Mass Isotopomer Analysis of Nucleosides Isolated from RNA and DNA

Dec 2, 2014 - Department of Chemistry, Faculty of Sciences and Technology, University of Coimbra, Coimbra 3004-535, Portugal. •S Supporting Informat...
0 downloads 0 Views 577KB Size
Download PDF
Article pubs.acs.org/ac

Mass Isotopomer Analysis of Nucleosides Isolated from RNA and DNA Using GC/MS Ines Miranda-Santos,⊥,‡,§ Silvia Gramacho,∥ Marta Pineiro,∥ Karla Martinez-Gomez,§,† Michel Fritz,§ Klaus Hollemeyer,§ Armindo Salvador,‡ and Elmar Heinzle*,§ ⊥

Department of Life Sciences, Faculty of Science and Technology, University of Coimbra, Coimbra 3000-456, Portugal CNC − Center for Neuroscience and Cell Biology, University of Coimbra, Coimbra 3004-504, Portugal § Biochemical Engineering Institute, University of Saarland, Saarbrücken, Saarland 66123, Germany ∥ Department of Chemistry, Faculty of Sciences and Technology, University of Coimbra, Coimbra 3004-535, Portugal ‡

S Supporting Information *

ABSTRACT: Nucleosides are biosynthesized from metabolites that are at key nodes of intermediary metabolism. Therefore, 13 C labeling patterns in nucleosides from ribonucleic acid (RNA) and deoxyribonucleic acid (DNA) in suitably designed isotopic tracer studies provide information on metabolic flux distributions of proliferating cells. Here, we present a gas chromatography (GC)-mass spectrometry (MS)based approach that permits one to exploit that potential. In order to elucidate positional isotopomers of nucleosides from RNA and DNA, we screened the fragmentation spectra of their trimethylsilyl derivatives. We identified the molecular ion moieties retained in the respective fragment ions, focusing particularly on the carbon backbone. Nucleosides fragmented at the N-glycosidic bond provide nucleobase and/or ribose or 2′deoxyribose fragment ions and fragments thereof. Nucleoside fragments composed of the nucleobase plus some carbons of the ribose ring were also observed. In total, we unequivocally assigned 31 fragments. The mass-isotopic distribution of the assigned fragments provides valuable information for later 13C metabolic flux analysis as indicated by a labeling experiment applying [1-13C]glucose in a yeast culture. etabolic studies, particularly metabolic flux analysis (MFA), rely on the selective enrichment of metabolites in stable isotopes, e.g., 13C, derived from supplied labeled substrates.1 Metabolites, and therefore their biosynthetic products, present patterns of labeling, isotopomers, in accordance with the particular metabolic fluxes that produced them. Amino acids, either free2 or derived from proteins,3 have been used extensively in metabolic studies based on tracers and mass spectrometry (MS),4,5 but nucleosides incorporated in ribonucleic acid (RNA) and deoxyribonucleic acid (DNA) have not. The following considerations led us to hypothesize that the assessment of the 13C distribution in nucleosides permits the estimation of metabolic fluxes of proliferating cells. The carbon backbone of pyrimidine nucleosides are derived from aspartate and carbamoyl phosphate, whereas those of purine nucleosides are derived from glycine, HCO3−, and N10-formyltetrahydrofolate. In both, the carbon backbone of the ribose moiety is derived from ribose-5-phosphate (Supporting Information, Figure S-1). Aspartate exchanges with oxaloacetate, which in turn is an intermediary of the tricarboxylic acid (TCA) cycle. Carbamoyl phosphate originates from HCO3−. Glycine and N10-formyltetrahydrofolate originate from serine, C1−C2, and C3, respectively. Ribose-5-phosphate is an intermediary of the

M

© XXXX American Chemical Society

pentose phosphate pathway (PPP).6 Enrichment in 15N may also be explored providing information about nitrogen metabolism related to the synthesis of carbamoyl phosphate used for the synthesis of pyrimidines, of glycine and glutamine used for purines, and of aspartate used for both types of nucleobases. In proliferating cells, nucleosides incorporated in RNA and DNA mainly originate from de novo biosynthesis. Hence, the flux distribution of proliferating cells fed with labeled substrates becomes imprinted in the nucleosides incorporated in nucleic acids. Labeling patterns in nucleosides of RNA and DNA are thus relevant for studies of the metabolism of cancer, embryonic and stem cells, and budding yeast or other microorganisms used in metabolic engineering but also infected cells (e.g., lymphocytes7 and erythrocytes8). The accuracy and precision of metabolic flux determination by MFA based on 13C tracer distribution improve with increasing resolution of the labeling patterns.4,9−11 MS only provides information about the total number of 13C atoms in the molecule−mass isotopomers. Fragment ions retain the Received: September 3, 2014 Accepted: December 2, 2014

