International Journal of Molecular Sciences Article Purine Nucleotide Alterations in Tumoral Cell Lines Maintained with Physiological Levels of Folic Acid Claudia Cano-Estrada 1,2 , Lidia de Benito-Gómez 3 , Paula Escudero-Ferruz 1,2 , Neus Ontiveros 1,2 , Daniel Iglesias-Serret 3,4 and José M. López 1,2, * 1 Institut de Neurociències, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Barcelona, Spain 2 Departament de Bioquímica i Biologia Molecular, Unitat de Bioquímica, Facultat de Medicina, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Barcelona, Spain 3 Departament de Ciències Fisiològiques, Facultat de Medicina i Ciències de la Salut, Universitat de Barcelona-IDIBELL (Institut d’Investigació Biomèdica de Bellvitge), 08907 L’Hospitalet de Llobregat, Barcelona, Spain 4 Facultat de Medicina, Universitat de Vic-Universitat Central de Catalunya (UVic-UCC), 08500 Vic, Barcelona, Spain * Correspondence: josemanuel.lopez@uab.cat Abstract: Most cancer cells have an increased synthesis of purine nucleotides to fulfil their en- hanced division rate. The de novo synthesis of purines requires folic acid in the form of N10 - formyltetrahydrofolate (10-formyl-THF). However, regular cell culture media contain very high, non-physiological concentrations of folic acid, which may have an impact on cell metabolism. Using cell culture media with physiological levels of folic acid (25 nM), we uncover purine alterations in sev- eral human cell lines. HEK293T, Jurkat, and A549 cells accumulate 50 -aminoimidazole-4-carboxamide ribonucleotide (ZMP), an intermediary of the de novo biosynthetic pathway, at physiological levels of folic acid, but not with the artificially high levels (2200 nM) present in regular media. Interestingly, HEK293T and Jurkat cells do not accumulate high levels of ZMP when AICAr, the precursor of ZMP, is added to medium containing 2200 nM folate; instead, ATP levels are increased, suggesting an Citation: Cano-Estrada, C.; de enhanced de novo synthesis. On the other hand, HeLa and EHEB cells do not accumulate ZMP at Benito-Gómez, L.; Escudero-Ferruz, physiological levels of folic acid, but they do accumulate in medium containing AICAr plus 2200 nM P.; Ontiveros, N.; Iglesias-Serret, D.; folate. Expression of SLC19A1, which encodes the reduced folate carrier (RFC), is increased in López, J.M. Purine Nucleotide HEK293T and Jurkat cells compared with HeLa and EHEB, and it is correlated with the total purine Alterations in Tumoral Cell Lines nucleotide content at high levels of folic acid or with ZMP accumulation at physiological levels of Maintained with Physiological Levels folic acid. In conclusion, tumoral cell lines show a heterogenous response to folate changes in the of Folic Acid. Int. J. Mol. Sci. 2023, 24, 12573. https://doi.org/10.3390/ media, some of them accumulating ZMP at physiological levels of folic acid. Further research is ijms241612573 needed to clarify the ZMP downstream targets and their impact on cell function. Academic Editors: Ronald Sluyter Keywords: purines; nucleotides; ZMP; AICAR; folic acid; SLC19A1; cancer and Rashid Giniatullin Received: 28 June 2023 Revised: 3 August 2023 Accepted: 4 August 2023 1. Introduction Published: 8 August 2023 The synthesis of nucleotides is essential for cells as a source of energetic and structural requirements. Both purines and pyrimidines are needed for cellular survival and prolifera- tion. There are two different pathways for the synthesis of purines in eukaryotic cells: the de novo and the salvage pathway [1]. Copyright: © 2023 by the authors. The de novo biosynthetic pathway has a very high energy demand. Just to synthesize Licensee MDPI, Basel, Switzerland. This article is an open access article one IMP, several molecules are needed: four ATP, two glutamines, two formats as 10-formyl- distributed under the terms and THF and one molecule of each glycine, aspartate, and carbon dioxide. Six different enzymes conditions of the Creative Commons catalyze the ten steps in this route to convert 5-phosphoribosyl-1-pyrophosphate (PRPP) Attribution (CC BY) license (https:// into IMP [1]. Interestingly, a network organization of these enzymes in the cytosol forms a creativecommons.org/licenses/by/ macromolecular complex, named purinosome, which is in close contact with the mitochon- 4.0/). dria [2]. As expected for an expensive metabolic pathway, purine synthesis is regulated at Int. J. Mol. Sci. 2023, 24, 12573. https://doi.org/10.3390/ijms241612573 https://www.mdpi.com/journal/ijms (PRPP) into IMP [1]. Interestingly, a network organization of these enzymes in the cytosol forms a macromolecular complex, named purinosome, which is in close contact with the Int. J. Mol. Sci. 2023, 24, 12573mitochondria [2]. As