‘Drugs from Bugs’: bacterial effector proteins as promising biological (immune-) therapeutics
Center for Molecular Biology of Inflammation (ZMBE), Institute of Infectiology, University of Münster, Münster, Germany
Correspondence: Christian Rüter, Center for Molecular Biology of Inflammation (ZMBE), Institute of Infectiology, University of Münster, von Esmarchstrasse 56, 48149 Münster, Germany. Tel.: +49 251 83 56477; fax: +49 251 83 56467; e-mail: firstname.lastname@example.org
Immune system malfunctions cause many of the most severe human diseases. The immune system has evolved primarily to control bacterial, viral, fungal, and parasitic infections. In turn, over millions of years of coevolution, microbial pathogens have evolved various mechanisms to control and modulate the host immune system for their own benefit and survival. For example, many bacterial pathogens use virulence proteins to modulate and exploit target cell mechanisms. Our understanding of these bacterial strategies opens novel possibilities to exploit ‘microbial knowledge’ to control excessive immune reactions. Gaining access to strategies of microbial pathogens could lead to potentially huge benefits for the therapy of inflammatory diseases. Most work on bacterial pathogen effector proteins has the long-term aim of neutralizing the infectious capabilities of the pathogen. However, attenuated pathogens and microbial products have been used for over a century with overwhelming success in the form of vaccines to induce specific immune responses that protect against the respective infectious diseases. In this review, we focus on bacterial effector and virulence proteins capable of modulating and suppressing distinct signaling pathways with potentially desirable immune-modulating effects for treating unrelated inflammatory diseases.
Autoimmune and chronic inflammatory diseases are based to a major extent on aberrant and overactive immune responses of the host against its own tissues and components. In autoimmunity, the patient's immune system is activated against the body's own proteins (Smedby et al., 2008). In inflammatory diseases, the overreaction of the immune system and its subsequent downstream signaling (TNF, IFN, etc.) cause tissue damage (McInnes & Schett, 2007). In recent years, therapeutic progress has been achieved by targeting the uncontrolled immune responses, and inhibiting innate immune responses is a primary goal for therapy of inflammatory diseases. Pharmacologic inhibition of the innate immune system by immunosuppressive drugs is efficacious and has been successfully applied to treat patients suffering from autoimmune diseases such as psoriasis, rheumatoid arthritis, and inflammatory bowel diseases (Cooper & Stroehla, 2003). Particularly, neutralization of individual cytokines (e.g. TNF, IL-12, IL-15, IFN) by antibodies has been highly effective in treating noninfectious autoimmune diseases (Rang et al., 2003). However, redundancy of cytokine effects can counteract therapeutic effects, and an intervention at a higher level of the involved signaling pathways might be even more effective (Yadav & Sarvetnick, 2003).
Despite remarkable successes, immunosuppressive agents are not without side effects and potential serious risks. Due to the need for systemic application and their nonselective mode of action, the control of novel or dormant microbial infections or the spread of malignant cells is severely hampered (Terme et al., 2008). As a consequence, for many patients, the application of immunosuppressive drugs is severely limited or impossible. In addition, the production of the most effective immunosuppressive agents is expensive and time-consuming. Therefore, a more cost-efficient, novel strategy for the therapy of autoimmune diseases that is targeted directly to the site of inflammation and without the need for systemic applications would be highly advantageous and would greatly lower the risks of deleterious side effects. In this regard, identifying and characterizing molecules from various pathogens that might be adapted for the benefit of the host is an expanding area of research that should offer a unique opportunity to uncover a large collection of natural modulators of inflammation with the potential for use as novel immunotherapeutics.
