Chlamydia trachomatis induces lncRNA MIAT upregulation to regulate mitochondria‐mediated host cell apoptosis and chlamydial development

Abstract Chlamydia trachomatis persistent infection is the leading cause of male prostatitis and female genital tract diseases. Inhibition of host cell apoptosis is the key to maintaining Chlamydia survival in vivo, and long noncoding RNAs (lncRNAs) play important roles in its developmental cycle and pathogenesis. However, it is not clear how lncRNAs regulate persistent Chlamydia infection. Here, using a microarray method, we identified 1718 lncRNAs and 1741 mRNAs differentially expressed in IFN‐γ‐induced persistent C. trachomatis infection. Subsequently, 10 upregulated and 5 downregulated differentially expressed lncRNAs were verified by qRT–PCR to confirm the reliability of the chip data. The GO and KEGG analyses revealed that differentially regulated transcripts were predominantly involved in various signalling pathways related to host immunity and apoptosis response. Targeted silencing of three lncRNAs (MIAT, ZEB1‐AS1 and IRF1) resulted in increased apoptosis rates. Furthermore, interference with lncRNA MIAT caused not only an obvious downregulation of the Bcl‐2/Bax ratio but also a marked release of cytochrome c, resulting in a significantly elevated level of caspase‐3 activation. Meanwhile, MIAT was involved in the regulation of chlamydial development during the persistent infection. Collectively, these observations shed light on the enormous complex lncRNA regulatory networks involved in mitochondria‐mediated host cell apoptosis and the growth and development of C. trachomatis.

One of the most successful studies was the IFNγ-induced persistent C. trachomatis infection model, 5 and it has been shown that under the treatment of IFNγ, C. trachomatis forms an atypical infective state in host cells. Although inclusion bodies still exist, they have no infective capacity. 6 Under the electron microscope, a large number of abnormally enlarged and loose reticulate bodies (RBs) were found in the inclusions, which we called abnormal bodies (ABs).
Mature elementary bodies (EBs) are not seen or are rarely seen. 7 Apoptosis is an important defence mechanism against pathogens. 8 The regulation of C. trachomatis infection on apoptosis is quite complex. This may be due to the combination of host immune clearance and Chlamydia immune escape. At present, it is believed that C. trachomatis has a significant time effect on the regulation of cell apoptosis, which can inhibit the apoptosis of host cells in the early stage of infection and induce the apoptosis of host cells in the late stage of infection. 9,10 Therefore, the molecular mechanism of C. trachomatis antiapoptotic activity may be one of the causes of persistent host infection.
Studies have shown that intracellular pathogens can regulate cell signalling pathways and achieve immune escape by regulating the expression profiles of host lncRNAs, 11 such as Rickettsial, 12 Mycobacterium tuberculosis, 13 Helicobacter pylori 14 and Listeria monocytogenes. 15 C. trachomatis, as an intracellular pathogen, causes physiological changes in the host similar to those of the virus and changes the expression profile of the host mRNA, thus promoting its infection. However, lncRNAs can affect the expression of host mRNA through direct binding of mRNA or competitive binding of miRNA, 16 which suggests that C. trachomatis may construct a complex intracellular regulatory network by changing the expression profile of host lncRNAs and thus fine-regulated cellular gene expression. To influence the process of persistent infection and C. trachomatis intracellular survival, the interaction mechanism remains to be further studied. The role of lncRNAs in bacterial infection has aroused widespread concern, 17,18 and the molecular mechanism of C. trachomatis pathogenesis has not been fully elucidated; therefore, the study of host lncRNAs in C. trachomatis infection may provide new ideas for the treatment of Chlamydia infection.
In the present study, transcriptomic profiles of lncRNAs and mRNAs This study may provide new insights into the pathogenic mechanisms associated with persistent C. trachomatis infection.
trachomatis E strain used in this study was cultured as described in the previous research. 19 The standard strain of C. trachomatis serovar E was obtained from ATCC (ATCC# VR-348B, BOUR, GenBank# JX559522).

