Group X secreted phospholipase A₂ induces lipid droplet formation and prolongs breast cancer cell survival

Tumor cells display progressive, oncogene-driven alterations in the metabolic pathways that supply energy and biosynthetic intermediates to enable their survival, growth and proliferation[1]. At the core of this metabolic reprogramming is a shift towards macromolecular biosynthesis, based largely on the use of mitochondrial metabolites as anabolic precursors, and supported by changes in lipid synthesis, degradation and signaling[2‐4]. The attainment of a lipogenic phenotype, characterized by the increased dependence of cancer cells on de novo fatty acid (FA) synthesis, is typical of many cancer cells[2]. The transformed properties of tumor cells can also depend on lipolytic remodeling[3, 5] and FA oxidation[6‐10]. The biochemical mechanisms governing the transformations of lipid metabolism in cancer cells, in particular the relationships between lipid synthesis, storage and use, and their importance in the neoplastic process are still largely unknown. Identifying the factors responsible for the modulation of lipid metabolism and signaling in cancer is important for understanding the disease and for devising more rational preventive and therapeutic approaches.

Secreted phospholipases A₂ (sPLA₂s) are lipolytic enzymes that act on membrane glycerophospholipids to liberate free FAs (FFAs) and lysophospholipids by catalyzing the hydrolysis of their sn-2 ester bond[11]. These low-molecular mass, disulfide-rich and Ca²⁺-dependent enzymes are secreted from a variety of cells and act in autocrine or paracrine manners on cell membranes and other extracellular phospholipids, including lipoprotein particles, surfactant and dietary lipids, microbial membranes and microvesicles[12]. The nine active sPLA₂ enzymes known in humans display different tissue expression patterns and specific enzymatic preferences for binding to different types of phospholipid membranes, suggesting distinct biological roles for each sPLA₂[13, 14]. The multitude of cellular effects of the released FAs and lysophospholipids, and of their numerous bioactive metabolites, further explain their involvement in a variety of physiological processes and diseases, including lipid digestion and remodeling, acute and chronic inflammatory diseases, cardiovascular diseases, reproduction and host defense against infections[12]. Recent studies have implicated various sPLA₂s in cancer and metabolic disorders[15].

Aberrant expression of various sPLA₂s in cancer cells has been associated with the pathology of colorectal, breast, gastric and prostate cancers[16, 17]. The most studied group IIA sPLA₂ has been proposed to have a pro-tumorigenic role in prostate[18] and esophageal cancer[19], but an anti-tumorigenic role in gastric cancer[20]. Its role in colorectal cancer is still controversial[12, 16, 21, 22]. The involvement of sPLA₂s in cancer and other diseases has been investigated in relation to their ability to release arachidonic acid (AA) from cell membranes and stimulate, either directly or in coordination with the cytosolic group IVA PLA₂ (cPLA₂α), the production of eicosanoids, including the mitogenic prostaglandin E2 (PGE2)[12, 14, 23]. Several studies have suggested a tumor-promoting role for the group III and X sPLA₂s in colorectal cancer, based on their ability to stimulate PGE2 synthesis and cell proliferation[24, 25]. However, the human group X (hGX) sPLA₂ stimulates colon cancer cell proliferation by a mechanism that is dependent on the released FFAs and lysophospholipids, but not on its potent stimulation of PGE2 synthesis[26]. The underlying mechanisms of the action of hGX sPLA₂ and other sPLA₂ enzymes in different cancers are not known and confirmation of their functional contribution to tumorigenesis awaits further studies.

