Abstract
Latent membrane protein 1 (LMP1) is the major transforming protein of Epstein-Barr virus (EBV) and is critical for EBV-induced B-cell transformation in vitro. Several B-cell malignancies are associated with latent LMP1-positive EBV infection, including Hodgkin and diffuse large B-cell lymphomas and HIV and post-transplant lymphoproliferative disorders. We have previously reported that promotion of B cell proliferation by LMP1 coincided with an induction of aerobic glycolysis. To further examine LMP1-induced metabolic reprogramming in B cells, we ectopically expressed LMP1 in an EBV-negative Burkitt lymphoma cell line preceding a targeted metabolic analysis. This analysis revealed that the most significant LMP1-induced metabolic changes were to fatty acids. In parallel, the same metabolic analysis was carried out to compare metabolic alterations in primary B cells following EBV-mediated B-cell growth transformation, which also revealed significant changes to fatty acid levels following establishment of lymphoblastoid cell lines (LCLs). Ectopic expression of LMP1 and EBV-mediated B-cell growth transformation induced fatty acid synthase (FASN) and increased lipid droplet formation. FASN is a crucial lipogenic enzyme responsible for de novo biogenesis of fatty acids in transformed cells. Furthermore, following treatment with C75, an inhibitor of lipogenesis, we observed preferential killing of LMP1-expressing B cells. Our findings demonstrate that ectopic expression of LMP1 and EBV-mediated B-cell growth transformation leads to induction of FASN, fatty acids and lipid droplet formation, possibly pointing to a reliance on lipogenesis. Therefore, the use of lipogenesis inhibitors could potentially be used in the treatment of LMP1+ EBV associated malignancies by targeting a LMP1-specific dependency on lipogenesis.
Importance Despite many attempts to develop novel therapies, EBV-specific therapies currently remain largely investigational and EBV-associated malignancies are often associated with a worse prognosis. Therefore, there is a clear demand for EBV-specific therapies for both prevention and treatment. Non-cancerous cells preferentially obtain fatty acids from dietary sources whereas cancer cells will often produce fatty acids themselves by de novo lipogenesis, often becoming dependent on the pathway for cell survival and proliferation. LMP1 and EBV-mediated B-cell growth transformation leads to induction of FASN, a key enzyme responsible for the catalysis of endogenous fatty acids. Preferential killing of LMP1-expressing B cells following inhibition of FASN suggests that the targeting LMP-induced lipogenesis programs could be an effective strategy in treating LMP1-positive EBV-associated malignancies. Importantly, targeting unique metabolic perturbations induced by EBV could be a way to explicitly target EBV-positive malignancies and distinguish their treatment from EBV-negative counterparts.
Introduction
The Epstein-Barr virus (EBV) is a double-stranded DNA human gammaherpesvirus that latently infects approximately 95% of the population worldwide (1). EBV was the first human tumor virus identified (2) and contributes to about 1.5% of all cases of human cancer worldwide (3). Latent membrane protein 1 (LMP1) is expressed in the majority of EBV-positive cancers, including: Hodgkin’s and diffuse large B-cell lymphomas; HIV and post-transplant lymphoproliferative disorders; nasopharyngeal and gastric carcinomas (11). In vitro, EBV is able to convert primary B-cells into immortalized lymphoblastoid cell lines (LCLs), and the EBV oncoprotein LMP1 is critical for this process (4, 5). LMP1 is a transmembrane protein containing two signaling domains: C-terminal-activating region 1 and 2 (CTAR1 and CTAR2). Through these two domains, LMP1 can mimic CD40 signaling to activate nuclear factor-κB (NF-κB), phosphoinositide 3-kinase (PI3K)/AKT and Ras – extracellular signal-regulated kinase (ERK) – mitogen-activated protein kinase (MAPK) pathways (6). The activation of these signaling pathways by LMP1 contribute to its ability to transform cells by altering the expression of a wide range of host gene targets (7). LMP1 has also been shown to promote aerobic glycolysis and metabolic reprogramming in B cell lymphomas and nasopharyngeal epithelial cells (8–14). The transition from a resting B-cell to a rapidly proliferating cell following EBV infection, and the presence of EBV in associated malignancies, entails major metabolic changes. The role of LMP1 in these processes is incompletely understood. To further examine LMP1-induced metabolic reprogramming in B cells, we ectopically expressed LMP1 in an EBV-negative Burkitt’s lymphoma cell line preceding a targeted relative quantitation of approximately 200 polar metabolites spanning 32 different classes. The top metabolites induced by LMP1 were fatty acids. In parallel, the same metabolic analysis was carried out to compare metabolic changes in primary B cells following EBV-mediated B-cell growth transformation, which also revealed large changes in fatty acid levels.
