Abstract The Epstein-Barr virus (EBV) infects nearly 90% of adults globally and is linked to over 200,000 annual cancer cases. Immunocompromised individuals from conditions such as primary immune disorders, HIV, or posttransplant immunosuppressive therapies are particularly vulnerable because of EBV’s transformative capability. EBV remodels B cell metabolism to support energy, biosynthetic precursors, and redox equivalents necessary for transformation. Most EBV-driven metabolic pathways center on mitochondria. However, how EBV regulates B cell mitochondrial function and metabolic fluxes remains unclear. Here, we show that EBV boosts cardiolipin (CL) biosynthesis, essential for mitochondrial cristae biogenesis, via EBV nuclear antigen 2/MYC-induced CL enzyme transactivation. Pharmacological and CRISPR genetic analyses underscore the essentiality of CL biosynthesis in EBV-transformed B cells. Metabolomic and isotopic tracing highlight CL’s role in sustaining respiration, one-carbon metabolism, and aspartate synthesis. Disrupting CL biosynthesis destabilizes mitochondrial matrix enzymes pivotal to these pathways. We demonstrate EBV-induced CL metabolism as a therapeutic target, offering synthetic lethal strategies against EBV-associated B cell malignancies. __________________________________________________________________ A human tumor virus drives mitochondrial lipid synthesis for transformation. INTRODUCTION The Epstein-Barr virus (EBV), infecting more than 95% of the adult population worldwide, is associated with a spectrum of diseases ranging from the benign infectious mononucleosis to more than 200,000 cases of cancer annually ([35]1). EBV has been implicated in various cancers, including Burkitt lymphoma (BL), Hodgkin lymphoma, diffuse large B cell lymphoma (DLBCL), nasopharyngeal carcinoma (NPC), and gastric carcinoma (GC). EBV^+ BL represents the most prevalent childhood cancer in Africa ([36]2–[37]7). If unchecked by T and natural killer cells, then EBV transforms human resting B cells into lymphoblasts, placing immunocompromised individuals at heightened risk for lymphoproliferative disorders ([38]8, [39]9). EBV remodels host metabolic pathways to support B cell transformation. This remodeling is primed by viral-encoded oncoproteins, including six EBV nuclear antigens (EBNA1, EBNA2, EBNA3A, EBNA3B, EBNA3C, and EBNA-LP) and two latent membrane proteins (LMP1 and LMP2), along with various viral noncoding RNAs. These viral factors initiate a cascade of metabolic reprogramming early in the infection process, leading to substantial reorganization in B cell architecture that favors cell growth and division. The early expression of EBNA2 and its primary target, MYC, in particular, activates multiple metabolic pathways, including oxidative phosphorylation (OXPHOS), one-carbon (1C) metabolism, and fatty acid biosynthesis, further compounded by the expression of LMP1 and LMP2, which mimic activated CD40 and B cell receptor pathways, respectively ([40]10, [41]11). The capability of EBV to reprogram host metabolism extends deeply into mitochondrial functions ([42]12–[43]20), where it ensures the continuous generation of adenosine 5′-triphosphate (ATP) and metabolic precursors critical for the synthesis of essential macromolecules and regulation of redox balance. Any perturbation in mitochondrial pathways, including OXPHOS, the tricarboxylic acid (TCA) cycle, and the mitochondrial 1C metabolism, is detrimental to the EBV transformation process ([44]12, [45]17, [46]21). The newly infected cell population that fails to transform into lymphoblastoid cell lines (LCLs) in culture is characterized by mitochondrial dysfunction ([47]17). However, how EBV modifies mitochondrial biogenesis during transformation remains incompletely understood. The inner mitochondrial membrane (IMM) serves as a crucial barrier and site for various key biochemical reactions. The IMM facilitates electron transport via the electron transport chain (ETC), which not only produces ATP but also modulates reactive oxygen species (ROS) levels. In addition, the IMM plays an essential role in biosynthetic pathways, housing enzymes for the TCA cycle and fatty acid β-oxidation. These pathways provide essential precursors for nucleotide, amino acid, and lipid biosynthesis, supporting cell growth and proliferation ([48]22). Cristae are folds within the IMM that increase its surface area to enhance these biochemical processes ([49]22, [50]23). Cardiolipin (CL) is a glycerophospholipid, which constitutes about 20% of the total mass of phospholipids of IMM. CL is essential for shaping the mitochondrial cristae and binding to ETC complexes in the IMM ([51]24, [52]25). Defects in CL metabolism led to compromised respiration and disruptions in the TCA cycle ([53]24–[54]26). CL also modulates apoptosis by controlling cytochrome c release ([55]27). In the current study, we demonstrate that EBV de novo infection in B cells activates the CL biosynthetic pathway to support cristae biogenesis, which plays critical role in sustaining extensive metabolic network reprogramming. Our findings highlight CL biosynthesis as a key metabolic vulnerability in EBV-transformed B cells and provide a foundational basis for developing synthetic lethal strategies to treat EBV-induced B cell lymphoproliferative disorders. RESULTS EBV up-regulates CL biosynthesis and remodels B cell mitochondrial ultrastructure A central hub for metabolic regulation in cells is the mitochondrion. However, how EBV regulates mitochondrial remodeling is not fully understood. After EBV infection, there was a marked boost in mitochondrial biogenesis, which was demonstrated by a significant increase in the MitoTracker Green signal and mitochondrial DNA abundance (fig. S1, A and B). Analyzing a published RNA sequencing (RNA-seq) dataset of B cells infected with EBV reveals a significant increase in the transcription of genes involved in the ETC ([56]28). Cells at day 2 postinfection (2 DPI) exhibited the highest transcriptional changes in these ETC genes (fig. S1C). Gene ontology analysis of differentially expressed metabolic genes at 2 DPI reveals terms including cellular respiration and respiratory chain complex assembly (fig. S1D). We therefore used blue native polyacrylamide gel electrophoresis (BN-PAGE) to visualize mitochondrial respirasome assembly during EBV infection in B cells. Compared to uninfected human resting B cells, ETC supercomplexes in EBV-infected cells were readily detected from 20 μg of mitochondrial proteins starting from 2 DPI and onward (fig. S1E). From 14 DPI, an additional increase in complex I and III[2] + IV + V supercomplexes was observed (fig. S1E). Consistently, increases in mitochondrial membrane potential, basal, and maximal oxygen consumption rates (OCRs), as well as ATP production, were significantly elevated from 2 DPI (fig. S1, F to H). These findings suggest that EBV strongly regulates mitochondrial biogenesis, ETC supercomplexes assembly, and respiration at the early stage of infection. We next investigated how EBV infection alters the mitochondrial ultrastructure to support B cell transformation. To capture dynamic change of the mitochondrial ultrastructure upon EBV infection, we used a transmission electron microscopy (TEM) to analyze human B cells, either uninfected or infected with EBV, which were collected at 4, 7, and 28 DPI. The infected cells appeared to be much larger in size with substantially increased cytosolic space ([57]Fig. 1A). We observed a large boost in mitochondrial cristae biogenesis, as indicated by both the increased number per mitochondrion and extended length of cristae ([58]Fig. 1, B and C). Cristae are distinguished by their membrane invagination structures that project into the mitochondrial matrix. These cristae greatly expand the IMM’s surface area and serve as a highly selective barrier that hosts the ETC and ATP synthesis, playing a crucial role in cellular energy production and metabolic regulation ([59]23). The pronounced increase in mitochondrial cristae biogenesis during EBV infection sparked our interest in mitochondrial lipid changes. Synthesized in the IMM, CL is crucial for stabilizing mitochondrial supercomplexes and shaping cristae architecture, both of which are vital for efficient mitochondrial respiration (fig. S2A) ([60]24) and necessary for EBV transformation ([61]17). We conducted lipidomic profiling of uninfected and newly EBV-infected B cells at 7 and 10 DPI, time points associated with noticeable increases in mitochondrial cristae biogenesis, using liquid chromatography–mass spectrometry (LC-MS). The abundance of each lipid species was normalized to the total ion count for each sample, as shown in fig. S2B. Our analysis showed significant increases in the levels of various CL species and its precursors, phosphatidic acid and phosphatidylglycerol (PG), in EBV-infected B cells at both 7 and 10 DPI compared to uninfected resting B cells ([62]Fig. 1D; fig. S2, C and D; and table S1). Fig. 1. EBV regulates CL biosynthesis to sustain mitochondrial remodeling during B cell transformation. [63]Fig. 1. [64]Open in a new tab (A) TEM analysis of newly EBV-infected B cells collected at 0, 4, 7, and 28 DPI. Scale bars, 500 nm. (B) Statistical analysis of cristae number per mitochondrial from five random TEM pictures as shown in (A). P values were calculated using one-way analysis of variance (ANOVA) with Tukey’s multiple comparisons test. (C) Statistical analysis of cristae length from five random TEM pictures as shown in (A). P values were calculated using one-way ANOVA with Tukey’s multiple comparisons test. (D) Heatmap visualization of z-scores of CL species detected in lipidomics analysis of uninfected or newly infected primary B cells by EBV, collected at 7 or 10 DPI. The total lipids were extracted from 5 million cells of each sample and quantified by LC-MS. The abundance of individual lipid species was normalized against the total ion count of the sample. The z-score describes the SD variation from the mean value of each lipid. Individual data from n = 4 samples are shown. P values are indicated as follows: *P < 0.05, **P < 0.01, and ****P < 0.0001. The CL biosynthetic pathway involves a cascade of enzymatic reactions, including key enzymes such as protein tyrosine phosphatase mitochondrial 1 (PTPMT1) and CL synthase 1 (CRLS1) ([65]24). These enzymes work sequentially to convert phosphatidic acid into nascent CL ([66]Fig. 2A and fig. S2A). We found that EBV up-regulated CRLS1 transcription and translation by 2 DPI, coinciding with the heightened expression of EBNA2 and MYC. PTPMT1 protein expression was activated at 2 DPI and largely enhanced from 7 DPI ([67]Fig. 2B and fig. S3A). Upon naïve B cells infection, EBV initiates a prelatency phase within the first 3 days. During this early phase, EBV expresses EBNA2 and EBNA-LP, which activates c-MYC ([68]12, [69]21, [70]29–[71]31). EBNA2 and MYC jointly remodel B cell metabolism before the first mitosis ([72]Fig. 2C) ([73]12, [74]32). We therefore explored the role of EBNA2 and EBNA-LP in initiating CL biosynthesis. We infected primary human B cells with either the B95.8 strain, ultraviolet (UV)–inactivated B95.8 strain or the nontransforming P3HR-1 strain of EBV ([75]33, [76]34). Notably, the P3HR-1 strain lacks EBNA2 and most EBNA-LP open reading frames making it a tool virus to study EBNA2 