Alpha-Ketoglutarate or 5-HMF: Single Compounds Effectively Eliminate Leukemia Cells via Caspase-3 Apoptosis and Antioxidative Pathways

Background: We recently showed that a combined solution containing alpha-ketoglutarate (aKG) and 5-hydroxymethyl-furfural (5-HMF) has a solid antitumoral effect on the Jurkat cell line due to the fact of its antioxidative, caspase-3 and apoptosis activities, but no negative effect on human fibroblasts was obtained. The question arises how the single compounds, aKG and 5-HMF, affect peroxynitrite (ONOO−) and nitration of tyrosine residues, Jurkat cell proliferation and caspase-activated apoptosis. Methods: The ONOO− luminol-induced chemiluminescence reaction was used to measure the ONOO− scavenging function of aKG or 5-HMF, and their protection against nitration of tyrosine residues on bovine serum albumin was estimated with the ELISA technique. The Jurkat cell line was cultivated in the absence or presence of aKG or 5-HMF solutions between 0 and 3.5 µM aKG or 0 and 4 µM 5-HMF. Jurkat cells were tested for cell proliferation, mitochondrial activity and caspase-activated apoptosis. Results: aKG showed a concentration-dependent reduction in ONOO−, resulting in a 90% elimination of ONOO− using 200 mM aKG. In addition, 20 and 200 mM 5-HMF were able to reduce ONOO− only by 20%, while lower concentrations of 5-HMF remained stable in the presence of ONOO−. Nitration of tyrosine residues was inhibited 4 fold more effectively with 5-HMF compared to aKG measuring the IC50%. Both substances, aKG and 5-HMF, were shown to cause a reduction in Jurkat cell growth that was dependent on the dose and incubation time. The aKG effectively reduced Jurkat cell growth down to 50% after 48 and 72 h of incubation using the highest concentration of 3.5 µM, and 1, 1.6, 2, 3 and 4 µM 5-HMF inhibited any cell growth within (i) 24 h; 1.6, 2, 3 and 4 µM 5-HMF within 48 h (ii); 2, 3 and 4 µM 5-HMF within 72 h (iii). Furthermore, 4 µM was able to eliminate the starting cell number of 20,000 cells after 48 and 72 h down to 11,233 cells. The mitochondrial activity measurements supported the data on aKG or 5-HMF regarding cell growth in Jurkat cells, in both a dose- and incubation-time-dependent manner: the highest concentration of 3.5 µM aKG reduced the mitochondrial activity over 24 h (67.7%), 48 h (57.9%) and 72 h (46.8%) of incubation with Jurkat cells compared to the control incubation without aKG (100%). 5-HMF was more effective compared to aKG; the mitochondrial activity in the presence of 4 µM 5-HMF decreased after 24 h down to 68.4%, after 48 h to 42.9% and after 72 h to 32.0%. Moreover, 1.7 and 3.4 µM aKG had no effect on caspase-3-activated apoptosis (0.58% and 0.56%) in the Jurkat cell line. However, 2 and 4 µM 5-HMF increased the caspase-3-activated apoptosis up to 22.1% and 42.5% compared to the control (2.9%). A combined solution of 1.7 µM aKG + 0.7 µM 5-HMF showed a higher caspase-3-activated apoptosis (15.7%) compared to 1.7 µM aKG or 2 µM 5-HMF alone. In addition, 3.5 µM µg/mL aKG + 1.7 µM 5-HMF induced caspase-activated apoptosis up to 55.6% compared to 4.5% or 35.6% caspase-3 activity using 3.5 µM aKG or 4 µM 5-HMF. Conclusion: Both substances showed high antioxidative potential in eliminating either peroxynitrite or nitration of tyrosine residues, which results in a better inhibition of cell growth and mitochondrial activity of 5-HMF compared to aKG. However, caspase-3-activated apoptosis measurements revealed that the combination of both substances synergistically is the most effective compared to single compounds.


