Patients infected with human immunodeficiency virus (HIV) are more predisposed to developing cancer, including glioblastoma (GBM) [1
] and ten percent of patients with acquired immune deficiency syndrome (AIDS) have brain tumors. Although the majority of these tumors are central nervous system lymphomas, glioma tumors arise as well [3
]. Multiple medical reports in HIV/AIDS patients indicate that GBM occurs at a higher frequency (5.4- to 45-fold increase) [4
] and at a younger age [8
] in individuals at various stages of HIV infection than in the general population. On average, GBM tumors appear approximately 3 years after initial HIV infection [2
]. Additionally, the median survival rate in GBM-HIV-infected patients is shorter (an average of 8 months) than in GBM-non-infected patients (an average of 14 months) despite receiving the same treatment [2
]. CD4+ cell count at initial diagnosis of GBM is not correlated with survival suggesting that increased aggressive tumor behavior is not a direct outcome of immune deficiency [8
The nature of the GBM–HIV relationship is not well understood. The stimulatory effect of HIV infection on glioma tumor development has been associated with reduced immune surveillance [11
]. However, immune incompetence has not been clearly shown to underlay glioma tumor development and progression [7
]. While the incidence of some malignancies among HIV-infected individuals has declined with Highly Active Antiretroviral Therapy (HAART), it remains elevated compared with uninfected population suggesting that overall immune deficiency is not the only cause [15
]. Moreover, HIV is not found in glioma tissues of patients diagnosed with GBM [2
]. It has been shown that some human GBM cell lines have the ability to suppress HIV infection by secreting molecules that inhibit HIV attachment to target cells, while other GBM cell lines do not [18
]. The predominant HIV target cells in the brain are microglia and macrophages, while other cells such as astrocytes, oligodendrocytes, neurons and microvascular cells tend to be mostly resistant [19
It has been shown that HIV infection promotes the development of glial tumors through a set of factors that includes the activation of oncogenes, the impairment of immune defenses and the production of growth factors and cytokines capable of inducing astrocytosis [21
]. Additionally, tumor cells can be directly exposed to HIV proteins such as gp120, which can be secreted by infiltrated and infected microglia and astrocytes [25
HIV is believed to enter and infect the central nervous system through the interaction between the envelope protein gp120 and the CCR5 or CXCR4 receptors expressed on macrophages but some neurons and astrocytes are thought to express these receptors as well, which can result in their subsequent infection [3
]. GBM cells can also express CXCR4 and CCR5 and activation of these receptors is known to promote cell survival and cell cycle progression [27
]. We hypothesize that despite some innate resistance to HIV infection, glioma cells can interact with the HIV envelope protein gp120 and this interaction promotes cell proliferation and tumor growth.
In this study, we investigated the effect of the HIV envelope protein gp120 on glioma cell growth and survival. We showed that continuous treatment of the U87 and A172 glioma cells and the primary human glioma cell line 965 with gp120 for a period of 10 days resulted in increased proliferation, migration and survival. Through a combination of western blot and metabolomics analysis we also detected the activation of glycolysis as well as protein and fatty acid synthesis metabolic pathways in these cells. Finally, using a HIVgp120tg/Gl261 mouse glioma implantation model, we demonstrated that animals expressing gp120 in their brain develop bigger tumors and have shorter median survival than their wild type littermates (WT).
Our study demonstrated that the HIV glycoprotein gp120 promotes proliferation, migration, survival and stimulates glycolysis in glioma cell lines. Increased glycolysis, also known as the Warburg effect, is characteristic of malignancy [41
]. Upregulated glycolysis promotes unconstrained proliferation and invasion of tumor cells, providing the required glycolytic intermediary precursors for DNA, protein and lipid synthesis [43
]. In Figure 9
we summarized the metabolic pathways that are altered in glioma cells in response to gp120. We observed an upregulation of the key glycolytic enzymes HXK, GAPDH and ENO2 in glioma cells treated with gp120. Despite the fact that we did not observe an upregulation of PKM2 protein levels—the enzyme that catalyzes the final step of glycolysis, converting phosphoenolpyruvate (PEP) to pyruvate—we found a significant increase in PKM2 activity and pyruvate synthesis, as well as an increase in the glycolytic index. These findings indicate that gp120 stimulates the glycolytic pathway in glioma cells.
