This article is mentioned in:

Introduction

Lung cancer is one of the most malignant types of cancer in humans, and according to the 2015 National Institutes of Health statistics, non-small cell lung cancer (NSCLC) constitutes ~83.9% of lung cancer cases in the USA (1,2). Radiotherapy is one of the principal therapeutic strategies in the treatment of lung cancer (3). However, resistance to radiation results in a poor outcome following radiotherapy (4,5). Thus, identifying novel methods to avoid radiation resistance and improve the efficacy of radiotherapy is a major challenge in the treatment of lung cancer.

Erastin was originally identified when Dolma et al (6) screened small molecules that are synthetically lethal in cancer cells expressing RAS (6,7). Erastin induces iron-dependent cell death that is morphologically, biochemically and genetically different from apoptosis, necrosis, necroptosis and autophagy, and this form of cell death is termed ferroptosis (8–11). Recently, Yu et al (12) reported that erastin enhances sensitivity of AML cells to chemotherapy. Another study reported that the inhibition of ferroptosis increases sorafenib resistance in hepatocellular carcinoma cells (13). Ivanov et al (14,15) described that iron-containing water improves the efficiency of radiotherapy and deferoxamine (DFOM) reduces the efficiency of radiotherapy in animals with gliomas. These data indicated that ferroptosis participates in the regulation of response to chemotherapy and radiotherapy in cancer cells.

According to the Lung Cancer Mutation Consortium database, oncogenic K-ras mutations were identified in 23% of patients with lung cancer between 2009–2015 (16). Erastin is a RAS-selective lethal (RSL) compound, which can trigger ferroptosis (17). Herein, we hypothesized that erastin may influence radiosensitivity in NSCLC by triggering ferroptosis. In the present study, two K-ras-expressing NSCLC cell lines (A549 and H460) were selected to conduct the experiments (18,19). The present study aimed to verify that erastin can decrease, at least partially, the radioresistance of NSCLC cells and attempted to perform a preliminary investigation regarding the molecular mechanism.

Materials and methods

Reagents

The RPMI-1640 culture medium was purchased from Gibco (Thermo Fisher Scientific, Inc., Waltham, MA, USA). Fetal bovine serum (FBS) was purchased from Hangzhou Sijiqing Biological Engineering Materials Co., Ltd (Hangzhou, China). Dimethyl sulfoxide (DMSO), DFOM, Z-VAD-FMK, trypsin and trypan blue were purchased from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany). Primers for glutathione peroxidase 4 (GPX4), small interfering (si)RNA of GPX4, Lipofectamine® 2000, TRIzol®, OPTI-MEM I, MMLV reverse transcriptase, Taq DNA polymerase and Oligo dT primers were purchased from Invitrogen (Thermo Fisher Scientific, Inc.). Erastin and olaparib were purchased from Selleck Chemicals (Houston, TX, USA). Protein molecular weight standards were purchased from Fermentas (Thermo Fisher Scientific, Inc.). Protein lysis buffer and the Bicinchoninic Acid (BCA) protein assay kit were purchased from the Beyotime Institute of Biotechnology (Wuhan, China). Protease inhibitors were purchased from Roche Diagnostics (Basel, Switzerland). Rabbit anti-GPX4 (catalog no. ab125066; 1:1,000) and anti-β-actin (catalog no. ab8226; 1:500) were purchased from Abcam (Cambridge, MA, USA). Horseradish peroxidase (HRP)-conjugated goat anti-rabbit secondary antibody (catalog no. TA130023; 1:3,000) was purchased from OriGene Technologies, Inc. (Beijing, China). All antibodies were diluted in bovine serum albumin. The Enhanced Chemiluminescence (ECL) chemiluminescence reagent was purchased from Thermo Fisher Scientific, Inc.

Cell culture

The human NSCLC cell lines A549 and H460 were purchased from the American Type Culture Collection (Manassas, VA, USA). Cells were cultured in RPMI-1640 medium supplemented with 10% FBS, 100 IU penicillin and 100 µg/ml streptomycin, and incubated at 37°C in a 5% CO2 humidified incubator.

