Abstract
The Janus kinase-signal transducer and activator of transcription-3 (JAK-STAT3) signaling pathway is a key regulator of cell growth, motility, migration, invasion and apoptosis in mammalian cells. Infection with intracellular pathogens of the genus Chlamydia can inhibit host cell apoptosis, and here we asked whether the JAK-STAT3 pathway participates in chlamydial anti-apoptotic activity. We found that, compared with uninfected cells, levels of JAK1 and STAT3 mRNA as well as total and phosphorylated JAK1 and STAT3 protein were significantly increased in Chlamydia psittaci-infected HeLa cells. Moreover, the apoptosis rate of infected cells was higher after treatment with the tyrosine kinase inhibitor AG-490 (2-cyano-3-(3,4-dihydroxyphenyl)-N-(phenylmethyl)-2-propenamide). Immunoblotting of apoptosis-related proteins showed that C. psittaci infection reduces Bax, but increases Bcl-2, protein levels, resulting in reduced activation of caspase-3, caspase-7, caspase-9 and PARP; AG490 attenuates these effects. Together, our data suggest that the JAK/STAT3 signaling pathway facilitates the anti-apoptotic effect of C. psittaci infection by reducing the Bax/Bcl-2 apoptotic switch ratio, and by inhibiting the intracellular activation of key pro-apoptotic enzymes.
INTRODUCTION
Chlamydia psittaci is an obligate intracellular bacterium that causes zoonotic diseases, such as pneumonia, conjunctivitis, abortion in female livestock and reduced survival rate in young animals, in a wide range of mammalian hosts (Kaleta and Taday 2003). It can also infect humans through close contact with infected mammals or birds, and human-to-human transmission has also been suggested (Stenzel, Pestka and Choszcz 2014; Ran et al.2017). In humans, C. psittaci can cause psittacosis, community-acquired pneumonia, chronic obstructive pulmonary disease, encephalitis, endocarditis and arthritic diseases, and it is believed to be a possible etiologic agent of lymphomas of the ocular adnexa (Lee et al.2014). Chlamydia spp. infections are often acute, but persistent infection is also an important characteristic of the host–parasite interaction, and plays a significant role in spreading the organism within animal populations (Schoborg and Borel 2014; Zhu et al.2014). Persistence can be a by-product of inhibition of apoptosis and then offer an alternative pathogenic mechanism for chlamydial scarring resulting in chronic chlamydiosis and serious sequelae (Dean and Powers 2001).
The induction of apoptotic resistance is thought to be an important immune escape mechanism allowing Chlamydia spp. to acquire nutrition and energy from host cells and thus to maintain their life cycles. Several anti-apoptotic mechanisms associated with chlamydial infection have been reported. Rajalingam and others (Rajalingam et al.2008; Du et al.2011) found that infection with C. trachomatis could upregulate the anti-apoptotic Bcl-2 family member Mcl-1 by activation of the Raf/MEK/ERK and PI3K/AKT pathways. Du et al. (2013) showed that Chlamydia spp. inhibits host cell apoptosis through induction of Bag-1 via the MAPK/ERK pathway, while Matsuo et al. (2013) found that Protochlamydia sp. R18, a so-called primitive chlamydial species, induces host cell apoptosis through mitochondrial dysfunction mediated by a chlamydial protease-like activity factor.
