Thymoquinone affects hypoxia-inducible factor-1α expression in pancreatic cancer cells via HSP90 and PI3K/AKT/mTOR pathways

Abstract

BACKGROUND

Pancreatic cancer (PC) is associated with some of the worst prognoses of all major cancers. Thymoquinone (TQ) has a long history in traditional medical practice and is known for its anti-cancer, anti-inflammatory, anti-fibrosis and antioxidant pharmacological activities. Recent studies on hypoxia-inducible factor-1α (HIF-1α) and PC have shown that HIF-1α affects the occurrence and development of PC in many aspects. In addition, TQ could inhibit the development of renal cancer by decreasing the expression of HIF-1α. Therefore, we speculate whether TQ affects HIF-1α expression in PC cells and explore the mechanism.

AIM

To elucidate the effect of TQ in PC cells and the regulatory mechanism of HIF-1α expression.

METHODS

Cell counting kit-8 assay, Transwell assay and flow cytometry were performed to detect the effects of TQ on the proliferative activity, migration and invasion ability and apoptosis of PANC-1 cells and normal pancreatic duct epithelial (hTERT-HPNE) cells. Quantitative real-time polymerase chain reaction and western blot assay were performed to detect the expression of HIF-1α mRNA and protein in PC cells. The effects of TQ on the HIF-1α protein initial expression pathway and ubiquitination degradation in PANC-1 cells were examined by western blot assay and co-immunoprecipitation.

RESULTS

TQ significantly inhibited proliferative activity, migration, and invasion ability and promoted apoptosis of PANC-1 cells; however, no significant effects on hTERT-HPNE cells were observed. TQ significantly reduced the mRNA and protein expression levels of HIF-1α in PANC-1, AsPC-1, and BxPC-3 cells. TQ significantly inhibited the expression of the HIF-1α initial expression pathway (PI3K/AKT/mTOR) related proteins, and promoted the ubiquitination degradation of the HIF-1α protein in PANC-1 cells. TQ had no effect on the hydroxylation and von Hippel Lindau protein mediated ubiquitination degradation of the HIF-1α protein but affected the stability of the HIF-1α protein by inhibiting the interaction between HIF-1α and HSP90, thus promoting its ubiquitination degradation.

CONCLUSION

The regulatory mechanism of TQ on HIF-1α protein expression in PC cells was mainly to promote the ubiquitination degradation of the HIF-1α protein by inhibiting the interaction between HIF-1α and HSP90; Secondly, TQ reduced the initial expression of HIF-1α protein by inhibiting the PI3K/AKT/mTOR pathway.

Key Words: Thymoquinone; Pancreatic cancer; Hypoxia-inducible factor-1α; PI3K/AKT/mTOR; HSP90

Core Tip: Thymoquinone (TQ) could significantly inhibit the proliferative activity, invasion and migration ability and promote the apoptosis of pancreatic cancer (PC) cells. TQ could inhibit hypoxia-inducible factor-1α (HIF-1α) expression in PC cells. The mechanism of TQ on HIF-1α protein expression in PC cells was to promote the ubiquitination degradation of HIF-1α protein by inhibiting the interaction between HIF-1α and HSP90 and reduce the initial expression of HIF-1α protein by inhibiting the PI3K/AKT/mTOR pathway.

INTRODUCTION

Pancreatic cancer (PC) is associated with some of the worst prognoses of all major cancers. According to the latest Global Cancer Report for 2020, PC accounts for 2.6% of new cancer cases worldwide and 4.7% of cancer-related deaths[1]. According to the data released by the American Cancer Society in 2023, PC accounts for 3% of all new cancer cases, and the morbidity in women and men rank 8^th and 10^th respectively, while PC-related deaths account for 8% of all cancer deaths, ranking 4^th for both men and women[2]. In the European Union countries, the incidence and mortality rates of PC rank 8th and 6th, respectively, among all malignancies. According to the current rising trend of PC incidence, that of 2040 will increase by about 30% compared to the current rates[3]. The early metastasis, low operation rate, and poor drug treatment effect in PC are the main reasons that PC has such a high mortality rate. Therefore, recent researches have aimed to present new options for treatment, including those based in traditional medicine.

Plant-derived drugs are often low toxicity, easily accessible, and widely applicable, and their extracts have played an increasingly important role in cancer prevention and treatment. Nearly 80% of cancer treatment drugs are extracts or derivatives of traditional medicinal plant ingredients[4]. With the rapid development of modern medical technology, scientists have begun to focus on traditional drugs that can be combined with modern medicine, identifying and extracting their useful active ingredients. Thymoquinone (TQ) extracted from black seed is a recent and typical example of the successful development of traditional drugs. Black seed has a long history in traditional medical practice and is known for its anti-cancer, anti-inflammatory, anti-fibrosis, and antioxidant pharmacological activities[5].

Hypoxia-inducible factor-1 (HIF-1), first reported in 1991, is an oxygen-dependent transcription factor and a key factor in regulating cell adaptation to hypoxia[6]. It is composed of an oxygen-regulated α subunit (HIF-1α) and a constitutionally expressed β subunit (HIF-1β), and HIF-1α is the key subunit for its function. Recent studies on HIF-1α and PC have shown that HIF-1α affects the occurrence and development of PC in many aspects, such as cancer cell survival and proliferation[7], apoptosis[8], autophagy[9], invasion and metastasis[10], metabolic reprogramming[11], tumor stem cells[12], chemotherapy and radiotherapy resistance[13], angiogenesis[14], immune evasion and fibroplasia[15,16].

Studies on the effect of TQ on HIF-1α are rare, with only two articles published so far. Lee et al[17] demonstrated that TQ promoted apoptosis of renal cancer cells through the HIF-1α-mediated glycolysis pathway and that the underlying mechanisms might be related to the inhibition of ubiquitination and HSP90 to reduce HIF-1α stability[17]. Amin et al[18] revealed the potential role of TQ as a HIF-1α activator to reduce cell death, modulate autophagy, and decrease the infarct volume after ischemic stroke onset. The neuroprotective effect of TQ is achieved by decreasing inflammation and increasing angiogenesis and neurogenesis via induction of the HIF-1α/vascular endothelial growth factor/nuclear factor erythroid 2-related factor 2/heme oxygenase 1/tyrosine receptor kinase B/phosphoinositide 3-kinases pathway[18]. At present, no other study has investigated the effect of TQ on HIF-1α in PC. Given TQ’s potential for extensive anti-cancer, anti-fibrosis, and anti-oxidation/pro-oxidation effect and the important features of interstitial fibrosis and hypoxia in PC, this study aimed to investigate the effects of TQ on HIF-1α in PC and the related underlying mechanisms.

MATERIALS AND METHODS

Materials

The following materials were used in this study: TQ (HY-D0803; MCE; New Jersey; United States); 26S proteasome inhibitor (HY-132598; MCE); geldanamycin (GA; HY-15230; MCE); cycloheximide (CHX) (S7418; SELLECK; Houston; United States); dimethyl sulfoxide (DMSO; Sigma; St. Louis; United States); fetal bovine serum (FBS; GIBCO; Invitrogen; Carlsbad; United States); Binding Buffer (XP2; Omega; Norcross; United States); trypsin-EDTA (GNM25200; Gino Biomedical Technology Co, Ltd; Hangzhou; China); Trypsin Solution without EDTA (C0205; Beyotime Biotech Co., Ltd; Shanghai; China); IP cell lysate (AS1003; Aspen Biotechnology Co., Ltd; Wuhan; China); SDS-PAGE gel preparation kit (AS1012; Aspen Biotechnology Co., Ltd; Wuhan; China); RPMI-1640 medium (GIBCO); Trizol (15596026; Ambion; Austin; United States); iScript® II Q RT SuperMix for quantitative real-time polymerase chain reaction (qPCR) (+gDNA wiper) (R233-01; VAZYME; Nangjing; China); HiScript II Q Select RT SuperMix for qPCR (+gDNA wiper) (R233-01; VAZYME); SYBR Green Master Mix (Q111-02; VAZYME); Taq Plus DNA Polymerase (ET105-01; TIANGEN; Beijing; China); DL2000 DNA Marker (MD114-02; TIANGEN); primary antibodies: Anti-HIF1α (20960-1-AP; PROTEINTECH; Chicago; United States), anti-OH-HIF1α (3434T; CST; Danvers; United States), anti-phosphatidyl inositol-4,5-bisphosphate-3-kinase-p85α (PI3K-p85α; 60225-1-Ig; PROTEINTECH), anti-protein kinase B (Akt; 60203-2-Ig; PROTEINTECH), anti-mammalian target of rapamycin (mTOR; 66888-1-Ig; PROTEINTECH), anti-S6 kinase (S6K; 14485-1-AP; PROTEINTECH), anti-eukaryotic translation initiation factor 4E binding protein p70 S6 kinase (4E-BP1; 60246-1-Ig; PROTEINTECH), anti-eukaryotic translation initiation factor 4E (eIF-4E; 66655-1-Ig; PROTEINTECH), and anti-von Hippel Lindau (VHL) (16538-1-AP; PROTEINTECH); Cell counting kit-8 (CCK-8) kit (Beyotime Biotech Co., Ltd; Shanghai; China); Annexin V-FITC Apoptosis Detection Kit (AO2001-02P-G; Tianjin Sanjian Biotechnology Co., Ltd; Tianjin; China); phosphate-buffered saline (PBS, Gino Biomedical Technology Co, Ltd; Hangzhou; China); and tris-buffered saline (TBS; Gino Biomedical Technology Co, Ltd).