HIV infects immune cells by binding to CD4, CCR5 and CXCR4 through its envelope protein gp120 [44
]. Current evidence suggests that most, if not all, GBMs express CXCR4 and CCR5 [27
] and these receptors are related to the survival, invasiveness, proliferation and resistance to the radio- and chemotherapy of glioma tumors [27
]. It has been shown that activation of CCR5 and CXCR4 promotes a global shift towards anabolic metabolism and increased cell proliferation: increased glucose uptake, ATP production and enhanced glycolysis, associated with extracellular acidification [51
]. Numerous viruses have been shown to cause significant alterations in the metabolism of the host cell, including HIV, hepatitis C, influenza, herpes simplex and human cytomegalovirus [54
]. The production of the enveloped viruses requires additional fatty acid and nucleotide synthesis to progeny virions and these biosynthetic pathways are supported by the upregulation of glycolysis and the TCA cycle in a host cell [55
]. It was noted that these changes in metabolic activity of infected cell are similar to oncogenic transformation [58
], suggesting that the biosynthetic needs of a virally infected cell are similar to those of proliferating cells. Based on these reports, we assume that the gp120-induced activation of glycolysis and cell proliferation might result from activation of cell surface signaling molecules, such as CXCR4 or CCR5 and further-downstream signaling activation in glioma cells. It was reported that activation of CXCR4 and CCR5 leads to the activation of AKT [60
] and promotes tumor proliferation [63
]. Activated AKT phosphorylates multiple downstream targets including Glycogen Synthase Kinases (GSK3) and Mammalian Target of Rapamycin (mTOR) involved in glycogen metabolism, glucose homeostasis and protein synthesis regulation [64
]. GSK3 is over activated in GBMs and the level of GSK3 phosphorylation is associated with increased tumor growth [66
]. It is plausible that gp120 binding to CXCR4 and CCR5 activates AKT/GSK3 signaling resulting in upregulation of glycolysis in glioma cells. Reported downregulation of Phosphatase and Tensin Homolog (PTEN) in investigated glioma cell lines [68
], upregulation of AKT signaling and high level of activity of GSK3, might provide the mechanistic ground for gp120-driven proliferation and metabolic switch specifically in gliomas but not in other investigated cancer cell lines. However, the detailed evaluation of underlying signaling mechanism is a subject for future studies.
PKM2, which we found to be activated in response to gp120 in glioma cells, is another critical player in the metabolic reprogramming of cancer cells. PKM2 can be translocated into the nucleus and induce cellular proliferation [71
] and has been shown to be activated in cancerous tissues [72
]. PKM2 is found as dimeric and tetrameric forms in which the tetramer has a high affinity and the dimer has a low affinity for phosphoenolpyruvate (PEP) [74
]. Cancer cells preferentially express the less active form of PKM2 [75
], leading to accumulation of glycolytic intermediates in upstream pathways and facilitating the formation of cell-building components. The dimer/tetramer ratio acts as a sensor that regulates metabolic synthesis and energy production from mitochondria [76
]. Since our study revealed an increase in PKM2 activity in gp120-treated glioma cells without upregulation of the PKM2 protein, we propose that gp120 can affect the PKM2 dimer/tetramer ratio for the coordination of glycolysis with the cell cycle. This conclusion is supported by literature reports indicating that both fructose 1,6-bisphosphate and PEP, the product metabolites of HXK and ENO2, upregulated by gp120 in our study, promote reassociation of the dimer into a tetramer [77
] and facilitate entry into the Krebs cycle and a decrease in lactate production. This is consistent with our results indicating activation of pyruvate kinase together with an insignificant increase in lactate.