Signal transducer and activator of transcription-3 (STAT3) is a key member of the Janus kinase (JAK)–signal transducer and activator of transcription (STAT) pathway, which is involved in numerous physiological processes (Hirano, Ishihara and Hibi 2000). STAT3 proteins are transcription factors that exist in a latent, monomeric form in the cytoplasm. They are activated by a variety of ligands, such as Interleukin 6 (IL-6), Tumor necrosis factor α (TNF-α) and Vascular endothelial growth factor (VEGF) (Hirano, Ishihara and Hibi 2000), which bind to the corresponding cell surface receptor, resulting in receptor dimerization and the subsequent activation of JAK tyrosine kinases. Specific tyrosine residues on the receptor are then phosphorylated by activated JAKs and serve as docking sites for cytoplasmic transcription factors of the STAT family. STAT3 is phosphorylated by JAK1 and JAK2, then dimerizes, subsequently leaves the receptor and translocates to the nucleus, where it binds specific DNA sequences, promoting the transcription of target genes (Shuai and Liu 2003). Recent studies have identified STAT3 activation as a key event in regulating cell growth, motility, migration, invasion and angiogenesis (Hirano, Ishihara and Hibi 2000; Xiong et al.2008; Subramaniam et al.2013; Habibie et al.2014). It is also widely reported that the JAK-STAT3 signaling pathway is aberrantly expressed and constitutively activated in a broad range of human malignancies, and can inhibit tumor cell apoptosis (Subramaniam et al.2013; Habibie et al.2014; Gaykalova et al.2015). Inhibitors that block the activity of JAK/STAT3 signaling pathways are emerging as new drugs with demonstrated efficacy in various pathological models, one of them is Tyrphostin AG-490. AG490 (2-cyano-3-(3,4-dihydroxyphenyl)-N-(phenylmethyl)-2-propenamide) is a well-known ATP-competitive inhibitor, which blocks the downstream counterpart substrates belonging to STAT family, and widely used to characterize the role of JAK/STAT3 pathway in cell signaling and many authors use AG 490 as a specific inhibitor of this pathway (Xiong et al.2008; Goncalves et al.2010; Fofaria and Srivastava 2015).
After infection, chlamydiae can induce host cells to produce a series of cytokines, including IL-6, TNF-α and Interferons (IFNs) (Cerny et al.2015). It is also known that there is cross-talk between different cytokine-signaling cascades involved in the regulation of the JAK/STAT pathways. For example, IL-6 stimulation can activate both STAT1 and STAT3 (Shuai and Liu 2003). Phosphorylated active STAT3 (p-STAT3) exerts an anti-apoptotic effect, while phosphorylated STAT1 (p-STAT1) has been described as a pathogen suppressor because of its function as an immunosurveillance mediator and its pro-apoptosis activity (Bowman et al.2000; Lim et al.2016). Lad et al. (2005) have confirmed that JAK/STAT1 pathway can be activated by C. trachomatis infection and then inhibit Chlamydia growth. Therefore, we hypothesize that chlamydial infection can also activate the JAK-STAT3 signal pathway, which may be responsible for the associated inhibition of apoptosis.
MATERIALS AND METHODS
Chlamydia psittaci and cell culture
Human cervical epithelial HeLa 229 cells (ATCC CCL-2.1) were used for culturing C. psittaci. HeLa 229 cells were maintained in Dulbecco Modified Eagle Medium (DMEM) (Hyclone, Logan, USA) supplemented with 10% (v/v) fetal bovine serum (FBS) (Gibco BRL, Gaithersburg, USA) and 2 mmol/l L-glutamine (Sigma-Aldrich, Munich, Germany) at 37˚C in an incubator supplied with 5% CO2. Chlamydia psittaci 6BC (ATCC VR-125) was propagated in confluent HeLa 229 cell monolayers in complete growth medium with 10% FBS and 2 μg/ml cycloheximide (Sigma-Aldrich) as described previously (Cai et al.2016; Liang et al.2016). Chlamydial elementary bodies (EBs) were harvested at 48 h post-infection (p.i.). Cells containing mature EBs were centrifuged at 1000 rpm for 10 min and then washed twice with phosphate-buffered saline (PBS) (pH 7.4). Cell pellets were resuspended in sucrose-phosphate-glutamate buffer (0.2 mol/l sucrose, 3.8 mmol/l KH2PO4, 6.7 mmol/l Na2HPO4, 5 mmol/l L-glutamic acid, pH 7.4), disrupted by sonication and stored at –80˚C until use.
Infection and induction of apoptosis
For experiments designed to examine whether the JAK-STAT signaling pathway participates in C. psittaci anti-apoptosis activity, HeLa 229 monolayers were grown in six-well cell culture plates at a density of 1.5 × 106 cells per well, and infected with C. psittaci at a multiplicity of infection (MOI) of 3. After 2-h adsorption, inocula were removed and the medium was replaced with complete DMEM containing 10% FBS and 2 μg/ml cycloheximide with or without 65 μM AG490 (Sigma-Aldrich, Munich, Germany). After 19-h incubation, 1 μM staurosporine (STS, Sigma-Aldrich) was added to induce apoptosis over a period of 5 h (Du et al.2013).
