Abstract

   Tuberculosis (TB) is the leading cause of death from a single
   infectious agent. The estimated total global TB deaths in 2019 were 1.4
   million. The decline in TB incidence rate is very slow, while the
   burden of noncommunicable diseases (NCDs) is exponentially increasing
   in low- and middle-income countries, where the prevention and treatment
   of TB disease remains a great burden, and there is enough empirical
   evidence (scientific evidence) to justify a greater research emphasis
   on the syndemic interaction between TB and NCDs. The current study was
   proposed to build a disease-gene network based on overlapping TB with
   NCDs (overlapping means genes involved in TB and other/s NCDs), such as
   Parkinson’s disease, cardiovascular disease, diabetes mellitus,
   rheumatoid arthritis, and lung cancer. We compared the TB-associated
   genes with genes of its overlapping NCDs to determine the gene-disease
   relationship. Next, we constructed the gene interaction network of
   disease-genes by integrating curated and experimentally validated
   interactions in humans and find the 13 highly clustered modules in the
   network, which contains a total of 86 hub genes that are commonly
   associated with TB and its overlapping NCDs, which are largely involved
   in the Inflammatory response, cellular response to cytokine stimulus,
   response to cytokine, cytokine-mediated signaling pathway, defense
   response, response to stress and immune system process. Moreover, the
   identified hub genes and their respective drugs were exploited to build
   a bipartite network that assists in deciphering the drug-target
   interaction, highlighting the influential roles of these drugs on
   apparently unrelated targets and pathways. Targeting these hub proteins
   by using drugs combination or drug repurposing approaches will improve
   the clinical conditions in comorbidity, enhance the potency of a few
   drugs, and give a synergistic effect with better outcomes. Thus,
   understanding the Mycobacterium tuberculosis (Mtb) infection and
   associated NCDs is a high priority to contain its short and long-term
   effects on human health. Our network-based analysis opens a new horizon
   for more personalized treatment, drug-repurposing opportunities,
   investigates new targets, multidrug treatment, and can uncover several
   side effects of unrelated drugs for TB and its overlapping NCDs.

   Keywords: Network Biology, Network Medicines, Disease-disease
   relationship, Disease-target interaction, MTB and NCDs

Introduction

   Tuberculosis (TB), a communicable disease caused by bacillus
   Mycobacterium tuberculosis, is the leading cause of death from a single
   infectious agent. Globally, an estimated 10.0 million people developed
   tuberculosis in 2020 ([42]WHO Global Tuberculosis Report-2021). Among
   these cases, 56% of individuals were men aged ≥15 years, 32% were
   women, and 12% were children aged <15 years. Most affected people were
   from the region of South-East Asia (44%), Africa (25%), and the Western
   Pacific (18%). A total of 1.5 million people died from TB in 2020
   (including 214,000 people with HIV). Worldwide, TB is the 13th leading
   cause of death and the second leading infectious killer after COVID-19
   (above HIV/AIDS).

   TB is still considered a deadly disease, particularly in high TB burden
   countries like India, China, Indonesia, Philippines, Pakistan, Nigeria,
   Bangladesh, and South Africa ([43]Bhatia et al., 2020). WHO reports
   reflect that the TB incidence rate decline is very slow, while the
   burden of noncommunicable diseases (NCDs) is exponentially increasing
   worldwide ([44]WHO Global Tuberculosis Report, 2020; [45]WHO
   Noncommunicable Diseases Progress Monitor, 2020).

   In the long term, tuberculosis may lead to collapse in immune
   surveillance, enhancing one’s susceptibility to non-communicable
   diseases (NCDs), which together contribute to two-thirds of the
   worldwide mortality ([46]Marais et al., 2013; [47]Peltzer, 2018).
   Emerging empirical evidence justifies the convergence of TB with NCDs
   such as Parkinson’s disease (PD) ([48]Shen et al., 2016),
   cardiovascular diseases (CVD) ([49]Huaman et al., 2015), diabetes
   mellitus (DM) ([50]Menon et al., 2016), rheumatoid arthritis (RA)
   ([51]Carmona et al., 2003), and lung cancer (LC) ([52]Chai and Shi,
   2020).