The group X sPLA₂ displays the greatest potency among mammalian sPLA₂s in hydrolyzing the phosphatidylcholine (PC)-rich extracellular leaflet of mammalian plasma membranes and of lipoprotein particles[12, 13]. Besides AA, the enzyme also releases numerous other monounsaturated and polyunsaturated FFAs, which could influence lipid metabolism and tumorigenesis in a variety of ways[3, 7, 12, 27, 28]. FFAs can be remodeled into membrane phospholipids, catabolized through mitochondrial FA oxidation or esterified into triacylglycerols (TAGs) and stored in lipid droplets (LDs)[29‐31]. Several enzymes regulating FFA availability through synthesis, such as fatty acid synthase (FAS) and acetyl-CoA carboxylase (ACC)[2], and through lipolysis[3, 5] have been clearly associated with cancer. In addition, there is increasing evidence for an important role for mitochondrial FA oxidation in tumorigenesis[6, 8‐10]. Interestingly, several recent reports have revealed that group X sPLA₂ affects lipid metabolism in various physiological and pathophysiological settings, including steroid hormone synthesis in adrenal glands[32], lipid digestion in the gut and diet-induced obesity[33]. Its recently proposed role in adipogenesis in mice has been associated with down-regulation of the expression of several genes important for lipid synthesis and adipogenesis, including sterol regulatory element-binding protein-1 (SREBP-1) and FAS[34]. Additionally, the group X sPLA₂ hydrolyzes serum low-density lipoprotein (LDL) and stimulates lipid accumulation and foam cell formation from macrophages[12]. The possible associations between sPLA₂s and basic lipid metabolism, such as fatty acid oxidation and synthesis, TAG synthesis and lipolysis, in the context of cell fate and tumorigenesis have, however, not been explored.

Altered lipid metabolism, including lipogenesis, β-oxidation and phospholipid remodeling, contributes to the transformed phenotype of breast cancer[2, 4, 35]. The involvement of sPLA₂s in breast cancer has not been studied, and there are only a few reports correlating the increased expression of group IIA sPLA₂ with advanced cancer and decreased patient survival[17, 36]. The aim of this study was to determine whether hGX sPLA₂ affects breast cancer cell growth and survival, and to delineate the underlying mechanism of action. We show for the first time that hGX sPLA₂ induces LD formation in the highly tumorigenic MDA-MB-231 breast cancer cells in an enzyme activity-dependent manner, thereby stimulating cell proliferation and significantly prolonging cell survival under serum deprivation-induced stress. Our results suggest that FFAs, in particular oleic acid (OA), released from membrane phospholipids by the action of hGX sPLA₂, are in large part responsible for LD biogenesis and cell survival. We also demonstrate that the mechanism of hGX-induced cell survival and lipid accumulation is associated with alterations in the expression of key lipogenic and β-oxidation enzymes, and modulation of AMP-activated protein kinase (AMPK) and protein B/Akt kinase signaling pathways. The pro-tumorigenic effects induced by hGX sPLA₂ were abolished by etomoxir, suggesting a critical role for β-oxidation in hGX-induced LD formation and cell survival in breast cancer cells.

We have demonstrated here that hGX sPLA₂-mediated phospholipid hydrolysis induces LD formation and alters lipid metabolism in triple negative breast cancer cells, stimulating their proliferation and prolonging cell survival during serum deprivation. Several mammalian sPLA₂s have been shown to stimulate cell proliferation in cancer cells[18, 19, 22, 25]. An anti-apoptotic role in growth factor-deprived cells has been shown for the group IIA in kidney fibroblasts[65], and for the group III and X sPLA₂s in neuronal cultures[66, 67]. The enzymatic activity-dependent mechanisms reported[18, 22, 65] have usually been associated with the pleiotropic actions of AA-derived eicosanoids[25] and rarely with lysophospholipids[26, 66]. The involvement of signaling pathways triggered by sPLA₂ receptors on the cell surface has been also suggested in some cases[68, 69]. However, the exact molecular mechanisms involved in the effects of sPLA₂ on cell fate have not been elucidated nor has the relevance of these activities in mammalian pathophysiology been delineated[14]. This is not surprising given the differential tissue expression patterns of sPLA₂s, the vast variety of extracellular target membranes and the plethora of bioactive products that are released from cell membranes in response to the action of sPLA₂s[12, 26]. hGX sPLA₂ has been shown to induce colon cancer cell proliferation by releasing a complex mixture of mitogenic FAs, lysophospholipids and eicosanoids, but, surprisingly, the proliferative effects of hGX sPLA₂ were not dependent on the mitogenic activity of AA-derived prostaglandin or LPC-derived LPA signaling[26]. Information about the possible involvement of sPLA₂s in the modulation of cellular metabolism, rather than direct bioactive lipid signaling, is only beginning to emerge[15]. Here we show that hGX sPLA₂ acts through the products of its hydrolysis and induces significant alterations in fatty acid metabolism and storage in breast cancer cells. These changes result primarily in prevention of serum withdrawal-induced cell death rather than in stimulation of cell proliferation. The previously reported mitogenic effects of sPLA₂s in colon cancer and other cells were also modest[18, 19, 22, 25, 26], suggesting that the positive effects of sPLA₂s on cell proliferation could, at least in some of these studies, be in fact a consequence of underlying changes in basic lipid metabolism and a pro-survival action, which is most evident under stressful conditions for the cell.

