Cyclic di-AMP-mediated interaction between Mycobacterium tuberculosis ΔcnpB and macrophages implicates a novel strategy for improving BCG vaccination

Cyclic di-AMP-mediated interaction between Mycobacterium tuberculosis ΔcnpB and macrophages... Abstract Cyclic di-AMP (c-di-AMP) has been shown to play an important role in bacterial physiology and pathogen–host interactions. We previously reported that deletion of the sole c-di-AMP phosphodiesterase-encoding gene (cnpB) in Mycobacterium tuberculosis (Mtb) led to significant virulence attenuation. In this study, we found that ΔcnpB of M. bovisbacillus Calmette-Guerin (BCG) was unable to secrete c-di-AMP, which differs from Mtb ΔcnpB. We infected bone marrow-derived macrophages (BMDMs) with c-di-AMP-associated mutants generated from both Mtb and BCG. Our results showed that upon infection with Mtb ΔcnpB, BMDMs of wildtype mice secreted a large amount of interferon-β (IFN-β) post-infection similarly as we reported previously. In contrast, the response was less pronounced with BMDMs isolated from cGAS−/− mice and was nearly abolished with BMDMs prepared from STING−/− mice. Deletion of the region of difference 1 (RD1) locus in Mtb ΔcnpB did not alter the c-di-AMP secretion of ΔcnpB but eliminated the IFN-β production in the infected cells. In contrast, neither BCG ΔcnpB nor a recombinant BCG ΔcnpB with a pRD1 cosmid induced a type I interferon response. Interestingly, multiple studies have demonstrated that type I IFN enhances BCG’s immunity. Thus, amending BCG based on our findings might improve BCG vaccination. Mycobacterium tuberculosis, BCG, c-di-AMP, RD1, IFN-β, macrophage INTRODUCTION Tuberculosis (TB) remains a major threat to public health. However, the pathogenesis of the etiologic agent, Mycobacterium tuberculosis (Mtb), is still not fully understood. Mycobacterium bovisbacillus Calmette-Guerin (BCG) is the only available TB vaccine, but inadequately controls the TB epidemic. The molecular basis of BCG’s limitation in TB prevention is still elusive. Thus, it is essential to understand how Mtb and BCG interact with the host in order to develop a better vaccination strategy. Cyclic di-AMP (c-di-AMP) was first recognized as a new second messenger in 2008 (Witte et al.2008). This di-nucleotide is produced from ATP by diadenylate cyclases and cleaved to pApA or AMP by distinct c-di-AMP phosphodiesterases (Bai et al.2013; Corrigan et al.2013; Manikandan et al.2014). The diadenylate cyclase domain is widespread in bacteria and archaea, but has not been found in eukaryotic cells (Romling 2008; Corrigan et al.2013). Presence of active diadenylate cyclase has been experimentally demonstrated in several bacterial species (Witte et al.2008; Woodward, Iavarone and Portnoy 2010; Corrigan et al.2011; Kamegaya, Kuroda and Hayakawa 2011; Luo and Helmann 2012; Bai et al.2013; Barker et al.2013; Mehne et al.2013), and the enzyme is essential for bacterial viability of Bacillus subtilis (Luo and Helmann 2012, Mehne et al.2013), Staphylococcus aureus (Corrigan et al.2011), Streptococcus pneumoniae (Bai et al.2013) and Listeria monocytogenes at certain growth conditions (Woodward et al.2010; Witte et al.2013; Whiteley, Pollock and Portnoy 2015). Additionally, c-di-AMP is associated with bacterial infections (Cron et al.2011; Pozzi et al.2012; Schwartz et al.2012; Yamamoto et al.2012; Witte et al.2013; Dey et al.2015), peptidoglycan homeostasis (Corrigan et al.2011; Luo and Helmann 2012), antibiotic resistance (Banerjee et al.2010; Corrigan et al.2011; Griffiths and O’Neill 2012; Luo and Helmann 2012; Witte et al.2013) and bacterial morphology and/or biofilm formation (Corrigan et al.2011; Pozzi et al.2012; Bai et al.2013; Yang et al.2014; Peng et al.2016). Furthermore, c-di-AMP produced by L. monocytogenes, Chlamydia trachomatis and Mtb promotes a type I interferon (IFNα/β) response mediated by host effectors STING and DDX41 through an IRF-3-regulated signaling pathway (Woodward et al.2010; Burdette et al.2011; Jin et al.2011; Sauer et al.2011; Bowie 2012; Parvatiyar et al.2012; Barker et al.2013; Yang et al.2014; Dey et al.2015). We previously identified DisA and CnpB as diadenylate cyclase and c-di-AMP phosphodiesterase, respectively, in Mtb (Bai et al.2012; Yang et al.2014) and demonstrated that deletion of disA abolished Mtb's c-di-AMP production. More interestingly, deletion of cnpB elevated bacterial c-di-AMP accumulation and secretion, induced a strong host type I IFN response and significantly attenuated the virulence in a mouse infection model (Yang et al.2014; Dey et al.2017). A similar observation was reported with a disA-overexpressing Mtb strain (Dey et al.2015). Together, these observations suggest that elevated c-di-AMP levels play a protective role during Mtb infection. It has been shown that both BCG and Mtb ΔRD1 (region of difference 1) fail to trigger a relevant host type I IFN response compared to Mtb wildtype (WT) (Novikov et al.2011; Dey et al.2015). Interestingly, addition of type I IFN improves BCG’s immunogenicity in vitro (Giacomini et al.2009), and boosting BCG vaccination with type I IFN protects mice against Mycobacterium lepraemurium infection (Guerrero et al.2015). Additionally, a recombinant BCG vaccine with ESAT-6 secretion complex-1 (ESX-1) of Mycobacterium marinum induces type I IFN and enhances T cell immunity against Mtb challenge (Groschel et al.2017). These observations implicate that the defect in type I IFN induction is one of the reasons for the failure of BCG. It has been well established that Mtb ESX-1 is required for the type I IFN response both in vitro and in vivo (Stanley et al.2007). ESX-1 is partially encoded by genes in RD1, a 9.5-kb locus in the Mtb genome but missing in BCG (Mahairas et al.1996; Behr et al.1999; Gordon et al.1999). RD1 contains nine genes, including esxA and esxB, which encode for ESAT-6 and CFP-10, respectively. Earlier reports support that ESAT-6 is essential for rupturing the phagosomal membrane and releasing the mycobacterial components into the cytosol of the host cell (van der Wel, Hava and Houben 2007; Houben et al.2012; Simeone et al.2012, 2015). However, a recent study suggests that ESX-1 but not ESAT-6 is needed to permeabilize the phagosomal membrane (Conrad et al.2017). Nonetheless, it is unknown whether this process is required for the phagosome-to-cytosol translocation of c-di-AMP as a small molecule produced by Mtb. In the current study, we demonstrate that c-di-AMP can be secreted by Mtb ΔcnpB but not BCG ΔcnpB. Both c-di-AMP secretion machinery and RD1 are required for Mtb to induce the c-di-AMP-mediated host response by the infected macrophages, which is observed from a type I IFN response in infected macrophages. Since type I IFN improves BCG’s immunogenicity, amending BCG to induce type I IFN response based on the c-di-AMP signaling network of Mtb will likely improve BCG’s efficacy. MATERIALS AND METHODS Bacterial strains and growth conditions The bacterial strains used in this study are listed in Table 1. Mtb H37Rv, BCG (Pasteur strain; Trudeau Institute) and their derivatives were grown at 37°C in mycomedium (Middlebrook 7H9 medium (BD) supplemented with 0.5% glycerol, 10% oleic acid-albumin-dextrose-catalase (OADC), and 0.05% Tween-80), Sauton's medium (Bai, Schaak and McDonough 2009) or Middlebrook 7H10 agar supplemented with 10% OADC and 0.01% cycloheximide. Fresh cultures were inoculated from frozen stocks for every experiment. Cultures were grown in tissue culture flasks standing with ambient air. Escherichia coli strains were grown in Luria-Bertani (LB) broth or on LB agar plates. Mycobacterium smegmatis mc2155 was grown in mycomedium. All cultures were grown at 37°C, except that M. smegmatis was grown at 30°C. Kanamycin at 25 μg/ml, hygromycin at 50 μg/ml or zeocin at 100 μg/ml were added when necessary. Table 1. Bacterial strains and plasmids used in this study. Strain or plasmid  Descriptiona  Source  Bacterial strain  Mycobacterium tuberculosis   H37Rv strain  WT  ATCC   Mt008  ΔdisA; Hygr  (Yang et al.2014)   Mt012  Mt008::pGB086; Hygr; Kanr  (Yang et al.2014)   Mt013  ΔcnpB; Hygr  (Yang et al., 2014)   Mt016  Mt013:pGB120; Hygr; Kanr  (Yang et al.2014)   Mt017  Unmarked ΔcnpB  This study   Mt020  ΔdisAΔcnpB; Hygr  This study   Mt023  ΔRD1; Hygr  This study   Mt024  ΔcnpBΔRD1; Hygr  This study  Mycobacterium bovis BCG   Pasteur strain  WT  Trudeau Institute   ΔdisA  ΔdisA; Hygr  This study   mb042  ΔdisA::pGB086; Hygr; Kanr  This study   mb069  ΔcnpB; Hygr  This study   mb072  mb069::pGB120; Hygr; Kanr  This study   mb076  Unmarked ΔcnpB  This study   mb092  mb076(pGB192); Hygr; Kanr  This study   mb096  mb076::pMBC1260; Kanr  This study   mb097  mb076::pGB218; Hygr; Kanr  This study   mb098  ΔdisAΔcnpB; Hygr  This study   Mt030  BCG WT::pRD1; Hygr  This study   Mt031  mb076::pRD1; Hygr  This study  Phage   pMBC1206  Deletion of disA  (Yang et al.2014)   pGB139  Deletion of cnpB  (Yang et al.2014)   RD1  Deletion of RD1  (Hsu et al.2003)  Plasmid   pMBC283  Multiple-copy expression vector with tuf promoter; Kanr; Hygr; Apr  (Bai et al.2011)   pMBC1260  Single-copy expression vector with Rv0805 promoter; Kanr; Apr  (Yang et al.2014)   pGB086  Complementation of ΔdisA; Kanr; Apr  (Yang et al.2014)   pGB120  Complementation of ΔcnpB; Kanr; Apr  (Yang et al.2014)   pGB192  pMBC283 carrying Rv1877 ORF; Kanr; Hygr; Apr  This study   pGB218  pMBC1260 carrying Rv3877 ORF; Kanr; Apr  This study   pYO11  Unmark the hyg marker in the mutants; Zeor  (Rosenberg et al.2015)   pRD1  Cosmid flanking RD1  (Pym et al.2003)  Strain or plasmid  Descriptiona  Source  Bacterial strain  Mycobacterium tuberculosis   H37Rv strain  WT  ATCC   Mt008  ΔdisA; Hygr  (Yang et al.2014)   Mt012  Mt008::pGB086; Hygr; Kanr  (Yang et al.2014)   Mt013  ΔcnpB; Hygr  (Yang et al., 2014)   Mt016  Mt013:pGB120; Hygr; Kanr  (Yang et al.2014)   Mt017  Unmarked ΔcnpB  This study   Mt020  ΔdisAΔcnpB; Hygr  This study   Mt023  ΔRD1; Hygr  This study   Mt024  ΔcnpBΔRD1; Hygr  This study  Mycobacterium bovis BCG   Pasteur strain  WT  Trudeau Institute   ΔdisA  ΔdisA; Hygr  This study   mb042  ΔdisA::pGB086; Hygr; Kanr  This study   mb069  ΔcnpB; Hygr  This study   mb072  mb069::pGB120; Hygr; Kanr  This study   mb076  Unmarked ΔcnpB  This study   mb092  mb076(pGB192); Hygr; Kanr  This study   mb096  mb076::pMBC1260; Kanr  This study   mb097  mb076::pGB218; Hygr; Kanr  This study   mb098  ΔdisAΔcnpB; Hygr  This study   Mt030  BCG WT::pRD1; Hygr  This study   Mt031  mb076::pRD1; Hygr  This study  Phage   pMBC1206  Deletion of disA  (Yang et al.2014)   pGB139  Deletion of cnpB  (Yang et al.2014)   RD1  Deletion of RD1  (Hsu et al.2003)  Plasmid   pMBC283  Multiple-copy expression vector with tuf promoter; Kanr; Hygr; Apr  (Bai et al.2011)   pMBC1260  Single-copy expression vector with Rv0805 promoter; Kanr; Apr  (Yang et al.2014)   pGB086  Complementation of ΔdisA; Kanr; Apr  (Yang et al.2014)   pGB120  Complementation of ΔcnpB; Kanr; Apr  (Yang et al.2014)   pGB192  pMBC283 carrying Rv1877 ORF; Kanr; Hygr; Apr  This study   pGB218  pMBC1260 carrying Rv3877 ORF; Kanr; Apr  This study   pYO11  Unmark the hyg marker in the mutants; Zeor  (Rosenberg et al.2015)   pRD1  Cosmid flanking RD1  (Pym et al.2003)  aKanr, kanamycin resistance; Hygr, hygromycin resistance; Apr, ampicillin resistance; Zeor, zeocin resistance. View Large Table 1. Bacterial strains and plasmids used in this study. Strain or plasmid  Descriptiona  Source  Bacterial strain  Mycobacterium tuberculosis   H37Rv strain  WT  ATCC   Mt008  ΔdisA; Hygr  (Yang et al.2014)   Mt012  Mt008::pGB086; Hygr; Kanr  (Yang et al.2014)   Mt013  ΔcnpB; Hygr  (Yang et al., 2014)   Mt016  Mt013:pGB120; Hygr; Kanr  (Yang et al.2014)   Mt017  Unmarked ΔcnpB  This study   Mt020  ΔdisAΔcnpB; Hygr  This study   Mt023  ΔRD1; Hygr  This study   Mt024  ΔcnpBΔRD1; Hygr  This study  Mycobacterium bovis BCG   Pasteur strain  WT  Trudeau Institute   ΔdisA  ΔdisA; Hygr  This study   mb042  ΔdisA::pGB086; Hygr; Kanr  This study   mb069  ΔcnpB; Hygr  This study   mb072  mb069::pGB120; Hygr; Kanr  This study   mb076  Unmarked ΔcnpB  This study   mb092  mb076(pGB192); Hygr; Kanr  This study   mb096  mb076::pMBC1260; Kanr  This study   mb097  mb076::pGB218; Hygr; Kanr  This study   mb098  ΔdisAΔcnpB; Hygr  This study   Mt030  BCG WT::pRD1; Hygr  This study   Mt031  mb076::pRD1; Hygr  This study  Phage   pMBC1206  Deletion of disA  (Yang et al.2014)   pGB139  Deletion of cnpB  (Yang et al.2014)   RD1  Deletion of RD1  (Hsu et al.2003)  Plasmid   pMBC283  Multiple-copy expression vector with tuf promoter; Kanr; Hygr; Apr  (Bai et al.2011)   pMBC1260  Single-copy expression vector with Rv0805 promoter; Kanr; Apr  (Yang et al.2014)   pGB086  Complementation of ΔdisA; Kanr; Apr  (Yang et al.2014)   pGB120  Complementation of ΔcnpB; Kanr; Apr  (Yang et al.2014)   pGB192  pMBC283 carrying Rv1877 ORF; Kanr; Hygr; Apr  This study   pGB218  pMBC1260 carrying Rv3877 ORF; Kanr; Apr  This study   pYO11  Unmark the hyg marker in the mutants; Zeor  (Rosenberg et al.2015)   pRD1  Cosmid flanking RD1  (Pym et al.2003)  Strain or plasmid  Descriptiona  Source  Bacterial strain  Mycobacterium tuberculosis   H37Rv strain  WT  ATCC   Mt008  ΔdisA; Hygr  (Yang et al.2014)   Mt012  Mt008::pGB086; Hygr; Kanr  (Yang et al.2014)   Mt013  ΔcnpB; Hygr  (Yang et al., 2014)   Mt016  Mt013:pGB120; Hygr; Kanr  (Yang et al.2014)   Mt017  Unmarked ΔcnpB  This study   Mt020  ΔdisAΔcnpB; Hygr  This study   Mt023  ΔRD1; Hygr  This study   Mt024  ΔcnpBΔRD1; Hygr  This study  Mycobacterium bovis BCG   Pasteur strain  WT  Trudeau Institute   ΔdisA  ΔdisA; Hygr  This study   mb042  ΔdisA::pGB086; Hygr; Kanr  This study   mb069  ΔcnpB; Hygr  This study   mb072  mb069::pGB120; Hygr; Kanr  This study   mb076  Unmarked ΔcnpB  This study   mb092  mb076(pGB192); Hygr; Kanr  This study   mb096  mb076::pMBC1260; Kanr  This study   mb097  mb076::pGB218; Hygr; Kanr  This study   mb098  