Bacterial Protein Reshapes Host Defense toward Antiviral Responses

Abstract

In this issue of Molecular Cell, Penn et al. (2018) report the protein interactome between Mycobacterium tuberculosis (Mtb) secreted effectors and macrophage cytosolic proteins. This Resource reveals that the interaction of Mtb effector LpqN with host CBL counteracts antibacterial defense but causes a reciprocal enhancement of antiviral defense.

Cells are packed with proteins that interact with each other to modulate signals required for homeostasis in normal physiology or during adaptation in a disease state. As proteins rarely act alone, their activities are regulated in large part by protein-protein interactions (PPIs). The importance of PPIs is highlighted by the fact that they represent frequent targets for therapeutic intervention. Following infection by pathogens, hosts rearrange homeostatic PPIs to activate various immune pathways, such as Toll-like receptor signaling, NF-κB signaling, phagosome maturation, and the apoptotic pathway, to recognize and ultimately clear invading pathogens. As a countermeasure, pathogens have evolved diverse strategies to survive hostile host environments (Durmuş Tekir et al., 2012). One such example is the intimate interaction between immune determinants of the host and virulence factors of the pathogen to modulate and ultimately evade host immunity. Secretion of microbial proteins that interfere with essential intracellular PPIs and disturb host immune pathways is a common evasion strategy for many viruses (Shapira et al., 2009) and bacteria (Bhavsar et al., 2007). Post-invasion, many Gram-negative bacteria use the type III secretion system (T3SS) as an injection device to directly deliver multiple virulence proteins, referred to as effectors, into the cytosol of infected cells for the disruption of host PPIs (Galán et al., 2014). In this issue of Molecular Cell, Penn et al. (2018) identified a strategy used by Mycobacterium tuberculosis (Mtb) that is functionally analogous to the T3SS and mapped the extensive PPIs between Mtb secreted effectors and macrophage cytosolic proteins (Penn et al., 2018). This study provides valuable insight into the biological functions of Mtb virulence factors, highlights the protein interactome critical to Mtb intracellular survival, and improves overall understanding of the molecular basis of Mtb pathogenicity.

While virus-host PPIs that mediate immune evasion and pathogenesis are well studied, the characterization of bacteria-host PPIs has lagged behind. There is increasing evidence that Mtb remodels host intracellular activities by altering its PPIs (Bhavsar et al., 2007). The Mtb-host protein interactome has previously been characterized using a yeast two-hybrid (Y2H) screen system (Nicod et al., 2017). Secreted effectors are of particular importance as they are often required for full Mtb virulence (Feltcher et al., 2010). For instance, the secreted Mtb PE_PGRS47 effector resides in the cytosol of macrophages where it inhibits MHC II-mediated antigen presentation, partially suppressing anti-Mtb autophagic activity (Saini et al., 2016). Analysis of protein interactome datasets has shown that the Mtb EsxH interaction with host ESCRT (endosome sorting complexes required for transport) interferes with antigen-specific CD4⁺ T cell proliferation (Portal-Celhay et al., 2016). Although the Y2H-mediated genetic approach has been widely used for studying protein interactomes, false-positive interactions and the physiological difference between mammalian and yeast cells are major pitfalls of this system.

Penn et al. (2018) employed a systems-level approach through the combination of affinity purification and mass spectrometry (AP-MS) to further investigate the Mtb-host protein interactome. This study identified 105 Mtb secreted effector candidates that were known to be Sec transporter dependent or independent as well as ESX type VII secretion system substrates. When MiST bioinformatics were applied to the AP-MS analysis, a high-confidence interactome of 34 Mtb secreted effectors and 187 host proteins was discovered. Subsequently, Penn et al. (2018) performed computational analyses using CORUM and STRING databases to interpret this protein interactome and identified many host interactors functioning in various regulatory pathways such as COP II vesicular trafficking complex, CCT chaperone complex, STRIPAK signal transduction complex, and RNAP II-mediated transcriptional elongation. The authors also observed that most host interactors underwent post-translational modifications such as phosphorylation and/or ubiquitination in response to Mtb infection, implying their functional alteration. The physiological relevance of these PPIs was evaluated by selecting and testing ten Mtb effectors interacting with host proteins that were involved in antibacterial immunity and/or underwent post-translational modifications.

Specifically, Penn et al. (2018) identified the Mtb LpqN as an essential effector for in vivo survival, and thus the LpqN-deficient Mtb mutant (ΔlpqN) was growth defective within the host (Figure 1). Notably, this growth defect was not seen in anoxic or normal in vitro conditions, implying that LpqN is a virulence factor required for intracellular survival. Furthermore, authors used AP-MS and discovered the host E3 ubiquitin ligase CBL (casitas B-lineage lymphoma) as an interactor of LpqN. CBL includes three distinct gene products in mammals: CBL, CBL-B, and CBL-C. Extensive studies of the CBL family have established that it functions as a protein tyrosine kinase (PTK)-selective E3 ubiquitin ligase (Mohapatra et al., 2013). CBL-dependent ubiquitination targets PTKs for degradation either by facilitating their entry into lysosomes or by promoting proteasomal degradation. In this study, Penn et al. (2018) showed that the growth of Mtb ΔlpqN was markedly attenuated in wild-type mouse bone marrow-derived macrophages (BMDMs) but was intact in Cbl^−/− mouse BMDM. Of note, Mtb with intact LpqN was relatively insensitive to CBL-mediated antibacterial activity, a property not directly associated with its known E3 ubiquitin ligase function. This indicates that Mtb secretes LpqN effector to circumvent CBL-mediated antibacterial activity in an E3 ligase-independent manner (Figure 1A). While the precise mechanism of CBL-mediated antibacterial activity and LpqN-mediated evasion remain elusive, their functional interaction may be a potential future therapeutic target.

Figure 1. Mtb Secretes LpqN to Evade Host CBL-Mediated Antibacterial Response.

(A and B) Host macrophages use CBL to balance antibacterial and antiviral responses. (A) After infection, wild-type Mtb ensures its intracellular survival by secreting LpqN that physically interacts with CBL and disturbs CBL-mediated antibacterial activity. This subsequently induces host antiviral defense. (B) LpqN-deficient Mtb (ΔlpqN) that cannot block the CBL-mediated antibacterial response is growth defective within the wild-type host.

Penn et al. (2018) intriguingly showed that while Cbl^−/− mouse BMDMs lost their antibacterial activity against Mtb, they reciprocally enhanced their antiviral ability by increasing expression of the type I interferon, IFN-β. The authors postulate that Mtb actively releases bacterial DNA into the host in order to trigger the cGAS/STING-mediated antiviral response, which may counteract antibacterial activity. As CBL regulates the balance between cell-intrinsic antibacterial and antiviral responses, the LpqN-CBL interaction may prolong or amplify such a signal and create a permissive intracellular environment for Mtb growth. On the other hand, this also suggests that during infection, hosts have evolved to deploy either a virus- or bacteria-specific defense but not both (Figure 1).