Hacking the host: exploitation of macrophage polarization by intracellular bacterial pathogens

ABSTRACT

Macrophages play an integral role in host defenses against intracellular bacterial pathogens. A remarkable plasticity allows for adaptation to the needs of the host to orchestrate versatile innate immune responses to a variety of microbial threats. Several bacterial pathogens have adapted to macrophage plasticity and modulate the classical (M1) or alternative (M2) activation bias towards a polarization state that increases fitness for intracellular survival. Here, we summarize the current understanding of the host macrophage and intracellular bacterial interface; highlighting the roles of M1/M2 polarization in host defense and the mechanisms employed by several important intracellular pathogens to modulate macrophage polarization to favor persistence or proliferation. Understanding macrophage polarization in the context of disease caused by different bacterial pathogens is important for the identification of targets for therapeutic intervention.

We summarize the roles of M1/M2 polarization in host defense and the mechanisms employed by several important intracellular pathogens to modulate macrophage polarization to favor persistence or proliferation.

INTRODUCTION TO MACROPHAGES IN THE CONTEXT OF BACTERIAL INFECTION

Macrophages are highly adaptable mononuclear cells of the innate immune system that play critical roles as effectors and regulators of tissue development and homeostasis, defense against pathogens and resolution of tissue damage. Similar to granulocytes and neutrophils, macrophages originate from bone marrow hematopoietic myeloid precursors (Wynn, Chawla and Pollard 2013). Monocytes circulating in blood can differentiate into activated macrophages in tissues following migration in response to chemokine gradients. Mature tissue macrophages are now understood to primarily originate during embryonic development and not from circulating monocytes (Epelman, Lavine and Randolph 2014). Tissue compartments are normally populated with resident tissue macrophages, including populations with trained immunity that are maintained independently of monocytes or bone marrow progenitors (Yao et al. 2018). Tissue resident macrophages are the first phagocytes to encounter invading pathogens at these sites and are essentially present in every tissue or organ. These tissue resident macrophages are named according to their location and include osteoclasts (bone), microglia (central nervous system), alveolar (lung), Kupffer (liver) and histiocytes (connective tissue) among others.

Macrophages are adherent with high plasticity and, as opposed to other immune cells, are neither clonally restricted nor antigen specific. They express a diverse repertoire of surface molecules involved in pathogen recognition, phagocytosis, antigen presentation and activation of antimicrobial activity. Macrophages recognize bacteria through complement receptors, scavenger receptors and other pattern recognition receptors (PRR) located on the cellular surface including Toll-like receptors (TLR) and C type lectin receptors (CLR). Activation of macrophages through these PRR following pathogen exposure activates production of cytokines and chemokines that direct stimulatory or suppressive types of immune responses. Macrophages are important antigen presenting cells (APC) and display Class I and II major histocompatibility complex (MHC) molecules that present antigens to the acquired immune system. The Fc and complement receptors mediate phagocytosis of opsonized pathogens. Following phagocytic uptake, intracellular bacteria are degraded in the phagolysosome. The resulting microbial products are further processed into peptides and presented through MHC II molecules to the adaptive immune cells or directed into the autophagy pathway for proteasomal processing and presentation on MHC I.

In addition to orchestrating innate and adaptive immune responses following infection, macrophages employ multiple antimicrobial mechanisms to eliminate or reduce growth of intracellular bacteria (for review see (Flannagan, Cosío and Grinstein 2009; Weiss and Schaible 2015). These include the cathepsins, cathelicidins, acid hydrolases, lysozyme, cationic peptides, and the production of reactive oxygen species (ROS) and reactive nitrogen species (RNS). Among these, the RNS nitric oxide (NO), catalyzed by inducible nitric oxide synthetase (iNOS), is especially effective in damaging DNA of intracellular bacteria.

THE ROLE OF M1 AND M2 MACROPHAGE POLARIZATION DURING BACTERIAL INFECTION

Macrophages are heterogeneous, and their function depends on the microenvironment of cytokines and inflammatory molecules to which they are exposed (Wynn, Chawla and Pollard 2013; Epelman, Lavine and Randolph 2014; Okabe and Medzhitov 2014; Xue et al. 2014). This microenvironment reflects the healthy or damaged state of the tissue compartment as dictated by metabolic status, exposure to pathogens or pathogen-derived molecules and the presence of immune effector molecules (e.g. cytokines) and immune cells. The balance of these factors in the microenvironment can promote different polarization states of the macrophages along a continuum that includes the M1 and M2 polarization states as well as an increasingly complex list of M2 subgroups (Murray et al. 2014). Classical activation with Th1-related stimuli (e.g. IFNγ and LPS, or TNF-α) promotes M1 polarization through glycolysis driven metabolic pathways (Mills et al. 2016). Conversely, alternative activation by Th2-related cytokines (e.g. IL-4, IL-10 or TGFβ) promotes M2 polarization, resulting from metabolic changes downstream from oxidative phosphorylation and fatty acid oxidation pathways (Huang et al. 2014).