ΔdisAΔcnpB; Hygr  This study   Mt030  BCG WT::pRD1; Hygr  This study   Mt031  mb076::pRD1; Hygr  This study  Phage   pMBC1206  Deletion of disA  (Yang et al.2014)   pGB139  Deletion of cnpB  (Yang et al.2014)   RD1  Deletion of RD1  (Hsu et al.2003)  Plasmid   pMBC283  Multiple-copy expression vector with tuf promoter; Kanr; Hygr; Apr  (Bai et al.2011)   pMBC1260  Single-copy expression vector with Rv0805 promoter; Kanr; Apr  (Yang et al.2014)   pGB086  Complementation of ΔdisA; Kanr; Apr  (Yang et al.2014)   pGB120  Complementation of ΔcnpB; Kanr; Apr  (Yang et al.2014)   pGB192  pMBC283 carrying Rv1877 ORF; Kanr; Hygr; Apr  This study   pGB218  pMBC1260 carrying Rv3877 ORF; Kanr; Apr  This study   pYO11  Unmark the hyg marker in the mutants; Zeor  (Rosenberg et al.2015)   pRD1  Cosmid flanking RD1  (Pym et al.2003)  aKanr, kanamycin resistance; Hygr, hygromycin resistance; Apr, ampicillin resistance; Zeor, zeocin resistance. View Large Construction and complementation of mutants in Mtb and BCG BCG ΔdisA and ΔcnpB were generated by using phages prepared with pMBC1206 and pGB132 (Table 1), respectively, which were previously used to delete disA and cnpB in Mtb by using a phage-based approach (Yang et al.2014). The mutants were selected with plates containing 50 μg/ml hygromycin and screened by PCR using primers as listed in Table S1, Supporting Information. Subsequently, ΔdisA and ΔcnpB of BCG were transformed with pGB086 and pGB120, respectively, for complementation similar to what we have performed to complement the relevant mutants in Mtb (Yang et al.2014). The hygromycin resistance marker in both ΔcnpB of Mtb and BCG was unmarked using pYO11 (provided by Dr Jeffery Cox) (Rosenberg et al.2015). Briefly, hygromycin-resistant ΔcnpB was transformed with pYO11 and selected with Middlebrook 7H10 plates containing 100 μg/ml zeocin. Zeocin-resistant transformants were passaged in mycomedium without antibiotics for at least three times to cure pYO11. Bacteria sensitive to both hygromycin and zeocin were confirmed as unmarked ΔcnpB by PCR using primers JY213 and JY214. The PCR products from the unmarked strains exhibit a significantly reduced size compared to those from the WT and the hygromycin-resistant ΔcnpB strains. ΔdisAΔcnpB of Mtb and BCG were then generated from the unmarked ΔcnpB strains by transducing the phage for deletion of disA similarly as we previously described (Yang et al.2014). Mtb ΔRD1 and ΔcnpBΔRD1 were generated using the phage-based method (Yang et al.2014). Briefly, a phage used to delete RD1 (provided by Dr William Jacobs, Jr) (Hsu et al.2003) was transduced into WT and the unmarked ΔcnpB mutant of Mtb. Replacement of RD1 with a hygromycin resistance marker in both genetic backgrounds was confirmed by PCR with primers JY407, JY002, JY003 and JY408 (Table S1, Supporting Information). Deletion of RD1 was further verified by PCR amplification of an esxA fragment using primers KM1325 and KM1326 (Table S1, Supporting Information). WT control exhibits a 250-bp product, whereas the mutants lack this fragment. Transformation of BCG strains with cosmid pRD1 BCG WT and ΔcnpB were transformed with cosmid pRD1 (provided by Dr Keith Derbyshire) using electroporation at 2.5 kV, 25 μF and 1000 Ω (Flint et al.2004). This cosmid harbors a 32-kb segment (Rv3861-Rv3885) flanking the 9.5-kb RD1 locus (Rv3871-Rv3879c) (Pym et al.2003). The recombinant strains were selected on a Middlebrook 7H10 plate containing 50 μg/ml hygromycin and were verified by PCR using primers specific to whiB6 (Rv3862), esxA (Rv3875) and Rv3879c (Table S1, Supporting Information). DNA isolated from BCG WT was used as a negative control for the PCR reaction. Detection of c-di-AMP c-di-AMP was extracted and measured by using an enzyme-linked immunosorbent assay (ELISA) as we described earlier (Underwood et al.2014; Yang et al.2014). Briefly, each bacterial strain was grown in 5 ml of Sauton's medium for 7 days. The optical density at 600 nm (OD600) was determined. Bacteria were harvested by centrifugation, and the supernatant was used for secreted c-di-AMP determination. Each bacterial pellet was resuspended in 0.5 ml 50 mM Tris-HCl (pH 8.0) and disrupted with 1 mm glass beads using a mini bead-beater (BioSpec). Bacterial debris was removed by centrifugation and the supernatant was boiled for 5 min and subsequently used as the bacterial c-di-AMP sample. c-di-AMP levels were quantified by using the competitive ELISA (Underwood et al.2014; Yang et al.2014). Results of both bacterial and secreted samples were normalized by respective OD600 readings. Preparation and infection of BMDMs This study was carried out in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Institutional Animal Care and Use Committee of Albany Medical College (Permit Number: 13-12006). WT mice were purchased from Taconic, and STING−/− and cGAS−/− mice were provided by Drs Lei Jin (Jin et al.2008) and Herbert Virgin (Schoggins et al.2014), respectively. The knockout mice were bred and maintained in the Animal Resources Facility at Albany Medical College. Bone marrow-derived macrophages (BMDMs) were generated as previously described (Jin et al.2011), which was approved by the Albany Medical College IACUC. Briefly, bone marrow cells were harvested from mouse femurs and cultured in complete RPMI-1640 supplemented 10% fetal bovine serum (FBS), 2 mM L-glutamine, 1 mM sodium pyruvate, 1× HEPES [4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid], 1× non-essential amino acids, 50 μM β-mercaptoethanol, 100 U/ml penicillin, 100 μg/ml streptomycin and 20 ng/ml M-CSF. All reagents were purchased from Gibco. BMDM cells were seeded in 12-well plates at 5 × 105 cells/ml and 0.5 ml/well in the supplemented RPMI-1640 medium without antibiotics. Cells were infected with Mtb, BCG or their derivatives at a multiplicity of infection (MOI) of 2:1. At 4 h post-infection, infection medium was replaced with fresh complete RPMI-1640 medium after washing thrice with PBS. At 1 and 2 days post-infection, culture supernatant was collected to determine cytokine production. Cytokine detection IFN-β in tissue culture supernatant was detected using ELISA with a Verikine-HS Mouse IFN-β serum ELISA kit (PBL Assay Science) following the manufacturer's manual similarly as we reported earlier (Yang et al.2014). Statistical analysis Comparisons of the c-di-AMP levels between WT, BCG and their derivatives were analyzed using a two-tailed t-test performed using Prism 5 (GraphPad Software). IFN-β levels were analyzed using a two-way ANOVA (GraphPad Software). P values of <0.05 were considered to be statistically significant. For IFN-β levels, each bacterial mutant was compared to the parental strain in infection of the BMDMs isolated from the same mouse. RESULTS DisA is likely the sole diadenylate cyclase in Mtb Increasing evidence has revealed that c-di-AMP participates in multiple critical intracellular bioprocesses. It has been reported that diadenylate cyclase-encoding genes are essential for several Gram-positive bacteria including S. aureus (Corrigan et al.2011), S. pneumoniae (Bai et al.2013) and L. monocytogenes (Woodward et al.2010; Witte et al.2013), each of which possesses a single diadenylate cyclase. A recent study has revealed that diadenylate cyclase is conditionally essential for L. monocytogenes when the bacterium is grown in rich media or within host cells, but not in minimal media that do not trigger (p)ppGpp synthesis (Whiteley et al.2015). B. subtilis encodes three diadenylate cyclases, which can be individually disrupted but cannot be deleted together when bacteria were grown in regular cultural media (Luo and Helmann 2012; Mehne et al.2013). A triple mutant of B. subtilis diandenylate cyclases was generated only in low K+ media (Gundlach et al.2017). DisA is the only known protein that possesses a diadenylate cyclase domain in Mtb. Surprisingly, we could easily delete disA from the Mtb genome, and the mutant grew well in various mycobacterial culture media and intracellularly (Yang et al.2014). We hypothesized that either the role of c-di-AMP in Mtb differs from that in Gram-positive bacteria or Mtb encodes more than one diadenylate cyclase, similar to B. subtilis. We generated a ΔdisAΔcnpB mutant in Mtb, which was unable to accumulate detectable c-di-AMP, similar to ΔdisA (Fig. 1A). This result indicates that DisA is very likely the sole diadenylate cyclase in this pathogen and that c-di-AMP is not required for the viability of Mtb. The role of c-di-AMP in Mtb is distinct from that in the Gram-positive bacteria that have been characterized for c-di-AMP to date. Figure 1. View largeDownload slide Determination of bacterial (A and B) and secreted (C and D) c-di-AMP. WT and the indicated derivatives of Mtb (A and C) and BCG (B and D), respectively, were grown in Sauton's broth for 7 days. Bacteria were harvested by centrifugation. The c-di-AMP levels in the supernatant (secreted) and in the lysate prepared from the pellet (bacterial) were determined using ELISA. The data shown are the means of three independent experiments. The error bars indicate the standard errors of mean (SEMs). *P < 0.05; **P < 0.01; ***P < 0.001. P values are statistical analyses resulted from a two-tailed t-test. Note the detectable limit of the assay is ∼10 nM. ΔdisA/C and ΔcnpB/C are the complemented strains for ΔdisA and ΔcnpB, respectively. Figure 1. View largeDownload slide Determination of bacterial (A and B) and secreted (C and D) c-di-AMP. WT and the indicated derivatives of Mtb (A and C) and BCG (B and D), respectively, were grown in Sauton's broth for 7 days. Bacteria were harvested by centrifugation. The c-di-AMP levels in the supernatant (secreted) and in the lysate prepared from the pellet (bacterial) were determined using ELISA. The data shown are the means of three independent experiments. The error bars indicate the standard errors of mean (SEMs). *P < 0.05; **P < 0.01; ***P < 0.001. P values are statistical analyses resulted from a two-tailed t-test. Note the detectable limit of the assay is ∼10 nM. ΔdisA/C and ΔcnpB/C are the complemented strains for ΔdisA and ΔcnpB, respectively. BCG differs from Mtb in c-di-AMP secretion In order to explore c-di-AMP-mediated biological differences between Mtb and BCG, we generated ΔdisA, ΔcnpB and ΔdisAΔcnpB mutants in BCG and compared these mutants with WT BCG as well as the Mtb strains in c-di-AMP production and secretion. We detected similar levels of c-di-AMP within BCG strains compared to those within the relevant Mtb strains (Fig. 1A and B). In contrast, secreted c-di-AMP was detected from Mtb ΔcnpB, but not BCG ΔcnpB (Fig. 1C and D). This result indicates that both DisA and CnpB are functional in BCG; however, the c-di-AMP secretion machinery of BCG is defective, which differs from Mtb. In L. monocytogenes, both MdrM and MdrT are capable of secreting c-di-AMP (Woodward et al.2010; Schwartz et al.2012; Tadmor et al.2014). These proteins are multidrug efflux pumps. We compared all the efflux pump-encoding genes of Mtb to those in the BCG genome by sequence alignment. Most efflux pumps of BCG are nearly identical in amino acid sequence to their orthologs in Mtb except the proteins encoded by Rv3877 and Rv1877 (Cole et al.1998; Brosch et al.2007). Rv3877 is located in the RD1 region (Behr et al.1999), which is absent in the BCG genome. In order to determine whether RD1 plays a role in c-di-AMP secretion, we deleted RD1 in both WT and ΔcnpB of Mtb and examined c-di-AMP secretion by the mutants. As a result, these mutants displayed no differences in c-di-AMP accumulation or secretion compared to their parental strains (Fig. S1A, Supporting Information). Consistently, we transformed either a plasmid expressing Rv3877 (not shown) or a cosmid flanking RD1 (pRD1) into BCG ΔcnpB (Fig. S1B, Supporting Information), and these recombinant strains were unable to secrete c-di-AMP. These observations suggest that Rv3877 does not play a role in c-di-AMP secretion. Additionally, the Rv1877 ortholog in BCG introduces a frameshift due to a single base insertion (*-c), which splits Rv1877 into two segments (Fig. S2, Supporting Information). We expressed Mtb Rv1877 in BCG ΔcnpB but could not detect secreted c-di-AMP from this recombinant strain (not shown), indicating that either Mtb utilizes other efflux pumps or it has a different mechanism for c-di-AMP secretion compared to L. monocytogenes. Both c-di-AMP secretion machinery and RD1 are needed in the c-di-AMP-mediated type I IFN response It is known that BCG and Mtb RD1 mutant are defective in inducing a type I IFN response (Giacomini et al.2009; Dey et al.2015). Interestingly, it has been shown that type I IFN improves BCG’s immunogenicity (Giacomini et al.2009), and boosting BCG vaccination with type I IFN provides better protection in mice challenged with M. lepraemurium (Guerrero et al.2015). Therefore, modifying BCG to be able to induce optimal levels of type I IFN response might improve BCG’s efficacy. In this study, we explored the molecular basis for BCG’s failure in the type I IFN response. It has been well established that c-di-AMP produced by intracellular bacteria stimulates a type I IFN response (Woodward et al.2010; Sauer et al.2011; Schwartz et al.2012; Barker et al.2013; Archer, Durack and Portnoy 2014; Dey et al.2015). Our previous study revealed that Mtb ΔcnpB secretes a large amount of c-di-AMP and induces high levels of IFN-β during a macrophage infection (Yang et al.2014), which is consistent with a recent report (Dey et al.2017) and the observation from a disA-overexpressing Mtb strain (Dey et al.2015). Several recent reports show that during Mtb infection, Mtb DNA rather than c-di-AMP is a major trigger for the type I IFN response, which is also cGAS-dependent (Collins et al.2015; Wassermann et al.2015; Watson et al.2015). Therefore, the IFN-β response we previously observed is either due to enhanced DNA secretion or partially