Identification and Analysis of Autophagy-Related Genes as Diagnostic Markers and Potential Therapeutic Targets for Tuberculosis Through Bioinformatics

doi:10.21203/rs.3.rs-4670127/v1

Download PDF

Research Article

Identification and Analysis of Autophagy-Related Genes as Diagnostic Markers and Potential Therapeutic Targets for Tuberculosis Through Bioinformatics

https://doi.org/10.21203/rs.3.rs-4670127/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background: According to the World Health Organization, Mycobacterium tuberculosis (Mtb) infections affect approximately 25% of the world's population. There is mounting evidence linking autophagy and immunological dysregulation to TB, according to many research. As a result, this research set out to discover TB-related autophagy-related biomarkers, gene regulatory networks, and prospective treatment targets.

Methods: We used five autophagy databases to get genes linked to autophagy and GEO databases to get genes connected to TB. Then, functional modules associated with autophagy were obtained by analyzing them using weighted gene coexpression network analysis (WGCNA). Both GO and KEGG were used to examine the ATGs of important modules. Limma, an R tool, was used to identify differentially expressed ATGs, and the external datasets of GSE19435 were used to further confirm their identification. We used DE-ATGs and a protein-protein interaction (PPI) network to search the hub genes. CIBERSORT was used to estimate the kinds and amounts of immune cells. After that, we built a drug-gene interaction network and a network that included messenger RNA, small RNA, DNA, and ceRNA. At last, the differential expression of hub ATGs was confirmed by RT-qPCR, Immunohistochemistry (IHC), and Western blotting (WB). The diagnostic usefulness of hub ATGs was evaluated using receiver operating characteristic (ROC) curve analysis.

Results: Including 508 ATGs, four of the nine modules strongly linked with TB were deemed essential. Three hub genes—IL1B, CAPS1, and STAT1—were identified by intersection out of twenty-two DE-ATGs discovered by differential expression analysis. Research into immune cell infiltration found that TB patients had an increased proportion of plasma cells, CD8 T cells, and M0 macrophages. A competitive endogenous RNA (ceRNA) network utilized 10 long non-coding RNAs (lncRNAs) and 2 miRNAs. Then, the IL1B-targeted drug CANAKINUMAB was assessed using this network. During bioinformatics analysis, three hub genes—IL1B, CAPS1, and STAT1—were validated using WB/RT-qPCR/IHC in mouse and macrophage infection models. In most cases, the new findings corroborated the old ones.

Conclusion: We found that IL1B, CASP1, and STAT1 are important biomarkers for TB. As a result, these crucial hub genes may hold promise as TB treatment targets.

Tuberculosis(TB)

immune infiltration

bioinformatic analysis

autophagy

Tuberculosis (TB), caused by bacteria of the Mycobacterium tuberculosis complex, is one of the oldest diseases known to affect humans. Tuberculosis continues to cause considerable morbidity and mortality globally. According to WHO, an estimated 10.6 million(95% CI: 9.9–11 million) people became newly sick with tuberculosis in 2021. Geographically, in 2021, most people who developed TB were in the WHO regions of South-East Asia (45%), Africa (23%) and the Western Pacific (18%), with smaller proportions in the Eastern Mediterranean (8.1%), the Americas (2.9%) and Europe (2.2%). The 30 countries with a high TB burden account for 87% of all estimated incident cases worldwide. Eight of these countries accounted for more than two-thirds of the global total: India (28%), Indonesia (9.2%), China (7.4%), the Philippines (7.0%), Pakistan (5.8%), Nigeria (4.4%), Bangladesh (3.6%) and the Democratic Republic of the Congo (2.9%)[1].

M. tuberculosis are the bacteria responsible for the airborne illness known as TB. The contagious bacilli may be inhaled by a person's cough, which can speak, sing, and sneeze[2]. Although most people's immune systems can rid themselves of Mycobacterium TB, an estimated 20–25 percent of those exposed get the infection. Infected individuals with M. tuberculosis have a 5% chance of developing active illness within 2 years, but the host immune system controls 90–95% of infections, causing them to stay dormant. About 5–10% of these people may have a reactivation of tuberculosis at some point in their lives[3]. Macrophages are a diverse variety of white blood cells originating from monocytes and are found in many different bodily regions. Their functional role is phagocytosis of fixed and free-floating pathogens and cellular detritus[4]. The major cell type responsible for the first absorption of Mycobacterium TB during infection is lung macrophages[5]. MTB activates Macrophages when interacting with many surface receptors, including Dectin-1, NOD2, toll-like receptor 2 (TLR2), TLR4, and TLR4. Cytokines such as TNF-α, IL-12, and IL-1B have anti-M. tuberculosis characteristics are released due to this activation[6–10].

Regular respiratory cells engage in autophagy as a survival strategy and as part of the body's cellular equilibrium. Autophagy is an essential cellular process that degrades microbes, damaged organelles, and proteins by lysosomal degradation. Microbes include viruses, bacteria, fungi, and protists/protozoa. Autophagy may be broadly classified into three categories, each with its unique entry mechanism into lysosomes: macroautophagy, microautophagy, and chaperone-mediated autophagy[11]. During the formation of autophagosomes, the well-studied kind of autophagy, macroautophagy, often called "autophagy," starts. Afterward, during the elongation phase, autophagic vacuoles or autophagosomes with a distinctive double-membrane shape are formed when the phagosomal membrane engulfs the cytoplasm after surrounding and targeting it. Afterward, a cascade of lysosomal degradation enzymes breaks down the autolysosome's encapsulated contents when the autophagosome and lysosome fuse to produce an autolysosome. Once released into the cytoplasm, the biosynthetic route uses the digested contents[12]. M. tuberculosis can multiply and stay inside infected macrophages by blocking phagosome acidification and phagosome lysosome fusion. M. tuberculosis clearance has been linked to autophagy, and in vitro, investigations have shown that intracellular mycobacterium tuberculosis survival is decreased when autophagy is induced by fasting or rapamycin therapy. Conversely, infection-induced macrophages that consumed BECN1 and ATG7 could better prevent phagosomal lysosomal fusion[13].

