Qiliqiangxin attenuates hypoxia‐induced injury in primary rat cardiac microvascular endothelial cells via promoting HIF‐1α‐dependent glycolysis

Abstract Protection of cardiac microvascular endothelial cells (CMECs) against hypoxia injury is an important therapeutic strategy for treating ischaemic cardiovascular disease. In this study, we investigated the effects of qiliqiangxin (QL) on primary rat CMECs exposed to hypoxia and the underlying mechanisms. Rat CMECs were successfully isolated and passaged to the second generation. CMECs that were pre‐treated with QL (0.5 mg/mL) and/or HIF‐1α siRNA were cultured in a three‐gas hypoxic incubator chamber (5% CO2, 1% O2, 94% N2) for 12 hours. Firstly, we demonstrated that compared with hypoxia group, QL effectively promoted the proliferation while attenuated the apoptosis, improved mitochondrial function and reduced ROS generation in hypoxic CMECs in a HIF‐1α‐dependent manner. Meanwhile, QL also promoted angiogenesis of CMECs via HIF‐1α/VEGF signalling pathway. Moreover, QL improved glucose utilization and metabolism and increased ATP production by up‐regulating HIF‐1α and a series of glycolysis‐relevant enzymes, including glucose transport 1 (GLUT1), hexokinase 2 (HK2), 6‐phosphofructokinase 1 (PFK1), pyruvate kinase M2 (PKM2) and lactate dehydrogenase A (LDHA). Our findings indicate that QL can protect CMECs against hypoxia injury via promoting glycolysis in a HIF‐1α‐dependent manner. Lastly, the results suggested that QL‐dependent enhancement of HIF‐1α protein expression in hypoxic CMECs was associated with the regulation of AMPK/mTOR/HIF‐1α pathway, and we speculated that QL also improved HIF‐1α stabilization through down‐regulating prolyl hydroxylases 3 (PHD3) expression.

the exchange of substance and energy between microcirculation and myocardial tissue and thus serve as the essential elements responsible for maintaining the normal myocardial tissue metabolism. 2 Hypoxia-induced CMECs injury is considered as an initiating process and pathological basis of various cardiovascular diseases, 3 and protecting CMECs from hypoxia insult might thus be an important therapeutic strategy for treating various cardiovascular diseases. 4 Hypoxia can insult various endothelial cell biological activities such as cell growth, survival, migration and energy metabolism, which may lead to angiogenesis impairment, redundant reactive oxygen species (ROS) generation, mitochondrial dysfunction and energy metabolism disturbance to some extent.
Hypoxia-inducible factor-1a (HIF-1a) is an important oxygen balance regulator which plays critical roles in the transportation and utilization of oxygen. 5 HIF-1a mediates hypoxia-induced adaptive changes in various cell types by activating various downstream target genes responsible for oxygen delivery, glucose metabolism, 6  and COX4I2). 9 Besides delivering nutrients and oxygen to organs and tissues, endothelium also plays a critical role in regulating tissue glucose metabolism via HIF-1a transcriptional regulation of GLUT1 expression. 10,11 Previous studies showed that up-regulated HIF-1a activity was linked with reduced the apoptosis of various cell types including cardiomyocytes and human umbilical vein endothelial cells (HUVECs). 12,13 On the contrary, suppressed HIF-1a expression was related to reduction in angiogenesis and glycolysis, increase in ROS production and induction of apoptosis or death of endothelial cells, cancer cells and other cell types. [14][15][16][17][18] Taken together, HIF-1a is a key regulator of endothelium survival under hypoxic condition, and up-regulation of HIF-1a might be a key method of protecting endothelial cells from hypoxic injury. Qiliqiangxin (QL), a traditional Chinese medicine compound preparation, is composed of 11 Chinese herbal medicines (such as astragalus, ginseng and aconite). Protective effects of QL were demonstrated by a multicenter randomized double-blind clinical study including 512 heart failure patients. 19 It was also reported that QL could improve cardiomyocyte metabolism and inhibit cardiomyocyte apoptosis. [20][21][22][23] Our previous work found that QL could promote cardiac angiogenesis and up-regulate HIF-1a expression in failing heart and hypoxic CMECs via NRG-1/ErbB-PI3K/Akt/mTOR pathway, thus promoting cardiac angiogenesis. 24 Isolation and culture of primary CMECs were performed with explant method, and the procedures were as follows: the rats were killed by cervical dislocation and the heart was quickly excised and rinsed with 4°C pre-colding PBS. After cutting off the upper part of the great vessels and atrial tissues, the epicardium and endocardium were gently torn. The remaining ventricular tissues were cut into 1 mm 3 small pieces and plated onto a 10-cm culture dish pre-coated with in 1 mL of foetal bovine serum (FBS). Tissues were then cultured in 37°C, 5% CO2 incubator for 4 hours and for another 48 hours in high-glucose Dulbecco's modified Eagle's medium (highglucose DMEM) (Gibco, #1791920) supplemented with 10% FBS.
After 72 hours, when the endothelial cells reached a confluency of 80%, the tissue pieces were removed and the cells were detached using 0.25% trypsin and passaged. The second generation of cells were used for experiments.