contributed by the large amount of c-di-AMP. Furthermore, it has been shown that RD1 is required for releasing mycobacterial components from the phagosome into the cytosol (Conrad et al.2017). However, it is unknown whether Mtb-secreted c-di-AMP can be directly released from the phagosome or it is dependent upon RD1. Thus, we sought to address two major questions: (i) Is the elevated type I IFN response to Mtb ΔcnpB infection mediated by DNA or c-di-AMP? and (ii) Is RD1 required for the ΔcnpB-induced type I IFN response? In this study, we isolated BMDMs from WT, STING−/− and cGAS−/− mice and examined the IFN-β production by the cells infected with Mtb, BCG and their derivatives, respectively. Similar results were observed between the samples collected at 1 day (Fig. 2) and 2 day (not shown) post-infection. From BMDMs of WT mice, Mtb ΔdisA stimulated less IFN-β than the WT strain (Fig. 2A). In contrast, Mtb ΔcnpB induced a considerably larger amount of IFN-β than WT, similarly as we reported earlier (Yang et al.2014). Moreover, ΔdisAΔcnpB failed to trigger a type I response (Fig. 2A). These results indicate an association between c-di-AMP and the type I IFN response. By comparing infected BMDMs isolated from WT, STING−/− and cGAS−/− mice, we found that IFN-β production induced by Mtb WT and ΔcnpB was nearly abolished in the infected STING−/− cells, and cGAS−/− cells exhibited reduced IFN-β production in a c-di-AMP-dependent pattern (Fig. 2). This result is consistent with a recent report by Dey et al. using a transposon mutant of cnpB (Dey et al.2017). Overall, our results indicate that DNA is a major player of the Mtb-induced type I IFN response, which is consistent with several recent reports (Collins et al.2015; Wassermann et al.2015; Watson et al.2015). Meanwhile, c-di-AMP also contributes directly to a substantial type I IFN response during Mtb infection, which is also coherent with an earlier observation (Dey et al.2015). Figure 2. View largeDownload slide IFN-β secretion by BMDMs isolated from WT (A), STING−/− (B) and cGAS−/− (C) mice. BMDMs were infected with the indicated bacterial strains for 4 h. After thoroughly washing, cells were incubated for 24 h, and the supernatant was then collected to determine IFN-β. The data shown are the means of three independent experiments. The error bars indicate the SEMs. *P < 0.05; **P < 0.01; ***P < 0.001. P values are statistical analyses resulted from a two-tailed t-test to compare each bacterial mutant to the parental strain in infection of the BMDMs isolated from the same mouse. Figure 2. View largeDownload slide IFN-β secretion by BMDMs isolated from WT (A), STING−/− (B) and cGAS−/− (C) mice. BMDMs were infected with the indicated bacterial strains for 4 h. After thoroughly washing, cells were incubated for 24 h, and the supernatant was then collected to determine IFN-β. The data shown are the means of three independent experiments. The error bars indicate the SEMs. *P < 0.05; **P < 0.01; ***P < 0.001. P values are statistical analyses resulted from a two-tailed t-test to compare each bacterial mutant to the parental strain in infection of the BMDMs isolated from the same mouse. RD1 has been shown to be essential for Mtb to permeabilize the phagosomal membrane of the infected cells (de Jonge et al.2007; Stanley et al.2007; De Leon et al.2012). In this study, we explored whether RD1 is required for the Mtb ΔcnpB-induced type I IFN response, as c-di-AMP is a small molecule, which may penetrate phagosomal membrane directly. Remarkably, IFN-β production was dramatically reduced by infection with ΔRD1 generated in either the WT or the ΔcnpB genetic background (Fig. 2), even though ΔcnpBΔRD1 still secretes c-di-AMP as potently as the ΔcnpB mutant (Fig. S1, Supporting Information). This result indicates that RD1 is required for the Mtb ΔcnpB-induced type I IFN response. With BCG strains, BCG WT failed to trigger an IFN-β response, which is consistent with other reports (Giacomini et al.2009; Dey et al.2015). Furthermore, neither BCG ΔcnpB nor BCG ΔcnpB bearing the pRD1 cosmid elicited a type I IFN response, suggesting that the absence of RD1 is not the only reason for BCG’s defect in the type I IFN induction. Collectively, we conclude that both the c-di-AMP secretion machinery and RD1 are required for c-di-AMP-mediated response in the infected macrophages. DISCUSSION In this study, we demonstrate that both the c-di-AMP secretion machinery and RD1 are required for Mtb to induce the host type I IFN response. We propose a model for this response based on our observations together with several earlier reports (Manzanillo et al.2012; Collins et al.2015; Wassermann et al.2015; Watson et al.2015; Conrad et al.2017) (Fig. 3). In this model, c-di-AMP is secreted from Mtb into an infected phagosome by a currently unknown mechanism. Proteins encoded by the ESX-1 facilitate c-di-AMP releasing from the phagosome into the cytosol of the infected macrophage. Meanwhile, Mtb DNA is also released through the permeabilized phagosome into the cytosol and activates cGAS to produce cyclic GMP-AMP (cGAMP). Both c-di-AMP and cGAMP are then sensed by STING, which induces the type I IFN response. Figure 3. View largeDownload slide Working model for the host type I IFN response to Mtb c-di-AMP. In a phagosome of the infected macrophage, Mtb secretes c-di-AMP with a currently unknown mechanism. In the meantime, the phagosomal membrane was ruptured by an ESX-1-associated mechanism. Mtb components including DNA and c-di-AMP are released into the cytosol. Mtb DNA activates cGAS to produce cGAMP. Both cGAMP and c-di-AMP are detected by STING, which subsequently induces the host type I IFN response. Figure 3. View largeDownload slide Working model for the host type I IFN response to Mtb c-di-AMP. In a phagosome of the infected macrophage, Mtb secretes c-di-AMP with a currently unknown mechanism. In the meantime, the phagosomal membrane was ruptured by an ESX-1-associated mechanism. Mtb components including DNA and c-di-AMP are released into the cytosol. Mtb DNA activates cGAS to produce cGAMP. Both cGAMP and c-di-AMP are detected by STING, which subsequently induces the host type I IFN response. Three mycobacterial components have been shown to induce a type I IFN response: (i) c-di-AMP directly detected by STING (Yang et al.2014; Dey et al.2015); (ii) DNA through a cGAS—STING signaling pathway (Manzanillo et al.2012; Collins et al.2015; Wassermann et al.2015; Watson et al.2015); and (iii) peptidoglycan through an NOD2 signaling pathway (Pandey et al.2009). We have demonstrated that the elevated IFN-β production induced by Mtb ΔcnpB is unlikely resulted from the NOD2 signaling pathway by using BMDMs isolated from NOD2−/− mice (Yang et al.2014). In this study, we show that the majority of IFN-β induced by Mtb ΔcnpB is elicited by the DNA-sensing pathway through cGAS, indicating that ΔcnpB enhanced Mtb DNA release after 1 and 2 days post-infection, which is consistent with the recent reports regarding the host response to mycobacterial DNA (Collins et al.2015; Wassermann et al.2015; Watson et al.2015) and c-di-AMP-mediated type I IFN induction by L. monocytogenes (Hansen et al.2014). Additionally, Mtb ΔcnpB also induces a small amount of IFN-β in a STING-dependent but cGAS-independent manner, suggesting that this proportion of IFN-β is directly stimulated by c-di-AMP. As we pointed out in an earlier report, secreted c-di-AMP that we detected from Mtb ΔcnpB is unlikely due to bacterial lysis (Yang et al.2014). Meanwhile, mycobacterial DNA sensed by cGAS has also been considered to originate from living but not degraded bacteria (Majlessi and Brosch 2015). Based on the results of the type I IFN response of BMDMs isolated from WT, STING−/− and cGAS−/− mice, c-di-AMP possibly enhances DNA secretion by Mtb. It is also likely that Mtb DNA and c-di-AMP share the same secretion machinery, which we will explore in the future. Several studies have demonstrated that type I IFN increased host susceptibility to Mtb infection (Manca et al.2001, 2005; Antonelli et al.2010; Dorhoi et al.2014). On the other hand, there is also evidence that type I IFN is protective in certain circumstances (Desvignes, Wolf and Ernst 2012). Thus, it is likely that a balance of this cytokine is required for optimal protection during Mtb infection (Wiens and Ernst 2016). Nonetheless, the recombinant BCG vaccine with ESX-1 of M. marinum induces type I interferon and stronger T cell response, which provides better protection against Mtb challenge (Groschel et al.2017), suggesting that the type I IFN induced during vaccination is beneficial to the host. Most interestingly, we found that BCG ΔcnpB was unable to secrete c-di-AMP. This observation is highly significant from the vaccine perspective, since the limitation of BCG in the prevention of the TB epidemic is still not fully understood. Manipulation of c-di-AMP secretion from BCG and release from the phagosome into host cytosol during vaccination will enable BCG to induce substantial levels of type I IFN locally, which might improve BCG’s immunogenicity based on multiple vaccine studies (Giacomini et al.2009; Guerrero et al.2015; Conrad et al.2017; Groschel et al.2017). This ‘localized’ response could be optimized by the levels of c-di-AMP production and should not significantly enhance the susceptibility of Mtb infection. Furthermore, c-di-AMP has been explored as an adjuvant for vaccinations and induces strong humoral and cellular immune responses (Ebensen et al.2011; Sanchez et al.2014; Skrnjug et al.2014). Therefore, an engineered BCG based on ΔcnpB and amended c-di-AMP secretion machinery may elevate the vaccine efficacy. In our future studies, we will generate and examine such recombinant BCG strains as novel vaccines in protection against TB. SUPPLEMENTARY DATA Supplementary data are available at FEMSPD online. Acknowledgements We thank Tiffany Zarrella and Dr Gwendowlyn Knapp for critical reading of the manuscript; Dr Lei Jin for providing STING−/− mice and the technical assistance; Dr Herbert Virgin for providing cGAS−/− mice; Dr Williams Jacobs, Jr. for providing the phage to delete RD1 locus; Dr Keith Derbyshire for providing the cosmid harboring the RD1 locus; and Dr Jeffery Cox for providing pYO11 plasmid. We are grateful to the Biosafety Level 3 Core Facility of Albany Medical College. FUNDING This project is partly supported by a Scientist Development Grant of the American Heart Association 12SDG12080067 to GB. Conflict of interest. None declared. Footnotes Present address: Shanghai Institute for Advanced Immunochemical Studies, 99 Haike Road, Pudong District, Shanghai, China. REFERENCES Antonelli LR, Gigliotti Rothfuchs A, Goncalves R et al.   Intranasal Poly-IC treatment exacerbates tuberculosis in mice through the pulmonary recruitment of a pathogen-permissive monocyte/macrophage population. J Clin Invest  2010; 120: 1674– 82. Google Scholar CrossRef Search ADS PubMed  Archer KA, Durack J, Portnoy DA. STING-dependent type I IFN production inhibits cell-mediated immunity to Listeria monocytogenes. PLoS Pathog  2014; 10: e1003861. Google Scholar CrossRef Search ADS PubMed  Bai G, Schaak DD, McDonough KA. cAMP levels within Mycobacterium tuberculosis and Mycobacterium bovis BCG increase upon infection of macrophages. FEMS Immunol Med Microbiol  2009; 55: 68– 73. Google Scholar CrossRef Search ADS PubMed  Bai G, Schaak DD, Smith EA et al.   Dysregulation of serine biosynthesis contributes to the growth defect of a Mycobacterium tuberculosis crp mutant. Mol Microbiol  2011; 82: 180– 98. Google Scholar CrossRef Search ADS PubMed  Bai Y, Yang J, Eisele LE et al.   Two DHH subfamily 1 proteins in Streptococcus pneumoniae possess cyclic di-AMP phosphodiesterase activity and affect bacterial growth and virulence. J Bacteriol  2013; 195: 5123– 32. Google Scholar CrossRef Search ADS PubMed  Bai Y, Yang J, Zhou X et al.   Mycobacterium tuberculosis Rv3586 (DacA) is a diadenylate cyclase that converts ATP or ADP into c-di-AMP. PLoS One  2012; 7: e35206. Google Scholar CrossRef Search ADS PubMed  Banerjee R, Gretes M, Harlem C et al.   A mecA-negative strain of methicillin-resistant Staphylococcus aureus with high-level beta-lactam resistance contains mutations in three genes. Antimicrob Agents Ch  2010; 54: 4900– 2. Google Scholar CrossRef Search ADS   Barker JR, Koestler BJ, Carpenter VK et al.   STING-dependent recognition of cyclic di-AMP mediates type I interferon responses during Chlamydia trachomatis infection. MBio  2013; 4: e00018– 00013. Google Scholar CrossRef Search ADS PubMed  Behr MA, Wilson MA, Gill WP et al.   Comparative genomics of BCG vaccines by whole-genome DNA microarray. Science  1999; 284: 1520– 3. Google Scholar CrossRef Search ADS PubMed  Bowie AG. Innate sensing of bacterial cyclic dinucleotides: more than just STING. Nat Immunol  2012; 13: 1137– 9. Google Scholar CrossRef Search ADS PubMed  Brosch R, Gordon SV, Garnier T et al.   Genome plasticity of BCG and impact on vaccine efficacy. P Natl Acad Sci USA  2007; 104: 5596– 601. Google Scholar CrossRef Search ADS   Burdette DL, Monroe KM, Sotelo-Troha K et al.   STING is a direct innate immune sensor of cyclic di-GMP. Nature  2011; 478: 515– 8. Google Scholar CrossRef Search ADS PubMed  Cole ST, Brosch R, Parkhill J et al.   Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature  1998; 393: 537– 44. Google Scholar CrossRef Search ADS PubMed  Collins AC, Cai H, Li T et al.   Cyclic GMP-AMP synthase is an innate immune DNA sensor for Mycobacterium tuberculosis. Cell Host Microbe  2015; 17: 820– 8. Google Scholar CrossRef Search ADS PubMed  Conrad WH, Osman MM, Shanahan JK et al.   Mycobacterial ESX-1 secretion system mediates host cell lysis through bacterium contact-dependent gross membrane disruptions. P Natl Acad Sci USA  2017; 114: 1371– 6. Google Scholar CrossRef Search ADS   Corrigan RM, Abbott JC, Burhenne H et al.   c-di-AMP is a new second messenger in Staphylococcus aureus with a role in controlling cell size and envelope stress. PLoS Pathog  2011; 7: e1002217. Google Scholar CrossRef Search ADS PubMed  Corrigan RM, Campeotto I, Jeganathan T et al.   