Abstract

   Endothelial cell (EC) dysfunction is key in initiating and progressing
   pulmonary hypertension (PH). EC dysfunction in PH leads to
   hyperproliferation and vascular remodeling of the pulmonary blood
   vessels. Increased glutaminolysis and altered cellular metabolism are
   pivotal in hyperproliferative cancer cells. However, whether a similar
   enhancement in glutamine metabolism is involved in the EC
   hyperproliferation and if this contributes to vascular remodeling
   during PH development is unresolved and was the focus of our study.
   Metabolic flux analysis showed elevated glutaminolysis and enhanced
   metabolic flux through the reductive tricarboxylic acid (TCA) cycle in
   pulmonary arterial ECs isolated from an ovine experimental model of PH
   (PH-PAECs). PH-PAECs also exhibited increased c-Myc protein levels, a
   master regulator of glutaminolysis. Therefore, we assessed the effect
   of increased c-Myc expression on metabolic reprogramming,
   glutaminolysis, and proliferation in control PAECs. Results from a
   comprehensive snapshot metabolomics investigation and metabolic flux
   analysis confirmed the reprogramming of mitochondrial metabolism,
   enhanced glutamine metabolism, and increased glycolysis in c-Myc
   overexpressing PAECs. Additionally, c-Myc overexpression impacted the
   ATP production rate, disrupted mitochondrial respiration, increased
   reactive oxygen species production, induced cell proliferation, and
   suppressed apoptosis. Functionally, these metabolic changes suppressed
   nitric oxide (NO) production. We also demonstrate that a small-molecule
   c-Myc inhibitor, 10058-F4, attenuates glutaminolysis, suppresses the
   reverse TCA cycle and glycolysis, and reverses the hyperproliferative
   phenotype, thereby restoring NO levels in PH-PAECs. We also demonstrate
   that directly targeting HIF-1α reverses the hyper-proliferative,
   anti-apoptotic phenotype in PH-PAECs. Thus, targeting c-Myc signaling
   and suppressing glutaminolysis or glycolysis could be a novel therapy
   for PH.

   Keywords: Pulmonary hypertension, Endothelial cells, Glutaminolysis,
   Metabolomics, Glycolysis, Proliferation

Graphical abstract

   Image 1
   [43]Open in a new tab

Highlights

     * •
       c-Myc levels are increased in a lamb model of PH.
     * •
       c-Myc disrupts mitochondrial function and increases ROS.
     * •
       Glycolysis and glutaminolysis are enhanced.
     * •
       This metabolic reprogramming suppresses NO generation.
     * •
       Inhibiting c-myc activity reverses the metabolic reprogramming and
       restores NO levels.

1. Introduction

   Pulmonary hypertension (PH) is a vascular lung disorder characterized
   by the occlusive remodeling of the pulmonary vessels, causing a
   progressive increase in vascular resistance [[44]1]. Injury to the
   endothelium is one of the primary initiators of PH development [[45]2].
   Prior work has shown that endothelial cells (ECs) are dysfunctional
   and/or damaged in PH patients [[46][3], [47][4], [48][5], [49][6]]. The
   emergence of hyperproliferative and antiapoptotic EC phenotypes is
   intimately involved in the vascular remodeling associated with PH
   [[50][7], [51][8], [52][9], [53][10]]. However, the underlying
   mechanisms that drive this process are poorly understood. It is
   becoming clear that metabolic programming, similar to that observed in
   cancer, is a critical component of PH development [[54]11]. The MYC
   oncogene encodes a transcription factor c-Myc that is well known for
   its role in regulating proliferation, growth and differentiation of
   many cell types, including ECs [[55]12,[56]13]. The major downstream
   effectors of c-Myc include those involved in metabolism, mitochondrial
   biogenesis, ribosome biogenesis, mRNA translation, and cell cycle
   regulation [[57]14]. c-Myc overexpression results in metabolic
   transformation leading to increased glucose uptake and lactate
   production, the Warburg effect [[58]15]. The Warburg effect is now
   regarded as a critical component of PH development [[59][16], [60][17],
   [61][18], [62][19]].

   c-Myc-driven proliferation is associated with a cellular addiction to
   glutamine for survival and growth [[63]20]. Dysregulated c-Myc
   expression has been shown to play a pivotal role in the development of
   cardiovascular disorders, including PH [[64]21,[65]22]. Emerging
   evidence has also shown c-Myc involvement in vascular development,
   suggesting an important role in EC growth and function [[66][23],
   [67][24], [68][25], [69][26]]. Further, the absence of c-Myc in human
   EC significantly affects cell growth and proliferation by activating
   senescence and compromising vascular homeostasis [[70]27]. Increased
   glutamine metabolism can promote vascular remodeling in animal models
   of PH [[71]28]. Many key regulators of glutamine metabolism have been
   identified in proliferating cells [[72][29], [73][30], [74][31],
   [75][32]]. However, the transcriptional program that promotes
   glutaminolysis and enhances glutamine addiction in dysfunctional EC
   during PH development remains elusive.

   Our results identified an increased glutamine metabolism in the PAECs
   isolated from an ovine experimental model of PH with increased
   pulmonary blood flow (PH-PAECs). The increase in glutamine metabolism
   is correlated with increased c-Myc. Results from our snapshot
   metabolomics and metabolic flux analysis demonstrated that
   c-Myc-dependent reprogramming of mitochondrial metabolism as associated
   with increased glutamine catabolism and was required to sustain the
   hyperproliferative antiapoptotic EC phenotype. Further, we show that
   increasing c-Myc expression in control PAECs was sufficient to mimic
   the EC PH phenotype with increases in glutaminolysis and glycolysis,
   suppressed mitochondrial bioenergetics, increased proliferation,
   decreased apoptosis, and reduced nitric oxide (NO) production.
   Critically, we could also demonstrate that the small-molecule inhibitor
   of c-Myc,10058-F4, reversed the metabolic reprogramming in PH-PAECs,
   attenuated the hyperproliferative antiapoptotic EC phenotype, and
   improved NO levels.

2. Materials and methods

2.1. Antibodies

   The antibodies against c-Myc, ASCT2, GLS1, GLS2, GLUD1/2, IDH2, ACO2,
   OGDH, β-actin, anti-mouse IgG and anti-rabbit IgG antibodies were
   purchased from Cell Signaling Technologies (Danvers, MA).

2.2. Cell culture and Western blot analysis

   As described in detail previously, primary cultures of pulmonary
   arterial endothelial cells were isolated and cultured from three
   independent Shunt and three age-matched Control lambs [[76][33],
   [77][34], [78][35], [79][36], [80][37]]. Control and PH-PAECs were used
   between passages 5 to 8 for all the experiments [[81]38]. PAECs were
   cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with
   10 % fetal calf serum and an antibiotic solution (500 IU penicillin,
   500 μg/ml streptomycin) at 37 °C and 5 % CO[2] in a humidified
   atmosphere. Adenoviruses Ad-c-Myc and Ad-GFP (control) were purchased
   from Vector Biolabs (Malvern, PA). Control PAECs were transduced with
   either adenoviruses Ad-c-Myc (MOI = 5) or Ad-GFP (MOI = 5), and the
   cells were allowed to grow for 48 h at 37 °C in an incubator with 5 %
   CO[2]. In other experiments, the PH-PAECs were treated with a c-Myc
   inhibitor (10058-F4, 20 μM, 16h) (Sigma-Aldrich, Burlington, MA, USA)
   or a HIF-1α inhibitor (CAY10585, 10 μM, 24h) (Fisher Scientific,
   Waltham, MA, USA). The Western blot method and analysis were performed
   following previously described protocol [[82]39].

