Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

Within the cells, glucose is mainly catalyzed through glycolysis and pentose phosphate pathway. Such a utilization of glucose is necessary for cellular adaptation against stressors. For examples, hypoxia induces a transcriptional factor, hypoxia inducible factor-1 (HIF-1), which induces GLUT-1, GAPDH, PDK-1, and LDH [1], and promotes anaerobic glycolysis to adapt hypoxia. Since hypoxia causes an induction of heme oxygenase [2, 3], the CO-producing enzyme, it is not unreasonable to hypothesize that the gas might also change the biotransformation of glucose. To measure the global picture of glucose utilization, we have herein established the flux analysis with 13C-labeled glucose using LC-MS/MS. Using this methodology, we aimed to examine the effects of CO on glucose utilization through the fluxome analysis.

2 Materials and Methods

2.1 Materials and Extraction of Metabolites

DHAP, 2-deoxy-G6P, Gal1P, GA3P, 3PG, and 6PG were purchased from Sigma. Fru1, 6P2, Ribo5P, and Xylu5P were purchased from Fluka. Fru6P, Glu1P, Glu6P, Pyr, PEP, were purchased from Wako. Lac was purchased from Tokyo Kasei Kogyo. Sed7P was purchased from GLYCOTEAM GmbH. CORM (Tricarbonyldichlororuthenium(II) dimer) and ruthenium (III) chloride (Ru) were purchased from Aldrich. U937 cells were cultured with RPMI-1640 medium supplemented with 10% FBS, penicillin and streptomycin, and grown in a CO2 incubator. U937 cells were plated at 5 × 106 cells/10 cm dish and cultured around 14 h. Then the cells were treated with 100 μM CORM or Ru as a control for 30 min. After the CORM or Ru treatment, medium was exchanged to glucose-free RPMI-1640 supplemented with 10% FBS, penicillin and streptomycin, and added 13C6-glucose at 2000 mg/L. After the addition of 13C6-glucose containing medium, cells were left for 5 min in the CO2 incubator and immediately centrifuged to collect them at 1,500 rpm for 2 min. Then cells were washed with PBS(–) once. Collected cells were immediately frozen in liquid N2 and stored at –80°C until extraction. 400 μl methanol containing internal standard (IS) (2-deoxy-glucose-6P: 0.3 μM) were added to harvested cells and voltexed. Then, 200 μL water was added and mixed for 30 min. After the addition of 400 μL chloroform and voltexed, centrifuged at 14,000×g for 30 min. Supernatant were filtrated with 5 kDa cut-off filter and dried with Speed-Vac. Dried samples were stored at –80°C till LC-MS/MS analysis.

2.2 LC-MS/MS

LC-MS/MS analysis was performed as previously reported with some modifications [4, 5]. Chromatographic separations were carried out on Synergi Hydro-RP (C18) column (150 × 2.0 mm, 4 μm, 80 Å, phenomenex). The chromato-graphic system was an Agilent 1100 series binary HPLC system (Agilent Technologies). Separations were performed under gradient conditions as described in Table 1. Column temperature was set to 40°C. Mobile phases consisted of 15 mM acetic acid/10 mM tributylamine/H2O (A) and 100% Methanol (B).

Table 1 Gradient profil e
Full size table

Mass spectrometry analysis was carried out on Applied Biosystems/API 3000 equipped with a TurboIon spray source. The MS was operated in the negative ion and multiple reactions monitoring (MRM) mode. Single analyte standard dissolved in a 10 mM acetic acid was infused at a flow rate of 5 μL/min for tuning compound dependent MS parameters. These parameters were as follows: ionspray voltage −4000 V, nebulizer gas (NEB) 14, curtain gas (GUS) 14, auxiliary gas temperature (TEM) 400, and collision gas (CAD) 4. MS parameters for each metabolite are given in Table 2. Extracted metabolites were diluted with 50 μL of H2O and the injected volume was 10 μL. Standard samples (1 μM) and extracted metabolites from cultured cells were measured by LC-MS/MS and detected peaks were compared. Areas of each metabolite were calculated and divided by IS peak area.

