2023 Volume 48 Issue 5 Pages 263-272
Glycation products are generated during the Maillard reaction, a non-enzymatic reaction between reducing sugars and the amino groups of proteins, which accumulate in the body with aging and cause many diseases. Herein, we have focused on dihydropyrazines (DHPs), which are glycation products formed by the dimerization of D-glucosamine or 5-aminolevulinic acid, and have reported that DHPs can produce several kinds of radicals and induce cytotoxicity via oxidative stress. To advance our understanding of DHP-mediated cytotoxicity, we selected a DHP, 3-hydro-2,2,5,6-tetramethylpyrazine (DHP-3), and two major Maillard reaction products, Nε-(carboxymethyl)-L-lysine (CML) and acrylamide, and performed comparative experiments focusing on their cytotoxicity and their ability to induce oxidative stress. The order of increasing cytotoxicity was DHP-3, acrylamide, and CML, and the LC50 value could be calculated only for DHP-3 (0.53 mM), indicating that DHP-3 is more toxic than the other Maillard reaction products. However, their toxicities were significantly lower than those of common toxic chemicals. Further, the results of their cytotoxicity assay were consistent with the results of intracellular reactive oxygen species production and activation of oxidative stress response signaling. These results indicate that the acute toxicity of Maillard reaction products is closely related to their ability to induce oxidative stress, and that DHP-3 is a particularly strong inducer of oxidative stress and thus exhibits high cytotoxicity among Maillard reaction products. In addition, we have shown that a comprehensive analysis comparing multiple Maillard reaction products is effective for elucidating their complex and diverse toxicities.
Carbohydrates, represented by glucose, are essential energy sources; however, their overconsumption leads to many diseases such as diabetes and metabolic syndrome. Glycation products are one of the key factors linking overconsumption of sugars and the development of such diseases. These products are generated through a non-enzymatic reaction between reducing sugars and the amino groups of proteins, which is called the Maillard reaction, and accumulate in the body with aging (Brownlee, 1995; Fournet et al., 2018; Tessier, 2010). Because harmful products of the Maillard reaction, including glycation products, have diverse and complex structures, it is difficult to evaluate and predict their toxicities.
Dihydropyrazines (DHPs) are glycation products formed by the dimerization of either D-glucosamine (Kashige et al., 1995) or 5-aminolevulinic acid (Teixeira et al., 2001). DHPs are ubiquitously present in the environment and in vivo, because many compounds with pyrazine skeletons, which are potential metabolites of DHPs, have been detected in foods and human urine (Joo and Ho, 1997; Maga, 1982; Zlatkis et al., 1973). Pyrazine and its derivatives, which result from the oxidation of DHPs, are also added to foods as flavoring agents. One such derivative, acetylpyrazine, is the most widely used flavoring agent in the United States, with a reported daily intake of 122 μg/day (Adams et al., 2002; Rizz, 1972). We have reported that DHPs are highly reactive, causing the production of various radicals (Yamaguchi et al., 2012), DNA strand breaks (Kashige et al., 2000; Yamaguchi et al., 1996), and growth inhibition and mutagenesis in Escherichia coli (Takechi et al., 2004). Experiments using mammalian cells have shown that 3-hydro-2,2,5,6-tetramethylpyrazine (DHP-3, Fig. 1A) is highly cytotoxic, especially among methyl-substituted DHPs, and induces cytotoxicity via oxidative stress (Ishida et al., 2012, 2014; Takechi et al., 2011, 2015). In addition to its adverse effects, recent studies have suggested that DHP-3 exerts anti-inflammatory effects by suppressing toll-like receptor 4 signaling (Esaki et al., 2020; Sawai et al., 2022).
Chemical structure of (A) 3-hydro-2,2,5,6-tetramethylpyrazine (DHP-3), (B) Nε-(carboxymethyl)-L-lysine (CML), and (C) acrylamide.
