Enhanced Hypoxia-Inducible Factor (HIF)-1α Stability Induced by 5-Hydroxymethyl-2-Furfural (5-HMF) Contributes to Protection against Hypoxia
We first reported the role of 5-hydroxymethyl-2-furfural (5-HMF) against hypoxia. Here, we studied the mechanism by using oxygen-dependent degradation domain (ODD)-Luc mice, which are a useful model to probe the stabilization of hypoxia-inducible factor 1α (HIF-1α). Compared with three other compounds that have been reported to have a role in stabilizing HIF-1α, 5-HMF caused stronger bioluminescence, which is indicative of HIF-1α stability in the brain and kidney of ODD-Luc mice. We further demonstrated that the HIF-1α protein accumulated in response to 5-HMF in the brains and kidneys of these mice, as well as in PC12 cells. Additionally, 5-HMF promoted the nuclear translocation of HIF-1α and the transcriptional activity of HIF-1, which was evaluated by detecting vascular endothelial growth factor (VEGF) mRNA expression. These results suggest that 5-HMF stabilized HIF-1α and increased its activity. Considering the role of proline hydroxylases (PHDs) in negatively regulating HIF-1α stability, we explored whether 5-HMF interacts with the substrates and cofactors of PHDs, such as 2-oxoglutarate (2-OG), Fe2+ and vitamin C (VC), which affects the activity of PHDs. The result revealed that 5-HMF did not interact with Fe2+ or 2-OG but interacted with VC. This interaction was confirmed by subsequent experiments, in which 5-HMF entered into cells and reduced the VC content. The enhanced stability of HIF-1α by 5-HMF was reversed by VC supplementation, and the improved survival of mice caused by 5-HMF under hypoxia was abrogated by VC supplementation. Thus, we demonstrated for the first time that 5-HMF increases HIF-1α stability by reducing the VC content, which mediates the protection against hypoxia.
Hypoxia is particularly damaging in the brain, where an oxygen deficiency can lead to stroke, cerebral palsy and epilepsy (1, 2, 3, 4). For the threat of hypoxia, prevention is far more important than treatment. In our recent study, we identified 5-hydroxymethyl-2-furfural (5-HMF) as a potential preventive drug against hypoxic diseases (5,6).
5-HMF is a product of the Maillard reaction and exists in plants, foods and beverages that contain carbohydrates, such as the roots of Polygonum multiflorum, dried fruits, bread, fruit juices, milk and coffee (7,8). The beneficial roles of 5-HMF include antioxidant activity, the inhibition of red blood cell sickling and the amelioration of hemorheology (4,9, 10, 11, 12). Among these benefits, the role of antisickling, as indicated by the high affinity of 5-HMF for sickle cell hemoglobin, has been granted a U.S. patent (patent number 7119208). We first reported the role of 5-HMF in protecting against hypoxia in 2011 (5,6). Nevertheless, the underlying mechanisms have not yet been elucidated. In this study, we demonstrated for the first time that 5-HMF stabilizes hypoxia-inducible factor 1a (HIF-1α) through an interaction with VC, which mediates the protection of 5-HMF against hypoxia.
HIF-1α is a subunit of hypoxia-inducible factor 1 (HIF-1) and determines HIF-1 activity. Under normoxia, HIF-1α is extremely unstable and is quickly degraded by the ubiquitin-proteasome system. Proline hydroxylases (PHDs) (also known as HIF prolyl hydroxylases [HPH]) use O2 and 2-oxoglutarate (2-OG) (also called α-ketoglutarate) as substrates and Fe2+ and vitamin C (VC) (also called ascorbic acid) as cofactors hydroxylate prolyl sites at residues 402 and 564 in the oxygen-dependent degradation domain (ODD) of HIF-1α. Then, the tumor suppressor protein von Hippel-Lindau (VHL), which is an E3 ubiquitin ligase complex, recognizes hydroxylated HIF-1α, and after polyubiquitination by ubiquitins, HIF-1α is marked for destruction and is ultimately degraded by hydrolases in proteasomes (13, 14, 15). Thus, it can be inferred that the reduction of O2, 2-OG, Fe2+ or VC will inhibit the activities of PHDs and then elevate HIF-1α stability. For example, under hypoxia, HIF-1α becomes stable because of a lack of O2 as a substrate of PHDs and, therefore, accumulates its protein. Subsequently, HIF-1α translocates from the cytoplasm into the nucleus and heterodimerizes with constitutively expressed HIF-1β, which is another HIF-1 subunit, to form the heterodimeric transcription factor HIF-1. HIF-1 binds to a conserved hypoxia-response element (HRE), which contains the core nucleotide sequence 5′-(A/G) CGTG-3′ in the promoter region of its target genes, to activate the transcription of HIF-1 target genes. HIF-1 target genes include vascular endothelial growth factor (VEGF), erythropoietin (EPO), glucose transporters (GLUTs), BCL2/adenovirus E1B 19 kd-interacting protein 3 (BNIP3)/BNIP3-like (BNIP3L), miRNAs, and so on, which mediate the adaptation to hypoxic stress (16-18).