| Small interfering RNA (siRNA) transfection
The second generation of endothelial cells were plated in 6-well plates and cultured overnight. Then, the cells were transfected with HIF-1a siRNA (100 nmol/L) or a negative control (NC) siRNA (100 nmol/L) using riboFECT TM CP Reagent and Buffer, according to the manufacturer's protocol. The sequence of the HIF-1a siRNA was TCGACAAGCTTAAGAAAGA. Both the siRNA and transfection reagent were purchased from RiboBio Co., Ltd. (Guangzhou, China).
Briefly, 10 lL siRNA/NC-RNA were diluted in 120 lL 1 9 riboFECT TM CP Buffer and 12 lL riboFECT TM CP Reagent, mixed fully and incubated for 20 minutes at room temperature. Then, this riboFECT TM CP mixture, together with 1858 lL DMEM, was added to each well and incubated at 37°C for 36 hours. Thereafter, cells were processed according to experimental protocols.

| Hypoxia treatment and experimental group
In our previous study, we found that HIF-1a expression peaked at 12 hours of hypoxia, so we performed our experiments at 12 hours of hypoxia in this study. Hypoxia was induced using a three-gas hypoxia incubator chamber (5% CO 2 , 1% O 2 and 94% N 2 ). For drug intervention, cultured CMECs were pre-treated with 0.5 mg/mL QL for 1 hour before hypoxia, which was prepared as in our previous study. 25 Cells were randomly divided into following groups: NC group; Hypoxia group; Hypoxia + HIF-1a siRNA group; Hypoxia + QL group; Hypoxia + QL + HIF-1a siRNA group. Another group of CMECs was incubated with or without AICAR (1 mmol/L, AMPK activator) in the absence or presence of QL.

| Determination of the levels of VEGF secretion
After the cell culture, supernatants were collected by centrifugation at 300 g for 10 minutes and the levels of VEGF secretion were detected using a commercial rat VEGF ELISA kit (Multi Sciences, Hangzhou, China) according to the manufacturer's instructions and calculated as pg/mL protein.

| Lactate dehydrogenase (LDHA) activity and glucose level
In order to estimate the level of glycolysis, cell culture supernatants were collected at the end of the experiment, and LDHA activity (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) and glucose level (BioSino Bio-Technology & Science Inc., Beijing, China) were measured according to the manufacturer instructions, respectively. | 2793 from 5 randomly chosen fields were photographed under a microscope (Olympus). Then, the fluorescence intensity of aggregates and monomers was analysed using ImageJ software using the same method as DHE staining, and the ratio of aggregates/monomers fluorescence intensity was calculated as an indicator of mitochondrial transmembrane potential.

| Measurement of intracellular ATP levels
2.12 | Total RNA extraction and quantitative

RT-PCR
Total RNAs were extracted using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA). One microgram of total RNA was reverse-transcribed to cDNA using a RT reagent kit (TaKaRa, Tokyo, Japan), which was then amplified with SYBR green dye on a CFX Connect TM Real-Time System (Bio-Rad Laboratories, Inc., California, USA).
The primer sequences are listed in Table 1. The relative quantification of mRNA levels was calculated with the software of the PCR system by standard 2 ÀMMCt relative quantification method.