Although the effects of DHP-3 on cells, primarily its cytotoxicity, are becoming clearer, we believe that a comparative experiment with other Maillard reaction products would advance our understanding of DHPs. In this study, we selected two major compounds, Nε-(carboxymethyl)-L-lysine (CML), an advanced glycation end product (AGE), and acrylamide, a highly toxic Maillard reaction product noted for carcinogenicity (Figs. 1B and C, respectively), for comparison with DHP-3 (Chen et al., 2022; Rifai and Saleh, 2020; Schleicher et al., 1997). CML is generated via the reaction of lysine with sugar and is used as a marker of AGEs in foods. It is abundant in meats, nuts, and grains cooked at high temperatures (Chen, 2021; Nowotny et al., 2018). The Takayama Study conducted in Japan reported daily intakes of CML were 2.75 and 2.58 mg/day for men and women, respectively (Nagata et al., 2020). In contrast, acrylamide is formed from asparagine, an amino acid that is abundant in potatoes and grains (Mottram et al., 2002). The daily intake of acrylamide is reported to be 0.3–0.8 μg/kg body weight/day by the Food and Agriculture Organization of the United Nations (FAO)/World Health Organization (WHO) and approximately 7 μg/day by a Japanese cohort study (Zha et al., 2021; FAO/WHO, 2002). We compared DHP-3 with these Maillard reaction products, with a focus on acute cytotoxicity and the ability to induce oxidative stress.
DHP-3 was freshly synthesized according to a previously reported method with slight modifications (Yamaguchi et al., 1996). CML and acrylamide were purchased from Nippi (Tokyo, Japan) and FUJIFILM Wako Pure Chemical (Osaka, Japan), respectively. DHP-3 was dissolved in dimethyl sulfoxide (DMSO), while CML and acrylamide were dissolved in phosphate-buffered saline (PBS; FUJIFILM Wako Pure Chemical, Cat#166-23555). All the other reagents used were of the highest commercially available grade.Cell culture
HeLa cells (JCRB9004) were obtained from the Japanese Collection of Research Bioresources Cell Bank (Osaka, Japan) and cultured in Dulbecco’s modified Eagle’s medium (DMEM; FUJIFILM Wako Pure Chemical, Cat#041-29775) supplemented with 10% fetal bovine serum (FBS), purchased from Corning (Corning, NY, USA, Cat#35-079-CV).Cell viability assay
Cell viability was evaluated using the Cell Counting Kit-8 (CCK-8; Dojin Molecular Technologies, Kumamoto, Japan), according to the manufacturer’s protocol. Briefly, HeLa cells were seeded in 96-well plates at a density of 1.0 × 104 cells/well. After 24 hr, the cells were treated with DHP-3 (0, 0.001, 0.005, 0.01, 0.05, 0.1, 0.25, 0.5, 0.75, and 1.0 mM), CML (0, 0.01, 0.05, 0.1, 0.5, 1.0, 2.5, 5.0, 7.5, and 10.0 mM), or acrylamide (0, 0.01, 0.05, 0.1, 0.5, 1.0, 2.5, 5.0, 7.5, and 10.0 mM) for 24 hr in the absence of FBS. The final concentration of DMSO was fixed at 1% in all experiments with DHP-3 treatment. After the treatment, the cells were incubated in DMEM containing 10% CCK-8 solution for 1 hr in a CO2 incubator to generate pigments. Absorbance was measured at 450 nm using a microplate reader (SUNRISE Remote; Tecan Group, Männedorf, Switzerland). Lethal concentration 50 (LC50) values were calculated using GraphPad Prism 5.04 software (GraphPad Software, La Jolla, CA, USA).Imaging of intracellular reactive oxygen species (ROS)
Intracellular ROS were visualized using a ROS Assay Kit (Dojin Molecular Technologies) according to the manufacturer’s protocol. Briefly, HeLa cells (1.0 × 106 cells) were seeded in a 35-mm dish 24 hr before the experiment. The cells were treated with 1 mM DHP-3, 10 mM CML, or 10 mM acrylamide for 3 hr, washed three times with Hanks’ Balanced Salt Solution (HBSS; FUJIFILM Wako Pure Chemical, Cat#085-09355), and further incubated in a Highly Sensitive DCFH-DA Working solution for 30 min. After washing twice with HBSS, the cells were kept in 1 mL of HBSS and analyzed using a C2 Confocal Laser Point Scanning Microscope (Nikon, Tokyo, Japan). A set of lasers and filters for the detection of fluorescein isothiocyanate isomer-I were utilized for ROS detection. Images were obtained using the NIS-Elements software (Nikon) and photographically processed with Affinity Photo 2 software (Serif Europe, Nottingham, UK).Reporter gene assay