Additionally, it has been shown that serine can bind to and activate PKM2 [79
]. Our study established that there is an increase in serine in gp120-treated glioma cells (Table 1
). Taken together with the increased in pyruvate kinase activity in these cells, we propose the existence of a serine activation loop in the glycolysis pathway (Figure 9
). The biosynthesis of serine starts with the oxidation of 3-phosphoglycerate, the metabolite produced by GAPDH. Upregulation of GAPDH, which we observed in our study, provides 3-phosphoglycerate as a primary source for the serine biosynthesis pathway. In addition to directing activation of PKM2, serine is crucial for cancer growth and oncogenic transformation due to its participation in the biosynthesis of purines, pyrimidines, sphingolipids and several amino acids, including glycine and cysteine [80
] and has been found to be upregulated in cancers [83
Our results revealed a reduction in the amino acid levels in cells treated with gp120. Together with the reduction in urea (Table 1
) and the upregulation of molecules involved in protein synthesis, previously identified by our group [32
], such as the RPS2, RPS13, RPS15 ribosomal proteins and number of initiation and elongation factors as eIF4A1, eIF4G1, PABPC1 and eEFG1, these findings indicate the predominance of anabolic processes over catabolic processes and active protein synthesis. It has been shown that glutamine uptake is essential for growth in a number of cancers [84
]. Glutamine, through glutamate, is a source of α-ketoglutarate in the TCA cycle, glutathione in redox homeostasis and citrate by reductive carboxylation to form lipids and glucosamines [90
]. Tumor cells tend to have a large pool of glutamate and this pool is maintained by the cell’s ability to convert glutamine into glutamate through glutaminases [91
]. The upregulated glutamate observed in gp120-treated glioma cells indicates the increased use of TCA cycle metabolites for synthesis of lipids and amino acids and replenishment of the TCA cycle through the glutamate–ɑ-ketoglutarate pathway. An increase in glutamate in gp120-treated cells can result from the increased uptake of glutamine and further conversion into glutamate, such as through the synthesis from 5-oxoproline, which is also upregulated in gp120-treated cells. Recent studies have shown that 5-oxoproline participates in the regulation of Na+
-dependent transport of glutamate [92
] and may contribute to the increased glutamate uptake by gp120-treated glioma cells.
We detected a 24-fold increase in proline in gp120-treated cells. Proline can be synthesized from glutamine [94
] and the metabolism of proline serves as a source of energy during stress, provides signaling reactive oxygen species for epigenetic reprogramming and regulates redox homeostasis [86
]. The critical role of proline biosynthesis in maintaining pyridine nucleotide levels by connecting the proline cycle to glycolysis and to the oxidative arm of the pentose phosphate pathway have also been shown [95
]. Proline biosynthesis activity has been associated with tumor cell growth, resistance to oxidative stress and energy production [96
]. Based on this we can assume, that the increase in proline synthesis in gp120-treated glioma cells might be associated to increased proliferation and survival through reprogramming of the glutamine and pyridine pathways.
Abnormal cellular lipid metabolism also plays an important role in cancer. Fatty acid biosynthesis is restricted to a subset of tissues, including liver, adipose and lactating breast tissues. However, reactivation of lipid biosynthesis has been reported in cancers [98
] and is associated with cancer growth and invasiveness [100
]. Fatty acids are the major building blocks for the synthesis of phosphoglycerides and are structural components of biological membranes, contributing to cancer cell proliferation. In our study, we detected the upregulation of stearic, oleic, myristic and palmitoleic fatty acids in glioma cells in response to gp120 which together with our other findings supports the increased proliferation and survival in gp120 treated glioma cells.
Another important finding from this study is related to the substantial 15-fold increase in tryptophan in gp120-treated cells. Many cancers drive tryptophan consumption [102
] and it has been shown that the primary product of tryptophan metabolism, kynurenine, is an endogenous ligand for the aryl hydrocarbon receptor, which mediates invasive tumor growth and the evasion of immunity [102
]. The autocrine binding of kynurenine to the aryl hydrocarbon receptor in cancer cells causes the transcriptional activation of genes related to tumor invasiveness [105
]. In patients with cancer, the upregulation of indoleamine 2,3-dioxygenase, an enzyme that generate kynurenine from tryptophan, is associated with a poor prognosis [106
]. The increased consumption of tryptophan observed in gp120-treated glioma cells may be related to the increased migration of these cells.