2.3. Snapshot and metabolic flux analysis

   Metabolites were extracted, derivatized, and analyzed by GC-MS
   following a previously described protocol [[83]40]. Metabolic flux
   assay was performed using stable isotopic tracers. Briefly, control
   PAECs, c-Myc overexpressing PAECs, PH-PAECs, and PH-PAECs treated with
   c-Myc inhibitor (20 μM) were cultured in 6-well plates with DMEM with
   either ^13C[6]-glucose or ^13C[5]-glutamine (Cambridge Isotope
   Laboratories, Tewksbury, MA) and 5 % FBS for 6h. Cellular extracts were
   derivatized following our previously published protocols and injected
   into the Agilent 8860 GC system connected to an Agilent 5977B GC/MSD
   (Agilent Technologies, Inc., Santa Clara, USA)
   [[84]19,[85]39,[86]41,[87]42]. The mass isotopomer distributions (MIDs)
   were analyzed by integrating metabolite ion fragments using Agilent
   MassHunter Quantitative Analysis Software, and the IsoCor software was
   used to analyze the natural abundance of isotopes. Next, data
   normalization was done using MetaboAnalyst online tool (log
   transformation and Pareto data scaling). Multivariate statistical
   analysis was done including principal component analysis (PCA) and
   partial-least-squares discrimination analysis (PLS-DA). We used
   correlation analysis and the clustering result was generated as a
   heatmap. The most relevant metabolic pathways were identified based on
   enrichment analysis procedures. A summary of enriched metabolites was
   displayed using the TCA cycle.

2.4. Measurement of extracellular acidification rate, oxygen consumption rate
and real-time cellular ATP rates

   The Seahorse XFe24 Analyzer (Agilent Technologies, Santa Clara, CA,
   USA) was used for measurement of extracellular acidification, oxygen
   consumption and real-time cellular ATP rates following previously
   published protocols [[88]39].

2.5. Analysis of the mitochondrial membrane potential and mitochondrial
reactive oxygen species

   Analysis of the mitochondrial membrane potential and mitochondrial
   reactive oxygen species (mt-ROS) were performed as previously described
   [[89]39].

2.6. Cell Proliferation Assays

   Cell counts were performed under microscope using a hemocytometer after
   adding 20 μl of cell suspension between the hemocytometer and cover
   glass. BrdU Cell Proliferation Assay was performed following previously
   published protocol [[90]39].

2.7. Resistance to TNF-mediated apoptosis

   Measurements were accomplished using the terminal deoxynucleotidyl
   transferase (Tdt)‐mediated dUTP nick‐end labeling (TUNEL) assay
   (Click-iT™ Plus TUNEL Assay, Thermo Fisher Scientific). TNFα (10 ng/mL)
   was added for 20h to PAECs grown on coverslips in multi-well plates.
   Microscopy images were captured on the Keyence microscope using a
   590 nm excitation and 615 nm excitation wavelengths using an x20
   objective and the images were analyzed with ImageJ [[91]39].

2.8. Cellular nitric oxide measurement

   Nitric Oxide (NO) levels in cultured PAECs were determined using
   fluorescent imaging in conjunction with DAF-FM diacetate, a
   cell-permeable fluorescent dye, according to previously published
   protocol [[92]43].

2.9. Statistical analysis

   For the cell biology studies, statistical analysis was performed using
   GraphPad Prism version 4.01. The mean ± SEM were calculated for all
   samples and significance was determined by the unpaired t-test. A
   statistically significant test result p < 0.05 was accepted. For the
   metabolomics studies, data were analyzed with the MetaboAnalyst
   platform, as previously described [[93]19,[94]39,[95]44].

3. Results

   Pulmonary hypertensive pulmonary arterial endothelial cells (PH-PAECs)
   exhibit increased glutaminolysis. The potential role of upregulated
   glutamine metabolism during PH development is now being investigated
   [[96]28,[97]45,[98]46]. Therefore, in this study, we investigated if
   PH-PAECs exhibited alterations in glutamine metabolism to replenish the
   TCA cycle intermediates necessary to maintain the hyperproliferative EC
   phenotype [[99]47]. To accomplish this, PH-PAECs were cultured with
   media containing ^13C[5]-glutamine for 6h and GC-MS was used to analyze
   the extracted TCA cycle intermediates. Both Principal Component
   Analysis (PCA) ([100]Fig. 1A) and the supervised Partial Least-Squares
   Discriminant Analysis (PLS-DA) ([101]Fig. 1B) plots showed a clear
   cluster distinction between control and PH groups. Heatmap was used to
   depict the changes in 15 ^13C[5]-glutamine-derived metabolites altered
   in PH-PAECs. All the TCA cycle metabolites were significantly higher in
   PH-PAECs ([102]Fig. 1C). Analysis of the isotopologue distribution from
   the ^13C[5]-glutamine flux analysis indicated enhanced glutamine import
   which subsequently increased the refueling of carbons into the TCA
   cycle ([103]Fig. 1D). Analysis of the mean isotopic enrichment of
   molecules revealed significant increases in TCA cycle intermediates in
   PH-PAECs ([104]Supplemental Fig. 1A–M). Next, we investigated if
   glutamine was metabolized via forward (oxidative) and/or reverse
   (reductive) TCA cycle metabolism. During oxidative phosphorylation,
   α-KG is converted to succinate by oxoglutarate dehydrogenase (OGDH,
   [105]Fig. 1E). In contrast, in the reductive carboxylation reaction,
   isocitrate dehydrogenase 2 (IDH2) and aconitase 2 (ACO2) convert α-KG
   to citrate ([106]Fig. 1E). Analysis of citrate formed through oxidative
   phosphorylation (m+4) and reductive carboxylation (labeled as m+5 from
   labeled glutamine because of the incorporation of an unlabeled CO[2] in
   the carboxylation reaction) showed increased fractional amount of both
   citrate (m+4; [107]Fig. 1F) and citrate (m+5; [108]Fig. 1G) in
   PH-PAECs. Further, in PH-PAECs, oxidative glutaminolysis generated
   increased M+4-labeled TCA cycle intermediates while reductive
   carboxylation generated increased M+5 citrate followed by M+3 malate,
   fumarate, succinate and aspartate ([109]Fig. 1D). Thus, both forward
   and reverse TCA cycle fluxes are upregulated in PH-PAECs.

Fig. 1.

   [110]Fig. 1
   [111]Open in a new tab

   Glutamine metabolism is increased in pulmonary hypertensive pulmonary
   arterial endothelial cells. Score plots of Principal Component Analysis
   (PCA, A) and Partial Least Squares Discriminant Analysis (PLS-DA, B)
   models for the ^13C[5]-glutamine flux data obtained by GC-MS, showing
   the metabolic profile differences between control (blue) and PH-PAECs
   (brown) groups. Hierarchical clustering heat map of the 15 differential
   metabolites, with the degree of change marked with brown
   (up-regulation) and blue (down-regulation) (C). Schematic model of
   glutamine metabolism in PH-PAECs with ^13C Isotopologue concentrations
   of TCA cycle metabolites determined by GC-MS (D). The dark brown arrows
   indicate oxidative carboxylation flux from glutamine. The light brown
   arrows indicate reductive carboxylation. Dark brown circles represent
   ^13C in the oxidative glutamine metabolism pathway. Light brown circles
   represent ^13C in the reductive carboxylation pathway. Grey circles
   represent ^12C carbon. The bar graphs show fractional enrichment from
   the ^13C[5]-glutamine tracer generated by IsoCor software. A schematic
   model of the enzymes and the fate of glutamine in the TCA cycle (E).
   The dark brown arrows indicate oxidative carboxylation, and the light
   brown arrows indicate reductive carboxylation. The fractions of citrate
   containing four ^13C carbons (F) or five ^13C carbons (G) after
   culturing with ^13C[5]-glutamine are shown. Data are mean ± SEM;
   Experiments were performed with PAECs isolated from three independent
   PH lambs and three independent age-matched control lambs.