Cell culture

The following cells were used: PANC-1 (20180418-01; BIOWING; Shanghai; China), cultured in a PANC-1 cell culture medium comprising 89% DMEM/F-12 +10% FBS + 1% P/S; AsPC-1 (20200612-02; BIOWING) and BxPC-3 (20201111-03; BIOWING), cultured respectively in AsPC-1 cell and BxPC-3 cell culture medium comprising 89% RPMI 1640 +10% FBS + 1% P/S; and normal pancreatic duct epithelial (hTERT-HPNE) (20190905-03; BIOWING) cultured in a hTERT-HPNE cell specific culture medium. Cell culture was performed under normoxic conditions (5% CO₂, 21%O₂, 37 °C) and hypoxic conditions (5% CO₂, 1% O₂, 37 °C). Briefly, 24 h after culture, the culture medium was changed, and the cells were collected after reaching 80% confluence. Next, the complete cell medium was mixed with trypsin-EDTA and heated in a 37 °C water bath for 15 min. Thereafter, the culture medium in a 25 cm² culture bottle was removed by gently washing the cell layer twice with 2 mL PBS solution, followed by removal of the PBS. Next, 1 mL of trypsin-EDTA was added to infiltrate the cell layer. Thereafter, 3 mL of complete medium was added to neutralize trypsin-EDTA, and then the cells were gently blown with a pipette to disperse them in the culture medium. Next, the culture medium containing cells in the culture bottle was placed in a centrifuge tube and centrifuged at 300 g for 5 min. The supernatant was discarded, the cells were re-suspended with a new complete medium, counted, and inoculated on a culture plate. The cells were then cultured in a 5% CO₂ incubator at 37 °C (under normoxic conditions). The O₂ parameter of the three-gas incubator was adjusted to 1%, N₂ was injected into the three-gas incubator, and cells requiring hypoxic treatment were cultured in the three-gas incubator under hypoxic conditions. Finally, following their adherence to the walls, the cells were treated according to the corresponding groups.

CCK-8

The cells in logarithmic growth stage (PANC-1, AsPC-1, BxPC-3, and hTERT-HPNE) were trypsinized and adjusted to a cell concentration of 1 × 10⁵/mL. Then, 100 μL of the cell suspension was inoculated into each well of a 96-well plate at a density of 1 × 10⁴ cells/well. The cultures were maintained under normoxia or hypoxia conditions until cell attachment. Subsequently, the medium in each group was replaced with 100 μL serum-free basic medium containing 1% BSA, followed by starvation for 12 h. Thereafter, the basic medium in each group was changed to 100 μL of respective drug samples at certain concentrations (TQ: 5, 10, 15, 20, 25, 30, and 35 μM). The control group received a solvent-containing medium, while corresponding blank wells without cells were included. Next, the cells were incubated for 4 h after adding 10 μL of CCK-8 solution to each well, and then the absorbance at 450 nm was measured using an enzyme-linked instrument. Cells treated with solvent (diluted ethanol) served as the control group, while blank zeroing wells represented the blank group. Cell survival rate under drug treatment was calculated according to the following formulae:

Inhibition rate % = [(control group - blank) - (experimental group - blank)]/(control group − blank) × 100%.

Proliferation rate % = (experimental group × blank)/(control group × blank) × 100%.

Cell scratch assay

In brief, parallel lines were evenly drawn at the bottom of a 6-well plate (each 1 cm apart). Next, 2.5 × 10⁵ cells /mL were prepared, and 2 mL was inoculated into each small well. When the cells covered the bottom of the well, a scratch was made using a gun tip in the direction perpendicular to the parallel lines and then the cells were washed wish PBS. Thereafter, serum-free medium was added, and pictures were taken using an inverted microscope. The intersection of scratches and parallel lines was recorded at this time as 0 h. Thereafter, the cells were cultured at 37 ℃ for 24 h in a three-gas incubator (with O₂ content adjusted to 1%), and then the intersection point noted at 0 h was photographed and recorded again.

Transwell migration assay

Cells were prepared into a 10⁵/mL suspension, and then 1 mL was centrifuged at 1500 g for 5 min, and the supernatant was discarded. Thereafter, serum-free medium was added, mixed, and then 200 μL was added into the Transwell chamber. Next, 500 μL of complete culture medium containing 10% FBS was placed in a 24-well plate, and then the Transwell chamber was placed in the plate and incubated in a three-gas incubator (with the O₂ content adjusted to 1%) at room temperature (20 ℃-25 ℃) until 8 h. Next, the chamber was removed and the medium was washed off with PBS. Thereafter, the cells were collected, fixed with paraformaldehyde for 20 min, washed twice with PBS, stained with 1% crystal violet solution for 10 min, and photographed under an inverted microscope.

Transwell invasion assay

The Matrigel matrix glue was diluted with medium at a 1:3 ratio, and then 50 μL was added to the Transwell chamber and dried for use. The other steps were the same as those for the Transwell migration assay described above.

Flow cytometry (apoptosis)

Cells were collected by centrifugation at 300 g at 4 ℃ for 5 min after digestion with pancreatic enzyme without EDTA. Next, the cells were washed twice with PBS and centrifuged at 300 g for 5 min, followed by collection of the precipitate and suspension with a 1 × Binding Buffer (300 μL).Thereafter, 5 μL of Annexin V-FITC (fluorescein) was added and incubated away from light for 10 min. Next, 5 μL of propidine iodide (PI) was added, followed by incubation away from light for 5 min. The cells were then analyzed by using a FACS software, with the following parameters: FITC, excitation at 494 nm and emission at 520 nm; PI, excitation at 493 nm and emission at 636 nm.

Flow cytometry (cell cycle)

Cells digested with EDTA-free trypsinwere were collected, gently blown into a single cell suspension, centrifuged at 300 g for 5 min, and the supernatant was discarded. The cells were then resuspended in 1 mL PBS, and the supernatant was discarded. Further, the cells were resuspended in 100 μL PBS, followed by addition of 300 μL pre-cooled ethanol, mixing, and fixation at 4 ℃ for 24 h. Thereafter, the cells were centrifuged at 300 g for 5 min, the supernatant was discarded, and then washed with PBS, followed by another resuspension. Next, 10 μL of propidium iodide storage solution and 10 μL of RNase A solution were added to 0.5 mL of staining buffer, mixed, and used to resuspend cells, followed by incubation at 37 ℃ for 30 min without light. Flow cytometry was performed on a computer, and the FACS software was used for data analysis.

qPCR

RNA was extracted by using Trizol, and the OD260, OD280, and OD260/OD280 (required to be between 1.8 and 2.0) values were determined to calculate the purity and concentration, followed by reverse transcription into cDNA (Tables 1 and 2), which was used for PCR or stored below −20 ℃ (7 d). The PCR primer sequences were shown in Table 3. The real-time fluorescence quantitative PCR reaction system and reaction procedures were shown in Tables 4 and 5, respectively. The PCR products were subjected to agarose gel electrophoresis, and the bands were analyzed using Image J software.

Table 1 Genomic DNA removal.

Constituent	Dose
Template RNA	1.76 μg
4 × gDNA wiper Mix	4 μL
RNase-free ddH2O	16 μL

Incubate at 42 ℃ for 2 min.

Table 2 Preparation of reverse transcriptional response system.

Constituent	Dose
5 × HiScript II Select qRT SuperMix II	4 μL
Reaction liquid from step 1	16 μL
Reverse transcriptase	1 μL

Incubate at 50 ℃ for 15 min; Incubate at 85 ℃ for 5 s; Incubate at 4 ℃ for 10 min.

Table 3 Primer sequence.

Gene	Primer	Sequence (5'-3')	PCR product
Homo GAPDH	Upstream	TCAAGAAGGTGGTGAAGCAGG	115 bp
	Downstream	TCAAAGGTGGAGGAGTGGGT
Homo HIF-1α	Upstream	GTGGCGAAGATGGTCAAGTC	116 bp
	Downstream	GGAGTGCCCTTGTTGAGGTGTT

HIF-1α: Hypoxia-inducible factor-1α; PCR: Polymerase chain reaction.

Table 4 Real-time fluorescence quantitative polymerase chain reaction system.

Constituent	Dose
cDNA	4 μL
Forward Primer (10 μM)	0.4 μL
Reverse Primer (10 μM)	0.4 μL
SYBR Green Master Mix	10 μL
50 × ROX Reference Dye 2	0.4 μL
Taq Plus DNA Polymerase	1 μL
RNase-free ddH2O	25 μL

Table 5 Real-time fluorescence quantitative polymerase chain reaction program.

Item	Temperature	Time	Cycle-index
Predegeneration	95 ℃	10 min	1
Denaturation	95 ℃	15 s	40
Annealing elongation	60 ℃	60 s	40
Melting curve acquisition	95 ℃	15 s	1
	60 ℃	60 s	1
	95 ℃	15 s	1

Western blot

Total protein was isolated, and its concentration determined by using the BCA protein concentration determination kit, and then subjected to SDS-PAGE (Tables 6 and 7). The proteins were then transferred onto a PVDF membrane, which was incubated with a sealing liquid for 1 h. Next, the membrane was incubated with a diluted primary antibody overnight at 4 ℃ and washed thrice with TBST (TBS with Tween-20) (5 min each). Thereafter, the membrane was incubated with a diluted secondary antibody at room temperature for 30 min and washed four times with TBST (5 min each) on a shaking bed. The ECL mixture was added to the protein side of the membrane, and then the membrane was exposed to a darkroom for development and fixing. The AlphaEaseFC software was used to analyze the optical density value of the target tape.