Introduction
Alpha-ketoglutarate (aKG) is commonly known for its function in the Krebs cycle and antioxidative capacity. Furthermore, the oxoglutarate dehydrogenase complex (OGDC) or α-ketoglutarate dehydrogenase complex is an enzyme complex that plays a major role in the citric acid cycle. aKG is a potent scavenger in eliminating peroxynitrite [1], a physiological nitrating agent derived from nitrosyl radial (NO • ) and superoxide anion radical (O 2 * − ; [2]), to succinate and nitrite, which can be easily converted back to NO • . 5-Hydroxymethylfurfural (HMF) is formed from reducing sugars in acidic environments when heated through the Maillard reaction. HMF has shown antioxidative, anti-allergic, antiinflammatory, antihypoxic, antisickling and antihyperuricemic effects. In addition, 5-HMF is known as a radical scavenger against hydroxyl radicals, alkyl radicals and superoxide anion radicals but also to inhibit oxidant enzymes, such as myeloperoxidase, and to increase the expression of glutathione and superoxide dismutase (SOD) in cell cultures [3].
We have recently shown that addition of a combined solution containing aKG + 5-HMF to leukemic cells exerted antitumoral, antiproliferative and apoptotic effects compared to HF-SAR [4]. This seems to be associated with the antioxidative capacity of both substances, including the prevention of oxidative protein modifications, but also protein nitration and oxidative DNA damage [5]. This is closely connected with increased nitrotyrosine levels, release of cytochrome c, activation of caspase-3 and alteration of the p53 gene.
During carcinogenesis, RONS, such as O 2 * − , are produced in a higher number, which are speculated to be involved in all steps including the initiation, promotion and activation of proto-oncogenes and tumor suppressor genes [6].
A newly presented meta-analysis correlated antioxidative activating enzymes, such as SOD, CAT, GPx and levels of RONS markers, namely, malondialdehyde (MDA) and 8-hydroxy-desoxoguanosine , between cancer patients and a healthy population [7]. Overall, MDA, a lipid peroxidation end product, and 8-OHdG, a nucleic acid oxidation marker, were markedly higher in a cancer group that also included CLL patients, whereas all antioxidative enzymes levels were lower compared to the healthy population. We estimated that the Jurkat cell line expresses higher carbonylated proteins in the cell membrane compared to normal cells [4]. Malondialdehyde is among the most expressed subgroups of these carbonyl groups on membrane proteins. Thus, we suggest that antioxidative substances, such as alpha-ketoglutarate and 5-HMF, can counteract the cell growth of leukemic cells directly in the mitochondrial activity in terms of aKG and 5-HMF. Additionally, both substances detoxify ammonia on one the hand by the transfer ionically with aKG to the glutamate-glutamine-oxalacetate pathway (GOT) or, on the other hand, covalently to 5-HMF.
aKG is required as a potent oxidant for the reductive carboxylation bidirectional in cancer cells with mitochondrial defects to operate IDH inversely and SDH directly [8] Under intracellular acidic conditions, which does take place intracellular in cancer cells, the nonenzymatic decarboxylation of aKG in the presence of H 2 O 2 and ONOO − cannot occur. Thus, nonenzymatically generated succinate decreases, which itself is able to stabilize HIFα [1].
Both aKG and 5-HMF are also involved in increasing SOD, GPx and CAT to reduce RONS in cell culture and cancer patients [3,9]. Free and bound carbonyl proteins levels were measured in patients with cancer before and after receiving both substances: aKG and 5-HMF [9][10][11]. Supplemented or iv-applicated aKG + 5-HMF effectively reduced the carbonylated proteins in cancer patients compared to the placebo groups [10]. Carbonyl proteins were also effectively reduced by the combination of aKG and HMF in cell culture using aKG + 5-HMF together.
The question arises over which substance of the combined solution of aKG + 5-HMF has (i) antiproliferative, (ii) caspase-3 and apoptosis effects on leukemic cells or (iii) an antinitrating effect on scavenging peroxynitrite to prevent nitro-tyrosine, or if they act in a synergistic manner. In addition to the oxidizing ability of ONOO − , nitration by nitrating agents of biomolecules, such as ONOO − , is involved directly in cytochrome c release, activation of caspase-3 and p53 alterations.

Cell Proliferation Experiments
After 24 h, no difference in the cell growth of Jurkat cells was obtained in the presence or absence of 5-HMF compared to the starting conditions. After 48 h of incubation, cell growth in the absence of 5-HMF increased up to 2.1 fold (421,245 ± 8122 cells) compared to starting conditions (20,000 cells). The usage of 1.6, 2, and 3 µM 5-HMF inhibited any cell growth of Jurkat cells after 48 h compared to the control, and 4 µM 5-HMF reduced cell growth after 48 h of incubation by 40%. After 72 h of incubation, no cell growth was estimated using 2 µM compared to the starting conditions, whereas 0 µg/mL 5-HMF increased 4.1 fold, 1 µM 5-HMF increased 2.7 fold and 1.6 µM 5-HMF increased 1.5 fold. The highest used 5-HMF concentrations, 3 and 4 µM, were able to reduce the cell growth of Jurkat cells by 33 and 44%, respectively.