   We next investigated how the enzymes involved in glutaminolysis
   metabolism in PH-PAECs. Western blot analysis identified increased
   levels of the glutamine transporter ASCT2 in PH-PAECs ([112]Fig. 2A).
   We also found that the enzymes involved in the glutaminolysis pathway:
   glutaminase 1 (GLS1, [113]Fig. 2B), glutaminase 2 (GLS2, [114]Fig. 2C),
   and glutamate dehydrogenase 1 and 2 (GLUD 1/2, [115]Fig. 2 D) were all
   upregulated in PH-PAECs. As Mitochondrial α-KG derived from glutamate
   can participate in the TCAcycle, supporting the oxidative
   phosphorylation or reductive carboxylation pathway ([116]Fig. 1E) we
   also evaluated the expression levels of enzymes active in oxidative
   phosphorylation (OGDH) and reductive carboxylation of glutamine (ACO2
   and IDH2). Western blot analysis revealed that OGDH ([117]Fig. 2E),
   IDH2 ([118]Fig. 2F), and ACO2 ([119]Fig. 2G) were increased in
   PH-PAECs. Taken together, our ^13C-metabolic flux analyses and Western
   Blot analysis confirmed a significantly increased glutaminolysis with
   both enhanced glutamine oxidation and reductive carboxylation in
   PH-PAECs.

Fig. 2.

   [120]Fig. 2
   [121]Open in a new tab

   Increased expression of enzymes responsible for glutaminolysis in
   pulmonary hypertensive pulmonary arterial endothelial cells. Western
   blot analysis revealed increased expression of glutamine transporter
   ASCT2 (A), as well as the glutaminolysis enzymes GLS1 (B), GLS2 (C) and
   GLUD 1/2 (D) in PH-PAECs. There was also increased expression of the
   enzymes responsible for both oxidative decarboxylation, OGDH (E) and
   reductive carboxylation, IDH2 (F) and ACO2 (G) in PH-PAECs.
   Representative images are shown. β-actin was used to normalize loading.
   Data are mean ± SEM; Experiments were performed with PAECs isolated
   from three independent PH lambs and three independent age-matched
   control lambs.

   c-Myc overexpression induces a hyperproliferative, antiapoptotic
   phenotype in control pulmonary arterial endothelial cells. c-Myc
   regulates cell growth and proliferation [[122]48,[123]49]. Beyond
   speculation, little information exists regarding the role of c-Myc in
   the hyperproliferative antiapoptotic EC phenotype and metabolic
   reprogramming associated with PH development [[124]50]. Therefore, we
   next investigated if increases in c-Myc expression correlated with
   these metabolic changes we observed in PH-PAECs. Western blot data
   demonstrated that c-Myc expression was increased in PH-PAECs compared
   to control PAECs ([125]Fig. 3A). To determine if the increase in c-Myc
   could be involved in the hyperproliferative anti-apoptotic phenotype in
   PH-PAECs [[126]47] we used an adenoviral expression construct to
   overexpress c-Myc in PAECs. Western blot analysis showed that a
   multiplicity of infection (MOI) = 5 achieved a significant increase in
   c-Myc protein levels ([127]Fig. 3B). Next, direct cell counts
   ([128]Fig. 3C) and the measurement of 5′-bromo-2′-deoxyuridine (BrdU)
   incorporation ([129]Fig. 3D) were used to confirm that c-Myc
   over-expression was sufficient to stimulate proliferation in control
   PAECs. Further, c-Myc over-expression also decreased the rate of
   TNF-mediated apoptosis ([130]Fig. 3E). Functionally, these metabolic
   changes resulted in a decrease in NO generation ([131]Fig. 3F).
   Together, these data implicate c-Myc in developing the
   hyperproliferative anti-apoptotic phenotype in PH-PAECs [[132]47].

Fig. 3.

   [133]Fig. 3
   [134]Open in a new tab

   c-Myc enhances cellular proliferation and decreases apoptosis in
   pulmonary arterial endothelial cells. Western blot analysis reveals
   increased c-Myc protein levels in PH-PAECs (A). Control PAECs were
   transduced with Ad-c-Myc (MOI = 5) or Ad-GFP (MOI = 5; control) for
   48h, and Western blot analysis was used to confirm increased c-Myc
   protein levels (B). Representative images are shown. β-actin was used
   to normalize loading. c-Myc overexpression enhances the proliferation
   of PAECs as determined by direct cell counting (C) and
   5-bromo-2′-deoxyuridine (BrdU) incorporation (D). The levels of
   apoptosis induced by TNFα (2 ng/ml, 12h) detected by TUNEL staining was
   decreased in c-Myc overexpressing cells (E). Scale bar = 20 μm. c-Myc
   overexpression decreases NO generation in PAECs, as shown by reductions
   in fluorescent intensity of the NO probe DAF-FM (F). Scale bar = 20 μm.
   Data are mean ± SEM; PH-PAECs were isolated from three independent PH
   lambs and three independent age-matched control lambs; all other
   experiments were performed with at least four biological replicates.

   c-Myc overexpression alters cell metabolism to favor glutaminolysis in
   control pulmonary arterial endothelial cells. We utilized an untargeted
   snapshot metabolomics profiling approach to investigate how c-Myc
   over-expression alters metabolism in control PAECs. Snapshot
   metabolomics results were analyzed using the Metaboanalyst platform.
   Both PCA ([135]Fig. 4A) and PLS-DA plots ([136]Fig. 4B) showed cluster
   differences between control and c-Myc overexpressing PAECs. Heatmap was
   used to depict the 95 metabolites altered by c-Myc overexpression
   ([137]Fig. 4C). Clustered heat map analysis of GC-MS data showed
   metabolites significantly decreased displayed in green. Metabolites
   that have significantly increased are displayed in red, with the
   brightness of each color corresponding to the magnitude of the
   difference compared to the average value. Pathway analysis was also
   used to identify the most significantly impacted metabolic pathways in
   c-Myc-overexpressing PAECs. These were identified as the purine
   metabolism pathway, the citric acid cycle (TCA cycle), the pentose
   phosphate pathway, the glyoxylate and dicarboxylate metabolism pathway,
   glycine, serine, and threonine metabolism, and aspartate, glutamine,
   and glutamate metabolism ([138]Fig. 4D). c-Myc overexpression increased
   the levels of guanine and inosine in the purine metabolism pathway
   ([139]Fig. 4E). Among the TCA intermediates, levels of citrate,
   succinate, and malate were elevated in c-Myc overexpressing PAECs
   ([140]Fig. 4F). The levels of ribose and ribose 5-phosphate were
   increased within the pentose phosphate pathway ([141]Fig. 4G).
   Overexpression of c-Myc also increased the levels of several amino
   acids, including glycine, serine, and cysteine ([142]Fig. 4H). Levels
   of glutamine, glutamate, and aspartate were also elevated ([143]Fig.
   4I). Further, Western blot analysis revealed a significant increase in
   glutamine transporter, ASCT2, expression level in c-Myc overexpressing
   PAECs ([144]Fig. 4J). The expression levels of enzymes involved in
   glutaminolysis pathway including GLS1 ([145]Fig. 4K), GLS2 ([146]Fig.
   4L), and GLUD1/2 ([147]Fig. 4M) were also increased in c-Myc
   overexpressing PAECs.

Fig. 4.