Establishment of the NSCLC radioresistant subtype cell lines

Exponentially growing NSCLC cells A549 and H460 were irradiated with 5 does of 6 Gy. Irradiation was performed with 6-MV X-rays generated by a Siemens Primus H high-energy linear accelerator (Siemens Healthineers, Erlangen, Germany), as described previously (20,21). There was a 7–9 day break in between each irradiation (21). The radiation field was 10×10 cm, the distance from the source to target was 100 cm and the absorbed dose rate was 0.2 Gy/min. The surviving sublines (A549-R and H460-R) were then passaged for three months at 37°C and the radiosensitivity was determined by colony formation assays.

Colony formation assay

A single cell suspension was prepared from the A549, A549-R, H460 and H460-R cells in their logarithmic growth phase using 0.25% trypsin for digestion. Cells were seeded in 6-well plates and received single dose irradiation of 0, 2, 4, 6 and 8 Gy after a 12 h culture adherence at 37°C in a 5% CO2 humidified incubator. Cell density was 200, 300, 600, 1,200 and 2,400 cells/well for each radiation dose. After 2 weeks, cells were fixed with 4% paraformaldehyde at 25°C for 20 min and were stained with Giemsa at 25°C for 15 min. Cell survival fraction (SF) was calculated based on the following formula: SF=colony numbers in experimental group/[plating number × plating efficiency (PE)]. PE=(colony numbers in the control group/plating number in control group) ×100%. A cell survival curve was obtained by the single-hit multi-target model [SF=l-(1-e−D0/D)N] using Sigma Plot 10.0 version (Systat Software, Inc, CA, USA). The parameters of cell survival fraction, a 2 Gy irradiation dose (survival fraction at 2 Gy irradiation, SF2), mean lethal dose value (D0), quasithreshold dose (Dq) and extrapolation number (N) were calculated. Higher values of SF2, D0 and Dq indicated an increased radioresistance (22).

Cell death assay

A549-R and H460-R cells were divided into four groups for the following treatments: i) control group treated with DMSO; ii) treated with 4 µM erastin alone for 12 h at 37°C; iii) treated with 2 Gy IR alone and cultured in medium without erastin for 12 h at 37°C; and iv) treated with 2 Gy IR and then cultured in medium with 4 µM erastin for 12 h. Following the indicated treatments, the cells were trypsinized and stained with trypan blue for 3 min at 25°C, followed by counting with a hemocytometer using a standard protocol (21). Cells stained blue were considered as dead cells.

In addition, A549-R and H460-R cells were divided into eight groups for the following treatments: i) control group; ii) treated with 100 µM DFOM; iii) treated with 50 µM Z-VAD-FMK; iv) treated with 10 µM olaparib; v) treated with 4 µM erastin; vi) treated with 4 µM erastin and 100 µM DFOM; vii) treated with 4 µM erastin and 50 µM Z-VAD-FMK; and viii) treated with 4 µM erastin and 10 µM olaparib. All groups were incubated with an equal volume DMSO, including the control group. The eight groups of cells were treated for 12 h at 37°C. Following the indicated treatments, the cells were trypsinized and stained with trypan blue for 3 min at 25°C, followed by counting with a hemocytometer using the standard protocol (23). Cells stained blue were considered as dead cells.

Western blotting

A549, A549-R, H460 and H460-R cell lysates were prepared using the protein lysis buffer, according to the manufacturer's protocol. Following extraction, protein concentration was determined with a BCA assay. A total of 30 µg protein from each sample was loaded into a 10% SDS-PAGE, followed by a transfer to polyvinylidene fluoride membranes (EMD Millipore, Billerica, MA, USA). Membranes were blocked with 5% bovine serum albumin (catalog no. ST023-1000 g; Beyotime Institute of Biotechnology) for 2 h at 37°C. They were then incubated with rabbit anti-GPX4 (1:1,000) and anti-β-actin (1:500) antibodies at 4°C overnight. The polyvinylidene fluoride membrane was washed three times with TBS and Tween-20, 5 min each. Subsequently, the membranes were incubated with HRP-conjugated goat anti-rabbit (1:3,000) antibodies for 1 h at 25°C. The ECL reagent then was added to the membranes and the optical density of the bands was analyzed using the Image J V1.8.0 software (National Institute of Health, Bethesda, MD, USA).

siRNA transfection

siGPX4 (catalog no. AM16708) and the control siRNA (siMock; catalog no. AM4611) were obtained from Invitrogen (Thermo Fisher Scientific, Inc.). A solution composed of 1 ml OPTI-MEM, 6 µl Lipofectamine 2000 and 2 µl siRNA (the final concentration was 2 µl/ml) was prepared and the mixture (1 ml/well) was incubated in a 6-well dish for 20 min at 37°C. Following incubation, 200,000 cells were suspended in 1 ml RPMI-1640 culture medium with 20% FBS (Thermo Fisher Scientific, Inc.). Subsequently, they were transferred to each well and incubated for 72 h. GPX4 expression was measured by western blot analysis, as aforementioned.