Aerobic glycolysis is a well-established phenotype in cancer cells, and even though deregulated lipid metabolism has received less attention, it is just as ubiquitous as a hallmark of cancer (15). Non-transformed cells will preferentially obtain fatty acids from dietary sources for their metabolic needs versus de novo lipid synthesis (lipogenesis). However, despite access to these same dietary sources, cancer cells will often preferentially rely on endogenous fatty acids produced by de novo lipogenesis, often becoming dependent on the pathway for cell survival and proliferation. Fatty acids are essential for these processes as they are used as substrates for oxidation and energy production, membrane synthesis, energy storage and production of signaling molecules. Fatty acid synthase (FASN) is responsible for the catalysis of endogenous fatty acids and therefore is commonly upregulated in cancer cells (16–18). FASN condenses malonyl-CoA with acetyl-CoA, using NADPH as a reducing equivalent, to generate the 16-carbon fatty acid palmitate (19). In addition, upregulated glycolysis has been suggested as a mechanism for generating intermediates for fatty acid synthesis (20, 21). Once fatty acids are made, they can be converted to triglycerides and stored as lipid droplets for cellular energy storage (16). Lipid droplets can also contain phospholipids and sterols for membrane production (22).
There are two main pathways that transformed cells use to upregulate FASN, found at the levels of both transcription and post-translation. In the first case, FASN expression can be stimulated by the transcription factor sterol regulatory element-binding protein 1c (SREBP1c), which binds to and activates sterol regulatory elements (SREs) in the promoter region of FASN and other genes involved in lipogenesis (23, 24). SREBP1c is an isoform of the SREBF1 gene, which transcribes the two splice variants, SREBP-1a and SREBP-1c, that are encoded from alternative promoters and differ in their NH2-terminal domains (25). SREBP1c activity by controlled by its expression and/or nuclear maturation following hormone and nutritional regulation which converges upon PI3K/AKT, and ERK/MAPK signaling cascades (26, 27), amongst others. Nuclear maturation of SREBP1c is achieved through extensive regulation by proteolytic cleavage of it NH2 terminal domain in the endoplasmic reticulum (ER) membrane and Golgi. Firstly, SREBPs are retained in the ER membrane by Insulin-induced gene (Insig) where they associate with cleave activating protein (SCAP) when cellular sterol levels are abundant. Once sterol levels drop, SCAP is released allowing Insig to traffic SREBPs to the Golgi. Once there, SREBPs are cleaved by site-1 and site-2 proteases (S1P and S2P). This results in the release of the mature transcriptionally active N-terminus and enables its subsequent translocation into the nucleus where it can bind to SREs and transcribe lipogenesis genes, including FASN (28). At the post-translational level, increased FASN protein levels can be obtained through interaction with ubiquitin-specific peptidase 2a (USP2a), a ubiquitin-specific protease that can stabilize FASN by removing ubiquitin from the enzyme (29). These two main methods of FASN regulation do not have to be mutually exclusive, it is also possible that they concurrently take place in cancer cells.
In this study, we determined that ectopic expression of LMP1 and EBV-mediated B-cell growth transformation leads to induction of FASN, fatty acids and lipid droplet formation. This points to a potential reliance on lipogenesis as demonstrated by preferential killing of LMP1-expressing B cells following inhibition of lipogenesis. It is therefore conceivable that use of lipogenesis inhibitors could play a role in the treatment of LMP1+ EBV associated malignancies by targeting LMP-induced metabolic dependencies.
Results
Fatty acids are the top metabolites increased by LMP1
To identify cellular metabolites that can be altered by LMP1, we first ectopically expressed LMP1 in the EBV-negative Burkitt’s lymphoma cell line DG75. Cells were transduced with retroviral particles containing either pBABE (empty vector) or pBABE-HA-LMP1 (LMP1) vectors as described previously (9). Using this cell system, we then undertook a targeted approach to determine the relative quantities of approximately 200 polar metabolites spanning 32 different classes to examine LMP1-induced metabolic changes. These changes are summarized by heat map and principal component analysis (PCA) (Fig 1A and 1B). This analysis used mass spectrometry followed by hydrophilic interaction chromatography (HILAC). Peak areas, representing metabolite levels, were extracted using ThermoScientific Compound Discoverer 3.0. Metabolites were identified from a provided mass list, and by MS/MS fragmentation of each metabolite follow by searching the mzCloud database. Significant differences (q-value < 0.05) in proteins of least 1.5-fold between empty vector (pBABE) and LMP1 conditions (based on average value of the triplicate sample) were indicated as ‘True’ changes (the list of all affected metabolites is provided in the supplemental table 1). When comparing pBABE vs LMP1 cell lines and sorting fold change of metabolites in descending order, the top 13 ‘True’ metabolites (confirmed using pure compounds) induced by LMP1 were fatty acids. These fatty acids were largely saturated medium and long chain and were increased from 2.64 to 36.42-fold change (Fig 1C). Previously, we have shown Poly(ADP-Ribose) Polymerase 1 to be important in LMP1-induced aerobic glycolysis and accelerated cellular proliferation using the PARP inhibitor olaparib (9). Therefore, we included an olaparib treatment group in our metabolic analysis to examine whether PARP inhibition could offset LMP1-induced changes to cellular metabolites. When we sorted the fold change of metabolites between pBABE vs LMP1 in descending order as described above, we found a perfect inverse correlation in the vast majority of the fatty acids in our LMP1 untreated vs LMP1 olaparib group. In other words, 11 of the 13 fatty acids fatty acids that were most increased with ectopic expression of LMP1 were also the most decreased when these same cells were then treated with olaparib. These significant fold changes were in the range of −1.89 to −3.64 (Fig 1C). This may partly explain the ability of olaparib to blunt the proliferative advantage bestowed by LMP1 that we previously reported (9).