The polarization state of macrophages is an important determinant of the innate response to intracellular bacterial infection. M1 macrophages are pro-inflammatory and promote strong antimicrobial activity against intracellular pathogens as well as matrix degradation and tissue injury. Macrophages with an M1 phenotype are more efficient in the production of ROS and NO and therefore, are more efficient at killing bacteria in the intracellular compartment (Ezekowitz and Gordon 1984; Mosser and Handman 1992; Macmicking, Xie and Nathan 1997). They also produce pro-inflammatory cytokines such as TNFα, IL-1β, IL-6, IL-12 and IL-23. The chemokine receptor CCR7 is expressed by M1 macrophages and functions to promote homing to the lymph node as it is recruited by a gradient of ligands such as CCL19 and CCL21. M1 macrophages also direct recruitment of other immune cells, including neutrophils and Th1 cells, through production of pro-inflammatory chemokines such as CXCL8, CCL2, CXCL11, CXCL9 and CXCL10.

M2 macrophages have an immune regulatory role that functions to limit tissue damage. These cells produce anti-inflammatory cytokines such as IL-10 and TGFβ and promote fibrosis and wound healing. Increased production of arginase by M2 macrophages also inhibits NO due to enzyme competition among arginase and iNOS for catabolism of L-arginine. Tissue repair processes are also activated by M2 macrophages through multiple mechanisms such as production of PDGF (Platelet-derived growth factor), VEGF (Vascular endothelial growth factor), and EGF (Epidermal growth factor). The M2 cells also express the mannose receptor (MR), which signals production of chemokines, such as CCL17, CCL18, CCL22 and CCL24, that recruit Th2 lymphocytes, eosinophils, basophils, and T regulatory (Treg) cells (Mantovani et al. 2004).

There are well-established groups of M2 macrophages from M2a to M2d. The M2a macrophages are induced by IL-4 or IL-13 and contribute to tissue repair. The M2b subgroup is comprised of resting macrophages that are generated by activation with immune complexes and TLR agonists. M2c macrophages are induced by IL-10/TGFβ, IL-21 or glucocorticoids, and have defined anti-inflammatory and tissue remodeling roles (Martinez et al. 2008). The fourth group of M2 macrophages, M2d, are differentiated by co-stimulation with TLR and adenosine receptor agonists and produce IL-10 and VEGF (Grinberg et al. 2009; Ferrante et al. 2013).

Polarization states were initially understood to represent a process of a terminal differentiation. Due to a significant capacity for plasticity, macrophages can shift polarization states (e.g. macrophage repolarization of an M2 to an M1) depending on the environmental cues present (Hagemann et al. 2008; Duluc et al. 2009). There are cases where classically and alternatively activated macrophages are observed in the same microenvironment. Macrophages that demonstrate this paradoxical plasticity are considered to have an M3, or ‘switch’ phenotype (Malyshev and Malyshev 2015). A fourth group of polarized macrophages has been described as having an M4 polarization state that appears to be irreversible as currently understood (Gleissner et al. 2010). The M4 bias is driven by the platelet-derived chemokine CXCL4 and is characterized by distinct phenotypic and functional attributes that are related to atherosclerosis (Gleissner 2012). Additional plaque-specific macrophages have been described (i.e. Mox and Mhem) and have been reviewed elsewhere (Colin, Chinetti-Gbaguidi and Staels 2014).

HOW INTRACELLULAR PATHOGENS MODULATE MACROPHAGE POLARIZATION

Bacteria have evolved to become better pathogens by manipulating host defenses. Due to the primary role of the macrophage in host innate immunity to intracellular bacteria, evasion of macrophage function is especially important. Pathogenic bacteria produce different molecules to escape the phagosome and enter the cytoplasm, inhibit maturation of the phagosome, avoid the immune system, or limit the anti-bacterial activity of immune cells and molecules. The manipulation of macrophage polarization states is emerging as an important pathogenesis mechanism of intracellular bacteria (Fig. 1). In the following sections we will explore specific examples of host-involved factors for macrophage polarization after invasion by several important human pathogens. The manipulation of M1 and M2 states in microbial pathogenesis, including specific bacterial effector molecules employed by important pathogens, will be further discussed.