expected for an expensive metabolic pathway, purine synthesis is 2 of 18 regulated at the first step, in a reaction catalyzed by phosphoribosyl pyrophosphate ami- dotransferase (PPAT), through feedback inhibition by end products, purine ribonucleo- tides, obtained either the first step, inby the de novo a reaction or salvage catalyzed pathways [3]. The by phosphoribosyl salvage pathway, pyrophosphate which amidotransferase is energetically cheaper, is regulated by the enzymes hypoxanthine-guanine (PPAT), through feedback inhibition by end products, purine ribonucleotides, obtained phosphori- bosyltransferase either by the (HGPRT) de novo and adenine or salvage phosphoribosyltransferase pathways [3]. The salvage pathway,(APRT), which which form is energetically purinecheaper, nucleotides in one step reaction by using purine bases and PRPP as is regulated by the enzymes hypoxanthine-guanine phosphoribosyltransferase substrates (Fig- ure 1A). The complete deficiency of HGPRT activity causes Lesch–Nyhan (HGPRT) and adenine phosphoribosyltransferase (APRT), which form purine nucleotides disease (LND). Patients present in one stepsevere reaction neurological by using purineproblems bases such and PRPPas spasticity, cognitive as substrates (Figuredeficit, 1A). Thedysto- complete nia, and self-injurious behavior. At a biochemical level, LND patients have deficiency of HGPRT activity causes Lesch–Nyhan disease (LND). Patients present severe hyperuricemia because hypoxanthine neurological and guanine problems such ascannot be recycled, spasticity, and deficit, cognitive they aredystonia, degradedand to uric acid self-injurious instead [4,5]. Moreover, there is an increased acceleration of de novo behavior. At a biochemical level, LND patients have hyperuricemia because hypoxanthine purine synthesis, whichandtriesguanine to compensate cannot befor the deficient recycled, and theysalvage pathway,tofurther are degraded increasing uric acid the pro- instead [4,5]. Moreover, duction of uric there is anacid [6,7]. Recently, increased we have acceleration of dedescribed novo purine the synthesis, accumulation whichof 5-aminoimid- tries to compensate azole-4-carboxamide for the deficientribosidesalvage5′-monophosphate (ZMP, also pathway, further increasing named the AICAR), production an interme- of uric acid [6,7]. Re- diary of the de cently, wenovo have purine described biosynthetic pathway, the accumulation of in LND fibroblasts maintained with 5-aminoimidazole-4-carboxamide riboside 0 physiological levels of folic acid (FA) [8]. This alteration likely 5 -monophosphate (ZMP, also named AICAR), an intermediary of the de novo purineresults from the fact that conversion of ZMPpathway, biosynthetic to IMP byinthe LNDenzyme 5-aminoimidazole-4-carboxamide fibroblasts maintained with physiological ribonucleo- levels of folic tide formyltransferase/IMP acid (FA) [8]. This alteration cyclohydrolase (ATIC)from likely results is folate dependent the fact (Figure 1A). that conversion of ZMPA de-to IMP rivative byoftheZMP, the nucleoside enzyme AICAr, is present in the urine 5-aminoimidazole-4-carboxamide and the cerebrospinal ribonucleotide fluid formyltransferase/IMP cyclohydrolase of patients with LND,(ATIC) but not is folate dependent in control (Figuresuggesting individuals, 1A). A derivative of ZMP,alteration that a similar the nucleoside AICAr, may occur is present in vivo [8]. in the urine and the cerebrospinal fluid of patients with LND, but not in control individuals, suggesting that a similar alteration may occur in vivo [8]. (A) (B) Purine biosynthetic pathways 10-formyl-THF supply PRPP Folic acid PPAT 5-PRA SLC19A1 De novo GAR 10-formyl-THF pathway GART Purinosome FGAR Folic acid (THF) (mitochondrial ZTP attachment) Cytosolic Mitochondrial ZMP (AICAR) Folate cycle Folate cycle PRPP 10-formyl-THF GTP ATIC ATP FAICAR 5,10-meTHF GMP XMP IMP AS AMP 5,10-meTHF 10-formyl-THF Salvage PRPP HGPRT HGPRT PRPP APRT PYRIMIDINES formate PRPP pathway guanine hypoxanthine adenine (cytosolic) formate xanthine 10-formyl-THF uric acid PURINES Figure Figure 1. Schematic 1. Schematic diagrams diagrams of the metabolic of the metabolic pathways pathways studied.studied. (A) Scheme (A) Scheme of the of the de novo deand novo and salvagesalvage pathwayspathways ofbiosynthesis. of purine purine biosynthesis. Cancer Cancer cells cells can use canpathways both use both topathways maintainto maintain their high their purine high requirements. An increasedAn purine requirements. deincreased novo pathway with de novo a limited pathway availability with a limitedofavailability folic acid, in of the folic acid, form ofin10-formyl-THF (which mediates the form of 10-formyl-THF the transfer (which mediatesofthe 1C