   [148]Fig. 4
   [149]Open in a new tab

   c-Myc overexpression causes metabolic reprogramming in pulmonary
   arterial endothelial cells. PCA (A) and PLS-DA (B) score plots were
   generated using MetaboAnalyst based on intracellular metabolites in
   c-Myc overexpressing PAECs compared with GFP-expressing (control)
   PAECs. Grey: Control group; yellow: c-Myc overexpressing group. Heatmap
   illustrating alterations in metabolite levels in c-Myc overexpressing,
   compared to GFP-expressing (control) PAECs (C). Red indicates the
   relative upregulation, and Green indicates the relative downregulation
   of individual metabolites. A metabolic pathway analysis plot generated
   using MetaboAnalyst shows that c-Myc overexpression alters multiple
   metabolic pathways in PAECs (D). The x-axis represents the pathway
   impact value computed from pathway topological analysis, and the y-axis
   is the-log of the P-value obtained from pathway enrichment analysis
   (D). The most significantly changed pathways are characterized by a
   high-log(p) value and a high impact value (top right region). Among the
   purine bases, the levels of inosine and guanine were increased in c-Myc
   overexpressing PAECs (E). The over-expression of c-Myc significantly
   increased the TCA intermediates citrate, malate, and succinate (F).
   Within the pentose phosphate pathway, the levels of ribose-5-phosphate
   and ribose were increased in c-Myc overexpressing PAECs (G). c-Myc
   increased the levels of glycine, serine, and cysteine among the
   glyoxylate and dicarboxylate metabolism pathway and the glycine,
   serine, and threonine pathway (H). The over-expression of c-Myc
   increased the levels of aspartate, glutamine, and glutamate (I).
   Western blot analysis confirmed increases in ASCT2 (J), GLS1 (K), GLS2
   (L), GLUD (M) enzyme levels in c-Myc overexpressing PAECs.
   Representative images are shown. β-actin was used to normalize loading.
   Data are mean ± SEM; All experiments were performed with at least four
   biological replicates.

   Based on the increased levels of glutamine metabolizing enzymes, we
   next investigated if c-Myc over-expression mimicked the increase in the
   metabolic flux of glutamine through the TCA cycle we observed in
   PH-PAECs. Ad-c-Myc or Ad-GFP (control) expressing PAECs were cultured
   with media containing ^13C[5]-glutamine for 6h, and the TCA cycle
   intermediates were analyzed using GC-MS. PCA ([150]Fig. 5A) and PLS-DA
   plots ([151]Fig. 5B) showed a cluster difference between the control
   and c-Myc-overexpressing groups. Clustered heat map analysis of GC-MS
   data revealed increased TCA cycle metabolite levels in
   c-Myc-overexpressing PAECs ([152]Fig. 5C). Analysis of the isotopologue
   distribution from the ^13C[5]-glutamine flux indicated enhanced
   glutamine utilization which subsequently increased the refueling of
   carbons into the TCA cycle ([153]Fig. 5D). Analysis of the mean
   isotopic enrichment of molecules revealed significant increases in TCA
   cycle intermediates in c-Myc overexpressing PAECs ([154]Supplemental
   Fig. 2A–N). Importantly, analysis of citrate formed through oxidative
   phosphorylation (m+4) and reductive carboxylation (m+5) showed
   increased fractional amounts of both citrate (m+4; [155]Fig. 5E) and
   citrate (m+5; [156]Fig. 5F) in c-Myc overexpressing PAECs. Further, the
   c-Myc overexpressing PAECs presented increased oxidative glutaminolysis
   with elevated M+4-labeled TCA cycle intermediates and increased
   reductive carboxylation with elevated M+5 citrate followed by M+3
   malate, fumarate, succinate and aspartate ([157]Fig. 5D). Lastly, c-Myc
   overexpression increased OGDH ([158]Fig. 5G), IDH2 ([159]Fig. 5H) and
   ACO2 ([160]Fig. 5I) enzyme levels in PAECs confirming that c-Myc
   over-expression induces both increases in oxidative phosphorylation and
   reductive carboxylation.

Fig. 5.

   [161]Fig. 5
   [162]Open in a new tab

   Overexpression of c-Myc promotes metabolization of glutamine carbons in
   the TCA cycle. PCA (A) and PLS-DA (B) score plots were generated for
   the ^13C[5]-glutamine flux data showing the metabolic profile
   differences between GFP- (control) and c-Myc-overexpressing PAECs.
   Grey: Control group; yellow: c-Myc overexpressing group. Heatmap
   illustrating alterations in metabolite levels in c-Myc overexpressing,
   compared to GFP-expressing (control) PAECs (C). Red indicates the
   relative upregulation, and Green indicates the relative downregulation
   of the 15 differential metabolites (C). Schematic model of glutamine
   metabolism in PH-PAECs with ^13C Isotopologue concentrations of TCA
   cycle metabolites determined by GC-MS (D). The dark brown arrows
   indicate oxidative carboxylation flux from glutamine. The light brown
   arrows indicate reductive carboxylation. Dark brown circles represent
   ^13C in the oxidative glutamine metabolism pathway. Light brown circles
   represent ^13C in the reductive carboxylation pathway. Grey circles
   represent ^12C carbon. The bar graphs show fractional enrichment from
   the ^13C[5]-glutamine tracer generated by IsoCor software. The
   fractions of citrate containing four ^13C carbons (E) or five ^13C
   carbons (F) after culturing with ^13C[5]-glutamine are shown. Western
   blot analysis confirmed increases in OGDH (G), IDH2 (H), and ACO2 (I)
   enzyme levels in c-Myc overexpressing PAECs. Representative images are
   shown. β-actin was used to normalize loading. Data are mean ± SEM; All
   experiments were performed with at least four biological replicates.

   Overexpression of c-Myc activates glucose oxidation and increases
   glycolytic flux to favor Warburg phenotype. One of the hallmarks of PH
   is mitochondrial dysfunction coupled with a Warburg phenotype
   [[163]51,[164]52]. Thus, we investigated whether the c-Myc
   overexpressing PAECs also depended on mitochondrial glucose oxidation.
   To accomplish this, PAECs transduced with either Ad-c-Myc or Ad-GFP
   (control) were cultured with media containing ^13C[6]-glucose for 6h
   and GC-MS analyzed the extracted TCA cycle intermediates. Both PCA
   ([165]Supplemental Fig. 3A) and PLS-DA plots ([166]Supplemental Fig.
   3B) showed cluster difference between control and c-Myc-overexpressing
   groups. Heatmap was used to depict the 13 metabolites altered by c-Myc
   overexpression ([167]Supplemental Fig. 3C). Analysis of the
   isotopologue distribution from the ^13C[6]-glucose flux indicated
   increased glucose and lactate production ([168]Supplemental Fig. 3D).
   Analysis of the mean isotopic enrichment of molecules revealed
   significant increases in TCA cycle intermediates in c-Myc
   overexpressing PAECs when compared to control PAECs ([169]Supplemental
   Fig. 3E–Q). Importantly, c-Myc overexpression in PAECs significantly
   upregulated pyruvate ([170]Supplemental Fig. 3E) and lactate
   ([171]Supplemental Fig. 3F) metabolism, indicative of increased
   glycolysis. Further, our data show that c-Myc overexpression induces
   mitochondrial dysfunction as demonstrated by increases in mitochondrial
   reactive oxygen species (ROS) levels ([172]Fig. 6A) and reductions in
   the mitochondrial membrane potential ([173]Fig. 6B). Next, utilizing
   the Seahorse bioanalyzer, we performed ATP rate analyses to obtain
   real-time kinetic quantification of ATP production from oxidative
   phosphorylation and glycolysis ([174]Fig. 6C). Total ATP production
   rates were significantly reduced in c-Myc-overexpressing PAECs
   ([175]Fig. 6D). This was associated with a significant reduction in ATP
   generated by oxidative phosphorylation ([176]Fig. 6E) that was offset
   somewhat by increases in ATP generated by glycolysis ([177]Fig. 6F).
   Overall, these changes significantly reduced the ATP rate index
   ([178]Fig. 6G). Next, the mitostress test ([179]Fig. 7A) was used to
   confirm the ATP rate data. Although c-Myc over-expression did not alter
   the basal oxygen consumption rate (OCR, [180]Fig. 7B) or the OCR for
   ATP synthesis ([181]Fig. 7C) the reserve ([182]Fig. 7D) and maximal
   ([183]Fig. 7E) respiratory capacity were attenuated, and the Proton
   leak was enhanced ([184]Fig. 7F). Conversely, the glycostress test
   ([185]Fig. 7G) revealed that basal glycolysis ([186]Fig. 7H), as well
   as the reserve ([187]Fig. 7I) and maximum ([188]Fig. 7J) glycolytic
   capacity, were all increased confirming a Warburg phenotype is induced
   by c-Myc over-expression.