Comparing the mitochondrial activity with the cell growth of the Jurkat cells, Figure 6A shows a high linear relation (r 2 = 0.9051) of all aKG concentrations at different incubation times, nearly the same for 5-HMF (r 2 = 0.8518) as presented in Figure 6B. The slope in the linear function using aKG was even higher (k = 0.0047) compared to 5-HMF (k = 0.005). The 5-HMF clouds are shown more in the lower scales between 10 and 30 percent of cell growth, whereas aKG was randomly distributed between 30 and 70% of cell growth.

Discussion
The use of aKG and 5-HMF as main components in an anticancerogenic solution for treating prostatic tumor patients by intravenous application was recently presented [12,13]. The doubling time of PSA increased significantly as the primary outcome. Using the same solution, a reduction in tumor mass in non-small cancer lung patients and also a higher quality of life using the Karnowski index was obtained in the verum group compared to the placebo group [13]. The combination of aKG and 5-HMF showed also a better prevention against cigarette smoke-induced radical protein modification compared to ascorbic acid or its single compounds [14]. aKG regulates as substrate processes in the inner mitochondrial membrane such as alpha-ketoglutarate dehydrogenase (AKGDH) or HIFα [15], in addition to oxygen to inhibit growth factor (e.g., endothelial growth factor) in general. Numerous 2-oxo-glutarate-dependent dioxygenases (2OGDDs) are recurrently dysregulated in malignancies, and aKG is also required for its function in 2OGDDs in neoplasia [16]. Furthermore, aKG seems to be involved in subduing tumors in bladder cancer patients [17]. In both enzymes, namely, HIF-α and 2OGDD, oxygen functions in addition to 2-oxoglutarate (αKG) act as a substrate to decarboxylase aKG into forming succinate and CO 2 from O 2 [18]. 2OGDD controls the hydroxylation on the hypoxiainducible factor (HIF)-α subunit at the proline 402 or 564 of the prolyl hydroxylase domain (PHD), which plays an immense role in the regulation of hypoxia by promoting the von Hippel-Lindau protein, ubiquination and proteasomal degradation.
Nitrating agents, such as peroxynitrite, result in one way: by the excessive production of superoxide anion radicals (O 2 •− ) and nitrosyl radicals (NO • ). The nitration of organic compounds, such as (i) amino acids (e.g., tryptophane), (ii) carbohydrates, (iii) nucleic acids and (iv) fatty acids, is involved in normal metabolic pathways but also in the beginning or generation of several pathologies such as cardiovascular disease, hypo-and hypertension, neurodegenerative processes, such Morbus Parkinson's or Morbus Alzheimer's disease, inflammation, cancer-causing pathologies and septicemias [19]. Arginine, the main compound in the enzymatic production of NO • via the NOS pathway (NO-synthase), itself is able to be nitrated to nitro-arginine, which is a potent inhibitor for the endothelial, neuronal, macrophage and inducible NOS.