Table 6 Preparation of sodium dodecyl sulfate-polyacrylamide gel electrophoresis separation glue.

Reagent	Separation glue concentration (%)
	8%	10%	12%	15%	18%	20%
H2O	4.63 mL	4 mL	3.3 mL	2.3 mL	1.3 mL	0.63 mL
30% acrylamide (29: 1)	2.67 mL	3.3 mL	4 mL	5 mL	6 mL	6.67 mL
1.5M Tris-HCl (pH 8.8)	2.5 mL	2.5 mL	2.5 mL	2.5 mL	2.5 mL	2.5 mL
10%SDS	0.1 mL	0.1 mL	0.1 mL	0.1 mL	0.1 mL	0.1 mL
Ammonium persulfate	0.1 mL	0.1 mL	0.1 mL	0.1 mL	0.1 mL	0.1 mL
TEMED	5 μL	5 μL	5 μL	5 μL	5 μL	5 μL
Bulk volume	10 mL	10 mL	10 mL	10 mL	10 mL	10 mL

SDS: Sodium dodecyl sulfate.

Table 7 Preparation of sodium dodecyl sulfate-polyacrylamide gel electrophoresis concentrated glue.

Reagent	Concentration
H2O	2 mL	3 mL	4 mL	6 mL
30% acrylamide (29:1)	0.5 mL	0.75 mL	1 mL	1.5 mL
1M TRIS-HCl (pH 6.8)	0.5 mL	0.75 mL	1 mL	1.5 mL
10%SDS	40 μL	60 μL	80 μL	120 μL
Ammonium persulfate	30 μL	45 μL	60 μL	90 μL
TEMED	4 μL	6 μL	8 μL	12 μL
Bulk volume	3 mL	4.5 mL	6 mL	9 mL

SDS: Sodium dodecyl sulfate.

Construction of plasmid vectors

cDNA cloning vector of VHL gene (with restriction sites) was purchased and digested with EcoR I and BamH I (Table 8), and the digested products were recovered for electrophoresis detection and stored at −20 °C for future use. To obtain the vector fragments, bacteria were transformed with the digested products, cultured, and centrifuged, and the supernatant was discarded, followed by collection of the bacteria. Next, 250 μL of Solution A was added to suspend the bacteria, followed by the addition of 250 μL Solution B and 250 μL of Solution C. This mixture was then centrifuged at 12000 rpm for 10 min at room temperature, and then the supernatant was collected. Thereafter, 1/10 of the volume of the supernatant was added into an Endo-Remove Buffer, mixed well, and incubated in an ice bath for 10 min, followed by further incubation at 42 °C for 5–10 min, centrifugation at 12000 rpm for 5 min at room temperature, and collection of the supernatant into a new EP tube. Thereafter, anhydrous ethanol (3 times the volume of the mixture) was added, mixed well, and then set aside. Next, the mixture was added to the adsorption column P, centrifuged at 12000 rpm for 1 min at room temperature, and then the waste liquid was discarded. Next, 600 μL of Wash Buffer solution containing anhydrous ethanol was added, the waste liquid was discarded, and then the Wash Buffer residue was completely shaken off. Thereafter, the adsorption column P was placed in a new EP tube and then set aside. Next, an appropriate amount of Elution Buffer or ddH₂O was added to the middle of the adsorption column P, followed by centrifugation to obtain the plasmid. For carrier double enzyme digestion and recovery, EcoR I and BamH I enzyme digestion carriers (Table 9) were used to recover the enzyme digestion products for electrophoresis detection and stored at −20 ℃ for subsequent use. For construction of the recombinant overexpression vector, after the completion of treatment according to the processing flow of the connection system (Table 10), 10 μL connection product was added to 100 μL of competent cells, which were incubated in an ice bath, heat-stimulated, and ice bathed successively. After adding 300 μL of LB liquid medium, the shaking bed was shaken at room temperature for 1 h. Then, 100 μL of the suspension was coated on the LB/Kan plate and cultured overnight at 37 ℃. A single colony was inoculated into 15 mL of LB medium and shaken overnight at 37 ℃. For recombinant vector recovery and identification, overnight cultured bacteria were identified (Table 11), and recombinant vector PCR amplification was performed, and then plasmid DNA was sent to Wuhan Aspen Biotechnology Co., LTD for sequencing.

Table 8 Objective fragment double enzyme digestion process.

Constituent	Dose
1 × Tango Buffer	2.0 μL
BamH I	0.5 μL
EcoR I	0.5 μL
Objective fragment	8.0 μL
ddH2O	9.0 μL
Total dose	20.0 μL

Water bath at 37 ℃ for 2 h.

Table 9 Carrier double enzyme digestion process.

Constituent	Dose
1 × Tango Buffer	2.0 μL
BamH I	0.5 μL
EcoR I	0.5 μL
Carrier plasmid DNA	8.0 μL
ddH2O	9.0 μL
Total dose	20.0 μL

Water bath at 37 ℃ for 2 h.

Table 10 Connection system processing.

Constituent	Dose
2 × Sosoo Cloning MIX	5.0 μL
pLVX-SV40-IRES-EGFP-Puro	2.0 μL
Target fragment	2.0 μL
ddH2O	1.0 μL

Recombined at 50 ℃ for 30 min.

Table 11 Culture solution identified by polymerase chain reaction.

Constituent	Dose
EntilinkTM PCR Master Mix	25.0 μL
Forward Primer (10 μM)	2.0 μL
Reverse Primer (10 μM)	2.0 μL
Template	Appropriate amount
ddH2O	Up to 50 μL

37 ℃ water bath overnight. PCR: Polymerase chain reaction.

Construction of ShRNA vector

For shRNA sequence acquisition, a DNA sequence corresponding to the shRNA was provided by Wuhan Aspen Biotechnology Co., Ltd. The reaction system and reaction conditions for shRNA annealing are shown in Tables 12 and 13.

Table 12 shRNA annealing reaction system.

Constituent	Dose
10 × annealing buffer	5 μL
Forward Primer (10 μM)	10 μL
Reverse Primer (10 μM)	10 μL
ddH2O	Up to 50 μL

Table 13 shRNA annealing polymerase chain reaction conditions.

Procedure	Temperature	Time	Cycle-index
Predegeneration	95 ℃	2 min	10
Degeneration	95 ℃	30 s
Anneal	60 ℃	30 s
Final insulation	4 ℃	Unlimited

For electrophoresis detection, 1 g of agarose powder was added to 100 mL of 1 × TAE buffer solution and dissolved using a microwave. The temperature was reduced to about 60 °C, and 10 μL of type I nucleic acid dye was added to the rubber tank. Thereafter, 1 × TAE buffer solution was added to the electrophoresis tank, and then the marker and sample were added, and electrophoresis was performed at 120 V for 20 min. Thereafter, the rubber block was taken for detection under an ultraviolet detector.

For construction of pLVX-shRNA vector, bacteria were inoculated, cultured, centrifuged, and then the supernatant was discarded, followed by collection of the bacteria. Thereafter, 500 μL Solution A was added to suspend the bacteria, followed by the addition of 500 μL of Solution B and 700 μL of Solution C. Thereafter, the mixture was centrifuged at 12000 rpm for 10 min at room temperature and then the supernatant was collected. The supernatant was added to the adsorption column P, centrifuged at 12000 rpm for 1 min at room temperature, and then the waste liquid was discarded. Next, 600 μL of Wash Buffer solution containing anhydrous ethanol was added, and the waste liquid was discarded. Thereafter, Wash Buffer residue was completely shaken off. The adsorption column P was then placed in a new EP tube and set aside. Next, an appropriate amount of Elution Buffer or ddH₂O was added to the middle of the adsorption column P and centrifuged to obtain the plasmid.

The procedures for double-enzyme digestion and recovery process of the vector are the same as those for the VHL overexpressed (OE-VHL) vector construction (Table 9).

For construction of the recombinant plasmid, after completing the treatment according to the recombinant plasmid connection system (Table 14), the steps for the OE-VHL vector construction were followed.

Table 14 Recombinant plasmid linking system.

Constituent	Dose
10 × ligase Buffer	2.0 μL
pLVshRNA-EGFP(2A)Puro double enzyme digestion vector	2.0 μL
Target fragment	2.0 μL
Ligase	0.5 μL
ddH2O	1.0 μL

Recombination at 22 ℃ for 2 h connection.

Co-immunoprecipitation

To extract total cell protein, cells were collected after centrifugation, followed by addition of PBS buffer for re-suspension. The cells were then centrifuged, and the supernatant was discarded. The above procedure was repeated two times to precipitate the cells. Thereafter, immunoprecipitation (IP) cell lysate was added, shaken well, centrifuged after incubation in an ice bath, and the supernatant was collected. Next, 1/10 of the supernatant was taken as input, and the remaining sample was used for IP.

Protein concentration of the sample was determined using the BCA protein concentration determination kit. Magnetic beads were added to the IP cell lysate in a centrifuge tube, and the supernatant was collected after washing on a vertical shaker. For lysate pretreatment, common immunoglobulin G (IgG) and magnetic beads were added to the cell lysate supernatant and incubated at 4 °C for 30 min. For the IP process, antibodies were added to the cell lysate supernatant to precipitate one of the interacting proteins. The common IgG group was used as a negative control and incubated at 4 °C for 1 h. Next, magnetic beads were added, and the supernatant was collected after overnight incubation at 4 °C. Thereafter, the washing buffer was added, and the supernatant was collected after washing on a vertical shaker; this process was repeated three times. After the last washing, the supernatant was discarded, and 2 × SDS loading buffer containing thioethanol was added, boiled for 10 min, and the samples were stored at −20 °C. The procedures for SDS-PAGE, membrane transfer, antibody incubation, chemiluminescence detection, and results analysis were the same as those described under the western blot experiment.