transfer units), produces accumulation of 1C units), produces of 5 -ami- accumulation of noimidazole-4-carboxamide ribonucleotide 50 -aminoimidazole-4-carboxamide (ZMP, also named ribonucleotide AICAR), (ZMP, also namedanAICAR), intermediary of the de of the an intermediary novo pathway. (B) Supply of 10-formyl-THF for the de novo purine biosynthesis. Folic acid is trans- de novo pathway. (B) Supply of 10-formyl-THF for the de novo purine biosynthesis. Folic acid is ported into the cells by the reduced folate carrier (RFC) SLC19A1, converted to tetrahydrofolate transported into the cells by the reduced folate carrier (RFC) SLC19A1, converted to tetrahydrofolate (THF), and subsequently to 10-formyl-TFH through multiple reactions taking place in the cytosol (THF), and subsequently to 10-formyl-TFH through multiple reactions taking place in the cytosol and/or the mitochondria. To simplify, the amino acids necessary for the de novo purine biosynthesis or in the folate-mediated one-carbon metabolism do not appear in the figures. Abbreviations not appearing in the text: 5-PRA: 5-phosphoribosylamine, 5,10-meTHF: 5,10-methylenetetrahydrofolate. Arrows indicate direct reactions and dotted arrows indicate that several reactions are involved. Int. J. Mol. Sci. 2023, 24, 12573 3 of 18 Folic acid, in the form of 10-formyl-tetrahydrofolate (THF), mediates the transfer of 1C units in de novo purine synthesis in two different reactions: the conversion of GAR to FGAR catalyzed by 50 -phosphoribosyl-glycinamide transformylase (GART) and the conversion of ZMP to FAICAR catalyzed by ATIC (Figure 1A). Although both GART and ATIC are folate dependent, ATIC seems to limit the rate of folic acid depletion [9] for reasons that are still not clear. To obtain 10-formyl-THF, multiple reactions are involved in creating a network known as folate-mediated one-carbon metabolism (FOCM), which is compartmentalized inside the mitochondria and the cytosol (Figure 1B) [10]. It has been proposed that the mitochondrial folic acid cycle is the main contributor of 1C units for nucleotide biosynthesis in cancer cells [11]. However, recently, Lee et al., have reported that human cell lines expressing low levels of the reduced folate carrier (RFC) SLC19A1 preferentially use the cytosolic versus mitochondrial one-carbon flux when the cells are maintained with physiological levels of FA [12]. Because tumoral cells usually have an accelerated de novo purine biosynthesis to support a high proliferation rate [13–17], we assessed whether cancer cells maintained with physiological levels of FA, which are not deficient in HGPRT, also accumulate ZMP. Here, we show that Jurkat cells mimic the purine alterations previously reported in LND fibroblasts [8]. Moreover, ZMP accumulation was also present in A549 and HEK293T cell lines, but not in HeLa or EHEB, thus indicating differential responses of the cells to physiological levels of FA. Interestingly, the cell lines that accumulate ZMP have higher levels of the reduced folate carrier (RFC) SLC19A1 and an efficient metabolization of AICAr, the precursor of ZMP, in regular media with high folate levels. 2. Results 2.1. Physiological Levels of FA Induce ZMP Accumulation in Jurkat Cells Jurkat is a human cell line established from the peripheral blood of a patient with acute T-cell leukemia [18]. ATCC recommends culturing these cells with RPMI medium at a concentration between 1 × 105 and 1 × 106 cells/mL, not allowing the cell density to exceed 3 × 106 cells/mL [19]. One of the most fundamental parameters that any healthy cell must maintain is a high ratio of ATP to ADP, as well as a high ratio of ADP to AMP [20]. In preliminary experiments, we found that Jurkat cells maintained with regular RPMI at a high density (2 × 106 or 1.5 × 106 cells/mL) presented higher levels of AMP than ATP, as well as higher levels of GMP than GTP. When the culture was maintained at a density of 5 × 105 cells/mL, the ATP/ADP/AMP and GTP/GDP/GMP ratios were well preserved, as expected for healthy cells (Supplementary Materials Figure S1). Therefore, all the experiments with Jurkat cells were done at low density (5 × 105 cells/mL or below) to preserve an optimal purine nucleotide ratio. Recently, we have explained that human fibroblasts from LND patients accumulate ZMP when the cells are cultured with RPMI containing physiological levels of FA (25 nM), instead of the nonphysiologically high levels present in regular medium (2200 nM) [8]. We addressed whether Z-metabolites accumulate in Jurkat cells maintained in these conditions. Cells were cultured during 5 days with RPMI medium supplemented either with 2200 nM FA or 25 nM FA, and perchloric acid (PCA) extracts were obtained to determine Z- metabolites