Infectious diseases like tuberculosis continue to disproportionately affect poor regions in developing nations[14]. Thus, to decrease tuberculosis (TB) morbidity and mortality, early detection and treatment are crucial. Although there have been many improvements in TB diagnostics, no simple, accurate test can still be used at the point of treatment to confirm a diagnosis. Immunizing with the BCG vaccination may protect children against severe and disseminated forms of the illness. Infection rates can be decreased by 30%, and adults may also give some protection[15]. Consequently, there is an immediate need for an easy-to-use, sensitive, and precise diagnostic procedure. No biomarkers have yet been widely employed for tuberculosis diagnosis or treatment. This work presents a novel approach to TB diagnosis and prevention, employing bioinformatics technology to forecast possible indicators of tuberculosis.

Patients and datasets

We used the Gene Expression Omnibus[16] (GEO, https://ncbi.nlm.mih.gov/geo/), a public functional genomics database with various gene profiles in various diseases. We retrieved multiple GEO series, including mRNA (GSE144127[17] and GSE19435[18]), miRNA (GSE179816), and lncRNA (GSE163292[18, 19]) datasets. The GPL10558 platform was used to extract the data from GSE144127 microarray profiles included 101 healthy and 327 TB samples.The GSE19435 microarray profiles were generated using the GPL6947 platform and included 45 TB and 61 matched control samples. In addition, GSE144127 was used as an external dataset, while GSE19435 was used as training datasets (details in Table 1). GenomeStudio (Illumina) was used to perform quality control and generate signal intensity values. Then, we integrated the ATGs derived from HADb (http://www.autophagy.lu/index.html)[20], AUTOPHAGY DATABASE (http://www.tanpaku.org/autophagy/index.html), HAMdb (http://hamdb.scbdd.com/)[21], GSEA (https://www.gseamsigdb.org/gsea/index.jsp) and NCBI (https://www.ncbi.nlm.nih.gov/). A total of 2323 genes were extracted, among which 1906 autophagy-related genes were expressed in the GSE144127 dataset (See Supplemental Table S1).

Weighted Gene Co-Expression Network Analysis

We constructed gene co-expression networks using the R package "WGCNA"[22] and identified key functional modules. The correlation of ATGs between samples was evaluated based on module membership values>0.8 and gene significance values>0.5. A soft threshold of 9 was determined as suitable with an R²>0.9. Subsequently, a standard scale-free network was established. Pearson correlation analysis was conducted to examine the correlation of each gene module with TB disease and health characteristics. The correlation functional network was then utilized to identify tissue-specific markers. Key modules were identified by hierarchical clustering of genes using a dynamic tree-cutting strategy (shear height = 0.25). Gene information in the modules (see Supplementary Table S2).

Enrichment Analysis of Key Modules

We used significance ratings to sort the genes in the core modules from most to least important. Based on the results of the Gene Ontology (GO) annotation analysis and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis in the functionally related modules, functional enrichment analysis was then conducted to examine, identify, and interpret various biological functions (see Supplementary Table S3-S6). The R software's "Cluster Profiler" and "ggplot2" (p-value Cut off = 0.25) were used to view and analyze the data. The essential modules are more biologically significant and have better characterization than other modules.

Construction of protein–protein interaction network

The STRING database[23] (https://string-db.org/), a search tool for the retrieval of interacting genes/proteins, was utilized to construct the protein-protein interaction (PPI) network, encompassing both experimentally validated and computationally predicted interactions. The 508 ATGs in the key modules were imported into the STRING online platform to construct a PPI network with default Settings for PPI analysis. The analysis results were imported into Cytoscape (3.10.1) software. Using the Degree method, we used CytoHubba[24] to analyze the core module enrichment of the PPI network .

Identification of hub genes

The R tool "limma" was used to analyze the changes in gene expression of 508 ATGs in the core modules, whereas the R package "ggplot2" was employed to generate the volcano diagrams. As shown in Supplementary Table S7, we used the criteria | log2FC | > 0.5 and p< 0.05 to filter the gene expression differences. The top nodes of the PPI network were intersected with the differentially expressed genes to get the gene expression in TB. We intersected the top 20 nodes in the PPI network with DE-ATGs to discover the differentially expressed Hub genes in tuberculosis. Researchers used Receiver Operating Characteristic (ROC) curve analysis to determine which genes were most useful for diagnosis. The validation dataset (GSE19435) was subjected to ROC analysis using R's "pROC" package.

Immune cell infiltration analysis

To find immunological infiltration, the CIBERSORT[25] method was used. This deconvolution method can determine the proportion of 22 different types of immune cells using a huge number of tissue samples. Based on the gene expression data, the infiltration abundance of 22 immune cells per sample was computed using the 'CIBERSORT R script v1.04' R-based script, referenced in Supplementary Table S8. At the same time, we looked at how immune cell invasion was correlated with hub gene expression levels.

Construction of ceRNA network

We utilized the following procedures to construct the ceRNA network: (1) differential analysis of GSE163292 (lncRNA database) and GSE179816 (miRNA database) with the limma package in the R environment using | log2FC | > 0.5 and p < 0.05 as screening thresholds to obtain differentially expressed microRNAs (DE-miRNAs) and differentially expressed lncRNAs (DE-lncRNAs) (see Supplementary Table S9, S10); (2) using mirNet (https://www.mirnet.ca/)[26] to predict the lncRNAs that would be targeted by the DE-miRNAs that bind to the hub gene [the predicted lncRNAs were then taken to intersect with the DElncRNAs from (1) to construct the miRNA–lncRNA axis]; and (3) combining the multiple mRNA–miRNA and miRNA–lncRNA axes obtained from the above steps into a mRNA–miRNA– lncRNA regulatory axis according to the predicted bindsing trends (see Supplementary Table S11, S12). In addition, the ceRNA network was visualized using Cytoscape software.

Drug–Gene Interaction and Molecular Docking Analysis

The medications that are now available or might be beneficial were discovered by the Drug-Gene Interaction Database (DGIdb, www.dgidb.org)[27]. For the data visualizations made using Cytoscape, see Supplementary Table S13. The PubChem database[28] and the PDB database[29] were used to construct the molecular structures of the target proteins. Pymol[30], which can distinguish between water molecules and small-molecule ligands, was used to fine-tune the targets after the sequences had been modeled using the AlphaFold2[31] program. After inputting the protein sequences, we ran protein-protein docking using the HDOCK SERVER [32]application. Using the PyMOL tool, the docked complexes were viewed when the ideal docking arrangement was settled upon.