Statistical analysis

All data were statistically analyzed using the SPSS software (version 26.0). Measurement data were expressed as mean ± SD. A χ² test was performed to compare the counting data between two groups, a t-test was performed to compare the means between two groups, and a one-way analysis of variance was performed to compare the means of the multiple groups. A P value of < 0.05 in these analyses was considered statistically significant.

RESULTS

Effects of TQ on the proliferative activity of PC cells and normal pancreatic ductal epithelial cells

First, we analyzed whether TQ influenced the proliferative activity of human PC cells (PANC-1, AsPC-1, BxPC-3) and hTERT-HPNE cells under hypoxia as HIF-1α is present only under hypoxic conditions. To maintain consistency, the experiments were constructed in hypoxic conditions (O₂ content was 1%). CCK8 assay was performed to determine the effects of TQ on the proliferative activity of PANC-1, AsPC-1, and BxPC-3 cells. The drug treatment duration was 8 h. Experimental results confirmed that different concentrations of TQ (0, 5, 10, 15, 20, 25, 30, and 35 µM) showed different proliferation inhibition rates in the three PC cell lines and a concentration dependence (the higher the TQ concentration, the lower the cell proliferation rate) (Figure 1). The half maximal inhibitory concentration (IC₅₀) of TQ in PANC-1, AsPC-1 and BxPC-3 cells was 11, 11, and 26 μM, respectively. These IC₅₀ values were used in subsequent experiments. In addition, TQ had no significant effect on the viability of hTERT-HPNE cells (Figure 1D). These results indicated that TQ had no obvious inhibitory effect on the proliferation of hTERT-HPNE cells, but had a significant inhibitory effect on the proliferation of PC cells, suggesting that TQ could be used as a potential therapeutic agent to inhibit the proliferation of PC cells. Although TQ caused significant inhibition of the cytovirulence of the three human PC cell lines, because of limited funding, we selected the PANC-1 cells as the representative to verify the effects of TQ on cell migration, invasion, cell cycling, and apoptosis.

Figure 1 Cell counting kit-8 assay was performed to determine the effects of different concentrations of Thymoquinone on the proliferation rate of pancreatic cancer cells and pancreatic duct epithelial cells under hypoxia condition. A: Cell counting kit-8 (CCK-8) assay was performed to detect the effects of Thymoquinone (TQ) on the relative proliferation rate of PANC-1 cells under hypoxia condition; B: CCK-8 assay was performed to detect the effects of TQ on the relative proliferation rate of AsPC-1 cells under hypoxia condition; C: CCK-8 assay was performed to detect the effects of TQ on the relative proliferation rate of BxPC-3 cells under hypoxia condition; D: CCK-8 assay was performed to detect the effects of TQ on the relative proliferation rate of normal pancreatic duct epithelial cells under hypoxia condition. TQ: Thymoquinone.

Effects of TQ on migration and invasion of PC cells and normal pancreatic ductal epithelial cells

Considering the highly aggressive nature and active migration of PC cells, we further determined whether TQ could inhibit the migration and invasion ability of PC cells, and verified its effects on the migration and invasion of normal pancreatic ductal epithelial cells. Human PC cell line PANC-1 and human normal pancreatic ductal epithelial cell line hTERT-HPNE were used as experimental cells, and the experimental groups were divided into normal control (NC) and TQ groups. The experimental condition was hypoxia and drug treatment duration was 8 h. The experiment was repeated three times. The effects of TQ on the migration ability of PANC-1 cells and hTERT-HPNE were investigated using cell scratch and Transwell migration assays. The migration ability of the PANC-1 cells in the TQ group was significantly lower than that of the NC group (P < 0.01); however, there was no significant change in the migration ability of the hTERT-HPNE cells in the TQ or NC groups (Figure 2). In addition, we verified the effects of TQ on the invasion of PANC-1 cells and hTERT-HPNE cells using the Transwell invasion assay. The invasion ability of the PANC-1 cells in the TQ group was significantly decreased compared with that of the NC group (P < 0.01), but there was no significant difference in the invasion ability of the hTERT-HPNE cells in the TQ or NC groups (Figure 3). These findings indicated that TQ significantly inhibited the migration and invasion of PC cells, but had no significant effect on the migration and invasion of normal pancreatic ductal epithelial cells.

Figure 2 Effects of Thymoquinone on migration ability of normal pancreatic duct epithelial and PANC-1 cells under hypoxia condition. A: The migration ability of normal pancreatic duct epithelial (hTERT-HPNE) cells and PANC-1 cells in normal control (NC) group and Thymoquinone (TQ) group was detected by cell scratch tassay under hypoxia condition; B: Statistical analysis of migration ability of hTERT-HPNE cells and PANC-1 cells in NC group and TQ group under hypoxia condition (cell scratch tassay); C: The migration ability of hTERT-HPNE cells and PANC-1 cells in NC group and TQ group were detected by Transwell assay under hypoxia condition (× 100); D: Statistical analysis of migration ability of hTERT-HPNE cells and PANC-1 cells in NC group and TQ group under hypoxia condition (Transwell assay). ^aP < 0.01. NC: Normal control; TQ: Thymoquinone; hTERT-HPNE: Normal pancreatic duct epithelial.

Figure 3 Effects of Thymoquinone on invasion of normal pancreatic duct epithelial and PANC-1 cells under hypoxia condition (Transwell assay). A: The invasion ability of normal pancreatic duct epithelial (hTERT-HPNE) cells and PANC-1 cells in normal control (NC) group and Thymoquinone (TQ) group were detected by Transwell assay under hypoxia condition (× 100); B: Statistical analysis of invasion ability of hTERT-HPNE cells and PANC-1 cells in NC group and TQ group under hypoxia condition. ^aP < 0.01. NC: Normal control; TQ: Thymoquinone; hTERT-HPNE: Normal pancreatic duct epithelial.

Effects of TQ on cell cycle and apoptosis of PC cells and normal pancreatic ductal epithelial cells

To determine whether TQ could affect the cell cycle and apoptosis of PC cells and normal pancreatic ductal epithelial cells, flow cytometry was performed on PC cell line PANC-1 and normal pancreatic cell line hTERT-HPNE. The experimental grouping, experimental condition, and drug treatment duration were the same as above. The proportion of cells in the S phase of the cell cycle in the TQ group was significantly lower than that in the NC group (P < 0.01), while the proportion of cells in the G0/G1 phase was higher than that in the NC group (P < 0.01). The difference was that the G2/M phase ratio of the PANC-1 cells in the TQ group was significantly higher than that in the NC group (P < 0.01), but the G2/M phase ratio of the hTERT-HPNE cells in the TQ group had no significant change in comparison with that in the NC group (Figure 4). Since the S phase is mainly involved in DNA replication, while the G2/M phase is mainly involved in initiating cell cycle checkpoints and repairing damaged DNA, the possible mechanism of TQ's influence on PC cell cycle was that TQ inhibited the proliferation of cancer cells (reducing the proportion of S phase) and damaged the DNA of cancer cells (increasing the proportion of G2/M phase). Flow cytometry was performed to detect apoptosis; the apoptosis rate of the PANC-1 cells in the TQ group was found to be significantly higher than that in the NC group (P < 0.01), but there was no significant difference between the TQ and NC groups in terms of the apoptosis rate of hTERT-HPNE cells (Figure 5). These results indicated that TQ could significantly promote the apoptosis of PC cells, but had no significant effect on the apoptosis of hTERT-HPNE cells.

Figure 4 Effects of Thymoquinone on cell cycle of normal pancreatic duct epithelial cells and PANC-1 cells under hypoxia condition. A: The cell cycle changes of normal pancreatic duct epithelial (hTERT-HPNE) cells and PANC-1 cells in normal control (NC) group and Thymoquinone (TQ) group were analyzed by flow cytometry; B: Statistical analysis of cell cycle changes of hTERT-HPNE cells in NC group and TQ group under hypoxia condition; C: Statistical analysis of cell cycle changes of PANC-1 cells in NC group and TQ group under hypoxia condition. ^aP < 0.01. NC: Normal control; TQ: Thymoquinone; hTERT-HPNE: Normal pancreatic duct epithelial.

Figure 5 Effects of Thymoquinone on apoptosis of normal pancreatic duct epithelial cells and PANC-1 cells under hypoxia condition. A: Flow cytometry was performed to detect the apoptosis of normal pancreatic duct epithelial (hTERT-HPNE) cells and PANC-1 cells in normal control (NC) group and Thymoquinone (TQ) under hypoxia condition; B: Statistical analysis of apoptosis of hTERT-HPNE cells and PANC-1 cells in NC group and TQ group under hypoxia condition. ^aP < 0.01. NC: Normal control; TQ: Thymoquinone; hTERT-HPNE: Normal pancreatic duct epithelial.