Detection of apoptosis
Apoptosis was determined by flow cytometry using a Annexin V-FITC/PI Detection Kit (BioVision, Mountain View, California, USA) according to the manufacturer's instructions. Briefly, cells were collected, washed with PBS and resuspended in 500 μl binding buffer containing 5 μl propidium Iodide (PI) and 5 μl fluorescein isothiocyanate (FITC), and then incubated for 5 min in the dark at room temperature. The stained cells were then immediately analyzed with a FACSCalibur System flow cytometer (BD Biosciences, La Jolla, CA). For all assays, 10 000 events were counted and the percentage of apoptotic cells was determined using ModFit software (BD Biosciences).
Quantitative real-time PCR
HeLa 229 monolayers were infected with C. psittaci at an MOI of 3. After 2, 4, 6, 12, and 24 h p.i., 1 μM STS was added, and 5 h later, infected cells were removed and total RNA was extracted using a TRIzol Plus RNA Purification Kit (Tiangen, Beijing, China). Total RNA was reverse-transcribed into cDNA using the Transcriptor First Strand cDNA Synthesis Kit (Roche, Basel, Switzerland). cDNA was mixed with TaqMan universal PCR master mix (Thermo Fisher Scientific, MA, USA) and the respective TaqMan probes. The primers were designed as described (Tanaka et al.2012; Zhang et al.2014; Zhou et al.2014). For STAT3, 5΄-GCTTTTGTCAGCGATGGAGTA-3΄ (forward), 5΄-CCACCGCATCTCTACATTCAA-3΄ (reverse); for JAK1, 5΄-GTCACAACCTCTTTGCCCTGTAT-3΄ (forward), 5΄-CGGAGGGACATCTTGTCATCA-3΄ (reverse); for JAK2, 5΄-AGCCTATCGGCATGGAATATCT-3΄ (forward), 5΄- TAACACTGCCATCCCAAGACA-3΄ (reverse), for β-actin, 5΄-TGACGTGGACATCCGCAAAG-3΄ (forward), 5΄-CTGGAAGGTGGACAGCGAGG-3΄ (reverse). PCR was performed in triplicate using the ABI Prism 7000 sequence detection system (Applied Biosystems). The melting curves of the amplification products were analyzed to ensure single product formation. The relative gene expression ratios of the target genes were calculated based on the real-time amplification efficiency and cycle threshold deviation of a given sample versus the control (uninfected cells) in comparison to a reference gene, the β-actin gene.
Western blot analysis
HeLa 229 monolayers were infected with C. psittaci at an MOI of 3 for 23 h, then 150 μM of clasto-Lactacystin β-lactone (Sigma-Aldrich, Munich, Germany), dissolved in methyl acetate, was added, and after 1 h treatment, the cells were lysed in RIPA reagent (Beyotime, Hangzhou, China) containing protease inhibitors and Phosphatase Inhibitor Cocktail (Cwbio, Beijing, China). For quantitation of apoptosis-related proteins, 65 μM AG490 was added in complete DMEM. Equal amounts of protein (100–200 μg/lane) from whole cell lysates were separated by 8% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to nitrocellulose (Amersham, Amersham, UK). After blocking with 5% dry non-fat skimmed milk, the membranes were probed with specific primary antibodies, and then with the appropriate Horseradish peroxidase (HRP)-conjugated secondary antibodies. Target proteins were detected using an enhanced chemiluminescence detection kit (Beyotime, Shanghai, China). β-Actin was used as control. The GelGDoc2000 imaging system (Bio-Rad, Munich, Germany) was employed to analyze the bands, and the protein level was represented by the relative optical density (ROD) × area (mm2) of the band. Antibodies used in this study were purchased from Cell Signaling Technology (Danvers, MA, USA), including β-actin, JAK1 (Cat#3344), pJAK1Tyr1022/1023 (Cat#3331), STAT3 (Cat#4904), pSTAT3Tyr705 (Cat#9145), Bcl-2 (Cat#2870), Bax (Cat#5023), caspase-3 (Cat#9665), cleaved caspase-3 (Asp175) (Cat#9664), caspase-7 (Cat#12 827), cleaved caspase-7 (Asp198) (Cat#9491), caspase-9 (Cat#9502), cleaved caspase-9 (Asp330) (Cat#9501), PARP (Cat#9532) and cleaved PARP (Asp214) (Cat#5625).