   Many Infectious diseases have been reported to contribute to the
   development of PD ([53]Harris et al., 2012; [54]Vlajinac et al., 2013;
   [55]Tan et al., 2015). Patients with TB have been reported to have a
   1.38-fold higher risk of developing PD as compared to control subjects
   ([56]Shen et al., 2016). The related mechanisms are not known; however,
   it is thought that pro-inflammatory responses generated in TB may be a
   key driving process associated with PD’s pathogenesis ([57]Kaufmann and
   Dorhoi, 2013). In 2018, Anetta, et al. suggested that the mechanism of
   our immune cells (macrophages) for wipe out the TB infection might also
   be involved in Parkinson’s disease. Generally, mutation in LRRK2 gene
   make the LRRK2-protein overactive in Parkinson’s disease. The LRRK2
   prevents phagosomes from fusing with lysosomes in macrophages, making
   them less efficient at clearing Mtb. Deleting the LRRK2 gene or
   treating the cells with an LRRK2 blocker significantly reduced the Mtb
   infection. So, drugs developed to treat PD (LRRK2 inhibitors) might
   work for TB too ([58]Härtlova et al., 2018).

   Tuberculosis and NCDs may not only co-exist but also increases the risk
   of each other. Developing tuberculosis disease may indicate background
   dysregulation of immune responses (innate immunity) in susceptible
   hosts, as these same abnormal responses may also predispose to CVD
   ([59]Marais et al., 2013; [60]Huaman et al., 2015). The burden of both
   diseases is enormous across the world and augment the risk of each
   other. The potential mechanistic association of TB with CVD is based on
   persistent immune activation in TB. Antibodies to mycobacterial HSP65
   cross-reacting with self-antigens in human vessels leading to
   autoimmunity may also affect CVD risk ([61]Huaman et al., 2015). The
   convergence of both diseases is posing a greater challenge for
   treatment plans in overlapping TB and CVD.

   The burden of diabetes has also been a major health concern in South
   Asian countries, with an estimated rise of more than 151% between 2,000
   and 2020 ([62]Jayawardena et al., 2012; [63]Shrestha et al., 2020).
   There is a bidirectional connection between TB and DM, and their
   synergistic role in causing human disease is well recognized. There is
   very little information available about the exact mechanism of how
   diabetes comorbidity impacts health outcomes in TB patients. However,
   there is some evidence for the negative impact of diabetes comorbidity
   on the TB treatment outcome ([64]Dooley et al., 2009; [65]Wang et al.,
   2009; [66]Chiang et al., 2015), specifically for delays in treatment
   failures, mycobacterial clearance, death, relapse, and re-infection.

   Furthermore, It has been also seen that tuberculosis lead to impair the
   induction of glucose intolerance and worsening of glycaemic control in
   DM patients ([67]Melmed, 2011). TB also has a bidirectional
   epidemiological association with RA and has reported that patients with
   RA have a 4-fold higher risk of developing TB than the control
   population ([68]Carmona et al., 2003). In this double burden disease,
   on one side, immunological responses involving Th1 mediated activation
   of cytokines are key to protect against TB ([69]Barnes and Wizel, 2000;
   [70]Stenger, 2005; [71]Yasui, 2014), while on the other side,
   anti-rheumatic drugs (tDMARDs) that act against the host immune system
   are increasing the risk of TB in RA patients ([72]Lim et al., 2017).
   Moreover, several studies have reported the reactivation of TB in RA
   patients treated with anti–TNF-α agents ([73]Keane et al., 2001;
   [74]Ormerod, 2004; [75]Dixon et al., 2010).

   The overlapping of TB and lung cancer has attracted many researchers in
   the last few decades. Many studies have reported that TB is associated
   with cancer and increases the risk and mortality of lung cancer and
   vice versa ([76]Leung et al., 2013). However, data related to TB
   treatment of LC patients is still incomplete and inconsistent. The
   connection between tuberculosis and lung cancer is still not completely
   understood. Lung parenchyma tissue involved in both diseases, regular
   cough in lung cancer, morphological vascular variations, lymphocytosis
   mechanisms, and production of immune system mediators like interleukins
   are all among the factors leading to the hypothesis about the major
   role of tuberculosis in lung cancer ([77]Liang et al., 2009;
   [78]Brenner et al., 2011; [79]Bhatt et al., 2012). It has been shown
   that the inflammatory process is one of the potential factors of lung
   cancer, and the crucial inflammation-inducing factors are tuberculosis
   (TB), pneumonia, and chronic bronchitis, among which TB has a more
   profound role in the emergence of lung cancer ([80]Keikha and Esfahani,
   2018). Many studies reported that the induction of necrosis and
   apoptosis or TB reactivation might result in increasing TNF-α and IL-17
   that will either decreases the activity of P53 or increase the BCL-2
   expression, decrease Bax-T, and cause the inhibition of caspase-3
   expression due to decreasing the expression of mitochondria cytochrome
   oxidase ([81]Mariani et al., 2001; [82]Liuzzo et al., 2013). It is
   clear that the epidemiological shift creates a double disease burden in
   the affected population and is rising as a critical health problem
   globally. The intersection between TB and other NCDs poses
   pharmacological issues and a great challenge for the co-management and
   treatment, reflecting a need for a radical shift, emphasizing common
   treatment targets irrespective of vertical approaches focused on
   individual diseases.