Figure 1.

Schematic representation of the current understanding regarding macrophage polarization by intracellular pathogens. While a large number of bacteria drive polarization to an M2 phenotype (e.g. Shigella, Listeria, Legionella), other pathogens benefit from an M1 polarization (e.g. Salmonella, Ehrlichia and Chlamydia). Interestingly, Mycobacterium and Coxiella seems to differentially regulate polarization to either M1 or M2 phenotypes depending on their infection stage. To date, the role of macrophage polarization in the pathogenesis of Brucella and Burkholderia species is poorly described.

Mycobacterium tuberculosis (Mtb)

Tuberculosis (TB) has plagued mankind since ancient times and remains the leading cause of death from an infectious agent today. In 2018, TB caused 1.5 million deaths (Organization 2019) and is estimated to be a latent infection in approximately one-fourth of the global population. Risk of TB reactivation in those with latent infection is increasing due to conditions such as HIV co-infection, diabetes, ageing, malnourishment or other factors that compromise the immune system (Huante, Nusbaum and Endsley 2019). The rapid development of multi-drug resistant or even extremely drug resistant strains of Mtb, is also an important public health issue.

Alveolar macrophages are the first cells to encounter and phagocytize Mtb after inhalation of small droplets of aerosolized bacteria. In addition to be the first cells in contact with Mycobacteria, macrophages are also the preferred cellular host and site of replication. In the early stage of infection, Mtb infection promotes an M1 polarization state. Activation of signaling pathways downstream of PRR recognition (O'Halloran et al. 2014) triggers production of ROS and NO, which kill the internalized bacteria. M1 macrophages also participate in formation of the granuloma (Huang et al. 2015), a protective structure comprised of macrophages, multinucleated giant cells and T-lymphocytes. In an in vitro tuberculosis granuloma model, M1 macrophages were shown to promote granuloma formation. In TB patients, M1 macrophages were found in non-granulomatous lung tissues, while M2 macrophages were found predominantly in necrotic and non-necrotic granulomas (Huang et al. 2015). According to a computational model where the macrophage and granuloma polarization ratio are considered, the temporal dynamics of the granuloma polarization ratios are predicted to direct the tuberculosis granuloma outcome (Marino et al. 2015). Immune responses that change M1/M2 polarization then, could have an important impact on the development or maintenance of protective granuloma structures.

Tissue damage due to the M1-driven pro-inflammatory processes (e.g. IL-1β and IFN-γ) in response to Mtb infection initiate formation of granuloma structures. NF-κβ signaling was predicted by a computational-biology approach as a viable therapeutic target to promote M1 macrophage polarization in early infection (Marino et al. 2015). As chronic infection progresses; however, these M1 macrophages polarize toward an M2 type that regulates excessive inflammation and promotes tissue repair. This phenomenon of polarization plasticity can be reproduced in the in vitro tuberculosis granuloma model (Huang et al. 2015) and occurs concomitantly with metabolic changes in the macrophages. Following infection, Mtb-exposed macrophages shift from aerobic glycolysis to mitochondrial oxidative phosphorylation and glutamine metabolism over time (Shi et al. 2019).

The shift from M1 to M2 polarization is an important mechanism for the host to avoid extensive lung tissue damage. However, the loss of immune pressure by the M1- and Th1-derived molecules gives an advantage to Mtb, which thrives in a Th2/M2 immune environment (Verreck et al. 2004; Kahnert et al. 2006). The bias toward M1 macrophages and Th 1 cell-mediated immunity after BCG (Bacille Calmette-Guerin) vaccination has been proposed as a potential mechanism for protection from Mtb in the early years after immunization. Consistent with this postulate, ex vivo stimulation of PBMCs from 10 week old BCG-vaccinated infants with mycobacterial antigens activated transcription of M1 and downregulated M2, macrophage gene signatures (Fletcher et al. 2009). Among the multitude of mechanisms used by Mtb to survive in the macrophage, promotion of M2 polarization has been demonstrated to further contribute to evasion of host microbicidal activity.