Several lines of evidence demonstrate that the effects of hGX sPLA₂ on breast cancer cells are dependent on its enzymatic activity. First, the effects of recombinant hGX sPLA₂ on MDA-MB-231 cell proliferation (Figure 1A), cell survival (Figure 1B) and LD formation (Figure 2B) were prevented by the potent sPLA₂ inhibitor varespladib. Secondly, it also prevented the effects of ectopically expressed hGX sPLA₂. The transient expression of its catalytically impaired H48Q mutant did not affect MDA-MB-231 cell proliferation or survival upon serum withdrawal (Figures 1C and1D). Further, the fact that varespladib, an inhibitor with low cell membrane permeability, completely prevents the actions of exogenous and ectopically expressed hGX argues for an extracellular action of the enzyme. Thirdly, while the potency of the recombinant mGX sPLA₂ to stimulate cell proliferation was similar to that of the human enzyme, its H48Q mutant did not induce a significant change in MDA-MB-231 cell proliferation rate (Additional file1: Figure S1B). Fourthly, two other sPLA₂ enzymes, hGV sPLA₂[13] and a neurotoxic snake venom sPLA₂, AtxA(V31W)[46], each with high activity on mammalian cell membranes, prevented cell death and induced LD formation in a varespladib sensitive manner. The hGV sPLA₂ enzyme was less effective than hGX in inducing these cellular effects, which is consistent with previous results showing a better ability of the latter enzyme to act on plasma membranes of mammalian cells and release free FAs, such as arachidonic acid[13, 45]. In contrast, the hGIIA sPLA₂ enzyme, known for its inability to bind to PC-rich membranes and act on intact mammalian cells[45], was unable to induce LD formation or prevent MDA-MB-231 cell death. These facts lead to the conclusion that the LD formation and prevention of cell death induced by hGX in MDA-MB-231 cells are dependent on the ability of hGX to bind to and hydrolyze phospholipid membranes.