and/or EBNA-LP function ([77]35). To maintain consistent infection levels across samples, we adjusted for input viral genome copy numbers, basing these adjustments on quantitative polymerase chain reaction (qPCR) analyses of the virus stocks (fig. S3B). The immunoblot revealed that only the B95.8 strain, but neither the P3HR-1 nor UV-inactivated B95.8, was capable of triggering the expression of CRLS1 and PTPMT1 by 2 DPI ([78]Fig. 2D). This observation was made while ensuring equal levels of infection, as evidenced by comparable postinfection intracellular viral loads (fig. S3C). These findings indicate that EBV EBNA2 and/or EBNA-LP, rather than a broad response to EBV infection, are crucial for triggering CL biosynthesis. Fig. 2. EBV infection induces EBNA2 to regulate CL synthesis. [79]Fig. 2. [80]Open in a new tab (A) Schematic picture of CL biosynthesis and remodeling. MLCL, monolysocardiolipin. Key enzymes are shown in red. (B) Immunoblot analysis for indicated proteins in whole cell lysates (WCLs) in newly EBV-transformed human primary B cells, collected at indicated days after infection. (C) EBV-induced B cells immortalization in vitro model. Once the EBV inside the B cell during the early 3 DPI, two oncoproteins of the EBNA2 and EBNA-LP were initiated to establish a prelatency program. In the latency IIb, cells express six EBNAs. During latency III, cells express latency membrane protein, which contributes to B cells fully transforming into LCLs. ncRNAs, noncoding RNAs. (D) Immunoblot analysis of WCL in uninfected, B cell infected with B95.8, UV-inactivated B95.8, or P3HR-1 viruses, collected at 2 DPI. (E) ChIP-seq tracks of the indicated transcription factors, along with the activating histone epigenetic marks H3K27ac, H3K4me1, or H3K4me3, at the GM12878 LCL CRLS1 locus are shown. Also displayed are the long-range DNA linkages between an upstream minichromosome maintenance 8 (MCM8) locus enhancer and CRLS1, as defined by GM12878 ChIA-PET (chromatin interaction analysis with paired-end tags). Pol II, polymerase II. (F) Immunoblot analysis of EBNA2, PTPMT1, CRLS1, and actin expression in EBNA2-HT cells cultured in the presence or absence of 4HT (1 μM). (G) Immunoblot analysis of EBNA2, PTPMT1, CRLS1, MYC, and actin expression in P493-6 cells cultured with the indicated supplement. Doxycycline (Dox) was used at 1 μg/ml, and 4HT was used at 1 μM. (H) Immunoblot analysis of PTPMT1, CRLS1, MYC, or actin in WCL of Cas9^+ GM12878 LCL, GM12892 LCL, Rael BL, and P3HR-1 BL expressing control or the indicated MYC sgRNAs. All blots are representative of n = 3 biological replicates. In an effort to further understand the potential roles of EBNA2 in the activation of CRLS1, the rate-limiting enzyme in CL biosynthesis, we used publicly available resources including ENCODE GM12878 LCL chromatin immunoprecipitation sequencing (ChIP-seq) dataset ([81]36), EBV EBNA2 ChIP-seq dataset ([82]37), and long-range chromatin interaction analysis ([83]38). We found that EBNA2, along with its host targets MYC, co-occupy the promoter of CRLS1 ([84]Fig. 2E). ChIP-qPCR at 2 DPI, where both EBNA2 and MYC are highly expressed, confirmed the binding of EBNA2 and MYC to the CRLS1 and PTPMT1 promoters in newly infected primary B cells (fig. S3, D and E). To further elucidate the role of EBNA2 in regulating CRLS1, we used the 2-2-3 LCL cell line expressing an engineered EBNA2, where EBNA2 is fused to a modified estrogen receptor ligand-binding domain. Therefore, EBNA2 transcriptional activity can be regulated by the addition of 4-hydroxytamoxifen (4HT). With 4HT, EBNA2-4HT is stabilized and translocated to the nucleus and regulates transcription ([85]39). When EBNA2 activity is conditionally down-regulated by withdrawing 4HT for 48 hours, there is a noticeable reduction in the levels of CRLS1 and PTPMT1 ([86]Fig. 2F). EBNA2 transactivates MYC, and both proteins can synergistically induce EBV target genes ([87]37, [88]38). Previous research showed that conditional inactivation of EBNA2 led to a rapid down-regulation of MYC levels in LCLs ([89]12). ChIP-seq in LCLs showed that MYC and MAX, which form a heterodimer complex binding to E-box sites, co-occupied the LCL CRLS1 promoter ([90]Fig. 2E). MYC targets approximately 600 nuclear-encoded mitochondrial genes ([91]40, [92]41). To further distinguish the specific roles of EBNA2 and MYC in regulating CRLS1 and PTPMT1 expression in LCLs, we used the P493-6 LCL, which harbors a conditional 4HT-inducible EBNA2 allele and a heterologous MYC allele under Tet-OFF control (fig. S3F) ([93]42). Our findings revealed that reexpression of MYC alone was sufficient to restore CRLS1 and PTPMT1 expression following EBNA2 inactivation through 4HT withdrawal ([94]Fig. 2G), indicating a direct association between MYC levels and the expression of these genes. To probe MYC’s role further, we used CRISPR-Cas9 to knock out MYC in LCLs. We analyzed cells at an early time point following CRISPR editing before the onset of cytotoxic effects of MYC knockout (KO). We found that MYC KO resulted in a marked reduction in PTPMT1 and CRLS1 protein levels in GM12878 and GM12892 LCLs ([95]Fig. 2H). In Rael and P3HR-1 BLs, which lack EBNA2 expression but have high MYC levels, knocking out MYC similarly reduced the levels of PTPMT1 and CRLS1 ([96]Fig. 2H). This highlights MYC as a key regulator of CRLS1 and PTPMT1 expression, independent of EBNA2. We also noticed the co-occupancy of EBNA2 and MYC in an upstream enhancer (about 110 kbp away), marked by H3K27ac, H3K4me1, and EP300 co-occupancy, linked to the CRLS1 promoter by a chromatin long range loop ([97]Fig. 2E, rectangle). Using CRISPR interference (CRISPRi) on GM12878 LCL cells, we were able to investigate the role of CRLS1 enhancer in regulating CRLS1 expression. GM12878 LCL expressing dCas9 repressor and CRLS1 enhancer targeted single guide RNA (sgRNAs) showed significantly reduced CRLS1 expression at both mRNA and protein level (fig. S3, G and H). Collectively, our data reinforce the role of EBNA2 as the key viral oncoprotein responsible for up-regulating CL biosynthesis through a MYC-dependent mechanism. CL biosynthesis is critical for EBV B cell transformation and LCL survival PTPMT1 converts phosphatidylglycerophosphate (PGP) to PG. Subsequently, CRLS1 finalizes the process by converting PG and cytidine diphosphate–diacylglycerol (CDP-DAG) into CL ([98]Fig. 3A) ([99]43). Alexidine dihydrochloride (AD) is a selective PTPMT1 inhibitor ([100]44). AD inhibits the growth of EBV^+ NPC cell line C666-1 but not in nontransformed cell types such as GM05757, HNEpC, or NIH/3T3 ([101]45). AD treatment on newly EBV-infected B cells significantly hampered transformation and increased 7-aminoactinomycin D (7-AAD) cell death signal on a dose-dependent manner ([102]Fig. 3, B and C). Similar AD-induced growth defects were observed in GM12878 and the other three LCLs. The on-target effect of AD treatment was assessed using a CL fluorometric assay, which showed a significant reduction in CL content in mitochondria isolated from AD-treated GM12878 LCLs (fig. S4, A to D). To further confirm the AD effects, we used CRISPR-Cas9 system to knock out the PTPMT1 in three LCLs (GM12878, GM12881, and GM12892), two EBV^+ BLs (P3HR-1 and Rael), and Reh, an EBV^− cell line exhibiting lymphoblastic morphology that was isolated from tissue from a patient with acute lymphocytic leukemia (ALL). We achieved >95% of PTPMT1 KO efficiency with all three individual sgRNA ([103]Fig. 3D), which consistently diminished GM12878 growth ([104]Fig. 3E). PTPMT1 KO significantly reduced mitochondrial CL content (fig. S4E). Similar growth defects were observed in GM12881 Cas9^+ and GM12892 Cas9^+ LCLs expressing PTPMT1 sgRNAs (fig. S4, F and G). When replacing a protospacer adjacent motif (PAM) site–disabled PTPMT1 cDNA in PTPMT1 KO GM12878 cells ([105]Fig. 3F), growth defects led by endogenous PTPMT1 KO was rescued ([106]Fig. 3G). Propidium iodide cell cycle analysis demonstrated that PTPMT1 KO significantly increased the sub-G[1] population indicative of cell death, while the PAM site–mutated PTPMT1 can rescue the cell death effects (fig. S4, H and I). We further tested the effects of PTPMT1 KO in EBV^+ Rael and P3HR-1 BLs. Similar growth defects were observed in PTPMT1 KO Rael and P3HR-1 BLs ([107]Fig. 3, H and I, and fig. S4J). PTPMT1 KO did not cause evident growth defects in EBV^− Reh cells, which is consistent with the data in DepMap database ([108]46), suggesting that the presence of EBV may increase the dependency on CL biosynthesis ([109]Fig. 3, J and K). Collectively, these data suggest that the survival of EBV-transformed LCLs and BLs depends on a functional CL biosynthetic pathway. Fig. 3. Inhibiting CL biosynthesis impedes EBV-transformed B cells growth. [110]Fig. 3. [111]Open in a new tab (A) Schematics show that AD selectively inhibits PTPMT1, which converts PGP to PG, a CL precursor. (B) Live cell counts of EBV-infected primary B cells treated with dimethyl sulfoxide (DMSO) or 2 μM AD for 24 hours (n = 4 donors; means ± SD). P values were calculated using unpaired Student’s t test. MFI, mean fluorescence intensity. (C) Fluorescence-activated cell sorting (FACS) analysis of 7-AAD in EBV-infected 4 DPI primary B cells that treated with DMSO control or 1 or 2 μM AD for 24 hours. (D) Immunoblot analysis of PTPMT1 or actin in WCL of Cas9^+ GM12878 LCL expressing control or the indicated PTPMT1 sgRNAs. (E) Growth curves of Cas9^+ GM12878 LCL expressing control or the indicated PTPMT1 sgRNAs. (F) Immunoblot analysis of WCL from Cas9^+ GM12878 LCL following expression of the indicated green fluorescent protein (GFP) control or V5-PTPMT1^R rescue cDNAs and the indicated control or PTPMT1-targeting sgRNAs. (G) Growth curves of Cas9^+ GM12878 LCL following expression of the indicated GFP control or PTPMT1^R rescue cDNAs and the indicated control or PTPMT1-targeting sgRNAs. (H) Immunoblot analysis of PTPMT1 or actin in WCL of Cas9^+ Rael BL expressing control or the indicated PTPMT1 sgRNAs. (I) Growth curves of Cas9^+ Rael BL expressing control or the indicated PTPMT1 sgRNAs. (J) Immunoblot analysis of PTPMT1 or actin in WCL of Cas9^+ Reh ALL (EBV^−) expressing control or the indicated PTPMT1 sgRNAs. (K) Growth curves of Cas9^+ Reh ALL expressing control or the indicated PTPMT1 sgRNAs. Means ± SD values in (C), (E), (G), (I), and (K) were from n = 3 experiments. P values were calculated one-way ANOVA with Tukey’s test and indicated as follows: **P < 0.01 and ****P < 0.0001. In (G), ####P < 0.0001. n.s., not significant. All blots are representative of n = 3 biological replicates. Impaired CL biosynthesis disrupts mitochondrial respiration in EBV-transformed B cells EBV-driven B cell transformation relies on functional OXPHOS ([112]12, [113]32). Pharmacological inhibition of the ETC halts EBV-driven B cell transformation ([114]12). Given the essential role of CL in maintaining mitochondrial function, we first investigated whether disrupting CL biosynthesis affects mitochondrial respiration in cells newly infected with EBV. AD treatment significantly reduced the B cell mitochondrial membrane potential, as measured by tetramethylrhodamine (TMRM), at 4 DPI ([115]Fig. 4A and fig. S5A). Along with this, the OCRs of both basal and maximal respiration, as well as ATP production, were significantly reduced ([116]Fig. 4, B and