| Persistent C. trachomatis infection
The cultured HeLa cells were inoculated in 6-or 24-well plates and cultured at 37°C with 5% CO 2 overnight. Then, 1 ml of DMEM was added to each 6-well plate (24-

| RNA extraction and lncRNA microarray
Total RNA was isolated from IFNγ-induced C. trachomatis infection and IFNγ control samples using TRIzol reagent (Invitrogen). We employed Arraystar Human LncRNA Microarray V5.0, which was designed for the global profiling of human lncRNAs and protein-coding transcripts. Approximately 39,317 lncRNAs and 21,174 coding transcripts could be detected by third-generation lncRNA microarray.
The protocol was as follows: First, mRNA was purified from total RNA after the removal of rRNA (mRNA-ONLY™ Eukaryotic mRNA Isolation Kit, Epicentre). Second, each sample was amplified and transcribed into a fluorescent cRNA using a random primer method. Third, the labelled cRNAs were purified using an RNeasy Mini Kit (Qiagen), and the concentration and activity were detected by a NanoDrop ND-1000. Fourth, chip hybridization was performed. Fifth, the hybrid chip was washed, fixed and scanned (Agilent DNA Microarray Scanner (Part Number G2505C)). Differentially expressed lncRNAs and coding transcripts with statistical significance between the two groups were identified through p-value and fold change filtering.

| Quantitative real-time PCR
The total RNA of each sample was measured by a nucleic acidprotein analyser and then reverse-transcribed to cDNA by a reverse-transcription kit (Tiangen) according to the manufacturer's instructions. qRT-PCR was performed using SYBR Green premix (Tiangen) in a LightCycler 96 apparatus (Roche, Basel, Switzerland).
Amplification was performed for 600 s at 95°C, followed by 40 cycles of 95°C for 15 s and 60°C for 30 s. 18S rRNA acted as an internal control. Three parallel replicates were performed for each sample.

| Bioinformatics analysis
To display the differential expression of lncRNAs and mRNAs more intuitively, MeV V4.9.0 software was used in this study to cluster the data and obtain the heatmap. Then, Gene Ontology (GO) and Kyoto Encyclopaedia of Genes and Genomes (KEGG) analyses of chip data were performed using the R language package to evaluate the functions of differentially expressed genes. The lncRNA-miRNA-mRNA interaction network was constructed using the RNA-RNA database (http://starb ase.sysu.edu.cn/index.php) and visualized using Cytoscape 3.6.1 (https://cytos cape.org/).

| RNA interference
Small interfering RNAs (siRNAs) corresponding to the sequences of MIAT, ZEB1-AS1 and IRF1, which were used to inhibit endogenous expression of the above three lncRNAs, and the negative control (NC) siRNAs, which exhibited no downregulation of any HeLa cell genes, were synthesized by Ribio. Transfection was performed with Lipofectamine 3000 (Invitrogen). Cells were transfected with 50 nm of each siRNA.

| Immunofluorescence analysis
The cells were fixed with 4% paraformaldehyde for 30 min, and then 0.1% Triton X-100 was added and allowed to permeate for 10 min. DMEM containing 10% FBS was used for blocking for 1 h.
After washing with PBS twice, primary antibodies (rabbit anti-C. trachomatis, rabbit anti-cytochrome c) were diluted with DMEM at a ratio of 1:500, mixed and incubated in a 37°C incubator for

| Western blot analysis
Total protein samples were collected for 24 h after infection, and the protein concentration was determined by the BCA method. Then, SDS-PAGE was carried out. According to the molecular weight of the target protein, the transfer time was set to transfer the protein to the PVDF membrane. After the PVDF membrane was blocked with 5% non-fat milk, the primary antibody was incubated overnight, and then the secondary antibody was incubated after the membrane was washed. Finally, the results were visualized using an enhanced chemiluminescence western blot system G: BoxChemi XXX9 (Syngene, Cambridge, UK). The densities of the protein bands were analysed by Quantity One (Bio-Rad, USA).

| Statistical analysis
All data were presented as X ± SD. Data were analysed and visualized with SPSS 18.0 and GraphPad Prism 5.0. The statistical significance of differences between different groups was analysed with a two-tailed Student's t-test. Statistical significance was set at p < 0.05.

| Validation of persistent C. trachomatis infection in HeLa cells
In our analysis, we established a persistent C. trachomatis-infected

| Microarray data analysis of differentially expressed lncRNAs and mRNAs
To identify changes in the expression levels of lncRNAs and mRNAs in an in vitro model of persistent C. trachomatis infection, total RNA was extracted from IFNγ-induced persistent C. trachomatisinfected or IFNγ control groups at 12 hpi and 24 hpi. We em- which also provided the basis for subsequent screening and identification. Table S2 and Table S3 summarize the top 20 differentially expressed lncRNAs and mRNAs.