Systematic identification of conserved bacterial c-di-AMP receptor proteins. P Natl Acad Sci USA  2013; 110: 9084– 9. Google Scholar CrossRef Search ADS   Cron LE, Stol K, Burghout P et al.   Two DHH subfamily 1 proteins contribute to pneumococcal virulence and confer protection against pneumococcal disease. Infect Immun  2011; 79: 3697– 710. Google Scholar CrossRef Search ADS PubMed  de Jonge MI, Pehau-Arnaudet G, Fretz MM et al.   ESAT-6 from Mycobacterium tuberculosis dissociates from its putative chaperone CFP-10 under acidic conditions and exhibits membrane-lysing activity. J Bacteriol  2007; 189: 6028– 34. Google Scholar CrossRef Search ADS PubMed  De Leon J, Jiang G, Ma Y et al.   Mycobacterium tuberculosis ESAT-6 exhibits a unique membrane-interacting activity that is not found in its ortholog from non-pathogenic Mycobacterium smegmatis. J Biol Chem  2012; 287: 44184– 91. Google Scholar CrossRef Search ADS PubMed  Desvignes L, Wolf AJ, Ernst JD. Dynamic roles of type I and type II IFNs in early infection with Mycobacterium tuberculosis. J Immunol  2012; 188: 6205– 15. Google Scholar CrossRef Search ADS PubMed  Dey B, Dey RJ, Cheung LS et al.   A bacterial cyclic dinucleotide activates the cytosolic surveillance pathway and mediates innate resistance to tuberculosis. Nat Med  2015; 21: 401– 6. Google Scholar CrossRef Search ADS PubMed  Dey RJ, Dey B, Zheng Y et al.   Inhibition of innate immune cytosolic surveillance by an M. tuberculosis phosphodiesterase. Nat Chem Biol  2017; 13: 210– 7. Google Scholar CrossRef Search ADS PubMed  Dorhoi A, Yeremeev V, Nouailles G et al.   Type I IFN signaling triggers immunopathology in tuberculosis-susceptible mice by modulating lung phagocyte dynamics. Eur J Immunol  2014; 44: 2380– 93. Google Scholar CrossRef Search ADS PubMed  Ebensen T, Libanova R, Schulze K et al.   Bis-(3΄,5΄)-cyclic dimeric adenosine monophosphate: strong Th1/Th2/Th17 promoting mucosal adjuvant. Vaccine  2011; 29: 5210– 20. Google Scholar CrossRef Search ADS PubMed  Flint JL, Kowalski JC, Karnati PK et al.   The RD1 virulence locus of Mycobacterium tuberculosis regulates DNA transfer in Mycobacterium smegmatis. P Natl Acad Sci USA  2004; 101: 12598– 603. Google Scholar CrossRef Search ADS   Giacomini E, Remoli ME, Gafa V et al.   IFN-beta improves BCG immunogenicity by acting on DC maturation. J Leukoc Biol  2009; 85: 462– 8. Google Scholar CrossRef Search ADS PubMed  Gordon SV, Brosch R, Billault A et al.   Identification of variable regions in the genomes of tubercle bacilli using bacterial artificial chromosome arrays. Mol Microbiol  1999; 32: 643– 55. Google Scholar CrossRef Search ADS PubMed  Griffiths JM, O’Neill AJ. Loss of function of the GdpP protein leads to joint beta-lactam/glycopeptide tolerance in Staphylococcus aureus. Antimicrob Agents Ch  2012; 56: 579– 81. Google Scholar CrossRef Search ADS   Groschel MI, Sayes F, Shin SJ et al.   Recombinant BCG expressing ESX-1 of Mycobacterium marinum combines low virulence with cytosolic immune signaling and improved TB protection. Cell Rep  2017; 18: 2752– 65. Google Scholar CrossRef Search ADS PubMed  Guerrero GG, Rangel-Moreno J, Islas-Trujillo S et al.   Successive intramuscular boosting with IFN-alpha protects Mycobacterium bovis BCG-vaccinated mice against M. lepraemurium infection. Biomed Res Int  2015; 2015: 414027. Google Scholar CrossRef Search ADS PubMed  Gundlach J, Herzberg C, Kaever V et al.   Control of potassium homeostasis is an essential function of the second messenger cyclic di-AMP in Bacillus subtilis. Sci Signal  2017; 10: eaal3011. Google Scholar CrossRef Search ADS PubMed  Hansen K, Prabakaran T, Laustsen A et al.   Listeria monocytogenes induces IFNbeta expression through an IFI16-, cGAS- and STING-dependent pathway. EMBO J  2014; 33: 1654– 66. Google Scholar CrossRef Search ADS PubMed  Houben D, Demangel C, van Ingen J et al.   ESX-1-mediated translocation to the cytosol controls virulence of mycobacteria. Cell Microbiol  2012; 14: 1287– 98. Google Scholar CrossRef Search ADS PubMed  Hsu T, Hingley-Wilson SM, Chen B et al.   The primary mechanism of attenuation of bacillus Calmette-Guerin is a loss of secreted lytic function required for invasion of lung interstitial tissue. P Natl Acad Sci USA  2003; 100: 12420– 5. Google Scholar CrossRef Search ADS   Jin L, Hill KK, Filak H et al.   MPYS is required for IFN response factor 3 activation and type I IFN production in the response of cultured phagocytes to bacterial second messengers cyclic-di-AMP and cyclic-di-GMP. J Immunol  2011; 187: 2595– 601. Google Scholar CrossRef Search ADS PubMed  Jin L, Waterman PM, Jonscher KR et al.   MPYS, a novel membrane tetraspanner, is associated with major histocompatibility complex class II and mediates transduction of apoptotic signals. Mol Cell Biol  2008; 28: 5014– 26. Google Scholar CrossRef Search ADS PubMed  Kamegaya T, Kuroda K, Hayakawa Y. Identification of a Streptococcus pyogenes SF370 gene involved in production of c-di-AMP. Nagoya J Med Sci  2011; 73: 49– 57. Google Scholar PubMed  Luo Y, Helmann JD. Analysis of the role of Bacillus subtilis sigma(M) in beta-lactam resistance reveals an essential role for c-di-AMP in peptidoglycan homeostasis. Mol Microbiol  2012; 83: 623– 39. Google Scholar CrossRef Search ADS PubMed  Mahairas GG, Sabo PJ, Hickey MJ et al.   Molecular analysis of genetic differences between Mycobacterium bovis BCG and virulent M. bovis. J Bacteriol  1996; 178: 1274– 82. Google Scholar CrossRef Search ADS PubMed  Majlessi L, Brosch R. Mycobacterium tuberculosis meets the cytosol: the role of cGAS in anti-mycobacterial immunity. Cell Host Microbe  2015; 17: 733– 5. Google Scholar CrossRef Search ADS PubMed  Manca C, Tsenova L, Bergtold A et al.   Virulence of a Mycobacterium tuberculosis clinical isolate in mice is determined by failure to induce Th1 type immunity and is associated with induction of IFN-alpha /beta. P Natl Acad Sci USA  2001; 98: 5752– 7. Google Scholar CrossRef Search ADS   Manca C, Tsenova L, Freeman S et al.   Hypervirulent M. tuberculosis W/Beijing strains upregulate type I IFNs and increase expression of negative regulators of the Jak-Stat pathway. J Interferon Cytokine Res  2005; 25: 694– 701. Google Scholar CrossRef Search ADS PubMed  Manikandan K, Sabareesh V, Singh N et al.   Two-step synthesis and hydrolysis of cyclic di-AMP in Mycobacterium tuberculosis. PLoS One  2014; 9: e86096. Google Scholar CrossRef Search ADS PubMed  Manzanillo PS, Shiloh MU, Portnoy DA et al.   Mycobacterium tuberculosis activates the DNA-dependent cytosolic surveillance pathway within macrophages. Cell Host Microbe  2012; 11: 469– 80. Google Scholar CrossRef Search ADS PubMed  Mehne FM, Gunka K, Eilers H et al.   Cyclic di-AMP homeostasis in Bacillus subtilis: both lack and high-level accumulation of the nucleotide are detrimental for cell growth. J Biol Chem  2013; 288: 2004– 17. Google Scholar CrossRef Search ADS PubMed  Novikov A, Cardone M, Thompson R et al.   Mycobacterium tuberculosis triggers host type I IFN signaling to regulate IL-1beta production in human macrophages. J Immunol  2011; 187: 2540– 7. Google Scholar CrossRef Search ADS PubMed  Pandey AK, Yang Y, Jiang Z et al.   NOD2, RIP2 and IRF5 play a critical role in the type I interferon response to Mycobacterium tuberculosis. PLoS Pathog  2009; 5: e1000500. Google Scholar CrossRef Search ADS PubMed  Parvatiyar K, Zhang Z, Teles RM et al.   The helicase DDX41 recognizes the bacterial secondary messengers cyclic di-GMP and cyclic di-AMP to activate a type I interferon immune response. Nat Immunol  2012; 13: 1155– 61. Google Scholar CrossRef Search ADS PubMed  Peng X, Zhang Y, Bai G et al.   Cyclic di-AMP mediates biofilm formation. Mol Microbiol  2016; 99: 945– 59. Google Scholar CrossRef Search ADS PubMed  Pozzi C, Waters EM, Rudkin JK et al.   Methicillin resistance alters the biofilm phenotype and attenuates virulence in Staphylococcus aureus device-associated infections. PLoS Pathog  2012; 8: e1002626. Google Scholar CrossRef Search ADS PubMed  Pym AS, Brodin P, Majlessi L et al.   Recombinant BCG exporting ESAT-6 confers enhanced protection against tuberculosis. Nat Med  2003; 9: 533– 9. Google Scholar CrossRef Search ADS PubMed  Romling U. Great times for small molecules: c-di-AMP, a second messenger candidate in Bacteria and Archaea. Sci Signal  2008; 1: pe39. Google Scholar CrossRef Search ADS PubMed  Rosenberg OS, Dovala D, Li X et al.   Substrates control multimerization and activation of the multi-domain ATPase motor of type VII secretion. Cell  2015; 161: 501– 12. Google Scholar CrossRef Search ADS PubMed  Sanchez MV, Ebensen T, Schulze K et al.   Intranasal delivery of influenza rNP adjuvanted with c-di-AMP induces strong humoral and cellular immune responses and provides protection against virus challenge. PLoS One  2014; 9: e104824. Google Scholar CrossRef Search ADS PubMed  Sauer JD, Sotelo-Troha K, von Moltke J et al.   The N-ethyl-N-nitrosourea-induced Goldenticket mouse mutant reveals an essential function of Sting in the in vivo interferon response to Listeria monocytogenes and cyclic dinucleotides. Infect Immun  2011; 79: 688– 94. Google Scholar CrossRef Search ADS PubMed  Schoggins JW, MacDuff DA, Imanaka N et al.   Pan-viral specificity of IFN-induced genes reveals new roles for cGAS in innate immunity. Nature  2014; 505: 691– 5. Google Scholar CrossRef Search ADS PubMed  Schwartz KT, Carleton JD, Quillin SJ et al.   Hyperinduction of host beta interferon by a Listeria monocytogenes strain naturally overexpressing the multidrug efflux pump MdrT. Infect Immun  2012; 80: 1537– 45. Google Scholar CrossRef Search ADS PubMed  Simeone R, Bobard A, Lippmann J et al.   Phagosomal rupture by Mycobacterium tuberculosis results in toxicity and host cell death. PLoS Pathog  2012; 8: e1002507. Google Scholar CrossRef Search ADS PubMed  Simeone R, Sayes F, Song O et al.   Cytosolic access of Mycobacterium tuberculosis: critical impact of phagosomal acidification control and demonstration of occurrence in vivo. PLoS Pathog  2015; 11: e1004650. Google Scholar CrossRef Search ADS PubMed  Skrnjug I, Rueckert C, Libanova R et al.   The mucosal adjuvant cyclic di-AMP exerts immune stimulatory effects on dendritic cells and macrophages. PLoS One  2014; 9: e95728. Google Scholar CrossRef Search ADS PubMed  Stanley SA, Johndrow JE, Manzanillo P et al.   The Type I IFN response to infection with Mycobacterium tuberculosis requires ESX-1-mediated secretion and contributes to pathogenesis. J Immunol  2007; 178: 3143– 52. Google Scholar CrossRef Search ADS PubMed  Tadmor K, Pozniak Y, Burg Golani T et al.   Listeria monocytogenes MDR transporters are involved in LTA synthesis and triggering of innate immunity during infection. Front Cell Infect Microbiol  2014; 4: 16. Google Scholar CrossRef Search ADS PubMed  Underwood AJ, Zhang Y, Metzger DW et al.   Detection of cyclic di-AMP using a competitive ELISA with a unique pneumococcal cyclic di-AMP binding protein. J Microbiol Methods  2014; 107: 58– 62. Google Scholar CrossRef Search ADS PubMed  van der Wel N, Hava D, Houben D. M. tuberculosis and M. leprae translocate from the phagolysosome to the cytosol in myeloid cells. Cell  2007; 129: 1287– 98. Google Scholar CrossRef Search ADS PubMed  Wassermann R, Gulen MF, Sala C et al.   Mycobacterium tuberculosis differentially activates cGAS- and inflammasome-dependent intracellular immune responses through ESX-1. Cell Host Microbe  2015; 17: 799– 810. Google Scholar CrossRef Search ADS PubMed  Watson RO, Bell SL, MacDuff DA et al.   The cytosolic sensor cGAS detects Mycobacterium tuberculosis DNA to induce type I interferons and activate autophagy. Cell Host Microbe  2015; 17: 811– 9. Google Scholar CrossRef Search ADS PubMed  Whiteley AT, Pollock AJ, Portnoy DA. The PAMP c-di-AMP is essential for Listeria monocytogenes growth in rich but not minimal media due to a toxic increase in (p)ppGpp. Cell Host Microbe  2015; 17: 788– 98. Google Scholar CrossRef Search ADS PubMed  Wiens KE, Ernst JD. The mechanism for type I interferon induction by Mycobacterium tuberculosis is bacterial strain-dependent. PLoS Pathog  2016; 12: e1005809. Google Scholar CrossRef Search ADS PubMed  Witte CE, Whiteley AT, Burke TP et al.   Cyclic di-AMP is critical for Listeria monocytogenes growth, cell wall homeostasis, and establishment of infection. MBio  2013; 4: e00282– 00213. Google Scholar CrossRef Search ADS PubMed  Witte G, Hartung S, Buttner K et al.   Structural biochemistry of a bacterial checkpoint protein reveals diadenylate cyclase activity regulated by DNA recombination intermediates. Mol Cell  2008; 30: 167– 78. Google Scholar CrossRef Search ADS PubMed  Woodward JJ, Iavarone AT, Portnoy DA. c-di-AMP secreted by intracellular Listeria monocytogenes activates a host type I interferon response. Science  2010; 328: 1703– 5. Google Scholar CrossRef Search ADS PubMed  Yamamoto T, Hara H, Tsuchiya K et al.   Listeria monocytogenes strain-specific impairment of the TetR regulator underlies the drastic increase in cyclic di-AMP secretion and beta interferon-inducing ability. Infect Immun  2012; 80: 2323– 32. Google Scholar CrossRef Search ADS PubMed  Yang J, Bai Y, Zhang Y et al.   Deletion of the cyclic di-AMP phosphodiesterase gene (cnpB) in Mycobacterium tuberculosis leads to reduced virulence in a mouse model of infection. Mol Microbiol  2014; 93: 65– 79. Google Scholar CrossRef Search ADS PubMed  © FEMS 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Pathogens and Disease Oxford University Press

Cyclic di-AMP-mediated interaction between Mycobacterium tuberculosis ΔcnpB and macrophages implicates a novel strategy for improving BCG vaccination

Loading next page...