From the above description, we can easily draw an inference that the stability of HIF-1α determines the hypoxic adaptation. Therefore, drugs that target HIF-1α have the potential to prevent hypoxic diseases. Unfortunately, there is a lack of drugs to stabilize HIF-1α under normoxia and to prevent hypoxic injury. In our previous studies, we confirmed the protective role of 5-HMF against hypoxic injury in mice and in ECV304 cells (5,6). However, whether the role is correlated to HIF-1α stability is unknown. In this study, using ODD-Luc transgenic mice to evaluate agents that stabilize HIF-1α, we demonstrated for the first time that 5-HMF stabilizes HIF-1α under nor-moxia by reducing the VC content, and this effect is relieved by VC supplementation. Thus, 5-HMF has the potential to be developed as a drug to prevent hypoxic diseases in the future.
Materials and Methods
5-HMF, ferulic acid, coumarin, quercetin, CoCl2 and ascorbate oxidase were purchased from Sigma-Aldrich (St. Louis, MO, USA). Luciferin was obtained from Promega (Madison, WI, USA). The HIF-1α mouse monoclonal antibody was purchased from Novus Biologicals (Littleton, CO, USA). The mouse β-actin monoclonal antibody was acquired from Sigma-Aldrich. The rabbit Pro564-OH polyclonal antibody was purchased from Cell Signaling Technology (Boston, MA, USA). The Alexa 594 goat anti-mouse secondary antibody and 4′,6-diamidino-2-phenylindole (DAPI) were Molecular Probes products (Thermo Fisher Scientific, Waltham, MA, USA).
The ODD-Luc mice were acquired from The Jackson Laboratory (Bar Harbor, ME, USA). This mouse model expresses the oxygen-dependent degradation (ODD) domain of HIF-1α fused to the firefly luciferase gene. Fusing the ODD domain to luciferase renders the luciferase protein sensitive to oxygen-dependent hydroxylation and to subsequent ubiquitin-dependent degradation under normal physiological circumstances. The advantage of ODD-Luc mice is that these mice can provide a rapid noninvasive measurement of possible agonists of HIF-1 activity (19).
C57BL/6J male mice were purchased from the Animal Central of Academy of Military Medical Science (Beijing, China) and were used for survival experiments under hypoxia conditions. The animal protocol was approved by the Institutional Animal Care and Use Committee of the Academy of Military Medical Science and was consistent with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals (NIH publications no. 85-23) (20).
The PC12 cell line was obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). PC12 cells are a neuronal cell line derived from a rat pheochromocytoma, and these cells are often used as a model for neurons and a model for hypoxia because of their sensitivity to hypoxia. As described previously (19), briefly, cells were cultured in 75-cm2 tissue culture flasks (Corning, NY, USA) in Dulbecco’s modified Eagle medium (DMEM) (Gibco [Thermo Fisher Scientific]) and were maintained in DMEM containing 5% newborn calf serum and 10% horse serum supplemented with 100 U/mL penicillin and 100 pg/mL streptomycin in a moist atmosphere of 5% CO2 at 37°C. Experiments were performed 1–2 d after plating.
Luciferase Expression in ODD-Luc Mice
ODD-Luc mice were given an intraperitoneal injection of 5-HMF, CoCl2, ferulic acid, coumarin, quercetin or vehicle. One hour later, luciferin (50 mg/kg), which is a substrate of luciferase, was intraperitoneally injected into each mouse. Then, 5 min later, the mice were anesthetized by isoflurane for 5 min and then placed in a light-tight chamber equipped with an IVIS imaging camera (IVIS 50; Xenogen, Alameda, CA, USA). Photons were collected for a period of 10–30 s, and images were obtained and analyzed by using the Living Image software (Xenogen).