| Western blot analysis
The cells were washed with pre-colding PBS, lysed in lysis buffer on ice and then centrifuged at 14 000 g for 20 minutes at 4°C. A total of 10-24 lg of cell total protein were separated using 10% or 12% SDS-polyacrylamide gel electrophoresis and then trans-

| Statistical analysis
The data were analysed using the SPSS software package. For the comparison between 2 groups, the differences in the mean values were evaluated by Student's t test and the Mann-Whitney U test.
For the comparison among 3 or more groups, differences were determined by one-way ANOVA with Bonferroni post hoc test. A value of P < .05 was considered to be statistically significant. All data were expressed as the means AE SD.

| RESULTS
3.1 | QL promoted CMECs proliferation and inhibited apoptosis under hypoxic condition via up-regulation of HIF-1a In order to examine the effects of QL on CMECs exposed to hypoxic injury and the role of HIF-1a, the proliferation and apoptosis of hypoxic CMECs were analysed upon transfection of either HIF-1a siRNA or control siRNA. CCK-8 assay results demonstrated reduced cell proliferation post-hypoxia, which could be reversed by QL treatment, but this beneficial effect of QL was significantly attenuated by HIF-1a siRNA ( Figure 1A). Meanwhile, CMECs apoptosis was evaluated by caspase-3 activity and the Bcl-2/Bax ratio. Caspase-3 activity increased and the Bcl-2/Bax ratio decreased under hypoxia, while QL significantly reduced hypoxiainduced cell apoptosis, and this protective effect was partly abolished by cotreatment with HIF-1a siRNA ( Figure 1B,C). These data indicated that QL could promote CMECs proliferation and inhibit apoptosis under hypoxic condition, and these effects were at least partly due to HIF-1a up-regulation.
3.2 | QL alleviated mitochondrial dysfunction and reduced ROS production under hypoxic condition in a HIF-1a-dependent manner As hypoxia induces mitochondria dysfunction, we used JC-1 staining to measure mitochondrial transmembrane potential of CMECs.
The results showed that the decreased ratio of aggregates/monomers in hypoxia group could be reversed by QL treatment, and this beneficial effect was partly abolished by HIF-1a siRNA (Figure 2B). We also measured the intracellular ROS generation using the dihydroethidium probe, as hypoxia-induced mitochondria dysfunction leads to ROS production. After hypoxia exposure for 12 hours, the intracellular ROS were increased in CMECs.

| QL increased CMECs ATP production and improved CMECs glucose metabolism under hypoxic condition through glycolytic pathway
ATPlite assay kit was used to detect ATP production, and the results showed that QL increased ATP production in hypoxic CMECs, while HIF-1a siRNA could suppress this effect ( Figure 4A), suggesting that QL could increase ATP production in hypoxic CMECs in a HIF-1adependent manner.
To examine the effect of QL on glucose metabolism of hypoxic CMECs, we collected cell culture supernatants after hypoxia exposure and detected glucose level and LDHA activity of various groups.
The results showed that QL could decrease glucose level and increase LDHA activity through up-regulating HIF-1a activity ( QL promoted proliferation of hypoxic CMECs. Primary rat CMECs were exposed to hypoxia and/or QL for 12 h, and cell proliferation was detected using CCK-8 assay (n = 4/group). (B) and (C) after hypoxia for 12 h, cell apoptosis was measured by caspase-3 activity and Bcl-2/ Bax protein ratio (n = 4/group). Data were expressed as means AE SD. *P < .05; **P < .01 99% goes to the glycolytic pathway in ischaemic endothelial cells. 17 In this study, we found that, under hypoxic condition, the mRNA level of citrate synthase (CS), a key enzyme initiating Krebs cycle, did not significantly change. QL significantly up-regulated its expression, while HIF-1a siRNA had no effect, indicating that QL could enhance glucose aerobic glucose metabolism in endothelial cells exposed to hypoxia in a HIF-1a-independent pathway ( Figure 4G).