M. tuberculosis culture

M. tuberculosis-BCG was obtained from Shihezi University (Shihezi, China). BCG was grown in Roche medium. After 21 days, the rapidly increasing BCG vaccine was collected to make a homogeneous bacterial suspension, and the BCG concentration was determined using a McColl turbidimeter.

Cell culture and M. tuberculosis infection

Procell of Wuhan, China, supplied the macrophage RAW264.7 and mononuclear THP-1 cell lines. The 6-well plates were seeded with cells at a 2×10⁶ cells/well density in RPMI 1640 medium (Gibco, USA) or DMEM medium (Gibco, USA) with 10% FBS (Gibco, USA) before usage. Phorbol-12-myristate-13-acetate (PMA, Sigma-Aldrich, USA) was added to THP-1 cells during 24 hours of incubation at 37 ℃ with 5% CO₂ to stimulate adherent differentiation of macrophages. The optimal conditions for BCG infection were a MOI of 10 and 4 h of infection time for the cells. Bacteria outside of cells were rinsed with PBS (servicebio, China). Twenty-four hours after infection, THP-1/RAW264.7 cells were cultivated.

C57BL/6 mice infection models

We bought our male C57BL/6 mice from the Chinese company Henan Scebis Biotechnology Co., Ltd. From 230 to 250 grams, they were weighed. Injecting 5×10⁶ BCG into the tail veins of mice served as a TB infection model(n≥3), whereas PBS was used as a control group. A standard environment was maintained for the mice, which included food and water accessible at all times, a temperature of 23 ± 1°C, 60% humidity, and a 12:12 light/dark cycle with light from 08:30 to 20:30. After four weeks of intervention, the mice were sacrificed by severing their necks. Their lung tissues were preserved for use in future studies.

Real-time quantitative polymerasechain reaction

Complete RNA extraction was performed according to the manufacturer's instructions using the TransZol Up (TransGen Biotech, China). The next phase included utilizing a PrimeScript RT Reagent Kit to reverse-transcribe the RNA into cDNA with the help of TransGen Biotech of China. Amplifying cDNA using a QuantStudio real-time PCR machine developed by Thermo Fisher Scientific in the USA. The primer sequences for STAT1, CASP1, and IL1B are shown in Table 2. To find the relative expression of mRNA, the 2^−ΔŔCt method was used, using ACTB normalization.

Western Blotting and Immunohistochemistry

According to the manufacturer's recommendations, the RIPA lysis buffer (Sigma, USA) was used to extract total protein from cells and tissues. After that, the BCA assay (Beyotime, China) was used to quantify the total protein concentration. After electrophoretically transferring 30 ug of total proteins onto PVDF membranes, sealed with 5% BSA, the proteins were separated on a 10% polyacrylamide gel using SDS-PAGE. The membranes were incubated at 4 ℃ overnight with diluted primary antibodies to STAT1, IL1B, and CASP1. The dilutions were 1:1000 for CST, 1:1000 for Affinity, and 1:1000 for CASP1. After that, the membrane was left to incubate at room temperature for another four hours with secondary antibodies that were HRP-labelled (Jackson, United States). At last, the protein bands were quantified and examined with the help of ImageJ.

Statistical analysis

GraphPad Prism 8.0 and R 4.2.2 were used for statistical analysis. We found significant differences between groups using an unpaired t-test with a p-value below 0.05. Hub ATG diagnostic efficacy was assessed using ROC curve analysis.

Selection of Key Modules

In our study, nine modules (indicated by the colors of Brown, Pink, Turquorise, Red, Yellow, Green, Black, Blue, and Grey) were identified according to the average linkage hierarchical clustering (Figure 1A, see Supplemental Table S2 for the number of genes in each module). The heat map displays the results of the correlation analysis between modules and TB (Figure 1B). The genes in the Pink (0.47, p<0.001), Turquorise (0.19, p<0.001) and Black (0.12, p<0.001) module showed a positive correlation with the clinical traits of the tuberculosis group. However, the Red (-0.27, p<0.001), Yellow (-0.26, p<0.001) and Blue (-0.099, p<0.001) modules were negatively correlated with the clinical features of the tuberculosis group and positively related to those of the control group . At last, four key modules (Pink, Turquoise, Red and Yellow modules) were identified.

GO and KEGG Enrichment Analysis

To analyze the potential biological functions of these differentially expressed autophagy-related genes, we conducted GO and KEGG enrichment analysis by using R software. We visualized the top ten pathways in four key modules resulting from the GO and KEGG pathway analysis. The analysis resulted in 66 ATGs in the Pink model, as shown in Figures 3A, B. They mainly focused on the signaling pathways associated with NOD−like receptor signaling pathway, Necroptosis, Cytokine−cytokine receptor interaction, Phagosome and TNF signaling pathway. In addition, the analysis resulted in 219 ATGs in the Turquoise module, as shown in Figures 3C,D, which focused on Autophagy-animal and Tuberculosis. In the Red model, 87 ATGs resulted from the analysis, as shown in Figures 3E, F, which were mainly related to the Cellular senescence. Finally, in the Yellow module, the analysis resulted in 136 ATGs, as shown in Figures 3G,H, which weremainly enriched in Calcium signaling pathway, NOD−like receptor signaling pathway, HIF−1 signaling pathway, and AMPK signaling pathway. Detailed results of variance analysis are shown in Supplemental Table S3-S6.