Fig. 6.

   [189]Fig. 6
   [190]Open in a new tab

   Overexpression of c-Myc disrupts mitochondrial function and alters
   cellular ATP production rate in pulmonary arterial endothelial cells.
   The over-expression of c-Myc significantly increases mitochondrial ROS
   generation (A) and decreases mitochondrial membrane potential (B) in
   PAECs. The overexpression of c-Myc also decreases the total cellular
   ATP production rate (C&D). The mitochondrial ATP production rate is
   reduced (E), but the glycolytic ATP production rate increases (F).
   Overall, these changes decreased the ATP rate index (G). Data are
   mean ± SEM; All experiments were performed with at least six biological
   replicates.

Fig. 7.

   [191]Fig. 7
   [192]Open in a new tab

   Overexpression of c-Myc disrupts mitochondrial bioenergetics and
   stimulates cellular glycolysis in pulmonary arterial endothelial cells.
   c-Myc overexpression disrupts the bioenergetic profile for oxygen
   consumption rate (OCR) (A). Although basal respiration (B) and the OCR
   for ATP synthesis (C) remain unchanged, the reserve (D) and maximum (E)
   respiratory capacity are decreased. At the same time, the proton leak
   is increased (F). c-Myc overexpression stimulates the extracellular
   acidification rate (ECAR) profile in PAECs (G) such that basal
   glycolysis (H), the reserve (I), and maximal (J) glycolytic capacities
   are all enhanced. Data are mean ± SEM; All experiments were performed
   with at least ten biological replicates.

   c-Myc inhibition attenuates the flux of glutamine-derived carbon into
   the TCA cycle in PH-PAECs and reverses the hyperproliferative,
   antiapoptotic phenotype in pulmonary hypertensive pulmonary arterial
   endothelial cells (PH-PAECs). Next, we explored the effect of c-Myc
   inhibition on glutamine metabolism and the hyperproliferative
   anti-apoptotic phenotype in PH-PAECs. To accomplish this, PH-PAECs were
   treated with the c-Myc inhibitor 10058-F4 (20 μM, 16h) and media was
   changed to ^13C[5]-glutamine containing media 6h prior to metabolite
   extraction. Both PCA ([193]Fig. 8A) and PLS-DA plots ([194]Fig. 8B)
   revealed distinct cluster differences between the treated and untreated
   groups. Clustered heat map analysis of GC-MS data showed decreased
   levels of TCA cycle metabolites in PH-PAECs treated with 10058-F4
   ([195]Fig. 8C). Analysis of the isotopologue distribution from the
   ^13C[5]-glutamine flux revealed that 10058-F4 induced a significant
   decrease in glutamine utilization which subsequently decreased the
   refueling of carbons into the TCA cycle in PH-PAECs ([196]Fig. 8D).
   This was confirmed by analyzing the mean isotopic enrichment of
   molecules ([197]Supplemental Fig. 4A–M). Citrate formed through
   oxidative phosphorylation (m+4, [198]Fig. 8E) and reductive
   carboxylation (m+5, [199]Fig. 8F) was reduced by 10058-F4. Further,
   decreased M+4-labeled TCA cycle intermediates indicated reduced
   oxidative glutaminolysis while decreased M+3 malate, fumarate,
   succinate and aspartate implied attenuated reductive carboxylation
   ([200]Fig. 8D). Western blot analysis also revealed that the levels of
   ASCT2 ([201]Fig. 8G), GLS1 ([202]Fig. 8H), GLS2 ([203]Fig. 8I), GLUD
   ([204]Fig. 8J) in PH-PAECs. OGDH ([205]Fig. 8K), IDH2 ([206]Fig. 8L),
   ACO2 ([207]Fig. 8M) were reduced by 10058-F4 exposure. These results
   show that the inhibition of c-Myc significantly reduces glutamine
   uptake and attenuates glutamine-derived carbon's flux into the TCA
   cycle in PH-PAECs. 10058-F4 also reduced mitochondrial ROS levels
   ([208]Fig. 9A) and increased the mitochondrial membrane potential
   ([209]Fig. 9B) in PH-PAECs, suggesting a restoration of mitochondrial
   function. In addition, the glycostress test ([210]Fig. 9C) revealed
   that 10058-F4 decreased basal glycolysis ([211]Fig. 9D), the reserve
   ([212]Fig. 9E), and maximal ([213]Fig. 9F) glycolytic capacity.
   Reversing the metabolic reprogramming also reduced cell proliferation
   ([214]Fig. 9G and H), restored TNF-mediated apoptosis ([215]Fig. 9I),
   and enhanced NO generation ([216]Fig. 9J) in PH-PAECs. As c-Myc
   inhibition attenuated the Warburg effect in PH-PAECs we examined the
   effect of HIF-1 inhibition (CAY10585, 10 μM, 24h) on the PH EC
   phenotype. Inhibition of HIF-1α altered the extracellular acidification
   rate (ECAR) profile in PH-PAECs ([217]Fig. 10A). Interestingly, basal
   glycolysis ([218]Fig. 10B) was increased, but the reserve ([219]Fig.
   10C) and maximal ([220]Fig. 10D) glycolytic capacities were decreased.
   HIF-1α inhibition also reversed the PH-EC phenotype as shown by
   decreases in the proliferation of PH-PAECs as determined by direct cell
   counting ([221]Fig. 10E) and 5-bromo-2′-deoxyuridine (BrdU)
   incorporation ([222]Fig. 10F). HIF-1α inhibition also restored
   TNFα-induced apoptosis ([223]Fig. 10G). Thus, targeting the HIF-1α can
   reverse the PH-EC phenotype.

Fig. 8.

   [224]Fig. 8
   [225]Open in a new tab

   Inhibition of c-Myc attenuates glutamine metabolism in pulmonary
   hypertensive pulmonary arterial endothelial cells. PH-PAECs were
   treated or not with the c-Myc inhibitor 10058-F4 (20 μm, 16h). PCA (A)
   and PLS-DA (B) score plots were generated for the ^13C[5]-glutamine
   flux data showing the metabolic profile differences associated with
   c-Myc inhibition. Brown: PH-PAECs; Blue: c-Myc inhibition. Hierarchical
   clustering heat map of the 15 differential metabolites, with the degree
   of change marked with brown (up-regulation) and blue (down-regulation)
   with c-Myc inhibition (C). Schematic model of glutamine metabolism in
   10058-F4 treated PH-PAECs with ^13C Isotopologue concentrations of TCA
   cycle metabolites determined by GC-MS (D). The dark brown arrows
   indicate oxidative carboxylation flux from glutamine. The light brown
   arrows indicate reductive carboxylation. Dark brown circles represent
   ^13C in the oxidative glutamine metabolism pathway. Light brown circles
   represent ^13C in the reductive carboxylation pathway. Grey circles
   represent ^12C carbon. The bar graphs show fractional enrichment from
   the ^13C[5]-glutamine tracer generated by IsoCor software. The
   fractions of citrate containing four ^13C carbons (E) or five ^13C
   carbons (F) after culturing with ^13C[5]-glutamine are shown. Western
   blot analysis confirmed that c-Myc inhibition decreased ASCT2 (G), GLS1
   (H), GLS2 (I), GLUD (J), OGDH (K) IDH2 (L), and ACO2 (M) protein levels
   in PH-PAECs. Representative images are shown. β-actin was used to
   normalize loading. Data are mean ± SEM; All experiments were performed
   with at least four biological replicates.