using the Bratton-Marshall test. Significantly higher levels of Z-metabolites were detected in cells cultured with physiological levels of FA (25 nM) compared with 2200 nM FA (Figure 2A). However, the Bratton-Marshall test cannot distinguish between ZMP and other derivatives (AICA, AICAr, ZDP, ZTP) and could even detect other precursors in the de novo purine biosynthetic pathway, like AIR, CAIR or SAICAR [21]. Therefore, PCA extracts were analyzed using high-performance liquid chromatography (HPLC). As shown in Figure 2B, high levels of ZMP were detected in the cells maintained at 25 nM FA but were almost undetectable at 2200 nM FA. rimidine, but not purine, biosynthesis [22]. As shown in Figure 2B, Jurkat cells maintained in RPMI containing 350 µM UA and 25 nM FA (both physiological concentrations) pre sented similar levels of ZMP to cells maintained with only 25 nM FA. In conclusion, phys iological levels of folic acid induce ZMP accumulation in Jurkat cells independently of th Int. J. Mol. Sci. 2023, 24, 12573 presence of uric acid in the medium. Similar results were obtained when the 4 of data 18 wer expressed in µM (Supplementary Materials Figure S2). ✱ ✱✱ Figure Figure 2. Physiological 2. Physiological levelsofoffolic levels folicacid acid induce induce ZMP ZMPaccumulation accumulation in Jurkat cells. in Jurkat (A) (A) cells. Jurkat Jurkat cell cells were incubated for 5 days in RPMI medium containing 25 nM folic acid (FA), were incubated for 5 days in RPMI medium containing 25 nM folic acid (FA), 25 nM FA plus 350 µM 25 nM FA plus uric350 acid uric acid µM(UA), (UA),nM or 2200 or 2200 FA.nM FA.extracts Cell Cell extracts werewere obtained obtained with0.4 with 0.4 N N PCA PCAandandZ-metabolites Z-metabolites wer were determined by using the Bratton-Marshall test. The graphs determined by using the Bratton-Marshall test. The graphs represent the mean represent the mean ± SEM ± SEM of 6of independ 6 independent experiments, expressing the results as nmol Z-metabolites/mg ent experiments, expressing the results as nmol Z-metabolites/mg protein. * p < 0.05. One-wa protein. * p < 0.05. One-way ANOVA, ANOVA, uncorrected uncorrected Fisher’s Fisher’s LSD LSD test. (B) test. (B) Jurkat Jurkat cellscells were were treatedwith treated with different different media, media, as de as described in (A), and ZMP levels were determined by HPLC in cell scribed in (A), and ZMP levels were determined by HPLC in cell extracts. The graphs extracts. The graphs represent represent th the mean ± SEM of 6 independent experiments, expressing the results as nmol ZMP/mg protein. mean ± SEM of 6 independent experiments, expressing the results as nmol ZMP/mg protein. ** p ** p < 0.01. One-way ANOVA, uncorrected Fisher’s LSD test. 0.01. One-way ANOVA, uncorrected Fisher’s LSD test. Cantor et al.; reported that medium containing uric acid (UA) at a physiological 2.2.concentration Physiological(350 Levels µM)ofinhibits FA do Not Induce uridine AMPK Activation monophosphate in Jurkat synthase (UMPS) Cells thus altering pyrimidine, but not Intracellular purine, levels biosynthesis of the [22]. As shown most abundant purinein nucleotides Figure 2B, Jurkat werecells main- determined by tained in RPMI containing 350 µM UA and 25 nM FA (both physiological concentrations) HPLC (Table 1). When the results were expressed as nmol/mg protein, there were not sig presented similar levels of ZMP to cells maintained with only 25 nM FA. In conclusion, nificant differences physiological levels in the acid of folic levels of ATP, induce ADP, AMP,inGTP, ZMP accumulation JurkatGDP, and GMP inofthe cell cells independently treated the presence of uric acid in the medium. Similar results were obtained when levels with different media. There was a tendency for lower ATP the datainwere cells main tained with physiological expressed FA (25Materials in µM (Supplementary nM), that onlyS2). Figure reached statistical significance when th results were expressed as µM (Supplementary Materials Table S1). The AMP/ATP ratio 2.2. Physiological Levels of FA do Not Induce AMPK Activation in Jurkat Cells did not change with the different media (Figure 3A), as well as the adenylate energy Intracellular levels of the most abundant purine nucleotides were determined by HPLC charge (Figure 3B), which is defined as the ratio ([ATP] + 1/2 [ADP])/([ATP] + [ADP] (Table 1). When the results were expressed as nmol/mg protein, there were not significant [AMP]) by Atkinson differences andofWalton in the levels [23].AMP, ATP, ADP, AMPK GTP,was GDP,notandactivated in cells GMP in the Jurkat cellswith treated maintained with physiological different levelswas media. There of aFA or physiological tendency for lower ATPlevels of in levels FAcells plus UA (Figure maintained with3C). Th pAMPK/AMPK physiological FA ratio did not (25 nM), that change in the only reached cells maintained statistical significance with when different the resultsmedia, were when expressed analyzing as µMthe either (Supplementary cytoplasmicMaterials Table S1).