Construction of Protein-Protein Interaction Network and identification of hub genes

To determine the interactions among differentially expressed autophagy-related genes, we explored the interactions among the 508 ATGs of key modules using the STRING database with medium confidence and obtained a PPI network containing 488 nodes and 6817 edges (Figure 4A). Using CytoHubba, the internal analysis plug-in of Cytoscape, using Degree method, the final output of the top 20 nodes, the top 20 nodes contain a total of 172 edges (Figure 4B). The gene expression differences of 508 ATGs in key modules were analyzed with R package "limma", and the volcano map was drawn with R package "ggplot2", and differential genes were screened with log2FC > 0.5 or <-0.5 and rate-adjusted values (FDR) < 0.05. A total of 23 differential genes (DEGs) were obtained, of which 22 were up-regulated and 1 was down-regulated (Figure 4C, Table 3). Detailed results of variance analysis are shown in Supplemental Table S7. The top 20 node in PPI network analysis was interleaved with 22 DEGs, and a total of three key genes were obtained, including STAT1, CASP1 and IL1B (Figure 4D).

The changes in immune cell infiltration in Tuberculosis

To better comprehend the immune response to MTB-specific antigens in TB individuals, this study analyzed the immune cells that regulate the immune response to MTB-specific antigens through autophagy via immune infiltration analysis. We evaluated the relative abundance of infiltrative 22 immune cell subtypes in normal and TB samples using the CIBERSORT algorithm. Bar graphs and heat maps show the infiltration of multiple immune-cell subtypes in each sample(Figure 5A). Violin plots show the difference in the percentage of immune cells between the two groups of samples (Figure 5B). Compared with the normal sample, the TB sample of infiltration increases And Tells. CD4. Naive, NK cells, activated, and dendritic cells. The activated infiltration reduction. Finally, the heat map showed the correlation between hub genes and immune cell infiltration (Figure 5C). The results showed that hub gene expression correlated with immune cell infiltration in each subtype to varying degrees. In TB samples, IL1B was negatively correlated with CD8 T cell, plasma cells, NK cells resting, and CD4 T cell memory activated, but it was positively correlated with neutrophils.

Potential ceRNA network of hub genes

To reveal the potential post-transcriptional regulatory mechanisms of these 3 hub genes, we screened differentially expressed miRNAs and lncRNAs during TB development and constructed a ceRNA network. DE-miRNAs and DE-lncRNAs were identified and screened in R software using the limma package, with |log2FC| > 0.5 and p < 0.05 set as significance thresholds. The volcano plots of DE-miRNAs and DE-lncRNAs are shown in Figure 6A, B. Subsequently, using miRNet to predict DE-miRNA-targeted mRNAs and DE-miRNA-targeted lncRNAs, a total of 1474 target mRNAs and 119 lncRNAs targeting these 9 DE-miRNAs were identified. The predicted 1474 DE-miRNA-targeted mRNAs were interspersed with the 23 differentially expressed genes, while 119 lncRNAs targeting DE-miRNAs were intersected with 2157 DE-lncRNAs. Finally, mRNA-miRNA and miRNA-lncRNA binding pairs were integrated to create a competitive ceRNA regulatory network of 2 mRNAs, 2 miRNAs and 10 lncRNAs (Figures 6C).

Drug–Gene Interaction and Molecular Docking Analysis

We used the DGIdb database to find 84 targeted medicines to create a drug-gene interaction network (Figure 7A). Based on an interaction value ≥3, eight candidate medicines interacting with STAT1, CASP1, or IL1B were chosen (Table 4). CANAKINUMAB, an IL1B target, is clinically authorized. After that, we predicted 11 IL1B/CANAKINUMAB molecular binding sites. All binding sites were hydrogen bonds (Figure 7B).

Identification and Validation of 3 Hub ATGs in External Datasets

To validate the reliability of STAT1, CASP1, and IL1B, we further confirmed their high expression in blood specimens from TB patients in an external dataset GSE19435 (Figure 8A-C). Their area under the curve (AUC) values were found to be 0.881 (STAT1), 0.829 (IL1B), and 0.960 (CASP1) by receiver operating characteristic (ROC) curves, all of which demonstrated high diagnostic efficacy (Figure 8D-F).

Validation of hub genes

The results of RT-qPCR demonstrated that the mRNA levels of STAT1, CASP1, and IL1B were elevated in BCG-infected human-derived macrophage THP-1, whereas in murine-derived macrophage RAW264.7, the expression of STAT1 and IL1B was elevated, but not that of CASP1. This discrepancy may be attributed to the cell source (Figure 8G-L). Immunohistochemical staining (IHC) demonstrated a notable upregulation of STAT1, CASP1, and IL1B expression in the lung tissues of TB-infected mice following BCG infection (Figure 9A). Western blot results were in alignment with RT-qPCR results (Figure 9B). The in vivo and in vitro test results typically corroborated the bioinformatics findings.

Humans and the mycobacterium TB bacterium lived side by side for an extremely long period. Mycobacterium tuberculosis is a serious danger to world health; an estimated 1.7 billion individuals are infected[33]. To prevent tuberculosis (TB) from starting, progressing, and spreading early diagnosis and tailored treatment are essential. Currently, no clinical trial data nor dependable biomarkers exist to assist in diagnosing or treating latent tuberculosis. In recent years, host-directed therapy (HDT) has arisen as a novel adjuvant treatment for tuberculosis (TB). The goals of HDT are to decrease lung damage, avoid drug resistance, and shorten treatment time by increasing autophagy and antimicrobial peptide synthesis, among other macrophage effector mechanisms[34].New research suggests autophagy activators that destroy bacteria might be a great way to treat tuberculosis[35].

Autophagy is a cellular process regulating cell growth, survival, development, and death. It is strongly associated with inflammatory and immune reactions[36]. Host cells can activate autophagy to control the escape of M. tuberculosis and reduce the bacterial load in infected cells. Furthermore, the activation of autophagy also contributes to the control of inflammation and is regulated by various proinflammatory cytokines[37]. In this study, differentially expressed genes (DEGs) from the GEO database were analyzed, autophagy-related genes (ATGs) were extracted from five autophagy databases, and 1906 ATGs were identified. To construct the protein-protein interaction (PPI) network, 508 ATGs were selected from the key modules using weighted gene co-expression network analysis (WGCNA). Three hub genes, STAT1, IL1B, and CASP1, were identified by intersecting the top 20 nodes of the PPI network with the 22 DE-ATGs identified by differential analysis. A bioinformatics study, double validation of the BCG-infected macrophage and animal infection models, and initial raw data all pointed to a significant shift in the expression of the three hub genes compared to control and TB samples. Using the validation dataset for these three genes, a diagnostic model was built. Potentially useful for tuberculosis (TB) infection diagnosis, the AUC values show a high degree of prediction accuracy.