C). Notably, this inhibition of OXPHOS does not attribute to decreased mitochondrial biogenesis; the level of MitoTracker Green was significantly increased after AD treatment ([117]Fig. 4D and fig. S5, C and D), which might link to a compensatory mechanism as described previously ([118]47–[119]49). AD treatment significantly reduced the extracellular acidification rate (ECAR) in cells at 4 DPI, indicating a potential down-regulation of glycolysis (fig. S5E). Given that CL contributes to mitochondrial cristae, we hypothesize that AD treatment may disrupt cristae structure in newly EBV-infected B cells. We therefore performed TEM analysis on cells at 4 DPI that were treated with dimethyl sulfoxide (DMSO) or AD for 24 hours ([120]Fig. 4E and fig. S5F). In DMSO-treated cells, the cristae are observed as shelf-like invaginations extending into the mitochondrial matrix ([121]Fig. 4E, top left, and fig. S5F). They generally display a lamellar shape but sometimes appear to be short sphere, which is probably due to ongoing cristae remodeling (asterisks). By contrast, in AD-treated cells ([122]Fig. 4E), the number of “empty” mitochondria lacking cristae was greatly increased (purple arrows). In some mitochondria, we observed a highly disorganized cristae structure, which is usually located at one side of the mitochondria (orange arrows) and multiple long and distorted tube-like structures originate from it (orange arrows). We speculate that these empty mitochondria may be newly generated following drug treatment, lacking the capability to fully develop their internal structures due to disrupted CL biosynthesis. In addition, the mitochondria with disorganized cristae (orange arrows) may be preexisting ones formed before the drug treatment, which had developed cristae but became disorganized because of the treatment. Besides the defects in mitochondria, we also observed an increased number of intracellular vesicles, endoplasmic reticulum, and Golgi apparatus in AD-treated cells (green arrows). Fig. 4. CL inhibition disrupts respiration, cristae biogenesis, and metabolome. [123]Fig. 4. [124]Open in a new tab (A) FACS analysis of TMRM MFI in EBV-infected primary B cells at 4 DPI treated with DMSO or 2 μM AD for 24 hours. (B and C) Seahorse OCR linked to basal respiration, maximal respiration, and ATP production in EBV-infected primary B cells at 4 DPI treated with DMSO or 2 μM AD for 24 hours. Oligo, oligomycin A; FCCP, carbonyl cyanide p-trifluoromethoxyphenylhydrazone; Rot/Anti, rotenone/antimycin A. (D) FACS analysis of MitoTracker Green MFI in EBV-infected primary B cells at 4 DPI treated with DMSO or 2 μM AD for 24 hours. (E) TEM analysis of EBV-infected primary B cells at 4 DPI treated with DMSO or 2 μM AD for 24 hours. Scale bars, 500 nm. (F) FACS analysis of MitoSOX Green fluorescence in primary B cells at 4 DPI treated with DMSO or 1 μM AD for 24 hours. (G) Newly infected B cells at 4 DPI were treated with DMSO, 2 μM AD, or 0.1 μM PierA for an additional 24 hours (h). LC-MS was used to quantify intracellular metabolites. (H) Volcano plot comparing metabolomic profiles of primary B cells at 4 DPI treated with 2 μM AD versus DMSO for 24 hours (n = 3 experiments). Hcy, homocysteine; PRPP, phosphoribosyl pyrophosphate; SAH, S-adenosyl-homocysteine; αKG, α-ketoglutarate. (I) Metabolic pathway analysis highlighting pathways significantly up-regulated or down-regulated by AD treatment. The x axis shows pathway impact values from MetaboAnalyst 3.0 topological analysis; the y axis shows −log[10] of P value from pathway enrichment analysis. (J) Absolute number of live EBV-infected primary B cells at 4 DPI treated with DMSO or 1 μM AD, supplemented with H[2]O, 2 mM GSH, or 1.5 mM NAC for 48 hours. Means ± SD values in (A) to (D), (F), (H), and (J) were from n = 3 experiments. P values were calculated using an unpaired Student’s t test. **P < 0.01, ***P < 0.001, and ****P < 0.0001. In (J), ###P < 0.001. CL stabilizes the ETC supercomplexes located on the cristae ([125]23, [126]24, [127]43, [128]50). We then investigated whether PTPMT1 inhibition destabilizes the respiration supercomplexes in EBV-transformed LCLs. We knocked out PTPMT1 in GM12878 cells and then tested the abundance of ETC proteins from whole cell lysates using an OXPHOS antibody cocktail. Overall, succinate dehydrogenase complex subunit B (SDHB) in complex II, ubiquinol-cytochrome c reductase core protein 2 (UQCRC2) in complex III, cytochrome c oxidase subunit II (COX II) in complex IV, and ATP synthase F1 subunit alpha (ATP5A) in complex V remained stable in PTPMT1 KO GM12878 LCLs. We observed a noticeable down-regulation of NADH:ubiquinone oxidoreductase subunit B8 (NDUFB8) in complex I in PTPMT1 KO cells expressing the PTPMT1 sgRNAs (fig. S5G). We further investigated whether PTPMT1 inhibition might affect the assembly of ETC supercomplexes. Therefore, we isolated mitochondria from control and PTPMT1 KO GM12878 LCLs and performed BN-PAGE and then immunoblotting for ETC complexes. In this experiment, immunoblotting with the anti-OXPHOS antibody cocktail revealed an additional unknown band with a size of around 550 kDa, as indicated by an arrow (fig. S5H). Other than this, we did not observe any obvious differences between the control and PTPMT1 KO cells using anti-OXPHOS antibody cocktail. However, when using individual antibodies targeting individual ETC complex, we found a notable reduction of complex II and I/III supercomplexes suggesting that PTPMT1 KO may compromise high-order respirasome assembly in EBV LCLs (fig. S5H). Deficiencies in the ETC can lead to increased production of ROS ([129]51). We therefore use MitoSOX to measure mitochondrial ROS in newly infected B cells. We observed increased mitochondrial ROS in AD-treated B cells, indicating elevated oxidative stress ([130]Fig. 4F). To elucidate the metabolic impacts of AD treatment on newly EBV-infected B cells, we further used an LC-MS–based untargeted metabolomic analysis. We used piericidin A (PierA) as a comparative control, which is a selective inhibitor of complex I. Newly isolated human primary B cells were infected with EBV B95.8 at a multiplicity of infection of 1. At 4 DPI, the cells were treated with DMSO, AD, or PierA for an additional 24 hours. Subsequent LC-MS–based metabolomic analyses of these samples unveiled significant variances in intracellular metabolites across cells treated with DMSO, AD, or PierA ([131]Fig. 4G, fig. S6, and table S2). The principal components analysis plot suggests that both AD and PierA treatments induce metabolic profiles that are distinctly different from the DMSO control, with some variations between AD and PierA treatments themselves, indicating both shared and unique effects on cellular metabolism (fig. S7A). Compared to DMSO-treated cells, AD treatment led to an up-regulation of metabolites within the B cell malate aspartate shuttle (MAS), the TCA cycle, the pentose phosphate pathway (PPP), and purine and pyrimidine metabolism. In contrast, it down-regulated metabolites associated with steroid synthesis, as well as folate, methionine, betaine, and nicotinamide metabolism ([132]Fig. 4, H and I), most of which are known dependent on NADPH [reduced form of nicotinamide adenine dinucleotide phosphate (NADP^+)] availability and are critical for EBV B cell transformation ([133]12–[134]14, [135]52). PierA treatment, on the other hand, significantly down-regulated metabolites in pyrimidine metabolism and the TCA cycle, while up-regulating those in purine metabolism and contributing to the Warburg effect (fig. S7, B and C). Despite some overall correlation in the metabolic profiles between AD- and PierA-treated cells, we identified notable exceptions: levels of citrate, succinate, itaconate, orotate, and glucosamine were increased in AD-treated cells but were significantly decreased in cells treated with PierA (fig. S7D). Notably, AD treatment significantly reduced NADPH levels by 7-fold and increased NADP^+ by 1.5-fold, leading to a marked decrease in the NADPH/NADP^+ ratio (fig. S6). This change may disrupt cellular redox balance and impair biosynthetic pathways, including steroid biosynthesis and folate metabolism ([136]Fig. 4I). Glutathione (GSH) supplementation reduced AD-induced cell mortality effectively, whereas N-acetylcysteine (NAC) was less effective ([137]Fig. 4J). Unlike PierA treatment, AD did not affect NADH [reduced form of nicotinamide adenine dinucleotide (oxidized form) (NAD^+)]/NAD^+ ratio (fig. S6), suggesting a targeted disruption of anabolic processes with little impact on catabolic processes such as TCA cycle. This specificity indicates CL’s role in modulating anabolism and redox defense. Impaired CL biosynthesis disrupts mitochondrial 1C metabolism in EBV-transformed B cells 1C metabolism encompasses folate and methionine metabolism, facilitating the transfer of 1C units across various metabolic pathways. This process is essential for DNA synthesis, repair, and methylation, ultimately affecting cell growth and genetic regulation ([138]53). EBV B cell transformation activates and heavily depends on this metabolic pathway, primarily contributing to NADPH production and the synthesis of purines and thymidylate ([139]Fig. 5A) ([140]12, [141]53). Our findings indicate that AD treatment in newly infected B cells down-regulates folate and methionine metabolism ([142]Fig. 4, H and J), leading to an accumulation of intracellular serine and folate, while the downstream product 5-methyl–tetrahydrofuran (THF) is significantly reduced ([143]Fig. 5B). Fig. 5. Inhibiting CL biosynthesis disrupts 1C metabolism by destabilizing mitochondrial 1C enzymes in EBV-transformed B cells. [144]Fig. 5. [145]Open in a new tab (A) Schematic of 1C metabolism. Enzymes are shown in blue. DHFR, dihydrofolate reductase; SHMT1/2, serine hydroxymethyltransferase 1/2; MTHFR: methylenetetrahydrofolate reductase; MTHFD1/2: methylenetetrahydrofolate dehydrogenase 1/2; MTHFD1L: methylenetetrahydrofolate dehydrogenase 1-like; ALDH1L2: aldehyde dehydrogenase 1 family member L2. GSSG, oxidized glutathione; IMP, inosine monophosphate. (B) Bar chart showing levels of serine, folate, and 5-methyl-THF. a.u., arbitrary units. (C) Immunoblot analysis for MTHFD2, SHMT2, SHMT1, ALDH1L2, TOMM20, and actin in WCL of uninfected or EBV-infected primary B cells at 4 DPI, treated with DMSO, 2 μM AD, or 0.1 μM PierA for 24 hours. Results are representative of three biological experiments. (D) Quantitative real-time PCR (qRT-PCR) for MTHFD2, SHMT2, or ALDH1L2 mRNA abundance in EBV-infected primary B cells at 4 DPI, treated with DMSO, AD, or PierA as indicated in (C). (E) Immunoblot analysis for SHMT2, SHMT1 ALDH1L2, MTHFD2, and actin in WCL of GM12878 LCL and GM12881 LCL expressing control or PTPMT1 sgRNAs. Results are representative of three experiments. (F) Absolute live number of GM12878 LCL, GM15892 LCL, MUTU I BL, or Daudi BL, EBV^− BCBL1, and EBV^− Reh treated with DMSO, 1 μM AD, 10 μM SHIN1, or 1 μM AD + 10 μM SHIN1 for 48 hours. Means ± SD values in (B), (D), and (F) were from n = 3 experiments. P values were calculated one-way ANOVA with Tukey’s test. *P < 0.05, ***P < 0.001, and ****P < 0.0001. Blots in (C) and (E) were representative of at least n = 3 experiments. A key portion of 1C metabolism occurs in mitochondria, where key enzymes such as serine hydroxymethyltransferase 2 (SHMT2) and