The host-signaling pathways that modulate macrophage polarization in response to Mtb exposure are not well characterized to date. A role for the IRAK-M signaling intermediate in polarization of monocytes, macrophages, and lung epithelial cells has been described. IRAK-M functions as a negative regulator of PAMP-TLR signaling in these cell populations. This kinase inhibits phosphorylation of the IRAK-1 and 4 kinases, leading to restriction of immune-mediated tissue damage (Kobayashi et al. 2002). IRAK-M is proven to drive M2 polarization during Mtb infection, leading to less tissue damage, but facilitating mycobacterial survival (Shen et al. 2017). Consistent with this observation, IRAK-M was positively correlated with bacterial load in Mtb-infected macrophages and in human lung tissue. Moreover, IRAK-M knockdown induced an M1 polarization type (Shen et al. 2017).

Roles for B cells and cytokines in the polarization of macrophages following Mtb infection have been described. In both human and murine systems, exposure to Mtb has been shown to activate production of B cell-mediated type I interferon that modulates macrophage polarization towards regulatory/anti-inflammatory or M2, states (Bénard et al. 2018). Activation of this B cell IFN-I signature following Mtb exposure is STAT-dependent and is further demonstrated to be positively regulated by activation of the cGAS/STING and Mincle PRR pathways. The anti-inflammatory cytokine IL-37 is also involved in M2 polarization in the setting of Mtb infection. Patients with TB have increased levels of IL-37, and those levels are reduced after treatment. Increased expression of IL-37 induced a shift towards M2, along with an upregulation of TGFβ, arginase-1 and IL-10 transcription (Huang et al. 2015).

Host microRNAs further regulate the polarization of M1 toward M2 following Mtb infection. The microRNA-26a (miR-26a) has been shown to target the KLF4 transcription factor (Sahu et al. 2017), which functions to induce M2 polarization (Liao et al. 2011) and prevent Mtb movement into lysosomes. Downregulation of miR-26a during Mtb infection activates KLF4 transcription, promoting M2 polarization as evidenced by increased arginase and decreased iNOS production. Further, Mtb infection has been shown to activate CREB-dependent synthesis of C/EBPβ as a mechanism for increased arginase production associated with the M2 bias (Sahu et al. 2017).

The importance of M2 polarization for Mtb survival in host macrophages is reflected by the stronger polarization bias driven by virulent strains. The ESAT-6 molecule produced by virulent strains of Mtb induces a stronger M2 polarization (e.g. STAT3, STAT6, arginase-1 and KLF4), in comparison to the attenuated (i.e. H37Ra) strain that lacks ESAT-6, which induces an M1 bias (i.e. STAT1, iNOS and NICD) (Lim et al. 2016). Treatment of H37Ra infected macrophages with ESAT-6 also shifts the polarization from M1 to M2, demonstrating the importance of ESAT-6 in macrophage polarization during Mtb infection.

Virulent Mtb is also known to induce less apoptosis and more necrosis than attenuated Mycobacteria (Briken and Miller 2008; Behar et al. 2011). Cell death via necrosis, when combined with induction of an M2 anti-inflammatory environment, provides a favorable environment for Mtb proliferation. Apoptosis and M1-modulated environments promote stronger T cell responses and less favorable conditions for Mtb growth. Interestingly, in M1-infected macrophages, the endoplasmic reticulum (ER) stress is upregulated and this upregulation controls infection through an apoptosis-dependent mechanism. This ER stress-driven apoptosis response is more often observed in M1, as compared to M2, macrophages (Lim et al. 2016).

Microbial serine proteases could also influence the macrophage shift toward M2 polarization following activation of the protease-activated receptors (PARs) in the cell. Human monocytes differentiated ex vivo from tuberculin-reactive donors were used to demonstrate that serine-proteases (thrombin and trypsin) triggered the proteolytic activation of PAR1/2 receptors, inducing IL-4 release and upregulation of MR (CD206), another well described M2a phenotype marker (García-González et al. 2019). Mtb heat shock proteins appear to have a growing role in macrophage polarization following infection. A bacterial orthologue of mammalian heat shock protein 70 (DnaK) produced by Mtb can polarize murine macrophages to an M2-like phenotype (Lopes et al. 2014). This polarization is driven in an IL-10 dependent manner that increases the expression of M2 markers Ym1 and Fizz, as well as TGFβ (Lopes et al. 2016). Similarly, the Mtb heat shock protein (Mtbhsp60) also promotes M2 polarization by activating expression of host IL-10 (Parveen et al. 2013). Moreover, Mtb Heat shock protein 16.3, (Hsp 16.3) plays a role for Mtb survival during latent infection (Yuan, Crane and Barry 1996). Hsp16.3 is upregulated by the dosR regulon of Mtb and is involved in M2 polarization through CCRL2 and CX3CR1 chemokine receptors (Zhang et al. 2019).