Moreover, it results in the oxidation of thiols peroxynitrite nitrate and oxidize tyrosine residues. The estimation of nitro tyrosine residues on several proteins is evidence of extensively generated peroxynitrite in vivo and in vitro [20,21]. We recently showed that aKG is able to recycle nonenzymatically NO • from peroxynitrite only under conditions over pH 7 in a concentration-dependent manner and to prevent nitration of tyrosine residues on mitochondrial and cytosolic proteins [1]. At lower pH, aKG was not able to interact with ONOO − . Here, we presented, for the first time, that 5-HMF is not as good at reducing peroxynitrite directly compared to aKG, but nitration of tyrosine to nitro-tyrosine was better blocked by 5-HMF compared to aKG. Combining both substances may implicate these different interactions with ONOO − in a more effective way. However, at low concentrations (0-20 mM aKG + 5-HMF), no synergistic effect was obtained in the elimination of ONOO − . Even at higher dose (i.e., 200 mM aKG + 200 mM 5-HMF), a better synergistic effect to eliminate peroxynitrite was estimated compared to single compounds. Using the estimation of nitro-tyrosine as a sensor for excessive peroxynitrite, 5-HMF was 4-5-fold more effective at inhibiting nitration compared to aKG. Even the combined substances showed a better reduction at the highest used concentration compared to the single compounds indicating again the different mechanism of aKG and 5-HMF. The same synergistic effect was also estimated using other radical induced measurements such as cigarette smoke radicals on proteins generating carbonyl proteins [4]. In addition to more than 4000 radicals in cigarette smoke, there is peroxynitrite, which can oxidize and/or nitrate lipids, proteins and nucleic acids. In the same report, we have shown that a combination of aKG and 5-HMF effectively decreased cell growth and mitochondrial activity of Jurkat cell lines by their reduction in oxidative stress. Here, we present measurements of the single components for the first time. 5-HMF was able to inhibit the concentration-and incubation-time-dependent cell growth: after 72 h of incubation using 500 µg/mL 5-HMF, cell growth was significantly lower compared to the starting cell counts. Usage of aKG also reduced cell growth, but in a slightly lower manner. It is known that aKG is involved in the regulation of oxygen levels via the HIF alpha pathway as a substrate in addition to oxygen, vitamin C and ferric ions. Therefore, we estimated with aKG alone a reduction in the mitochondrial activity of Jurkat cells lines which correlated highly positive with cell growth in a linear function: the higher the aKG concentration the lower the cell growth and mitochondrial activity. We also found a similar correlation with 5-HMF, but the slope was lower compared to aKG. It was recently shown that 5-HMF can stabilize HIF alpha [22], and therefore use other pathways to inhibit cell growth and cytotoxic activity. These pathways were characterized by [23], which increased enzyme activities of SOD, catalase (CAT) and glutathione peroxidase (GPx) to protect cells from oxidative damage and to induce high antiproliferative activity on cancer cells.
The aKG was shown to induce apoptosis via caspase-9 and JNK mechanism by 7% and 12.1% using 5 and 200 mM [4], and we also observed apoptosis effects, which were not significantly different compared to the control using 0-3.4 mM solutions (3.4 mM = 500 µg/mL). Caspase-3 activity was not increased by all used aKG concentrations, indicating that the JNK and caspase pathways JNK and caspase-9 were used. 5-HMF showed a massive caspase-3 activity of 22% and 45% using 250 µg/mL and 500 µg/mL concentration, inducing apoptosis of Jurkat cells up to 10% and 35%, but we obtained the highest caspase-3 activity and apoptosis using the combined solution 500 µg/mL aKG + 166.7 µg/mL of 51% and 55%. NO • is involved in human leukemia [24] by a different pathway, e.g., apoptotic or nonapoptotic. Inducible NOS (iNOS) is regulated via transcriptional pathways and calcium and calmodulin. Under normal physiological conditions iNOS is not present on cells. During inflammation the expression of iNOS increase in the presence of cytokines, generated by macrophages. Hypoxia itself is a main promotor for the induction of iNOS, which is directly switched with the HIF-alpha pathway. It is known that under hypoxic conditions, the generation of radicals increases. Radicals, such as NO • , regulates cytostatic pathways against human malignant and nonmalignant cells [25], which remains unclear, but it seems that this might be concentration dependent. It was recently shown [26] that Jurkat cells exposed to NO • implicated a decrease in cardiolipin. Loss of cardiolipin massively affected the mitochondrial energy cycle process, which reduced the respiratory chain complex activities accompanied by a massive release of radicals, generated from dysregulated inner-mitochondrial enzymes such as AKGDH and/or low mitochondrial transmembrane potential. Activation of caspase-9 and caspase-3 follows after the release of cytochrome c into the cytosol. It seems that NO • plays a crucial role in