 
/lp/ou_press/cyclic-di-amp-mediated-interaction-between-mycobacterium-tuberculosis-h5DhVCesNm
Publisher
John Wiley & Sons Inc.
Copyright
© FEMS 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
2049-632X
eISSN
2049-632X
D.O.I.
10.1093/femspd/fty008
Publisher site
See Article on Publisher Site

Abstract

Abstract Cyclic di-AMP (c-di-AMP) has been shown to play an important role in bacterial physiology and pathogen–host interactions. We previously reported that deletion of the sole c-di-AMP phosphodiesterase-encoding gene (cnpB) in Mycobacterium tuberculosis (Mtb) led to significant virulence attenuation. In this study, we found that ΔcnpB of M. bovisbacillus Calmette-Guerin (BCG) was unable to secrete c-di-AMP, which differs from Mtb ΔcnpB. We infected bone marrow-derived macrophages (BMDMs) with c-di-AMP-associated mutants generated from both Mtb and BCG. Our results showed that upon infection with Mtb ΔcnpB, BMDMs of wildtype mice secreted a large amount of interferon-β (IFN-β) post-infection similarly as we reported previously. In contrast, the response was less pronounced with BMDMs isolated from cGAS−/− mice and was nearly abolished with BMDMs prepared from STING−/− mice. Deletion of the region of difference 1 (RD1) locus in Mtb ΔcnpB did not alter the c-di-AMP secretion of ΔcnpB but eliminated the IFN-β production in the infected cells. In contrast, neither BCG ΔcnpB nor a recombinant BCG ΔcnpB with a pRD1 cosmid induced a type I interferon response. Interestingly, multiple studies have demonstrated that type I IFN enhances BCG’s immunity. Thus, amending BCG based on our findings might improve BCG vaccination. Mycobacterium tuberculosis, BCG, c-di-AMP, RD1, IFN-β, macrophage INTRODUCTION Tuberculosis (TB) remains a major threat to public health. However, the pathogenesis of the etiologic agent, Mycobacterium tuberculosis (Mtb), is still not fully understood. Mycobacterium bovisbacillus Calmette-Guerin (BCG) is the only available TB vaccine, but inadequately controls the TB epidemic. The molecular basis of BCG’s limitation in TB prevention is still elusive. Thus, it is essential to understand how Mtb and BCG interact with the host in order to develop a better vaccination strategy. Cyclic di-AMP (c-di-AMP) was first recognized as a new second messenger in 2008 (Witte et al.2008). This di-nucleotide is produced from ATP by diadenylate cyclases and cleaved to pApA or AMP by distinct c-di-AMP phosphodiesterases (Bai et al.2013; Corrigan et al.2013; Manikandan et al.2014). The diadenylate cyclase domain is widespread in bacteria and archaea, but has not been found in eukaryotic cells (Romling 2008; Corrigan et al.2013). Presence of active diadenylate cyclase has been experimentally demonstrated in several bacterial species (Witte et al.2008; Woodward, Iavarone and Portnoy 2010; Corrigan et al.2011; Kamegaya, Kuroda and Hayakawa 2011; Luo and Helmann 2012; Bai et al.2013; Barker et al.2013; Mehne et al.2013), and the enzyme is essential for bacterial viability of Bacillus subtilis (Luo and Helmann 2012, Mehne et al.2013), Staphylococcus aureus (Corrigan et al.2011), Streptococcus pneumoniae (Bai et al.2013) and Listeria monocytogenes at certain growth conditions (Woodward et al.2010; Witte et al.2013; Whiteley, Pollock and Portnoy 2015). Additionally, c-di-AMP is associated with bacterial infections (Cron et al.2011; Pozzi et al.2012; Schwartz et al.2012; Yamamoto et al.2012; Witte et al.2013; Dey et al.2015), peptidoglycan homeostasis (Corrigan et al.2011; Luo and Helmann 2012), antibiotic resistance (Banerjee et al.2010; Corrigan et al.2011; Griffiths and O’Neill 2012; Luo and Helmann 2012; Witte et al.2013) and bacterial morphology and/or biofilm formation (Corrigan et al.2011; Pozzi et al.2012; Bai et al.2013; Yang et al.2014; Peng et al.2016). Furthermore, c-di-AMP produced by L. monocytogenes, Chlamydia trachomatis and Mtb promotes a type I interferon (IFNα/β) response mediated by host effectors STING and DDX41 through an IRF-3-regulated signaling pathway (Woodward et al.2010; Burdette et al.2011; Jin et al.2011; Sauer et al.2011; Bowie 2012; Parvatiyar et al.2012; Barker et al.2013; Yang et al.2014; Dey et al.2015). We previously identified DisA and CnpB as diadenylate cyclase and c-di-AMP phosphodiesterase, respectively, in Mtb (Bai et al.2012; Yang et al.2014) and demonstrated that deletion of disA abolished Mtb's c-di-AMP production. More interestingly, deletion of cnpB elevated bacterial c-di-AMP accumulation and secretion, induced a strong host type I IFN response and significantly attenuated the virulence in a mouse infection model (Yang et al.2014; Dey et al.2017). A similar observation was reported with a disA-overexpressing Mtb strain (Dey et al.2015). Together, these observations suggest that elevated c-di-AMP levels play a protective role during Mtb infection. It has been shown that both BCG and Mtb ΔRD1 (region of difference 1) fail to trigger a relevant host type I IFN response compared to Mtb wildtype (WT) (Novikov et al.2011; Dey et al.2015). Interestingly, addition of type I IFN improves BCG’s immunogenicity in vitro (Giacomini et al.2009), and boosting BCG vaccination with type I IFN protects mice against Mycobacterium lepraemurium infection (Guerrero et al.2015). Additionally, a recombinant BCG vaccine with ESAT-6 secretion complex-1 (ESX-1) of Mycobacterium marinum induces type I IFN and enhances T cell immunity against Mtb challenge (Groschel et al.2017). These observations implicate that the defect in type I IFN induction is one of the reasons for the failure of BCG. It has been well established that Mtb ESX-1 is required for the type I IFN response both in vitro and in vivo (Stanley et al.2007). ESX-1 is partially encoded by genes in RD1, a 9.5-kb locus in the Mtb genome but missing in BCG (Mahairas et al.1996; Behr et al.1999; Gordon et al.1999). RD1 contains nine genes, including esxA and esxB, which encode for ESAT-6 and CFP-10, respectively. Earlier reports support that ESAT-6 is essential for rupturing the phagosomal membrane and releasing the mycobacterial components into the cytosol of the host cell (van der Wel, Hava and Houben 2007; Houben et al.2012; Simeone et al.2012, 2015). However, a recent study suggests that ESX-1 but not ESAT-6 is needed to permeabilize the phagosomal membrane (Conrad et al.2017). Nonetheless, it is unknown whether this process is required for the phagosome-to-cytosol translocation of c-di-AMP as a small molecule produced by Mtb. In the current study, we demonstrate that c-di-AMP can be secreted by Mtb ΔcnpB but not BCG ΔcnpB. Both c-di-AMP secretion machinery and RD1 are required for Mtb to induce the c-di-AMP-mediated host response by the infected macrophages, which is observed from a type I IFN response in infected macrophages. Since type I IFN improves BCG’s immunogenicity, amending BCG to induce type I IFN response based on the c-di-AMP signaling network of Mtb will likely improve BCG’s efficacy. MATERIALS AND METHODS Bacterial strains and growth conditions The bacterial strains used in this study are listed in Table 1. Mtb H37Rv, BCG (Pasteur strain; Trudeau Institute) and their derivatives were grown at 37°C in mycomedium (Middlebrook 7H9 medium (BD) supplemented with 0.5% glycerol, 10% oleic acid-albumin-dextrose-catalase (OADC), and 0.05% Tween-80), Sauton's medium (Bai, Schaak and McDonough 2009) or Middlebrook 7H10 agar supplemented with 10% OADC and 0.01% cycloheximide. Fresh cultures were inoculated from frozen stocks for every experiment. Cultures were grown in tissue culture flasks standing with ambient air. Escherichia coli strains were grown in Luria-Bertani (LB) broth or on LB agar plates. Mycobacterium smegmatis mc2155 was grown in mycomedium. All cultures were grown at 37°C, except that M. smegmatis was grown at 30°C. Kanamycin at 25 μg/ml, hygromycin at 50 μg/ml or zeocin at 100 μg/ml were added when necessary. Table 1. Bacterial strains and plasmids used in this study. Strain or plasmid  Descriptiona  Source  Bacterial strain  Mycobacterium tuberculosis   H37Rv strain  WT  ATCC   Mt008  ΔdisA; Hygr  (Yang et al.2014)   Mt012  Mt008::pGB086; Hygr; Kanr  (Yang et al.2014)   Mt013  ΔcnpB; Hygr  (Yang et al., 2014)   Mt016  Mt013:pGB120; Hygr; Kanr  (Yang et al.2014)   Mt017  Unmarked ΔcnpB  This study   Mt020  ΔdisAΔcnpB; Hygr  This study   Mt023  ΔRD1; Hygr  This study   Mt024  ΔcnpBΔRD1; Hygr  This study  Mycobacterium bovis BCG   Pasteur strain  WT  Trudeau Institute   ΔdisA  ΔdisA; Hygr  This study   mb042  ΔdisA::pGB086; Hygr; Kanr  This study   mb069  ΔcnpB; Hygr  This study   mb072  mb069::pGB120; Hygr; Kanr  This study   mb076  Unmarked ΔcnpB  This study   mb092  mb076(pGB192); Hygr; Kanr  This study   mb096  mb076::pMBC1260; Kanr  This study   mb097  mb076::pGB218; Hygr; Kanr  This study   mb098  ΔdisAΔcnpB; Hygr  This study   Mt030  BCG WT::pRD1; Hygr  This study   Mt031  mb076::pRD1; Hygr  This study  Phage   pMBC1206  Deletion of disA  (Yang et al.2014)   pGB139  Deletion of cnpB  (Yang et al.2014)   RD1  Deletion of RD1  (Hsu et al.2003)  Plasmid   pMBC283  Multiple-copy expression vector with tuf promoter; Kanr; Hygr; Apr  (Bai et al.2011)   pMBC1260  Single-copy expression vector with Rv0805 promoter; Kanr; Apr  (Yang et al.2014)   pGB086  Complementation of ΔdisA; Kanr; Apr  (Yang et al.2014)   pGB120  Complementation of ΔcnpB; Kanr; Apr  (Yang et al.2014)   pGB192  pMBC283 carrying Rv1877 ORF; Kanr; Hygr; Apr  This study   pGB218  pMBC1260 carrying Rv3877 ORF; Kanr; Apr  This study   pYO11  Unmark the hyg marker in the mutants; Zeor  (Rosenberg et al.2015)   pRD1  Cosmid flanking RD1  (Pym et al.2003)  Strain or plasmid  Descriptiona  Source  Bacterial strain  Mycobacterium tuberculosis   H37Rv strain  WT  ATCC   Mt008  ΔdisA; Hygr  (Yang et al.2014)   Mt012  Mt008::pGB086; Hygr; Kanr  (Yang et al.2014)   Mt013  ΔcnpB; Hygr  (Yang et al., 2014)   Mt016  Mt013:pGB120; Hygr; Kanr  (Yang et al.2014)   Mt017  Unmarked ΔcnpB  This study   Mt020  ΔdisAΔcnpB; Hygr  This study   Mt023  ΔRD1; Hygr  This study   Mt024  ΔcnpBΔRD1; Hygr  This study  Mycobacterium bovis BCG   Pasteur strain  WT  Trudeau Institute   ΔdisA  ΔdisA; Hygr  This study   mb042  ΔdisA::pGB086; Hygr; Kanr  This study   mb069  ΔcnpB; Hygr  This study   mb072  mb069::pGB120; Hygr; Kanr  This study   mb076  Unmarked ΔcnpB  This study   mb092  mb076(pGB192); Hygr; Kanr  This study   mb096  mb076::pMBC1260; Kanr  This study   mb097  mb076::pGB218; Hygr; Kanr  This study   mb098  ΔdisAΔcnpB; Hygr  This study   Mt030  BCG WT::pRD1; Hygr  This study   Mt031  mb076::pRD1; Hygr  This study  Phage   pMBC1206  Deletion of disA  (Yang et al.2014)   pGB139  Deletion of cnpB  (Yang et al.2014)   RD1  Deletion of RD1  (Hsu et al.2003)  Plasmid   pMBC283  Multiple-copy expression vector with tuf promoter; Kanr; Hygr; Apr  (Bai et al.2011)   pMBC1260  Single-copy expression vector with Rv0805 promoter; Kanr; Apr  (Yang et al.2014)   pGB086  Complementation of ΔdisA; Kanr; Apr  (Yang et al.2014)   pGB120  Complementation of ΔcnpB; Kanr; Apr  (Yang et al.2014)   pGB192  pMBC283 carrying Rv1877 ORF; Kanr; Hygr; Apr  This study   pGB218  pMBC1260 carrying Rv3877 ORF; Kanr; Apr  This study   pYO11  Unmark the hyg marker in the mutants; Zeor  (Rosenberg et al.2015)   pRD1  Cosmid flanking RD1  (Pym et al.2003)  aKanr, kanamycin resistance; Hygr, hygromycin resistance; Apr, ampicillin resistance; Zeor, zeocin resistance. View Large Table 1. Bacterial strains and plasmids used in this study. Strain or plasmid  Descriptiona  Source  Bacterial strain  Mycobacterium tuberculosis   H37Rv strain  WT  ATCC   Mt008  ΔdisA; Hygr  (Yang et al.2014)   Mt012  Mt008::pGB086; Hygr; Kanr  (Yang et al.2014)   Mt013  ΔcnpB; Hygr  (Yang et al., 2014)   Mt016  Mt013:pGB120; Hygr; Kanr  (Yang et al.2014)   Mt017  Unmarked ΔcnpB  This study   Mt020  ΔdisAΔcnpB; Hygr  This study   Mt023  ΔRD1; Hygr  This study   Mt024  ΔcnpBΔRD1; Hygr  This study  Mycobacterium bovis BCG   Pasteur strain  WT  Trudeau Institute   ΔdisA  ΔdisA; Hygr  This study   mb042  ΔdisA::pGB086; Hygr; Kanr  This study   mb069  ΔcnpB; Hygr  This study   mb072  mb069::pGB120; Hygr; Kanr  This study   mb076  Unmarked ΔcnpB  This study   mb092  mb076(pGB192); Hygr; Kanr  This study   mb096  mb076::pMBC1260; Kanr  This study   mb097  mb076::pGB218; Hygr; Kanr  This study   mb098  ΔdisAΔcnpB; Hygr  This study   Mt030  BCG WT::pRD1; Hygr  This study   Mt031  mb076::pRD1; Hygr  This study  Phage   pMBC1206  Deletion of disA  (Yang et al.2014)   pGB139  Deletion of cnpB  (Yang et al.2014)   RD1  Deletion of RD1  (Hsu et al.2003)  Plasmid   pMBC283  Multiple-copy expression vector with tuf promoter; Kanr; Hygr; Apr  (Bai et al.2011)   pMBC1260  Single-copy expression vector with Rv0805 promoter; Kanr; Apr  (Yang et al.2014)   pGB086  Complementation of ΔdisA; Kanr; Apr  (Yang et al.2014)   pGB120  Complementation of ΔcnpB; Kanr; Apr  (Yang et al.2014)   pGB192  pMBC283 carrying Rv1877 ORF; Kanr; Hygr; Apr  This study   pGB218  pMBC1260 carrying Rv3877 ORF; Kanr; Apr  This study   pYO11  Unmark the hyg marker in the mutants; Zeor  (Rosenberg et al.2015)   pRD1  Cosmid flanking RD1  (Pym et al.2003)  Strain or plasmid  Descriptiona  Source  Bacterial strain  Mycobacterium tuberculosis   H37Rv strain  WT  ATCC   Mt008  ΔdisA; Hygr  (Yang et al.2014)   Mt012  Mt008::pGB086; Hygr; Kanr  (Yang et al.2014)   Mt013  ΔcnpB; Hygr  (Yang et al., 2014)   Mt016  Mt013:pGB120; Hygr; Kanr  (Yang et al.2014)   Mt017  Unmarked ΔcnpB  This study   Mt020  ΔdisAΔcnpB; Hygr  This study   Mt023  ΔRD1; Hygr  This study   Mt024  ΔcnpBΔRD1; Hygr  This study  Mycobacterium bovis BCG   Pasteur strain  WT  Trudeau Institute   ΔdisA  ΔdisA; Hygr  This study   mb042  ΔdisA::pGB086; Hygr; Kanr  This study   mb069  ΔcnpB; Hygr  This study   mb072  mb069::pGB120; Hygr; Kanr  This study   mb076  Unmarked ΔcnpB  This study   mb092  mb076(pGB192); Hygr; Kanr  This study   mb096  