   Recently, Gysi et al. implemented a network-medicine and
   drug-repurposing approach to identify repurposable drugs for COVID-19
   ([83]Morselli Gysi et al., 2021). Sakle et al. have used a network
   pharmacology-based approach to prove that Caesalpinia pulcherima (CP)
   is a multi-target herb for the betterment of clinical uses for the
   treatment of breast cancer ([84]Sakle et al., 2020). Besides, Azuaje,
   et al. had provided systemic insights into cardiovascular effects of
   non-cardiovascular drugs by combining different sources of drug and
   protein interaction information to assemble the myocardial infarction
   drug-target interactome network ([85]Azuaje et al., 2011) In another
   similar study, Kim et al. has suggested that network-based drug-disease
   proximity offers a novel perspective into a drug’s therapeutic effect
   in the Systemic Sclerosis (SSc) disease and that could be applied to
   drug combinations or drug repositioning ([86]Kim et al., 2020).

   Network analysis is uniquely suited to approach based on the
   theoretical paradigm and methodological tools to research, describe,
   explore, and understand structural and relational aspects of human
   health and diseases ([87]Luke and Harris, 2007). Network-based studies
   are emerging as an important tool to determine the disease
   susceptibility genes and their relationship with different diseases.
   These studies have also improved our understanding of drug targets and
   their effects and suggested new drug targets, therapeutics, and
   therapeutic management approaches in severe diseases ([88]Berger and
   Iyengar, 2009). Analysis of networks is significantly contributing to
   the genesis of systems pharmacology.

   The current study was proposed to build a disease network based on the
   overlapping of TB with other NCDs, namely PD, CVD, DM, RA, and LC. The
   disease network was analyzed to identify the TB-associated genes that
   are commonly associated with other NCDs and determine the gene-disease
   relationship. Next, we constructed the gene interaction network of each
   disease independently by integrating curated and experimentally
   validated interactions in humans ([89]Barabási and Oltvai, 2004). All
   the gene interaction networks were merged into a single large network
   using the graph union operation, and the network’s structural
   properties were distinguished through the behavior of the topological
   parameters followed by modules identification because modules in a
   large network are functionally and statistically significant
   interacting clusters of nodes that resemble community organizations.
   Next, we generate and analyzed the drug-target interactome network,
   which integrates data about clinically relevant drug-drug and
   drug-target interactions. The resulting network lays the basis for a
   broader picture of the drug-target interaction landscape. The overall
   study offers new opportunities for understanding the biological basis
   of treatment efficacy and targeted and multidrug therapy in TB and its
   overlapping NCDs.

Material and Methods

   The schematic workflow of this study is represented in [90]Figure 1.

FIGURE 1.

   [91]FIGURE 1
   [92]Open in a new tab

   The schematic representation of workflow and methodology used in this
   study.

Collection of Disease-Associated Genes

   Disease-associated genes of Tuberculosis (TB), along with its
   associated non-communicable diseases, namely Parkinson disease (PD),
   cardiovascular disease (CVD), diabetes mellitus (DM), rheumatoid
   arthritis (RA), and lung cancer (LC), were obtained from the DisGeNet
   (v7.0), a database comprehensively integrated expert-curated. DisGeNET
   contains a compilation of genes associated to diseases, that taken from
   several publicly available databases including, UniProt/SwissProt,
   Cancer Genome Interpreter (CGI), Comparative Toxicogenomics Database™
   (CTD™), Orphanet, Mouse Genome Database (MGD), PsyGeNET, Genomics
   England, ClinGen, and Rat Genome Database (RGD) ([93]Piñero et al.,
   2017).

   The gene-disease correlation was analyzed and selected only those genes
   with many publications supporting the association (PubMed references