Numerous studies have shown that hGX sPLA₂ is the most potent of the mammalian sPLA₂ enzymes in hydrolyzing PC-rich phospholipid vesicles, plasma membranes and lipoprotein particles, thus releasing large amounts of lysophospholipids and unsaturated FAs, including oleic, linoleic and arachidonic acids[12, 13, 26, 43]. Thus, lipoprotein particles are an important target for hGX sPLA₂ hydrolysis during cell culture in the presence of serum; however, the major source of lipid for hGX-induced LD generation in serum-deprived cells must be the cell membranes of MDA-MB-231 cells. hGX sPLA₂ may act directly on the plasma membrane of the cells and/or on microvesicles being actively released and recycled by MDA-MB-231 cells[70], as well as on apoptotic cells during starvation[71]. Regardless of the source of lipid, the results of this study indicate that of the products typically released upon hGX sPLA₂ membrane hydrolysis OA is largely responsible for the metabolic and signaling alterations that support its pro-tumorigenic effects. Exogenous OA is known to induce a PI3K/Akt-dependent proliferation, stimulate LD formation and prevent serum withdrawal-induced apoptosis in MDA-MB-231 cells[27, 28, 42, 47]. hGX sPLA₂ is shown in the present work to stimulate cell proliferation and increase the survival of serum-deprived MDA-MB-231 cells (Figures 1A and5C and5D). Further, exogenous hGX and OA are both shown to activate AMPK in proliferating cells (Figures 8A and8B), strongly suggesting that OA is one of the major mediators of the pro-tumorigenic effects of hGX. Importantly, the effects of OA are not restricted to breast cancer cells, since there is ample evidence that OA feeds into the TAG synthesis pathway and stimulates LD formation, cell growth and survival in different non-adipose cells, even channeling saturated FAs to TAGs to prevent their apoptotic effects[31, 55]. In cells exposed to excess lipids, the removal of FFAs through increased TAG accumulation and β-oxidation appears to be a general cellular response to the lipotoxic effects of FA overload[31]. Thus, besides promoting TAG synthesis, OA also prevented palmitate-induced apoptosis in skeletal muscle cells by stimulating β-oxidation through elevation of the expression of CPT1, activation of AMPK and repression of the activity of ACC[72]. Similarly, hGX significantly increased the levels of two important β-oxidation enzymes, CPT1A and VLCAD, in MDA-MB-231 cells (Figure 7), in parallel with the high rate of LD formation, activation of AMPK (Figures 8A and8B) and suppression of the induction of lipogenic enzymes, including ACC1 (Figure 7). Nevertheless, it is highly likely that, besides OA, other products of hGX phospholipid hydrolysis contribute to its effects in breast cancer cells, either by feeding metabolic pathways or by triggering cell signaling to various degrees[73]. Our results indicate that cPLA₂α activation and LPA signaling (Additional file5: Figure S4) are not important for the effects of hGX on MDA-MB-231 cells. However, the ability of rapamycin and indomethacin to partially suppress hGX-induced LD formation points to a possible role for AA in supporting LD formation through mTOR activation[49] and COX-dependent prostaglandin synthesis, respectively. Nevertheless, the contribution of AA-mediated signaling mechanisms to the changes in lipid metabolism induced by hGX sPLA₂ in MDA-MB-231 cells is clearly minimal. Altogether, the results presented in this study suggest that FFAs, in particular OA, liberated from membrane phospholipids by the enzymatic activity of hGX sPLA₂ are responsible for the observed alterations in lipid metabolism and the pro-survival effects induced by hGX in MDA-MB-231 breast cancer cells.

LDs, the intracellular neutral lipid storehouses until recently regarded as inert energy depots, are now regarded as complex organelles not only involved in the metabolic regulation of lipolysis and lipogenesis, but also in cell survival, apoptosis and cancer[3, 5, 31]. hGX sPLA₂ induced robust TAG synthesis and LD formation in proliferating MDA-MB-231 cells (Figures 2C and2E), but the effects on cell proliferation were modest (Additional file1: Figure S1A). On the other hand, although LD formation was less pronounced in serum-deprived cells, the increase in cell proliferation (Figure 1A) and, in particular, the reduction in apoptosis (Figure 1B) were more significant. This suggests a mechanism by which the formed LDs provide energy, building blocks or signaling molecules to sustain cell survival during energy stress[3, 5, 31]. Consistent with this, although the LDs accumulated in hGX-treated proliferating cells exhibited a minimal immediate proliferative effect (Figures 1A and Additional file1: Figure S1A), they conferred to the cells a marked survival advantage during long-term starvation in the absence of the sPLA₂ (Figure 3B). The hGX-induced LD accumulation was accompanied by increased levels of perilipin 2 mRNA, while a decrease in its transcriptional level was observed 24 h after the cells were switched to serum-free medium (Figure 7A). This is in line with its suggested role in promoting TAG accumulation and blocking lipolysis[3, 59], as well as with the reported correlation between TAG amount and perilipin 2 expression[57]. Since the transcription of β-oxidation genes was elevated almost in parallel with that of perilipin 2, it is conceivable that the FFAs released by hGX from membrane phospholipids are immediately partitioned between β-oxidation and TAG synthesis, which may contribute to cell survival by minimizing FFA toxicity[55]. However, since hGX-induced LDs were sufficient to prevent cell death in the absence of the sPLA₂ (Figure 3B), the FFAs released following LD lipolysis are probably also involved in the hGX-induced changes in cell metabolism and survival[3, 5]. Indeed, a cycle of FFA esterification and TAG lipolysis was required for FA-induced PPAR-mediated signaling responsible for mitochondrial gene expression and oxidative phosphorylation in cardiomyocytes[74]. Furthermore, PPAR activation by lipolytic FFAs modulated mitochondrial gene expression in brown adipose tissue, matching FA oxidation with supply[75]. In line with this, the hGX-induced alterations in gene expression were augmented when proliferating cells were switched to serum-free and sPLA₂-free medium (Figure 7A), suggesting that they form the basis for the metabolic adaptations that enable the positive effects of hGX on cell survival. Under these conditions, the pro-survival effects of the pre-formed LDs were abolished if high concentrations of etomoxir were used to block β-oxidation and LD breakdown (Figures 6D and6E), suggesting that TAG lipolysis followed by β-oxidation is critical for the pro-survival effects of hGX-induced LDs in MDA-MB-231 cells.