The luciferase reporter plasmids, pGL4.37 [luc2P/ARE/Hygro] and pNL3.2.NF-κB-RE [NlucP/NF-κB-RE/Hygro] were purchased from Promega (Madison, WI, USA) to analyze the activation of antioxidant response element (ARE) by nuclear factor erythroid 2-related factor 2 (Nrf2) and that of nuclear factor-κB (NF-κB) response element, respectively. HeLa cells (1.0 × 104 cells) were seeded in a 96-well white plate and cultured for 24 hr. Reporter plasmids (0.3 μg/well) and HilyMax transfection reagent (Dojin Molecular Technologies, 0.9 μL/well) were mixed in Opti-MEM I reduced Serum Medium (Opti-MEM; Thermo Fisher Scientific, Waltham, MA, USA) and incubated for 15 min at room temperature. The plasmid-transfection reagent complex was added to the culture medium and incubated for 4 hr after which the medium was replaced with fresh DMEM containing 10% FBS. The transfected cells were then cultured for 24 hr and treated with DHP-3 (0.05, 0.1, 0.25, 0.5, 0.75, and 1.0 mM), CML (0.5, 1.0, 2.5, 5.0, 7.5, and 10.0 mM), or acrylamide (0.5, 1.0, 2.5, 5.0, 7.5, and 10.0 mM) for 12 hr in the absence of FBS. Reporter activity of pGL4.37 and pNL3.2.NF-κB-RE was measured using Luciferase Assay systems and Nano-Glo Luciferase Assay system, respectively (both from Promega, Cat#E1500 and N1110, respectively). Chemiluminescence was measured using an Infinite M200 Pro microplate reader (Tecan group), and each signal was normalized to that of DMSO- or PBS-treated cells.Immunoblotting
HeLa cells (1.0 × 106 cells) were seeded in a 35-mm dish one day prior to the experiment and treated with 1 mM DHP-3, 10 mM CML, or 10 mM acrylamide for 0, 6, 12, and 24 hr in the absence of FBS. After washing with PBS, cells were lysed in 200 μL of radio-immunoprecipitation assay buffer (FUJIFILM Wako Pure Chemical, Cat#182-02451). The protein concentration of the lysates was determined using the Quick Start Bradford 1 × Dye Reagent protein assay kit (Bio-Rad Laboratories, Hercules, CA, USA) with bovine serum albumin as a standard. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis was used to separate 20 μg of each lysate, and the separated samples were electronically blotted onto a polyvinylidene difluoride membrane (FUJIFILM Wako Pure Chemical, Cat#034-25663). The blots were washed with Tris-buffered saline containing 0.1% Tween 20 (TBS-T) and incubated in TBS-T containing 5% skim milk for 30 min. After the blocking step, the blots were incubated with rabbit anti-Nrf2 antibody (Proteintech, Rosemont, IL, USA, Cat#16396-1-AP), rabbit anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody (Proteintech, Cat# 10494-1-AP), and rabbit anti-heme oxygenase 1 (HO-1) antibody (GeneTex, Alton Pkwy Irvine, CA, USA, Cat#GTX101147), all of which had been diluted 2,000 times with TBS-T containing 5% skim milk at 4°C overnight. The blots were washed with TBS-T for 15 min and incubated for 1 hr with horseradish peroxidase-conjugated goat anti-rabbit IgG (Proteintech, Cat#SA00001-2). After 15 min of washing with TBS-T, the signal was visualized using EzWestLumi plus (ATTO, Tokyo, Japan) and the iBright Imaging System (Thermo Fisher Scientific) as the substrate and detector, respectively. The relative expression levels of Nrf2 and HO-1 were calculated as % GAPDH, and each value was standardized with the mean value of 0 hr treatment.Statistical analysis
The analysis was conducted with GraphPad Prism 5.04 software. Significant changes in the protein levels of Nrf2 and HO-1 compared to 0 hr treatment were determined by one-way analysis of variance followed by Dunnett’s test (P < 0.05).