LMP1 induces FASN and lipogenesis
Due to the fact that fatty acids were the dominate metabolite class induced by LMP1, we sought to pursue a potential enzyme responsible. FASN catalyzes de novo lipogenesis and is commonly upregulated across many different cancers (16–18). Furthermore, a recent study demonstrated that LMP1 is able to upregulate FASN and lipogenesis in NPC (30). We therefore wanted to determine if LMP1 could induce FASN and lipogenesis in B cells. Using western blotting, we showed that ectopic expression of LMP1 increased FASN protein levels around 2.5-fold as compared to empty vector control (Fig 2A and 2B). To determine if the LMP1-mediated increased fatty acids and FASN levels were inducing lipogenesis, we employed Nile Red staining, a potent and specific lipid droplet stain. Lipid droplets are small cytoplasmic organelles that can store fatty acids, providing available energy as well as cellular membrane material (16). Under serum-deprived conditions, we stained pBABE and LMP1 cells with Nile red followed by FACS analysis. We found that LMP1 led to an increase in Nile Red staining (Fig 2C), which was then further quantified using a fluorescent plate reader (Fig 2D). The somewhat modest increases in FASN and lipid droplet formation should be viewed in the context of the BL background used for the ectopic expression of LMP1, as alteration in lipid metabolism has been noted to be a feature of BL and most likely blunted the effect of LMP1 (31).
EBV-immortalization of B cells leads to significant increases in metabolic cofactors and fatty acids
Our initial analysis into LMP1-mediated metabolic changes revealed that fatty acids were the major metabolites increased. However, we wanted to extend our examination of LMP1’s role in metabolic remodeling of the cell in the broader context of EBV-immortalization of B cells. To do this, using the same metabolic analysis as described for ectopic expression of LMP1, we infected primary B cells with EBV, resulting in their transformation into LCLs, a process in which LMP1 is critical (4, 5). Both primary B cells and their corresponding matched LCLs (60dpi) were extracted for metabolite analysis (the list of all affected metabolites is provided in the supplemental table 1). These changes are summarized by heat map and principal component analysis (PCA) (Fig 3A and 3B). Interestingly, the highest metabolites induced (50-70-fold change) following immortalization of B cells was nicotinamide (NAM), nicotinic acid and nicotinamide adenine dinucleotide (NAD) (Fig 3C). NAM and nicotinic acid are both precursors of NAD and nicotinamide adenine dinucleotide phosphate (NADP), which are both coenzymes in wide-ranging enzymatic oxidation-reduction reactions, including glycolysis, the citric acid cycle, and the electron transport chain (32). Of note, the reduced form of NADP, NADPH, is the critical reducing equivalent used by FASN to synthesize long chain fatty acids (33). NAD+ is also an essential cofactor for Poly(ADP-Ribose) Polymerase 1 (34) which we have previously shown to be important in EBV latency status and LMP1-mediated host gene activation (9, 35). Aside from these important metabolic cofactors, our metabolic analysis also revealed several fatty acids amongst the top proteins induced following EBV transformation. These increases were in the range of 3-20-fold change and were mainly in the class of long and very long chain polyunsaturated fatty acids, differing form our ectopic LMP1 analysis where the top fatty acids were mainly saturated and medium to long chain length (Fig 3C).