A

dx.doi.org/10.1021/ac503305w | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

respective 13C enrichment providing additional information. By exploring the fragmentation spectrum, the fragments can be assigned. Methods have been developed to elucidate the carbon composition of the fragments11,12 and to estimate fluxes13,14 using respective mass isotopomer distributions (MID) of each fragment. This approach has been exploited for amino acids2,9,15 but not hitherto for nucleosides. Here, we present (i) an optimized chromatographic method (gas chromatography, GC) for the identification and separation of nucleosides and (ii) an exhaustive assignment of carbon composition of fragments of nucleosides generated in the GC/ MS conditions used. We discuss the information that can be obtained from mass isotopomer analysis of fragments of nucleosides derived from RNA and DNA of proliferating cells, and we present an illustrative application to yeast batch cultures fed with [1-13C]glucose.

min. Supernatant was discarded; 1 mL of 70% ethanol was added and vortexed briefly to resuspend the pellet. After centrifugation at 16 000g, 4 °C for 5 min, supernatant was discarded, and the RNA pellet was dried and resuspended in 100 μL of RNase free water. RNA was analyzed on 1% agarose gel for integrity. RNA was quantified by absorbance at 260 nm using a Nanodrop 2000c device (Thermo Scientific, USA). The 260 nm/280 nm and 260 nm/230 nm ratios were examined for protein and solvent contamination. RNA samples were stored at −70 °C. DNA. Total DNA was isolated and purified following the protocol reported by Holm et al.17 Samples containing 7 × 109 cells (approximately 100 mL of culture, OD = 3.5) were washed with 5 mL of ice cold ultrapure water and spun down at 1500g, 4 °C for 5 min. The supernatant was discarded, and the pellet was resuspended in 2.1 mL of SCE buffer (1 M sorbitol, 0.1 M sodium acetate, 60 mM EDTA, pH 7). Cell suspension was treated with 150 μL of Zymolase solution (2000 U/mL Zymolyase 20T, 10% 2-mercaptoethanol in SCE buffer) and incubated overnight at 37 °C to promote spheroplast formation. The spheroplasts were spun down at 1500g for 1 min, and the resulting pellets were well drained. The pellet was slowly resuspended in 2.1 mL of a guanidine-HCl (GuHCl) solution (4.5 M GuHCl, 0.1 M EDTA, 0.15 M NaCl, 0.05% sodium lauryl sarcosinate, pH 8) and then incubated at 65 °C for 10 min with occasional swirling to lyse spheroplasts. After cooling to room temperature, 2.1 mL of cold ethanol was added, and the mixture was centrifuged at 16 000g for 5 min. The supernatant was discarded, and the pellet was then well drained. The pellet was initially slowly resuspended in 4.2 mL of TE buffer, 10× concentrated, and then vortexed for 2 min. RNA was removed by applying 2800 U of RNase A and 108 kU of RNase T1 at 37 °C for 1 h. This incubation was followed by incubation with 12 U of proteinase K at 37 °C for 30 min. Samples were then extracted twice with 1.7 volumes of phenol−chloroform−isoamyl-alcohol (P−C−I), 25:24:1, pH 8, pre-equilibrated with 0.5 M Tris-HCl pH 8. In detail, 7 mL of P−C−I was added, and the solution was vortexed for 1 min and centrifuged at 1600g for 5 min. After the second extraction, DNA was precipitated by adding sodium acetate, pH = 8, to a final concentration of 0.3 M, and by adding 2.5 volumes of ethanol. Samples were mixed by inversion and incubated at 70 °C for a minimum of 15 min. The DNA was spun down at 16 000g for 30 min, and supernatant was discarded. The DNA pellet was resuspended in ethanol 70% and spun down again at 16 000g for 30 min. The supernatant was discarded; the DNA pellet was well dried and resuspended in 100 μL of ultrapure water. DNA was analyzed on a 1% agarose gel for integrity and purity. When some vestigial amounts of RNA were still present, samples were reincubated with RNases A and T1 in TE buffer 10× concentrated, 37 °C overnight. After incubation, DNA was precipitated with sodium acetate and ethanol as specified above. DNA was quantified by absorbance at 260 nm using a Nanodrop 2000c device (Thermo Scientific, USA). The 260 nm/280 nm and 260 nm/230 nm ratios were examined for protein and solvent contamination. DNA samples were stored at −20 °C. Nucleic Acid Digestion. Nucleic acid digestion was modified starting from a previous protocol.18 Briefly, 100 μg of nucleic acids at a final concentration of 0.2 μg/μL was incubated overnight at 37 °C with 0.5 U/mL benzonase, 0.3 mU/mL phosphodiesterase, and 0.4 U/mL alkaline phosphatase in the digestion buffer composed of 10 mM NH4HCO3, 50