Effects of TQ on HIF-1α expression in PC cells

Since the effects of TQ on the proliferating activity, migration and invasion ability and apoptosis of PC cells were confirmed, we verified whether it had any effects on the expression of HIF-1α. PCR and western blot assay were performed to determine the mRNA and protein expression of HIF-1α respectively. PC cell lines (PANC-1, AsPC-1, and BxPC-3) were used as experimental cells and divided into NC and TQ groups. The experimental condition was hypoxia, and drug treatment duration was 8 h. The experiment was repeated three times. Compared with the NC group, the mRNA and protein expression levels of HIF1α in PANC-1, AsPC-1, and BxPC-3 cells in the TQ group were significantly decreased (P < 0.01) (Figure 6), indicating that TQ could inhibit the expression of HIF-1α at the transcriptional level. Since TQ significantly inhibited the expression of HIF-1α in three human PC cell lines, we then selected gemcitabine-resistant PANC-1 cells as the representative of human PC cell lines to further verify the specific mechanisms of TQ's influence on HIF-1α protein expression.

Figure 6 Effects of Thymoquinone on hypoxia-inducible factor-1αmRNA and protein expression in PANC-1, AsPC-1 and BxPC-3 cells under hypoxia condition. A: Statistical analysis of hypoxia-inducible factor-1α (HIF-1α) mRNA expression of PANC-1, AsPC-1 and BxPC-3 cells in normal control (NC) group and Thymoquinone (TQ) group under hypoxia condition; B: Western blot assay was performed to detect HIF-1α protein expression of PANC-1, AsPC-1 and BxPC-3 cells in NC group and TQ group under hypoxia condition; C: Statistical analysis of HIF-1α protein expression of PANC-1, AsPC-1 and BxPC-3 cells in NC group and TQ group under hypoxia condition. ^aP < 0.01. NC: Normal control; TQ: Thymoquinone; HIF-1α: Hypoxia-inducible factor-1α.

Mechanisms underlying the regulation of HIF-1α protein by TQ

Effects of TQ on PI3K/AKT/mTOR in the HIF-1α upscream pathway: The above experiments proved that TQ inhibited the expression of the HIF-1α protein in PC cells. The specific regulatory mechanisms of TQ affecting the HIF-1α protein expression were then considered. We analyzed the effects of TQ on PI3K-p85α (phosphoinositide 3-kinase-p85α), Akt (protein kinase B), mTOR (mammalian target of rapamycin), S6K (S6 kinase), 4E-BP1 (eukaryotic translation initiation factor 4E binding protein 1), and eIF-4E expression. Using PANC-1 as experimental cells, the experiment was divided into NC and TQ groups. The experimental condition was hypoxia, the drug treatment duration was 8 h, and the experiment was repeated three times. Compared with the NC group, the expression levels of p-PI3K-P85α, p-Akt, p-mTOR, p-S6K, p-4E-BP1, and eIF-4E in the TQ group were significantly down-regulated (Figure 7). This suggested that TQ might influence the initial expression of the HIF-1α protein in PANC-1 cells by inhibiting the PI3K/AKT/mTOR pathway.

Figure 7 Effects of Thymoquinone on PI3K/AKT/mTOR upstream pathway of hypoxia-inducible factor-1αprotein under hypoxia condition. A: Western blot assay was performed to detect the expression of p-PI3K-p85α, p-AKT, p-mTOR, p-S6K, p-4E-BP1 and eIF-4E of PANC-1 cells in normal control (NC) group and Thymoquinone (TQ) group under hypoxia condition; B: Statistical analysis of the expression of p-PI3K-p85α, p-AKT, p-mTOR, p-S6K, p-4E-BP1 and eIF-4E of PANC-1 cells in NC and TQ groups under hypoxia condition. ^aP < 0.05; ^bP < 0.01. NC: Normal control; TQ: Thymoquinone.

Effects of TQ on ubiquitination-mediated degradation of HIF-1α: Since the inhibitory effect of TQ on PI3K/AKT/mTOR, the upstream pathway of HIF-1α protein initial expression, was confirmed, the next step was to determine whether TQ had effects on the ubiquitination degradation of the HIF-1α protein in PC cells. MG132 (a proteasome inhibitor) could effectively block the proteolytic activity of the 26S proteasome complex, thus blocking the ubiquitination degradation of the HIF-1α protein[17]. Using PANC-1 as the experimental cells, the experiment was divided into NC, MG132 (dose 20 μM), TQ, and MG132+TQ groups under normoxic conditions with a drug treatment duration of 8 h. The experiment was repeated three times. Compared with the NC group, the expression level of the HIF-1α protein in the MG132 group was significantly increased (P < 0.01) (Figure 8), indicating that MG132 blocked the ubiquitination degradation of HIF-1α protein in PANC-1 cells under normoxic conditions. There was no significant difference in HIF-1α protein expression between the TQ and NC groups (Figure 8). This indicated that TQ had no effect on the final expression of the HIF-1α protein in PANC-1 cells under normoxic conditions (the change in protein level caused by the effect of TQ on the initial expression of HIF-1α protein would be degraded by ubiquitination under normoxic conditions). Compared with the MG132 group, the expression level of the HIF-1α protein in the MG132+TQ group was marginally but not significantly decreased (Figure 8), indicating that the MG132 group blocked the effects of TQ on the final expression level of the HIF-1α protein in PANC-1 cells under normoxic conditions. In other words, it blocked the effects of TQ on the ubiquitination degradation of HIF-1α protein. These results indicated that TQ affected the expression of the HIF-1α protein mainly by promoting its ubiquitination degradation.

Figure 8 Effects of Thymoquinone on ubiquitination degradation of hypoxia-inducible factor-1α protein under normoxia condition. A: Western blot assay was performed to detect hypoxia-inducible factor-1α (HIF-1α) expression of PANC-1 cells in normal control (NC) group, MG132 group, Thymoquinone (TQ) group and MG132+TQ group under normoxia condition; B: Statistical analysis of HIF-1α protein expression of PANC-1 cells in NC group, MG132 group, TQ group and MG132+TQ group under normoxia condition. ^aP < 0.01. NC: Normal control; TQ: Thymoquinone; HIF-1α: Hypoxia-inducible factor-1α.

Effects of TQ on hydroxylation of HIF-1α protein: Since TQ affects the expression level of the HIF-1α protein mainly by promoting its ubiquitination degradation, we identified the specific mechanism by which TQ affected the ubiquitination degradation. First, we examined whether TQ affected the hydroxylation of the HIF-1α protein. PANC-1 cells were used as experimental cells. The expression levels of HIF-1α protein and OH-HIF-1α protein (hydroxylated form of HIF-1α) in the NC and TQ groups were compared under normoxic and hypoxic conditions, respectively. The drug treatment duration was 8 h, and the experiment was repeated three times. There was almost no expression of the HIF-1α protein under normoxic conditions; however, the expression level of the HIF-1α protein was significantly increased under hypoxic conditions, and its expression level was significantly inhibited by TQ (P < 0.01) (Figure 9A and B). Correspondingly, the expression level of the OH-HIF-1α protein was significantly increased under normoxic conditions, but almost no expression of OH-HIF-1α protein was observed under hypoxic conditions (P < 0.01) (Figure 9A and C). This supported the notion that HIF-1α protein could undergo hydroxylation changes under normoxic conditions, before further ubiquitination and degradation. In addition, there was no change in the expression level of the OH-HIF-1α protein between the NC and TQ groups under normoxic conditions (Figure 9A and C), indicating that TQ had no effect on the hydroxylation of the HIF1α protein in PANC-1 cells. These results indicate that the ubiquitination degradation of HIF-1α protein by TQ in PANC-1 cells was not achieved by influencing the hydroxylation of the HIF-1α protein.

Figure 9 Effects of Thymoquinone on hydroxylation of hypoxia-inducible factor-1αprotein. A: Western blot assay was performed to detect hypoxia-inducible factor-1α (HIF-1α) and OH-HIF-1α expression of PANC-1 cells in normal control (NC) group and Thymoquinone (TQ) group under normoxia and hypoxia conditions; B: Statistical analysis of HIF-1α protein expression of PANC-1 cells in NC group and TQ group under normoxia and hypoxia conditions; C: Statistical analysis of OH-HIF-1α protein expression of PANC-1 cells in NC group and TQ group under normoxia and hypoxia conditions. ^aP < 0.01. NC: Normal control; TQ: Thymoquinone; HIF-1α: Hypoxia-inducible factor-1α.

Effects of TQ on ubiquitination degradation of HIF-1α protein mediated by VHL protein: We examined whether TQ affected the ubiquitination degradation of the HIF-1α protein by regulating the HIF-1α binding to the VHL protein (p-VHL). First, PANC-1 cells models of VHL overexpression (Lenti-VHL) and VHL knockdown (sh-VHL) were constructed by performing lentiviral vectors. After the cell models were constructed, the expression levels of the p-VHL in the cell models with Lenti-VHL and knockdown were verified. To verify the expression of the p-VHL in the OE-VHL cell model, the experimental cells were divided into control, overexpressed empty carrier (OE-NC), and OE-VHL groups. The experimental condition was normoxic, the treatment time was 8 h, and the experiment was repeated for 3 times. There was no significant difference in the expression levels of the p-VHL between the OE-NC and control groups, and the expression levels of the p-VHL in the OE-VHL group were significantly higher than those in the OE-NC group (P < 0.05) (Figure 10A and B), indicating that the OE-VHL cell model was effective. To verify the expression of the p-VHL in the sh-VHL cell model, we divided the experimental cells into control, knockdown empty vector (sh-NC), and sh-VHL groups, and the experimental condition, processing time and repetition times were the same as above. There was no significant difference in the expression levels of the p-VHL between the sh-NC group and the control group, and the expression levels of the p-VHL in the sh-VHL group were significantly lower than that in the sh-NC group (P < 0.01) (Figure 10C and D), indicating that the sh-VHL cell model was also effective.