mb076::pMBC1260; Kanr  This study   mb097  mb076::pGB218; Hygr; Kanr  This study   mb098  ΔdisAΔcnpB; Hygr  This study   Mt030  BCG WT::pRD1; Hygr  This study   Mt031  mb076::pRD1; Hygr  This study  Phage   pMBC1206  Deletion of disA  (Yang et al.2014)   pGB139  Deletion of cnpB  (Yang et al.2014)   RD1  Deletion of RD1  (Hsu et al.2003)  Plasmid   pMBC283  Multiple-copy expression vector with tuf promoter; Kanr; Hygr; Apr  (Bai et al.2011)   pMBC1260  Single-copy expression vector with Rv0805 promoter; Kanr; Apr  (Yang et al.2014)   pGB086  Complementation of ΔdisA; Kanr; Apr  (Yang et al.2014)   pGB120  Complementation of ΔcnpB; Kanr; Apr  (Yang et al.2014)   pGB192  pMBC283 carrying Rv1877 ORF; Kanr; Hygr; Apr  This study   pGB218  pMBC1260 carrying Rv3877 ORF; Kanr; Apr  This study   pYO11  Unmark the hyg marker in the mutants; Zeor  (Rosenberg et al.2015)   pRD1  Cosmid flanking RD1  (Pym et al.2003)  aKanr, kanamycin resistance; Hygr, hygromycin resistance; Apr, ampicillin resistance; Zeor, zeocin resistance. View Large Construction and complementation of mutants in Mtb and BCG BCG ΔdisA and ΔcnpB were generated by using phages prepared with pMBC1206 and pGB132 (Table 1), respectively, which were previously used to delete disA and cnpB in Mtb by using a phage-based approach (Yang et al.2014). The mutants were selected with plates containing 50 μg/ml hygromycin and screened by PCR using primers as listed in Table S1, Supporting Information. Subsequently, ΔdisA and ΔcnpB of BCG were transformed with pGB086 and pGB120, respectively, for complementation similar to what we have performed to complement the relevant mutants in Mtb (Yang et al.2014). The hygromycin resistance marker in both ΔcnpB of Mtb and BCG was unmarked using pYO11 (provided by Dr Jeffery Cox) (Rosenberg et al.2015). Briefly, hygromycin-resistant ΔcnpB was transformed with pYO11 and selected with Middlebrook 7H10 plates containing 100 μg/ml zeocin. Zeocin-resistant transformants were passaged in mycomedium without antibiotics for at least three times to cure pYO11. Bacteria sensitive to both hygromycin and zeocin were confirmed as unmarked ΔcnpB by PCR using primers JY213 and JY214. The PCR products from the unmarked strains exhibit a significantly reduced size compared to those from the WT and the hygromycin-resistant ΔcnpB strains. ΔdisAΔcnpB of Mtb and BCG were then generated from the unmarked ΔcnpB strains by transducing the phage for deletion of disA similarly as we previously described (Yang et al.2014). Mtb ΔRD1 and ΔcnpBΔRD1 were generated using the phage-based method (Yang et al.2014). Briefly, a phage used to delete RD1 (provided by Dr William Jacobs, Jr) (Hsu et al.2003) was transduced into WT and the unmarked ΔcnpB mutant of Mtb. Replacement of RD1 with a hygromycin resistance marker in both genetic backgrounds was confirmed by PCR with primers JY407, JY002, JY003 and JY408 (Table S1, Supporting Information). Deletion of RD1 was further verified by PCR amplification of an esxA fragment using primers KM1325 and KM1326 (Table S1, Supporting Information). WT control exhibits a 250-bp product, whereas the mutants lack this fragment. Transformation of BCG strains with cosmid pRD1 BCG WT and ΔcnpB were transformed with cosmid pRD1 (provided by Dr Keith Derbyshire) using electroporation at 2.5 kV, 25 μF and 1000 Ω (Flint et al.2004). This cosmid harbors a 32-kb segment (Rv3861-Rv3885) flanking the 9.5-kb RD1 locus (Rv3871-Rv3879c) (Pym et al.2003). The recombinant strains were selected on a Middlebrook 7H10 plate containing 50 μg/ml hygromycin and were verified by PCR using primers specific to whiB6 (Rv3862), esxA (Rv3875) and Rv3879c (Table S1, Supporting Information). DNA isolated from BCG WT was used as a negative control for the PCR reaction. Detection of c-di-AMP c-di-AMP was extracted and measured by using an enzyme-linked immunosorbent assay (ELISA) as we described earlier (Underwood et al.2014; Yang et al.2014). Briefly, each bacterial strain was grown in 5 ml of Sauton's medium for 7 days. The optical density at 600 nm (OD600) was determined. Bacteria were harvested by centrifugation, and the supernatant was used for secreted c-di-AMP determination. Each bacterial pellet was resuspended in 0.5 ml 50 mM Tris-HCl (pH 8.0) and disrupted with 1 mm glass beads using a mini bead-beater (BioSpec). Bacterial debris was removed by centrifugation and the supernatant was boiled for 5 min and subsequently used as the bacterial c-di-AMP sample. c-di-AMP levels were quantified by using the competitive ELISA (Underwood et al.2014; Yang et al.2014). Results of both bacterial and secreted samples were normalized by respective OD600 readings. Preparation and infection of BMDMs This study was carried out in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Institutional Animal Care and Use Committee of Albany Medical College (Permit Number: 13-12006). WT mice were purchased from Taconic, and STING−/− and cGAS−/− mice were provided by Drs Lei Jin (Jin et al.2008) and Herbert Virgin (Schoggins et al.2014), respectively. The knockout mice were bred and maintained in the Animal Resources Facility at Albany Medical College. Bone marrow-derived macrophages (BMDMs) were generated as previously described (Jin et al.2011), which was approved by the Albany Medical College IACUC. Briefly, bone marrow cells were harvested from mouse femurs and cultured in complete RPMI-1640 supplemented 10% fetal bovine serum (FBS), 2 mM L-glutamine, 1 mM sodium pyruvate, 1× HEPES [4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid], 1× non-essential amino acids, 50 μM β-mercaptoethanol, 100 U/ml penicillin, 100 μg/ml streptomycin and 20 ng/ml M-CSF. All reagents were purchased from Gibco. BMDM cells were seeded in 12-well plates at 5 × 105 cells/ml and 0.5 ml/well in the supplemented RPMI-1640 medium without antibiotics. Cells were infected with Mtb, BCG or their derivatives at a multiplicity of infection (MOI) of 2:1. At 4 h post-infection, infection medium was replaced with fresh complete RPMI-1640 medium after washing thrice with PBS. At 1 and 2 days post-infection, culture supernatant was collected to determine cytokine production. Cytokine detection IFN-β in tissue culture supernatant was detected using ELISA with a Verikine-HS Mouse IFN-β serum ELISA kit (PBL Assay Science) following the manufacturer's manual similarly as we reported earlier (Yang et al.2014). Statistical analysis Comparisons of the c-di-AMP levels between WT, BCG and their derivatives were analyzed using a two-tailed t-test performed using Prism 5 (GraphPad Software). IFN-β levels were analyzed using a two-way ANOVA (GraphPad Software). P values of <0.05 were considered to be statistically significant. For IFN-β levels, each bacterial mutant was compared to the parental strain in infection of the BMDMs isolated from the same mouse. RESULTS DisA is likely the sole diadenylate cyclase in Mtb Increasing evidence has revealed that c-di-AMP participates in multiple critical intracellular bioprocesses. It has been reported that diadenylate cyclase-encoding genes are essential for several Gram-positive bacteria including S. aureus (Corrigan et al.2011), S. pneumoniae (Bai et al.2013) and L. monocytogenes (Woodward et al.2010; Witte et al.2013), each of which possesses a single diadenylate cyclase. A recent study has revealed that diadenylate cyclase is conditionally essential for L. monocytogenes when the bacterium is grown in rich media or within host cells, but not in minimal media that do not trigger (p)ppGpp synthesis (Whiteley et al.2015). B. subtilis encodes three diadenylate cyclases, which can be individually disrupted but cannot be deleted together when bacteria were grown in regular cultural media (Luo and Helmann 2012; Mehne et al.2013). A triple mutant of B. subtilis diandenylate cyclases was generated only in low K+ media (Gundlach et al.2017). DisA is the only known protein that possesses a diadenylate cyclase domain in Mtb. Surprisingly, we could easily delete disA from the Mtb genome, and the mutant grew well in various mycobacterial culture media and intracellularly (Yang et al.2014). We hypothesized that either the role of c-di-AMP in Mtb differs from that in Gram-positive bacteria or Mtb encodes more than one diadenylate cyclase, similar to B. subtilis. We generated a ΔdisAΔcnpB mutant in Mtb, which was unable to accumulate detectable c-di-AMP, similar to ΔdisA (Fig. 1A). This result indicates that DisA is very likely the sole diadenylate cyclase in this pathogen and that c-di-AMP is not required for the viability of Mtb. The role of c-di-AMP in Mtb is distinct from that in the Gram-positive bacteria that have been characterized for c-di-AMP to date. Figure 1. View largeDownload slide Determination of bacterial (A and B) and secreted (C and D) c-di-AMP. WT and the indicated derivatives of Mtb (A and C) and BCG (B and D), respectively, were grown in Sauton's broth for 7 days. Bacteria were harvested by centrifugation. The c-di-AMP levels in the supernatant (secreted) and in the lysate prepared from the pellet (bacterial) were determined using ELISA. The data shown are the means of three independent experiments. The error bars indicate the standard errors of mean (SEMs). *P < 0.05; **P < 0.01; ***P < 0.001. P values are statistical analyses resulted from a two-tailed t-test. Note the detectable limit of the assay is ∼10 nM. ΔdisA/C and ΔcnpB/C are the complemented strains for ΔdisA and ΔcnpB, respectively. Figure 1. View largeDownload slide Determination of bacterial (A and B) and secreted (C and D) c-di-AMP. WT and the indicated derivatives of Mtb (A and C) and BCG (B and D), respectively, were grown in Sauton's broth for 7 days. Bacteria were harvested by centrifugation. The c-di-AMP levels in the supernatant (secreted) and in the lysate prepared from the pellet (bacterial) were determined using ELISA. The data shown are the means of three independent experiments. The error bars indicate the standard errors of mean (SEMs). *P < 0.05; **P < 0.01; ***P < 0.001. P values are statistical analyses resulted from a two-tailed t-test. Note the detectable limit of the assay is ∼10 nM. ΔdisA/C and ΔcnpB/C are the complemented strains for ΔdisA and ΔcnpB, respectively. BCG differs from Mtb in c-di-AMP secretion In order to explore c-di-AMP-mediated biological differences between Mtb and BCG, we generated ΔdisA, ΔcnpB and ΔdisAΔcnpB mutants in BCG and compared these mutants with WT BCG as well as the Mtb strains in c-di-AMP production and secretion. We detected similar levels of c-di-AMP within BCG strains compared to those within the relevant Mtb strains (Fig. 1A and B). In contrast, secreted c-di-AMP was detected from Mtb ΔcnpB, but not BCG ΔcnpB (Fig. 1C and D). This result indicates that both DisA and CnpB are functional in BCG; however, the c-di-AMP secretion machinery of BCG is defective, which differs from Mtb. In L. monocytogenes, both MdrM and MdrT are capable of secreting c-di-AMP (Woodward et al.2010; Schwartz et al.2012; Tadmor et al.2014). These proteins are multidrug efflux pumps. We compared all the efflux pump-encoding genes of Mtb to those in the BCG genome by sequence alignment. Most efflux pumps of BCG are nearly identical in amino acid sequence to their orthologs in Mtb except the proteins encoded by Rv3877 and Rv1877 (Cole et al.1998; Brosch et al.2007). Rv3877 is located in the RD1 region (Behr et al.1999), which is absent in the BCG genome. In order to determine whether RD1 plays a role in c-di-AMP secretion, we deleted RD1 in both WT and ΔcnpB of Mtb and examined c-di-AMP secretion by the mutants. As a result, these mutants displayed no differences in c-di-AMP accumulation or secretion compared to their parental strains (Fig. S1A, Supporting Information). Consistently, we transformed either a plasmid expressing Rv3877 (not shown) or a cosmid flanking RD1 (pRD1) into BCG ΔcnpB (Fig. S1B, Supporting Information), and these recombinant strains were unable to secrete c-di-AMP. These observations suggest that Rv3877 does not play a role in c-di-AMP secretion. Additionally, the Rv1877 ortholog in BCG introduces a frameshift due to a single base insertion (*-c), which splits Rv1877 into two segments (Fig. S2, Supporting Information). We expressed Mtb Rv1877 in BCG ΔcnpB but could not detect secreted c-di-AMP from this recombinant strain (not shown), indicating that either Mtb utilizes other efflux pumps or it has a different mechanism for c-di-AMP secretion compared to L. monocytogenes. Both c-di-AMP secretion machinery and RD1 are needed in the c-di-AMP-mediated type I IFN response It is known that BCG and Mtb RD1 mutant are defective in inducing a type I IFN response (Giacomini et al.2009; Dey et al.2015). Interestingly, it has been shown that type I IFN improves BCG’s immunogenicity (Giacomini et al.2009), and boosting BCG vaccination with type I IFN provides better protection in mice challenged with M. lepraemurium (Guerrero et al.2015). Therefore, modifying BCG to be able to induce optimal levels of type I IFN response might improve BCG’s efficacy. In this study, we explored the molecular basis for BCG’s failure in the type I IFN response. It has been well established that c-di-AMP produced by intracellular bacteria stimulates a type I IFN response (Woodward et al.2010; Sauer et al.2011; Schwartz et al.2012; Barker et al.2013; Archer, Durack and Portnoy 2014; Dey et al.2015). Our previous study revealed that Mtb ΔcnpB secretes a large amount of c-di-AMP and induces high levels of IFN-β during a macrophage infection (Yang et al.2014), which is consistent with a recent report (Dey et al.2017) and the observation from a disA-overexpressing Mtb strain (Dey et al.2015). Several recent reports show that during Mtb infection, Mtb DNA rather than c-di-AMP is a major trigger for the type I IFN response, which is also cGAS-dependent (Collins et al.2015; Wassermann et al.2015; Watson et