EBV-induced immortalization of B cells upregulates FASN and lipogenesis
As we had already determined that LMP1 could induce FASN and lipogenesis in B cells, and both our LMP1 and EBV-immortalization metabolite studies showed significant changes to fatty acids, we also wanted to examine the effect of EBV-induced immortalization of B cells on FASN and lipogenesis. We firstly extracted protein from primary B cells and their established LCLs before western blotting for FASN. What we found was a massive upregulation of FASN at the protein level in LCLs vs primary B cells (Fig 4D). Specifically, FASN in B cells was not or only barely detectable vs robust expression in LCLs. Under serum-deprived conditions, we then stained primary B cells and LCLs cells with Nile red followed by FACS analysis (Fig 4A and 4B). Similar to our FASN western blot results, we observed virtually no Nile Red staining in B cells, to the extent that they looked like unstained controls, and we observed strong staining in our LCLs, which was confirmed by confocal microscopy (Fig 4C). These data suggest that EBV-induced immortalization of B cells turns on a lipogenesis program as shown by substantial upregulation of fatty acids and their metabolic cofactors, FASN and lipogenesis.
LMP1+ B cells are more sensitive to FASN inhibition
Deregulated FASN and lipogenesis is a hallmark of cancer and cancer cells have been shown to become addicted to the FASN pathway and de novo lipogenesis. This has led to many attempts to target FASN in cancers. This steered us to examine if we could take advantage of a potential LMP1-mediated dependency on the FASN pathway by using FASN inhibitors to selectively kill LMP1-expressing cells. Using the FASN inhibitor C75, we generated dose response curves for LMP1-expressing cells vs empty vector control using percent of cell death as determined by a trypan blue exclusion assay. C75 dose concentrations were transformed to log10 prior to nonlinear regression analysis and EC50 values were estimated (Fig 5A). We calculated EC50 values of 72 and 36 uM for pBABE and LMP1, respectively, suggesting an increased sensitivity to FASN inhibition in cells expressing LMP1 and increased FASN levels. We then treated latency type I and III cells with C75. During various stages of B cell differentiation in vivo, EBV will express either the latency III,II or I program, which entails expression of different subsets of latency genes. Type I latency cells do not endogenously express LMP1 as opposed to latency type III (36, 37). Comparing two such cell types therefore offers a more physiologically relevant comparison between LMP1-positive and negative cells. Mutu I and III are EBV-infected BL cell lines that differ only in their EBV latency status (I vs III). When we treated the LMP1-expressing Mutu III cells with C75, we observed much higher cell death compared to Mutu I cells that do not express LMP1 (Fig 5B). Two LCL cell lines (LCL and GM12878) also demonstrated sensitivity to FASN inhibition with significant accumulation of cell death after 24 hrs (Fig 5C). Finally, we measured cell viability following FASN inhibition in primary B cells and LCLs established following EBV infection. Whereas B cell viability was unaffected by C75 treatment, LCLs showed a significant drop in viability of around 50% vs untreated control (Fig 5D). Cells were also dosed with palmitic acid, which is the predominant product FASN and was used to determine if the observed toxicity of FASN inhibition was due to lack of fatty acid synthesis or toxic build-up of precursors (38). LCLs responded to palmitic acid with a significant increase in cell viability, given both individually and in combination with C75. This demonstrates that the effects of C75 is due to a block on the synthesis of downstream fatty acid metabolites which are therefore required for the viability of B cells latently infected with EBV.
LMP1 stabilizes FASN protein levels
We next sought to determine the mechanisms LMP1 employs to upregulate FASN. Previous work has pointed to LMP1 driving expression of FASN through its upstream regulator SREBP1c, at least in the context of NPC (30). However, upon examining our previously published RNA-seq data (9) it did not suggest that SREBP1c was a factor upregulated by LMP1, and this was confirmed by RT-qPCR using primers against both the precursor and mature isoforms of SREBF, SREBFa and SREBFc (Fig 6A). In fact, both FASN and SREBFa were actually downregulated in LMP1+ cells vs LMP1-cells while SREBFc remained unchanged. FASN can also be stabilized at the protein level by USP2a, a ubiquitin-specific protease that functions by removing ubiquitin from FASN and thus prevents its degradation by the proteasome (29) (Fig 6B). Our RNA-seq dataset (9) suggested that USP2a is upregulated by LMP1 and this was confirmed by RT-qPCR (Fig 6A). Because of this, we then wanted to determine if LMP1 stabilized FASN at the protein level. To examine this, we used the protein synthesis inhibitor cycloheximide (CHX). Following treatment with CHX, we observed that FASN protein levels were more stable at 24 hours in our LMP1-expresssing cell line vs empty vector control (Fig 6C). This suggests that LMP1, potentially through increased expression of USP2a, can post-translationally stabilize FASN.