4. Materials and Methods
4.1. Cell Culture
U87 and A172 human glioma cell lines were obtained from the American Type Culture Collection (Manassas, VA, USA). The GL261 glioma cell line derived from C57BL/6 mice was obtained from the NCI (Frederick, MD, USA). The 965 primary human glioma cell line was obtained from a resected GBM tumor mass in the laboratory of Dr. Quinones-Hinojosa, Mayo Clinic. Clinical data for 965 primary cell lines have been described and analyzed in detail previously [68
]. Cells were cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal calf serum and 50 U/mL penicillin/50 µG/mL streptomycin and maintained in a humidified atmosphere of 5% CO2/95% air at 37 °C. Cultures used in this study underwent less than sixteen passages.
All procedures involving rodents were conducted in accordance with the National Institutes of Health regulations concerning the use and care of experimental animals. All procedures involving animals were approved by Universidad Central del Caribe Institutional Animal Care and Use Committee (protocol #036-214-14-01-PHA from 4 August 2016). All efforts were made to minimize suffering.
HIVgp120tg mice were kindly provided by Dr. Marcus Kaul (Sanford Burnham Prebys Medical Discovery Institute, San Francisco, CA, USA) [35
]. C57Bl/6 mice were purchased from the Jackson Laboratory. C57Bl/6 and HIV gp120tg mice were crossbred and animals heterozygous for HIV gp120 and wild type littermates were used as experimental and control groups correspondingly. Genotyping was performed according to the previously published protocols [36
4.3. Intracranial Implantation of Glioma Cells
All surgeries were performed under isoflurane anesthesia and all efforts were made to minimize suffering. GL261 glioma cells were implanted into the right cerebral hemisphere of 12–16 week old C57BL/6 mice. Implantation was performed according to the protocol that we described earlier [109
]. Briefly, mice were anesthetized with isoflurane and a midline incision was made on the scalp. At stereotaxic coordinates of bregma, 2 mm lateral, 1 mm caudal and 3 mm ventral a small burr hole (0.5 mm diameter) was drilled on the skull. 1 μL of cell suspension (2 × 104 cells/μL in PBS) was delivered at a depth of 3 mm over 2 min. Sixteen days following injection, animals were anesthetized with pentobarbital (50 mg/kg) and transcardially perfused with PBS followed by 4% paraformaldehyde (PFA). Brains were removed and postfixed in 4% PFA/PBS for 24 h at 4 °C, followed by 0.15 M, 0.5 M and 0.8 M sucrose at 4 °C until fully dehydrated. Brains were then frozen-embedded in Cryo-M-Bed embedding compound (Bright Instrument, Huntingdon, England) and cut using a Vibratome UltraPro 5000 cryostat (American Instrument, Haverhill, MA, USA).
4.4. Tumor Size Evaluation
15 μm coronal frozen sections encompassing the entire tumor were stained with Hematoxylin & Eosin. Tumor size was calculated as sum of tumor area × section thickness for each section containing a tumor.
4.5. Survival Analysis
GL261 glioma cells were implanted into HIVgp120tg and WT littermate mice. Animals were inspected daily and body weight loss of 15%, decreased activity/responsiveness, abnormal posture or any neurological disorders signs are to be a subject for euthanasia. Time between tumor bearing and animal death was recorded.
4.6. In Vitro Viability Assay
Glioma cells were plated in petri dishes at 200,000 cells per dish and incubated for 10 days with and without gp120 (100 ng/mL). The cells were then harvested, stained with trypan blue and the total number of live and dead cells determined by cell counting.
4.7. Migration Assay
Migration assays were performed using Fluoroblok inserts (8-µM pore size, VWR Scientific). Serum-starved cells (30,000) were placed on the insert membrane and the assays were performed following the addition of medium containing 5% serum to the lower compartment. After 5 h, the cells were fixed with methanol and stained with propidium iodide. The number of cells that had migrated to the lower compartment was determined by counting the number of fluorescent cells.
