase D activity leading to tumor cell growth and motility by lysophosphatidic acid production. J Cell Biol 2002;158:227–33. [15] Tokumura A, Majima E, Kariya Y, et al. Identification of human plasma lysophospholipase D, a lysophosphatidic acid-producing enzyme, as autotaxin, a multifunctional phosphodiesterase. J Biol Chem 2002;277:39436–42. [16] Choi JW, Lee CW, Chun J. Biological roles of lysophospholipid receptors revealed by genetic null mice: an update. Biochim Biophys Acta 2008;1781: 531–9. [17] Tanaka M, Okudaira S, Kishi Y, et al. Autotaxin stabilizes blood vessels and is required for embryonic vasculature by producing lysophosphatidic acid. J Biol Chem 2006;281:25822–30. [18] Aoki J, Inoue A, Okudaira S. Two pathways for lysophosphatidic acid production. Biochim Biophys Acta 2008;1781:513–8. [19] Prestwich GD, Gajewiak J, Zhang H, et al. Phosphatase-resistant analogues of lysophosphatidic acid: agonists promote healing, antagonists and autotaxin inhibitors treat cancer. Biochim Biophys Acta 2008;1781:588–94. [20] Baker DL, Morrison P, Miller B, et al. Plasma lysophosphatidic acid concentration and ovarian cancer. JAMA 2002;287:3081–2. [21] Westermann AM, Havik E, Postma FR, et al. Malignant effusions contain lysophosphatidic acid (LPA)-like activity. Ann Oncol 1998;9:437–42. [22] Chun J, Hla T, Lynch KR, et al. International Union of Basic and Clinical Pharmacology. LXXVIII. Lysophospholipid receptor nomenclature. Pharmacol Rev 2010;62:579–87. [23] Yanagida K, Ishii S. Non-Edg family LPA receptors: the cutting edge of LPA research. J Biochem 2011;150:223–32. [24] Frisca F, Sabbadini RA, Goldshmit Y, et al. Biological effects of lysophosphatidic acid in the nervous system. Int Rev Cell Mol Biol 2012;296:273–322. [25] Anliker B. Lysophospholipid G protein-coupled receptors. J Biol Chem 2004;279:20555–8. [26] Keller JN, Steiner MR, Holtsberg FW, et al. Lysophosphatidic acid-induced proliferation-related signals in astrocytes. J Neurochem 1997;69:1073–84. [27] Meyer zu Heringdorf D, Jakobs KH. Lysophospholipid receptors: signalling, pharmacology and regulation by lysophospholipid metabolism. Biochim Biophys Acta 2007;1768:923–40. [28] Prenzel N, Zwick E, Daub H, et al. EGF receptor transactivation by G-proteincoupled receptors requires metalloproteinase cleavage of proHB-EGF. Nature 1999;402:884–8. [29] Daub H, Weiss FU, Wallasch C, et al. Role of transactivation of the EGF receptor in signalling by G-protein-coupled receptors. Nature 1996;379:557–60. [30] Daub H, Wallasch C, Lankenau A, et al. Signal characteristics of G proteintransactivated EGF receptor. EMBO J 1997;16:7032–44. [31] Goldshmit Y, Munro K, Leong SY, et al. LPA receptor expression in the central nervous system in health and following injury. Cell Tissue Res 2010;341:23–32. [32] Lee Z, Cheng CT, Zhang H, et al. Role of LPA4/p2y9/GPR23 in negative regulation of cell motility. Mol Biol Cell 2008;19:5435–45.