■

EXPERIMENTAL SECTION Chemicals. Unless otherwise stated, chemicals were purchase from Sigma-Aldrich (St. Louis, MO, USA). [U-13C]glucose and [1-13C]glucose 99.0% purity were purchased from Cambridge Isotope Laboratories (Andover, MA, USA). RNase free pipet tips were purchased from Axigen, Corning Incorporated Life Sciences (Tewksbury, MA, USA). NMethyl-N-(trimethylsilyl) trifluoroacetamide (MSTFA) was purchased from Macherey-Nagel GmbH & Co (Düren, Germany). Zymolyase was purchased from Seigaku Corp., Japan. RNase A and RNase T1 were purchased from ThermoScientific (Fremont, CA, USA). Organism and Growth Conditions. Cells of S. cerevisiae strain CEN.PK113-7D were grown and harvested as described in detail in the Supporting Information. Nucleic Acids Extraction. RNA. Total RNA was isolated and purified following the basic protocol from Collart and Oliviero16 with some modifications. RNA extractions were handled in the sterile bench. Pipette tips were RNase free. Microcentrifuge tubes and flasks were autoclaved. All the solutions were prepared in RNase free water. RNase free water was prepared by incubating ultrapure water with diethylpyrocarbonate (DEPC) to a final concentration of 0.1% at 37 °C overnight and followed by autoclaving to remove DEPC. Samples containing 100 × 106 cells (approximately 1.5 mL of culture with optical density (OD) = 3.5) were washed with 1 mL of ice-cold RNase free water and spun down at 1500g, 4 °C for 5 min, and the supernatant was discarded. To disrupt cells, the cell pellet was resuspended in 400 μL of TES buffer (10 mM Tris-HCl, pH 7.5, 10 mM EDTA, 0.5% SDS). Immediately, 400 μL of acid phenol/chloroform 5:1, pH 5.1, preheated to 65 °C, was added and vigorously vortexed for 10 s to promote protein denaturation and prevent RNA degradation. Extensive cell lysis and protein denaturation was achieved by incubation at 65 °C for 60 min and vortexing for 10 s every 10 min. After cooling down on ice for 5 min, samples were centrifuged at 16 000g, 4 °C for 5 min. The aqueous phase was then extracted twice with an equal volume of chloroform. The aqueous upper phase was transferred to a clean microcentrifuge tube, and an equal volume of chloroform was added. Samples were mixed by vortexing for 20 s and centrifuged at 16 000g, 4 °C for 5 min. After the second extraction, RNA was precipitated by adding sodium acetate, pH 5.3, to a final concentration of 0.3 M, and by adding 2.5 volumes of ethanol. Samples were mixed by inversion, incubated at −20 °C for a minimum of 30 min, and centrifuged at 16 000g, 4 °C for 15 B

dx.doi.org/10.1021/ac503305w | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Figure 1. GC/MS chromatogram and spectra at the respective retention times of trimethylsilyl-nucleosides. The spectrum corresponding to each numbered chromatographic signal is shown in the respective panels (1−4); the solid boxes (□) in the spectra (1−4) identify the molecular ions corresponding to the parent molecules whose structures are embedded and presented in detail in the Supporting Information Chart S-1 by the respective structure number. The m/z values in bold note the assigned fragments generated from the respective parent ion; fragment structures are depicted underneath the respective spectrum and shown in detail in the Supporting Information Chart S-2 identified by the respective number between brackets; spectrum expansions at the assigned fragment regions are show in the Supporting Information Figure S-3. Nucleoside-(TMS)n identified in the chromatogram by (1−4) was prepared from standard compounds at the following concentrations: [uridine] = 0.08 mM, [adenosine] = 0.06 mM, [cytidine] = 0.36 mM, and [guanosine] = 0.08 mM.

mM NaCl, and 10 mM MgCl2, pH 7.9. Benzonase had to be subjected to buffer substitution as it was provided by the manufacturer in a solution containing glycerol; after microfiltration with a 3 kDa cutoff membrane, benzonase was recovered in 20 mM NH4HCO3, 20 mM NaCl, and 2 mM MgCl2, pH 8. Phosphodiesterase I was prepared from powder in 100 mM NH4HCO3, pH 8.9, and alkaline phosphatase in 10 mM NH4HCO3, 0.1 mM ZnCl2, and 1 mM MgCl2, pH 8. NH4HCO3 buffer was always prepared prior to digestion and with fresh ultrapure water. After digestion, a microfiltration with a cutoff membrane of 3 kDa was performed, and the filtrate was freeze-dried. Digestion works in any aqueous preparation of nucleic acids, i.e., irrespective of the method of extraction used. Derivatization of Nucleosides. Nucleosides, either standards or digested from nucleic acids, were derivatized with MSTFA (N-methyl-N-(trimethylsilyl)trifluoroacetamide).19