Biomarkers are essential for TB disease surveillance and treatment decision-making as they provide valuable information on infection status, disease severity, treatment response, and patient prognosis. STAT1 (Signal Transducer And Activator Of Transcription 1), a cytokine-responsive signal transducer and activator of transcription, is considered a crucial mediator in response to IFN-γ and host defence against Mycobacterium tuberculosis[38–40]. A recent study has demonstrated that CLEC4E and TLR4 agonists, in combination with anti-TB drugs, activate STAT1, increasing autophagy[41].STAT1 may be a target for autophagy-assisted therapy, offering novel TB treatment options[42, 43]. Thus, studying STAT1 and autophagy's reciprocal regulation in TB is promising.

STAT1 and autophagy's functions in TB require additional study[38–40].NLRP3's inflammatory vesicle, which plays a role in various inflammatory processes, contains CASP1, which hydrolyzes IL-1B[44–46]. Pro-inflammatory cytokine interleukin 1B is crucial to humans' innate and adaptive immune systems. Autophagy enhances IL1B transcription, processing, and release, a negative feedback mechanism that modulates inflammatory responses in vitro and in vivo. New anti-infective medicines may target autophagy genes[45]. Human tuberculosis causes infected organs to develop granulomatous inflammation. As TB develops, inflammatory cells infiltrate the granulomas' core structures differently[47, 48]. The composition and number of immune cells associated with TB status have been the focus of recent research that tries to discover and describe distinct granuloma[49–51]. The study examined transcriptome data to determine immune infiltration and found that the HC and TB groups had different proportions of 22 kinds of invading immune cells. Plasma cells and M0-type macrophages were found to be significantly different.

Because of their role in regulating gene networks, non-coding RNAs (ncRNAs, miRNAs, etc.) have emerged as promising therapeutic targets in the fight against tuberculosis (TB)[52].In most cases, long non-coding RNAs (lncRNAs) act as endogenous RNA, competing with miRNAs and mRNAs at the RNA level. MicroRNAs (miRNAs) are non-coding RNAs that control how the host cell fights against pathogens. They control inflammation, autophagy, apoptosis, and macrophage polarization via host signaling pathways that are differently regulated in response to mycobacterium tuberculosis infection[53–56].The potential of microRNAs and long non-coding RNAs (lncRNAs) as biomarkers for TB diagnosis and treatment targets is well acknowledged[52]. Consequently, we built a ceRNA regulation network using the chosen DE-ATGs, DE-lncRNAs, and DE-miRNAs. The network consists of ten lncRNAs, two microRNAs, and two messenger RNAs. We also considered using them as disease biomarkers. After villose and the second IL-1 blocker to be licensed for CAPS, CANAKINUMAB (ACZ885 Ilaris), a human IgGκ monoclonal antibody that targets IL1B, is the third IL-1 blocker to get approval for the treatment of autoimmune illnesses[57].There are additional IL-1-activated pathways that CANAKINUMAB leaves alone because of its strong specificity for IL1B. Without numerous injections or large dosages, CANAKINUMAB may block IL1B activity for lengthy durations due to its prolonged half-life[58]. So far, CANAKINUMAB has shown promise in treating symptoms and abnormalities linked to several rare autoinflammatory disorders, such as familial Mediterranean fever, cold pyrimidine-associated periodic syndrome (CAPS), and autoinflammatory relapsing fever syndrome (ARFS)[59, 60].

The pooled analysis of three randomized, double-blind studies showed that CANAKINUMAB did not impact an increased rate of Mtb reactivation[61]. These analyses insist that CANAKINUMAB could be a targeted drug for immunotherapy of tuberculosis. Nevertheless, there is a need for studies and evidence to confirm its efficacy and safety in TB patients. First, the expression of genes studied was examined only in macrophages and mice. It was not verified in a clinical sample of patients. Further research and verification in a clinical setting will allow one to understand reliable molecular biomarkers, which can be used in clinical practice.

A bioinformatics technique examined differentially expressed genes (DEGs) in TB. Three hub genes—IL1B, STAT1, and CASP1—were identified among genes associated with autophagy as components of this disease. In addition, we built a ceRNA network to study the relationship between immune cell infiltration and hub genes, and we analyzed immune cell infiltration. Lastly, we projected the therapeutic targets of hub genes and verified their expression via in vitro and in vivo tests. Based on our findings, IL1B, STAT1, and CASP1 might play an important role as TB progression indicators. This aids in identifying possible therapeutic targets and sheds light on the pathophysiology of tuberculosis (TB).

Mtb Mycobacterium tuberculosis

TB tuberculosis

WGCNA weighted gene coexpression network analysis

ATGs autophagy-related genes

DE Differently expressed

PPI protein-protein interaction

ROC receiver operating characteristic

ceRNA network competitive endogenous RNA network

GEO Gene Expression Omnibus

GO Gene ontology

KEGG Kyoto encyclopedia of genes and genomes

qRT-PCR Quantitative real-time polymerase chain reaction

Ethics statement: The animal study was reviewed and approved by the institutional review board.

Author contributions: SSJ contributed equally to this work. LTT and SSJ designed the search; LTT wrote the main manuscript text. Reem, DSQ and CCZ gave me comprehensive writing advice. LTT, SSJ, SYF and ZX performed statistical analysis; WJD and ZWJ primary responsibility for final content. All authors read and approved the final manuscript.

Funding:This study was funded by the National Natural Science Foundation of China (82360321 and 82060021), 2023 Corps Key Areas of Science and Technology Tackling Program Projects (2023AB053）, and 2022 Corps Guided Science and Technology Program Projects (2022ZD043).

Acknowledgment: The authors would like to thank all members of our laboratory for their work. Saeed EI-Ashram and Reem Atalla Alajm, supported by Project number (RSP2024R99), King Saud University, Riyadh, Saudi Arabia, for their support for this project.

Conflict of interest:The authors declare that the research was conducted without any commercial or financial relationships that could potentially create a conflict of interest.