To examine potential mechanisms of how EBV infection was causing upregulation of FASN, we again looked at factors effecting both expression and post-translational modifications of the enzyme. As with ectopic expression of LMP1, we looked into the SREBPs, the principal upstream regulators of FASN gene expression, and USP2a, the ubiquitin-specific protease that stabilizes FASN protein by decreasing its ubiquitination. We firstly used RT-qPCR to examine the gene expression of FASN and USP2a. When we compared the expression of these genes between primary B cells and LCLs, we found interesting results. Depending on the LCL (each generated form a different donor B cell) we found that either FASN expression was increased or USP2 expression, but never the two together (Fig 6D). Again, all of these LCLs robustly upregulate FASN at the protein level, suggesting that EBV will co-opt alternative pathways to achieve the same end result of increased FASN.
Discussion
The EBV-encoded oncoprotein LMP1 is expressed in several EBV associated malignancies including Hodgkin and post-transplant B cell lymphomas and NPC, for example. We and others have previously reported that LMP1 is able to stimulate aerobic glycolysis (‘Warburg’ effect) in cells (8–13). Our initial work was grounded in expression data where we observed that LMP1 could induce HIF-1α-dependent gene expression, alteration of cellular metabolism, and accelerated cellular proliferation (9). As a follow up to further investigate these LMP1-associated cellular metabolic changes, we used a targeted approach to examine the effects of both ectopic expression of LMP1, as well as EBV-mediated B-cell growth transformation, on host metabolites. From this initial analysis we observed that the top 15-20 metabolites significantly induced by LMP1 in the BL cell line DG75 were fatty acids (Figure). This was congruent with increased levels of FASN and lipid droplet formation as compared to empty vector controls (Figure). A recent study has specifically linked LMP1 to promotion of de novo lipogenesis, lipid droplet formation and increased FASN in NPC (30). This study went on to show that FASN overexpression is common in NPC, with high levels correlating significantly with LMP1 expression. Moreover, elevated FASN expression was associated with aggressive disease and poor survival in NPC patients. Interestingly, alteration of lipid metabolism was also observed in Burkitt Lymphoma following gene expression analysis. Based on this, adipophilin was identified as a novel marker of BL (31). This elevated level of lipid metabolism in BL might explain why we observed relatively minor changes to FASN levels and lipid droplet formation when we introduced LMP1 to the EBV-negative BL cell line DG75.
The increase in fatty acids by ectopic expression of LMP1 was offset following treatment with the PARP inhibitor olaparib (Figure 1). Previously we have shown that PARP1 is important in LMP1-induced aerobic glycolysis and accelerated cellular proliferation, both of which could be attenuated with PARP inhibition. PARP1 gene deletion and inhibition has been reported to enhance the lipid accumulation in liver and exacerbate high fat-induced obesity in mice (39, 40). However, a conflicting report concludes robust increases in PARP activity in livers of obese mice and non-alcoholic fatty liver disease (NAFLD) patients and that inhibition of PARP1 activation alleviates lipid accumulation and inflammation in fatty liver (41). Therefore, the role of PARP1 in lipid metabolism remains inconclusive, at least in the context of the liver and diet-induced obesity. As we have previously demonstrated that PARP1 can act as a coactivator HIF-1α-dependent gene expression, it is of interest to note that there is an emerging body of work that shows that HIF-1α can regulate lipid metabolism (42) including an ability to regulate FASN (43). How much of the LMP1-mediated changes we have observed to aerobic glycolysis and lipid metabolism is facilitated distinctly through PARP1, HIF-1α, a combination of the two, or completely independently of these factors, still needs to be elucidated.
In addition to examining LMP1-specific metabolic effects, we then examined metabolic changes following EBV-mediated B-cell growth transformation. Well we didn’t find that all the absolute highest fold changes in metabolites were fatty acids, as we did with ectopic expression of LMP1, we did find fatty acids being amongst the top metabolites altered. A recent study used proteomics to examine resting B cells and at nine time points after EBV infection. Their data pointed to induction of one-carbon (1C) metabolism being important for the EBV-mediated B-cell growth transformation process (44). This same analysis also revealed that EBV significantly upregulates fatty acid and cholesterol synthesis pathways. There are several key differences in this proteomics study versus our metabolomics approach. While we compared resting B cells with established LCLs around two months after infection, the above study also used several earlier timepoints. A follow up study using the same dataset suggested important roles for Epstein-Barr nuclear antigen 2 (EBNA2), SREBP and MYC in cholesterol and fatty acid pathways (45). The EBV-encoded transcription factor EBNA2 is produced early on in the infection phase (72hrs) (46), and the cholesterol and fatty acids synthesis pathways, including upregulation of FASN, were found to induced early in infection (96 hours). As LMP1 doesn’t appear until 3-7 days post infection (46), it lends to the questions as to its role in the induction of these pathways. The study referenced (45) indicated an important role for Rab13 role in possible trafficking of LMP1 to lipid raft signaling sites. It’s therefore possible, that the early changes to cholesterol and fatty acids synthesis pathways aid in the localization of LMP1 to cellular membranes to then enable LMP1 to maintain cholesterogenic and lipogenic programs at later timepoints by stimulating PI3K/AKT signaling cascades, for example.