methylenetetrahydrofolate dehydrogenase 2 (MTHFD2) catalyze reactions using folate as a 1C carrier ([146]53). In AD-treated newly EBV-infected cells, there was a significant reduction in the protein levels of SHMT2 and MTHFD2, key enzymes in the mitochondrial 1C pathway ([147]Fig. 5C). Moreover, our data reveal that the protein level of aldehyde dehydrogenase 1 family member L2 (ALDH1L2), which catalyzes the conversion of 10-formyl-THF to THF and CO[2] while generating NADPH, was reduced in cells treated with AD and PierA ([148]Fig. 5C). In comparison, the AD treatment did not affect the cytosolic SHMT1 or the mitochondrial outer membrane protein translocase of outer mitochondrial membrane 20 (TOMM20). Despite protein down-regulation, quantitative real-time PCR (qRT-PCR) further showed increased transcription of MTHFD2, SHMT2, and ALDH1L2 under AD treatment ([149]Fig. 5D). This transcriptional response may represent a compensatory response by the AD-treated cell to restore mitochondrial 1C. Immunoblot analysis revealed that SHMT2 and ALDH1L2 were consistently down-regulated in PTPMT1 KO GM12878 and GM12881 LCLs, while MTHFD2 was down-regulated in PTPMT1 KO GM12881 but unchanged in PTPMT1 KO GM12878 ([150]Fig. 5E). Given these findings, we further tested whether SHIN1, a dual inhibitor of SHMT1 and SHMT2 ([151]54), can synergize with AD to achieve synthetic lethality in EBV-transformed B cells. We assessed cell viability after 48 hours of drug treatment using two LCLs (GM12878 and GM15892), two EBV^+ BLs (Daudi and MUTU I), and two EBV^− cell lines [BCBL1 primary effusion lymphoma (PEL) and Reh]. Synthetic lethality was particularly evident in GM12878 and GM15892 LCLs, as well as MUTU I and Daudi BLs, but to a lesser extent in EBV^− BCBL1 PEL and Reh cells. While BCBL1 PEL cells were very sensitive to SHIN1, they showed little sensitivity to AD treatment ([152]Fig. 5F). Impaired CL biosynthesis leads to dysfunction of glutamate oxaloacetate transaminase 2–driven aspartate biosynthesis in EBV-transformed B cells Metabolomic analysis of intracellular amino acids in AD-treated, newly infected B cells revealed a marked decrease in the levels of aspartate and ornithine ([153]Fig. 6A and fig. S6). Aspartate plays a critical role in the growth and proliferation of cancer cells, serving not only as an amino acid for protein synthesis but also as a key precursor for the de novo synthesis of purines and pyrimidines ([154]55, [155]56). During the hyperproliferative phase of EBV B cell transformation, typically occurring between 3 and 7 DPI, the infected B cells divide approximately every 8 hours. This rapid division places a high demand on the synthesis of purines, pyrimidines, and proteins, which must rely on the sustainable aspartate flux. Fig. 6. CL biosynthesis sustains aspartate synthesis in EBV-transformed B cells. [156]Fig. 6. [157]Open in a new tab (A) Heatmap visualization of z-scores of amino acids detected in metabolomic analysis of uninfected or EBV-infected 4 DPI B cells, treated with DMSO and 2 μM AD. (B) FACS CFSE analysis of newly EBV-infected primary human B cells, treated with DMSO or 100 μM AOA for 5 days. All blots are representative of n = 3 replicates. (C) Immunoblot analysis for V5-tagged proteins and actin in WCL of Cas9^+ GM12878 LCL expressing GFP or SLC1A3. (D) Percentage of live cell number of GM12878 LCL expressing GFP or SLC1A3, treated with DMSO or 1 μM AD or 1 μM AD + 1 mM Asp for 48 hours. (E) U-^13C-glutamine tracing schematic showing B cells at 4 DPI treated with DMSO or 2 μM AD, with malate(m + 3) measuring reductive carboxylation and malate(m + 4) measuring glutamine oxidation. (F) Stacked bar charts of ^13C-glutamine–labeled malate or aspartate. (G) Fraction of m + 4 or m + 3 malate or Asp in DMSO- or AD-treated cells at 4 DPI. (H) Immunoblot analysis for GOT1, GOT2, TOMM20, and actin in WCL of indicated LCLs expressing control or PTPMT1 sgRNAs. (I) qRT-PCR analysis of LONP1 mRNA levels (relative to ACTB) in indicated LCLs expressing control or PTPMT1 sgRNAs. (J) Immunoblot analysis for PTPMT1, GOT2, and actin in WCL of GM12878 treated with DMSO or 0.5 μM LONP1-in-2 for indicated time. (K) Immunoblot analysis for PTPMT1, GOT2, and actin in WCL of GM12878 LCL expressing control or PTPMT1 sgRNAs, treated with DMSO or 0.5 μM LONP1-in-2 for 24 hours. Means ± SD values in (D) and (G) (n = 3 experiments) were analyzed using one-way ANOVA with Tukey’s test. (F) (n = 3 experiments) used two-way ANOVA with Bonferroni’s test. (I) (n = 3 experiments) used Student’s t test. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. All blots were representative of n = 3 experiments. While human plasma has low aspartate levels, certain cancer cells can uptake aspartate from their surroundings through the expression of aspartate membrane transporters ([158]55, [159]57). Solute carrier family 1 member 3 (SLC1A3), the primary transporter of aspartate ([160]55), exhibits minimal to negligible transcription levels during EBV transformation in primary human B cells, particularly when compared to more common amino acid transporters such as solute carrier family 1 member 3 (SLC1A5) (fig. S8A) ([161]28). The low SLC1A3 protein expression was even shut down once the newly infected B cells fully transformed into LCL (fig. S8B, 28 DPI). This suggests that aspartate synthesis plays a major role in sustaining rapid cellular proliferation in EBV-transformed B cells. Aspartate biosynthesis is intricately linked to cellular metabolism and energy production. The enzymes glutamate oxaloacetate transaminase 1 (GOT1) and GOT2, which function in the cytosol and mitochondria, respectively, play critical roles in this process. In EBV-infected B cells, there is a significant up-regulation of GOT1 and GOT2 expressions (fig. S8, A and B). In cells with intact mitochondrial respiration, the primary site of aspartate synthesis is the mitochondria, where GOT2 catalyzes the conversion of glutamate and oxaloacetate into aspartate (fig. S8C). This mitochondrial activity is crucial not only for producing aspartate but also for linking amino acid metabolism to the TCA cycle and the overall energy status of the cell ([162]55, [163]56, [164]58, [165]59). We therefore investigated the role of aspartate biosynthesis in EBV B cell transformation. We infected primary B cells with EBV for 2 hours to ensure equal viral entry. These cells were labeled with carboxyfluorescein diacetate succinimidyl ester (CFSE), a fluorescent dye commonly used for cell proliferation, and treated with DMSO or aminooxyacetic acid (AOA), a dual inhibitor of GOT1 and GOT2. EBV-driven cell proliferation was assessed by reduced CFSE intensity at 5 DPI using flow cytometry. We found that AOA inhibition of aspartate synthesis halted the EBV B cell transformation ([166]Fig. 6B), highlighting the critical role of aspartate synthesis in EBV transformation. To elucidate the role of CL biosynthesis in sustaining aspartate biosynthesis in EBV-transformed B cells, we next tested whether adding exogenous aspartate could rescue AD-induced cell death in GM12878 LCLs. Since EBV^+ LCLs only express minimal level of SLC1A3, we created a stable GM12878 LCL expressing SLC1A3. A control cell line expressing green fluorescent protein (GFP) was also established ([167]Fig. 6C). Those cells were treated with DMSO or AD. Aspartate (1 mM) was used to rescue the AD effects on cell growth defects. We found that exogenous aspartate was tentative to rescue the cells in the GFP group by 48 hours after AD treatment. This rescue effect was significantly enhanced in the SLC1A3-expressing cells ([168]Fig. 6D). However, the rescue was not complete, suggesting that CL inhibition–induced growth defects are at least partially associated with aspartate deficiency. Previous studies have shown that aspartate biosynthesis is closely regulated by functional mitochondrial respiration. Disruption in the NADH/NAD^+ ratio caused by complex I inhibition halts aspartate synthesis. This is because malate dehydrogenase 2 (MDH2) in the mitochondria uses NAD^+ to oxidize malate to oxaloacetate, which is the precursor of aspartate ([169]58, [170]59). In AD-treated newly infected B cells, NADH /NAD^+ ratio was not changed in comparison to that in DMSO-treated control cells (fig. S6), suggesting that AD treatment may use different mechanisms to inhibit aspartate synthesis. Two alternative cytosolic pathways related to pyruvate can compensate for aspartate biosynthesis when mitochondrial respiration is disrupted: (i) MDH1 reversely converts malate to oxaloacetate using NAD^+ as an electron acceptor, with GOT1 catalyzing reverse transamination to produce aspartate. Pyruvate may enhance this pathway by converting NADH to NAD^+ via lactate dehydrogenase (LDH). (ii) Pyruvate carboxylase (PC) can convert pyruvate to oxaloacetate (fig. S9A). Increased LDHA and PC expressions have been noted at the early stage of EBV transformation as shown by previous RNA-seq (fig. S8A) ([171]28). We therefore examined whether exogenous pyruvate could restore AD-induced aspartate deficiency. Adding exogenous pyruvate did not prevent cell death in GM12878 LCLs with AD treatment (fig. S9B), indicating neither PC nor MDH1 pathways could compensate for aspartate biosynthesis in EBV^+ LCLs under inhibition of CL biosynthesis. Our metabolomic analysis of newly infected B cells revealed that AD treatment increased TCA cycle metabolite levels, contrary to PierA treatment, which completely shut down the upper TCA cycle (fig. S6). Notably, a significant increase in glutamine, α-ketoglutarate (αKG), and citrate were observed in AD-treated cells (fig. S6). Given that AD treatment inhibited mitochondrial respiration in EBV-transformed B cells, the glutamine reductive carboxylation may be alternatively activated to generate citrate for aspartate biosynthesis in the cytosol ([172]60). To explore this, we conducted a parallel isotopic tracing study using [^13C[5]]-glutamine and [^13C[6]]-glucose. We collected newly transformed B cells at 4 DPI and treated them with either DMSO or AD for 16 hours. We then introduced [^13C[5]]-glutamine (m + 5) or [^13C[6]]-glucose (m + 6) to these cells for an additional 8-hour period to allow ^13C integration. The schematic in [173]Fig. 6E depicts the metabolic fate of [^13C[5]]-glutamine: Oxidation within the mitochondria generates m + 4–labeled compounds through the conventional TCA cycle, while reductive carboxylation produces m + 3–labeled compounds in the cytosol. Notably, after 8 hours of [^13C[5]]-glutamine tracing, approximately 69% of total aspartate was labeled with ^13C in DMSO-treated cells (fig. S9C). AD treatment significantly increased the levels of ^13C-labeled malate ([174]Fig. 6F), while the levels of ^13C-labeled aspartate decreased ([175]Fig. 6F). The fraction of m + 4 malate (produced through the oxidative TCA cycle) was significantly increased in AD-treated cells compared to DMSO-treated cells ([176]Fig. 6G). Conversely, the fraction of m + 3 malate (produced through reductive carboxylation) was reduced under AD treatment ([177]Fig. 6G), suggesting that glutamine reductive carboxylation was not increased after AD treatment. Collectively, the accumulation of the [^13C[5]]-glutamine–derived m + 4 TCA metabolites and down-regulation of the aspartate after AD treatment implicate a possibility of GOT2 deficiency. In addition, in the