Hypoxia Treatment and Animal Survival
C57BL/6J mice were used to detect the effect of 5-HMF or VC on animal survival under a hypoxia environment. The mice were pretreated with 5-HMF (100 mg/kg), VC (140 mg/kg) or vehicle (normal saline) by intraperitoneal injection for 1 h and then supplemented with vehicles or VC (140 mg/kg) for another 1 h. After that, they were placed in a hypoxia chamber (DYC-DWI; Guizhou Fenglei, Guizhou, China) containing 20% O2. For sublethal acute hypoxia, the oxygen levels were decreased at a velocity of approximately 1–2% O2/min and reached 6% O2 within 10 min. After exposed to hypoxia, animal survival was recorded. The survival time was defined with 10 min as the threshold value. The mice that lived over 10 min under hypoxia were considered survivors. Survival rate (%) = the number of mice that lived over 10 min/total number of mice.
Western Blot Analysis
The cortices and kidneys of ODD-Luc mice or PC12 cells were harvested, and the total proteins were extracted with radioimmunoprecipitation assay lysis buffer, which contained 50 mmol/L Tris-HCl (pH 7.4), 1% NP-40, 0.25% Nadeoxycholate, 150 mmol/L NaCl and 0.1% sodium dodecyl sulfate (SDS) supplemented with protease inhibitors (Roche, Indianapolis, IN, USA) and 1 mmol/L PMSF. The lysates were quantified by using a protein assay kit (Bio-Rad Laboratories, Hercules, CA, USA), separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a polyvinylpyrrolidone difluoride membrane (Millipore, Billerica, MA, USA). After blocking with 5% nonfat dry milk in Tris-buffered saline with Tween-20 (TBST) at room temperature for 1 h, membranes were incubated with a HIF-1α mouse monoclonal antibody (1:1,000; Novus Biologicals) and with a β-actin mouse monoclonal antibody (1:5,000; Sigma-Aldrich) in TBST, which contained 5% nonfat dry milk, at 4°C overnight. This overnight incubation was followed by incubation with a goat anti-mouse secondary antibody, which was conjugated with horseradish peroxidase (1:1,000; Santa Cruz Biotechnology, Santa Cruz, CA, USA) in TBST, which contained 5% nonfat dry milk. Immune complexes on the membrane were visualized by using an enhanced chemiluminescence (ECL) detection system (Amersham Biosciences, Piscataway, NJ, USA). β-actin was used as a loading control. Experiments were repeated at least three times. Scans were used for semiquantitative analysis.
Cells that were grown on glass coverslips were fixed with methanol at −20°C for 20 min and permeabilized with 0.5% Triton X-100 at room temperature for 30 min. After blocking nonspecific binding with 5% normal goat serum in PBS-0.3% Triton X-100 at room temperature for 1 h, the cells were incubated with an anti-HIF-1α antibody (1:500; Novus Biologicals). The primary antibody was detected by incubation with an Alexa Flour 594 goat anti-mouse secondary antibody (1:500; Molecular Probes [Thermo Fisher Scientific]) at room temperature for 1 h. All incubations were performed in a humidified chamber. The cells were visualized and photographed by using a confocal microscope (Leica TCS SP5; Leica, Heidelberg, Germany).
Reverse-Transcription-Polymerase Chain Reaction (RT-PCR)
For the analysis of VEGF mRNA, total RNA isolated from the cortices and kidneys of ODD-Luc mice was prepared by using the TRIzol® reagent (Invitrogen [Thermo Fisher Scientific]) according to the manufacturer’s instructions. Total RNA (2 µg) was amplified by a two-step protocol by using M-MLV reverse transcriptase and Taq polymerase. The PCR products were amplified by using the following program: primers for VEGF (forward: 5′-GACCCTGGTGGACATCTT-3′, reverse: 5′-TAGTTCCCGAAACCCTGA-3′) were used to yield a 452-base pair (bp) product for 28 cycles at 58°C; 18S rRNA primers (forward: 5′-GCCCG AAGCGTTTACTTTGAA-3′, reverse: 5′-GGTGAGGTTTCCCGTGTTGA-3′) were used to yield a 459-bp product for 28 cycles at 56°C. The mRNA expression level was semiquantified relative to the endogenous expression level of 18S rRNA as an internal control. PCR products were separated on a 1.5% (w/v) agarose gel and visualized by ethidium bromide staining. The images were acquired by using a MultiImage Light Cabinet (Alpha Innotech, San Leandro, CA, USA).