4.8. Cell Cycle Assays
Cells were harvested, fixed in 70% ethanol, re-suspended in PBS containing 1 µg/ml 7-AAD (Bio-Rad Laboratories, Hercules, CA, USA) and 0.2 mg/mL RNase A (Sigma-Aldrich, St. Louis, MO, USA), incubated for 30 min at 37 °C in the dark and analyzed with a FACSCanto II flow cytometer (Becton Dickinson, San Jose, CA, USA). The percentage of cells in G0/G1, S and G2/M phases was determined from the DNA content using FlowJo data analysis software v.10 (Ashland, OR, USA).
4.9. Western Blot Analysis
Clarified cell lysates separated on 10% SDS-PAGE gels were transferred to PVDF membranes and probed with mouse anti-ENO2 antibody (Santa Cruz Biotechnology, Dallas, TX, USA; #SC-21738) and rabbit polyclonal anti-PKM2, anti-GAPDH and anti-HXK antibodies (Cell Signaling, Danvers, MA, USA; #4053, #5174 and #2024, respectively), diluted 1:1000, followed by the secondary antibodies (Sigma-Aldrich, Saint Louis, MO, USA; #A9169). Detection was performed with enhanced chemiluminescence methodology (SuperSignal® West Dura Extended Duration Substrate; Pierce, Rockford, IL, USA) and the intensity of the signal was measured using a gel documentation system (Versa Doc Model 1000, Bio Rad). The intensity of the chemiluminescent signal was corrected for minor changes in protein content after densitometry analysis of the India ink-stained membrane.
4.10. Pyruvate Kinase, Hexokinase and Glyceraldehyde 3-Phosphate Dehydrogenase Activity Assays
Intracellular pyruvate, HXK and GAPDH activity was measured in cell lysates with the Pyruvate Kinase Activity Colorimetric/Fluorometric Assay kit (BioVision, Milpitas, CA, #K709-100), HXK Activity Fluorometric Assay Kit (Abcam, Cambridge, MA, USA, #ab211103) and GAPDH Activity Colorimetric Assay Kit (Abcam, Cambridge, MA, USA, #ab204732) according to the manufacturer’s protocol. In each experiment, 5,000 cells were used and their OD at 570 nm (for Pyruvate Kinase Assays), 450 nm (for GAPDH assays), or fluorescence at Ex/Em = 535/587 nm measured with a Perkin Elmer Wallac 1420 Victor2 Microplate Reader. Standard curves were used to determine the concentration of pyruvate, HXK, or GAPDH in the sample from the numeric colorimetric/fluorometric data. Enzyme’s activity was calculated as the amount of product produced in the sample in 10 min.
4.11. Glucose Uptake Assays
Glucose uptake was measured with the cell-based Glucose Uptake Assay Kit (Abcam, Cambridge, MA, USA, #ab204702) according to the manufacturer’s protocol. 50,000 cells were seeded on coverslips one day before starting the assay. Cells were incubated with glucose uptake mix containing fluorescent GluTracker reagent for 30 min and immediately visualized using an Olympus Fluoview FV1000 confocal microscope (Olympus, Japan) with 40× oil immersion objective and FITC excitation—emission filter set (absorption maximum at 494 nm and emission maximum of 521 nm). The fluorescent images were processed using ImageJ software.
4.12. Glycolysis and Extracellular Oxygen Consumption Assays
Comparative measurements were taken with Glycolysis Assay (Abcam, Cambridge, MA, USA, #ab197244) and Extracellular Oxygen Consumption Assay (Abcam, Cambridge, MA, USA, #ab197243) according to the manufacturer’s protocols. Cells were seeded in a 96-well plate at a density of 5000 cells/well. 24 h later the cell culturing medium was replaced with Glycolysis Assay Reagent or Extracellular Oxygen Consumption Reagent and the assay’s signals were measured simultaneously with a Perkin Elmer Wallac 1420 Victor2 Microplate Reader using Ex/Em = 380/615 nm.