There is increasing evidence that CPT1 activity and β-oxidation contribute to the metabolic adaptations that enable cancer cell growth and survival[7]. Accelerated β-oxidation protects cancer cells from cell death induced by starvation or matrix detachment[6, 8, 10, 35] by contributing ATP and generating NADPH to counteract the accumulation of ROS during metabolic stress[7‐9, 35]. Furthermore, the ability of etomoxir to block the positive effect of hGX on cancer cell survival (Figure 6) is in line with recent studies showing that etomoxir-mediated inhibition of β-oxidation leads to a reduction in cancer cell proliferation and increased sensitivity to cell death[9, 35, 53]. Additionally, apoptosis-induced mitochondrial damage leads to LD formation due to inhibition of β-oxidation and increased de novo lipid synthesis[76]. The opposite alterations in FA oxidation and synthesis induced by hGX sPLA₂ in MDA-MB-231 cells may thus counteract the apoptosis-related changes and avert cell death. Therefore, the increased levels of CPT1A and VLCAD in hGX-treated cells, together with the ability of etomoxir to abrogate hGX-induced cell survival and induce cell death in starved MDA-MB-231 cells, strongly suggest that β-oxidation, and in particular CPT1 activity, is necessary for the positive effects of hGX on MDA-MB-231 cell proliferation and survival following serum withdrawal.