Mycobacterium leprae

Leprosy is a neglected, chronic infectious disease, caused by the obligate intracellular bacteria Mycobacterium leprae. The earliest clinical manifestation of the disease and its associated prognosis is the characteristic paucibacillary tuberculoid form of the disease. This form of leprosy effectively restricts bacterial growth through activation of type 1 cytokines such as IFN-γ (Stefani et al. 2003) and through polarization of macrophages to the antimicrobial M1 phenotype (Salgame et al. 1991; Yamamura et al. 1992; Montoya et al. 2009). In contrast, multibacillary (or lepromatous) progressive leprosy skews polarization to an M2 state (Benoit, Desnues and Mege 2008), through production of IL-10 (Mege, Mehraj and Capo 2011).

In lesions from patients who experience spontaneous conversion from multibacillary to the paucibacillary and self-limited form of the disease, there is also a repolarization from M2 to an M1 program. Interestingly, exposure to live M. leprae promotes an M2 bias, regardless of whether the strain originated from tissue of a patient suffering from the tuberculous or lepromatous lesions (Ma et al. 2018).

In addition to the anti-inflammatory M2 polarization following M. leprae infection, exposure of healthy donor monocytes to M. leprae interferes with M1 maturation. The bacterial phenolic glycolipid PGL1 has been implicated as a mediator of M2 bias through in vitro studies where PGL1 was added exogenously to cultured macrophages. Similarly, this glycolipid decreased M1 macrophage cytokine/chemokine responses (Fallows et al. 2016). Moreover, the endothelial cells that allow the passage of monocytes into the infected tissues, despite lack of immune stimulation, normally trigger differentiation and polarization toward an M2 phenotype (He et al. 2012).

The process of apoptotic cell removal, or efferocytosis, is an important contributor to polarization. This role was demonstrated by polarizing macrophages from human PBMCs toward M1 or M2 phenotypes with GM-CSF or M-CSF respectively, and culturing them with M. leprae in the presence or absence of apoptotic bodies (De Oliveira Fulco et al. 2014). For M2 macrophages, there was no alteration of phenotype, except for an altered expression of TGF-β. Interestingly, efferocytosis induced a repolarization of M1 macrophages to M2 after exposure to bacteria. This response occurred simultaneously with increased phagocytic capacity, greater expression of CD16, IL-10, TGF-β and SRA-I and increased bacterial survival. The increased phagocytic activity of these M2 macrophages further resulted in accumulation of host-derived lipids, which promote mycobacterial growth (Ridley and Russell 1982; Cruz et al. 2008).

To counter this M2 skew that favors M. leprae fitness, the host directs M1 polarization through a Jagged 1 (JAG1)-dependent mechanism within the Notch signaling pathway. In tuberculous leprosy (the self-limiting form), JAG-1 is found localized preferentially to the vascular endothelium and promotes M1 polarization following activation of endothelial cells by IFN-γ (Kibbie et al. 2016).

Lastly, CXCL4-derived M4 macrophages have also been described in leprosy patients and are implicated as a factor associated with poor clinical outcomes (De Sousa et al. 2018). Analysis of tuberculoid and lepromatous patient samples by immunohistochemistry in one report described a predominance of an M4 macrophage phenotype in lesions of patients with lepromatous-multibacillary forms of disease compared to those with paucibacillary disease (De Sousa et al. 2018).

Salmonella

The Salmonella genus contains a wide spectrum of bacteria types including two species (S. enterica and S. bongori), 6 subspecies and over 3000 serovars. Salmonella enterica serovar Typhimurium is an important cause of human foodborne illness, responsible for 93.8 million cases of illness and 155,000 deaths per year globally (Scallan et al. 2011; Eng et al. 2015). In the United States alone, these bacteria cause over 1 million human cases and an economic burden of $3.7 billion annually (Scallan et al. 2011; Hoffmann, Maculloch and Batz 2015). The adaptability to diverse host cells permits a range of manifestations, from typhoid fever to gastrointestinal complications. This intracellular pathogen can infect several human cell types, including dendritic cells, macrophages, epithelial cells and M cells.