We freshly synthesized DHP-3 and compared its cytotoxicity to that of CML and acrylamide in HeLa cells. The cytotoxicity of DHP-3 was not observed up to 0.1 mM (100 μM) and finally became marked above 0.5 mM (Fig. 2A), and LC50 was calculated as 0.53 mM (n = 3), which is close to the milli-molar order. In contrast, although the concentrations of CML and acrylamide were set up to 10 times those of DHP-3, no cytotoxicity was observed in CML (Fig. 2B), and cytotoxicity was observed in acrylamide at concentrations close to 10 mM (Fig. 2C). These results indicate that the acute toxicity of DHP-3 is lower than that of common toxic chemicals but higher than that of the other Maillard reaction products.
Cell viability following treatment with each Maillard reaction product. The viability of HeLa cells after 24 hr treatment with the Maillard reaction products was determined using Cell Counting Kit-8. Each plot represents the mean ± S.E. of triplicate assays. Each experiment was repeated thrice, and a representative experiment is shown. The results for (A) 3-hydro-2,2,5,6-tetramethylpyrazine (DHP-3), (B) Nε-(carboxymethyl)-L-lysine (CML), and (C) acrylamide are presented.
Our previous studies showed that DHP-3 elicits cytotoxicity by inducing oxidative stress (Ishida et al., 2012, 2014; Takechi et al., 2011, 2015). Hence, we treated HeLa cells with Maillard reaction products, and intracellular ROS were detected using a fluorescent probe. Three-hour treatments with DHP-3 evoked ROS production in a dose-dependent manner; significant green signals were detected in 1 mM DHP-3-treated cells, but the signals were little in 0.5 mM DHP-3-treated cells. Interestingly, 10 mM acrylamide treatment resulted in marked ROS generation comparable to 1 mM DHP-3-treated cells, while the signal was weak in the 10 mM CML-treated group (Fig. 3). This result is consistent with our previous studies and suggests that the acute toxicity of the Maillard reaction products is closely related to their capacity to generate ROSs.
Generation of intracellular reactive oxygen species (ROS) following treatment with each Maillard reaction product. HeLa cells were treated with 0.5 mM/1 mM 3-hydro-2,2,5,6-tetramethylpyrazine (DHP-3), 10 mM Nε-(carboxymethyl)-L-lysine (CML), or 10 mM acrylamide for 3 hr, and the generated ROS was visualized using a fluorescent probe. The stained cells were observed using a confocal laser microscope to detect fluorescein isothiocyanate isomer-I. Each scale represents 200 µm.