Outside of the context of EBV-mediated B-cell growth transformation, there is also evidence of glucose-dependent de novo lipogenesis in B lymphocytes following lipopolysaccharide (LPS)-stimulated differentiation into Ig-secreting plasma cells (47). Specifically, this study pointed to ATP citrate lyase (ACLY) linking glucose metabolism into fatty acid and cholesterol synthesis during differentiation. This becomes especially interesting when taking into consideration the ability of EBV and LMP1 to induce both aerobic glycolysis and lipogenesis programs. One of the questions that arises from such studies is: are these metabolic changes unique to EBV-induced immortalization of B cells or are we observing the hijacking of pathways and metabolic remodeling used in the normal proliferation and differentiation of B cells? A study into primary effusion lymphoma (PEL) cells, which are a unique subset of human B-cell non-Hodgkin lymphomas cells latently infected with Kaposi’s sarcoma-associated herpesvirus (KSHV, another y-herpesvirus), showed that FASN expression and induction of fatty acid synthesis was necessary for the survival of latently infected PEL cells (48). Interestingly and related to the aforementioned question, these researchers stimulated resting B cells with LPS to determine if the differences in glycolysis and FASN were a consequence of proliferation, as PEL cells are continuously proliferating as lymphomas, rather than the transformed phenotype. While they did observe an elevated rate of glycolysis following LPS-stimulation of primary B cells, it was still significantly lower than that of vehicle-treated PEL cells. In addition, FASN did not substantially change in LPS-stimulated versus vehicle-treated primary B cells nor did LPS stimulation of PEL lead to any further increases in glycolysis or FASN compared with vehicle- treated PEL (48). These data potentially suggest that FASN activity is an independent phenotype of γ-herpesvirus, whether in the context of latently infected KSHV PEL or latently infected EBV NPCs or lymphomas, rather than a consequence of increased proliferation index.
We then went on to show that LMP1-expressing cells, including ectopically expressed LMP1; latency type III cell lines; or LCLs transformed from primary B cells were all more sensitive to FASN inhibition vs their corresponding LMP1-controls. We also observed that those type III latency cell lines with higher levels of FASN expression were more sensitive to FASN inhibition (Figure). Analysis of FASN expression in NPC patients found that higher levels of FASN expression significantly correlated with advanced primary tumor and distant lymph node metastasis (30). Latent infection of endothelial cells by KSHV led to a significant increase in long chain fatty acids as detected by a metabolic analysis. Fatty acid synthesis is required for the survival of latently infected endothelial cells, as inhibition of key enzymes in this pathway led to apoptosis of infected cells (49). We also observed that primary B cells, which express no or very little FASN protein, unsurprisingly were not sensitive to FASN inhibition. However, our LCLs transformed from primary B cells developed sensitivity to FASN inhibitors corresponding with induction of FASN and lipogenesis. A study reported that use of the lipoprotein lipase inhibitor orlistat resulted in apoptosis of B-cell chronic lymphocytic leukemia (CLL) cells without killing normal B cells from donors (50).
In conclusion, LMP1 is expressed in the majority of EBV-positive lymphomas. Despite many attempts to develop novel therapies, EBV-specific therapies currently remain largely investigational and EBV-associated malignancies are often associated with a worse prognosis. Therefore, there is a clear demand for EBV-specific therapies for both prevention and treatment. The work presented here suggests that targeting lipogenesis programs may be an effective strategy in the treatment of LMP1-positive EBV-associated malignancies. Further studies into the metabolic signaling pathways manipulated by EBV is critical to aid in the development of targeted, novel therapies against EBV-associated malignancies.
Materials and Methods
Cell culture and drug treatment
All cells were maintained at 37°C in a humidified 5% CO2 atmosphere in medium supplemented with 1% penicillin/streptomycin antibiotics. Lymphocyte cell lines (EBV-negative Burkitt’s lymphoma cell line DG75 ATCC CRL-2625 (DG75), EBV-positive latency III cell lines Mutu III, Mutu-LCL, GM12878 and EBV-positive latency I cell line Mutu I) were cultured in suspension in RPMI 1640 supplemented with fetal bovine serum at a concentration of 15%. Primary B cells were cultured in suspension in RPMI 1640 supplemented with fetal bovine serum at a concentration of 20%. 293T ATCC CRL-3216 (HEK 293T) cells were cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented with fetal bovine serum at a concentration of 10%. Olaparib (Selleck Chemical), cycloheximide (Sigma), MG-132 (Sigma) C75 (Sigma) and orlistat (Sigma) was dissolved in dimethyl sulfoxide (DMSO), and cells were treated for upon dilution in the appropriate media.