| AMPK/mTOR/HIF-1a signalling pathway was involved in QL-mediated induction of HIF-1a and glycolysis in hypoxic CMECs
It has been demonstrated that HIF-1a could be regulated at translational level via AMPK/mTOR pathway under hypoxia. We thus evaluated the role of AMPK/mTOR in QL-induced HIF-1a protein accumulation in CMECs. Our data showed that QL treatment significantly decreased AMPK phosphorylation and increased ATP production, mTOR phosphorylation, GLUT1 and PKM2 expressions in CMECs. Moreover, activating AMPK with AICAR (1 mmol/L, AMPK activator) prior to QL treatment markedly reversed these effects ( Figure 5A-D). These data indicated that AMPK could affect the HIF-1a activity and HIF-1a-associated anaerobic glycolysis. Hypoxia+QL+ HIF-1a siRNA group, however, the inhibition of QL on PHD3 expression was more significant than the promotion of HIF-1a ( Figure 6B).

| DISCUSSION
In the present study, we explore the protective effects and related mechanisms of QL on hypoxic CMECs in the absence or presence Normal cell metabolism depends on the regulatory delivery of fuel substrates and oxygen from vasculature to organs and tissues.
CMECs, a special cell type derived from coronary microvessels, play a major role in promoting angiogenesis and delivering oxygen to myocardium. The fact that HIF-1a is elevated under hypoxic conditions has been confirmed in most types of endothelial cells, for example, human/rat cardiac microvascular endothelial cells and human umbilical vein endothelial cells (HUVECs). [28][29][30] As is known, activated HIF-1a and its downstream effector VEGF are key regulators which can protect endothelial cells from hypoxia-induced impairment and promote angiogenesis and cell migration. 7 In the present study, however, with long-time hypoxia, the protective mechanism initiated by adaptive HIF-1a elevation seems decompensated and not sufficient. Our results showed that extreme hypoxia aggravated cell apoptosis by increasing ROS production and release, which caused hyperpermeability of the inner mitochondrial membrane and led to impaired proliferation and capillary-like tube formation of CMECs. Previous studies also found that, with hypoxia time gradually increased, endothelial cells oxidative stress, apoptosis and autophagy also increased. 31,32 In CMECs hypoxia-reoxygenation model, decreased cell viability, migration and angiogenesis were also observed, while HIF-1a and VEGF expressions were elevated. 33 Fortunately, in the present study, we found that QL remedy upregulated HIF-1a and VEGF expressions, which directly facilitated adaptation and survival of endothelial cells exposed to hypoxia.
Taken together, it can be concluded that QL promotes proliferation and angiogenesis of hypoxic CMECs through up-regulating HIF-1a/ VEGF signalling.
Endothelium, located between blood and tissue, is an essential determinant element of glucose metabolism in various whole organs including brain and heart. The oxygen-sensing transcription factor HIF-1a is a key regulator of endothelial metabolism. 10  glycolysis for ATP production, which can be enhanced by QL through up-regulating HIF-1a and its downstream genes.
The dependence of CMECs on glycolysis has multiple advantages. Firstly, most glycolytic key enzymes are located in the cytosol, which favours immediate ATP supplies for actin rearrangement to facilitate angiogenesis and vesicular secretion and generates required intermediates necessary for cell growth and migration. 36,37 In the present study, we confirmed that down-regulated HIF-1a and glycolytic key enzymes expressions in the Hypoxia+HIF-1a siRNA group could further reduce proliferation and tube formation of the inhibitory effect of QL on PHD3 was more significant than HIF-1a promotional effect in the Hypoxia + QL group.
In conclusion, our results indicate that QL can protect CMECs against hypoxic injury by promoting angiogenesis and proliferation, reducing ROS generation and apoptosis and increasing polarized mitochondrial membrane potential and ATP production. The beneficial effects of QL are mainly mediated by up-regulating HIF-1a activity and glycolysis. We also demonstrated that QL promotes the mRNA and protein expressions of HIF-1a in hypoxic CMECs by regulating AMPK/mTOR/HIF-1a pathway and deduced that QL maintains HIF-1a stabilization by decreasing PHD3 expression.

ACKNOWLEDG EMENTS
This study was supported by grants from the National Basic

CONFLI CT OF INTEREST
The authors confirm that there is no conflict of interests.