Exposure to oxidative stress activates the Nrf2 and NF-κB pathways in cells (Sivandzade et al., 2019). Once activated, Nrf2 translocates to the nucleus, binds to ARE, and upregulates many antioxidant genes. In contrast, NF-κB is involved in the inflammatory responses and aggravates oxidative stress, leading to cytotoxicity and apoptosis. Hence, we analyzed the effects of Maillard reaction products on these pathways using reporter gene assays to examine the association between the activation of these signals and cytotoxicity. Activation of the Nrf2 pathway was marked when cells were treated with DHP-3, rising in a concentration-dependent manner, reaching 5-fold at 0.5 mM, and then quickly declining to below control levels (Fig. 4A). In contrast, no activation was observed with CML and acrylamide, even though the cells were treated with a concentration 10 times higher than that of DHP-3 (Figs. 4B and C). In the case of the NF-κB pathway, there were slight differences depending on the Maillard reaction product: DHP-3 treatment showed activation at the lowest concentration among the three products, at 2–3 times of the control, whereas there was no marked activation in CML (Figs. 5A and B). In the case of acrylamide, the signal was not different from that of the control up to 5 mM, but at 7.5 and 10 mM, the signal increased rapidly, reaching approximately 20-fold (Fig. 5C). As both pathways were most activated at low concentrations in the case of DHP-3, which showed marked cytotoxicity, but not in CML, which showed no cytotoxicity, a correlation between the activation of these pathways and cytotoxicity was predicted. In contrast, acrylamide induced NF-κB activation and cytotoxicity only at high concentrations, suggesting that even though the Maillard reaction products induced the same oxidative stress, they had different effects on intracellular signaling. The results of the reporter gene assay suggested that DHP-3 transiently increased the activity of Nrf2 (Fig. 4). Therefore, we analyzed the time-dependent effects of the Maillard reaction products on the levels of Nrf2 and HO-1, a downstream factor of Nrf2, by fixing the concentration of DHP-3 at 1 mM, and the concentrations of CML and acrylamide at 10 mM. As shown in Fig. 6A, 12 hr treatment with 1 mM DHP-3 resulted in a significant 2–3 fold increase in Nrf2 levels while 24 hr treatment decreased Nrf2 levels to those of the control. HO-1 levels also tended to fluctuate with changes in Nrf2 levels, but the differences were not statistically significant. In contrast, treatment with neither 10 mM CML nor acrylamide had a significant effect on Nrf2 and HO-1 levels, although acrylamide altered Nrf2 levels in a time-dependent manner, similar to DHP-3 (Figs. 6B and C). Immunoblotting results also supported findings that DHP-3 is a stronger inducer of oxidative stress than CML and acrylamide, and suggested that the various Maillard reaction products have different effects on intracellular signaling pathways.
Effect of the Maillard reaction products on the Nrf2 pathway. HeLa cells were transfected with pGL4.37 [luc2P/ARE/Hygro], treated with each Maillard reaction product for 12 hr, and the luciferase activity was measured. Each plot represents the mean ± S.E. of triplicate assays. The results for (A) 3-hydro-2,2,5,6-tetramethylpyrazine (DHP-3), (B) Nε-(carboxymethyl)-L-lysine (CML), and (C) acrylamide are presented.
Effect of the Maillard reaction products on the NF-κB pathway. HeLa cells were transfected with pNL3.2.NF-κB-RE [NlucP/NF- κB-RE/Hygro], treated with each Maillard reaction product for 12 hr, and the luciferase activity was measured. Each plot represents the mean ± S.E. of triplicate assays. Only the 10 mM plot in (C) shows the mean + S.E. The results for (A) 3-hydro-2,2,5,6-tetramethylpyrazine (DHP-3), (B) Nε-(carboxymethyl)-L-lysine (CML), and (C) acrylamide are presented.
Time-dependent changes in Nrf2 and heme oxygenase 1 (HO-1) protein levels driven by Maillard reaction products. HeLa cells were treated with 1 mM 3-hydro-2,2,5,6-tetramethylpyrazine (DHP-3), 10 mM Nε-(carboxymethyl)-L-lysine (CML), or 10 mM acrylamide for 0, 6, 12, or 24 hr. Cell lysates (20 μg) were analyzed by immunoblotting, and Nrf2, HO-1, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were detected. The relative intensity was determined as % GAPDH and standardized with a mean value at 0 hr of treatment. Each bar represents the mean ± S.E. of three samples. Statistical significance was determined by analysis of variance, followed by Dunnett's test, compared to 0 hr (* P < 0.05). The results for (A) DHP-3, (B) CML, and (C) acrylamide are presented.
