Background
The β-galactoside α2-6-sialyltransferase ST6Gal-I catalyzes the addition of the negatively-charged sugar, sialic acid, to the termini of
N-linked glycans on selected cell surface or secreted proteins as they transit through the Golgi. ST6Gal-I elaborates an α2-6 linkage of sialic acid to galactose, and this enzyme appears to be the primary sialyltransferase responsible for this modification in most tissues [
1,
2]. Depending on the specific substrate targeted by ST6Gal-I, α2-6 sialylation can modulate protein conformation, oligomerization and/or receptor internalization (reviewed in [
3]). Another important function of α2-6 sialylation is to negatively regulate certain galectin-dependent cell responses [
4]. Galectins are lectins that bind galactose-containing glycans, and the addition of α2-6 sialic acid to galactose impedes the ability of most galectins to bind their targets [
4]. Given that many glycoprotein receptors are held on the cell surface through an interaction with the extracellular galectin lattice [
5‐
7], ST6Gal-I-mediated sialylation can block glycoprotein binding to the lattice, causing receptor internalization. Conversely, α2-6 sialylation enhances the surface retention of other types of receptor glycoproteins [
8], albeit through mechanisms not well-defined. These observations suggest that ST6Gal-I may play a role in regulating the complement of receptors on the cell surface, in addition to modulating the function of distinct glycoproteins through effects on receptor conformation and/or clustering.
ST6Gal-I is overexpressed in many different types of cancers including ovarian, breast, and colon carcinoma (reviewed in [
3,
4]), and ST6Gal-I upregulation is driven by oncogenic ras [
9,
10]. Elevated expression of ST6Gal-I has been correlated with a negative patient prognosis in breast and colorectal cancers [
11,
12]. Cell culture studies suggest that ST6Gal-I promotes cell migration and invasion, at least in part through altering the sialylation and function of the β1 integrin [
13‐
15]. More recently ST6Gal-I has also been identified as an inhibitor of several cell death pathways. For example, one important function of extracellular galectins is to induce apoptosis, and this activity is blocked by ST6Gal-I mediated sialylation of galectin substrates [
16‐
18]. Additionally, our group has shown that sialylation of the Fas and TNFR1 death receptors by ST6Gal-I hinders apoptotic signaling in response to their respective ligands, FasL and TNFα [
8,
19]. Finally, ST6Gal-I activity is associated with resistance to radiation treatment [
20].
In view of ST6Gal-I’s upregulation in cancer, as well as its emerging role as an inhibitor of cell death pathways, we investigated whether ST6Gal-I activity could influence the sensitivity of tumor cells to cisplatin. Cisplatin is the parent compound of the platinum family of chemotherapeutics commonly used in frontline ovarian cancer treatment. Cisplatin and other platinum derivatives (e.g., oxaliplatin, carboplatin) function by forming inter- and intra-strand crosslinks in DNA, leading to an apoptotic cell death. Resistance to platinum drugs represents a major treatment challenge in ovarian and other cancers. The vast majority of ovarian cancer patients have an initial response to platinum compounds, however up to 75% of patients will relapse, with most exhibiting drug resistant disease [
21]. The molecular events underlying resistance are complex, and it is likely that different tumor cells exhibit different mechanisms, or combinations of mechanisms, to escape cisplatin-induced apoptosis. At present, investigations into the mechanisms of tumor cell resistance to platinum agents have focused on drug import or export [
22], cytosolic inactivation (e.g. by glutathione and other antioxidants) [
23], compensatory DNA repair [
24], and defects in apoptotic signaling [
25]. The activation of caspases following DNA damage is important for cisplatin-induced cell death, therefore factors impinging on caspase activity can influence drug efficacy. As well, cisplatin may elicit cytotoxicity through mechanisms independent of DNA damage, as cisplatin is known to bind many molecules other than DNA, and can also modulate cytoskeletal structure [
26]. In the current study we describe a new mechanism for cisplatin resistance involving α2-6 sialylation of glycoproteins by the ST6Gal-I sialyltransferase.