(Figure or total extracts The AMP/ATP 3D). ratio did not change with the different media (Figure 3A), as well as the adenylate energy charge (Figure 3B), which is defined as the ratio ([ATP] + 1/2 [ADP])/([ATP] + [ADP] + [AMP]) by Atkinson and Walton [23]. AMPK was not activated in Jurkat cells maintained with physiological levels of FA or physiological levels of FA plus UA (Figure 3C). The pAMPK/AMPK ratio did not change in the cells maintained with different media, when analyzing either the cytoplasmic or total extracts (Figure 3D). Int. J. Mol. Sci. 2023, 24,Int. 12573 J. Mol. Sci. 2023, 24, x FOR PEER REVIEW 5 of 20 5 of 18 Table 1. Purine nucleotide levels of Jurkat cells incubated with different media. Table 1. Purine nucleotide levels of Jurkat cells incubated with different media. 25 nM FA + Condition 25 nM FA 2200 nM FA 25 UA 350 µM nM FA + Nucleotide Condition 25 nM FA 350 µM UA 2200 nM FA Mean ± SEM Mean ± SEM Mean ± SEM Nucleotide Mean ± SEM Mean ± SEM Mean ± SEM ATP 15.8 ± 4.6 10.6 ± 1.3 23.0 ± 5.9 ATP 15.8 ± 4.6 10.6 ± 1.3 23.0 ± 5.9 ADP 5.0 ± 2.2 1.8 ± 0.2 5.2 ± 0.7 ADP 5.0 ± 2.2 1.8 ± 0.2 5.2 ± 0.7 AMP 0.6 ± 0.2 0.3 ± 0.1 1.1 ± 0.3 AMP 0.6 ± 0.2 0.3 ± 0.1 1.1 ± 0.3 GTP 6.3 ± 1.6 5.3 ± 0.9 7.7 ± 1.9 GTP 6.3 ± 1.6 5.3 ± 0.9 7.7 ± 1.9 GDP GDP 1.2 ±1.2 0.4± 0.4 0.6 ± 0.6 0.1 ± 0.1 1.2 ± 0.3 1.2 ± 0.3 GMPGMP 0.5 ±0.5 0.2± 0.2 0.3 ± 0.3 0.1 ± 0.1 0.5 ± 0.2 0.5 ± 0.2 ZMPZMP 0.4± 1.2 ±1.2 ** 0.4 ** 0.2 ± 0.2 0.8 ± 0.8 0.1 ± 01 0.1 ± 01 Jurkat cells were incubated Jurkat for incubated cells were 5 days in for RPMI medium 5 days in RPMIcontaining 25 nM folic medium containing acidfolic 25 nM (FA), 25(FA), acid nM 25FAnMplus FA350 µM uric acid (UA),plus or 2200 350 µMnMuric FA.acid Cell extracts (UA), or 2200were obtained nM FA. with were Cell extracts 0.4 Nobtained PCA and purine with levels 0.4 N PCA anddetermined purine by HPLC. Resultslevels are the mean ±by determined SEM HPLC.of 6Results independent experiments are the mean ± SEM ofand are expressed 6 independent as nmol/mg experiments and areprotein. ATP > ADP > AMP expressed as nmol/mg and GTP > GDPprotein. > GMPATP > ADP ratios were> AMP and GTP conserved > GDPexperimental in every > GMP ratios were conserved condition in ZMP studied. every was significantly experimental increased condition in 25 nM studied. ZMP FA compared withwas 2200significantly nM FA. **increased p < 0.01.inOne-way 25 nM FAANOVA, compareduncorrected with 2200 nM FA. ** p < 0.01. One-way ANOVA, uncorrected Fisher’s LSD test. Fisher’s LSD test. AMP/ATP x 100 15 Percentage AMP/ATP Energy charge ratio 10 5 0 FA M FA FA M FA FA FA A A nM nM m nM m nM U U 35 25n 35 25n 25 25 00 00 22 22 0 0 + + Total Cytoplasmic pAMPK (63KDa) Actin (42 KDa) AMPK (63KDa) Actin (42 KDa) Figure 3. Physiological levels of folic acid do not induce AMPK activation in Jurkat cells. Jurkat cells were incubated for 5 days in Figure 3. Physiological RPMI levels medium of folic containing acid do not induce AMPK25activation nM folicinacidJurkat(FA), 25 nM cells. Jurkat FA plus cells were incubated 350 µM uric acid (UA), orfor2200 5 days nMin RPMI FA. medium containing Cell extracts 25 nM were folic acid with obtained (FA), 250.4 nMN FAPCA plus 350 andµMpurine uric acid (UA), or 2200 nM FA. Cell extracts were obtained with 0.4 N PCA and purine levels were levels were determined determined bybyHPLC. HPLC. (A)(A) AMP/ATP AMP/ATP ratio, expressed ratio, expressed as a percentage as a percentage and represented andas represented the mean as the mean ± SEM of 6 independent experiments. One-way ANOVA, uncorrected Fisher’s LSD test. (B) Cellular energy charge calculated as the ratio ([ATP] + 1/2 [ADP])/([ATP] + [ADP] + [AMP]) and represented as the mean ± SEM of 6 independent experiments. One-way ANOVA, uncorrected Fisher’s LSD test. (C) AMPK activity determined by Western blot. Jurkat cells were cultured in the above-mentioned conditions and whole-cell or cytoplasmic extracts obtained and analyzed by Western blot with antibodies against pAMPK (63 KDa), AMPK (63 kDa), and actin (42 KDa). (D) Quantification of AMPK activity. Bands obtained by Western blot were analyzed and quantified using the Image-J program. Each value obtained for pAMPK or AMPK was corrected by the corresponding levels of actin. The graph represents the mean ± SEM of 3 independent experiments. Total and cytoplasmic data were analyzed independently using one-way ANOVA, uncorrected Fisher’s LSD test. activity. Bands obtained