The central metabolic regulator AMPK responds to energy stress by suppressing ATP-consuming processes, including FA, cholesterol and TAG synthesis[61, 63], while stimulating ATP-producing processes, such as glycolysis, mitochondrial biogenesis and β-oxidation[3, 8, 61]. The acute effects of AMPK activation in most cell types include a direct inactivation of ACC, leading to suppression of FA synthesis, and also to a reciprocal stimulation of CPT1 activity and β-oxidation due to reduction in malonyl-CoA levels[61]. Reduced expression and activity of AMPK have been found in many cancers, including primary breast tumors[77]. A metabolic tumor suppressor role has been demonstrated recently for AMPK in lymphoma, where it negatively regulates the Warburg effect and limits cancer cell growth[78]. However, AMPK can also support cancer cell survival and invasiveness[8, 60], suggesting that its role in cancer is dependent on the cancer cell type and the pathophysiological context[8, 79]. In this study, we show that the activity of hGX sPLA₂ in invasive breast cancer cells leads to the activation of AMPK, suggesting that the kinase supports the pro-tumorigenic metabolic alterations induced by hGX sPLA₂. Elevated phosphorylation of AMPK was detected in hGX-treated cells after 48 h of cell proliferation (Figures 8A and8B) when neutral lipid accumulation reached maximal levels (Figure 2C) and the gene expression changes were significant (Figure 7A). Furthermore, etomoxir and triacsin C, which both attenuated hGX-induced LD formation, also prevented hGX-induced AMPK activation (Figures 8A and8B). This suggests that the energy stress caused by rapid cell growth and proliferation combined with extensive FA activation, TAG synthesis and LD biogenesis in hGX-treated MDA-MB-231 cells leads to AMPK activation[29]. Accordingly, by mimicking cellular low energy status and inducing a several-fold higher increase in the level of phosphorylated AMPK relative to hGX (Figures 8A and8B), the AMPK activator AICAR completely prevented hGX-induced LD formation (Figure 8C). This is consistent with the previously reported strong cytostatic effect of AICAR on MDA-MB-231 cells caused by suppression of DNA, protein and lipid synthesis[64]. It is thus possible that one of the important roles of AMPK in hGX-treated cells is to restore the energy balance by preventing further LD formation, by suppressing TAG synthesis, by phosphorylating glycerol-3-phosphate acyltransferase (GPAT)[63], and by stimulating lipolysis, presumably by activating adipose triglyceride lipase (ATGL/PNPLA2)[3, 62], as well as β-oxidation. Apart from these immediate effects on lipid metabolism, the observed long-lasting transcriptional adaptations induced by hGX in MDA-MB-231 cells (Figure 7A) could also be mediated by AMPK. Namely, AMPK blocks SREBP-1 activity by direct phosphorylation[80] or through inhibition of mTOR[50], thus suppressing the transcription of its target genes, including ACACA, FASN and SCD[4], but also lowering SREBP-1 expression by reducing its auto-loop regulation[80]. Additionally, hGX-released polyunsaturated FAs may directly suppress the expression of SREBP-1 and its target genes[34], including FASN and SCD, whose inhibition has been shown to induce AMPK activation[4]. Also, elevated AMPK activity may induce the expression and activity of peroxisome proliferator-activated receptor-γ co-activator 1α (PGC1α)[7, 61] to stimulate mitochondrial biogenesis and the transcription of β-oxidation genes, such as those encoding CPT1A and VLCAD. Similarly to the effects of hGX in MDA-MB-231 cells, increased rates of β-oxidation associated with AMPK phosphorylation, elevation of CPT1A mRNA and a decrease in lipogenesis due to inactivation of ACC have recently been implicated in the adipocyte-induced survival and metastasis of ovarian cancer cells[60]. Importantly, it has been shown that activation of AMPK in cancer cells during energy stress enables cell survival by blocking lipid synthesis through inactivation of ACC and elevating β-oxidation-dependent NADPH production to restore the redox balance[8]. Our results indicate that AMPK activation also supports survival of MDA-MB-231 cells, since AICAR displayed a strong anti-apoptotic effect in these cells (Figure 8E). Thus, the activation of AMPK by hGX in proliferating cells implicates AMPK in the coordination of the adaptation of MDA-MB-231 cell metabolism to the FAs derived from hGX membrane hydrolysis. Its association with the hGX sPLA₂-induced LD formation and cell survival, however, remains to be confirmed.