Methods
Cell lines
The Pa-1 ovarian cancer cell line was purchased commercially through ATCC (Manassas, VA). Pa-1 cells were cultured and grown in Dulbecco’s eagle's minimal essential medium (DMEM) with 4.5 g glucose supplemented with 10% fetal bovine serum (FBS)(Hyclone) and 1% antibacterial/antimycotic solution containing penicillin, streptomycin, and amphotericin B (Invitrogen). Pa-1 cells were previously found to express high endogenous levels of ST6Gal-I [
13]. To examine the effects of ST6Gal-I expression on cell response to cisplatin treatment a shRNA construct targeting ST6Gal-I as well as an empty vector control were introduced via a lentiviral vector (empty vector and shRNA-expressing lentiviral particles were purchased from Sigma). Pa-1 empty vector (EV) and ST6Gal-I shRNA-mediated knockdown (sh.ST6) lines are stable, polyclonal cell populations initially selected by puromycin at a concentration of 10 μg/ml, and then maintained in 0.5 μg/ml puromycin. The OV4 ovarian cancer cell line was a generous gift from Dr. Timothy Eberlein (Harvard, Cambridge, MA). OV4 cells were cultured and grown in Dulbecco's modified Eagle's MEM/Ham's F-12 50:50 (DMEM/F12) supplemented with 10% FBS and 1% antibiotic/antimycotic solution. OV4 cells lack detectable endogenous ST6Gal-I expression and we previously forced ST6Gal-I expression and an empty vector control by lentiviral transduction (MOI = 3) [
13]. Stable, polyclonal populations were isolated through puromycin selection. A2780ip2 and A2780cp20 cell lines were generous gifts from Dr. Anil Sood (MD Anderson Cancer Center). Lines were maintained in RPMI media (Cellgro) supplemented with 10% FBS and 1% antibiotic/antimycotic solution. A2780cp20 cells represent a cisplatin-resistant derivative cell line of A2780ip2 created by repeated cisplatin exposure as previously described [
27].
Immunofluorescence imaging
Cells were seeded onto 4-well chamber microscope slides (Beckin Dickinson) and allowed to adhere overnight. Cells were washed and fixed in 4% paraformaldehyde for 10 minutes followed by permeabilization in 5% Triton X-100 (in PBS) for 5 minutes. Cells were then incubated overnight at 4°C or 3 hr at room temperature with the ST6Gal-I antibody (polyclonal, R&D Systems, catalog # AF5924) and 3 hr at room temperature with anti-Golgi Matrix-130 (GM-130) (monoclonal, BD Transduction Laboratories). Following incubation with primary antibody, cells were washed and incubated with anti-goat Alexa-Fluor 594-conjugated or anti-mouse Alexa-Fluor 488 secondary antibody (Molecular Probes) for 30 minutes. Chambers were removed and DAPI-containing mounting solution, Vectamount (Vector Labs), was placed onto each well. Coverslips were added and the slides visualized under a Nikon Eclipe 80i fluorescence microscope fitted with a Photometrics CoolSNAP camera (Roper Scientific). Images were analyzed on NIS elements software. ST6Gal-I co-localization with GM-130 was imaged by confocal microscopy.
Western blot
Prior to lysis cells were grown in puromycin free media for at least one day, passaged normally, and plated onto 6 well tissue culture plates (Fisher) at a density of 7.5 × 105 cells per well. Cells were allowed to adhere overnight and then treated with cisplatin. Cells were lysed on ice in 50 mM Tris-HCl (pH 7.4) containing 1% Triton X-100, and a protease inhibitor cocktail (Roche Applied Bioscience). Cell lysate was kept on ice for 40 minutes vortexing regularly or lysates were sonicated using a sonicator model C-18 (Fisher). Protein concentrations of the lysates were determined using a modified Bradford Assay (Sigma). Proteins were then resolved by SDS-PAGE, and transferred to polyvinylidene difluoride membranes. Membranes were blocked with 5% nonfat dry milk in TBS containing 0.1% Tween 20 (TBST). Primary antibodies against ST6Gal-I (R&D Systems) or cleaved caspase-3 (Cell Signaling) were added to the membrane. Membranes were then washed and incubated with horseradish peroxidase-coupled secondary antibody (Amersham) and visualized with Immobilon enhanced chemiluminscence reagent (Millipore). Protein loading was evaluated by immunoblotting for either β-tubulin or β-actin (Cell Signaling).