The importance of macrophage function in the host response to S. Typhimurium is reflected by a multitude of strategies evolved to evade and manipulate macrophage function. The S. Typhimurium genome possesses multiple pathogenicity islands, which, when expressed, produce proteins used in the entry, evasion, modulation and persistence processes important for infection. Salmonella Pathogenicity Island 1 (SPI-1) is responsible for the entry of the bacterium into macrophages. Salmonella Pathogenicity Island 2 (SPI-2) contributes to the vacuole formation, immune response evasion, host cell modulation and replication of the bacteria in the intracellular compartment. SPI-2 effectors, which are translocated via the Type Three Secretion System (T3SS), play a known role in polarization toward an M2-like phenotype following S. Typhimurium infection (Jaslow et al. 2018). Recently, the SPI-2 effector protein SteE, also known as Salmonella anti-inflammatory response activator (SarA), was shown to regulate polarization. Interestingly, investigations with a mutant strain demonstrated that repression of M1, and polarization of M2, by SteE occur through independent mechanisms and are refractive to effects of exogenous IFN-y (Stapels et al. 2018). The mechanism of macrophage polarization is caused by activation of STAT3 transcriptional targets by SteE, inducing increased levels of IL-10 production. SteE was also shown to contribute to intracellular replication in vitro and to tissue bacterial loads in a murine model (Jaslow et al. 2018). This effector protein appears to be a primary factor in the induction of M2-like polarization for S. Typhimurium in both in vitro cell lines and in a murine model of infection.

This mechanism was recently characterized and showed how the M2 polarization is achieved via the STAT3 pathway. Once translocated successfully, SteE interacts with the host-pleiotropic serine/threonine kinase, GSK3. This interaction leads to the phosphorylation of SteE, and the GSK3 in the resulting complex is driven to interact with and phosphorylate the substrate transducer and activator of transcription 3 (STAT3) on tyrosine-705. STAT3 can then translocate to the nucleus and drive the anti-inflammatory macrophage polarization program (Panagi et al. 2020).

Investigations using a mouse model of persistent infection showed that by day 42 post challenge, most macrophages infected with S. Typhimurium were M1 as indicated by expression of iNOS and CXCL9/10 (Goldberg et al. 2018). CXCL10 is the IFN-γ-inducible protein 10, a chemokine secreted from cells stimulated with type I and II IFNs and LPS. The results of this persistent infection model differ from previously cited reports from short term in vitro studies. The intra-macrophage S. Typhimurium bacteria were found to contain increased mRNAs of genes regulating intracellular survival (phoP, sifBandsseJ), reactive oxygen species detoxification (katG), and sensing and detoxification of reactive oxygen species (hcp, ytfE, hmpA, norV). These transcriptional modifications allow the bacteria to survive intracellularly in a dormant state in an iNOS-producing cell by neutralizing host cell molecules that are normally toxic (Goldberg et al. 2018). Thus, other bacterial effector molecules enable the intracellular survival of S. Typhimurium independent of the macrophage polarization state. In some stages of infection, multiple mechanisms including M2 bias may collaborate to favor S. Typhimurium survival and propagation.

Coxiella burnetii

Q fever is an infectious disease caused by inhalation of the obligate intracellular bacteria Coxiella burnetii. This bacterium is a CDC category B select agent, which is transmitted as a zoonosis from aerosolization of infected secretions following close contact with ruminants. There are two types of Q fever: acute and chronic. Acute Q fever is characterized by flu-like and pulmonary symptoms. Protective immunity during the acute stage is associated with pro-inflammatory cytokine activation and development of granulomas in peripheral organs. The chronic form of Q fever is mainly manifested as endocarditis after months to years of acute infection. Those with chronic Q fever lack protective granulomas, which are replaced with lymphocyte infiltration and necrotic foci (Maurin and Raoult 1999). This chronic form is presented with more frequency in pregnancy or in patients with history of valvulopathy or immunosuppression.

Coxiella burnetii preferentially infects monocytes and macrophages and survives better in cells conditioned by an M2 polarization environment (Amara, Bechah and Mege 2012). Infection with C. burnetii stimulates an atypical M2 activation program in human macrophages, including in both monocyte-derived macrophages (MDMs) and alveolar macrophages (AM) (Benoit, Desnues and Mege 2008; Dragan, Kurten and Voth 2019). This atypical activation is characterized by expression of M2 factors including TGF-β1, IL-1ra, CCL18, mannose receptor and arginase. Interestingly, moderate expression of M1 factors including IL-6 and CXCL8 (IL-8) are still observed, although expression is significantly reduced compared to levels observed following exposure to classical M1 stimuli such as LPS (Benoit et al. 2008).