The derivatization was promoted in pyridine/MSTFA (50:50) at 80 °C for 30 min. In detail, standards of nucleosides were solubilized separately in pyridine/dimethylformamide (50:50). Guanosine and 2′-deoxyguanosine were incubated at 80 °C for 5−10 min, with occasional vortexing, until complete solubilization. Samples were diluted with an equal volume of MSTFA. Nucleosides from RNA and DNA digestion were extracted from the freeze-dried powder with 200 μL of pyridine/DMF (50:50). After incubation at 80 °C for 10 min with occasional vortexing, samples were centrifuged for 10 min at 16 000g. The supernatant was collected and dried under vacuum. Samples were first solubilized in pyridine, and an equal volume of MSTFA was then added. The volume was adjusted to obtain a final concentration of 0.8 mM in cytidine (RNA) and 0.2 mM in thymidine (DNA), keeping a 50:50 pyridine/MSTFA ratio. Derivatization was promoted as described above. C

dx.doi.org/10.1021/ac503305w | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Figure 2. GC/MS chromatogram and spectra at the respective retention times of trimethylsilyl derivatives 2′-deoxynucleosides. The spectrum corresponding to each numbered chromatographic signal is shown in the respective panels (5−9); the solid boxes (□) in the spectra (5−9) identify the molecular ions corresponding to the parent molecules whose structures are embedded and presented in detail in the Supporting Information Chart S-1 by the respective structure number. The m/z values in bold note the assigned fragments generated from the respective parent ion; fragment structures are depicted underneath the respective spectrum and shown in detail in the Supporting Information Chart S-3 identified by the respective number between brackets; spectrum expansions at the assigned fragment regions are shown in the Supporting Information Figure S-4. Ions identified in the chromatogram by (5−9) were prepared from standard 2′-deoxynucleosides at the following concentrations: [thymidine] = 0.16 mM, [2′-deoxycitidine] = 0.08 mM, [2′-deoxyadenosine] = 0.06 mM, and [2′-deoxyguanosine] = 0.06 mM.

Gas Chromatography/Mass Spectrometry. GC/MS measurements were carried out on a HP6890 GC system using a Mass Selective Detector 5973 (Agilent Technologies, Waldbronn, Germany). We used a HP5MS UI capillary column (5%-phenyl)-methylpolysiloxane Ultra Inert (60 m × 250 μm × 0.25 μm, Agilent Technologies, Waldbronn, Germany). For electron impact ion generation, a 70 eV electron beam was used followed by mass analysis using a quadrupole. The optimized conditions for the measurement of nucleosides were as follows. One μL of sample was injected splitless for 2 min using a PTV with an initial temperature of 130 °C and a temperature gradient of 12 °C/s, up to 320 °C. Carrier gas flow was helium at 1.1 mL/min. The temperature gradient for the separation of nucleosides was 130 °C held for 1 min and 10 °C/min up to 325 °C held for 4 min. Temperatures of ion source and transfer liner were 230 and 320 °C. Full scan (SCAN) ranging from 50

to 700 Da using a scan rate of 9 scans/s and selective ion monitoring (SIM) were carried out with optimized settings for each analyte.

■

RESULTS Extraction of DNA and RNA from Cells and Digestion. Cells of S. cerevisiae were cultivated as described and harvested in the mid exponential phase at OD = 3.5. For all the extractions both of RNA and DNA, the values of the 260 nm/ 280 nm ratio were in the range of 1.8−2.0 and the values of the 260 nm/230 nm ratio ranged between 2.0 and 2.3. On an average, we obtained 180 μg of RNA per each extraction of 100 × 106 cells, what represents approximately 4.5% of the biomass. On an average, we extracted 130 μg of DNA per each extraction of 7 × 109 cells, corresponding to 0.075% of the biomass. We consistently obtained high yields of pure material. This D

dx.doi.org/10.1021/ac503305w | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