Cell viability assay
Cells were plated in opaque-sided 96-well plates (Corning) at a density of 104 cells per well in 75 μl of media and allowed to adhere overnight. Cisplatin stock solutions were made by dissolving solid cisplatin in distilled water to a concentration of 2.5 mM and stored at 4°C with new solutions made monthly. On the day of each experiment, stock solutions were used to dilute cisplatin into media to obtain the desired concentrations, and then cells were grown in the cisplatin-containing media for 21 hours (Pa-1) or 24 hours (OV4). Cell viability was evaluated by determining ATP content using the CellGlo ATP quantification kit (Promega) following the manufacturer’s protocol. Luminescence was measured on a Synergy 2 plate reader (Biotek).
Cell selection with cisplatin
The polyclonal Pa-1 sh.ST6 population contains stable clones with varying levels of ST6Gal-I knockdown. Two cell flasks were grown in parallel; one with DMEM containing 10% FBS and 1% antibiotic/antimycotic solution (control), and the other with this media supplemented with 1 μM cisplatin. After a 3-week interval, greater than 90% of the cells grown in cisplatin-containing media had died, whereas the control population proliferated over this interval. At the end of the 3 week incubation in cisplatin, the remaining viable cells were resuspended in 10% FBS/DMEM lacking cisplatin, and cultures were expanded to gain a sufficient number of cells for lysis and western blot analysis.
Discussion
Despite the clear clinical significance of chemotherapy resistance, a single mechanism of resistance has not yet been established for all cases. The multifactorial nature of tumor cell resistance to cisplatin leaves open the possibility of novel mechanisms that remain undiscovered. In this study we show that forced expression of ST6Gal-I confers resistance to cisplatin, whereas ST6Gal-I knockdown conversely sensitizes cells to cisplatin. Furthermore, cells selected for resistance to cisplatin exhibit an upregulation in endogenous ST6Gal-I protein, suggesting that increased receptor α2-6 sialylation may provide tumor cells with a survival advantage. These findings illuminate a new mechanism for chemoresistance, and underscore the importance of the cellular glycosylation machinery in drug response. An aberrant glycan profile was one of the earliest identified characteristics of a cancer cell, and selective enrichment in α2-6 sialylation (relative to α2-3 sialylation), is a common feature of transformed cells [
3]. It is also known that platinum drug-resistant cells have abnormal glycosylation [
29,
30], and studies spanning more than two decades indicate that cisplatin treatment alters the sialic acid content of tumor cells [
31‐
33]. The link between these glycosylation changes and ST6Gal-I is currently unclear, however the inhibitory effect of ST6Gal-I on cisplatin-induced cell death is likely driven by the activity of variantly-sialylated surface receptors, given that ST6Gal-I modifies glycoproteins bound for the plasma membrane or secretion (and not cytosolic proteins). Interestingly, tumors expressing activating ras mutations [
34] or ras overexpression [
35] are typically resistant to cisplatin, and ST6Gal-I is one of the targets upregulated by ras signaling [
9,
10]. Furthermore, we recently reported that high ST6Gal-I expression correlates with expression of the cancer stem cell markers ALDH1 and CD133, suggesting that ST6Gal-I activity may contribute to stem-like cell behaviors including chemoresistance [
36].