4.13. ATP and Lactate Assays
ATP and L-Lactate were measured in cell lysates with the ATP Assay Kit (Abcam, Cambridge, MA, USA, #83355) and L-lactate Assay Kit (Abcam, Cambridge, MA, USA, #65331) according to the manufacturer’s protocol. Cultured cells or tissue lysates were used and their OD at 570 nm measured with a Perkin Elmer Wallac 1420 Victor2 Microplate Reader. A standard curve was used to determine the concentration of ATP and lactate in the sample from the numeric colorimetric data.
4.14. Metabolomics Analysis
Metabolites were extracted using an optimized protocol [110
]. Cultured cells were scraped in MeOH/H2O (85:15), sonicated and centrifuged at 13,000× g. The supernatant was dried in a SpeedVac (Savant AS160, Farmingdale, NY, USA), followed by methoxyamination in a 20-mg/mL methoxyamine hydrochloride in pyridine (Sigma-Aldrich, St. Louis, MO, USA) and trimethylsilylation in N-methyl-N-trimethylsilyl-trifluoroacetamide (MSTFA + 1% trimethylchlorosilane (TMCS), ThermoFisher Scientific, Waltham, MA, USA). The supernatants were dissolved 3/50 in hexane and applied to a GC-2010 gas chromatograph (Shimadzu Scientific, Columbia, MD, USA) with an AOC-20i auto-injector in split mode (split ratio, 15). The analytes were fractionated on a fused-silica capillary RXI-5MS column (0.25 mm inner diameter, 0.25 μm D.F.; 30 m; Restek Bellefonte, PA, USA). The oven temperature was set to increase from 100 °C to 290 °C at 8 °C/min. Mass spectra were obtained on a Shimadzu GCMS-QP2010 mass spectrometer (EI, 70 eV, ion source temperature, 200 °C) in scan mode between 35 and 700 amu. The data obtained were processed using GCMS Solution Post-run Analysis software (Shimadzu Corp) for metabolite identification by comparison with the NIST08 spectral mass library (National Institute of Standards and Technology, Gaithersburg, MD, USA) using the NIST MS Search Program 2.0, mass spectral databases and AMDIS Version 2.71 deconvolution software (Automated Mass Spectral Deconvolution and Identification System, www.amdis.net
]. Peaks were integrated (maximum peak number, 200; width time, 2 s; smoothing method, standard) and manually checked. The peak intensities were quantified as the fold-change relative to control and the statistical significances were analyzed using Student’s t
-values < 0.05 were considered significant. Metabolomics analysis was conducted using the IPA core analysis of metabolites (QIAGEN, Redwood City, CA, USA). Data sets containing metabolite identifiers (KEGG IDs) and metabolic enzyme protein identifiers (UNIPROT IDs), including their corresponding fold-change values relative to control, were mapped together with their corresponding objects in the Ingenuity Knowledge. Fisher’s exact test was used to calculate the probability that each ID set was enriched. Only the biological functions/pathways (Bonferroni’s corrected p
value < 0.05) were considered significantly enriched.
4.15. Protein Synthesis Assay
EZClick Global Protein Synthesis Assay Kit (BioVision Inc., Milpitas, CA, USA, #K715-100) was used according to the manufacturer’s protocol. Briefly, 500,000 cells were seeded in 60mm petri dishes 24 h prior to the assay followed by addition of the Protein Label for 1 h. Cells were harvested, fixed and stained with Fluorescent Azide followed by Flow Cytometry analysis, using a BD-FACSCanto II flow cytometer (Becton Dickinson, San Jose, CA, USA). Flow cytometry data was analyzed using FlowJo data analysis software v.10 (Ashland, OR, USA).
4.16. Chemicals and Reagents
Sodium monofluorophosphate was obtained from Santa Cruz Biotechnology, Inc., Dallas, TX, USA, cat. #SC-264320
4.17. Statistical Analysis
Results are expressed as mean ± standard deviation (SD). The statistical probability was calculated using GraphPad software. Unpaired t-tests or one-way ANOVA tests followed by Tukey’s post-hoc test were used to determine significance between groups. p-values < 0.05 were considered as significant.