Quantitative Real-time PCR (qRT-PCR)
The mRNA levels of HIF-1α and β-actin were measured using Power SYBR® Green (Applied Biosystems [Thermo Fisher Scientific]) and an ABI StepOnePlus RealTime PCR System (Applied Biosystems [Thermo Fisher Scientific]) according to the manufacturer’s instructions. Total RNA was isolated from PC12 cells by using the TRIzol® Reagent and performed according to the manufacturer’s protocol. An aliquot of 1 µg total mRNA was reverse-transcribed at 42°C for 1 h in a 10-µL reaction mixture. The qRT-PCR was performed in triplicate by using Power SYBR® Green and a real-time PCR system. The following HIF-1α primers were used: forward 5′-CCGCA GTGTG GCTAC AAGAA-3′, reverse 5′-GATGA GGAAT GGGTTCACAAATC-3′, 67 bp. The results were normalized by using β-actin as the internal control, with the following primers: forward 5′-CCAGT TCGCC ATGGATGAC-3′, reverse 5′-ATGCC GGAGCCGTTGTC-3′, 87 bp. Calculations were based on the “delta-delta method” by using the following equation: R (ratio) = 2−[ΔCT sample − ΔCT control] (21). The data were expressed as fold changes of the treatment groups relative to the control. β-actin was amplified as an internal control for qRT-PCR analysis.
Ultraviolet-Visible Spectroscopy Analysis
To detect the interaction of 5-HMF with Fe2+ (FeSO4), 2-OG or VC, an ultraviolet (UV)-visible spectrophotometer (Multiskan Go; Thermo Scientific [Thermo Fisher Scientific]) was used for the analyses. The solution of 20 µmol/L 5-HMF and 0, 2.5, 5, 10, 15 or 20 imol/L FeSO4, 2-OG or VC was mixed and incubated for 2 min at room temperature. Spectra were collected in the range of 200- to 500-nm wavelengths. The maximum UV absorptivity of 5-HMF is 285 nm.
To determine the VC content in the cytoplasm, sample preparation and measurements, as described previously in the literature (22,23), were as follows: briefly, after PC12 cells were treated with 5-HMF, the cells were permeabilized with 0.5% Triton X-100 in phosphate-buffered saline (PBS) for 10 min, and the supernatant was collected. Then, 200 µL of the supernatant was added to 1 mL 0.1 mmol/L NaH2PO4 (pH 5.6) and 750 µL ultrapure water and mixed. The absorbance at 265 nm (A265) was measured by using a spectrophotometer, followed by incubation with ascorbate oxidase (0.4 U, pH 5.6) for 3 min and measured again. The decrease in A265 was monitored. The VC content was calculated based on the decrease in A265.
High-Performance Liquid Chromatography (HPLC) Analysis of 5-HMF Content
The stock solution of 5-HMF (4 µmol/L) was prepared by dissolving 25.22 mg 5-HMF in 50 mL ultrapure water and stored at 4°C. Five concentrations of 5-HMF standard solutions (0.02, 0.10, 0.50, 2.52 and 12.61 mg/100 mL) were obtained by appropriately diluting the stock solution with mobile phase and were used to evaluate the linearity of the HPLC data (HP1100; Agilent, Burnsville, MN, USA). Chromatographic separation was performed by using a Diamonsil C18 column (250 × 4.6 mm inside diameter [i.d.], 5 µm; Dikma, Lake Forest, CA, USA). The mobile phase was a mixture of ultrapure water to acetonitrile (95:5, v/v) and was delivered at a flow rate of 1.0 mL/min. The detection wavelength was 285 nm.
Statistical analysis was performed by using the SPSS software. All the data were expressed as the means ± standard error (SE). A comparison between two samples was performed using a two-tail Student t test. p < 0.05 was considered statistically significant.