mneutral moiety). Knowing that, under the conditions applied, fragment ions carry only one charge allowed us to exclude the 28 out of 102 plausible fragment ions that did not fulfill this difference. In the second criterion, we intended to verify the molecular formula of the plausible fragment ions that passed the first criterion. SIM mode measurement of the remaining 74 selected fragment ions allowed the exclusion of a further 37 fragment ions because of mismatch of their isotopic distribution to the known natural one (Supporting Information Tables S-2 and S-3). A mismatch of larger than 15% led to exclusion. Using [U-13C]nucleosides and [U-13C]2′-deoxynucleosides, the number of carbons in the predicted fragments was assessed after cultivation of yeast on [U-13C]glucose. It was possible to identify the four derivatives of nucleosides and the five derivatives of 2′-deoxynucleosides. Among all the extracts, we selected those presenting chromatographic signals of nucleosides with intensities higher than 1 × 106. The fragments whose m/z of the most intense mass isotopomer did not shift to (m + n)/z (where n is the total number of carbons in the carbon backbone of the fragment) were discarded. The fragments that, besides (m + n)/z, presented another high intensity mass isotopomer, (m + x)/z (where x is any number of carbons in the carbon backbone), were also discarded because it may be two different fragment ions with the same mass but a different total number of carbons. The 31 fragments that passed the three above-mentioned criteria are shown in the expansion A− D in panels 1−9 of Supporting Information Figures S-3 and S-4 for nucleosides, respectively. The proposed structures for each fragment of nucleosides are presented in the Supporting Information Charts S-2 and S-3, respectively. Biological Application. Nucleosides of RNA and DNA isolated from cells grown on [1-13C]glucose were derivatized and measured in GC/MS in scan and SIM mode for the assigned fragments. We subtracted the contributions of the natural isotopic abundance of all atoms other than those belonging to the carbon backbones of interest.21 The relative intensity of the mass isotopomers m + 1, m + 2, and m + 3 was higher than that of the corresponding naturally labeled ones for all the molecular ions and the assigned fragments (Supporting Information Figure S-5). In the ions of the whole nucleosides ((1.a), (3.a), (2.a), (4.a), (8), (9)), the most abundant mass isotopomer was m + 1/z, and the second and third most abundant were m + 2 and m + 0 in the case of purine nucleosides and m + 0 and m + 2 in nucleosides of pyrimidine. A closer examination of the Supporting Information Figure S-5 results in the following remarkable observations: (i) purine nucleosides are highly labeled in one or two carbons, whereas pyrimidine nucleosides are predominantly labeled in one carbon. (ii) m + 0 of (1.e) and of (2.d) are more abundant than m + 0 of the respective molecular ions, showing that (1) and (2) are predominantly labeled in the nucleobase; however, because m + 1 of (1.e) and of (2.d) are more abundant than in naturally labeled compounds, the ribose moiety is also labeled. (iii) m + 1 and m + 2 of (3.c) and of (4.e) and m + 1 and m + 2 of (3.d) and (4.f) are less abundant than those of their molecular ions, suggesting an even distribution of 13C labeling among nucleobase and ribose moiety in (3) and (4). (iv) In (5.a), m + 0 is the most abundant mass isotopomer, whereas in (6) it is m + 1. Abundances of m + 2 in (6) and m + 1 in (7.a) are approximately the same as their respective m + 0 showing that the labeling is predominantly in the nucleobase moiety. (v) m + 1 and m + 2 of (8.b) and of (9.b) are less abundant than m + 1 and m + 2 of (8) and (9), and m + 2 of (8.b) and (9.b) are

procedure allowed us to extract in a single step the number of cells required to obtain the amount of nucleic acids needed for measurement. Common commercial kits use a smaller number of cells what would require operating multiple parallel extractions to obtain the same amount of material. The digestion, performed as described above, had a yield above 50%, resulting in a minimum of 50 nmol of each nucleoside and 2′deoxynucleoside per digestion. Detection and Identification of Nucleosides and 2′Deoxynuclosides. For the identification and assignment of the derivatized compounds in the GC/MS, a range of concentrations (40−800 μM) of each pure naturally labeled nucleoside and 2′-deoxynucleoside was independently measured in scan mode. The spectrum at the respective retention time (RT) was examined, and the presence of the molecular ion, identified by the corresponding m/z, confirmed the assignment of the compound. In the Supporting Information Table S-1, we present the relative RT and m/z values for all nucleosides and purine 2′-deoxynucleosides and for the two species observed for the pyrimidine 2′-deoxynucleosides. The applied standard mixture contained 80 μM uridine and guanosine, 60 μM adenosine, 720 μM cytidine, 160 μM thymidine, 80 μM 2′-deoxycytidine, and 60 μM 2′-deoxyadenosine and 2′-deoxyguanosine. The chromatograms and spectra of nucleosides are presented in GC panels of Figures 1 and 2, respectively. The 2′-deoxynucleoside derivatization products are shown in the Supporting Information Chart S-1. The derivatization process resulted in complete trimethylsilylation of all the purine and pyrimidine nucleosides as well as of the purine 2′deoxynucleosides. In the derivatization of 2′-deoxythymidine and 2′-deoxycytidine, pyridine promoted the cleavage of the Nglycosidic bond. Subsequently, the nucleobases reacted with MSTFA, resulting in the respective N-trimethylsilyl derivative of thymine and cytosine. In the Supporting Information Scheme S-1, we present a mechanistic proposal20 for the above-mentioned cleavage of the N-glycosidic bond. In the basic medium used for derivatization, deoxyribose and the pyrimidine ring suffered base catalyzed elimination. The results are two different trimethylsilylated species in the chromatogram, trimethylsilyldeoxyribose (5), and trimethylsilylpyrimidines (6 and 7). The molecular ion is observable for all the derivatized products, although at low intensity. Except in the case of trimethylsilylcytosine, the fragment resulting from the loss of a methyl group of the trimethylsilyl substituent is more intense than the molecular ion and can be used for the identification. Mass Spectra Analysis and Fragment Assignment. To assign carbon composition of fragments, three criteria were used: (i) matching of fragment ions to plausible structures, (ii) mismatch between the experimental and the theoretical natural isotopic distribution of each proposed fragment ion, and (iii) the shift to (m + n)/z (n equals the number of carbons) of the most intense mass isotopomer of the fragment ions in [U-13C]nucleosides and [U-13C]2′-deoxynucleosides. The spectra at the corresponding RT of each compound were analyzed in detail, seeking for ions that could result from fragmentation of the molecular ion. The complete spectra of the derivatized nucleosides are presented in Figure 1, panels 1− 4, and Figure 2, panels 5−9, respectively. Following the classical rules of fragmentation, we calculated the m/z difference between the molecular ion and every plausible fragment ion (m molecular ion − m fragment ion = E