parallel [U^13C[6]]-glucose tracing experiment (fig. S9D), approximately 30% of total aspartate was derived from glucose (fig. S9, E and F), further indicating that glutamine anaplerosis predominantly fuels aspartate biosynthesis in newly infected B cells. AD treatment significantly reduced [^13C[6]]-glucose–derived aspartate and acetyl–coenzyme A (CoA) (fig. S9G). Along with the decreased ECAR observed in AD-treated cells at 4 DPI (fig. S5E), these findings suggest that inhibition of CL biosynthesis may also suppress EBV-driven glycolysis in newly infected B cells. We used CRISPR-Cas9 to individually knock out GOT1 and GOT2 in Cas9^+ GM12878 and GM12881 LCLs. Knocking out GOT1 with two different sgRNAs slightly reduced cell growth in both cell lines (fig. S9, H and I). In contrast, GOT2 KO resulted in significant cell death (fig. S9, J and K), underscoring the essential role of GOT2 in maintaining cell viability in LCLs. Our immunoblot analyses showed a marked reduction in GOT2 protein levels without corresponding decreases in its mRNA in PTPMT1 KO LCLs, while GOT1 protein levels remained unchanged ([178]Fig. 6H and fig. S9L). We hypothesize that inhibiting CL biosynthesis may lead to a mitochondrial unfolded protein response (mtUPR), activating mitochondrial proteases that degrade improperly assembled proteins. Notably, PTPMT1 KO LCLs displayed elevated mRNA levels of Lon peptidase 1 (LONP1) ([179]Fig. 6I), one of the major mitochondrial proteases activated by the mtUPR, which degrades unassembled matrix protein complexes ([180]61). Treating GM12878 LCLs with 0.5 μM LONP1-in-2, a selective LONP1 inhibitor at concentrations below 10 μM ([181]62), resulted in an increase in PTPMT1 and GOT2 levels ([182]Fig. 6J). Furthermore, in PTPMT1 KO GM12878 cells, LONP1 inhibition stabilized GOT2 expression ([183]Fig. 6K), indicating that LONP1 is the mitoprotease responsible for regulating GOT2 turnover. Collectively, our findings suggest that CL biosynthesis is crucial for GOT2 stability and thereby essential for sustaining mitochondrial aspartate biosynthesis in EBV^+ LCLs. DISCUSSION Shortly after entry, EBV actively initiates aerobic glycolysis in B cells through the translational activation of all glycolytic enzymes and quickly down-regulates thioredoxin-interacting protein (TXNIP), a strong glucose transporter 1 (GLUT1) negative regulator. Glycolytic enzymes, including hexokinase 2 (HK2), are up-regulated by 2 DPI. Glucose consumption and lactate release are evident at 2 DPI and peak at 4 DPI ([184]12). The reliance on glucose for viral transformation is further underscored by the impaired growth of B cells cultured in galactose-containing medium, which forces cells to rely on mitochondrial oxidative metabolism rather than glycolysis. The induction of the Warburg effect, along with the up-regulation of OXPHOS, is critical in supporting the proliferation of B cells during viral transformation ([185]12). In addition, EBV early infection also boosts key fatty acid synthesis enzymes, acetyl-CoA carboxylase 1 (ACC1) and fatty acid synthase (FASN), to convert acetyl-CoA derived from glycolysis into palmitate, facilitating lipogenesis ([186]13). In this study, we show that EBV activates CL biosynthesis within 4 DPI. We speculate that the increase in glycerol-3-phosphate from glycolysis and acyl groups from lipogenesis may drive this process, leading to mitochondrial remodeling. Notably, our ECAR measurements and U^13C-glucose tracing experiments suggest that inhibiting CL synthesis suppresses glycolysis in EBV-infected B cells, linking CL biosynthesis to both OXPHOS and glycolysis. This underscores CL’s role in coordinating the metabolic network in EBV-infected B cells. Previous research shows that HK2 localizes to the outer mitochondrial membrane via its interaction with voltage-dependent anion channel 1 (VDAC1). VDAC1 is crucial for the transport of ions, ATP/adenosine diphosphate, metabolites, and respiratory substrates across the outer membrane ([187]63–[188]65). HK2 depends on mitochondrial ATP transported through VDAC1 to phosphorylate glucose, the first step of glycolysis ([189]64). We speculate that a reduction in mitochondrial ATP following CL biosynthesis inhibition may impair HK2 activity, down-regulating glycolysis and ultimately contributing to transformation failure. However, the metabolic snapshot observed in newly transformed EBV-infected B cells treated with AD likely reflects a combination of direct metabolic effects by CL biogenesis inhibition and indirect consequences stemming from transformation defects. Key metabolic pathways, including OXPHOS, glycolysis, and 1C metabolism, play crucial roles in EBV-driven transformation ([190]32). Inhibiting CL biosynthesis disrupts these pathways, halting the transformation process, which may, in turn, contribute to the observed metabolic alterations. For instance, cross-talk between 1C metabolism and epigenetic regulation has been shown to affect the transformation process ([191]14). Disruptions in 1C metabolism alter the availability of metabolites such as S-adenosylmethionine, a key methyl donor required for DNA and histone methylation, which is essential for epigenetic regulation of the EBV genome and viral oncogene expression ([192]14). This disruption in 1C metabolism may lead to dysregulation of viral oncogene expression critical for successful transformation. In addition, our metabolomic data showed that AD treatment significantly increased αKG levels in newly infected cells at 4 DPI. αKG serves as a crucial cofactor for demethylases, including ten-eleven translocation 2 (TET2), a DNA demethylase ([193]66). These demethylases play essential roles in regulating EBV latent genome ([194]67, [195]68). We speculate that AD-induced αKG elevation may enhance the activity of demethylases, thereby influencing the methylation program of the EBV genome during de novo infection. This could suppress transformation and ultimately feedback into metabolic regulation. MYC is a crucial regulator of metabolism and mitochondrial biogenesis ([196]40, [197]41, [198]69). In newly infected B cells, MYC is up-regulated by the EBNA2 superenhancer ([199]38), peaking around 2 DPI ([200]12), coinciding with the activation of CRLS1 and PTPMT1. We found that MYC is sufficient to drive CRLS1 and PTPMT1 expression in absence of EBNA2, supporting CL biosynthesis. Thus, we propose that EBNA2 first activates MYC, which, in turn, predominantly drives CRLS1 and PTPMT1. However, EBNA2 likely has a broader role in regulating CL biosynthesis during early infection. As a non–DNA binding transcriptional coactivator, EBNA2 reprograms chromatin structure, enabling transcription factors such as early B cell factor 1 (EBF1) and recombination signal binding protein for immunoglobulin kappa J region (RBP-jκ) to access previously inaccessible regions of the genome ([201]37, [202]70). This implies that EBNA2 not only up-regulates MYC but also actively remodels chromatin around CL biosynthesis genes, enhancing the expression of key metabolic regulators. EBNA2 is known to interact with superenhancers and modify chromatin topology through phase separation, reorganizing nuclear architecture to promote the transcription of genes crucial for cell proliferation and metabolic shifts ([203]38). We found that inhibiting an EBNA2-mediated enhancer suppressed CRLS1 expression in LCLs, further supporting EBNA2’s regulatory role. CL contributes to the structural integrity and functionality of mitochondrial cristae. While there is ongoing debate regarding CL’s cone-shaped structure’s direct contribution to cristae curvature, it interacts with numerous mitochondrial protein complexes to support cristae formation. The Mitochondrial Contact Site and Cristae Organizing System (MICOS) play crucial roles in shaping and maintaining cristae structure ([204]71). Notably, EBV transcriptionally up-regulates MICOS genes, including MIC60, during B cell transformation ([205]28). The loss of CL is associated with MICOS disruption, leading to cristae structure loss and mitochondrial dysfunction, as observed in fibroblasts from patients with Barth syndrome ([206]71). How EBV-driven CL biosynthesis coordinates with MICOS to support cristae biogenesis remains to be investigated. Furthermore, the dimerization of F1Fo-ATP synthase and the integration of complex II into the respirasome are crucial for cristae curvature formation. CL is known to bind to and stabilize all ETC complexes and ATP synthase (complex V), modulating their stability and function ([207]72–[208]76). Our findings indicate that PTPMT1 KO results in a highly disordered tubular structure in the mitochondria of newly infected B cells. We are tempted to speculate that disruptions in the assembly and integration of protein complexes into the cristae may contribute to the observed disorder. Supporting this hypothesis, our BN-PAGE data revealed a deficiency in complex II, where the assembly of the succinate dehydrogenase complex flavoprotein subunit A (SDHA) subunit was disrupted because of PTPMT1 KO in EBV-transformed B cells. We found that blocking CL biosynthesis leads to the down-regulation of mitochondrial 1C enzymes driven by EBV, resulting in a significant reduction of NADPH production. This finding reveals a previously unrecognized role of CL in maintaining mitochondrial 1C metabolism. As mentioned above, CL’s ability to bind to and interact with a wide variety of mitochondrial protein complexes was well documented. The interaction between CL and proteins involves strong hydrophilic interactions facilitating the binding between its negative charged glycerol head group and various amino acid residues of the protein ([209]25). We speculate that during EBV B cell transformation, CL in the IMM probably binds to mitochondrial 1C enzymes, which may promote 1C enzymes to form metabolons to meet the increased demand for nucleotides and NADPH. Inhibiting CL biosynthesis may disrupt the assembly of these protein complexes and lead to the activation of mtUPR. This response activates mitochondrial proteases involved in the degradation of unassembled 1C enzymes ([210]61). Further biochemical and structural studies are essential to thoroughly understand the interactions between CL and mitochondrial 1C enzymes and to elucidate how these interactions contribute to EBV-induced lymphomagenesis. Our research has demonstrated that aspartate, crucial for protein and nucleotide synthesis during EBV-driven cell proliferation, is predominantly synthesized in mitochondria from glutamate and oxaloacetate by GOT2. This pathway represents a significant metabolic vulnerability in EBV-transformed B cells. Notably, CL plays a vital role in stabilizing GOT2 throughout the transformation process, thereby ensuring sustainable aspartate synthesis. While some of the most aggressive cancers, characterized by compromised respiration, depend on GOT1-driven pathways to compensate for aspartate synthesis, our findings from pyruvate rescue experiments and isotopic tracing suggest that EBV-transformed B cells lack the ability to use these major alternative pathways. This implies that directly targeting GOT2 could be considered as an effective strategy for treating lymphomas associated with EBV B cell transformation. Furthermore, our data reveal that in the initial stages of B cell transformation, glutamine oxidation via the