Retroviral transduction
Plasmid constructs hemagglutinin (HA)-tagged full-length LMP1, pBABE, pVSV-G, and pGag/Pol were kindly provided by Nancy Raab-Traub (UNC, Chapel Hill, NC) and were described previously [59]. Retroviral particles were generated using the Fugene 6 reagent (Promega) to simultaneously transfect subconfluent monolayers of 293T cells with 1μg pBABE (vector) or HA-LMP1, 250 ng pVSV-G, and 750 ng pGal/Pol according to the manufacturer’s instructions. Supernatant containing lentivirus was collected at 48- and 72-h post-transfection and filtered through a 0.45 μM filter. DG75 cells were transduced by seeding 5×105 cells in 6-well plates in 500 μl medium and adding 500 μl of medium containing retroviral particles. The transduced cells were placed under long-term selection in medium containing 1 μg/ml puromycin.
EBV Infection of Primary B cells
De-identified, purified human B cells were obtained from the Human Immunology Core of the University of Pennsylvania under an Institutional Review Board-approved protocol and were isolated using the RosetteSep Human B Cell Enrichment Cocktail (StemCell Technologies) as per protocol. Lymphocyte transformation was carried out using EBV-containing supernatants from the B95-8 cell line, distributed by ATCC as VR-1492TM. 1 vial of ATCC VR-1492TM (1 mL) was used per transformation flask. Per flask 2 mL Iscove’s modified DME (IMDM) with 20% FBS was added to 1 mL lymphocyte suspension containing 1-6 x 106 primary B cells. The B cell suspension was then gently mixed and incubated in flasks at 37°C under a 5% CO2 atmosphere. 7 days after initiation, flasks were observed for signs of transformation before 4 mL of fresh medium (IMDM with 20% FBS and 1X penicillin/streptomycin) was added. Fresh media was added every 3-4 days thereafter, and transformed cultures were expanded to establish lymphoblast cell lines. As part of the expansion process, cells were transferred from a densely populated T-25 flask to a T-75 flask and IMDM with 10% FBS and penicillin/streptomycin was added to obtain a cell density of 1-3 x 105 cells per mL. Cell density was maintained in the T-75 flask at 1-3 x 105 cells per mL by addition of culture medium.
Targeted relative metabolites quantitation
Cells were pelleted by centrifugation at 2,000 rpm for 5 min, 4 °C, washed cells twice in ice-cold PBS. Samples were extracted using cold extraction solution containing 80% methanol/20% water/0.2 uM heavy internal standard mix (MSK-A2-1.2 Cambridge Isotope Laboratories, Inc) using 2 million cells in 500 uL. Samples were vortexed thoroughly for 30 sec and placed on dry ice for at least 15 min. Samples were then spun at max speed (>13,000 rpm) for 15 min at 4 °C to pellet any debris. LC-MS analysis was performed on a Thermo Fisher Scientific Q Exactive HF-X mass spectrometer equipped with a HESI II probe and coupled to a Thermo Fisher Scientific Vanquish Horizon UHPLC system. Polar metabolites were extracted using 80% methanol and separated at 0.2 ml/min by HILIC chromatography at 45 °C on a ZIC-pHILIC 2.1 inner diameter x 150-mm column using 20 mM ammonium carbonate, 0.1% ammonium hydroxide, pH 9.2, and acetonitrile with a gradient of 0 min, 85% B; 2 min, 85% B; 17 min, 20% B; 17.1 min, 85% B; and 26 min, 85% B. Relevant MS parameters were as follows: sheath gas, 40; auxiliary gas, 10; sweep gas, 1; auxiliary gas heater temperature, 350 °C; spray voltage, 3.5 kV for the positive mode and 3.2 kV for the negative mode; capillary temperature, 325 °C; and funnel RF level at 40. A sample pool (quality control) was generated by combining an equal volume of each sample and analyzed using a full MS scan at the start, middle, and end of the run sequence. For full MS analyses, data were acquired with polarity switching at: scan range 65 to 975 m/z; 120,000 resolution; automated gain control (AGC) target of 1E6; and maximum injection time (IT) of 100 ms. Data-de-pendent MS/MS was performed without polarity switching; a full MS scan was acquired as described above, followed by MS/MS of the 10 most abundant ions at 15,000 resolution, AGC target of 5E4, maximum IT of 50 ms, isolation width of 1.0 m/z, and stepped collision energy of 20, 40, and 60. Metabolite identification and quantitation were performed using Com-pound Discoverer 3.0. Metabolites were identified from a mass list of 206 verified compounds (high confidence identifications) as well as by searching the MS/MS data against the mzCloud database and accepting tentative identifications with a mini-mum score of 50.