Figure 10  Expression of von Hippel Lindau protein in von Hippel Lindau overexpression and von Hippel Lindau knockdown cell models under normoxia condition. A: Western blot assay was performed to detect von Hippel Lindau (VHL) protein expression in cell models of control group, overexpressed empty carrier (OE-NC) group and VHL overexpressed (OE-VHL) group under normoxia condition; B: Statistical analysis of VHL protein expression in cell models of control group, OE-NC group and OE-VHL group under normoxia condition; C: Western blot assay was performed to detect VHL protein expression in cell models of control group, knockdown empty vector (sh-NC) group and VHL knockdown (sh-VHL) group under normoxia condition; D: Statistical analysis of VHL protein expression in cell models of control group, sh-NC group and sh-VHL group under normoxia condition. ^aP < 0.05; ^bP < 0.01. VHL: von Hippel Lindau; OE-NC: Overexpressed empty carrier; OE-VHL: VHL overexpressed; sh-NC: Knockdown empty vector; sh-VHL: VHL knockdown.

As the successful construction of Lenti-VHL and knockdown cell models was completed, we divided the experiment into the Lenti-VHL, sh-VHL, Lenti-VHL+TQ, and sh-VHL+TQ groups. Experimental condition, drug treatment duration and numbers of repetition were the same as above. The expression levels of the p-VHL in groups with Lenti-VHL (Lenti-VHL group and Lenti-VHL+TQ group) were significantly higher than those in groups with sh-VHL (sh-VHL group and sh-VHL+TQ group) (P < 0.01) (Figure 11A and B). In addition, there was no significant change in the expression levels of p-VHL in the Lenti-VHL+TQ group compared with the Lenti-VHL group. Compared with the sh-VHL group, there was no significant difference in the expression levels of p-VHL in the sh-VHL+TQ group (Figure 11A and B), indicating that TQ had no significant effect on the expression of the p-VHL. Regarding the effects of VHL and TQ on the HIF-1α protein expression, the expression levels of HIF-1α protein in the groups with sh-VHL (sh-VHL group) were significantly higher than that in the groups with Lenti-VHL group (P < 0.01) (Figure 11A and C). This proved that ubiquitin degradation occurred only after the HIF-1α binding to the p-VHL under normoxic conditions. Compared with the sh-VHL group, the expression levels of the HIF-1α protein in the sh-VHL+TQ group were significantly decreased (P < 0.05) (Figure 11A and C), indicating that sh-VHL did not block the effects of TQ on HIF-1α protein expression. These results indicated that the effects of TQ on HIF-1α ubiquitination degradation were independent of the VHL pathway.

Figure 11  Effects of Thymoquinone on ubiquitination degradation of hypoxia-inducible factor-1α protein mediated by von Hippel Lindau under normoxia condition. A: Western blot analysis was performed to detect the expression of von Hippel Lindau (VHL) protein and hypoxia-inducible factor-1α (HIF-1α) protein in Lenti-VHL group, Lenti-VHL + Thymoquinone (TQ) group, VHL knockdown (sh-VHL) group and sh-VHL+TQ group under normoxia condition; B: Statistical analysis of VHL protein expression in the Lenti-VHL group, Lenti-VHL+TQ group, sh-VHL group and sh-VHL+TQ group under normoxia condition; C: Statistical analysis of HIF-1α expression in Lenti-VHL group, Lenti-VHL+TQ group, sh-VHL group and sh-VHL+TQ group under normoxia condition. ^aP < 0.05; ^bP < 0.01. TQ: Thymoquinone; VHL: von Hippel Lindau; sh-VHL: VHL knockdown; HIF-1α: Hypoxia-inducible factor-1α.

Effects of TQ on interaction between HIF-1α and HSP90: In addition to ubiquitination mediated by the p-VHL, HIF-1α degradation is also related to HSP90-mediated stability[19]. After establishing that the effects of TQ on HIF-1α protein ubiquitination degradation were independent of the p-VHL pathway, we investigated whether the ubiquitination degradation of the HIF-1α protein by TQ was related to the stability mediated by HSP90. First, the interaction between HIF-1α and HSP90 was verified using IP. PANC-1 cells were used as the experimental cells, and the experiment was divided into MG132 and MG132+TQ groups. The experimental condition was normoxic and the drug treatment duration was 8 h. The total cell lysate was incubated with normal serum and anti-HSP90 antibody, respectively, and the protein complex was detected by western blotting after IP. The immune complex (IP) of HIF-1α and HSP90 showed significantly positive results compared with the negative control of IG, indicating the interaction between HIF-1α and HSP90 (Figure 12A). In the presence of MG132, TQ still reduced the immune complex (IP) of HIF-1α, while the immune complex of HSP90 remained unchanged, suggesting that TQ inhibited the interaction of HIF-1α with HSP90 and was not affected by proteasome inhibitors. GA is a small molecule inhibitor of HSP90 that competitively binds to the N-domain adenosine triphosphate (ATP) binding site of HSP90, thereby disrupting the chaperone activity of HSP90[20]. To verify whether the ubiquitination degradation of the HIF-1α protein by TQ was dependent on the stability of the HIF-1α protein mediated by HSP90, we divided the experimental cells into four groups: The NC, TQ, GA (using a dose of 10 μM), and TQ+GA groups. The experimental condition was hypoxic, drug treatment duration was 8 h, and the experiment was repeated three times. Compared with the NC group, the expression level of the HIF-1α protein in the TQ group was significantly decreased (P < 0.01) (Figure 12B and C). However, there was no significant difference in the expression level of the HIF-1α protein between the GA and TQ+GA groups (P < 0.01) (Figure 12B and C), suggesting that GA blocked the effects of TQ on the HIF-1α protein expression. Specifically, the effects of TQ on the interaction of HIF-1α and HSP90 were blocked. Based on the above findings on the effects of TQ on the HIF-1α protein ubiquitination degradation, the effects of TQ on HIF-1α protein ubiquitination degradation were related to the stability of HIF-1α protein mediated by HSP90. In other words, TQ promoted the ubiquitination degradation of HIF-1α protein by inhibiting the interaction between HIF-1α and HSP90.

Figure 12  Effects of Thymoquinone on interaction between hypoxia-inducible factor-1α and HSP90. A: Immunoprecipitation assay was performed to detect the hypoxia-inducible factor-1α (HIF-1α) and HSP90 immunoprecipitation of PANC-1 cells in MG132 group and MG132 + Thymoquinone (TQ) group under normoxia condition; B: Western blot assay was performed to detect HIF-1α expression of PANC-1 cells in normal control (NC) group, TQ group, geldanamycin (GA) group and TQ+GA group under hypoxia condition; C: Statistical analysis of HIF-1α protein expression of PANC-1 cells in NC group, TQ group, GA group and TQ+GA group under hypoxia condition. ^aP < 0.01. NC: Normal control; TQ: Thymoquinone; GA: Geldanamycin; HIF-1α: Hypoxia-inducible factor-1α.

Further verification of initial expression and ubiquitination degradation of HIF-1α protein by TQ

Since preliminary confirmation that TQ might affect the final expression of HIF-1α protein by inhibiting the initial expression of HIF-1α protein and promoting the ubiquitination degradation of HIF-1α protein, we independently verified the two influence pathways in order to make the conclusion more reliable. First, we investigated the effects of TQ on the initial expression of HIF-1α protein. PANC-1 cells were first pretreated with CHX (100 μM) for 3 h under normoxic conditions to block the initial expression of HIF-1α protein and gradually degrade it to achieve "destocking". The experimental cells were then divided into MG132 and MG132+TQ groups, and the changes in the HIF-1α protein expression level were detected by western blot assay at 0 h, 0.5 h, 1 h, 2 h and 4 h, respectively. The expression levels of the HIF-1α protein gradually increased in the MG132 group (Figure 13A and B), indicating that when the ubiquitination degradation of the HIF-1α protein was blocked, the initial expression of the HIF-1α protein led to the gradual increase in its expression levels. Although the expression levels of HIF-1α protein in the MG132+TQ group also gradually increased, the increase trend was significantly slower than that in the MG132 group (Figure 13C and D), which further confirmed that TQ inhibited the initial expression of the HIF-1α protein.

Figure 13  Further verification of the effects of Thymoquinone on the initial expression of hypoxia-inducible factor-1α protein. A: Western blot assay was performed to detect hypoxia-inducible factor-1α (HIF-1α) expression in PANC-1 cells treated with MG132 at 0 h, 0.5 h, 1 h, 2 h and 4 h under normoxia condition; B: Statistical analysis of HIF-1α protein expression in PANC-1 cells treated with MG132 at 0 h, 0.5 h, 1 h, 2 h and 4 h under normoxia condition; C: Western blot assay was performed to detect HIF-1α expression in PANC-1 cells treated with MG132 + Thymoquinone (TQ) at 0 h, 0.5 h, 1 h, 2 h and 4 h under normoxia condition; D: Statistical analysis of HIF-1α protein expression in PANC-1 cells treated with MG132+TQ at 0 h, 0.5 h, 1 h, 2 h and 4 h under normoxia condition. TQ: Thymoquinone; HIF-1α: Hypoxia-inducible factor-1α.