Ethics approval and consent to participate：The present study was approved by Animal Experimental Ethical Inspection of FirstAffiliated Hospital of Shihezi University.All data used in this study are from publicly available databases, and no additional ethical approval and informed consent are required.

Competing interests: The authors declare no competing interests.

WHO: Global tuberculosis report. In.; 2022.
Migliori GB, Nardell E, Yedilbayev A, D'Ambrosio L, Centis R, Tadolini M, van den Boom M, Ehsani S, Sotgiu G, Dara M: Reducing tuberculosis transmission: a consensus document from the World Health Organization Regional Office for Europe. EUR RESPIR J 2019, 53(6).
WHO consolidated guidelines on tuberculosis: Module 3: diagnosis – rapid diagnostics for tuberculosis detection. Geneva: World Health Organization; 2021.
Gordon S, Taylor PR: Monocyte and macrophage heterogeneity. NAT REV IMMUNOL 2005, 5(12):953-964.
Weiss G, Schaible UE: Macrophage defense mechanisms against intracellular bacteria. IMMUNOL REV 2015, 264(1):182-203.
Shukla S, Richardson ET, Drage MG, Boom WH, Harding CV: Mycobacterium tuberculosis Lipoprotein and Lipoglycan Binding to Toll-Like Receptor 2 Correlates with Agonist Activity and Functional Outcomes. INFECT IMMUN 2018, 86(10).
Drage MG, Pecora ND, Hise AG, Febbraio M, Silverstein RL, Golenbock DT, Boom WH, Harding CV: TLR2 and its co-receptors determine responses of macrophages and dendritic cells to lipoproteins of Mycobacterium tuberculosis. CELL IMMUNOL 2009, 258(1):29-37.
Windheim M, Lang C, Peggie M, Plater LA, Cohen P: Molecular mechanisms involved in the regulation of cytokine production by muramyl dipeptide. BIOCHEM J 2007, 404(2):179-190.
Chen X, Xiao Z, Xie X, Liu X, Jiang M, Yuan C, Yang L, Hu J: TNF-α-Induced NOD2 and RIP2 Contribute to the Up-Regulation of Cytokines Induced by MDP in Monocytic THP-1 Cells. J CELL BIOCHEM 2018, 119(7):5072-5081.
Ochoa AE, Congel JH, Corley JM, Janssen WJ, Nick JA, Malcolm KC, Hisert KB: Dectin-1-Independent Macrophage Phagocytosis of Mycobacterium abscessus. INT J MOL SCI 2023, 24(13).
Mizushima N, Komatsu M: Autophagy: renovation of cells and tissues. CELL 2011, 147(4):728-741.
Racanelli AC, Kikkers SA, Choi A, Cloonan SM: Autophagy and inflammation in chronic respiratory disease. AUTOPHAGY 2018, 14(2):221-232.
Singh SB, Davis AS, Taylor GA, Deretic V: Human IRGM induces autophagy to eliminate intracellular mycobacteria. SCIENCE 2006, 313(5792):1438-1441.
Sharma D, Sharma J, Deo N, Bisht D: Prevalence and risk factors of tuberculosis in developing countries through health care workers. MICROB PATHOGENESIS 2018, 124:279-283.
Kruk ME, Gage AD, Arsenault C, Jordan K, Leslie HH, Roder-DeWan S, Adeyi O, Barker P, Daelmans B, Doubova SV et al: High-quality health systems in the Sustainable Development Goals era: time for a revolution. LANCET GLOB HEALTH 2018, 6(11):e1196-e1252.
Clough E, Barrett T: The Gene Expression Omnibus Database. Methods Mol Biol 2016, 1418:93-110.
Hoang LT, Jain P, Pillay TD, Tolosa-Wright M, Niazi U, Takwoingi Y, Halliday A, Berrocal-Almanza LC, Deeks JJ, Beverley P et al: Transcriptomic signatures for diagnosing tuberculosis in clinical practice: a prospective, multicentre cohort study. LANCET INFECT DIS 2021, 21(3):366-375.
Berry MP, Graham CM, McNab FW, Xu Z, Bloch SA, Oni T, Wilkinson KA, Banchereau R, Skinner J, Wilkinson RJ et al: An interferon-inducible neutrophil-driven blood transcriptional signature in human tuberculosis. NATURE 2010, 466(7309):973-977.
Li ZB, Han YS, Wei LL, Shi LY, Yi WJ, Chen J, Huang H, Jiang TT, Li JC: Screening and identification of plasma lncRNAs uc.48+ and NR_105053 as potential novel biomarkers for cured pulmonary tuberculosis. INT J INFECT DIS 2020, 92:141-150.
Chen Y, Zeng A, He S, He S, Li C, Mei W, Lu Q: Autophagy-Related Genes in Atherosclerosis. J HEALTHC ENG 2021, 2021:6402206.
Wang NN, Dong J, Zhang L, Ouyang D, Cheng Y, Chen AF, Lu AP, Cao DS: HAMdb: a database of human autophagy modulators with specific pathway and disease information. J CHEMINFORMATICS 2018, 10(1):34.
Langfelder P, Horvath S: WGCNA: an R package for weighted correlation network analysis. BMC BIOINFORMATICS 2008, 9:559.
Szklarczyk D, Gable AL, Nastou KC, Lyon D, Kirsch R, Pyysalo S, Doncheva NT, Legeay M, Fang T, Bork P et al: The STRING database in 2021: customizable protein-protein networks, and functional characterization of user-uploaded gene/measurement sets. NUCLEIC ACIDS RES 2021, 49(D1):D605-D612.
Chin CH, Chen SH, Wu HH, Ho CW, Ko MT, Lin CY: cytoHubba: identifying hub objects and sub-networks from complex interactome. BMC Syst Biol 2014, 8 Suppl 4(Suppl 4):S11.
Chen B, Khodadoust MS, Liu CL, Newman AM, Alizadeh AA: Profiling Tumor Infiltrating Immune Cells with CIBERSORT. Methods Mol Biol 2018, 1711:243-259.
Chang L, Zhou G, Soufan O, Xia J: miRNet 2.0: network-based visual analytics for miRNA functional analysis and systems biology. NUCLEIC ACIDS RES 2020, 48(W1):W244-W251.
Wagner AH, Coffman AC, Ainscough BJ, Spies NC, Skidmore ZL, Campbell KM, Krysiak K, Pan D, McMichael JF, Eldred JM et al: DGIdb 2.0: mining clinically relevant drug-gene interactions. NUCLEIC ACIDS RES 2016, 44(D1):D1036-D1044.
Wang Y, Xiao J, Suzek TO, Zhang J, Wang J, Zhou Z, Han L, Karapetyan K, Dracheva S, Shoemaker BA et al: PubChem's BioAssay Database. NUCLEIC ACIDS RES 2012, 40(Database issue):D400-D412.
Karuppasamy MP, Venkateswaran S, Subbiah P: PDB-2-PBv3.0: An updated protein block database. J BIOINF COMPUT BIOL 2020, 18(2):2050009.
Seeliger D, de Groot BL: Ligand docking and binding site analysis with PyMOL and Autodock/Vina. J COMPUT AID MOL DES 2010, 24(5):417-422.