In this study, we performed comparative experiments among DHP-3, which we have been studying continuously over the years, and the major Maillard reaction products, CML and acrylamide. A significant difference in acute toxicity to HeLa cells was observed between DHP-3 and the other Maillard reaction products in the order DHP-3 > acrylamide > CML (Fig. 2). DHP-3 had an LC50 value of 0.53 mM. Since the LC50 value obtained in the previous study was 0.89 mM (Miyauchi et al., 2021), the freshly synthesized DHP-3 in this study seemed to be slightly more reactive. The acute toxicity of acrylamide and CML was very low, which is consistent with the results of previous studies (Boesten et al., 2014; Shan et al., 2014). These acute toxicity results agreed with those for ROS production (Fig. 3) and activation of the Nrf2/NF-κB pathways (Figs. 4, 5, and 6), suggesting that the acute toxicity of Maillard reaction products is mainly determined by their ability to induce oxidative stress. As shown in Fig. 1, DHP-3 exhibits higher hydrophobicity than CML and acrylamide and can thus easily permeate cell membranes, which may be another factor contributing to its ability to induce oxidative stress and acute toxicity. DHP-3 showed the most acute cytotoxicity among the three Maillard reaction products; however, its LC50 was on the order of near milli-molar, making it difficult to evaluate the Maillard reaction product as a general toxic chemical.
The receptor for AGEs (RAGE), localized at the plasma membrane, is deeply involved in the toxic expression of CML and other AGEs (Lander et al., 1997; Sparvero et al., 2009). Although the basal expression level of RAGE is very low in organs other than the lung, activation of the AGEs-RAGE axis by the binding of AGEs increases the expression level via positive feedback through the NF-κB pathway, which exacerbates oxidative stress and results in many inflammatory diseases (Bierhaus et al., 2005; Li and Schmidt, 1997). Because RAGE is not detectable in HeLa cells (Miyauchi et al., 2021), no acute toxicity or ROS production by CML was observed in this experiment; however, it is possible that these results would be marked in a model with RAGE induction. In contrast, acrylamide is believed to be cytotoxic independent of RAGE, which is similar to DHP-3 (Friedman, 2015; Miyauchi et al., 2021). Direct DNA damage via oxidative stress is thought to be a major contributor to acrylamide toxicity. The high ROS production (Fig. 3) and activation of the NF-κB pathway (Fig. 5C) observed in this study supports this view. However, the effect on the Nrf2 pathway was quite different from that of DHP-3 (Figs. 4 and 6), indicating that the Maillard reaction products have distinct mechanisms of acute toxicity for each compound. In the future, the chronic effects of treatment at lower concentrations should be examined to further understand their toxicity.
Our recent studies have revealed that DHP-3 exhibits anti-inflammatory effects (Esaki et al., 2020; Sawai et al., 2022). In these studies, human hepatocarcinoma-derived HepG2 cells were pre-treated with lipopolysaccharide (LPS), which activates inflammatory signals, and then further treated with DHP-3. DHP-3 markedly suppressed LPS-dependent activation of the NF-κB pathway by inhibiting the phosphorylation of the p65 subunit and its nuclear translocator. In contrast, in this study, the results of the reporter gene assay suggested that DHP-3 induced 2–3-fold activation of the NF-κB pathway (Fig. 5A). These results were consistent with those of previous studies on the anti-inflammatory responses observed on pre-treatment with LPS. Future studies should focus on analyzing the direct effect of DHP-3 on the NF-κB pathway, as it is difficult to draw conclusions based on gene reporter assay results alone.
Comparative experiments with other Maillard reaction products clearly showed that DHP-3 exhibits a particularly high capacity to induce oxidative stress, leading to marked cytotoxicity. Considering that a variety of Maillard reaction products are constantly produced in the body and taken up from foods, it is necessary to estimate the total toxicity of these compounds. We hope to develop a method for estimating the total toxicities of Maillard reaction products by furthering our comparative approach to the toxicological analysis of DHPs.
This work was supported in part by JSPS KAKENHI (Grant-in-Aid for Early-Career Scientists), Grant Number 22K15330 (Recipient YM). The authors thank Mr. Masahide Isoda, Ms. Akane Katayama, Ms. Yumi Yoshida, Ms. Saki Imada, Mr. Taiki Iwata, Ms. Yumiko Makihara, and Mr. Toshiki Yatsuki for helping us in a prat of experiment.Conflict of interest
The authors declare that there is no conflict of interest.