5-HMF Increased HIF-1α Stability in a Dose- and Time-Dependent Manner in ODD-Luc Transgenic Mice
5-HMF Promoted HIF-1α Stability and Nuclear Translocation in PC12 Cells
By immunofluorescence staining, we observed that HIF-1α gathered to the nucleus and partly entered into the nucleus in the cells that were treated with 5-HMF; in contrast, HIF-1α was scattered in the cytoplasm in the control (Figure 2C). By counting the number of cells where HIF-1α entered into their nuclei and the total number of cells that expressed HIF-1α, we obtained the ratio of HIF-1α nuclear translocation. The statistical data revealed that 5-HMF significantly increased the ratio of HIF-1α nuclear translocation (Figure 2D). In short, the above results show that 5-HMF promotes HIF-1α stability and its nuclear translocation.
5-HMF Increased HIF-1 Transcriptional Activity by the Induction of VEGF mRNA Expression in ODD-Luc Transgenic Mice
5-HMF Reacted with VC In Vitro and Reduced the VC Content in PC12 Cells
Increased HIF-1α Stability by 5-HMF Was Reversed by VC Supplementation in PC12 Cells and in ODD-Luc Mice
Increased Survival of Mice by 5-HMF Under Hypoxia Was Reversed by VC Supplementation
We first reported the role of 5-HMF against hypoxia in 2011 (5,6); however, the potential mechanism had not yet been elucidated. By using ODD-Luc mice, which are an ideal model for identifying and characterizing small molecular inhibitors of PHD (19,25), we found the role of 5-HMF in stabilizing HIF-1α. The ODD-Luc mice were generated by Safran et al. in 2006 (24). The mice provide a rapid noninvasive measurement of possible agonists of HIF-1 activity and are considered an ideal model that is suited for identifying and characterizing small molecular inhibitors of PHDs that would increase HIF-1 activity (25). We tested several compounds to find out the best one stabilizing HIF-1α using these ODD-Luc mice. Ferulic acid and coumarin increased the activity of HIF-1α in kidney but not in brain, and quercetin hardly played a role in stability of HIF-1α. Compared with the other three compounds, 5-HMF displayed a more significant role in stabilizing HIF-1α. The stabilized HIF-1α was concentrated in the kidney area and part of the brain regions, and the stabilization was dose and time dependent. These results indicate that 5-HMF shows not only a strong role in stabilizing HIF-1α, but also a relatively wide distribution. In particular, the increased HIF-1α activity in brain hints that 5-HMF can cross the blood-brain barrier, which suggests a more promising clinical application. To verify the effect of 5-HMF on HIF-1α, we further measured the protein expression of HIF-1α in the cortex and kidney of ODD-Luc mice by Western blot assay. In both the cortex and the kidney, 5-HMF increased HIF-1α expression. Thus, we obtained a preliminary conclusion that 5-HMF improved HIF-1α stability at the animal level.
By using hypoxia-sensitive PC12 cells, we confirmed the role of 5-HMF in stabilizing HIF-1α and explored the potential HIF-1α stabilization pathway of 5-HMF. The HIF-1α protein was accumulated by 5-HMF in a dose-dependent manner. To determine whether this effect was due to the enhanced HIF-1α stability or due to the increased protein de novo synthesis, we detected the level of HIF-1α mRNA by quantitative real-time PCR after the cells were treated with 5-HMF. Although there were no significant differences between 5-HMF and the control, there was a subtle trend of increasing HIF-1α mRNA with increasing 5-HMF dosage. This result could indicate that the action of 5-HMF may not solely affect PHD function and may increase HIF-1α levels.
In addition to regulating HIF-1α stability, 5-HMF was also involved in promoting the nuclear translocation of HIF-1α. One viewpoint is that the rapid nuclear translocation of HIF-1α is an efficient way to escape from proteasomal degradation (26); another viewpoint is that nuclear translocation is not necessary for HIF-1α stabilization because both “nuclear” and “cytoplasmic” proteasomes are fully competent for HIF-1α degradation in an oxygen-dependent manner (27). We believe that the nuclear translocation of HIF-1α is one active mechanism of HIF-1α and is a necessary process to form a heterodimer with HIF-1β in the nucleus. Perhaps HIF-1α cannot be degraded in the nucleus once the heterodimer is formed. Regardless, the nuclear translocation of HIF-1α is a premise to form the heterodimeric transcription factor HIF-1. In view of this result, we tested HIF-1α translocation by using immunofluorescence staining. By confocal scanning, we directly observed the nuclear translocation of HIF-1α, which was promoted by 5-HMF. The increased ratio of HIF-1α nuclear translocation suggested the enhanced activity of HIF-1α. As some literature previously reported, once HIF-1α is stable, this protein can rapidly translocate into the nucleus and activate the transcriptional activity of HIF-1 (28,29).