Fig. 9.

   [226]Fig. 9
   [227]Open in a new tab

   Inhibition of c-Myc reverses the Warburg and hyperproliferative
   antiapoptotic phenotype in pulmonary hypertensive pulmonary arterial
   endothelial cells. PH-PAECs were treated or not with the c-Myc
   inhibitor (10058-F4, 20 μm, 16h). c-Myc inhibition significantly
   decreases mitochondrial ROS generation (C) and increases the
   mitochondrial membrane potential (B) in PH-PAECs. Inhibition of c-Myc
   reduced the extracellular acidification rate (ECAR) profile in PH-PAECs
   (C) such that basal glycolysis (D) and the reserve (E) and maximal (F)
   glycolytic capacity were all decreased. c-Myc inhibition also reduced
   the proliferation of PH-PAECs as determined by direct cell counting (G)
   and 5-bromo-2′-deoxyuridine (BrdU) incorporation (H). The levels of
   apoptosis induced by TNFα (2 ng/ml, 12h) detected by TUNEL staining was
   increased by c-Myc inhibition (I). Scale bars: 20 μm. Inhibition of
   c-Myc restores NO levels in PH-PAECs, as shown by increases in
   fluorescent intensity of the NO probe DAF-FM (J). Scale bars: 20 μm.
   Data are mean ± SEM; All experiments were performed with at least four
   biological replicates.

Fig. 10.

   [228]Fig. 10
   [229]Open in a new tab

   Inhibition of hypoxia-inducible factor 1α affects glycolysis and
   reverses the hyperproliferative antiapoptotic phenotype in pulmonary
   hypertensive pulmonary arterial endothelial cells. PH-PAECs were
   treated or not with a HIF-1α inhibitor (CAY10585, 10 μM, 24h).
   Inhibition of HIF-1α impacted the extracellular acidification rate
   (ECAR) profile in PH-PAECs (A) such that basal glycolysis was increased
   (B) while the reserve (C) and maximal (D) glycolytic capacity were
   decreased. HIF-1α inhibition reduced the proliferation of PH-PAECs as
   determined by direct cell counting (E) and 5-bromo-2′-deoxyuridine
   (BrdU) incorporation (F). The levels of apoptosis induced by TNFα
   (2 ng/ml, 12h) detected by TUNEL staining was increased by HIF-1α
   inhibition in PH-PAECs (G). Scale bars: 50 μm. Data are mean ± SEM; All
   experiments were performed with at least six biological replicates.

4. Discussion

   Previously, we have identified metabolic reprogramming [[230]53] and
   the development of the hyperproliferative antiapoptotic phenotype in
   PH-PAECs [[231]47]. However, the underlying molecular mechanism
   responsible for these events is unclear. Abnormal glutamine metabolism
   can significantly impact cellular pathways, including redox
   homeostasis, cell proliferation, autophagy, and the synthesis of
   biological macromolecules, and it is emerging as a mechanism driving
   the progression of PH [[232]46,[233]54]. The data we present in this
   study support a critical role for abnormal glutamine metabolism in the
   development of PH. Several critical findings arise from our studies.
   First, PH-PAECs isolated from a lamb model of PH showed increased
   glutamine oxidation and glutamine-dependent reductive carboxylation
   that correlated with increased c-Myc expression. Second, an untargeted
   snapshot metabolomics approach showed that the overexpression of c-Myc
   impacted cellular metabolism, particularly glutamine metabolism in
   control PAECs. Third, ^13C metabolic flux analyses showed that the
   overexpression of c-Myc increased glucose oxidation, glutamine
   oxidation, and glutamine-dependent reductive carboxylation. Fourth, the
   overexpression of c-Myc attenuated mitochondrial function and increased
   glycolysis, producing a Warburg phenotype that drives the development
   of a hyperproliferative antiapoptotic endothelial phenotype. Fifth,
   c-Myc inhibition in PH-PAECs reversed the metabolic reprogramming in
   PH-PAECs, attenuated cellular proliferation, and restored TNF-mediated
   apoptosis.

   Although numerous studies have demonstrated that c-Myc overexpression
   is linked to cellular hyperproliferation, tumor development, and cancer
   progression, minimal studies have investigated the role of c-Myc in PH
   development [[234][55], [235][56], [236][57], [237][58], [238][59]]. We
   previously reported hyperproliferation and decreased apoptosis in PAECs
   isolated from a lamb model of PH [[239]47]. This phenotype was linked
   to endothelial dysfunction and impaired vasodilation
   [[240]34,[241]60,[242]61]. Emerging evidence implicates metabolic
   reprogramming in the development of this hyperproliferative
   antiapoptotic phenotype [[243]62,[244]63]. However, the potential role
   of glutamine metabolism in this process has not been elucidated. Our
   ^13C metabolic flux analyses showed enhanced glutamine oxidation and
   glutamine-dependent reductive carboxylation in PH-PAECs. Enhanced
   glutamine oxidation was confirmed by analysis of citrate m+6
   isotopologue that could only have arisen from glutamine-derived m+4
   oxaloacetate (forward TCA cycle reactions). While the increased
   glutamine-dependent reductive carboxylation was characterized by the
   elevated citrate m+5 isotopologue that can only be generated by
   reductive carboxylation of glutamine-derived α-ketoglutarate.
   Simultaneously, the levels of key enzymes of the TCA cycle involved in
   glutamine oxidation, OGDH, and reductive carboxylation, including IDH2
   and ACO2, were also upregulated in PH-PAECs. Together, these data
   suggest that mitochondrial metabolism reprogramming is essential for
   providing the anaplerotic precursors required for macromolecule
   synthesis in hyperproliferative PH-PAECs. While glucose-derived carbon
   was shunted for lactate production, despite ample oxidative glucose
   metabolism, our data indicate that the endothelium in PH requires the
   reductive glutamine-dependent pathway to manifest the
   hyperproliferative phenotype. This is like the metabolic reprogramming
   observed in cancer. Further, this metabolic reprogramming comes at the
   expense of the ECs ability to generate NO. Endothelium-derived NO is an
   essential mediator of pulmonary vascular reactivity [[245]64].
   Previously, we have shown that PH-PAECs have decreased NO levels
   [[246]65]. While other studies have reported decreased endothelial NO
   synthase expression in pulmonary arteries in patients with PH
   [[247][65], [248][66], [249][67]]. Moreover, there is growing evidence
   of an inverse relationship between glutamine levels and NO production
   [[250]68]. Higher glutamine levels have been shown to attenuate NO
   production and influence endothelial-dependent relaxation in vitro
   [[251]69]. However, this is the first study to unravel the molecular
   mechanism demonstrating the link between glutamine metabolism and NO
   synthesis. Our data also suggest that reversing the glutaminolysis in
   PH could restore NO signaling and could be a new therapeutic target to
   enhance the current drugs used in PH, such as prostacyclin, which
   stimulate NO generation.