by Western blot were analyzed and quantified using the Image-J program. Each value obtained for pAMPK or AMPK was corrected by the corresponding levels of actin. The graph represents the mean ± SEM of 3 independent experiments. Total and cytoplasmic data were analyzed independently using one-way ANOVA, uncorrected Fisher’s LSD test. Int. J. Mol. Sci. 2023, 24, 12573 6 of 18 2.3. Physiological Levels of FA Decrease Tetrazolium Dye Reduction by Jurkat Cells without Changes in Mitochondrial Membrane Potential Next, we measured the capacity 2.3. Physiological of the Levels of FA reduction Decrease of tetrazolium Tetrazolium Dye Reductiondyeby(MTT assay) Jurkat Cells by without Jurkat cells cultured Changesfor 5 days in different in Mitochondrial Membrane media. A significant decrease in tetrazolium Potential dye reduction wasNext,observed in cells the we measured maintained capacity ofwith physiological the reduction levels ofdye of tetrazolium FA, either (MTT assay) by with or without the presence Jurkat of UA, cells cultured for compared with cells 5 days in different maintained media. in medium A significant decreasecontain- in tetrazolium ing high levelsdye reduction of FA (Figurewas observed 4A). in cells maintained No significant differences with physiological were obtained in levels of FA, either with the doubling or (Figure time of the cells without 4B). the presence Moreover, of UA, compared Jurkat with cells maintained cells maintained with different in medium media containing did not present any change in mitochondrial membrane potential, measured as DilC signal by time high levels of FA (Figure 4A). No significant differences were obtained in the doubling of the cells (Figure 4B). Moreover, Jurkat cells maintained with different media did not cell cytometry (Figure 4C). As a positive control of mitochondrial depolarization, cells present any change in mitochondrial membrane potential, measured as DilC signal by cell were incubatedcytometry with the (Figure disrupter 4C).CCCP. Lee et control As a positive al.; haveof reported that Jurkat mitochondrial cells main- depolarization, cells were tained with physiological incubated with the disrupter CCCP. Lee et al.; have reported that Jurkat cells [12]. levels of FA have lower levels of mitochondrial NADPH maintained This could explain the decreased with physiological activity levels of FAdetected have lowerinlevels the MTT assay (see Discussion). of mitochondrial NADPH [12]. This could explain the decreased activity detected in the MTT assay (see Discussion). ✱ ✱ Tetrazolium dye reduction Doubling time (hours) (arbitrary units) FA M FA FA FA M FA FA 22 UA 22 UA nM nM m nM nM m 35 25n 35 25n 25 00 25 00 0 0 + + Physiological Figure 4.levels Figure 4. Physiological levels of folic acid of folic acid decrease decrease tetrazolium tetrazolium dye reductiondye reduction (MTT assay) (MTT byassay) Jurkat by Jurkat cells, without altering the mitochondrial membrane potential. (A) cells, without altering the mitochondrial membrane potential. (A) Jurkat cells were incubated for 5 Jurkat cells were incubated for 5 days in RPMI medium containing 25 nM FA, 25 nM days in RPMI medium containing 25 nM FA, 25 nM FA plus 350 µM UA, or 2200 nM FA and MTTFA plus 350 µM UA, or 2200 nM FA and MTT assay was assay was performed performedinasMaterial as described describedand in Material Methods. andThe Methods. graphThe graph represents represents the mean SEM ± SEM the± mean of 3 independentofexperiments, 3 independentusing experiments, 2200 nM using FA 2200 nM FA as acondition as a reference reference condition for normalization for normalization of the of the data and data and expressing the expressing the resultsunits results as arbitrary as arbitrary (AU). units * p < (AU). * p < 0.05. One-way 0.05. One-way ANOVA,ANOVA, uncorrecteduncorrected Fisher’s LSD test. (B) Jurkat cells were treated with different media, as described in (A), and dou-doubling Fisher’s LSD test. (B) Jurkat cells were treated with different media, as described in (A), and time was using bling time was obtained obtained theusing formulathe formula described described in Materials in Materials and Methods. and Methods. This This graphgraph represents repre- mean ± sents the mean ± SEM of 5 independent experiments. One-way ANOVA, uncorrected Fisher’s LSDLSD test. the SEM of 5 independent experiments. One-way ANOVA, uncorrected Fisher’s (C) Jurkat test. (C) Jurkat cells cellsin cultured cultured RPMIin RPMI medium medium with 25 with nM25FA nMorFA or 2200 2200 nMnM FAFA were were incubatedwith incubated with DilC or DilC or DilC plus DilC plus CCCP CCCP and signal and signal intensity intensity waswasquantified quantified by bycytometry. cytometry. Results represent Results the ± SEM the mean represent mean ± SEM of 3ofindependent 3 independent experiments. Unpaired t-test experiments.Unpaired t-test comparing comparing2200 2200nMnM vs.vs. 2525 nMnM FAFAwith or without with CCCP or without CCCP addition. addition. 