Bacterial effector proteins as promising biological therapeutics
The use of pathogens as therapeutics is well established through the exposure of people to live or attenuated pathogens as vaccines for infectious disease. An extension of this strategy is using the potentially desirable immune-modulating effects of pathogen infection for treating unrelated inflammatory diseases. One of the most fascinating and widespread pathogenicity modules in Gram-negative pathogens is the type III secretion system (T3SS) that targets essential cytoplasmic processes of the host cell by directly injecting effector proteins into the cytoplasm via a molecular injection machine (‘molecular syringe’; Cornelis, 2002). Some of these effector proteins interfere with signaling mechanisms, including those triggering innate immune responses. In this regard, pathogenic Gram-negative bacteria harbor a wide range of modulators that target a considerable number of signaling mechanisms such as MAPKs (mitogen-activated protein kinases) signaling cascades, ubiquitin signaling pathways, or pathways leading to repression of NF-κB (nuclear factor kappa B) activation (Matsumoto & Young, 2009). Immune signaling in mammals involves a complex system of host molecules triggered by a variety of extracellular and intracellular signals. However, the involved immune signaling pathways follow an ‘hourglass’ shape (Fig. 1; Beutler, 2004). A multiplicity of incoming signals such as PAMPs (pathogen-associated molecular patterns) or DAMPS (danger-associated molecular patterns) are sensed by dozens of receptors (e.g. Toll-like receptors, cytokine receptors, CD receptors) and transduced by numerous adaptor proteins to the central NF-κB and MAPK signaling pathways (Hoebe et al., 2003). From here, the signaling paths grow in complexity again, and incoming signals are amplified by several protein kinases, which collectively lead to the activation of numerous transcription factors (e.g. p65, AP-1, and IRF3), which are responsible for the production of hundreds of host molecules comprising the immune response of the host (Beutler, 2004). As the majority of the immune signal traverses this choke point, it is predestined for therapeutic intervention, and moreover, also pathogenic bacteria have exploited these circumstances to manipulate the inflammatory signaling of the host (Fig. 1; O'Sullivan et al., 2007). Up to 100 different bacterial effector proteins have been discovered, and a huge diversity of biochemical activities such as (de-)phosphorylation, (de-)ubiquitinylation, proteolysis, AMPylation, ribosylation, lyase reactivity, O-GlcNAcylation, or adenylate cyclase activity have been described (Fig. 1; Dean, 2011).
Of particular interest for a potential effective immune-therapeutic intervention of inflammatory disorders are those effectors that redirect host signaling networks and downregulate the immune response (Table 1). Several T3SS effectors of Escherichia coli, Salmonella, Yersinia, and Shigella target the activation of NF-κB and its translocation to the nucleus to promote or prevent inflammatory gene expression. The NF-κB/Rel transcription factors are present in the cytosol in an inactive state, complexed with the inhibitory IκB proteins. Activation of NF-κB occurs via phosphorylation of IκBα, resulting in the ubiquitinylation and proteasome-dependent degradation of the inhibitory protein. The active NF-κB dimers are phosphorylated and translocated to the nucleus (Ghosh & Karin, 2002). In addition to the NF-κB pathway, the p38 MAP kinase participates in a signaling cascade controlling cellular responses to cytokines and stress (Keshet & Seger, 2010). In this regard, recent studies uncovered ubiquitin ligase activities of the Shigella and Salmonella LPX-effector family. The regulated destruction of proteins via the ubiquitin proteasome pathway governs many cellular processes, including NF-κB or MAPK signal transduction pathways (Okuda et al., 2005, Rohde et al., 2007). Particularly, the ubiquitination of several host cell factors by effectors of the LPX-family such as SspH1 or IpaH9.8 leads to modulation of NF-κB gene transcription and in turn to inhibition of cytokine production (e.g. IL-8) and reduction of the host inflammatory response (Haraga & Miller, 2003; Ashida et al., 2010), making these two effectors promising candidates as immune-therapeutic agents (Fig. 1). In almost the same manner, the Shigella T3SS effector OspF exhibits impressive anti-inflammatory capacities. This effector protein harbors a phosphothreonine lyase activity that irreversibly inactivates MAP kinases p38 and ERK, implementing a highly selective inhibition of immune response of epithelial cells (Li et al., 2007; Zhu et al., 2007). OspG, another T3SS effector of Shigella, inhibits TNF-induced NF-κB activation by preventing ubiquitination of IκBα, thereby sequestering NF-κB in the cytosol and preventing inflammatory responses (Kim et al., 2005). Enteropathogenic E. coli (EPEC) encode the NleH1 and NleH2 T3SS effectors (Garcia-Angulo et al., 2008; Gao et al., 2009). Both proteins share sequence similarity with the Shigella flexneri OspG effector. NleH1 and NleH2 bind ribosomal protein S3 (RPS3), a newly discovered NF-κB subunit (Gao et al., 2009). RPS3 binds to the p65 NF-κB subunit and increases the affinity of NF-κB for a subset of target genes. However, only NleH1, but not NleH2, prevents the association of RPS3 with NF-κB in the nucleus (Wan et al., 2011).