TCA cycle is the primary source of aspartate synthesis. These insights suggest that cotargeting CL biosynthesis and glutaminolysis may offer innovative approaches to inducing aspartate depletion, presenting a promising avenue for the treatment of EBV-associated lymphomas. A primary limitation of our study is its reliance on an in vitro transformation model, which may not fully capture the dynamics of the in vivo lymphoid microenvironment affecting EBV-driven mitochondrial biogenesis. This gap has been partially bridged by a recent study using a humanized mouse model, which presents a promising approach to further exploring the role of CL synthesis in EBV transformation in a more physiologically relevant setting ([211]52). Our findings also indicate that AD significantly curtails EBV B cell transformation and exhibits a potent synthetic lethal effect when combined with SHIN1, an SHMT inhibitor. This finding supports the potential for in vivo validation using a mouse xenograft model. However, the intraperitoneal administration of AD poses challenges due to its unknown pharmacokinetics and safety profile when used systemically; it has traditionally been used as an oral rinse to prevent gingivitis ([212]77). Pilot mouse studies will be critical to determine its feasibility for systemic use. Given the rapid clearance of SHIN1 in vivo and the subsequent development of SHIN2 with improved pharmacokinetic properties ([213]14, [214]78), we are inspired to develop a generation of PTPMT1 inhibitors. These could similarly be optimized for stability and efficacy in vivo, ensuring that our approach remains adaptable and responsive to the unique challenges of systemic drug delivery in EBV lymphoma treatments. While our investigations have centered on B cell models, EBV also infects epithelial cells and, less frequently, T and natural killer cells. It remains to be explored whether CL biosynthesis plays a crucial role in other EBV-related diseases, such as NPC and GC. Moreover, our focus on peripheral blood B cells leaves open the possibility that CL phenotypes might differ in EBV-infected germinal center or memory B cells within secondary lymphoid tissues. In summary, we demonstrate that early EBV infection in B cells leverages CL biosynthesis pathways to establish specialized mitochondria, facilitating extensive metabolic network remodeling. CL stabilizes key enzymes involved in mitochondrial aspartate biosynthesis and 1C metabolism. In addition, our isotopic tracing data reveals that glutamine serves as the primary carbon source fueling aspartate synthesis, thereby supporting EBV B cell transformation ([215]Fig. 7). Targeting CL biosynthesis and using CL-related synergistic combinations could serve as a promising therapeutic approach for treating EBV-associated cancers. Fig. 7. Schematic model of EBV-driven CL biosynthesis sustains cristae biogenesis and metabolic remodeling. [216]Fig. 7. [217]Open in a new tab MATERIALS AND METHODS Cell lines and reagents EBV^+ P3HR-1 and Daudi BL cell lines were obtained from American Type Culture Collection (ATCC). EBV^+ MUTU I BL was obtained from B. Gewurz (originally a gift from J. Sample). EBV^+ Rael BL was obtained from L. Guilino-Roth. EBV^+ GM12878, GM12892, GM12881, GM14890, and GM15892 LCLs were obtained from the Coriell Institute. 2-3-3 E2HT LCLs, expressing EBNA2 fused to an estrogen receptor 4HT-binding domain, were obtained from B. Zhao. In the presence of 1 μM 4HT (SML1666, Sigma-Aldrich), EBNA2-HT localizes to the nucleus; without 4HT, it relocalizes to the cytosol and becomes destabilized. To remove 4HT, cells underwent five washes with 4HT-free medium, the last two washes lasting 30 min each before reseeding at 300,000 cells/ml. Cells were then grown for an additional 48 hours before cell lysate preparation. The conditional P493-6 LCLs, a gift from B. Gewurz, contain a conditional EBNA2-HT allele and an exogenous Tet-OFF MYC allele. Cells were washed three times with phosphate-buffered saline (PBS), then seeded at 0.3 million/ml in RPMI 1640 medium with 10% doxycycline-free fetal bovine serum (FBS), and treated under various conditions for 48 hours as specified. To culture P493-6 cells at a low MYC state, they were grown without 4HT and with 1 μg/ml doxycycline (HY-N0565, MedChemExpress). For a high MYC BL-like state, both doxycycline and 4HT were removed. Reh ALL cell line was obtained from B. Gewurz. Human embryonic kidney (HEK) 293T cell line was obtained from ATCC. All the BL, ALL, DLBCL, LCL, and PEL cell lines, as well as primary B cells, were cultured in RPMI 1640 medium (Gibco, Life Technologies) with 10% FBS (Gibco) or with 10% dialyzed FBS where indicated (Gibco). HEK-293T cells were cultured in Dulbecco’s modified Eagle’s medium (Gibco) with 10% FBS. To obtain the stable Streptococcus pyogenes Cas9 expression, cell lines were treated with lentiviral transduction and blasticidin (5 μg/ml; ant-bi-1, InvivoGen) selection. To select the transduced cells, puromycin (3 μg/ml; A11138-03, Gibco) or hygromycin (100 μg/ml; 10687010, Thermo Fisher Scientific) were added to the postinfection cells. All cells were cultured at 37°C in 5% CO[2] incubator. GM12878, GM12892, GM12881, and GM15892 LCLs were treated with 1 μM AD (A4542, APExBIO) at a density of 2 × 10^5/ml for 48 hours, refreshing AD after 24 hours and diluting the cells 1:1 with culture medium. Primary B cells treated with 2 μM AD for 24 hours at a density of 3 × 10^5 cells/ml. The AD treatment details are specified in figure legends. For antioxidant rescue analysis with NAC (7874, Tocris) and GSH (70-18-8, Millipore), GM12878 cells were cotreated with 1 μM AD and either 1.5 mM NAC or 2 mM GSH, refreshing treatments after 24 hours, and cell viability was measured at 48 hours after the first treatment. GM12878 LCLs expressing GFP or SLC1A3, after puromycin selection, received DMSO, 1 μM AD, or 1 μM AD plus 1 mM l-aspartate (A1330000, Sigma-Aldrich) for 48 hours. For pyruvate supplementation studies, GM12878 LCL cells were treated with DMSO or 1 μM AD and 0, 1, or 5 mM sodium pyruvate (P5280, Sigma-Aldrich) for 48 hours. Synthetic lethal analyses were performed using GM12878 LCL, GM15892 LCL, MUTU I BL, Daudi BL, and Reh ALL, treated with DMSO, 1 μM AD, 10 μM SHIN1, or a combination of 1 μM AD and 10 μM SHIN1 for 48 hours. Cas9^+ GM12878 cells expressing either control or PTPMT1 sgRNAs were seeded at a density of 0.2 × 10^6 cells/ml on days 8 to 10 after puromycin selection, a window at which we began to observe GOT2 degradation without cell death. LONP1-in-2 (HY-153034, MedChemExpress), a LONP1 selective inhibitor ([218]62), was applied at a concentration of 0.5 μM for 12 hours, with DMSO used as the vehicle control. Human primary B cells isolation Discarded, deidentified leukocyte fractions left over from platelet donations were obtained from the Brigham and Women’s Hospital Blood Bank. Blood cells were collected from platelet donors following institutional guidelines. Since these were deidentified samples, the gender was unknown. Our studies on primary human blood cells were approved by the Tufts University Institutional Review Board (Tufts IRB: STUDY00004385). Primary human B cells were isolated by negative selection using RosetteSep Human B Cell Enrichment and EasySep Human B cell enrichment kits (STEMCELL Technologies), according to the manufacturers’ protocols. B cell purity was confirmed by plasma membrane CD19 expression through flow cytometry. Cells were then cultured with RPMI 1640 medium with 10% FBS. EBV production and concentration The EBV B95-8 strain was generated from B95-8 cells engineered for inducible Z Transactivator (ZTA) expression (a gift from B. Gewurz). The activation of EBV lytic cycle was achieved by treating the cells with 1 μM 4HT for 24 hours. Subsequently, the 4HT was removed, and the cells were cultured in RPMI 1640 medium supplemented with 10% FBS, devoid of 4HT, for an additional 96 hours. The viral supernatants obtained were then cleared of producer cells by passing through a 0.45-μm filter. The viral titer was assessed using a transformation assay. Similarly, the P3HR-1 strain of EBV was obtained from a P3HR-1 ZHT/RHT cell line that expresses 4HT-inducible Z transactivator protein fused to a modified hydroxy-tamoxifen (HT)-responsive receptor (ZHT) and R transactivator protein fused to a modified hydroxy-tamoxifen (HT)-responsive receptor (RHT) alleles, generously provided by B. Gewurz. The induction process involved treating P3HR1 ZHT/RHT cells with 1 μM 4HT for a 24-hour period. Afterward, the culture medium was replaced with fresh RPMI 1640/FBS medium, and the cultures were allowed to incubate for 96 hours to collect the virus-rich supernatants. These supernatants were then filtered using a 0.45-μm filter for purification. The supernatant was transferred to an ultracentrifuge tube (326823, Beckman Coulter) and centrifuged at 25,000 rpm for 2 hours at 4°C in an ultracentrifuge (OPTIMA XPN-100, Beckman Coulter). The viral pellet was resuspended and aliquoted in PBS with 2% dialyzed FBS and stored at −80°C until infection. The genomic DNA of virus was quantified by PCR targeting the EBV DNA polymerase (BALF5) gene from the extracted viral genome. This quantification was used to standardize the virus amounts for cell infection experiments. Transmission electron microscopy A mixture of 2.5% paraformaldehyde, 5% glutaraldehyde, and 0.06% picric acid in 0.2 M cacodylate buffer was freshly prepared before use, and then the above mixture was diluted 1:1 with distilled H[2]O. To fix primary B cells, 1 million cells were collected and washed one time with Dulbecco’s PBS (14190, Gibco), removed the residue buffer, and then gently added the diluted fixative to the primary B cells. The cells were fixed at room temperature for 1 hour. The cells were then postfixed for 30 min in 1% osmium tetroxide (OsO[4])/1.5% potassium ferrocyanide (KFeCN[6]), washed three times in water, and incubated in 1% aqueous uranyl acetate for 30 min. Samples were then washed twice in water and dehydrated in grades of alcohol (5 min each; 50, 70, 95, and 2× 100%). Cells were removed from the dish in propyleneoxide, pelleted at 3000 rpm for 3 min, and infiltrated for 2 hours in a 1:1 mixture of propyleneoxide and TAAB Epon (Marivac Canada Inc., St. Laurent, Canada). Samples were subsequently embedded in TAAB Epon and polymerized at 60°C for 48 hours. Ultrathin sections (about 60 nm) were cut on a Reichert Ultracut-S microtome, picked up on to copper grids stained with lead citrate, and examined in a JEOL 1200EX TEM or a TecnaiG^2 Spirit BioTWIN. Images were recorded with an AMT 2000 charge-coupled device camera. CRISPR-Cas9 editing CRISPR-Cas9 KO was performed in cells with stably Cas9 expression, using Broad Institute Brunello or Avana library sgRNA sequences as listed in table S3. CRISPRi was performed in GM12878 LCL expressing dCas9 fused with Krüppel-associated box domain (KRAB) ([219]79). sgRNAs against targeting upstream CRLS1 enhancers were designed using the sgRNA Designer from the Genetic Perturbation Platform at the Broad Institute. sgRNA oligos were synthesized from Integrated DNA Technologies and cloned into the pLentiGuide-Puro vector (Addgene plasmid #52963, a gift from F. Zhang). Lentiviruses were produced in HEK-293T cells by cotransfection of pLentiGuid-puro with psPAX2 and vesicular stomatitis virus