Bryant P, Pozzati G, Elofsson A: Improved prediction of protein-protein interactions using AlphaFold2. NAT COMMUN 2022, 13(1):1265.
Yan Y, Tao H, He J, Huang SY: The HDOCK server for integrated protein-protein docking. NAT PROTOC 2020, 15(5):1829-1852.
Gina P, Davids M, Dheda K: Manipulation of autophagy for host-directed tuberculosis therapy. Afr J Thorac Crit Care Med 2019, 25(2).
Kolloli A, Subbian S: Host-Directed Therapeutic Strategies for Tuberculosis. FRONT MED-LAUSANNE 2017, 4:171.
Kimmey JM, Stallings CL: Bacterial Pathogens versus Autophagy: Implications for Therapeutic Interventions. TRENDS MOL MED 2016, 22(12):1060-1076.
Deretic V: Autophagy in tuberculosis. CSH PERSPECT MED 2014, 4(11):a18481.
Mizushima N, Noda T, Yoshimori T, Tanaka Y, Ishii T, George MD, Klionsky DJ, Ohsumi M, Ohsumi Y: A protein conjugation system essential for autophagy. NATURE 1998, 395(6700):395-398.
Meraz MA, White JM, Sheehan KC, Bach EA, Rodig SJ, Dighe AS, Kaplan DH, Riley JK, Greenlund AC, Campbell D et al: Targeted disruption of the Stat1 gene in mice reveals unexpected physiologic specificity in the JAK-STAT signaling pathway. CELL 1996, 84(3):431-442.
Durbin JE, Hackenmiller R, Simon MC, Levy DE: Targeted disruption of the mouse Stat1 gene results in compromised innate immunity to viral disease. CELL 1996, 84(3):443-450.
Esquivel-Solís H, Quiñones-Falconi F, Zarain-Herzberg A, Amieva-Fernández RI, López-Vidal Y: Impaired activation of Stat1 and c-Jun as a possible defect in macrophages of patients with active tuberculosis. CLIN EXP IMMUNOL 2009, 158(1):45-54.
Popov SV, Mukhomedzyanov AV, Voronkov NS, Derkachev IA, Boshchenko AA, Fu F, Sufianova GZ, Khlestkina MS, Maslov LN: Regulation of autophagy of the heart in ischemia and reperfusion. APOPTOSIS 2023, 28(1-2):55-80.
Pahari S, Negi S, Aqdas M, Arnett E, Schlesinger LS, Agrewala JN: Induction of autophagy through CLEC4E in combination with TLR4: an innovative strategy to restrict the survival of Mycobacterium tuberculosis. AUTOPHAGY 2020, 16(6):1021-1043.
Goldberg AA, Nkengfac B, Sanchez A, Moroz N, Qureshi ST, Koromilas AE, Wang S, Burelle Y, Hussain SN, Kristof AS: Regulation of ULK1 Expression and Autophagy by STAT1. J BIOL CHEM 2017, 292(5):1899-1909.
Dupont N, Jiang S, Pilli M, Ornatowski W, Bhattacharya D, Deretic V: Autophagy-based unconventional secretory pathway for extracellular delivery of IL-1β. EMBO J 2011, 30(23):4701-4711.
Harris J: Autophagy and IL-1 Family Cytokines. FRONT IMMUNOL 2013, 4:83.
Harris J, Hartman M, Roche C, Zeng SG, O'Shea A, Sharp FA, Lambe EM, Creagh EM, Golenbock DT, Tschopp J et al: Autophagy controls IL-1beta secretion by targeting pro-IL-1beta for degradation. J BIOL CHEM 2011, 286(11):9587-9597.
Guler R, Ozturk M, Sabeel S, Motaung B, Parihar SP, Thienemann F, Brombacher F: Targeting Molecular Inflammatory Pathways in Granuloma as Host-Directed Therapies for Tuberculosis. FRONT IMMUNOL 2021, 12:733853.
Ulrichs T, Kosmiadi GA, Trusov V, Jörg S, Pradl L, Titukhina M, Mishenko V, Gushina N, Kaufmann SH: Human tuberculous granulomas induce peripheral lymphoid follicle-like structures to orchestrate local host defence in the lung. J PATHOL 2004, 204(2):217-228.
Ramakrishnan L: Revisiting the role of the granuloma in tuberculosis. NAT REV IMMUNOL 2012, 12(5):352-366.
Liu H, Zhou Z, Deng H, Tian Z, Wu Z, Liu X, Ren Z, Jiang Z: Trim65 attenuates isoproterenol-induced cardiac hypertrophy by promoting autophagy and ameliorating mitochondrial dysfunction via the Jak1/Stat1 signaling pathway. EUR J PHARMACOL 2023, 949:175735.
Qin C, Liu Q, Hu ZW, Zhou LQ, Shang K, Bosco DB, Wu LJ, Tian DS, Wang W: Microglial TLR4-dependent autophagy induces ischemic white matter damage via STAT1/6 pathway. THERANOSTICS 2018, 8(19):5434-5451.
Almatroudi A: Non-Coding RNAs in Tuberculosis Epidemiology: Platforms and Approaches for Investigating the Genome's Dark Matter. INT J MOL SCI 2022, 23(8).
Behar SM, Martin CJ, Booty MG, Nishimura T, Zhao X, Gan HX, Divangahi M, Remold HG: Apoptosis is an innate defense function of macrophages against Mycobacterium tuberculosis. MUCOSAL IMMUNOL 2011, 4(3):279-287.
Kim YS, Silwal P, Kim SY, Yoshimori T, Jo EK: Autophagy-activating strategies to promote innate defense against mycobacteria. EXP MOL MED 2019, 51(12):1-10.
Ke Z, Lu J, Zhu J, Yang Z, Jin Z, Yuan L: Down-regulation of lincRNA-EPS regulates apoptosis and autophagy in BCG-infected RAW264.7 macrophages via JNK/MAPK signaling pathway. INFECT GENET EVOL 2020, 77:104077.
Li M, Cui J, Niu W, Huang J, Feng T, Sun B, Yao H: Long non-coding PCED1B-AS1 regulates macrophage apoptosis and autophagy by sponging miR-155 in active tuberculosis. BIOCHEM BIOPH RES CO 2019, 509(3):803-809.
Dhimolea E: Canakinumab. MABS-AUSTIN 2010, 2(1):3-13.
Alten R, Gram H, Joosten LA, van den Berg WB, Sieper J, Wassenberg S, Burmester G, van Riel P, Diaz-Lorente M, Bruin GJ et al: The human anti-IL-1 beta monoclonal antibody ACZ885 is effective in joint inflammation models in mice and in a proof-of-concept study in patients with rheumatoid arthritis. ARTHRITIS RES THER 2008, 10(3):R67.
De Benedetti F, Gattorno M, Anton J, Ben-Chetrit E, Frenkel J, Hoffman HM, Koné-Paut I, Lachmann HJ, Ozen S, Simon A et al: Canakinumab for the Treatment of Autoinflammatory Recurrent Fever Syndromes. NEW ENGL J MED 2018, 378(20):1908-1919.
Cantarini L, Lopalco G, Caso F, Costa L, Iannone F, Lapadula G, Anelli MG, Franceschini R, Menicacci C, Galeazzi M et al: Effectiveness and tuberculosis-related safety profile of interleukin-1 blocking agents in the management of Behçet's disease. AUTOIMMUN REV 2015, 14(1):1-9.
Houben RM, Dodd PJ: The Global Burden of Latent Tuberculosis Infection: A Re-estimation Using Mathematical Modelling. PLOS MED 2016, 13(10):e1002152.