To further assess whether the transcription activity of HIF-1 is activated by 5-HMF, we measured the mRNA expression of VEGF, which is a downstream gene of HIF-1, in the cortex and the kidney of ODD-Luc mice by RT-PCR. Consistent with the above results, 5-HMF improved the level of VEGF mRNA in both the cortex and the kidney. Thus far, we can say that 5-HMF promotes HIF-1α stability, which therefore leads to the enhanced transcriptional activity of HIF-1.
The stability of HIF-1α is determined by the activity of PHDs. To determine the mechanism by which 5-HMF stabilizes HIF-1α, we analyzed the interaction of 5-HMF with Fe2+, 2-OG or VC in vitro, which determines the activity of PHDs. The result showed that 5-HMF did not react with Fe2+ or 2-OG but did react with VC. However, we still do not know the product that is generated from the reaction. In addition, we demonstrated that 5-HMF could enter the cells in a dose-dependent manner and reduce the intra-cellular VC content. This result suggests that 5-HMF reacted with VC in cells and thereby reduced the VC content.
Because 5-HMF reduced the VC content, could 5-HMF affect PHD activity? As we know, increased PHD activity can cause the prolyl hydroxylation of HIF-1α and its subsequent degradation. In view of the difficulty in measuring PHD activity directly, we tested the level of hydroxylated 564 prolyl residues of HIF-1α to indirectly reflect the activity of PHDs. We detected the reduced prolyl hydroxylation of HIF-1α in those cells that were treated with 5-HMF, whereas VC supplementation reversed this effect of 5-HMF. Next, we verified the reverse effect of VC on 5-HMF-induced HIF-1α stability separately in PC12 cells and in the ODD-Luc mice. Thus, we can speculate that PHD activity was inhibited by 5-HMF, which might be achieved through a reduction in VC content by the interaction of 5-HMF with VC, and thus, the stability of HIF-1α was improved. A recent finding that VC plays a significant role in supporting HIF-hydroxylase function and thereby modulates the cell survival response was consistent with our speculation (30).
To verify the effect of HIF-1α stability on the protection against hypoxia, we used C57BL/6J mice to test the survival of mice under hypoxia. The survival time and the survival rate of the mice that were pretreated with 5-HMF both increased, most likely because of the improved HIF-1α stability by 5-HMF. VC supplementation reversed this effect exerted by 5-HMF, which illustrated the role of HIF-1α stability in hypoxia protection. This role is contrary to the role of VC in tumors, in which HIF-1 is an important regulator of the adaptation of the tumor to its hypoxic microenvironment, and VC administration can significantly reduce tumor growth rates (31). This observation indicates that the same mechanism can lead to opposite effects due to different pathophysiological states. In this study, pretreatment with 5-HMF increased the stability of HIF-1α that subsequently translocated to the nucleus and activated the transcription of its downstream genes, including VEGF, EPO, GLUTs, and so on, which can enhance hypoxia tolerance (16, 17, 18,32) and ultimately increased survival time and survival rate of the mice. The reverse effect of VC supplementation on survival further confirmed that the role of 5-HMF in promoting survival under hypoxia was probably mediated by regulation on the HIF-1α pathway.
In summary, we identified a molecular mechanism by which 5-HMF increases HIF-1α stability in normoxia, which may play a critical role in the protection of 5-HMF against hypoxia. The regulation of HIF-1α stability by 5-HMF is an important pathophysiological mechanism that may provide a novel therapeutic approach to hypoxic diseases.
The authors declare that they have no competing interests as defined by Molecular Medicine, or other interests that might be perceived to influence the results and discussion reported in this paper.
This work was supported by the National Basic Research Programs of China (2012CB518200, 2011CB910800) and grants from the Natural Science Foundation of China (81071066, 81000856 and 31271211).
- 20.Institute of Laboratory Animal Resources (U.S.), Committee on Care and Use of Laboratory Animals. (1985) Guide for the Care and Use of Laboratory Animals Rev. 1985. Bethesda (MD): NIH. 83 pp. (NIH publication; no. 85-23).Google Scholar
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, and provide a link to the Creative Commons license. You do not have permission under this license to share adapted material derived from this article or parts of it.
The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
To view a copy of this license, visit (https://doi.org/creativecommons.org/licenses/by-nc-nd/4.0/)