al.2015). Therefore, the IFN-β response we previously observed is either due to enhanced DNA secretion or partially contributed by the large amount of c-di-AMP. Furthermore, it has been shown that RD1 is required for releasing mycobacterial components from the phagosome into the cytosol (Conrad et al.2017). However, it is unknown whether Mtb-secreted c-di-AMP can be directly released from the phagosome or it is dependent upon RD1. Thus, we sought to address two major questions: (i) Is the elevated type I IFN response to Mtb ΔcnpB infection mediated by DNA or c-di-AMP? and (ii) Is RD1 required for the ΔcnpB-induced type I IFN response? In this study, we isolated BMDMs from WT, STING−/− and cGAS−/− mice and examined the IFN-β production by the cells infected with Mtb, BCG and their derivatives, respectively. Similar results were observed between the samples collected at 1 day (Fig. 2) and 2 day (not shown) post-infection. From BMDMs of WT mice, Mtb ΔdisA stimulated less IFN-β than the WT strain (Fig. 2A). In contrast, Mtb ΔcnpB induced a considerably larger amount of IFN-β than WT, similarly as we reported earlier (Yang et al.2014). Moreover, ΔdisAΔcnpB failed to trigger a type I response (Fig. 2A). These results indicate an association between c-di-AMP and the type I IFN response. By comparing infected BMDMs isolated from WT, STING−/− and cGAS−/− mice, we found that IFN-β production induced by Mtb WT and ΔcnpB was nearly abolished in the infected STING−/− cells, and cGAS−/− cells exhibited reduced IFN-β production in a c-di-AMP-dependent pattern (Fig. 2). This result is consistent with a recent report by Dey et al. using a transposon mutant of cnpB (Dey et al.2017). Overall, our results indicate that DNA is a major player of the Mtb-induced type I IFN response, which is consistent with several recent reports (Collins et al.2015; Wassermann et al.2015; Watson et al.2015). Meanwhile, c-di-AMP also contributes directly to a substantial type I IFN response during Mtb infection, which is also coherent with an earlier observation (Dey et al.2015). Figure 2. View largeDownload slide IFN-β secretion by BMDMs isolated from WT (A), STING−/− (B) and cGAS−/− (C) mice. BMDMs were infected with the indicated bacterial strains for 4 h. After thoroughly washing, cells were incubated for 24 h, and the supernatant was then collected to determine IFN-β. The data shown are the means of three independent experiments. The error bars indicate the SEMs. *P < 0.05; **P < 0.01; ***P < 0.001. P values are statistical analyses resulted from a two-tailed t-test to compare each bacterial mutant to the parental strain in infection of the BMDMs isolated from the same mouse. Figure 2. View largeDownload slide IFN-β secretion by BMDMs isolated from WT (A), STING−/− (B) and cGAS−/− (C) mice. BMDMs were infected with the indicated bacterial strains for 4 h. After thoroughly washing, cells were incubated for 24 h, and the supernatant was then collected to determine IFN-β. The data shown are the means of three independent experiments. The error bars indicate the SEMs. *P < 0.05; **P < 0.01; ***P < 0.001. P values are statistical analyses resulted from a two-tailed t-test to compare each bacterial mutant to the parental strain in infection of the BMDMs isolated from the same mouse. RD1 has been shown to be essential for Mtb to permeabilize the phagosomal membrane of the infected cells (de Jonge et al.2007; Stanley et al.2007; De Leon et al.2012). In this study, we explored whether RD1 is required for the Mtb ΔcnpB-induced type I IFN response, as c-di-AMP is a small molecule, which may penetrate phagosomal membrane directly. Remarkably, IFN-β production was dramatically reduced by infection with ΔRD1 generated in either the WT or the ΔcnpB genetic background (Fig. 2), even though ΔcnpBΔRD1 still secretes c-di-AMP as potently as the ΔcnpB mutant (Fig. S1, Supporting Information). This result indicates that RD1 is required for the Mtb ΔcnpB-induced type I IFN response. With BCG strains, BCG WT failed to trigger an IFN-β response, which is consistent with other reports (Giacomini et al.2009; Dey et al.2015). Furthermore, neither BCG ΔcnpB nor BCG ΔcnpB bearing the pRD1 cosmid elicited a type I IFN response, suggesting that the absence of RD1 is not the only reason for BCG’s defect in the type I IFN induction. Collectively, we conclude that both the c-di-AMP secretion machinery and RD1 are required for c-di-AMP-mediated response in the infected macrophages. DISCUSSION In this study, we demonstrate that both the c-di-AMP secretion machinery and RD1 are required for Mtb to induce the host type I IFN response. We propose a model for this response based on our observations together with several earlier reports (Manzanillo et al.2012; Collins et al.2015; Wassermann et al.2015; Watson et al.2015; Conrad et al.2017) (Fig. 3). In this model, c-di-AMP is secreted from Mtb into an infected phagosome by a currently unknown mechanism. Proteins encoded by the ESX-1 facilitate c-di-AMP releasing from the phagosome into the cytosol of the infected macrophage. Meanwhile, Mtb DNA is also released through the permeabilized phagosome into the cytosol and activates cGAS to produce cyclic GMP-AMP (cGAMP). Both c-di-AMP and cGAMP are then sensed by STING, which induces the type I IFN response. Figure 3. View largeDownload slide Working model for the host type I IFN response to Mtb c-di-AMP. In a phagosome of the infected macrophage, Mtb secretes c-di-AMP with a currently unknown mechanism. In the meantime, the phagosomal membrane was ruptured by an ESX-1-associated mechanism. Mtb components including DNA and c-di-AMP are released into the cytosol. Mtb DNA activates cGAS to produce cGAMP. Both cGAMP and c-di-AMP are detected by STING, which subsequently induces the host type I IFN response. Figure 3. View largeDownload slide Working model for the host type I IFN response to Mtb c-di-AMP. In a phagosome of the infected macrophage, Mtb secretes c-di-AMP with a currently unknown mechanism. In the meantime, the phagosomal membrane was ruptured by an ESX-1-associated mechanism. Mtb components including DNA and c-di-AMP are released into the cytosol. Mtb DNA activates cGAS to produce cGAMP. Both cGAMP and c-di-AMP are detected by STING, which subsequently induces the host type I IFN response. Three mycobacterial components have been shown to induce a type I IFN response: (i) c-di-AMP directly detected by STING (Yang et al.2014; Dey et al.2015); (ii) DNA through a cGAS—STING signaling pathway (Manzanillo et al.2012; Collins et al.2015; Wassermann et al.2015; Watson et al.2015); and (iii) peptidoglycan through an NOD2 signaling pathway (Pandey et al.2009). We have demonstrated that the elevated IFN-β production induced by Mtb ΔcnpB is unlikely resulted from the NOD2 signaling pathway by using BMDMs isolated from NOD2−/− mice (Yang et al.2014). In this study, we show that the majority of IFN-β induced by Mtb ΔcnpB is elicited by the DNA-sensing pathway through cGAS, indicating that ΔcnpB enhanced Mtb DNA release after 1 and 2 days post-infection, which is consistent with the recent reports regarding the host response to mycobacterial DNA (Collins et al.2015; Wassermann et al.2015; Watson et al.2015) and c-di-AMP-mediated type I IFN induction by L. monocytogenes (Hansen et al.2014). Additionally, Mtb ΔcnpB also induces a small amount of IFN-β in a STING-dependent but cGAS-independent manner, suggesting that this proportion of IFN-β is directly stimulated by c-di-AMP. As we pointed out in an earlier report, secreted c-di-AMP that we detected from Mtb ΔcnpB is unlikely due to bacterial lysis (Yang et al.2014). Meanwhile, mycobacterial DNA sensed by cGAS has also been considered to originate from living but not degraded bacteria (Majlessi and Brosch 2015). Based on the results of the type I IFN response of BMDMs isolated from WT, STING−/− and cGAS−/− mice, c-di-AMP possibly enhances DNA secretion by Mtb. It is also likely that Mtb DNA and c-di-AMP share the same secretion machinery, which we will explore in the future. Several studies have demonstrated that type I IFN increased host susceptibility to Mtb infection (Manca et al.2001, 2005; Antonelli et al.2010; Dorhoi et al.2014). On the other hand, there is also evidence that type I IFN is protective in certain circumstances (Desvignes, Wolf and Ernst 2012). Thus, it is likely that a balance of this cytokine is required for optimal protection during Mtb infection (Wiens and Ernst 2016). Nonetheless, the recombinant BCG vaccine with ESX-1 of M. marinum induces type I interferon and stronger T cell response, which provides better protection against Mtb challenge (Groschel et al.2017), suggesting that the type I IFN induced during vaccination is beneficial to the host. Most interestingly, we found that BCG ΔcnpB was unable to secrete c-di-AMP. This observation is highly significant from the vaccine perspective, since the limitation of BCG in the prevention of the TB epidemic is still not fully understood. Manipulation of c-di-AMP secretion from BCG and release from the phagosome into host cytosol during vaccination will enable BCG to induce substantial levels of type I IFN locally, which might improve BCG’s immunogenicity based on multiple vaccine studies (Giacomini et al.2009; Guerrero et al.2015; Conrad et al.2017; Groschel et al.2017). This ‘localized’ response could be optimized by the levels of c-di-AMP production and should not significantly enhance the susceptibility of Mtb infection. Furthermore, c-di-AMP has been explored as an adjuvant for vaccinations and induces strong humoral and cellular immune responses (Ebensen et al.2011; Sanchez et al.2014; Skrnjug et al.2014). Therefore, an engineered BCG based on ΔcnpB and amended c-di-AMP secretion machinery may elevate the vaccine efficacy. In our future studies, we will generate and examine such recombinant BCG strains as novel vaccines in protection against TB. SUPPLEMENTARY DATA Supplementary data are available at FEMSPD online. Acknowledgements We thank Tiffany Zarrella and Dr Gwendowlyn Knapp for critical reading of the manuscript; Dr Lei Jin for providing STING−/− mice and the technical assistance; Dr Herbert Virgin for providing cGAS−/− mice; Dr Williams Jacobs, Jr. for providing the phage to delete RD1 locus; Dr Keith Derbyshire for providing the cosmid harboring the RD1 locus; and Dr Jeffery Cox for providing pYO11 plasmid. We are grateful to the Biosafety Level 3 Core Facility of Albany Medical College. FUNDING This project is partly supported by a Scientist Development Grant of the American Heart Association 12SDG12080067 to GB. Conflict of interest. None declared. Footnotes Present address: Shanghai Institute for Advanced Immunochemical Studies, 99 Haike Road, Pudong District, Shanghai, China. REFERENCES Antonelli LR, Gigliotti Rothfuchs A, Goncalves R et al.   Intranasal Poly-IC treatment exacerbates tuberculosis in mice through the pulmonary recruitment of a pathogen-permissive monocyte/macrophage population. J Clin Invest  2010; 120: 1674– 82. Google Scholar CrossRef Search ADS PubMed  Archer KA, Durack J, Portnoy DA. STING-dependent type I IFN production inhibits cell-mediated immunity to Listeria monocytogenes. PLoS Pathog  2014; 10: e1003861. Google Scholar CrossRef Search ADS PubMed  Bai G, Schaak DD, McDonough KA. cAMP levels within Mycobacterium tuberculosis and Mycobacterium bovis BCG increase upon infection of macrophages. FEMS Immunol Med Microbiol  2009; 55: 68– 73. Google Scholar CrossRef Search ADS PubMed  Bai G, Schaak DD, Smith EA et al.   Dysregulation of serine biosynthesis contributes to the growth defect of a Mycobacterium tuberculosis crp mutant. Mol Microbiol  2011; 82: 180– 98. Google Scholar CrossRef Search ADS PubMed  Bai Y, Yang J, Eisele LE et al.   Two DHH subfamily 1 proteins in Streptococcus pneumoniae possess cyclic di-AMP phosphodiesterase activity and affect bacterial growth and virulence. J Bacteriol  2013; 195: 5123– 32. Google Scholar CrossRef Search ADS PubMed  Bai Y, Yang J, Zhou X et al.   Mycobacterium tuberculosis Rv3586 (DacA) is a diadenylate cyclase that converts ATP or ADP into c-di-AMP. PLoS One  2012; 7: e35206. Google Scholar CrossRef Search ADS PubMed  Banerjee R, Gretes M, Harlem C et al.   A mecA-negative strain of methicillin-resistant Staphylococcus aureus with high-level beta-lactam resistance contains mutations in three genes. Antimicrob Agents Ch  2010; 54: 4900– 2. Google Scholar CrossRef Search ADS   Barker JR, Koestler BJ, Carpenter VK et al.   STING-dependent recognition of cyclic di-AMP mediates type I interferon responses during Chlamydia trachomatis infection. MBio  2013; 4: e00018– 00013. Google Scholar CrossRef Search ADS PubMed  Behr MA, Wilson MA, Gill WP et al.   Comparative genomics of BCG vaccines by whole-genome DNA microarray. Science  1999; 284: 1520– 3. Google Scholar CrossRef Search ADS PubMed  Bowie AG. Innate sensing of bacterial cyclic dinucleotides: more than just STING. Nat Immunol  2012; 13: 1137– 9. Google Scholar CrossRef Search ADS PubMed  Brosch R, Gordon SV, Garnier T et al.   Genome plasticity of BCG and impact on vaccine efficacy. P Natl Acad Sci USA  2007; 104: 5596– 601. Google Scholar CrossRef Search ADS   Burdette DL, Monroe KM, Sotelo-Troha K et al.   STING is a direct innate immune sensor of cyclic di-GMP. Nature  2011; 478: 515– 8. Google Scholar CrossRef Search ADS PubMed  Cole ST, Brosch R, Parkhill J et al.   Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature  1998; 393: 537– 44. Google Scholar CrossRef Search ADS PubMed  Collins AC, Cai H, Li T et al.   Cyclic GMP-AMP synthase is an innate immune DNA sensor for Mycobacterium tuberculosis. Cell Host Microbe  2015; 17: 820– 8. Google Scholar CrossRef Search ADS PubMed  Conrad WH, Osman MM, Shanahan JK et al.   