   Results from our metabolic flux analysis suggested there was likely a
   master regulator controlling glutamine metabolism during PH
   development. Previous studies in other systems had identified c-Myc
   [[252]56,[253]70], and we were able to identify increased levels of
   c-Myc in the PH-PAECs. The c-Myc protein (a helix-loop-helix leucine
   zipper family transcription factor) forms a dimeric structure with the
   Myc-Associated factor X (MAX) protein to regulate the expression of a
   broad number of genes [[254][71], [255][72], [256][73]]. There is a
   strong connection between c-Myc and cellular metabolism suggesting it
   plays an essential role in regulating global metabolic reprogramming
   [[257]56]. To understand the cellular and molecular effects of
   increased c-Myc expression in EC, we overexpressed c-Myc in control
   PAECs and found that this could mimic the metabolic changes we observed
   in PH-PAECs, suggesting that c-Myc is a key regulator of PH
   development. Indeed, numerous prior studies have documented that c-Myc
   controls multiple cellular pathways and mediates a broad
   transcriptional response that regulates many aspects of cellular growth
   and proliferation [[258][74], [259][75], [260][76], [261][77],
   [262][78]]. In support of these effects in EC, our snapshot
   metabolomics data revealed that c-Myc overexpression altered core
   cellular metabolism, increasing the production of TCAcycle
   intermediates required for macromolecule biosynthesis. Utilizing
   pathway analysis, we identified the six most significantly impacted
   metabolic pathways in c-Myc-overexpressing PAECs. These were identified
   as purine metabolism pathway, the citric acid cycle (TCA cycle), the
   pentose phosphate pathway, the glyoxylate and dicarboxylate metabolism
   pathway, glycine, serine, and threonine metabolism, and aspartate,
   glutamine, and glutamate metabolism. Within the purine metabolism
   pathway, levels of guanine and inosine were increased. Purine is
   essential for nucleic acid synthesis and is directly regulated by c-Myc
   [[263]79]. Importantly, purine metabolites are increased in PH patients
   and correlate with the severity of right ventricular-pulmonary vascular
   dysfunction [[264]80]. Purine synthesis suppression also reduces the
   development and progression of PH [[265]81]. Among the citric acid
   cycle, citrate, malate, and succinate were increased. This aligns well
   with work from Zhao et al. who reported that TCA cycle intermediates
   including citrate, succinate, and succinyl carnitine are significantly
   increased in lung tissue from patients with PH [[266]82]. Within the
   pentose phosphate pathway, the levels of ribose-5-phosphate and ribose
   were increased. Again, this aligns with prior work. Previously, we
   reported that elevated levels of pentose phosphate pathway metabolites,
   including ribose 5-phosphate, are associated with increased glucose
   uptake and utilization in PH [[267]83] while other studies have shown
   that c-Myc regulates the pentose phosphate pathway and plays an
   essential role in cell growth [[268]84]. The glyoxylate and
   dicarboxylate metabolism pathway and glycine, serine, and threonine
   metabolism pathway are interconnected sharing metabolites including
   glycine, serine and cysteine. Glycine and serine play roles in
   metabolic processes regulated by c-Myc [[269]85]. Glycine is known to
   be elevated in PH patients [[270]86]. Serine and glycine, which are
   required for one-carbon metabolism, are biosynthetically linked and can
   be used to supply pyruvate for cell proliferation [[271]87]. Finally,
   the pathway analysis identified increased metabolism of aspartate,
   glutamine, and glutamate. Again, this is supported by prior work, which
   identified high labeled fractions of aspartate and glutamate in
   c-Myc-overexpressing cells, suggesting that c-Myc promotes the
   synthesis of these amino acids from glutamine [[272]88]. Importantly, a
   significant increase in circulating glutamine levels has been
   demonstrated in PH patients, as has increased glutamine metabolism
   [[273]89,[274]90]. Holistically, our data show that c-Myc
   overexpression in PAECs regulates various aspects of cellular
   metabolism, providing the macromolecules and energy required to
   replicate DNA, undergo cell division, and increase cell mass all traits
   that are required to form a hyperproliferative phenotype. Intriguingly,
   among the potential candidate pathways, the glutamine metabolism
   pathway stands out as the most relevant to PH because c-Myc plays a
   significant role in promoting cell growth and proliferation by
   influencing glutamine metabolism and, consequently, cellular biomass
   production.

   It is now commonly accepted that, like many cancers, PH development is
   associated with a switch from oxidative phosphorylation to aerobic
   glycolysis, the Warburg effect [[275]51]. Interestingly, like PH-PAECs
   [[276]53], c-Myc overexpressing PAECs exhibited increased glycolysis as
   demonstrated by converting a majority of glucose-derived pyruvate to
   lactate [[277]15]. Glucose is an essential metabolic substrate and the
   central energy molecule in mammalian cells [[278]91]. Metabolic flux
   analysis using ^13C[6]-glucose as a tracer confirmed this Warburg
   effect with increased glycolytic flux and production of pyruvate and
   lactate in c-Myc overexpressing PAECs. This is an expected finding as
   c-Myc has been shown to activate the transcription of many glycolytic
   genes through binding to the classical E-box sequence (CACGTG)
   [[279]92] and drives overall metabolism to activate the cell cycle
   machinery essential for cell proliferation [[280]93], including glucose
   transporter, Hexokinase 2, and lactate dehydrogenase A
   [[281]94,[282]95]. Using a glycolysis stress test, we confirmed that
   glycolysis was significantly higher in the c-Myc overexpressing PAECs
   indicating that the Warburg effect is at least partially under c-Myc
   control. Further, our data provide evidence of increased ROS levels and
   decreased mitochondrial membrane potential in c-Myc overexpressing
   PAECs. Although, it has been previously shown in other systems that
   c-Myc overexpression increases intercellular ROS Levels [[283]96], this
   is the first study to link c-Myc to increased ROS levels in vascular
   ECs. Elevated ROS levels contribute to ROS-associated damage in DNA,
   proteins, and lipids. In addition, a strong correlation exists between
   ROS production and the loss of mitochondrial membrane potential in the
   development of PH [[284]39,[285]53,[286]97]. This occurs through
   depolarizing the inner mitochondrial membrane potential, which impairs
   ATP generation [[287]98,[288]99].

   When we evaluated the spectra for additional c-Myc dependent
   ^13C-enriched intermediary metabolites, we found increased levels of
   ^13C-labeled TCA cycle intermediates and metabolites, including
   aspartate. It has been documented that respiration plays a significant
   role in proliferating cells by providing electron acceptors for
   aspartate synthesis [[289]100]. Aspartate is a critical amino acid
   required to convert IMP to AMP in de novo purine synthesis and provides
   the carbon backbone for de novo pyrimidine synthesis. Thus, high levels
   of aspartate are necessary to produce the nucleotides needed for DNA
   synthesis in highly proliferating cells. Our results show additional
   evidence of association between increased glucose metabolism and
   nucleotide synthesis in c-Myc-expressing PAECs. In addition to glucose,
   glutamine is a significant nutrient essential for cell proliferation.
   c-Myc induction of glutamine metabolism is important for cell survival
   under glucose- or oxygen-deprived conditions [[290]101]. However, our
   metabolomic tracking studies using ^13C-labeled glucose and glutamine
   tracers in c-Myc overexpressing PAECs and PH-PAECs demonstrate the
   ability of c-Myc to drive a glucose-independent cycle using glutamine
   as the substrate. The c-Myc overexpressing PAECs and PH-PAECs may use a
   TCA cycle with hybrid intermediates containing both glucose and
   glutamine carbons. However, further research will be needed to define
   the role of c-Myc in the production of the TCA cycle hybrid
   intermediates.

   Previously, it has been reported in other systems that c-Myc stimulates
   genes involved in glutamine metabolism at the transcriptional and
   posttranscriptional levels [[291]32]. Indeed, results from our
   ^13C[5]-glutamine flux analysis show that overexpression of c-Myc in
   control PAECs stimulated glutamine metabolism. Interestingly, this
   resulted in enhanced glutamine oxidation and glutamine-dependent
   reductive carboxylation, as observed in PH-PAECs. Further, our Western
   blot results validated that c-Myc overexpressing PAECs show enhanced
   expression of amino acid transporters solute carrier family A1 member 5
   (ASCT2/SLC1A5) likely via the binding of c-Myc to promoter elements of
   ASCT2, promoting glutamine uptake [[292]32]. Our Western blot results
   also show upregulated glutaminase 1, glutaminase 2, and glutamate
   dehydrogenase 1/2 enzyme levels in c-Myc overexpressing PAECs
   suggesting increased glutaminolysis. c-Myc-mediated enhanced
   translation and activation of glutaminase 1, glutaminase 2, and
   glutamate dehydrogenase 1/2 enzymes have been reported in different
   cancers [[293]84,[294][102], [295][103], [296][104]]. Still, to our
   knowledge, this is the first report that these enzymes are increased in
   PH-PAECs. Our data suggest that the multiple metabolic events
   coordinated by c-Myc replenish TCA cycle intermediates.