dx.doi.org/10.1021/ac503305w | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

less abundant than their respective m + 0, showing that also in (8) and (9) labeling is present in both moieties but prevalent in the nucleobase. (vi) m + 1 of (1.f), (2.e), (3.e), (4.g), and (5.b) and m + 1 and m + 2 of (2.b) and (4.b) are less abundant than m + 1 of (1.e), (2.d), (3.d), (4.f), and (5.a) and m + 1 and m + 2 of (2.a) and (4.a), which means that C5′ is labeled. (vii) m + 1 and m + 2 of (2.c) and of (4.c) are more abundant than m + 1 and m + 2 of (2.b) and of (4.b), suggesting that there is a predominance of labeling in either C2′ or C1′. (viii) m + 1 and m + 2 of (4.e) are much less abundant than m + 1 and m + 2 of (4.d) and of (4.c), and m + 1 and m + 2 of (4.d) is slightly lower than that of (4.c), showing that either C1′ or C2′ are extensively labeled and perhaps a slight predominance in C1′. (ix) m + 1 and m + 2 of (1.d) and of (1.b) are less abundant than m + 1 and m + 2 of (1.c) suggesting that, unlike purines, in nucleosides of uridine C1′ and C2′ should not be labeled. (x) m + 0 of (2.b) is more abundant than m + 0 of (2.c), showing that C3′ of (2) should not be labeled.

spectra and of the isotopic abundances of naturally and of uniformly 13C labeled nucleosides. In uridine, adenosine, and guanosine, we observed fragments corresponding to the loss of the single carbon, C5′ (Figure 1 and Supporting Information Chart S-2 (1.c), (2.b), and (4.b)). In all the nucleosides, we observed fragments corresponding to the fragmentation at the ribose ring keeping the N-glycosidic bond intact (Figure 1 and Supporting Information Chart S-2 (1.b), (1.d), (2.c), (3.b), (4.c), and (4.d)). The fragmentation at the N-glycosidic bond is common to nucleosides and to the purine 2′-deoxynucleosides originating different fragments. We observed fragments retaining the whole ribose moiety (Figure 1 and Supporting Information Chart S-2 (1.e), (2.d), (3.d), and (4.f)), fragments formed by the loss of C5′ of the ribose moiety (Figure 1 and the Supporting Information Chart S-2 (1.f), (2.e), (3.e), and (4.g)), and fragments keeping the nucleobase moiety (Figure 1 and Supporting Information Chart S-2 (3.c) and (4.e) and Figure 2 and Supporting Information Chart 3 (8.b) and (9.b)). Though ions of the ribose and 2′-deoxyribose moieties smaller than 245 m/z and 157 m/z could be observed in the spectra, fragments could not be unequivocally assigned to them. This is a common difficulty when a fragment ion can be assigned distinct moieties of the molecule having the same atom composition and, thus, the same chemical formula12 and m/z. This difficulty can be overcome using selectively labeled molecules or using a MS/MS system. We also could not assign any fragment to nucleobase moieties, which prevented a complete elucidation of their positional isotopomers. However, altogether, the mass isotopomers of nucleobase moiety ions, ribose and 2′-deoxyribose moiety fragment ions, and the moieties retaining the complete nucleobase and some ribose carbons provide rich information for flux estimation. For 13C metabolic flux analysis, the transfer of carbon in a metabolic network is simulated yielding isotopomer distributions of metabolites that are converted to any MID and then compared with experimental data for flux estimation.4,14,29,30 Upon the analysis of the mass spectra of nucleosides extracted from yeast fed with naturally labeled glucose, we selected as useful fragments those presented in the Supporting Information Charts S-2 and S-3. Fragments (1.c) and (4.f) are overlapped by an unidentified signal at the (m + n)/z mass isotopomer, but they are useful nevertheless because they provide unique information and the overlap can be avoided by an adequate choice of substrate labeling. As a proof of principle, we present a qualitative analysis of the metabolic pattern implied by the labeling of the fragments of nucleosides derived from the RNA and DNA of yeast cells fed with [1-13C]glucose (Supporting Information Figure S-5). This qualitative analysis is based on the known carbon interchanges through the metabolic reactions of the reconstructed network (Supporting Information Figure S-6), which includes glycolysis, PPP, TCA cycle, and nucleoside biosynthesis. The 13C tracer supplied in [1-13C]glucose was specifically distributed in the carbon backbone of nucleosides derived from nucleic acids, tracing back the metabolic pathway profile of the cell. Fragments generated from the same parent molecular ion but retaining different carbon moieties presented different mass isotopomer abundances as a consequence of the metabolic flux distribution. On the other hand, fragments retaining the same carbon moieties but derived from different nucleosides presented the same mass isotopomer abundances (Supporting Information Figure S-5 (1.e), (2.d), and (3.d);