Subsequently, we verified the effect of TQ on the ubiquitination degradation of the HIF-1α protein. sh-VHL type PANC-1 cells were used as experimental cells, so that the HIF-1α protein would not be degraded by ubiquitination through the p-VHL pathway even under normoxic conditions, and the possibility of the HIF-1α protein being degraded via other mechanisms under normoxic conditions was preserved. We first treated PANC-1 cells with CHX for 0, 1, 2, and 3 h. There was no significant change in the expression of the HIF-1α protein (Figure 14A and B), indicating that while CHX blocked the initial expression of the HIF-1α protein, no degradation occurred, and the protein was always in a stable state. Additionally, we added CHX and TQ to the sh-VHL type PANC-1 cells for 0 h, 1 h, 2 h, and 3 h respectively. The expression level of the HIF-1α protein gradually decreased (P < 0.01) (Figure 14C and D), indicating that under the premise that the initial expression of the HIF-1α protein was blocked, TQ still promoted its degradation, independent of the p-VHL pathway mediated ubiquitination.

Figure 14  Further validation of the effects of Thymoquinone on ubiquitination degradation of hypoxia-inducible factor-1αprotein. A: Western blot assay was performed to detect hypoxia-inducible factor-1α (HIF-1α) expression in von Hippel Lindau (VHL) knockdown PANC-1 cells treated with cycloheximide (CHX) at 0 h, 1 h, 2 h and 3 h under normoxia condition; B: Statistical analysis of HIF-1α protein expression in VHL knockdown PANC-1 cells treated with CHX at 0 h, 1 h, 2 h and 3 h under normoxia condition; C: Western blot assay was performed to detect HIF-1α expression in VHL knockdown PANC-1 cells treated with CHX + Thymoquinone (TQ) at 0 h, 1 h, 2 h and 3 h under normoxia condition; D: Statistical analysis of HIF-1α expression in VHL knockdown PANC-1 cells treated with CHX+TQ at 0 h, 1 h, 2 h and 3 h under normoxia conditions. ^aP < 0.01. CHX: Cycloheximide; VHL: von Hippel Lindau; sh-VHL: VHL knockdown; HIF-1α: Hypoxia-inducible factor-1α.

DISCUSSION

Over the past two decades, the number of PC patients diagnosed each year has doubled worldwide. In 2017, 441000 cases of PC were reported worldwide, up from 196000 in 1990[21]. PC, which rarely occurs before the age of 40, is a disease that carries an increased risk with age. The changing age structure of the global population and improved diagnosis are the main reasons for the increasing incidence of PC. This is also one of the reasons that the incidence is significantly higher in developed countries than in developing countries. Although the overall five-year survival rate for PC has increased from 5% in 1990 to 9% in 2019, according to the national cancer databases in Europe and the United States, the survival rate for PC remains extremely low[22,23]. The low survival rate for PC is partly due to the fact that most patients with PC are at an advanced stage when they are diagnosed, and less than 20% of patients have the opportunity to undergo surgery. In contrast, the 5-year survival rate of patients who undergo radical surgery is only about 15%-25%[24]. Although surgical treatment is still the main treatment method for long-term survival of PC, the low survival rate after surgery makes it necessary to identify new treatments to improve the prognosis of patients with PC. Drug therapy is an important means to improve the survival rate of PC. TQ has become a promising option for anti-cancer drugs in recent years because of its non-significant toxicity to normal cells and strong lethality to tumor cells. Its anti-cancer effect on PC has also been widely studied.

HIF-1 is composed of HIF-1α and HIF-1β. There are two transactivation domains (TADs) in the HIF-1α subunit structure, one at the C-end of the peptide chain and the other at the N-end of the peptide chain. Both domains are inhibitory domains that inhibit HIF-1α activity under normal oxygen conditions. An oxygen-dependent degradation domain is involved in the degradation of HIF-α, and an N-terminal TAD, as a positive regulator of the transcription process, regulates the downstream target gene expression. HIF-1α is the key subunit involved in HIF-1 functions. Under the action of growth factors, it is synthesized through the PI3K/Akt/mTOR pathway and the rat sarcoma/rapidly accelerated fibrosarcoma/mitogen-activated protein kinase/extracellular regulatory protein kinase pathway (commonly known as the RAS/RAF/MEK/ERK pathway). Under normal oxygen levels, HIF-1α-specific proline and asparagine residues form a β-fold-like conformation (hydroxylation) in the presence of prolyl hydroxylase (PHD) and are then recognized and bound to the E3 ligase complex p-VHL. Together, they form the ubiquitin-proteolytic enzyme complex, which is subsequently ubiquitinated and degraded in the 26S proteasome. When the concentration is lower than 5%, the residues of proline and asparagine of HIF-1α are not hydroxylated by PHD; thus, the subsequent ubiquitination-mediated degradation process are not carried out, and HIF-1α accumulates in cells. With the help of coactivators, such as cyclic adenosine phosphate reaction element binding protein and acetyltransferase (p300), HIF-1α forms a heterodimer with HIF-1β, which enters the nucleus and binds to the hypoxia response element of target genes to exert biological effects. In addition, HSP90 stabilizes HIF-1α, and this process is independent of p-VHL-mediated ubiquitination. The absence of HSP90 or disruption of its interaction with HIF-1α has been shown to lead to the destabilization of ubiquitination-mediated HIF-1α degradation[25]. At present, no studies have investigated the interaction between TQ and HIF-1α in PC. Notably, Mu et al[26] found that TQ enhances the anti-tumor effect of gemcitabine by blocking the PI3K/AKT/mTOR signaling pathway in PC cells[26], and Lee et al[17] found that TQ inhibits ubiquitination and influences HSP90 for reducing HIF-1α stability[17]. Based on these previous findings, we analyzed the relationship between TQ and HIF-1α and observed the effects of TQ on the synthesis, stabilization, and degradation pathways of HIF-1α in PC.

In our study, we found that TQ promoted apoptosis and inhibited the proliferation and invasion of PC cells, but it had no significant effect on normal pancreatic ductal epithelial cells. This finding suggested that TQ was a potentially safe drug against PC. Moreover, we demonstrated that TQ inhibited HIF-1α expression in PC. TQ affected the stability and ubiquitination-mediated degradation of HIF-1α by affecting the interaction between HSP90 and HIF-1α as well as the initional expression of HIF-1α by affecting the PI3K/AKT/mTOR pathway. In addition, the effects of TQ on HIF-1α were only slightly reduced in the presence of proteasome inhibitors (Figure 8). After CHX treatment under normoxia condition to achieve "HIF-1α destocking," we found that the expression of HIF-1α in PC cells treated with TQ and MG132 was significantly lower than it was in cells treated with MG132 alone (Figure 13). This suggested that the promotion of ubiquitination degradation of HIF-1α protein by TQ was the main pathway to affect the final expression of HIF-1α protein, while the inhibition of PI3K/AKT/mTOR signaling pathway to reduce the initial expression of HIF-1α protein was a secondary pathway to affect the expression of HIF-1α protein.

Molecular chaperones are a class of molecules that can help these proteins acquire a functionally active form by interacting with them. Once the final active structure of these proteins is formed, the molecular chaperones are separated from these bound proteins. As one of the most important molecular chaperones in human cells, HSP90 was discovered due to its specific elevated expression in heat shock response. HSP90, which accounts for 1%-2% of cellular proteins and increases to 4%-6% in stressed cells, is composed of three domains: The N-terminal dimerization domain, which is responsible for binding ATP, and is connected to the intermediate domain through unstructured charged connectors; the C-terminal domain, which is responsible for the inherent dimerization of proteins; and the N-terminal domain , which is transiently dimerized by binding ATP[27]. In normal cells, HSP90 exists mainly in an unbound state, while cancer cells are highly dependent on HSP90 for survival[28]. HSP90 plays an important role in cancer cell survival, not only because cancer cells rely on HSP90-assisted signaling pathways, but also because unstable carcinogenic mutations may increase tumor cell dependence on HSP90[29].

Nagaraju et al[30] used GA to interfere with the key chaperone protein molecule HSP90 of the HIF-1α protein, thereby affecting the expression activity of HIF-1α to enhance the radiotherapeutic sensitivity of PC[30]. Daunys et al[31] found that the addition of HSP90 inhibitors (ICPD47 and ICPD62) to human PC cell lines MIA PaCa-2 and PANC-1 could have a synergistic effect with gemcitabine, 5-fluorouracil, and doxorubicin chemotherapeutic drugs[31]. Ghadban et al[32] evaluated the effects of three different HSP90 inhibitors (17-AAG, 17-DMAG, and 17-AEPGA) on proliferation and apoptosis of human PC cell lines 5061, 5072, 5156, and L3.6pl. It was found that HSP90 inhibitors could inhibit the proliferation and promote apoptosis of all the above cell lines, especially against gemcitabine and 5-fluorouracil resistant cell lines 5061 and 5072, which confirmed that HSP90 was crucial for establishing the chemotherapeutic sensitivity of PC[32]. Adachi et al[33] confirmed that 17-AAG (an HSP90 inhibitor) could significantly reduce the cell viability of human PC cell lines (KP3, BxPc-3, AsPc-1). This mechanism might be related to the desensitization of the oncogenic protein epidermal growth factor receptor-associated with HSP90[33]. Xue et al[34] confirmed that the HSP90 inhibitor, Y306zh, could inhibit cell proliferation of the human PC cell line, Miapaca2, halt the cell cycle to the G2/M phase, and inhibit tumor growth in nude mouse PC transplant models[34]. Li et al[35] demonstrated that epigallocin gallate, as a novel HSP90 inhibitor, could promote the apoptosis of the human PC cell line, Mia Paca-2, in a dose-dependent manner, possibly because it directly bound to the C-terminal domain of HSP90 and reduced the level of carcinogenic protein in HSP90 clients[35]. Similarly, triptolide blocked the binding of Cdc37 (an important companion protein of HSP90 that binds HSP90 to client proteins) to HSP90, leading to the degradation of HSP90 client proteins. This induced apoptosis of human PC cell line PANC-1 and inhibited tumor growth in xenografted mouse models[36]. Other studies have revealed that green tea extract could induce apoptosis and growth inhibition of human PC ductal cell line HPAF-II, which was related to the down-regulation of HSP90 expression, thus inhibiting the phosphorylation of Akt, the target protein of HSP90, and the expression of P53[37]. Bobrov et al[38] used the dual inhibitor of HSP90 and topoisomerase I, STA-12-8666, to improve the survival rate of mice with PC[38]. Gu et al[39] demonstrated that Withaferin A and its analogue (a naturally occurring steroid ester that can directly bind to HSP90 and lead to degradation of the HSP90 client protein) could promote apoptosis and loss of activity of PANC-1 cells. The possible mechanism was the degradation of HSP90 client proteins Akt and Cdk4 induced by a proteasome-dependent pathway in PC cells[39]. HSP90 is crucial to the occurrence and development of PC because many carcinogenic proteins must bind to HSP90 in order to exert biological effects, and HSP90 inhibitors mostly play anti-cancer effects by inhibiting the binding of HSP90 and carcinogenic proteins. In our experiment, TQ was found to inhibit the interaction between HIF-1α and HSP90, thus destabilizing the HIF-1α protein and eliminating the cancer-promoting effect of HIF-1α. TQ was therefore potentially a promising new type of HSP90 inhibitor, and its dissociation of HSP90 and other carcinogenic proteins would be further confirmed.