2.4. ZMP Levels in Different Human Cell Lines Maintained with Physiological Folate Next, we addressed whether ZMP accumulation detected in Jurkat cells also occurred in other human cell lines. As shown in Figure 5A, significant levels of ZMP, expressed as nmol/mg protein, were detected in HEK293T and A549 cell lines maintained with RPMI containing 25 nM FA, but not in HeLa and EHEB cells. HEK293T accumulated higher levels of ZMP than Jurkat and A549 and, accordingly, ZTP was detected only in HEK293T cells maintained with physiological folate (Figure 5D). None of the cell lines analyzed showed accumulation of ZMP or ZTP when maintained with RPMI containing 2200 nM FA (Figure 5A,D). ATP levels were decreased in HEK293T cells maintained with physiological FA (Figure 5B). No significant differences were observed for other nucleotides comparing high FA condition with physiological FA within each cell line (Figure 5C,E,F,H,I). cells maintained with physiological folate (Figure 5D). None of the cell lines analyzed showed accumulation of ZMP or ZTP when maintained with RPMI containing 2200 nM FA (Figure 5A,D). ATP levels were decreased in HEK293T cells maintained with physio- logical FA (Figure 5B). No significant differences were observed for other nucleotides com- Int. J. Mol. Sci. 2023, 24, 12573 7 of 18 paring high FA condition with physiological FA within each cell line (Figure 5C,E,F,H,I). (A) (B) C) RPMI 2200 nM FA ✱✱ ✱ RPMI 25 nM FA nmol/ mg prot nmol/ mg prot ✱✱ (D) (E) (F) nmol/ mg prot nmol/ mg prot nmol/mg prot Doubling time ✱ ✱ ✱ (G) ✱ (H) (I) 90 80 Doubling time (hours) 70 60 nmol/ mg prot nmol/ mg prot 40 30 20 10 0 HEK293T Jurkat A549 HeLa EHEB Cell line ✱ (J) ✱ Figure 5. Physiological levels of folic acid induce ZMP accumulation in different tumoral cell lines. Figure 5. Physiological (A–F,H,I) levels HEK293T, Jurkat, of folic A549, HeLaacid and induce EHEB cells ZMPwereaccumulation incubated for 5 in different days in RPMI tumoral containingcell lines (A–F,H,I) 25 nM HEK293T, FA or 2200 nMJurkat, FA. CellA549, HeLa extracts and were EHEBwith obtained cells0.4were incubated N PCA for 5 days and nucleotides in RPMI determined by contain- ing 25 nM Results HPLC. FA or 2200were nM FA. Cell expressed extracts protein. as nmol/mg were obtained with The graphs 0.4 N the represent PCA and±nucleotides mean SEM of at deter- minedleast by3 HPLC. independent experiments, Results expressing were expressed the results protein. as nmol/mg as nmol nucleotide/mg p < mean protein. *the The graphs represent 0.05; ± SEM ** p < 0.01. Unpaired t-test comparing 2200 nM vs. 25 nM FA within each cell line. of at least 3 independent experiments, expressing the results as nmol nucleotide/mg protein. * p < (G) Doubling 0.05;time ** p of the cells < 0.01. was obtained Unpaired t-testusing the formula comparing 2200 described nM vs. in25Materials and Methods. nM FA within Unpaired each cell t-testDoubling line. (G) timecomparing of the cells 2200 nMobtained was vs. 25 nM using FA within theeach cell line. formula One-wayin described ANOVA, uncorrected Materials Fisher’s LSD and Methods. Unpaired t- test comparing each condition between different cell lines. * p < 0.05. (J) Total purine concentration, calculated as the summatory of adenylates plus guanylates, in each cell line cultured with RPMI containing 2200 nM or 25 nM FA. Results were expressed as nmol/mg protein. Differences between cell lines were analyzed with a one-way ANOVA, uncorrected Fisher’s LSD test. Differences within cell lines, comparing 2200 nM vs. 25 nM folic acid, were analyzed with an unpaired t-test. * p < 0.05. Int. J. Mol. Sci. 2023, 24, 12573 8 of 18 The doubling time of HEK293T, Jurkat, and A549 cells maintained with 2200 nM FA was similar (22–23 h), higher in HeLa (29 h), and very high in EHEB cells (63 h) (Figure 5G). The differences between EHEB and the other cell lines were statistically significant. The total content of purine nucleotides (adenylates plus guanylates) in cells maintained with 2200 nM FA was higher in HEK293T, Jurkat, and A549 than in HeLa and EHEB (Figure 5J, dotted bars). The difference between HEK293T and HeLa was statistically significant. Cell lines maintained with physiological levels of FA presented a similar content of purines (Figure 5J, solid bars). Note that HEK293T, Jurkat, and A549 cell lines decreased the total purine content at 25 nM FA compared to 2200 nM FA, being statistically significant in HEK293T (Figure 5J). In conclusion, physiological levels of FA induce ZMP accumulation in some tumoral cell lines (HEK293T, Jurkat, A549) but not in others (HeLa, EHEB). 