■

DISCUSSION This work opens up ways to the analysis of isotopic labeling patterns in the nucleosides from cellular nucleic acids to eventually obtain information about metabolic flux distributions in proliferating cells. Labeling of DNA and RNA with stable isotopes has been used before to determine cell proliferation rates in tumor22 and infected cells, like lymphocytes.7,23,24 In such studies, cell proliferation was determined only from the total 13C enrichment in nucleosides. Most commonly, 13Cglucose, 2H-glucose,25 and 13C-dNTP have been used allowing one to also distinguish the contribution of the de novo synthesis and salvage pathways to the biosynthesis of DNA. However, in contrast to our study, these studies did not exploit the selective 13 C enrichment of nucleosides incorporated in RNA and DNA to determine relative metabolic fluxes. PPP fluxes, both the oxidative and the nonoxidative branch, are tracked in the mass isotopomer abundances of nucleosides fragments retaining ribose and 2′-deoxyribose moieties. The mass isotopomer abundances of the nucleobase purine fragments allow one to study the serine−glycine metabolic system. The serine−glycine system feeds purine nucleobase biosynthesis both directly via glycine and indirectly via formyltetrahydrofolate that belongs to the one-carbon metabolism system and is therefore in exchange with methyltetrahydrofolate that is a product of serine conversion to glycine. It has been shown that cancer cell proliferation is dependent on serine/glycine 26 and that serine supports one carbon metabolism for nucleotide biosynthesis.27 The application of our methodology will allow one to trace back the pathway(s) originating from serine, glycine, and one-carbon incorporated in purine nucleobases in nucleotide biosynthesis, helping one to clarify the interplay between serine, glycine, and the one-carbon metabolism. The flux distribution between glycolysis and PPP and between TCA cycle and pyruvate carboxylase (and the glyoxylate cycle when present) is tracked in the MID of pyrimidine nucleobases. The determination of positional isotopomers improves the accuracy and precision of 13C-metabolic flux analysis calculations.28 Previous work has elucidated the fragmentation of amino acids9,11 and of intermediary metabolites10 but not that of nucleosides. Here, we present the assignment of the fragment formulas of nucleosides leading to a partial determination of 13C positional isotopomers. We unequivocally assigned 31 fragments from the analysis of the fragmentation F

dx.doi.org/10.1021/ac503305w | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

■

(1.d) and (3.b)), giving evidence for the accuracy of the determination of the MID. The predominance of labeling in nucleobases indicates their precursor molecules were synthesized via glycolysis rather than via the oxidative branch of PPP that would lead to complete elimination of 13C in the C1-decarboxylation step. The presence of labeling of C5′ and C1′ in the ribose and 2′deoxyribose moieties reveals a substantial activity of the nonoxidative branch of PPP toward pentose phosphate synthesis. The presence of labeling in one or two carbons of purine nucleobases gives evidence of de novo biosynthesis of serine from 3-phosphoglycerate and its conversion to glycine and methyltetrahydrofolate. The occurrence of several rounds of TCA cycle would create multilabeled oxaloacetate and aspartate and would result in high abundance of m + 3 mass isotopomers of pyrimidine nucleobases. However, we observed the prevalence of the m + 1 mass isotopomer in pyrimidine nucleobases excluding extensive TCA cycle activity. Together with the observed low activity of the oxidative branch of PPP, the prevalence of m + 1 mass isotopomer of pyrimidine nucleobase reveals that the flux of pyruvate carboxylase is high compared to that of pyruvate dehydrogenase, a clear indication of anaplerotic replenishing of the oxaloacetate pool. The selectively labeled substrate must be carefully chosen for optimal estimation of metabolic fluxes of interest. The results above highlight the potential of our approach to determine relative flux distribution of metabolic pathways (glycolysis, PPP, TCA cycle, and anaplerotic pathways related to the TCA cycle and the serine−glycine system and homocysteine−cysteine system) in conditions when cell proliferation and RNA and DNA biosynthesis play a pivotal role, e.g., in cancer and stem and embryonic cells.