PI3K /AKT/mTOR is a classical intracellular signal transduction pathway that is critical for regulating cell survival, growth, migration, and metabolism[40]. Identifying specific molecular alterations is a key factor in guiding biomarker-based targeted therapies and in determining the best treatment strategy for individual cancer patients. The PI3K/AKT/mTOR pathway is a key downstream effector of RAS, and RAS activation is the most prominent genetic change in PC[41]. Targeting PI3K/AKT/mTOR as a key downstream signaling pathway may be a more effective treatment option for PC[42].

That's exactly what previous studies have shown. Firstly, the proliferation, invasion, and migration of PC cells are mostly related to the abnormal activation of the PI3K/AKT/mTOR pathway[43]. For example, Xiao et al[44] confirmed that fisetin could inhibit the growth, invasion, and migration of PANC-1 cells and tumor growth in xenotransplantation models of nude mice by inhibiting the PI3K/AKT/mTOR cascade[44]. Chen et al[45] confirmed that serine/threonine kinase 3 inhibited the proliferation, accelerated apoptosis, enhanced invasion and migration of human PC cell lines, BXPC-3 and PANC-1, by inhibiting the PI3K/AKT/mTOR pathway, and inhibited tumor growth in xenotransplantation models of nude mice[45]. In the relationship between urolitin and PC, researchers such as Totiger et al[46] confirmed that urolitin could inhibit the proliferation of human PC cell lines, MiaPaCa-2, BxPC-3, and PANC-1, by targeting the PI3K/AKT/mTOR pathway and slowing down tumor growth in xenotransplantation models of nude mice[46]. Secondly, PI3K/AKT/mTOR is also involved in the autophagy of PC cells. Qian et al[47] confirmed that quercetin could promote apoptosis and autophagy of human PC cell lines, SW1990 and PANC-1, and inhibit tumor growth in xenotransplantation models of nude mice, the mechanism of which might be related to the inactivation of the PI3K/AKT/mTOR pathway[47]. As an inhibitor of protein kinase C-δ, rottlerin could induce autophagy of cancer cells and lead to apoptosis by inhibiting the PI3K/AKT/mTOR pathway in human PC stem cells[48]. Third, the PI3K/AKT/mTOR pathway is also closely related to the chemotherapeutic sensitivity of gemcitabine to PC. Lan et al[49] demonstrated that silencing receptors for advanced glycation end-product expression could inhibit the PI3K/AKT/mTOR axis in human PC cell lines, MIA Paca-3 and MIA Paca-2, thereby significantly increasing the cytotoxicity induced by giscitabine[49]. Tian et al[50] revealed that arsenic trioxide and gemcitabine could jointly inhibit the PI3K/AKT/mTOR pathway, thereby synergistically promoting epithelial mesenchymal transformation and apoptosis of PC and improving the chemotherapeutic sensitivity of gemcitabine[50]. Fourth, glycolysis is an important mechanism for PC cells to survive the hypoxic and hypoglycemic environment, and the activation and inactivation of the PI3K/AKT/mTOR signaling pathway is crucial for glycolysis. Li et al[51] demonstrated that TRIM59 could activate the PI3K/AKT/mTOR signaling pathway, resulting in the human PC cell line, BxPC-3, enhancing glycolysis and promoting liver metastasis of pancreatic tumors in nude mice[51]. As an inhibitor of mTOR, evolimus could also reduce the “Warburg effect” of PC cells by blocking the PI3K/AKT/mTOR pathway and overcoming gemcitabine resistance in in vivo and in vitro experiments[52]. Fifth, the PI3K/AKT/mTOR pathway is also closely related to the angiogenesis of pancreatic tumors. Using in vitro and in vivo angiogenesis measurements, He et al[53] found that hispidulin inhibited VEGF (vascular endothelial growth factor)-induced aortic ring micro-vascular germination and corneal neovascularization, as well as the growth of transplanted tumors in mice. The mechanism was that hispidulin could inhibit the PI3K/AKT/mTOR pathway mediated by vascular endothelial growth factor receptor 2 to prevent angiogenesis in PC[53]. Sixth, the PI3K/AKT/mTOR pathway is also closely related to the immune regulation of the PC microenvironment. Xu et al[54] found that saikosaponin could reduce the polarization of M3 macrophages by down-regulating the phosphorylation of STAT2 and the PI3K/AKT/mTOR signaling pathway, which could directly inhibit the apoptosis and invasion of human PC cell lines, Panc02 and PancH7, inhibit the growth of transplanted tumors in nude mice, and regulate the immune microenvironment[54]. Our study found that TQ affected the initial expression of the HIF-1α protein by inhibiting the PI3K/AKT/mTOR pathway, promoting apoptosis of PC cells, and inhibiting proliferation, migration, and invasion. On the one hand, this may be related to the inhibition of the expression of the HIF-1α protein, which leads to the reduction of its carcinogenic effects. However, its relation to the inhibition of the PI3K/AKT/mTOR pathway, a key signaling pathway for the pathogenesis of PC, cannot be ruled out.

Overall, our study confirmed that TQ inhibited the expression of HIF-1α in PC cells, and systematically explored the specific mechanism of TQ affecting the expression of the HIF-1α protein. In other words, TQ mainly inhibited the interaction between HIF-1α and HSP90 to cause the instability of the HIF-1α protein and promote its ubiquitination degradation, and secondly affected the initial expression of the HIF-1α protein by inhibiting the PI3K/AKT/mTOR pathway. In addition, TQ inhibited the proliferative activity, invasion, and migration of PC cells and promoted the apoptosis of cancer cells. On the one hand, TQ might inhibit the expression of the HIF-1α protein, thus weakening its cancer-promoting effect. However, its relation to the inhibition of the PI3K/AKT/mTOR pathway, the classical pathogenic pathway of PC, cannot be ruled out. Although we tested the proliferative activity of TQ on cancer cells and its effects on HIF-1α mRNA and protein expression in three human PC cell lines (PANC-1, AsPC-1, and BxPC-3), only one of the cancer cell lines, PANC-1, was selected for follow-up experiments. First, this was due to financial constraints; second, PANC-1 was the most representative of PC cell lines and the most widely used in cell experiments. In addition, we quantified the effects of TQ on the proliferation, migration, invasion, and apoptosis of PC cells under hypoxic conditions. On the one hand, this was to be consistent with subsequent main experiments, because the effects of TQ on HIF-1α protein expression were mainly determined under hypoxic conditions. However, the effects of TQ on the viability, apoptosis, migration, and invasion of PC cells under normoxic conditions had been verified by other researchers[26,55-57], and repeated experiments were of little significance and further verification was not needed.

CONCLUSION

TQ could significantly inhibit the proliferative activity, invasion and migration ability and promote the apoptosis of PC cells, which might be related to the inhibition of HIF-1α protein expression. The regulatory mechanism of TQ on HIF-1α protein expression in PC cells was mainly to promote the ubiquitination degradation of HIF-1α protein by inhibiting the interaction between HIF-1α and HSP90; Secondly, TQ reduced the initial expression of HIF-1α protein by inhibiting the PI3K/AKT/mTOR pathway (Figure 15).

Figure 15  Full text summary. HIF-1α: Hypoxia-inducible factor-1α; HIF-1β: Hypoxia-inducible factor-1β; PI3K: Phosphatidyl inositol-4,5-bisphosphate-3-kinase; Akt: Protein kinase B; mTOR: Mammalian target of rapamycin; 4E-BP1: Eukaryotic translation initiation factor 4E binding protein 1; eIF-4E: Eukaryotic translation initiation factor 4E; S6K: S6 kinase; Ras: Rat sarcoma; Raf: Rapidly accelerated fibrosarcoma; MAPK: Mitogen-activated protein kinases; ERK: Extracellular signal-regulated kinase; MEK: Mitogen extracellular signal-regulated kinas; MNK: MAP kinase interacting kinase; HRE: Hypoxia response elements; HSP90: Heat shock protein 90; pVHL: von Hippel-Lindau protein; Mdm2: Murine double-minute 2.

ACKNOWLEDGEMENTS

The authors are grateful to the Health Commission of Qinghai Province for their help.