2.5. Differential Effect of AICAr in Purine Nucleotide Levels of Tumoral Cell Lines ZMP could accumulate in cells if there is increased de novo purine biosynthesis and limited availability of folic acid. To assess whether the differential accumulation of ZMP between cell lines was due to a different flux of de novo purine biosynthetic pathway, cells were incubated with RPMI containing 2200 nM FA with or without 2 mM AICAr for 24 h. AICAr is a nucleoside that, after incorporation into the cells, is phosphorylated to ZMP (AICAR) by adenosine kinase [24]. Therefore, cells with an accelerated de novo purine synthesis and high levels of FA should metabolize ZMP into ATP and/or GTP. As shown in Figure 6A, HEK293T cells treated with AICAr did not accumulate ZMP and Jurkat cells only did so at very low levels, but both presented higher levels of ATP compared with non- treated cells (Figure 6B). Instead, HeLa and EHEB cells treated with AICAr accumulated high levels of ZMP (Figure 6A), without any significant change in the ATP pool (Figure 6B). ZTP was detected in the cells that accumulated high levels of ZMP (Figure 6D). A459 cells showed an intermediate response, accumulating higher levels of ZMP than Jurkat but lower than HeLa and EHEB (Figure 6A). In conclusion, ZMP accumulation in the cell lines treated with 2 mM AICAr plus 2200 nM FA (Figure 6A) was inversely correlated with ZMP accumulation at physiological levels of FA (Figure 5A). GTP, GDP, GMP, ADP and AMP levels did not change significantly with AICAr treatment (Figure 6C,E,F,H,I). AICAr treatment increased the total purine content of HEK293T, Jurkat and A549 cell lines, but reduced the purine levels of HeLa and EHEB (Figure 6G). AICAr treatment also decreased the cell growth, between 30 and 60% in all the cell lines (Supplementary Materials Figure S3). It is well known that AICAr inhibits cell cycle and cell growth in an AMPK-independent manner [25–27]. Both biological effects, the increase in purines and cell growth inhibition, can occur simultaneously in cells treated with AICAr [27]. HeLa and EHEB cells treated with AICAr plus 2200 nM folic acid accumulate ZMP, contrary to HEK293T cells. This could be explained by an increased uptake of AICAr and/or phosphorylation into ZMP by adenosine kinase in HeLa and EHEB cells. However, intracellular levels of AICAr (Supplementary Materials Figure S4) and cytosolic adenosine kinase expression (Supplementary Materials Figure S5) were similar in all the cell lines except Jurkat, which expressed lower levels of adenosine kinase and significantly higher levels of intracellular AICAr. Therefore, ZMP accumulation induced by AICAr treatment (Figure 6A) is not correlated with AICAr intracellular levels or adenosine kinase expression. Altogether, the previous results suggest that HEK293T cells preferentially use the de novo purine biosynthetic pathway to fulfil their purine nucleotide requirements, as opposed to HeLa or EHEB. However, we should be cautious with this conclusion because we have not measured nucleotidases or purine recycling in the different cell lines, and this is certainly a limitation of our work. More studies are necessary to know the relative fluxes of the de novo and salvage pathways, as well as the nucleotidase activity, in the different cell lines. Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW 9 of 20 Int. J. Mol. Sci. 2023, 24, 12573 9 of 18 ✱✱ ✱ ✱ ✱ ✱✱✱✱ ✱ ✱✱✱ Variation in total purine content induced by AICAr ✱✱ ✱ ✱ 100 ✱ 50 0 -50 -100 3T t 9 a EB ka eL 54 29 EH r A H Ju EK H Cell line Figure 6. Purine alterations in different cell lines treated with 2 mM AICAr. HEK293T, Jurkat, A549, Figure 6. Purine alterations in different cell lines treated with 2 mM AICAr. HEK293T, Jurkat, A549, HeLa and EHEB cells were maintained for 4 days in RPMI containing 2200 nM FA and then incubated HeLa and EHEB cells were maintained for 4 days in RPMI containing 2200 nM FA and then incu- withbated same the medium with the with or with same medium without 2 mM AICAr or without 2 mM AICArfor 24for h. 24 (A–F,H,I) Cell Cell h. (A–F,H,I) extracts were extracts obtained were withobtained 0.4 N PCAwithand 0.4 purine N PCAlevels determined and purine levels by HPLC. Results determined by HPLC.wereResults expressed wereasexpressed nmol/mgasprotein. Thenmol/mg protein. The graphs represent thegraphs meanrepresent ± SEM of theatmean least±3SEM of at least 3experiments, independent independent expressing experiments,theex-results pressing as nmol the results as nmol nucleotide/mg nucleotide/mg protein. * p < 0.05;protein. * p < 0.05; ** p < 0.01; *** p**