Mycobacterial ESX-1 secretion system mediates host cell lysis through bacterium contact-dependent gross membrane disruptions. P Natl Acad Sci USA  2017; 114: 1371– 6. Google Scholar CrossRef Search ADS   Corrigan RM, Abbott JC, Burhenne H et al.   c-di-AMP is a new second messenger in Staphylococcus aureus with a role in controlling cell size and envelope stress. PLoS Pathog  2011; 7: e1002217. Google Scholar CrossRef Search ADS PubMed  Corrigan RM, Campeotto I, Jeganathan T et al.   Systematic identification of conserved bacterial c-di-AMP receptor proteins. P Natl Acad Sci USA  2013; 110: 9084– 9. Google Scholar CrossRef Search ADS   Cron LE, Stol K, Burghout P et al.   Two DHH subfamily 1 proteins contribute to pneumococcal virulence and confer protection against pneumococcal disease. Infect Immun  2011; 79: 3697– 710. Google Scholar CrossRef Search ADS PubMed  de Jonge MI, Pehau-Arnaudet G, Fretz MM et al.   ESAT-6 from Mycobacterium tuberculosis dissociates from its putative chaperone CFP-10 under acidic conditions and exhibits membrane-lysing activity. J Bacteriol  2007; 189: 6028– 34. Google Scholar CrossRef Search ADS PubMed  De Leon J, Jiang G, Ma Y et al.   Mycobacterium tuberculosis ESAT-6 exhibits a unique membrane-interacting activity that is not found in its ortholog from non-pathogenic Mycobacterium smegmatis. J Biol Chem  2012; 287: 44184– 91. Google Scholar CrossRef Search ADS PubMed  Desvignes L, Wolf AJ, Ernst JD. Dynamic roles of type I and type II IFNs in early infection with Mycobacterium tuberculosis. J Immunol  2012; 188: 6205– 15. Google Scholar CrossRef Search ADS PubMed  Dey B, Dey RJ, Cheung LS et al.   A bacterial cyclic dinucleotide activates the cytosolic surveillance pathway and mediates innate resistance to tuberculosis. Nat Med  2015; 21: 401– 6. Google Scholar CrossRef Search ADS PubMed  Dey RJ, Dey B, Zheng Y et al.   Inhibition of innate immune cytosolic surveillance by an M. tuberculosis phosphodiesterase. Nat Chem Biol  2017; 13: 210– 7. Google Scholar CrossRef Search ADS PubMed  Dorhoi A, Yeremeev V, Nouailles G et al.   Type I IFN signaling triggers immunopathology in tuberculosis-susceptible mice by modulating lung phagocyte dynamics. Eur J Immunol  2014; 44: 2380– 93. Google Scholar CrossRef Search ADS PubMed  Ebensen T, Libanova R, Schulze K et al.   Bis-(3΄,5΄)-cyclic dimeric adenosine monophosphate: strong Th1/Th2/Th17 promoting mucosal adjuvant. Vaccine  2011; 29: 5210– 20. Google Scholar CrossRef Search ADS PubMed  Flint JL, Kowalski JC, Karnati PK et al.   The RD1 virulence locus of Mycobacterium tuberculosis regulates DNA transfer in Mycobacterium smegmatis. P Natl Acad Sci USA  2004; 101: 12598– 603. Google Scholar CrossRef Search ADS   Giacomini E, Remoli ME, Gafa V et al.   IFN-beta improves BCG immunogenicity by acting on DC maturation. J Leukoc Biol  2009; 85: 462– 8. Google Scholar CrossRef Search ADS PubMed  Gordon SV, Brosch R, Billault A et al.   Identification of variable regions in the genomes of tubercle bacilli using bacterial artificial chromosome arrays. Mol Microbiol  1999; 32: 643– 55. Google Scholar CrossRef Search ADS PubMed  Griffiths JM, O’Neill AJ. Loss of function of the GdpP protein leads to joint beta-lactam/glycopeptide tolerance in Staphylococcus aureus. Antimicrob Agents Ch  2012; 56: 579– 81. Google Scholar CrossRef Search ADS   Groschel MI, Sayes F, Shin SJ et al.   Recombinant BCG expressing ESX-1 of Mycobacterium marinum combines low virulence with cytosolic immune signaling and improved TB protection. Cell Rep  2017; 18: 2752– 65. Google Scholar CrossRef Search ADS PubMed  Guerrero GG, Rangel-Moreno J, Islas-Trujillo S et al.   Successive intramuscular boosting with IFN-alpha protects Mycobacterium bovis BCG-vaccinated mice against M. lepraemurium infection. Biomed Res Int  2015; 2015: 414027. Google Scholar CrossRef Search ADS PubMed  Gundlach J, Herzberg C, Kaever V et al.   Control of potassium homeostasis is an essential function of the second messenger cyclic di-AMP in Bacillus subtilis. Sci Signal  2017; 10: eaal3011. Google Scholar CrossRef Search ADS PubMed  Hansen K, Prabakaran T, Laustsen A et al.   Listeria monocytogenes induces IFNbeta expression through an IFI16-, cGAS- and STING-dependent pathway. EMBO J  2014; 33: 1654– 66. Google Scholar CrossRef Search ADS PubMed  Houben D, Demangel C, van Ingen J et al.   ESX-1-mediated translocation to the cytosol controls virulence of mycobacteria. Cell Microbiol  2012; 14: 1287– 98. Google Scholar CrossRef Search ADS PubMed  Hsu T, Hingley-Wilson SM, Chen B et al.   The primary mechanism of attenuation of bacillus Calmette-Guerin is a loss of secreted lytic function required for invasion of lung interstitial tissue. P Natl Acad Sci USA  2003; 100: 12420– 5. Google Scholar CrossRef Search ADS   Jin L, Hill KK, Filak H et al.   MPYS is required for IFN response factor 3 activation and type I IFN production in the response of cultured phagocytes to bacterial second messengers cyclic-di-AMP and cyclic-di-GMP. J Immunol  2011; 187: 2595– 601. Google Scholar CrossRef Search ADS PubMed  Jin L, Waterman PM, Jonscher KR et al.   MPYS, a novel membrane tetraspanner, is associated with major histocompatibility complex class II and mediates transduction of apoptotic signals. Mol Cell Biol  2008; 28: 5014– 26. Google Scholar CrossRef Search ADS PubMed  Kamegaya T, Kuroda K, Hayakawa Y. Identification of a Streptococcus pyogenes SF370 gene involved in production of c-di-AMP. Nagoya J Med Sci  2011; 73: 49– 57. Google Scholar PubMed  Luo Y, Helmann JD. Analysis of the role of Bacillus subtilis sigma(M) in beta-lactam resistance reveals an essential role for c-di-AMP in peptidoglycan homeostasis. Mol Microbiol  2012; 83: 623– 39. Google Scholar CrossRef Search ADS PubMed  Mahairas GG, Sabo PJ, Hickey MJ et al.   Molecular analysis of genetic differences between Mycobacterium bovis BCG and virulent M. bovis. J Bacteriol  1996; 178: 1274– 82. Google Scholar CrossRef Search ADS PubMed  Majlessi L, Brosch R. Mycobacterium tuberculosis meets the cytosol: the role of cGAS in anti-mycobacterial immunity. Cell Host Microbe  2015; 17: 733– 5. Google Scholar CrossRef Search ADS PubMed  Manca C, Tsenova L, Bergtold A et al.   Virulence of a Mycobacterium tuberculosis clinical isolate in mice is determined by failure to induce Th1 type immunity and is associated with induction of IFN-alpha /beta. P Natl Acad Sci USA  2001; 98: 5752– 7. Google Scholar CrossRef Search ADS   Manca C, Tsenova L, Freeman S et al.   Hypervirulent M. tuberculosis W/Beijing strains upregulate type I IFNs and increase expression of negative regulators of the Jak-Stat pathway. J Interferon Cytokine Res  2005; 25: 694– 701. Google Scholar CrossRef Search ADS PubMed  Manikandan K, Sabareesh V, Singh N et al.   Two-step synthesis and hydrolysis of cyclic di-AMP in Mycobacterium tuberculosis. PLoS One  2014; 9: e86096. Google Scholar CrossRef Search ADS PubMed  Manzanillo PS, Shiloh MU, Portnoy DA et al.   Mycobacterium tuberculosis activates the DNA-dependent cytosolic surveillance pathway within macrophages. Cell Host Microbe  2012; 11: 469– 80. Google Scholar CrossRef Search ADS PubMed  Mehne FM, Gunka K, Eilers H et al.   Cyclic di-AMP homeostasis in Bacillus subtilis: both lack and high-level accumulation of the nucleotide are detrimental for cell growth. J Biol Chem  2013; 288: 2004– 17. Google Scholar CrossRef Search ADS PubMed  Novikov A, Cardone M, Thompson R et al.   Mycobacterium tuberculosis triggers host type I IFN signaling to regulate IL-1beta production in human macrophages. J Immunol  2011; 187: 2540– 7. Google Scholar CrossRef Search ADS PubMed  Pandey AK, Yang Y, Jiang Z et al.   NOD2, RIP2 and IRF5 play a critical role in the type I interferon response to Mycobacterium tuberculosis. PLoS Pathog  2009; 5: e1000500. Google Scholar CrossRef Search ADS PubMed  Parvatiyar K, Zhang Z, Teles RM et al.   The helicase DDX41 recognizes the bacterial secondary messengers cyclic di-GMP and cyclic di-AMP to activate a type I interferon immune response. Nat Immunol  2012; 13: 1155– 61. Google Scholar CrossRef Search ADS PubMed  Peng X, Zhang Y, Bai G et al.   Cyclic di-AMP mediates biofilm formation. Mol Microbiol  2016; 99: 945– 59. Google Scholar CrossRef Search ADS PubMed  Pozzi C, Waters EM, Rudkin JK et al.   Methicillin resistance alters the biofilm phenotype and attenuates virulence in Staphylococcus aureus device-associated infections. PLoS Pathog  2012; 8: e1002626. Google Scholar CrossRef Search ADS PubMed  Pym AS, Brodin P, Majlessi L et al.   Recombinant BCG exporting ESAT-6 confers enhanced protection against tuberculosis. Nat Med  2003; 9: 533– 9. Google Scholar CrossRef Search ADS PubMed  Romling U. Great times for small molecules: c-di-AMP, a second messenger candidate in Bacteria and Archaea. Sci Signal  2008; 1: pe39. Google Scholar CrossRef Search ADS PubMed  Rosenberg OS, Dovala D, Li X et al.   Substrates control multimerization and activation of the multi-domain ATPase motor of type VII secretion. Cell  2015; 161: 501– 12. Google Scholar CrossRef Search ADS PubMed  Sanchez MV, Ebensen T, Schulze K et al.   Intranasal delivery of influenza rNP adjuvanted with c-di-AMP induces strong humoral and cellular immune responses and provides protection against virus challenge. PLoS One  2014; 9: e104824. Google Scholar CrossRef Search ADS PubMed  Sauer JD, Sotelo-Troha K, von Moltke J et al.   The N-ethyl-N-nitrosourea-induced Goldenticket mouse mutant reveals an essential function of Sting in the in vivo interferon response to Listeria monocytogenes and cyclic dinucleotides. Infect Immun  2011; 79: 688– 94. Google Scholar CrossRef Search ADS PubMed  Schoggins JW, MacDuff DA, Imanaka N et al.   Pan-viral specificity of IFN-induced genes reveals new roles for cGAS in innate immunity. Nature  2014; 505: 691– 5. Google Scholar CrossRef Search ADS PubMed  Schwartz KT, Carleton JD, Quillin SJ et al.   Hyperinduction of host beta interferon by a Listeria monocytogenes strain naturally overexpressing the multidrug efflux pump MdrT. Infect Immun  2012; 80: 1537– 45. Google Scholar CrossRef Search ADS PubMed  Simeone R, Bobard A, Lippmann J et al.   Phagosomal rupture by Mycobacterium tuberculosis results in toxicity and host cell death. PLoS Pathog  2012; 8: e1002507. Google Scholar CrossRef Search ADS PubMed  Simeone R, Sayes F, Song O et al.   Cytosolic access of Mycobacterium tuberculosis: critical impact of phagosomal acidification control and demonstration of occurrence in vivo. PLoS Pathog  2015; 11: e1004650. Google Scholar CrossRef Search ADS PubMed  Skrnjug I, Rueckert C, Libanova R et al.   The mucosal adjuvant cyclic di-AMP exerts immune stimulatory effects on dendritic cells and macrophages. PLoS One  2014; 9: e95728. Google Scholar CrossRef Search ADS PubMed  Stanley SA, Johndrow JE, Manzanillo P et al.   The Type I IFN response to infection with Mycobacterium tuberculosis requires ESX-1-mediated secretion and contributes to pathogenesis. J Immunol  2007; 178: 3143– 52. Google Scholar CrossRef Search ADS PubMed  Tadmor K, Pozniak Y, Burg Golani T et al.   Listeria monocytogenes MDR transporters are involved in LTA synthesis and triggering of innate immunity during infection. Front Cell Infect Microbiol  2014; 4: 16. Google Scholar CrossRef Search ADS PubMed  Underwood AJ, Zhang Y, Metzger DW et al.   Detection of cyclic di-AMP using a competitive ELISA with a unique pneumococcal cyclic di-AMP binding protein. J Microbiol Methods  2014; 107: 58– 62. Google Scholar CrossRef Search ADS PubMed  van der Wel N, Hava D, Houben D. M. tuberculosis and M. leprae translocate from the phagolysosome to the cytosol in myeloid cells. Cell  2007; 129: 1287– 98. Google Scholar CrossRef Search ADS PubMed  Wassermann R, Gulen MF, Sala C et al.   Mycobacterium tuberculosis differentially activates cGAS- and inflammasome-dependent intracellular immune responses through ESX-1. Cell Host Microbe  2015; 17: 799– 810. Google Scholar CrossRef Search ADS PubMed  Watson RO, Bell SL, MacDuff DA et al.   The cytosolic sensor cGAS detects Mycobacterium tuberculosis DNA to induce type I interferons and activate autophagy. Cell Host Microbe  2015; 17: 811– 9. Google Scholar CrossRef Search ADS PubMed  Whiteley AT, Pollock AJ, Portnoy DA. The PAMP c-di-AMP is essential for Listeria monocytogenes growth in rich but not minimal media due to a toxic increase in (p)ppGpp. Cell Host Microbe  2015; 17: 788– 98. Google Scholar CrossRef Search ADS PubMed  Wiens KE, Ernst JD. The mechanism for type I interferon induction by Mycobacterium tuberculosis is bacterial strain-dependent. PLoS Pathog  2016; 12: e1005809. Google Scholar CrossRef Search ADS PubMed  Witte CE, Whiteley AT, Burke TP et al.   Cyclic di-AMP is critical for Listeria monocytogenes growth, cell wall homeostasis, and establishment of infection. MBio  2013; 4: e00282– 00213. Google Scholar CrossRef Search ADS PubMed  Witte G, Hartung S, Buttner K et al.   Structural biochemistry of a bacterial checkpoint protein reveals diadenylate cyclase activity regulated by DNA recombination intermediates. Mol Cell  2008; 30: 167– 78. Google Scholar CrossRef Search ADS PubMed  Woodward JJ, Iavarone AT, Portnoy DA. c-di-AMP secreted by intracellular Listeria monocytogenes activates a host type I interferon response. Science  2010; 328: 1703– 5. Google Scholar CrossRef Search ADS PubMed  Yamamoto T, Hara H, Tsuchiya K et al.   Listeria monocytogenes strain-specific impairment of the TetR regulator underlies the drastic increase in cyclic di-AMP secretion and beta interferon-inducing ability. Infect Immun  2012; 80: 2323– 32. Google Scholar CrossRef Search ADS PubMed  Yang J, Bai Y, Zhang Y et al.   Deletion of the cyclic di-AMP phosphodiesterase gene (cnpB) in Mycobacterium tuberculosis leads to reduced virulence in a mouse model of infection. Mol Microbiol  2014; 93: 65– 79. Google Scholar CrossRef Search ADS PubMed  © FEMS 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

Pathogens and DiseaseOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off