   The c-Myc transcription factor can influence the electron transport
   chain (ETC) in several ways. Previously, we demonstrated that
   mitochondrial dysfunction in PH-PAECs correlates with deficiencies in
   the activities of Complexes I and II [[297]38]. Further, we identified
   decreased ETC Complex I assembly and linked these changes to disrupted
   mitochondrial bioenergetics. An impaired TCA cycle can directly affect
   the ETC by limiting the supply of electron carriers, which reduces ATP
   production and triggers feedback mechanisms that further impact both
   pathways [[298]105]. Importantly, upregulated c-Myc can directly induce
   reactive oxygen species (ROS) production potentially affecting the ETC
   [[299]106]. It has been suggested that approximately half of the ROS
   generated by c-Myc overexpression can be accounted for by the induction
   of CYP2C9 [[300]107]. Although, experimental findings from this study
   and others suggest that c-Myc might also influence the redox state and
   ETC activity via the activation of reverse TCA [[301]108]. Our data
   implicate reverse TCA as being activated by c-Myc overexpression.
   However, the full implications and regulation of the reverse TCA cycle
   are still not fully understood, particularly in PH development.
   Therefore, further studies will be required to understand the molecular
   mechanisms and the role of reverse TCA in ROS generation, as well as
   how this disrupts ETC activity and assembly in PH, and how these all
   factor into the overall metabolic reprogramming associated with PH.

   Our study also explored the effect of a small-molecule c-Myc
   inhibitor,10058-F4, on PH-PAECs as a pharmacological therapy to
   suppress vascular cell proliferation and attenuate PH progression. Our
   results show that 10058-F4 decreased cellular glycolysis, attenuated
   mitochondrial ROS production, increased mitochondrial membrane
   potential and reversed the hyperproliferative antiapoptotic EC
   phenotype in PH-PAECs. In addition, our ^13C[5]-glutamine flux analysis
   showed that 10058-F4 reduced glutamine derived carbons from entering
   the TCA cycle, attenuating glutaminolysis. Further, inhibition of c-Myc
   suppressed the expression of glutamine transporters (ASCT2/SLC1A5),
   glutaminase 1, glutaminase 2 and glutamate dehydrogenase 1/2 enzyme
   levels. Subsequently, the levels of enzymes of the TCA cycle involved
   in glutamine oxidation and reductive carboxylation including IDH2, ACO2
   and OGDH were also downregulated by c-Myc inhibition. Over the last
   decade, several c-Myc inhibitors have been identified to suppress cell
   proliferation in diverse cancer cell lines [[302][109], [303][110],
   [304][111], [305][112]]. However, the pharmacological inhibition of
   vascular cell hyperproliferation utilizing c-Myc inhibitor has not been
   explored. Results from our study suggest that future therapies
   targeting c-Myc signaling could be beneficial for preventing PH disease
   progression. However, it should be noted that a significant concern
   with utilizing c-Myc inhibitor is the possible side effects resulting
   from inactivating a master regulator essential for normal cell survival
   and proliferation [[306]113]. However, several conventional small
   molecules, including 10058-F4, have been identified to inhibit
   c-Myc/Max dimerization, and these could be exploited to suppress
   PH-PAECs hyperproliferation [[307]114] and potentially restore NO
   signaling.

   Previously, studies have shown a complex relationship between c-Myc and
   HIF-1α during PH development [[308]115]. c-Myc has been implicated in
   increased HIF activity through both transcriptional and
   post-transcriptional mechanisms [[309]116,[310]117]. Previously, we
   reported that PH-PAECs have significantly higher levels of HIF-1α
   protein [[311]118]. In this study, we investigated the effect of the
   small-molecule HIF-1α inhibitor CAY10585 on PH-PAECs as a potential
   pharmacological therapy to suppress glycolysis and inhibit cell
   proliferation. Our results show that CAY10585 treatment was unable to
   reduce cellular basal glycolysis. This is likely because the glycolytic
   pathway is a multifaceted metabolic pathway and signaling hub involving
   multiple regulators. Previously, we and others have reported that
   Hypoxia-inducible factor 2α (HIF-2α) also regulates glycolysis and
   plays a role in the development of PH [[312]119,[313]120]. Although,
   both HIF-1α and HIF-2α play important roles in regulating
   transcriptional responses that affect the development of PH, recent
   studies have shown that HIF-2α plays a significant role in the
   development of PH in PAECs, and that HIF-2α is highly expressed in the
   endothelium of the pulmonary vasculature [[314]121,[315]122]. Thus, it
   is possible that in the absence of HIF-1α, the HIF-2α subunit was able
   to partially compensate at least at the level of basal glycolysis, as
   our data show that CAY10585 significantly decreases both the reserve
   and maximal glycolytic capacity in PH-PAECs. Importantly, the
   inhibition of HIF-1α reversed the hyperproliferative antiapoptotic EC
   phenotype in PH-PAECs. Further, we have previously demonstrated that
   CAY10585 was able to restore NO production in PH-PAECs [[316]118]. This
   links metabolic reprogramming to the loss of NO signaling and
   endothelial function in PH development, and thus targeting HIF-1α with
   inhibitors shows promise for treating PH. However, further research is
   needed to fully understand the intricate relationship between c-Myc,
   HIF-1α, and HIF-2α, their synergistic interaction, and antagonistic
   effects during PH development.

   In conclusion, our results demonstrate the broad effects of elevated
   c-Myc on glucose and glutamine carbon flux that increases glycolysis
   and glutaminolysis resulting in the modulation of TCA cycle
   intermediates. Further, the inhibition of c-Myc in PH-PAECs reduces
   both glycolysis and glutaminolysis, reversing the hyperproliferative
   antiapoptotic EC phenotype associated with PH development. Although, in
   vivo studies demonstrate that inhibition of c-Myc stimulates
   anti-proliferative effect, further research is essential to identify
   the regulation mechanism of the c-Myc interactome, and to bring the
   potential c-Myc inhibitors to clinical practice. Nevertheless,
   targeting and reducing glutaminolysis could be a potential therapeutic
   strategy to treat PH.

CRediT authorship contribution statement

   Manivannan Yegambaram: Writing – original draft, Methodology,
   Investigation, Formal analysis, Data curation. Xutong Sun: Writing –
   review & editing, Investigation, Formal analysis, Data curation. Qing
   Lu: Writing – review & editing, Investigation, Formal analysis, Data
   curation. Alejandro Garcia Flores: Writing – review & editing,
   Investigation, Formal analysis, Data curation. Marina Zemskova:
   Investigation, Formal analysis, Data curation. Jamie Soto:
   Investigation, Formal analysis, Data curation. Adam Rauckhorst: Writing
   – review & editing, Methodology. Emin Maltepe: Writing – review &
   editing. Ting Wang: Writing – review & editing, Supervision, Funding
   acquisition, Conceptualization. Jeffrey R. Fineman: Writing – review &
   editing, Supervision, Funding acquisition, Conceptualization. Stephen
   M. Black: Writing – review & editing, Supervision, Project
   administration, Funding acquisition, Data curation, Conceptualization.

Funding

   This research was supported in part by HL60190 (SMB), HL137282
   (SMB/JRF), HL134610 (SMB/TW), HL142212 (SMB/TW), HL146369 (SMB/TW/JRF),
   and UG3HG013615 (SMB).

Declaration of competing interest

   None.

Acknowledgement