Statistical analysis

Statistical analysis was performed using SPSS 17.0 (SPSS Inc., Chicago, IL, USA). Data are presented as the mean ± standard error of the mean. Significant differences between the groups were determined by unpaired Student's t-test and by one-way analysis of variance followed by a Bonferroni post hoc test for multiple group comparisons (Figs. 3, 4 and 7). P<0.05 was considered to indicate a statistically significant difference.

Figure 3. Cell death measurements following the administration of different treatments. The cell death fraction in the erastin + IR group was significantly increased, compared with the erastin or IR groups. (A) Cell death fraction in A549-R cells. (B) Cell death fraction in H460-R cells. Data are presented as the mean ± standard deviation. ***P<0.001. R, radioresistant; IR, irradiation; DMSO, dimethyl sulfoxide.

Figure 4. Treatment with erastin induces notable cell death. Treating cells with DFOM, Z-VAD-FMK and olaparib alone exhibited no significant effect on cell death. When cells were treated with DFOM and erastin together, cell death was decreased, compared with treatment with erastin alone. However, when cells were treated with Z-VAD-FMK and erastin, cell death exhibited no significance difference to the group treated with erastin alone. Additionally, olaparib had no significant effect on erastin-induced cell death. (A) Cell death fraction in A549-R cells. (B) Cell death fraction in H460-R cells. Data are presented as the mean ± standard deviation. ***P<0.001. R, radioresistant; DFOM, deferoxamine.

Figure 7. siRNA inhibition of GPX4 expression and cell death examination. Inhibiting GPX4 expression induced significant cell death. Treating siGPX4 group cells with DFOM significantly decreased cell death. However, treating cells with Z-VAD-FMK and olaparib did not significantly change the cell death fraction. (A) Cell death fraction in A549-R cells. (B) Cell death fraction in H460-R cells. Data are presented as the mean ± standard deviation. ***P<0.001. siRNA, small interfering RNA; GPX4, glutathione peroxidase 4; R, radioresistant; DFOM, deferoxamine; DMSO, dimethyl sulfoxide.

Results

Radiosensitivity change in A549-R and H460-R cells

To confirm that the radioresistant cell line was successfully established, a colony formation assay was performed to determine the radiosensitivity of A549-R and H460-R cells. The results demonstrated that A549-R and H460-R cells had increased survival fractions, compared with A549 and H460 cells, respectively. Cell survival curves were used to analyze the results. Therefore, the shoulder area under the survival curve was increased in H460-R and A549-R cells (Fig. 1). Compared with the A549 and H460 cells, A549-R and H460-R cells, respectively, had increased D0, Dq, N and SF2 values (Table I), which represented increased radioresistance. Collectively, these results demonstrated that radioresistant (A549-R and H460-R) cells were successfully established.

Figure 1. Clone formation assay determining the radiosensitivity of A549-R and H460-R cells. The results demonstrated that A549-R and H460-R cells had increased survival fractions, compared with A549 and H460 cells, respectively. The shoulder area under the survival curve was increased in H460-R and A549-R cells. (A) Radiosensitivity of A549 and A549-R cells. (B) Radiosensitivity of H460 and H460-R cells. Data are presented as the mean ± standard deviation. *P<0.05, **P<0.01, ***P<0.001 A549-R vs. A549 or H460-R vs. H460. R, radioresistant; IR, irradiation.

Table I. Parameters of radiosensitivity in different cell lines.

Table I.

Parameters of radiosensitivity in different cell lines.

Parameter	A549	A549-R	H460	H460-R
D0	1.84	3.12	1.74	2.81
Dq	1.50	2.17	1.70	2.40
N	2.43	2.80	3.14	3.92
SF2	0.59	0.85	0.52	0.81

[i] D₀, mean lethal dose; D_q, quasi-threshold dose; N, extrapolation number; SF₂, surviving fraction at 2 Gy; R, radioresistant.