Coxiella burnetii survives without replication in monocytes; however, activation with the M2-polarizing cytokine IL-4 promotes bacterial propagation (Ghigo et al. 2003). In further support of permissiveness by M2, C. burnetii replication is restricted in alveolar macrophages (AM) of IL4^−/− mice, while AM from mice deficient in M1 molecules, such as Nos2^−/− or Ifng^−/− mice, are more permissive to replication (Fernandes et al. 2016). In contrast to this polarization displayed by AM, murine bone marrow-derived macrophages (BMDMs) lack polarization (M0) during both the non-infected phase or the phase II infection phase with C. burnetii (Cockrell et al. 2017).

In addition to activation of cytokines that drive classical polarization paradigms, C. burnetii infection also upregulates expression of host Vanin-1. This molecule is a host tissue epithelial cell membrane-anchored pantetheinase, an enzyme that controls tissue inflammation, regulates thymus homing, promotes granuloma formation and regulates macrophage polarization. Vanin-1-deficient mice have decreased granuloma formation in both the liver and spleen, while bacterial clearance is not affected (Meghari et al. 2007). The Vanin-1-deficient mice also display increased resistance to a systemic oxidative stress and reduced inflammatory reactions. As granuloma formation depends on macrophage recruitment and activation, these deficient animals also present decreased expression of iNOS and MCP-1 (M1 markers) and increased IL-10 and arginase (M2 markers). These results demonstrate that modulation of Vanin-1 polarizes the macrophages toward M2 during C. burnetii infection, affecting granuloma formation and promoting greater bacterial dissemination.

The polarization state observed following infection depends on the type of Q fever (acute or chronic). During chronic Q fever, macrophages are unable to control C. burnetii replication due to inhibition of microbicidal competence associated with overproduction of IL-10 (Honstettre et al. 2003). Similarly, overexpression of IL-10 in transgenic mice enables bacterial persistence, mimicking chronic Q fever (Meghari et al. 2008). In support of these observations, IL-10 and another important M2 cytokine, TGF-β, are released by monocytes from patients with Q fever endocarditis (Capo et al. 1996). Also during acute Q fever, it has been observed that C. burnetii stimulated an M1 profile in monocytes, resulting in control of bacterial replication (Benoit et al. 2008). In contrast, an atypical M2 polarization of macrophages that is associated with moderate replication and persistence of the bacteria is observed in acute Q fever (Benoit, Desnues and Mege 2008).

Patients with valvulopathy, a risk factor for chronic Q fever and endocarditis, have greater numbers of circulating apoptotic leukocytes during chronic Q fever, and depending on the presence of valvulopathy, uptake of apoptotic lymphocytes by monocytes and macrophages also redirects macrophage polarization towards the non-protective M2 phenotype, thereby increasing C. burnetii replication (Benoit, Desnues and Mege 2008). Monocytes/macrophages become polarized towards M2 following experimental cell contact with these apoptotic cells. Moreover, in vitro neutralization of M2 cytokines (e.g. IL-10 and TGF-β) prevented replication of Coxiella. Interestingly, in contrast to apoptotic uptake, binding of necrotic cells to monocytes or macrophages polarized the cells toward M1 and led to killing of bacteria. These results have led to speculation that, in the presence of valvulopathy, the circulating apoptotic leukocytes promote an M2 bias that favors C. burnetii replication and Q fever endocarditis. In the absence of valvulopathy, necrosis directs polarization towards a protective M1 response and C. burnetii killing, leading only to acute Q fever (Benoit et al. 2008). To date, however, the specific bacterial effector molecules used by Coxiella to drive M2 polarization are not defined.

Chlamydia

The Chlamydia genus contains twelve species including four that are pathogenic to humans: Chlamydia trachomatis, C. pneumoniae, C. abortus and C. psittaci. Over 1.7 million cases of chlamydia were reported by the US CDC in 2018, which represents a 3% rise from the previous year. Chlamydia trachomatis alone is the leading cause of sexually transmitted bacterial infections, as well as the leading infectious cause of blindness. Chlamydiae are Gram-negative obligate intracellular bacteria that infect epithelial cells, macrophages, fibroblasts and dendritic cells (Ying et al. 2007). Chlamydia infections generally occur via two routes: sexual contact, ocular exposures, or transmission through the respiratory tract. These parasitic bacteria lack normal biosynthetic pathways, forcing them to use their host cell machinery for energy production and other metabolic tasks. In this regard, they resemble a virus more than a bacterium (Becker 1996).

Chlamydia pneumoniae infects the respiratory tract and patients generally display mild or no symptoms. However, it can result in more severe upper and lower respiratory complications, such as pharyngitis, bronchitis, sinusitis and pneumonia (Kuo et al. 1995; Porritt and Crother 2016). Chlamydia can infect and survive in all macrophage types, including the hostile intracellular environment of M1. If placed in previously polarized cells, however, replication only occurs in M2 macrophages. It has been shown that C. pneumoniae possess a preferred tropism for M2-like macrophages over M0 or M1-like macrophages (Buchacher et al. 2015).