Table 1: Tuberculosis Data Sampling

	Sample	Control	Tuberculosis	All
mRNA	GSE144127	327	101	428
mRNA	GSE19435	12	21	33
miRNA	GSE179816	4	4	8
lncRNA	GSE163292	4	12	16

Table 2: Primer Sequences for STAT1, CASP1, and IL1B

Gene	Forward	Reverse
IL1B-homo	AGCTACGAATCTCCGACCAC	CGTTATCCCATGTGTCGAAGAA
IL1B-mouse	GAAATGCCACCTTTTGACAGTG	TGGATGCTCTCATCAGGACAG
CASP1-homo	TTTCCGCAAGGTTCGATTTTCA	GGCATCTGCGCTCTACCATC
CASP1-mouse	AATACAACCACTCGTACACGTC	AGCTCCAACCCTCGGAGAAA
STAT1-homo	CAGCTTGACTCAAAATTCCTGGA	TGAAGATTACGCTTGCTTTTCC
STAT1-mouse	TCACAGTGGTTCGAGCTTCAG	GCAAACGAGACATCATAGGCA

Table 3: The DE-ATGs of TB.

Regulation

DE-ATGs

Upregulated (n = 13)

AIM2、IRF1、STAT1、PSME2、LAP3、MYOF、CASP1、CD274、TNFSF10、TRIM22、IFI30、TNFSF13B、PLSCR1、IL1B、ALDH1A1、XAF1、IFI6、MT2A、NOD2、TLR5、ISG15、IFI27

Downregulated (n = 4)

GPR183

Table4: Drug-Interacting Hub Genes (interaction score≥3）.

Gene	Drug	regulatory approval	interaction score
IL1B	CANAKINUMAB	Approved	5.782358225
IL1B	GEVOKIZUMAB	Not Approved	3.469414935
CASP1	BELNACASAN	Not Approved	4.536927223
CASP1	CHEMBL:CHEMBL415893	Not Approved	4.536927223
CASP1	VERMISTATIN	Not Approved	4.536927223
STAT1	GUTTIFERONE K	Not Approved	9.830008983
STAT1	PICOPLATIN	Not Approved	9.830008983
STAT1	GARCINOL	Not Approved	4.915004491

No competing interests reported.

Download PDF

Editor assigned by journal
08 Jul, 2024
Submission checks completed at journal
08 Jul, 2024
First submitted to journal
01 Jul, 2024

You are reading this latest preprint version

Identification and Analysis of Autophagy-Related Genes as Diagnostic Markers and Potential Therapeutic Targets for Tuberculosis Through Bioinformatics

Status:

Version 1

Abstract

Figures

Background

Materials and methods

Patients and datasets

Weighted Gene Co-Expression Network Analysis

Enrichment Analysis of Key Modules

Construction of protein–protein interaction network

Identification of hub genes

Immune cell infiltration analysis

Construction of ceRNA network

Drug–Gene Interaction and Molecular Docking Analysis

M. tuberculosis culture

Cell culture and M. tuberculosis infection

C57BL/6 mice infection models

Real-time quantitative polymerasechain reaction

Western Blotting and Immunohistochemistry

Statistical analysis

Results

Selection of Key Modules

GO and KEGG Enrichment Analysis

Construction of Protein-Protein Interaction Network and identification of hub genes

The changes in immune cell infiltration in Tuberculosis

Potential ceRNA network of hub genes

Drug–Gene Interaction and Molecular Docking Analysis

Identification and Validation of 3 Hub ATGs in External Datasets

Validation of hub genes

Discussion

Abbreviations

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1