Erastin enhances radiosensitivity in A549-R and H460-R cells

A colony formation assay was performed to investigate the effect of erastin on radiosensitivity of A549-R and H460-R cells. The results demonstrated that erastin-treated groups had reduced survival fractions, compared with the control groups. Cell survival curves were used to analyze the results. Additionally, the shoulder area under the survival curves was reduced in the erastin-treated groups, compared with the control groups (Fig. 2). Compared with the control group cells, the erastin group cells had decreased D0, Dq, N and SF2 values (Table II), which indicated decreased radioresistance. The result demonstrated that erastin enhances the sensitivity to radiation therapy in the radioresistant NSCLC cells.

Figure 2. Colony formation assay investigating the effect of erastin on radiosensitivity of A549-R and H460-R. The results demonstrated that groups treated with erastin had reduced survival fractions, compared with the control groups. The shoulder area under the survival curves was reduced in erastin groups. (A) Radiosensitivity of A549-R cells. (B) Radiosensitivity of H460-R cells. Data are presented as the mean ± standard deviation. ***P<0.001 vs. control. R, radioresistant; IR, irradiation.

Table II. Parameters of radiosensitivity in different groups.

Table II.

Parameters of radiosensitivity in different groups.

	A549-R		H460-R
Parameter	Control	Erastin	Control	Erastin
D0	2.97	2.21	2.85	2.28
Dq	2.08	1.45	2.39	1.61
N	2.71	1.93	3.82	2.20
SF2	0.85	0.63	0.93	0.69

[i] D₀, mean lethal dose; D_q, quasi-threshold dose; N, extrapolation number; SF₂, surviving fraction at 2 Gy; R, radioresistant.

Erastin and IR exhibit a combined effect on killing cells

Following indicated treatment, cell death was measured. In A549-R cells, cell death fraction induced by erastin + IR was 60.2±2.7%. The cell death percentage induced by erastin alone was 32.5±1.9%, while the IR-induced cell death fraction was 19.1±2.5%. In H460-R cells, the cell death fraction in the erastin + IR group was significantly increased, compared with the erastin or IR groups (39.4±4.5% vs. 21.2±2.2% or 14.9±2.1%, respectively) (Fig. 3). These data revealed that erastin and IR exhibited a combined effect on killing cells.

Erastin-induced cell death could partially be rescued by DFOM instead of Z-VAD-FMK and olaparib

DFOM is an iron chelator, which can inhibit ferroptosis (24). Z-VAD-FMK is a caspase-inhibitor, which can inhibit caspase-dependent apoptosis (25). Olaparib is a poly(ADP-ribose) polymerase 1 (PARP-1) inhibitor that can inhibit PARP-1-dependent apoptosis (26). Cell death was detected following treatment with DFOM (100 µM), Z-VAD-FMK (50 µM) and olaparib (10 µM) alone or with erastin (4 µM) for 12 h at 37°C. Treatment with erastin induced notable cell death. Treating cells with DFOM, Z-VAD-FMK and olaparib alone had no notable effect on cell death. When cells were treated with DFOM and erastin together, cell death was significantly decreased, compared with treated with erastin alone. However, when cells were treated with Z-VAD-FMK and erastin, cell death exhibited no significant difference, compared with treated with erastin alone. Additionally, olaparib had no significant effect on erastin-induced cell death (Fig. 4). The results indicated that DFOM partially rescued erastin-induced cell death, while Z-VAD-FMK and olaparib had no significant influence on it.

GPX4 expression is increased in the radioresistant cells and erastin inhibits GPX4 expression in the radioresistant cells

According to Yang et al (24), one of the mechanisms for erastin to induce cell death is that erastin inactivates GPX enzymes to trigger ferroptosis. GPX4 expression was detected in the radioresistant cells and GPX4 expression level was increased in A549-R and H460-R cells, compared with A549 and H460 cells, respectively (Fig. 5A). The radioresistant cells were treated with erastin and the expression of GPX4 was detected. The results demonstrated that GPX4 expression was suppressed by erastin in the A549-R and H460-R cells, compared with the control groups (Fig. 5B). Therefore, it was inferred that erastin increases radiosensitivity in NSCLC cells through inhibiting GPX4.