898
W. Ng et al. / Journal of Clinical Neuroscience 21 (2014) 893–898
[33] Tabuchi S, Kume K, Aihara M, et al. Expression of lysophosphatidic acid receptor in rat astrocytes: mitogenic effect and expression of neurotrophic genes. Neurochem Res 2000;25:573–82. [34] Savaskan NE, Rocha L, Kotter MR, et al. Autotaxin (NPP-2) in the brain: cell type-specific expression and regulation during development and after neurotrauma. Cell Mol Life Sci 2007;64:230–43. [35] Steiner MR, Holtsberg FW, Keller JN, et al. Lysophosphatidic acid induction of neuronal apoptosis and necrosis. Ann N Y Acad Sci 2000;905:132–41. [36] Stracke ML, Krutzsch HC, Unsworth EJ, et al. Identification, purification, and partial sequence analysis of autotaxin, a novel motility-stimulating protein. J Biol Chem 1992;267:2524–9. [37] Omidfar K, Shirvani Z. Single domain antibodies: a new concept for epidermal growth factor receptor and EGFRvIII targeting. DNA Cell Biol 2012;31: 1015–26. [38] Gan HK, Kaye AH, Luwor RB. The EGFRvIII variant in glioblastoma multiforme. J Clin Neurosci 2009;16:748–54. [39] Gschwind A, Zwick E, Prenzel N, et al. Cell communication networks: epidermal growth factor receptor transactivation as the paradigm for interreceptor signal transmission. Oncogene 2001;20:1594–600. [40] Hernández M, Barrero MJ, Crespo MS, et al. Lysophosphatidic acid inhibits Ca2+ signaling in response to epidermal growth factor receptor stimulation in human astrocytoma cells by a mechanism involving phospholipase C(gamma) and a G(alphai) protein. J Neurochem 2000;75:1575–82. [41] Lee HY, Bae GU, Jung ID, et al. Autotaxin promotes motility via G proteincoupled phosphoinositide 3-kinase gamma in human melanoma cells. FEBS Lett 2002;515:137–40. [42] Schleicher SM, Thotala DK, Linkous AG, et al. Autotaxin and LPA receptors represent potential molecular targets for the radiosensitization of murine glioma through effects on tumor vasculature. PLoS One 2011;6:e22182. [43] Shah BH, Neithardt A, Chu DB, et al. Role of EGF receptor transactivation in phosphoinositide 3-kinase-dependent activation of MAP kinase by GPCRs. J Cell Physiol 2006;206:47–57. [44] Ren XD, Kiosses WB, Schwartz MA. Regulation of the small GTP-binding protein Rho by cell adhesion and the cytoskeleton. EMBO J 1999;18: 578–85. [45] Yang SY, Lee J, Park CG, et al. Expression of autotaxin (NPP-2) is closely linked to invasiveness of breast cancer cells. Clin Exp Metastasis 2002;19:603–8. [46] Boucharaba A, Serre CM, Guglielmi J, et al. The type 1 lysophosphatidic acid receptor is a target for therapy in bone metastases. Proc Nat Acad Sci USA 2006;103:9643–8.
[47] Harper K, Arsenault D, Boulay-Jean S, et al. Autotaxin promotes cancer invasion via the lysophosphatidic acid receptor 4: participation of the cyclic AMP/EPAC/Rac1 signaling pathway in invadopodia formation. Cancer Res 2010;70:4634–43. [48] Frisca F, Crombie DE, Dottori M, et al. Rho/ROCK pathway is essential to the expansion, differentiation, and morphological rearrangements of human neural stem/progenitor cells induced by lysophosphatidic acid. J Lipid Res 2013;54:1192–206. [49] Hoelzinger DB, Demuth T, Berens ME. Autocrine factors that sustain glioma invasion and paracrine biology in the brain microenvironment. JNCI J Natl Cancer Inst 2007;99:1583–93. [50] Clair T, Lee HY, Liotta LA, et al. Autotaxin is an exoenzyme possessing 50 -nucleotide phosphodiesterase/ATP pyrophosphatase and ATPase activities. J Biol Chem 1997;272:996–1001. [51] Seasholtz TM, Radeff-Huang J, Sagi SA, et al. Rho-mediated cytoskeletal rearrangement in response to LPA is functionally antagonized by Rac1 and PIP2. J Neurochem 2004;91:501–12. [52] Khalil BD, El-Sibai M. Rho GTPases in primary brain tumor malignancy and invasion. J Neurooncol 2012. [53] Jongsma M, Matas-Rico E, Rzadkowski A, et al. LPA is a chemorepellent for B16 melanoma cells: action through the cAMP-elevating LPA5 receptor. PLoS One 2011;6:e29260. [54] Kim EK, Yun SJ, Do KH, et al. Lysophosphatidic acid induces cell migration through the selective activation of Akt1. Exp Mol Med 2008;40:445–52. [55] Nam SW, Clair T, Kim YS, et al. Autotaxin (NPP-2), a metastasis-enhancing motogen, is an angiogenic factor. Cancer Res 2001;61:6938–44. [56] Tigyi G. Aiming drug discovery at lysophosphatidic acid targets. Br J Pharmacol 2010;161:241–70. [57] Zhang H, Xu X, Gajewiak J, et al. Dual activity lysophosphatidic acid receptor pan-antagonist/autotaxin inhibitor reduces breast cancer cell migration in vitro and causes tumor regression in vivo. Cancer Res 2009;69:5441–9. [58] Baker DL, Fujiwara Y, Pigg KR, et al. Carba analogs of cyclic phosphatidic acid are selective inhibitors of autotaxin and cancer cell invasion and metastasis. J Biol Chem 2006;281:22786–93. [59] Saunders LP, Ouellette A, Bandle R, et al. Identification of small-molecule inhibitors of autotaxin that inhibit melanoma cell migration and invasion. Mol Cancer Ther 2008;7:3352–62. [60] Xu X, Yang G, Zhang H, et al. Evaluating dual activity LPA receptor panantagonist/autotaxin inhibitors as anti-cancer agents in vivo using engineered human tumors. Prostaglandins Other Lipid Mediat 2009;89:140–6.