Table 1. Examples of the vast array of bacterial pathogen effector proteins interfering with host immune signaling pathways. This effectors hold promise as a rich and important source of future bacteria-derived immune-modulatory molecules
E3 ubiquitin ligase
Inhibits cytokine production (IL-8)
E3 ubiquitin ligase
Suppresses immune responses
ERK and p38 MAPKs
Inhibits MAPK signaling
E2 ubiquitin conjugating enzymes
Inhibits nuclear NF-κB translocation
Inhibits RPS3 nuclear translocation
Inhibits NF-κB activation and IL-8 production
FADD, GAPDH, RIPK1, TRADD,
Suppresses NF-κB activation; Inhibits apoptosis
NF-κB p65 subunit
Cleaves and inactivates p65
Akt and FAK
Inhibits Akt signaling
Rho GAP mimicry
Inhibits NF-κB and caspase1 activation
IκB and MAPKK6
Inhibits MAPK and NF-κB signaling
Inhibits pro-inflammatory cytokines
The NF-κB pathway is also inhibited by the EPEC effector NleC. This effector protein is a potent metalloprotease that is responsible for proteolytic cleavage and inactivation of the p65 subunit (Pearson et al., 2011). Furthermore, NleE from EPEC and OspZ from Shigella can block selectively nuclear translocation of the p65 subunit and c-Rel, but not p50 or STAT1/2 (Newton et al., 2010). NleE is a methyltransferase that modifies the Npl4 zinc finger (NZF) domains of TAB 2/TAB 3, interfering with TAB 2/3 binding to ubiquitin chains and inhibiting both p65 nuclear translocation and IκBα degradation to block NF-κB activation (Newton et al., 2010; Zhang et al., 2012). Therefore, NleE provides great properties for the treatment of a number of inflammatory diseases, where aberrant IL-8 production can lead to chronic inflammatory conditions, as described for rheumatoid arthritis, inflammatory bowel disease, palmoplantar pustulosis (PPP), or psoriasis (Skov et al., 2008).
The NleB effector is highly conserved among the attaching/effacing pathogens (EHEC, EPEC, Citrobacter rodentium). EPEC NleB1 binds to death domain-containing proteins including FADD, RIPK1, and TRADD. NleB possesses an N-acetylglucosamine (GlcNAc) transferase activity that modifies arginine residues in these proteins to inhibit death receptor-induced apoptosis (Pearson et al., 2013). EPEC NleB-catalyzed modification of Arg residues of target proteins also results in blocked assembly of the TNF receptor 1 (TNFR1) complex, through GlcNAcylation of the TNFR1-associated death domain protein (TRADD), thus inhibiting apoptosis and NF-κB signaling (Li et al., 2013). NleB from C. rodentium suppresses NF-κB activation by binding the mammalian glycolysis enzyme glyceraldehyde 3-phosphate dehydrogenase (GAPDH). GAPDH normally binds the TNF receptor-associated factor 2 (TRAF2) and regulates TRAF2 polyubiquitination. Consequently, NleB-mediated GAPDH O-GlcNAcylation disrupts the TRAF2-GAPDH interaction to suppress TRAF2 polyubiquitination and NF-κB activation (Gao et al., 2013). This mechanism may be an interesting target against which a new drug class could be developed to inhibit dysfunctions of NF-κB signaling pathways.
Additionally, several Yersinia outer proteins (Yops) are known to have inhibitory functions on host immune responses (Viboud & Bliska, 2005). Some effectors of pathogenic Yersinia mimic host activities, such as YopH, a highly active phosphotyrosine phosphatase that interferes with focal adhesion kinase (FAK) and Akt (also known as Protein Kinase B) signaling, resulting in inhibition of phagocytosis and suppression of innate and adaptive immune response (Persson et al., 1997; Gerke et al., 2005). YopE, another Yersinia effector protein, also inhibits MAPK, NF-κB, and additionally caspase-1 activation by inactivating Rho GTPases (Von Pawel-Rammingen et al., 2000; Viboud et al., 2006). One of the most potent anti-inflammatory effector proteins of pathogenic Yersinia is YopP/YopJ, because it can block MAPK signaling just as well as NF-κB signaling (Matsumoto & Young, 2009). YopP/J acts as an acetyltransferase through the acetylation of MAPKK6 and through de-ubiquitination of IκBα, which in turn inhibits p65 translocation to the nucleus as well as MAPK signaling pathway (Zhou et al., 2005; Mukherjee et al., 2006).