One predominant surface receptor known to modulate cisplatin sensitivity is the Fas death receptor. Fas is activated by binding to FasL, which in turn causes receptor internalization, formation of the Death Inducing Signaling Complex (DISC), followed by activation of apoptotic caspases. Caspase activation is also a critical downstream event following cisplatin-induced DNA damage, and cisplatin-resistant cells exhibit attenuated activation of caspases 3, 8 and 9 [
35]. Cisplatin is reported to cause clustering and activation of the Fas receptor in a ligand independent manner [
36], as well as increased Fas expression [
37‐
43]. Additionally, cisplatin stimulates the aggregation of Fas into lipid rafts [
44,
45], which is correspondingly important for Fas internalization and apoptotic signaling [
46]. In mice with subcutaneous tumors formed from syngeneic Lewis lung carcinoma cells, one intraperitoneal dose of cisplatin induced a dramatic increase in Fas expression in the tumors, and also stimulated tumor regression [
43]. In this same study, the anti-tumor effects of cisplatin were abrogated in mice deficient in FasL. These results implicate cisplatin-induced Fas upregulation in promoting tumor cell death [
43], and further suggest that in order to acquire cisplatin-resistance, tumor cells may evolve mechanisms to disable Fas signaling. Our prior studies demonstrated that Fas is a ST6Gal-I substrate, and that increased α2-6 sialylation of Fas functions to inhibit Fas receptor internalization and DISC formation [
8], effectively shutting off Fas apoptotic signaling. Hence, α2-6 sialylated Fas isoforms could play a part in cisplatin resistance.
Another potential mechanism for ST6Gal-I-mediated cisplatin-resistance may involve the differential sialylation of one or more drug transporters. Many cisplatin-resistant cell lines show reduced accumulation of cisplatin [
47], pointing to dysfunctions in cell surface transporters that control either drug uptake or efflux. Defective glycosylation of ATP binding cassette (ABC) transporters has been suggested to contribute to cancer development, and possibly, chemoresistance [
48]. Liang et al. reported that in epidermoid carcinoma cells selected for resistance to cisplatin, the MRP1 transporter (also known as ABCC1) was aberrantly glycosylated, and this was associated with mislocalization to intracellular compartments and reduced cell surface expression [
29]. Similarly, altered N-glycosylation of MRP1 and MRP4 was correlated with cisplatin and oxaliplatin resistance in ovarian cancer cells [
49]. In this latter study,
N-glycosylation defects were linked to reduced levels of two glycosyltransferases: (i) N-acetylglucosamine-1-phosphate transferase, gamma subunit (GNPTG) and (ii) mannosyl (alpha-1,6)-glycoprotein beta-1,6-N-acetyl-glucosaminyltransferase (MGAT5) [
49].
N-glycans are also known to be crucial for the stability of the ABCG2 transporter in the endoplasmic reticulum [
30,
50]. These findings indicate the importance of glycosylation in transporter function, and suggest that studies of variant transporter sialylation may be a fruitful area for future research.
Although the mechanisms underlying the effects of ST6Gal-I activity on cisplatin sensitivity are not yet understood, the current study adds to the body of literature implicating this enzyme as a major contributor to tumor cell survival. In addition to conferring cisplatin resistance, ST6Gal-I-mediated receptor sialylation blocks apoptotic signaling by the Fas [
8] and TNFR1 [
19] death receptors, and also inhibits galectin-induced cell death [
16‐
18]. Taken together, these results suggest that ST6Gal-I may be a promising clinical target, and that inhibition of ST6Gal-I expression or activity could be employed to sensitize tumor cells to platinum drugs, increasing therapeutic efficacy.
Competing interests
The authors have no conflicts to disclose.
Authors’ contributions
MJS developed the methodology, acquired and interpreted data, and drafted the manuscript. AFS and JWW acquired data. ESS acquired data and aided in study design. CNL conceived and designed the study. SLB conceived and designed the study, developed the methodology, interpreted data, edited the manuscript, and oversaw the study. All authors have read and approved the final manuscript.