Figure 5. GPX4 expression detection in the radioresistant cells. GPX4 protein expression level was increased in A549-R and H460-R cells. Following treatment with erastin, GPX4 expression was markedly decreased. (A) GPX4 expression in the parental and radioresistant cells. (B) GPX4 expression in the control group and erastin group cells. R, radioresistant; GPX4, glutathione peroxidase 4.

Knocking down GPX4 expression radiosensitizes NSCLCs cell to radiation in the radioresistant cell lines

It was inferred that high levels of GPX4 may contribute to radioresistance in NSCLC cells. Therefore, siRNA was used to knockdown GPX4 expression in A549-R and H460-R cells. The radiosensitivity of the GPX4-knockdown cells (siGPX4) and control cells (siMock) was subsequently detected with a colony formation assay and the survival curves were calculated. In A549-R cells, siGPX4 cells exhibited decreased radioresistance, compared with siMock cells. In H460-R cells, similar results were achieved, demonstrating that knocking down the GPX4 expression resulted in decreased radioresistance (Fig. 6). The radiosensitivity parameters D0, Dq, N and SF2 are presented in Table III. Based on the results, it was concluded that knocking down the GPX4 expression resulted in decreased radioresistance.

Figure 6. siRNA was used to knockdown GPX4 expression in radioresistant cells, and the radiosensitivity of the GPX4-knockdown cells (siGPX4) and control cells (siMock) was determined with a colony formation assay. siGPX4 cells exhibited decreased radioresistance, compared with siMock cells. (A) GPX4 expression in A549-R and H460-R cells. (B) Radiosensitivity of A549-R cells. (C) Radiosensitivity of H460-R cells. Data are presented as the mean ± standard deviation. **P<0.01 and ***P<0.001. siRNA, small interfering RNA; GPX4, glutathione peroxidase 4; R, radioresistant; IR, irradiation.

Table III. Parameters of radiosensitivity in different groups.

Table III.

Parameters of radiosensitivity in different groups.

	A549-R		H460-R
Parameter	siMock	siGPX4	siMock	siGPX4
D0	3.11	2.27	2.83	1.87
Dq	2.15	1.50	2.37	1.66
N	2.77	1.97	3.78	2.83
SF2	0.87	0.65	0.92	0.69

[i] D₀, mean lethal dose; D_q, quasi-threshold dose; N, extrapolation number; SF₂, surviving fraction at 2 Gy; si, small interfering; R, radioresistant; GPX4, glutathione peroxidase 4.

Cell death induced by the knockdown of GPX4 could be partially rescued by DFOM, yet not rescued by Z-VAD-FMK and olaparib

Inhibiting GPX4 expression in A549-R and H460-R cells induced notable cell death. In A549-R cells, the cell death fraction was 56.03±3.63% in the siGPX4 group. In H460-R cells, the cell death fraction was 64.70±1.80% in the siGPX4 group. Treating siGPX4 group cells with DFOM decreased cell death fraction to 17.5±3.78% in A549-R cells. In H460-R cells, the cell death fraction was decreased to 16.00±3.05% following treatment with DFOM. However, treating cells with Z-VAD-FMK and olaparib did not significantly alter the cell death fraction (Fig. 7). Cell death fractions refer to the DMSO control. Collectively, these results indicated that the knockdown GPX4 decreases radioresistance in NSCLC cells by inducing ferroptosis.

Discussion

Erastin is a type of RSL, which can activate a lethal pathway that differs from other forms of cell death, such as apoptosis (6,7). Erastin-induced cell death exhibits unique morphological, biochemical and genetic features, and the term ferroptosis was proposed to describe this type of cell death (8). Iron chelators, including DFOM, can decrease the iron uptake of cells and inhibit ferroptosis (9,10,27). However, apoptotic inhibitors, including Z-VAD-FMK, a caspase inhibitor, and olaparib, a PARP inhibitor, have no effect on ferroptosis (12,28).