| GO and KEGG analyses of differentially expressed mRNAs
During persistent C. trachomatis infection, uncovering the links between core functions or pathways with these differentially expressed mRNAs will further help us obtain deep insight into the key roles of these genes. The differentially expressed mRNAs showed an obvious trend of amplification at 24 hpi compared with 12 hpi; therefore, the 24 hpi chip data were analysed by GO and KEGG analysis. GO analysis was performed by biological process (BP), cellular components (CC) and molecular function (MF) and was composed of three parts. We found that the upregulated differential mRNAs were mainly enriched in biological processes related to nucleic acid metabolic process and RNA metabolic process, cellular components related to nuclear part and nuclear lumen, and molecular functions related to RNA binding and protein binding ( Figure 3A, C, E).
However, the downregulated differentially expressed mRNAs were mainly concentrated in biological processes related to cell fate specification and inclusion body assembly, cellular components related to intraciliary transport particle B and intrinsic components of the membrane, and molecular functions related to RAGE receptor binding and NAD binding ( Figure 3B, D, F).
Pathway analysis of 1741 differentially expressed mRNAs was performed using KEGG analysis. In the persistent infection group, the upregulated differential mRNAs were mostly enriched in the NOD-like receptor signalling pathway, herpes simplex infection, the TNF signalling pathway and the RIG-I-like receptor signalling pathway ( Figure 3G). In addition, the downregulated differential mRNAs were mostly enriched in the TGFβ, calcium and Hippo signalling pathways ( Figure 3H). These results suggested that C. trachomatis might evade host immunity by affecting host signal transduction in persistent infection.

| Validation of differentially expressed lncRNAs by quantitative real-time PCR
The differential expression levels of lncRNAs screened in our micro-  Figure 2E). showed that the apoptosis rates of the Ct-siMIAT, Ct-siZEB1-AS1

| LncRNA-miRNA-mRNA interaction network analysis
The ceRNA regulatory mechanism is very important between lncR-NAs and mRNAs. Therefore, the analysis of specific RNA-RNA interactions has become the key to exploring the mechanism of interaction between Chlamydia and the host. We predicted miRNAs and mRNAs of three lncRNAs (MIAT, ZEB1-AS1 and IRF1) based on ENCORI and compared these mRNAs with the differentially expressed mRNAs that were identified in our chip results. To make the network more concise and effective, we further filtered the key mRNAs closely related with cell fate and the interferon response pathway ( Figure 3G, H). A total of 29 mRNAs were selected for further lncRNA-miRNA-mRNA network construction ( Figure 6A). These ceRNA regulatory relationships may be more important in the pathogenesis of persistent C. trachomatis infection considering the complexity amongst the lncRNAs and mRNAs. Furthermore, the ClueGo Plugin was used to analyse the targeted mRNAs of the three lncRNAs. Interestingly, we found that MIAT was involved in the response to mitochondrial depolarization ( Figure 6B). Therefore, the following work will focus on the specific mechanism of MIAT.

| The antiapoptotic effects of MIAT are related to the mitochondrial pathway
In recent years, many studies have shown that the mitochondrial pathway in the cell survival mechanism plays a vital role, 26,27 and in numerous apoptosis control genes, the Bcl-2 and caspase families are of the most concern. 28,29 Amongst them, the Bcl-2 and Bax genes are known to regulate the apoptosis regulation function of a counterpart pair of the most important control genes. Caspase-3 is the most critical apoptotic executive protease during the process of apoptosis. [30][31][32] To further understand the antiapoptotic mechanism of MIAT in persistent infection, we tested the classical apoptosisrelated molecules Bax, Bcl-2 and cleaved caspase-3 in the Ct-siMIAT group. Compared with the Ct-siNC group, the Bcl-2/Bax ratio of the Ct-siMIAT group was downregulated, and the ratio of cleaved caspase-3 to the internal reference was increased (p < 0.05) ( Figure 7A).