glycoprotein packaging. At 24 hours after transfection, cell culture medium was changed to RPMI 1640 + 10% FBS. Two rounds of lentiviral transduction were performed at 48 and 72 hours after plasmid transfection. Cells were selected by puromycin (3 μg/ml), which was added 48 hours after transduction. Depletion of target gene–encoded protein expression was confirmed by immunoblot. cDNA rescue The PTPMT1 cDNA with silent PAM site mutations was purchased from Integrated DNA Technologies and was inserted into pLX-TRC313 (a gift from J. Doench) by HifiAssembly (New England Biolabs). Cas9^+ GM12878 with stable C-terminal V5 epitope–tagged PTPMT1 cDNA expression was established by lentiviral transduction and hygromycin selection as described above. Ten days after hygromycin selection, PTPMT1-V5 expression was confirmed by immunoblot. The sequence of PTPMT1 rescue cDNA is listed below. sgPTPMT1 targeting sequences are highlighted in underlined bold. PAM sequences are underlined. Mutation sites are indicated in bold italics. Overlapping sequences for HifiAssembly reaction are highlighted in underlined. PTPMT1^R cDNA, TCTTCCATTTCAGGTGTCGTGAGGCTAGCATGGCGGCCACCGCGCTGCTGGAGGCCGGCCTGGCGCGGGTG CTCTTCTACCCGACGCTGCTCTACACCCTGTTCCGCGGGAAGGTGCCGGGTCGGGCGCACCGGGACTGGTA CCACCGCATCGACCCCACCGTGCTGCTGGGCGCGCTGCCGTTGCGGAGCTTGACGCGCCAGCTGGTACAGG ACGAGAACGTGCGCGGGGTGATCACCATGAACGAAGAGTACGAGACGAGGTTCCTGTGCAACTCTTCACAG GAGTGGAAGAGACTAGGAGTCGAGCAGCTGCGGCTCAGCACAGTAGACATGACTGGGATCCCCACCTTGGA CAACCTCCAGAAGGGAGTCCAATTTGCTCTCAAGTACCAGTCGCTGGGCCAGTGTGTTTACGTGCATTGTA AGGCTGGGCGCTCCAGGAGTGCCACTATGGTGGCAGCATACCTGATTCAGGTGCACAAATGGAGTCCAGAG GAGGCTGTAAGAGCCATCGCCAAGATCCGGTCATACATCCACATCAGGCCTGGCCAGCTGGATGTTCTTAA AGAGTTCCACAAGCAGATTACTGCACGGGCAACAAAGGATGGGACTTTTGTCATTTCAAAGACAGATATCG GTAAGCCTATCCCTAACCCTC. Mitochondria isolation The mitochondria were isolated following the manufacturer’s protocols of Mitochondria Isolation Kit for Cultured Cells (89874, Thermo Fisher Scientific). Specifically, 20 million cells were harvested into a 2.0-ml microcentrifuge tube and centrifuged at 850g for 2 min, the supernatant was removed, and then the cell pellet was added with 800 μl of Mitochondria Isolation Reagent A containing EDTA-free protease inhibitor (cOmplete, MilliporeSigma). The mixture was gently vortexed and incubated on ice for exactly 2 min. Ten microliters of Mitochondria Isolation Reagent B was then added to the mixture and incubated on ice, fully vortexing the mixture at every minute. After 5 min, 800 μl of Mitochondria Isolation Reagent C containing EDTA-free protease inhibitor was included. The tube was gently inverted several times before centrifuging at 700g for 10 min at 4°C. The supernatant was transferred to a new, 2.0-ml tube and centrifuged at 12,000g for 15 min at 4°C. The obtained mitochondrial pellet was resuspended in 100 μl of Mitochondria Isolation Reagent C containing EDTA-free protease inhibitor. The protein content of mitochondria was further measured by Qubit 4 Fluorometer ([220]Q33226, Thermo Fisher Scientific) with Qubit Protein Assay kit ([221]Q33212, Thermo Fisher Scientific). CL assay CL of isolated mitochondria was determined using the CL Assay Kit (#ab241036, Abcam). GM12878 cells were treated with either vehicle control or 2 μM AD for 24 hours. Isolated mitochondria from Cas9^+ GM12878 cells expressing control or PTPMT1-targeting sgRNAs were also analyzed. Cells were collected before the cell death caused by inhibition of CL synthesis. Mitochondria from those cells were isolated as described. Twenty micrograms of mitochondria from each sample were resuspended in CL assay buffer and mixed with CL probe following the protocol. The fluorescence at excitation/emission of 340/480 nm was recorded after incubating at room temperature for 10 min. Protein concentration was determined for normalization, and the mitochondrial CL content was displayed as nanomoles per milligram of protein. BN gel electrophoresis and immunodetection Sample preparation The BN gel electrophoresis and Immunoblot was conducted as previously described ([222]80). The NativePAGE sample prep kit (BN2008, Thermo Fisher Scientific) was used to make the mitochondrial Sample buffer cocktail. Fifty micrograms of mitochondrial protein was mixed with 5 μl of 4× NativePAGE sample buffer, 8 μl of 5% digitonin, and 7 μl of water in the kit. Then, we incubated the solubilized mitochondria on ice for 20 min. After that, solubilized mitochondria were then centrifuged at 20,000g for 10 min at 4°C. Fifteen microliters of supernatant was transferred into a new tube and fully mixed with 2 μl of Coomassie G-250 sample additive in kit. Electrophoresis Each well (NativePAGE, 3 to 12% gradient gel, BN2011, BX10, Thermo Fisher Scientific) was gently washed with 1 ml of dark blue 1× cathode buffer (BN2002, Thermo Fisher Scientific). Subsequently, 15 μl of mitochondrial sample was loaded into the gel. The inner chamber was filled with 1× dark blue cathode buffer, and 600 ml of 1× running buffer (BN2002, Thermo Fisher Scientific) was added to the outer chamber. Electrophoresis took place in an XCell SureLock Electrophoresis-Cell (EI0001, Novex) at a constant voltage of 150 V for 30 min, with the current limited to 15 mA. Following this, the buffer in the inner chamber was replaced with light blue buffer (created by mixing 20 ml of dark blue 1× cathode buffer with 180 ml of distilled water). Electrophoresis continued at 250 V for 60 min. Immunoblot The BN-PAGE gel was gently washed with water for 5 min to remove the cathode buffer. It was then placed in bicarbonate transfer buffer (10 mM NaHCO[3] and 3 mM Na[2]CO[3]) for a 15-min incubation. The polyvinylidene difluoride membrane was activated by immersion in 100% methanol for 2 s and then washed with water for 5 min, followed by a 5-min incubation in bicarbonate transfer buffer. The transfer was carried out at a constant current of 300 mA for 1 hour in the cold room. Following the transfer, the membrane was rinsed with PBS and then fixed with 8% acetic acid for 5 min. The membrane was subsequently washed three times with water, each for 5 min. To remove the Coomassie blue, the membrane was shaken three times with methanol, each for 5 min. This was followed by three water washes, each for 5 min. The membrane was incubated with 5% milk in PBS–Tween 20 (PBST) for blocking for 1 hour. The membrane was washed three times with PBST, each for 5 min, and then incubated with primary antibodies for the ETC complexes I, II, III, and IV and OXPHOS overnight at 4°C. The next day, the membrane was washed three times with PBS, each for 5 min, followed by a 1-hour incubation with the secondary antibody, and then washed three times with PBST for 5 min each. Chemiluminescent detection was performed on the membranes by LI-COR XF system. Lipidomic profiling analysis The intracellular lipidomic profiling was performed as described ([223]81). Newly isolated human B cells were mock infected or infected with EBV at a multiplicity of infection of 1 for 7 and 10 days. B cells were counted and pelleted at 1200 rpm for 5 min at 4°C with an equal number of cells in each sample. Lipidomic profiling was performed as described previously ([224]81, [225]82). They were then resuspended in 200 μl of high-performance LC (HPLC)–grade water (270733, Sigma-Aldrich) and mixed vigorously with 2.5 ml of HPLC-grade methanol (A456, Thermo Fisher Scientific) in glass tubes. Following this, 5 ml of methyl tert-butyl ether (1634-04-4, Supelco) was added, and the samples were agitated for 1 hour at room temperature. To separate phases, 1.5 ml of water was added, and after vigorous vortexing, the samples were centrifuged at 1000g for 10 min at room temperature. The upper phase was then dried a speed vacuum concentrator (Savant SPD 1010, Thermo Fisher Scientific) for 4 hours at room temperature and stored at −80°C. For analysis, samples were reconstituted in 35 μl of a 1:1 mixture of LC-MS–grade isopropanol and methanol and subjected to LC-MS as previously outlined, using a high-resolution hybrid QExactive HF Orbitrap mass spectrometer (Thermo Fisher Scientific) set to data-dependent acquisition mode (top 8) with the capability of switching between positive and negative ion polarities. Lipid species identification and quantification were performed using the LipidSearch 4.1.30 software (Thermo Fisher Scientific), leveraging an internal database comprising ≥20 major lipid classes and ≥80 subclasses. For verifying signal linearity, a pooled sample was created by combining 5 μl from each sample, which was then diluted with a 1:1 mixture of isopropanol and methanol to generate dilutions of 0.3× and 0.1×, alongside a blank. These dilutions underwent analysis, and for each lipid species within this series, the Pearson correlation coefficient between ion count and sample concentration was computed. Only lipids exhibiting a correlation coefficient (r) greater than 0.9 were retained for final analysis. The abundance of individual lipid species was normalized against the total ion count of the sample. Using R, lipids were categorized by class, and the total ion intensity for each lipid class in each sample was calculated. Intracellular metabolite profiling The intracellular metabolite profiling was performed as described ([226]83). Newly infected B cells collected at 4 DPI were washed three times with PBS and counted. Six million cells were seeded into a T25 flask with 20 ml RPMI 1640 medium with 10% dialyzed FBS. Cells were incubated with DMSO, 2 μM AD, or 100 nM of PierA (2738-64-9, MedChemExpress) for 24 hours. The cells were counted and washed three times with prechilled PBS. The cell pellet was fully resuspended with 100 μl of PBS by vortex, and the metabolism was quenched by adding 3.3 ml of dry ice-cold 80% aqueous methanol (A456, Thermo Fisher Scientific) and kept at −80°C overnight. The lysate was centrifuged at 21,000g for 15 min at 4°C. The supernatants were obtained and dried by a speed vacuum concentrator (Savant SPD 1010, Thermo Fisher Scientific) for 4 hours at room temperature. Samples were resuspended using 20 μl of HPLC-grade water for MS. Five to 7 μl was injected and analyzed using a hybrid 6500 QTRAP triple quadrupole mass spectrometer (AB/SCIEX) coupled to a Prominence Ultra-fast liquid chromatography HPLC system (Shimadzu) via selected reaction monitoring (SRM) of a total of 300 endogenous water soluble metabolites for steady-state analyses of samples and 150 endogenous metabolites for ^13C/^15N isotopomer flux tracing. Some metabolites were targeted in both positive and negative ion modes for a total of 311 SRM transitions using positive/negative ion polarity switching. Electrospray ionization voltage was +4950 V in positive ion mode and –4500 V in negative ion mode. The dwell time was 3 ms per SRM transition, and the total cycle time was 1.55 s. Approximately 9 to 12 data points were acquired per detected metabolite. Samples were delivered to the mass spectrometer via hydrophilic interaction chromatography using a 4.6-mm inside diameter by 10-cm Amide XBridge column (Waters) at 400 μl/min. Gradients were run starting from 85% buffer B (HPLC-grade acetonitrile) to 42% B from 0 to 5 min; 42% B to 0% B from 5 to 16 min; 0% B was held from 16 to 24 min; 0% B to 85% B from 24 to 25 min; 85% B was held for 7 min to reequilibrate the column. Buffer A was composed of 20 mM ammonium hydroxide/20 