the activation of apoptotic processes in human leukemic cells and a loss of NO • will help human leukemic cells to survive. A loss of NO can be speculated by the formation of ONOO − . aKG itself is able to recycle NO • from peroxynitrite, but at normal pH. Hypoxia leads to acidic conditions and aKG is not able to counteract with peroxynitrite. A loss of NO • in favor of ONOO − seems to increase the cell survival of leukemic cells. As a consequence of oxidative metabolism O 2 •− radical production is increased, e.g., by dysregulated iNOS and/or AKGDH, which can be easily converted by SOD to hydrogen peroxide and eliminated by CAT and/or GPx. Additional dysregulation of either SOD, CAT or GPx leaves more O 2 •− to counteract NO • , leading to the formation of peroxynitrite which itself oxidizes thiols and nitrate proteins on tyrosine residues. We suggest that a balanced combination of 5-HMF and aKG are potential antioxidants in lowering peroxynitrite formation directly by an enzymatic upregulation of SOD, CAT and GPx with 5-HMF and a reduction in peroxynitrite to nitrite and/or NO • in the presence of aKG. The aKG/5-HMF combination is able to inhibit nitration of tyrosine proteins synergistically, which leads to an increase in caspase activity and apoptosis in Jurkat cells in a concentration-dependent manner. This results into a decrease in cell growth in Jurkat cells as summarized in Figure 8 and Box 1. (2) alpha-ketoglutarate (mitochondrial or cytosolic) is removed by peroxynitrite under loss of molecular oxygen; (3) RONS react with proteins or amino acids forming increased CP or NP modulating proteins, such as the p53 and p21 pathways or SOD, CAT, GPx, NOS and HIFα, but also carbohydrates, nucleic acids and fatty acids; (4) HIF-alpha increase because of the loss of alphaketoglutarate or oxygen; (5) SOD, CAT and GPx is modified and or inactivated by RONS; (6) cell growth is initiated over HIFα; (7) apoptosis is reduced and Cell death decrease. Alpha-ketoglutarate added to leukemic cells: 1: Addition of alpha-ketoglutarate interferes with peroxides such as peroxynitrite; (2) recycling NO and accumulation of succinate; (3) succinate effects positively on HIFα, low cell growth, higher apoptosis and cell death; (4) reduction in RONS, such as ONOOor O 2 •− , implicate no modified proteins, amino acids, such as the p53 and p21 pathways, or SOD, CAT, GPx, NOS and HIFα, but also carbohydrates, nucleic acids and fatty acids and implicate again step 3. 5-HMF added to leukemic cells: 1: Addition of 5-HMF implicate saturation of molecular oxygen effecting positively HIFα. 2: Reduction in RONS, such as ONOO or O 2 •− implicate no modified proteins, amino acids, such as the p53 and p21 pathways or SOD, CAT, GPx, NOS and HIFα, but also carbohydrates, nucleic acids and fatty acids and implicate again step 3.

Preparation of ONOO −
ONOO − was prepared according to Hughes and Nicklin [27] by incipient mixing of equal volumes of 0.7 M H 2 O 2 solution in 0.6 M HCl and 0.6 M NaNO 2 on ice, followed immediately by termination of the reaction with 1.5 M NaOH. Surplus H 2 O 2 was removed by the addition of a pinch of MnO 2 and subsequent filtration of the suspension. The ONOOconcentration was determined spectrophotometrically at 302 nm with an extinction coefficient of 1670 M −1 cm −1 . Aliquots were stored at −70 • C until measurements. measured using the extinction of 450 nm in a spectrophotometric reader (Spectra Max Pro 384; Molecular Devices; San Jose, CA 95134, USA).

Caspase-3 Activity Measurements
Caspase-3-activated apoptosis was carried out as described in a recently published paper [4]. After incubation for 72 h in the absence or in the presence of 1.7 or 3.5 µM of aKG, 2 or 4 µM 5-HMF, 1.7 µg/mL aKG + 0.7 µM 5-HMF and 3.5 µM aKG + 1.3 µM 5-HMF, cells were separated by centrifugation at 3500× g. After several washing steps using cold PBS, 500 µL Cytofix-Cytoperm was added to the cells on ice at −20 • C. Antibody-FITC was used after incubation and the washing step was examination of the caspase-3 activity at 495/519 nm, respectively.

Statistical Analysis
t-Tests and ANOVA tests were used for comparing groups using SPSS 25 (SPSS Inc., Chicago, IL, USA). All values are given as the mean values and standard deviations. Statistical significance was considered to be at p < 0.05, with high significance at p < 0.01.

Conclusions
The combination of aKG and 5-HMF definitely reduces oxidative modifications, which are a hallmark in carcinogenesis and tumoral growth, including downregulation of antioxidative substances and antioxidative regulating enzymes of the Krebs cycle in the mitochondria but also in the cytosol. Both substances synergistically increase the antioxidative capacity, leading to a caspase-activated process, which eliminates Jurkat cells. The usage of this combination in solutions was effective in other clinical trials [10,12,13] without any increase in side effects; therefore, we suggest that the application of an IV solution containing both substances does have a potential role as a supporting therapy for leukemic patients in addition to the standard therapies used. Further clinical studies are therefore needed.