Decreased MMP is one of the important factors leading to apoptosis
and is considered to be the first step of the apoptosis cascade. 33 To further investigate whether MIAT affected mitochondrial per-

| Inhibition of MIAT promotes C. trachomatis growth and development
During persistent Chlamydia infection, in addition to changes in morphology, the expression profiles of genes involved in cell biological processes such as membrane structure, energy metabolism and development cycle regulation also show different degrees of change. [34][35][36] Abnormal bodies showed decreased metabolic activity and replication ability in cells. German researchers studied the regulation of mitochondrial fission regulators by host miRNAs, which, in turn, severely affected chlamydial growth and had a marked effect on the mitochondrial network. 37 A reduced mitochondrial membrane potential has been shown to correlate with the reduced antiviral response. 38 Because the functional lncRNA MIAT can cause decreases F I G U R E 4 Validation of the microarray data using qRT-PCR. All data are representative of three independent experiments in the host cell mitochondrial membrane potential, we hypothesized that MIAT may influence chlamydial development. Therefore, we examined the effect of MIAT on C. trachomatis development. When MIAT was silenced, we found that the size of the Chlamydia inclusions was larger at 24 hpi. To investigate whether the decrease in mitochondrial membrane potential is conducive to the transformation of RB to EB, thus affecting the development of Chlamydia, we treated CCCP for 20 min as the positive control and then collected the progeny Chlamydia for reinfection counts ( Figure 8A). Each group was given an identical volume to infect fresh HeLa cell monolayers.
Compared with the Ct-siNC group, the Ct-siMIAT group progeny had (5.18 ± 0.25) × 10 6 IFUs, which was higher than that of the Ct-siNC group, at (3.26 ± 0.35) × 10 6 IFUs. The IFUs of C. trachomatis were significantly increased (p < 0.01) ( Figure 8B). The average diameter of the inclusions in the Ct-siMIAT group was 1.5-fold that of the negative control group (p < 0.01) ( Figure 8C). Except for their larger size,

| DISCUSS ION
Persistent C. trachomatis infection is considered a survival strategy to evade host immune effects in vitro and in vivo. [39][40][41] Most previous studies have been based on the interaction between bacterial virulence proteins and host coding genes. [42][43][44] However, interestingly, 98.5% of the human genome is noncoding, and lncRNAs have emerged as important regulators of gene expression. A large amount of evidence has shown that lncRNAs are involved in the regulation of apoptosis pathways. 45,46 However, apoptosis is an important host defence mechanism against pathogens, and the regulation of C. trachomatis infection on apoptosis is quite complex. [47][48][49] Currently, it is believed that C. trachomatis can proliferate itself by resisting host cell apoptosis, and some studies have also shown that the enlargement of Chlamydia inclusions would promote cell apoptosis. 50 In this study, we combined our previous work 51,52 to propose the scientific hypothesis that persistent C. trachomatis infection-related infection, most of which had the same differential expression ( Figure   S1 and Table S4). From these chip data, we determined that a large number of differentially expressed lncRNAs were induced by C. trachomatis and that the pORF5 plasmid protein is an important virulence factor for C. trachomatis. Based on these findings, we believe that the lncRNA profile of the pORF5 deletion strain can explain whether the antiapoptotic activity of lncRNA induced by C. trachomatis requires the involvement of pORF5. This will help us further study the key molecules responsible for C. trachomatis infection.
Numerous studies have shown that lncRNAs affect pathogen survival by regulating cell metabolism and innate immunity. Subuddhi promoted interferon responses by acting as positive feedback for RIG-I signalling and then controlled Hantaan virus replication. 60 In adapting to intracellular life, Chlamydia has lost energy-generating capability but has modified nucleotide transport functions to exploit the host ATP. 61 In this study, we found that MIAT can influ- is generally believed that lncRNA localization in the nucleus is transcriptional regulation and plays a role in the nucleus by internuclear factors. 64 Through the lncATLAS database, we found that MIAT accounted for 99.1% of nuclear localization; thus, we speculated that it might play an important role in the regulation of chromatin, transcriptional regulation and variable shear regulation. LncRNAs located in the cytoplasm may regulate gene expression at the posttranscriptional level through the ceRNA mechanism 65 ; for example, ZEB1-AS1 can affect the expression of host mRNA through direct binding of mRNA or competitive binding of miRNA. 66 According to the analysis, the majority of IRF1 was located in the cytoplasm and exosomes. Therefore, the study of its cellular sublocalization becomes a key step to explore the mechanism in depth. The regulation of lncRNAs in persistent C. trachomatis infection is a synergistic effect of multiple mechanisms. Our study implies that MIAT is not only a previously unknown molecule for C. trachomatis growth suppression following persistent infection but also a potential target for future therapy against C. trachomatis-mediated chronic inflammation.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflicts of interest with the contents of this article.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon request.