Our results with etomoxir and bezafibrate, modulators of β-oxidation, suggest that β-oxidation supports the process of hGX-induced LD biogenesis in MDA-MB-231 cells, regardless of their metabolic and proliferative status (Figures 6A and6C). It is, however, not clear how β-oxidation can support LD formation. Presumably, elevated β-oxidation may provide ATP and NADPH[7] for the energetically expensive process of LD formation[29], which, besides TAG synthesis, also requires alterations in FA, cholesterol and phospholipid synthesis and remodeling[30]. Although the simultaneous activity of FA synthesis and oxidation is controversial[7], a high β-oxidation flux could contribute to the cytosolic pool of acetyl-CoA molecules for de novo FA synthesis. Thus, despite the increased level of FFAs released by the sPLA₂ from phospholipids and from TAGs through lipolysis, a low level of FA synthesis is probably still necessary for maintaining the proper FA composition of cell membranes and the membranes of LDs, in particular in proliferating cells[1, 30]. Additionally, hGX may stimulate a cycle of FA esterification and lipolysis, as suggested for OA in MDA-MB-231 cells[29, 42]. Since FA/TAG cycling requires high ACS activity, at the expense of ATP, to provide a continuous supply of FA-CoA, it may also contribute to the observed hGX-induced activation of AMPK[29]. In line with this, besides etomoxir, the ACS inhibitor triacsin C also partially blocked hGX-induced LD formation (Additional file5: Figure S4) and AMPK activation (Figures 8A and8B) in proliferating cells. We may thus speculate that, by supplying FFAs, hGX stimulates β-oxidation that in turn supports the anabolic branch of FA/TAG cycling, resulting in net LD accumulation and thus filling the LD energy reserves that can be used to support cell survival. Interestingly, recent studies revealing that mitochondria form contact sites with nascent LDs and participate in phospholipid and TAG synthesis during their biogenesis[30] are in line with a possible association between β-oxidation and LD formation. It is thus likely that hGX sPLA₂ modulates the balance between the catabolic branch of glycerolipid metabolism, including TAG lipolysis and β-oxidation, and the anabolic processes, such as de novo FA synthesis, phospholipid remodeling and TAG synthesis. sPLA₂ phospholipid hydrolysis, which may feed a variety of lipids into both branches, would thus induce metabolic alterations that lead to net LD accumulation and enable the pro-survival activity of hGX in MDA-MB-231 cells during prolonged serum deprivation.

The metabolic transformations induced by hGX sPLA₂ in the highly invasive breast cancer cells, that include increased accumulation of cytosolic LDs, up-regulated β-oxidation and suppressed lipogenesis, resemble the effects of omental fat pad adipocytes that provide lipids for ovarian cancer cells to enable their growth and survival at the metastatic site[60]. Interestingly, when MDA-MB-231 and T-47D cells were exposed to the same primary human fat pad adipocytes they also accumulated large amounts of LDs and displayed increased invasive properties[60]. This suggests that the effects of hGX sPLA₂ identified in this study could be pathophysiologically relevant. hGX sPLA₂ may be secreted not only from breast cancer cells, but also from different cells in the tumor microenvironment, including inflammatory cells[12] and adipocytes[34], at primary tumor sites or at lipid-rich metastatic sites. It may then act in an autocrine or paracrine manner on cellular and extracellular phospholipids[12] to alter the availability of FFAs and induce metabolic transformations in cancer cells to support their survival, growth and metastatic potential. Furthermore, alterations in lipid metabolism and lipid accumulation within LDs in non-adipose tissue have been recognized as a major risk factor for the development of cancer and also other chronic diseases, such as metabolic syndrome, cardiovascular disease and diabetes[4, 31]. Thus, the present study raises the possibility that modulation of cellular lipid metabolism by hGX and other sPLA₂s may also contribute to some of these debilitating diseases.

Several sPLA₂s have been previously shown to affect the fate of cancer and other cells, however, their mechanisms of action at the cellular level are still unclear, have not been causally linked to eicosanoid or other lipid signaling, and have never been related to lipid droplets or alterations in basic lipid metabolism. We show in this study that hGX sPLA₂, through the products of its enzymatic activity, induces LD formation and alters lipid metabolism in triple negative breast cancer cells, stimulating their proliferation and prolonging cell survival during growth factor deprivation. We provide evidence that the pro-tumorigenic effects of hGX are associated with activation of AMPK, suppression of lipogenesis and activation of β-oxidation, which is critical for the survival of hGX-treated MDA-MB-231 cells, most probably by contributing to the restoration of the energy and redox balance[1, 7]. The results also suggest the intriguing possibility that hGX-induced elevated β-oxidation also supports the anabolic branch of FA/TAG cycling, leading to a net LD accumulation that in turn enables prolonged cell survival. Finally, the ability of hGX sPLA₂ to act as a modulator of basic lipid metabolism and cancer cell survival is established. This could have important implications in elucidating the role of hGX and other sPLA₂s, such as hGV and hGIII, in cancer and human pathophysiology in general.