Lipid droplet fluorescence staining
Nile Red fluorescence staining was assessed with the Lipid Droplets Fluorescence Assay Kit according to the manufacturer’s protocol (Cayman Chemical, Ann Arbor, Ml, USA). One day before staining assay, cells were incubated in serum free medium. As a positive control, cells in completed medium were treated overnight with Oleic Acid provided from assay kit at 1:2000 dilution. For lipid droplets staining and quantification using a plate reader, cells were fixed with 1X assay fixative, washed with PBS and then stained with working solution of Hoechst 33342 (1 ug/ml)) and Nile Red (1:1000). The fluorescence of cells was determined using a GloMax plate reader (Promega). Hoechst 33342 fluorescence was measured with an excitation of 355 nm and an emission of 460 nm, while Nile Red fluorescence was determined using a 485 nm excitation and 535 nm emission. Differences in cell number were corrected by using Hoechst 33342 fluorescence signal to normalize the Nile Red signal in each well. For flow cytometric analysis, cells were only stained with Nile Red (1:1000). Analysis was carried out using a FACS Calibur flow cytometer (Becton Dickinson) and CellQuest software, and the cell population was analyzed using FlowJo software. Confocal microscopy images were taken on a Leica TCS SP8 MP multiphoton microscope.
Western Blot Analysis
Cell lysates were prepared in radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris-HCI, pH 7.4, 150 mM NaCI, 0.25% deoxycholic acid, 1% NP-40, 1 mM EDTA) supplemented with IX protease inhibitor cocktail (Thermo Scientific). Protein extracts were obtained by centrifugation at 3,000×g for 10 minutes at 4°C. For nuclear fractionation, nuclear soluble and chromatinbound protein fractions were extracted from cells using the Subcellular Protein Fractionation Kit for Cultured Cells kit (Invitrogen) according to manufacturer’s instructions. The bicinchoninic (BCA) protein assay (Pierce) was used to determine protein concentration. Lysates were boiled in 2x SDS-PAGE sample buffer containing 2.5% β-mercaptoethanol, resolved on a 4 to 20% polyacrylamide gradient Mini-Protean TGX precast gel (Bio-Rad), and transferred to an Immobilon-P membrane (Millipore). Membranes were blocked for 1 h at room temperature and incubated overnight with primary antibodies recognizing LMP1 (Abeam ab78113), FASN (ab22759) Ub (ab7780) and Actin (Sigma A2066), as recommended per the manufacturer. Membranes were washed, incubated for 1 h with the appropriate secondary antibody, either goat anti-rabbit IgG-HRP (Santa Cruz sc-2030) or rabbit anti-mouse IgG-HRP (Thermo Scientific 31430). Membranes were then washed and detected by enhanced chemiluminescence.
RT-qPCR
For reverse transcription quantitative PCR (RT-qPCR), RNA was extracted from 2 × 106 cells using TRIzol (Thermo Fisher Scientific) according to the manufacturer’s instructions. SuperScript II reverse transcriptase (Invitrogen) was used to generate randomly primed cDNA from 1 μg of total RNA. A 50-ng cDNA sample was analyzed in triplicate by quantitative PCR using the ABI StepOnePlus system, with a master mix containing 1X Maxima SYBR Green and 0.25 μM primers.
Data were analyzed by the ΔΔCT method relative 18s and normalized to untreated controls. Primers are available upon request.
Cell Viability Assay
Cell viability was measured using the CellTiter-Glo Luminescent Cell Viability Assay (Promega). 100 μl of cells in culture medium per well were plated in 96-well opaque-walled plates. The plate and samples were equilibrated by placing at room temperature for approximately 30 minutes. 100 μl of CellTiter-Glo Reagent was added to 100μl of medium containing cells. Plate contents for were then mixed for 2 minutes on an orbital shaker to induce cell lysis. Finally, the plate was incubated at room temperature for 10 minutes to stabilize luminescent signal before luminescence was recorded on a GloMax plate reader (Promega).
Dose-response curves
Dose concentrations were transformed to log10 prior to nonlinear regression analysis using GraphPad Prism version 8.00 for Mac OS X, GraphPad Software, La Jolla California USA, www.graphpad.com. Specifically, % dead cells based on live/dead counting using a Countess II FL Automated Cell Counter (ThermoFisher) following incubation with trypan blue was used as the Y value response.
Acknowledgements
This work was supported by National Institute of Health grant R01AI130209 to IT from the National Institute of Allergy and Infectious Diseases. The metabolomics analysis was performed at The Wistar Institute Proteomics and Metabolomics Shared Resource on a Thermo Q-Exactive HF-X mass spectrometer purchased with NIH grant S10 OD023586.