mM ammonium acetate (pH 9.0) in 95:5 water:acetonitrile. Peak areas from the total ion current for each metabolite SRM transition were integrated using MultiQuant v3.0.2 software (AB/SCIEX). Metabolites with P < 0.05, log[2](fold change) > 1 or < −1 were used for pathway analysis using MetaboAnalyst 5.0 ([227]www.metaboanalyst.ca/MetaboAnalyst/ModuleView.xhtml). U-^13C-glutamine tracing and U^13C-glucose glucose tracing EBV-infected primary B cells were collected at 4 DPI, treated with DMSO or 2 μM AD for 16 hours, and then U-^13C-glutamine or U-^13C-glucose was applied to the cells. For U-^13C-glutamine tracing, 10 million cells were cultured in glutamine-free RPMI 1640 medium containing 10% dialyzed FBS and 2 mM U-^13C-glutamine (184161-19-1, Cambridge Isotope Laboratories) for 8 hours. For U-^13C-glucose tracing, 10 million cells were cultured in glucose-free RPMI 1640 medium containing 10% dialyzed FBS and 11.11 mM U-^13C-glucose (110187-42-3, Cambridge Isotope Laboratories) for 8 hours. Cell samples were collected and processed as mentioned in intracellular metabolite profiling. Metabolic flux analysis was performed as described ([228]84). Seahorse mitochondrial stress test The Agilent Seahorse XF Assay was conducted as described previously ([229]12). Specifically, the sensor cartridge was first hydrated with water overnight and incubated with XF Calibrant for 1 hour. Twelve microliters of Cell-Tak solution (1.3 ml of 0.1 M sodium bicarbonate, 11.2 μl of 0.1 M NaOH, and 22.4 μl of Cell-Tak solution) was added to each well of the V7-PS 96-well cell culture plate. The Cell-Tak solution was washed twice with sterile water, and 0.25 million primary B cells (resuspension in 180 μl of RPMI 1640 medium with 10% FBS and 5 mM pyruvate) were seed on a Seahorse plate. Then, the cells were placed in a non-CO[2] incubator at 37°C for 30 min. The OCR and ECAR were simultaneously recorded by a Seahorse XFe96 Analyzer (Agilent). The cells were sequentially probed by 20 μl of 3.5 μM oligomycin A, 20 μl of 2 μM carbonyl cyanide-p-trifluoromethoxyphenylhydrazone, and 20 μl of 100 nM Rotenone/Antimycin A. For Seahorse in GM12878 cells, cells were seeded as 0.1 million per well in a 96-well cell culture plate. Data were analyzed by Seahorse Wave Desktop software (Agilent). Flow cytometry analysis The mitochondrial mass was determined by the MitoTracker Green FM (M7514, Thermo Fisher Scientific), and the mitochondrial membrane potential was determined by the fluorescence intensity of TMRM (T668, Thermo Fisher Scientific) following the manual. Cells (1 × 10^6) were collected and resuspended in 500 μl of cell culture medium with 1.5 μl of 100 μM MitoTracker Green or TMRM. Cells were then incubated in 37°C incubator for 30 min. Then, cells were washed once with 1× PBS and resuspended in PBS buffer with 2% FBS for fluorescence-activated cell sorting (FACS). For cell viability analysis, cells were washed and resuspended with PBS buffer with 2% FBS. Then, cells were incubated with 1 μM 7-AAD (A1310, Invitrogen) for 5 min before analyzing. For CFSE (C345544, Invitrogen) cell proliferation staining, 10 million primary B cells were resuspended in PBS with 0.1% bovine serum albumin, and then the cells were mixed with the same volume of 1 μM CFSE for 10 min at 37°C. Cells were then neutralized by prechilled 10% FBS RPMI 1640 medium for 5 min. After washing the cell with culture medium, cells were resuspended and infected with virus. One hour after infection, cells were treated with 100 μM AOA, and CFSE were analyzed at 5 DPI. For cell cycle analysis, cells were fixed with ice-cold 70% ethanol for at least 24 hours. At the day for analysis, cells were centrifuged at 3000 rpm for 10 min to remove ethanol and resuspended in PBS buffer. Cells were incubated in 1 ml of staining buffer [propidium iodide (5 μg/ml), ribonuclease A (40 μg/ml), and 0.1% Triton X-100 in PBS) at room temperature for 30 min in dark. For MitoSOX analysis, 1 million cells were collected and washed with Hank’s balanced salt solution (Gibco) buffer and then incubated with 1 μM MitoSOX Green ([230]M36006, Thermo Fisher Scientific) for 30 min. Cells were gently washed with warm Hank’s balanced salt solution buffer. Flow cytometry was performed on a BD FACSCalibur instrument. Data were analyzed with FlowJo V10. Quantitative real-time PCR Extraction of total RNA from cells was conducted by the RNeasy Mini Kit (QIAGEN). Ribonuclease-free deoxyribonuclease set (QIAGEN) was used to remove genomic DNA, and SYBR Green RNA-to-CT 1-step kit (Applied Biosystems) was applied to assemble the qRT-PCR reaction. Primer sequences are listed in table S3. Samples were run in technical triplicates. The relative expression was calculated using the 2^−ΔΔCt method, and the data were normalized to internal control β-actin mRNA levels. RNA-seq and gene ontology analysis RNA-seq data of newly EBV-infected B cells were obtained from [231]GSE125974 ([232]28). The raw read count table from the RNA-seq analysis was obtained from Wang et al. ([233]28). DESeq2 was used to evaluate differential expression ([234]85) between 0 and 2 DPI samples. DESeq2 applies a negative binomial distribution to account for overdispersion in transcriptome datasets and uses a conservative heuristic-based analysis. Each differential expression analysis involved pairwise comparisons between the experimental and control groups. Differentially expressed genes were identified using a P < 0.05 and an absolute fold change of >2 as cutoffs. Differentially expressed genes were further filtered using a curated gene list focused on metabolism ([235]58). The differentially expressed metabolic genes were subjected to Enrichr analysis, which was used to perform gene set enrichment analysis on the selected gene subset. The algorithm used to calculate combined scores has been described previously ([236]86). DNA extraction and qPCR The total DNA from the intracellular was extracted by the Blood & Cell Culture DNA Mini Kit (#13362, QIAGEN). Extracted DNA was diluted to 10 ng/μl and quantified by qPCR for the EBV BALF5 in table S3. For quantification of intracellular EBV genome copy number, standard curves of BALF5 were set as serial dilution of a pHAGE-BALF5 miniprep DNA at 25 ng/μl. Viral DNA copy number was calculated according to the inputting sample C[t] and the standard curve, as described previously ([237]87). For quantification of mitochondrial DNA, the total DNA samples were used for qPCR analysis of mitochondrially encoded NADH:Ubiquinone oxidoreductase core subunit 1 (MT-ND1). The relative expression was calculated using the 2^−ΔΔCt method, and the data were normalized to β-actin. The fold change was calculated by normalizing data points to uninfected control (0 DPI) DNA levels. Chromatin immunoprecipitation Cells (2 × 10^7) were fixed with 10 ml of 1% formaldehyde in RPMI 1640 + 10% FBS for 10 min at room temperature. The cross-linking reaction was then quenched by adding 1.425 ml of 1 M glycine for 10 min at room temperature. Cells were washed three times with ice-cold PBS. Then, cells were incubated with 2 ml of lysis buffer [50 mM tris (pH 8.1), 10 mM EDTA, 1% SDS, and protease inhibitor cocktail] for 20 min on ice. Chromatin was fragmented with Bioruptor (Diagenode, USA) with 30-s on/30-s off condition for 30 cycles. Fragmented chromatin was then diluted with dilution buffer [16.7 mM tris (pH 8.1), 1.2 mM EDTA, 167 mM NaCl, 1.1% Triton X-100, 0.01% SDS, and protease inhibitor cocktail] and incubated with 5 μg of the ChIP-grade antibody of interest (see table S3) at 4°C overnight. Immunocomplexes were precipitated by addition of 100 μl of prewashed protein A or G magnetic beads for 1 hour at 4°C. Beads were isolated using a magnet and washed two times with 10 ml of low salt buffer [20 mM tris (pH 8.1), 2 mM EDTA, 150 mM NaCl, 1% Triton X-100, and 0.1% SDS], two times with 10 ml of high salt buffer [20 mM tris (pH 8.1), 2 mM EDTA, 500 mM NaCl, 1% Triton X-100, and 0.1% SDS], once with 10 ml of lithium chloride buffer [10 mM tris (pH 8.1), 1 mM EDTA, 0.25 M LiCl, 1% NP-40, and 1% deoxycholic acid], and once with 10 ml of Tris-EDTA buffer. DNA was eluted using freshly prepared elution buffer (1%SDS and 100 mM NaHCO[3] in H[2]O). A total of 250 μl (per 100 μl beads) of elution buffer was added and incubated at room temperature for 10 min. The eluted DNA was then reverse cross-linked by treating with 20 U of protease K at 65°C overnight. After reverse cross-linking, DNA was purified by the QIAquick PCR purification kit (QIAGEN) according to the protocol. ChIP assay–purified DNA was quantitated by qPCR, and values were normalized to the percentage of input DNA. Primers for qPCR are listed in table S3. Western blot analysis Immunoblot analysis was performed according to the previous instructions ([238]14). Cell lysates were prepared by incubating cells in 1× Laemmli buffer at 95°C for 5 min. For the target protein anchoring to the membrane such as SLC1A3 and the ETC, cell lysates were incubated at 37°C for 1 hour. Lysate samples were separated by SDS-PAGE, transferred onto the nitrocellulose membranes, blocked with 5% milk in Tris-buffered saline with Tween 20 buffer for 1 hour, and then probed with relevant primary antibodies at 4°C overnight. The next day, the membranes were incubated with secondary antibody for 1 hour. Blots were then developed by incubation with enhanced chemiluminescence (Millipore), and images were captured by LI-COR Fx system. Bands intensities were measured where indicated by Image Studio Lite version 5.2. All antibodies used in this study were listed in table S3. Cell viability analysis Cell viability was determined using the Countess 3 Automatic Cell Counter with trypan blue staining (15250061, Thermo Fisher Scientific) to distinguish live from dead cells. Trypan blue selectively stains nonviable cells, allowing for accurate calculation of live cell counts. For growth curve analysis, cells were initially seeded at a density of 2 × 10^5 cells/ml in 12-well tissue culture plates to ensure optimal growth conditions. Cell counts were taken at each time point to monitor growth over the experimental period. For CRISPR-Cas9 KO cells, seeding was performed on day 5 after puromycin selection, after KO efficiency was confirmed by immunoblotting of the targeted gene. This ensured that any observed changes in growth rate or viability could be attributed to the KO. To prevent overconfluency, the cells were regularly split, and the absolute live cell numbers were adjusted on the basis of the splitting (dilution) factors. This adjustment ensured accurate tracking of cell proliferation and viability without the confounding effects of nutrient depletion or cell overgrowth. The growth curve analysis was performed by calculating the total live cell count at each time point, taking into account the dilution factor from each passage. Statistical analysis Unless otherwise indicated, all bar graphs and line graphs represent the arithmetic mean of three independent experiments (n = 3), with error bars denoting SDs. Data were analyzed using two-tailed unpaired Student’s t test or analysis of variance (ANOVA) with the appropriate posttest using GraphPad Prism7 software. Metabolic pathway analysis was performed using MetaboAnalyst 3.0. Graphics Figures were drawn with GraphPad, BioRender, Microsoft PowerPoint, and ggplot2 in R. Acknowledgments