Statistical analyses
Results were expressed as the mean ± SD. One-way analysis of variance (ANOVA) (www.physics.csbsju.edu/stats/anova.html) was performed on the mRNA quantification, normalized protein level and rate of apoptotic cell data from multiple groups, while a two-tailed Student t-test (Microsoft Excel) was used to compare two given groups. Results were considered significant if the P value was <0.05.
RESULTS
Chlamydia psittaci infection has an anti-apoptotic effect on host cells
The protein kinase inhibitor staurosporine (STS) is known to induce apoptosis in mammalian cells by activation of caspase-3 (Sergeeva et al.2017). Thus, when HeLa cells were treated with STS, the proportion of apoptotic cells in the culture, as judged by expression of annexin V and permeability to PI, increased almost 3-fold (quadrant Q2 in Fig. 1A). However, infection of STS-treated cells with C. psittaci resulted in a marked reduction in apoptotic cells, from 19.1% to 10.7% of the population. Indeed, even in STS untreated HeLa cells, C. psittaci infection led to a reduction in the rate of apoptosis (Fig. 1).
The effect of C. psittaci on host cell apoptosis. HeLa 229 monolayers were infected with C. psittaci at a MOI of 3. After incubation for 19 h, 1 μM staurosporine (STS) was added and cells were analyzed 5 h later using a FITC-annexin V/PI flow cytometric assay. (A) Representative FACS plot of apoptotic cells in each group. The apoptotic cells are represented by those in Q2. (B) The proportion of apoptotic cells from pooled data of three independent experiments, where error bars show SD. The proportion of apoptotic cells were first analyzed by ANOVA and reanalyzed between two given groups (that displayed significant differences under ANOVA) by using a two-tailed Student t test (**, P < 0.01).
The JAK/STAT3 signal pathway is implicated in the anti-apoptotic effect of Chlamydia psittaci
Two series of experiments were performed to examine whether the JAK/STAT3 signal pathway is involved in the anti-apoptotic effect of C. psittaci in HeLa cells. In the first series, we measured the levels of mRNA and phosphorylated proteins of key members of the pathway following infection. Compared with uninfected HeLa cells, JAK1 mRNA levels were significantly upregulated in C. psittaci-infected cells at 4 h p.i., and STAT3 was upregulated at 6 and 12 h p.i., while there was no obvious change in JAK2 mRNA levels throughout the whole infection process; when HeLa cells were treated with STS, the JAK1 and STAT3 mRNA levels were also upregulated a little bit in C. psittaci-infected cells (Fig. 2). We also examined the levels of total and phosphorylated JAK1 and STAT3 proteins, and found that these were also increased in C. psittaci-infected HeLa cells than uninfected ones (Fig. 3).
The mRNA expression levels of STAT3, JAK1 and JAK2 in C. psittaci-infected cells. HeLa 229 monolayers were infected with C. psittaci at an MOI of 3. After 2, 4, 6, 12, and 24 h p.i., 1 μM STS was added, and 5 h later total RNA was extracted from infected cells and then reverse transcribed into cDNA; mRNA levels for STAT3 (A), JAK1 (B) and JAK2 (C) were evaluated by real-time PCR using specific primers. Each mRNA level was normalized against that of the housekeeping gene β-actin. Results obtained from three independent experiments are expressed as mean ± SD. The data were first analyzed by ANOVA and reanalyzed between two given groups (that displayed significant differences under ANOVA) by using a two-tailed Student t test (**, P < 0.01).
The protein expression levels of STAT3, p-STAT3, JAK1 and p-JAK1 in C. psittaci-infected HeLa cells. HeLa 229 monolayers were infected with C. psittaci at an MOI of 3. After incubation for 19 h, 1 μM STS was added, and 5 h later, the cells were lysed with RIPA reagent with clasto-Lactacystin β-lactone treatment, and the expression of STAT3, p-STAT3, JAK1 and p-JAK1 was evaluated by western blot (A). The expression of β-actin was used as an internal control. The relative expression levels of STAT3, p-STAT3 (B), JAK1 and p-JAK1 (C) were quantified by densitometry and normalized to the control level. Results obtained from three independent experiments are expressed as mean ± SD. The data were first analyzed by ANOVA and reanalyzed between two given groups (that displayed significant differences under ANOVA) by using a two-tailed Student t test (**, P < 0.01).