Delivery of intracellular immune-modulating proteins
A hindrance for the potential therapeutic use of any intracellular-acting virulence protein is the difficulties in translocating the effector protein through the cell membrane for its delivery to the cytosol in the absence of the undesirable parent pathogen. However, these problems can be bypassed with the help of transduction technologies. Intracellular delivery and therapeutic application of hydrophilic molecules such as DNA, proteins, or oligonucleotides are often limited due to the poor permeability of the plasma membrane of eukaryotic cells (Alberts et al., 2002). Several strategies have been investigated to enable intracellular drug therapy involving viral and nonviral drug delivery systems, but most of these applications were restricted due to lack of stability of the complexes or degradation of the cargo (Afonin et al., 2006). A promising strategy to overcome these limitations is the conjugation of therapeutic compounds to cell-penetrating peptides (CPP), which are able to cross eukaryotic cell plasma membranes (Langel, 2005). CPPs are a heterogeneous group of relatively short peptides (10–30 amino acids) that share the ability to translocate into the cytoplasm of eukaryotic host cells (Hawiger, 1997). With reference to cellular import of cargo molecules, the protein transduction domains (PTD) of Tat, Pep1, penetratin, poly-arginine, or Transportan are known for their cell-penetrating properties (Stewart et al., 2008). These PTD can deliver covalently or noncovalently linked bioactive cargos, such as the described anti-inflammatory effector molecules into the host cell cytoplasm (Fig. 2). In addition to facilitating transport across the plasma membrane, applications of targeting/homing peptides to the fusion proteins can enable the constructs to cross physiological barriers (Fig 2). To date, various homing peptides have been identified by in vivo phage display and proven to act as cell type-specific or tissue-specific cell-penetrating agents (Langel, 2010). These peptides are able to concentrate a cargo inside the cells of several target tissues, such as neurons, bone marrow, kidney, liver, cancer, lung, spleen, or muscle (Jung et al., 2012). A potential targeted therapy might enhance significantly the efficacy and decrease the toxicity of potential therapeutic effector proteins in comparison with conventional, systemically applied drugs.
Apart from using CPP technology to deliver effector proteins, we found that T3SS effectors such as the translocated intimin receptor (Tir) of ATEC/EPEC or YopM of Y. enterocolitica are able to penetrate target cell membranes and remain functional, independent of additional factors (Michgehl et al., 2006; Rüter et al., 2010). The identification of the Yersinia outer protein M (YopM) as a cell-penetrating effector protein (CPE) established a new class within the heterogeneous group of CPPs (Rüter et al., 2010; Scharnert et al., 2013). The protein shares the ability of known CPPs to translocate across eukaryotic plasma membranes, and it has the capacity to deliver molecular cargos such as GFP intracellularly (Rüter et al., 2010). However, YopM does not show sequence homologies to known CPPs, indicating it may possess unique features that promote its uptake. YopM interferes with important signaling pathways in host target cells and, moreover, modulates the expression of immune signaling molecules (McPhee et al., 2010; Rüter et al., 2010). Cell-penetrating YopM is functional and efficiently downregulates the transcription of pro-inflammatory cytokines (e.g. TNF, IL-12, IL-15, and IL-18). Taken together, YopM can be considered as a novel ‘self-delivering’ immunotherapeutic molecule. So far, there are several reports of CPP conjugates that can attenuate inflammatory disease-associated responses, particularly in models of rheumatoid arthritis. One example of this is an apoptosis-inducing fusion protein that, when administered to rabbit joints, reduces the amount of synovial inflammation (Mi et al., 2003). Another example is a fusion protein derived from the NEMO-binding domain (NBD) and the PTD of Tat, which was used to block NF-κB activation with the consequence of reduced cytokine production and inhibited joint inflammation (Dai et al., 2004). Although these are very promising approaches, in contrast to these engineered peptides, YopM has been optimized during pathogen–host coevolution to reduce several pro-inflammatory cytokines simultaneously. These effects could be beneficial, especially for the treatment of autoimmune diseases characterized by elevated levels of pro-inflammatory cytokines.
The use of bacterial proteins to treat anti-inflammatory disorders is a highly innovative and cutting-edge translational approach and bears the potential to add a new class of biologics to the treatment strategies against inflammatory diseases. However, it is likely that the bacterial-derived proteins will induce antibody responses that might neutralize their efficacy and therefore might also introduce limitations for continuous or long-term application. Nevertheless, nearly all proteins currently used as human therapeutics induce antibodies in patients, and several strategies such as epitope removal, humanization, or PEGylation are used to reduce the immunogenicity of those biologicals (Schellekens, 2010). Taken together, the diversity of molecular actions employed by bacterial effector proteins demonstrates how pathogenic genomes/proteomes could provide a rich toolkit for therapeutic or biotechnological applications. Moreover, a possible topical application of anti-inflammatory biologics (effector proteins) would be surely advantageous, because current reagents are administered systemically, are patient-restricted in their application, cause numerous detrimental side effects, and harbor the risk to facilitate serious infections.
We would like to thank the group of Infectiology for critical reading and useful comments on the review. Work in the author's laboratories has been supported in part by the Deutsche Forschungsgemeinschaft – DFG (SFB 1009 TP B03, PA689/13-1), by a grant of the Interdisciplinary Clinical Research Center (IZKF; SchMA2/014/13), and by an Innovative Medical Research grant of the Medical Faculty of the University of Münster (IMF; I-RÜ111106).