Table 2 MS parameters
Full size table

3 Results

3.1 Measurement of 13C-Labeled Metabolites by LC-MS/MS

To assess the effect of carbon monoxide on glucose metabolism, we tried to establish a method for measuring the global flux of 13C6-labeled glucose by modifying the method previously reported by Luo et al. [4, 5]. First, we measured standard samples and confirmed the detection of each metabolite. Detected peaks of each standard are shown in Fig. 1. Sed7P, Ribo5P, Ribu5P, Xylu5P, Glu6P, Fru6P, Glu1P, Gal1P, GA3P, DHAP, Pyr, Lac and 2-deoxy-G6P were detected in period 1 and 6PG, Fru1, 6P2, 3PG, and PEP were detected in period 2. MS scan for Glu1P/Gal1P and Glu6P/Fru6P detected all other metabolites but in lower sensitivity (Fig. 1d). Next, we confirmed the detection of 13C-labeled metabolites. We measured the cells cultured without 13C6-glucose as a negative control and confirmed that they did not give peaks of 13C-labeled metabolites (data not shown). Then, we measured the cells cultured with 13C6-glucose and checked the overlap of signals derived from 13C-labeled metabolites and non-labeled metabolites (Fig. 1). With these results, we confirmed that metabolites derived from 13C6-glucose were measured correctly.

Fig. 1
figure 14_1_117031_1_En

Chromatograms of standard samples (a–k) and metabolites in extracted samples (l–w). Each chromatogram shows the result of LS-MS/MS analysis of 1 μM standard samples (a–k). Retention times are shown at the bottom of each chromatogram. Arrows (a–k) indicate peaks of detected standard samples. (d) Solid line indicates MS scan of Gal1P/Glu1P and dashed line indicates MS scan of Glu6P/Fru6P. Solid lines indicate [12C] metabolites and dashed lines indicate [13C] metabolites (l–u). Solid line indicates MS scan of Gal1P/Glu1P and dashed line indicates MS scan of Glu6P/Fru6P (v, w). [12C] metabolites and [13C] metabolites are shown in (v) and (w), respectively. Arrows indicate detected [13C] labeled in each metabolite (l–w)

Full size image

3.2 Effect of CO on Glucose Metabolism

To assess the effect of CO on glucose metabolism, we measured and compared the global flux of 13C6-glucose between CORM treated and Ru treated U937 cells (Fig. 2). This result shows that the amount of 6PG, Xylu5P was significantly increased and Sed7P and Ribu5P were modestly increased by CORM treatment, they were located on PPP. On the other hand, GA3P and 3PG were significantly decreased and PEP was modestly decreased by CORM treatment, they were located on the glycolytic pathway. These results indicate that CO suppressed flux to the glycolytic pathway, and increased flux to PPP.

Fig. 2
figure 14_2_117031_1_En

Measurement of each [13C] metabolite in Ru (black bars) or CORM (gray bars) treated U937 cells. Each value was calculated as relative value when the amount of each metabolite in Ru-treated cells was defined as 100. Bars indicate means ± S.E , *P < 0.05, #P < 0.07, n = 7

Full size image

4 Discussion

CO is produced by HO in response to a variety of stress, and is known to contribute to cell protection [2, 3]. In this report, we showed that CO worked as a putative metabolic regulator for glucose biotransformation, although the mechanisms remain unknown. Because activation of PPP produces NADPH by G6PD [6], CO might regulate redox conditions within cells, functioning against stresses. On the other hand, CO appears to suppress glycolysis. Since the cells were viable under the current experimental conditions, alternative mechanisms for cellular energetic to maintain ATP in the CO-exposed cells should be further examined. Collectively, the current study provided a useful method to follow the fate of glucose in different types of cells including not only those of mammals but of microorganisms which are exposed to varied stress conditions.