In the second series of experiments, we tested the effect of JAK inhibitor AG490 on HeLa cell apoptosis by flow cytometry. The results showed that there are more apoptotic cells in AG490-treated C. psittaci-infected cells than in untreated cells, and the difference was also found in uninfected groups (Fig. 4).
The effect of JAK inhibitor AG490 on host cell apoptosis after C. psittaci infection. HeLa 229 monolayers were infected with C. psittaci at an MOI of 3, and treated with 65 μM AG490. After incubation for 24 h, cells were analyzed using a FITC-annexin V/PI flow cytometric assay. (A) Representative FACS plot of apoptotic cells in each group. The apoptotic cells are represented by those in Q2. (B) The proportion of apoptotic cells from pooled data of three independent experiments, where error bars show SD. The data were first analyzed by ANOVA and reanalyzed between two given groups (that displayed significant differences under ANOVA) by using a two-tailed Student t test (**, P < 0.01).
Chlamydia psittaci downregulates Bax and upregulates Bcl-2 via the JAK/STAT3 signaling pathway
We investigated the expression of pro-apoptotic protein Bax and anti-apoptotic protein Bcl-2 by immunoblotting. The results showed that Bax was downregulated, but Bcl-2 was upregulated in C. psittaci-infected cells either with or without STS treatment; and AG490 could upregulated Bax but downregulated Bcl-2 (Fig. 5A and B). The ratio of Bax/Bcl-2 in AG490-treated C. psittaci-infected HeLa cells was much higher than that in AG490 untreated infected cells either with or without STS treatment, even in uninfected HeLa cells (Fig. 5C).
Expression levels of Bax and Bcl-2 proteins in C. psittaci-infected HeLa cells. HeLa 229 monolayers were infected with C. psittaci at an MOI of 3, and treated with 65 μM AG490, as shown in the key. After incubation for 24 h, the cells were lysed with RIPA reagent with clasto-Lactacystin β-lactone treatment, and the expression of Bax and Bcl-2 was evaluated by western blot (A). The expression of β-actin was used as an internal control. The relative expression levels of Bax and Bcl-2 (B) were quantified by densitometry and normalized to the control, and the Bax/Bcl-2 ratios were evaluated (C). Results obtained from three independent experiments are expressed as mean ± SD. The data were first analyzed by ANOVA and reanalyzed between two given groups (that displayed significant differences under ANOVA) by using a two-tailed Student t test (**, P < 0.01).
Chlamydia psittaci reduces caspase-3, caspase-7, caspase-9 and PARP activation via the JAK/STAT3 signal pathway
We also examined the effect of C. psittaci infection on the expression and activation of caspase-3, caspase-7, caspase-9 and PARP, enzymes that are directly involved in apoptosis. As shown in Fig. 6, C. psittaci infection led to a reduction in levels of activated caspase-3, caspase-7, caspase-9 and PARP, and AG490 was able to attenuate these effects.
Protein expression level of caspase-3, caspase-7, caspase-9 and PARP in C. psittaci-infected HeLa cells. HeLa 229 monolayers were infected with C. psittaci at an MOI of 3, and treated with 65 μM AG490 with or without 1 μM STS, as indicated. After incubation for 24 h, the cells were lysed with RIPA reagent with clasto-Lactacystin β-lactone treatment, and the expression of total and activated caspase-3, caspase-7, caspase-9 and PARP was evaluated by western blot (A). The expression of β-actin was used as an internal control. The relative expression levels of caspase-3 (B), caspase-7 (C), caspase-9 (D) and PARP (E) and their cleaved versions were quantified by densitometry and normalized to the control. Results obtained from three independent experiments are expressed as mean ± SD. The data were first analyzed by ANOVA and reanalyzed between two given groups (that displayed significant differences under ANOVA) by using a two-tailed Student t test (**, P < 0.01).
DISCUSSION
After invading through the upper respiratory tract, Chlamydia psittaci reproduces locally in mononuclear macrophages, then spreads to the lungs and other organs through blood circulation, where the infection leads to a wide variety of diseases. However, no effective vaccine is available, and the pathogenic and host immune evasion mechanisms remain unclear. An improved understanding of these mechanisms is important for the development of treatments aimed at controlling C. psittaci infection.