Ferroptosis is a form of cell death that is morphologically, biochemically and genetically different from apoptosis, necrosis and autophagy. Ferroptosis has previously been observed in RSL-treated cells (8–10). The classic characteristics of apoptosis include mitochondrial cytochrome c release, caspase activation and chromatin fragmentation, which are absent in RSL-induced cell death (7,8). However, increased levels of intracellular reactive oxygen species is commonly observed in RSL-treated cells (9,10). Additionally, RSL-induced cell death can be prevented by iron chelation or the suppression of iron uptake (7,8,24,29). Ferroptosis can be activated by small activators in Ras-expressing cancer cells (6,7,24). Therefore, the Ras-expressing A549 and H460 cells were selected for use in the present study. Firstly, radioresistant NSCLC cells were established (A549-R and H460-R). Subsequently, erastin was used to treat A549-R and H460-R cells. Treatment with erastin increased the sensitivity to radiation in A549-R and H460-R cells. Erastin and IR exhibited a combined effect on inducing cell death. DFOM partially rescued cell death induced by erastin while Z-VAD-FMK and olaparib had no notable effect on erastin-induced cell death. Collectively, the results indicated that erastin increased radiosensitivity in NSCLC by inducing ferroptosis, at least partially.

GPX4 expression has been demonstrated to be increased in radioresistant cells and GPX4 is a target molecular of erastin (24). In the present study, GPX4 expression was knocked-down in the radioresistant cells, and decreasing GPX4 expression was observed to significantly induce cell death. Cell death induced by GPX4-inhibition could be partially rescued by DFOM, and was not affected by treatment with Z-VAD-FMK and olaparib. The results indicated that knockdown of GPX4 induced ferroptosis. The knockdown of GPX4 in the radioresistance NSCLC cells resulted in increased cell death, which indicated that GPX4 may be involved in the regulation of radiosensitivity in NSCLC cells by modulating ferroptosis. Therefore, we hypothesized that erastin decreases radioresistance in NSCLC cells partially, by GPX4-mediated ferroptosis.

Ferroptosis has become the focus of an increasing number of studies. For example, in human hepatocellular carcinoma cells and other cancer cell lines, sorafenib has been demonstrated to induce ferroptosis (30,31). Treatment with iron-containing water in tumor-bearing animals prior to undergoing radiotherapy also suppressed tumor growth by inducing ferroptosis (14,15). Yu et al (12) also reported that ferroptosis induced enhanced sensitivity of acute myeloid leukemia cells to chemotherapy. In the present study, similar results were obtained by demonstrating that erastin radiosensitized NSCLC cells by inducing ferroptosis. These studies demonstrated that ferroptosis may participate in regulating sensitivity to chemotherapy and radiotherapy in different types of tumors.

The mechanism of ferroptosis has not been completely elucidated. However, GPX4 has demonstrated to be an essential in ferroptosis (24). One basic function of GPX4 is reducing the peroxides by consuming glutathione (GSH) (32). In 2008, Seiler et al (33) reported that a non-apoptotic form of cell death could be induced by deleting GPX4, and this form of cell death was mediated by lipid oxidation. In 2012, another study demonstrated that inhibiting the cystine/glutamate antiporter, system XC−, would elicit a novel form of cell death. System XC-is a transporter that provides cells with cysteine, which is an important material for GSH biosynthesis (8). GPX4 is the most important GSH-dependent enzyme in inducing this form of cell death (24). They named this novel form of cell death ferroptosis (10,24). Although Kagan et al (34) demonstrated that one of the earliest events following GPX4 deletion is cardiolipin oxidation and phospholipid oxidation, the downstream mechanisms following GPX4 deletion remain largely unknown, and this mechanisms requires further study.

The present study, when taken in conjunction with the previous research, provides evidence of the therapeutic possibility of triggering ferroptosis in cancer cells. However, further research is required to understand more about ferroptosis, before it can be used as a cancer therapeutic in a clinical setting.

Acknowledgements

The authors would like to thank Professor Gang Wu and Professor Gang Peng for their advice and suggestions during the experiments.

Funding

The present study was supported by the Affiliated Hospital of Guangdong Medical College (grant no. 2017031795/1814).

Availability of data and materials

All data generated or analyzed during this study are included in this published article.

Authors' contributions

XP performed the colony formation assays and cell death assays. DJ established the radioresistant cell lines and performed the western blot analysis. YY and DY cultured the cells and participated in revising the manuscript. HZ and DZ analyzed the data. GP, ZC and ZY designed and instructed the study. SL, ZL and XP contributed to writing and revising the manuscript. GP gave final approval of the version to be published. All authors read and approved the final manuscript.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References