■

REFERENCES

(1) Niklas, J.; Schneider, K.; Heinzle, E. Curr. Opin. Biotechnol. 2010, 21, 63−69. (2) Wittmann, C.; Hans, M.; Heinzle, E. Anal. Biochem. 2002, 307, 379−382. (3) Dauner, M.; Sauer, U. Biotechnol. Prog. 2000, 16, 642−649. (4) Wittmann, C.; Heinzle, E. Biotechnol. Bioeng. 1999, 62, 739−750. (5) Krömer, J.; Quek, L.; Nielsen, L. Aust. Biochem. 2009, 40, 17−20. (6) Michal, G.Schomburg, D., Ed. Biochemical Pathways: An Atlas of Biochemistry and Molecular Biology; 2nd Edition, John Wiley & Sons: Hoboken, New Jersey, 2012. (7) Asquith, B.; Borghans, J. A. M.; Ganusov, V. V.; Macallan, D. C. Trends Immunol. 2009, 30, 182−189. (8) Salcedo-Sora, J. E.; Caamano-Gutierrez, E.; Ward, S. A.; Biagini, G. A. Trends Parasitol. 2014, 30, 170−175. (9) Antoniewicz, M. R.; Kelleher, J. K.; Stephanopoulos, G. Anal. Chem. 2007, 79, 7554−7559. (10) Rühl, M.; Rupp, B.; Nöh, K.; Wiechert, W.; Sauer, U.; Zamboni, N. Biotechnol. Bioeng. 2012, 109, 763−771. (11) Choi, J.; Grossbach, M. T.; Antoniewicz, M. R. Anal. Chem. 2012, 84, 4628−4632. (12) Wegner, A.; Weindl, D.; Jäger, C.; Sapcariu, S. C.; Dong, X.; Stephanopoulos, G.; Hiller, K. Anal. Chem. 2014, 86, 2221−2228. (13) Weitzel, M.; Nöh, K.; Dalman, T.; Niedenführ, S.; Stute, B.; Wiechert, W. Bioinformatics 2013, 29, 143−145. (14) Wiechert, W. Metab. Eng. 2001, 3, 195−206. (15) Christensen, B.; Nielsen, J. Metab. Eng. 1999, 1, 282−290. (16) Collart, M. A.; Oliviero, S. Current Protocols in Molecular Biology; John Wiley and Sons, Inc.: New York, 2001; Chapter 13, Unit 13.12. (17) Holm, C.; Meeks-Wagner, D. W.; Fangman, W. L.; Botstein, D. Gene 1986, 42, 169−173. (18) Quinlivan, E. P.; Gregory, J. F. Anal. Biochem. 2008, 373, 383− 385. (19) Orata, F. Derivatization reactions and reagents for gas chromatography analysis; Mohd, M. A., Ed.; InTech: Rijeka, Croatia, 2012; pp 83−109. (20) Brook, A. Acc. Chem. Res. 1974, 7, 77−84. (21) Van Winden, W. A.; Wittmann, C.; Heinzle, E.; Heijnen, J. J. Biotechnol. Bioeng. 2002, 80, 477−479. (22) Defoiche, J.; Zhang, Y.; Lagneaux, L.; Pettengell, R.; Hegedus, A.; Willems, L.; Macallan, D. C. Clin. Chem. 2009, 55, 1824−1833. (23) Ho, D. D.; Neumann, A. U.; Perelson, A. S.; Chen, W.; Leonard, J. M.; Markowitz, M. Nature 1995, 373, 123−126. (24) Wei, X.; Ghosh, S. K.; Taylor, M. E.; Johnson, V. A.; Emini, E. A.; Deutsch, P.; Lifson, J. D.; Bonhoeffer, S.; Nowak, M. A.; Hahn, B. H. Nature 1995, 373, 117−122. (25) Macallan, D. C.; Asquith, B.; Zhang, Y.; de Lara, C.; Ghattas, H.; Defoiche, J.; Beverley, P. C. L. Nat. Protoc. 2009, 4, 1313−1327. (26) Locasale, J. W. Nat. Rev. Cancer. 2013, 13, 572−583. (27) Labuschagne, C. F.; van den Broek, N. J. F.; Mackay, G. M.; Vousden, K. H.; Maddocks, O. D. K. Cell Rep. 2014, 7, 1248−1258. (28) Choi, J.; Antoniewicz, M. R. Metab. Eng. 2011, 13, 225−233. (29) Antoniewicz, M. R.; Kelleher, J. K.; Stephanopoulos, G. Metab. Eng. 2007, 9, 68−86. (30) Yang, T. H.; Wittmann, C.; Heinzle, E. E. Metab. Eng. 2004, 6, 256−267.

ASSOCIATED CONTENT

S Supporting Information *

(i) Methods of yeast cultivation, (ii) figures on the origin of nucleoside carbons (Figure S-1), sampling scheme (Figure S2), and fragment abundances in the labeling experiment (Figure S-3), (iii) schemes of the proposed mechanism of cleavage of the N-glycosidic bond (Scheme S-1) and fragment structures (Scheme S-2), (iv) charts of trimethylsilyl derivatives (Chart S1), structures of fragments of trimethylsilyl derivatives (Charts S-2 and S-3), and (v) tables with retention times and corresponding molecular ions (Table S-1) and natural mass isotopomer ratios (Table S-2). This material is available free of charge via the Internet at http://pubs.acs.org.

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Present Address †

K.M.-G.: Boehringer-Ingelheim, Vet-Medica, Calle 30, 44940 Jalisco, México. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We acknowledge fellowship SFRH/BD/18158/2004/dj9e to I.M.-S. and grants PEst-C/SAU/LA0001/2013-2014 (to Center for Neuroscience and Cell Biology) and PEst-OE/ QUI/UI0313/2014 (to Coimbra Chemistry Center) from “Fundaçaõ para a Ciência e a Tecnologia (FCT-Portugal)”. G

dx.doi.org/10.1021/ac503305w | Anal. Chem. XXXX, XXX, XXX−XXX