After infecting host cells, Chlamydia spp. can inhibit host cell apoptosis (West et al.2016). Our flow cytometry results confirmed that this in the case of C. psittaci and showed that the bacterium could inhibit HeLa cell apoptosis with or without staurosporine induction. To test whether the JAK/STAT3 signaling pathway plays a role in this anti-apoptosis activity, we first analyzed the mRNA levels of STAT3, JAK1 and JAK2 in C. psittaci-infected cells and showed upregulation of JAK1 and STAT3, but not JAK2, gene expression. Similarly, the levels of total and phosphorylated JAK1 and STAT3 proteins were significantly enhanced in infected cells. Furthermore, treatment with the JAK-specific inhibitor AG490 significantly increased the cell apoptosis rate after C. psittaci infection. Together, these results show that the JAK/STAT3 signaling pathway is activated and therefore implicated in the anti-apoptotic mechanism induced by C. psittaci.
The Bcl-2 family can be divided into anti-apoptosis proteins and pro-apoptosis proteins (Volkmann et al.2014). Bcl-2 itself is an anti-apoptosis protein and can block all early stages of apoptosis, including the activation of caspase family proteins. In contrast, Bax is a pro-apoptosis protein that can promote several aspects of apoptosis (Volkmann et al.2014). In order to exclude the proteolytic activity of CPAF (chlamydial proteasome-like activity factor), the cell samples were harvested after clasto-Lactacystin β-lactone treatment, conditions under which no CPAF processing can take place (Chen et al.2012; Tan and Sütterlin 2014). We found that C. psittaci infection could downregulate Bax, but upregulate Bcl-2, and that AG490 reduces these effects, which further confirms that C. psittaci inhibits apoptosis and demonstrates a role in this process for the JAK/STAT3 signaling pathway by regulation of Bax and Bcl-2 expression. Other authors (Seo et al.2015; Qiao et al.2016) have reported that, when the JAK/STAT3 pathway is inhibited, apoptosis is stimulated in breast cancer cells and squamous-cell carcinoma cells, as indicated by an increase in Bax/Bcl-2 ratios.
Apoptosis is usually accompanied by the release of cytochrome c, and is followed by activation of caspase (cysteine aspartate protease) family proteins and a change in membrane permeability of mitochondria (Nam et al.2016). Once caspase-3 is activated, a cascade effect ensues, and apoptosis is irreversible. Thus, caspase-3 is considered to be critical for triggering apoptosis. Caspase-7, like caspase-3, is an executioner caspase involved in the degradation of many cellular products during apoptosis (Slee, Adrain and Martin 2001; Du et al.2017), while caspase-9 is an initiator caspase, which activates executioner caspases (Kim, Srivastava and Kim 2015). PARP (poly[ADP-ribose] polymerase) is a nuclear protein that is activated by DNA damage and can initiate apoptosis, but is also itself a caspase substrate, being cleaved by caspase-9, for example (Esmaeili et al.2016; Zheng et al.2016). We found that there is a significant reduction in the activation of caspase-3, caspase-7, caspase-9 and PARP concomitent with C. psittaci infection, which is consistent with its anti-apoptotic effect. The fact that AG490 attenuates these enzyme effects again indicates that the JAK/STAT3 pathway facilitates C. psittaci-induced anti-apoptosis. Ma and others (Li et al.2015; Ma et al.2015) have reported similar outcomes in tumor cells: when JAK/STAT signaling was inhibited, caspase-3, caspase-7 and caspase-9 activation was increased, thus expediting tumor cell apoptosis.
In summary, our findings strongly suggest that the JAK/STAT3 signaling pathway participates in the anti-apoptotic effect of C. psittaci infection in HeLa cells by reducing the Bax/Bcl-2 expression ratio, and by inhibiting the intracellular activation of the pro-apoptotic enzymes caspase-3, caspase-7, caspase-9 and PARP. Further study is required to determine whether the JAK/STAT3 pathway is the only pathway involved or whether multiple signaling pathways regulate the effects of C. psittaci infection.
FUNDING
This work was supported by the National Natural Science Foundation of China (grant no. 81572011, 31600150), the Natural Science Foundation of Hunan Province (2016JJ3103) and the Construct Program of the Key Discipline (Public Health and Preventive Medicine) in Hunan Province.
Conflict of interest. None declared.