A role for inhibitors of apoptosis proteins (IAPs) in M1/M2 polarization has been suggested through studies of C. pneumoniae. An IAP-dependent polarization to the M1 phenotype was demonstrated in macrophages infected with C. pneumoniae. Furthermore, it was revealed that IAP deficiency caused macrophages to be refractory for immune stimulation with C. pneumoniae and promoted repolarization from M1 to M2 effector macrophages (Nadella et al. 2017).

Observations of C. muridarum infection, commonly used in animal models of human Chlamydia infections, demonstrated that infection of M0 macrophages did not induce M2 nor M1 macrophage-associated genes. In agreement with the observations of C. pneumoniae, C. muridarum is able to survive in all polarized states and yet preferentially replicates in M2-like macrophages (Gracey et al. 2013).

Nutrient availability may also play a mechanistic role in polarization following Chlamydia infection. IFN-γ, an M1 macrophage polarizing cytokine and marker, plays a role in modulating the growth and persistence of Chlamydia. This occurs via the IFN-γ induced enzyme indoleamine 2,3-dioxygenase and its depletion of the essential amino acid tryptophan. Using C. trachomatis, it was proposed that the IFN-γ-mediated chlamydial persistence and growth inhibition is due to differences in the intracellular tryptophan pool reserves between polarized and non-polarized macrophages (Kane et al. 1999).

Rickettsiales

The Rickettsiales order contain endosymbiont alphaproteobacteria that rely on eukaryotic cells for survival, infecting arthropods, nematodes and mammals. Several Rickettsiales species across genera (Ehrlichia, Orientia and Rickettsia spp.) modulate macrophage polarization, directly or indirectly, as a mechanism of bacterial pathogenesis.

Ehrlichia

The Ehrlichia genus is comprised of seven species, including the important human pathogen Ehrlichia chaffeensis. Transmission of E. chaffeensis occurs primarily via infected ticks and causes human monocytotropic ehrlichiosis (HME). The manifestations of HME range from subclinical disease to life-threatening conditions including multi-organ failure. Ehrlichia chaffeensis is an obligate intracellular pathogen that preferentially infects macrophages. Like other rickettsial species, it lacks lipopolysaccharide and peptidoglycan in its membrane.

After a tick bite, the bacterium infects a mononuclear phagocyte via a receptor-mediated process. The surface protein EtpE binds to a GPI-anchored protein, DNase X and initiates entry in the host cell (Mohan Kumar et al. 2013). The pathogen replicates in host cell cytoplasmic vacuoles and forms morulae microcolonies (Ismail and McBride 2017). Multiple mechanisms are employed by the bacteria to subvert the host defenses in order to survive and replicate in these modified vacuoles (Paddock and Childs 2003; Luo et al. 2011). Host cell signaling pathway manipulation by E. chaffeensis has been well characterized, which allows it to survive, establish colonization, downregulate the immune response and invade neighboring cells (Rikihisa 2006; Nandi et al. 2009; Rikihisa 2010; Thomas, Popov and Walker 2010; Dunphy, Luo and McBride 2013; Alves et al. 2014; Yan et al. 2018).

A recent study employing E. muris and Ixodes ovatus ehrlichia (IOE) infection models demonstrated that macrophage polarization from lethal and sublethal infections is achieved via an mTORC1-dependent manner. Ehrlichia muris is a human pathogen, while IOE causes disease in mice similar to the human disease from an E. chaffeensis infection (Munderloh et al. 2009). mTORC1 is a protein complex that has roles in metabolism, redox reactions, and protein synthesis regulation. Once activated, mTORC1 will phosphorylate autophagy-related protein 13, stopping it from joining the ULK1 kinase complex. This causes the ULK1 kinase complex to be unavailable to the pre-autophagosomal structure, thus leading to autophagy inhibition.

A mild and lethal ehrlichiosis murine model of E. muris and IOE infection, respectively, was used to investigate macrophage polarization outcomes. The IOE infected mice produced an M1-like polarization of macrophages caused by mTORC1 activation, while the E. muris infected mice favored an M2-like macrophage polarization that was linked to the suppression of mTORC1 activation (Haloulet al. 2019). In the setting of E. muris, these results provide in vivo evidence that inhibition of autophagy (mTORC1 activation) promotes M1 macrophage polarization while activation of autophagy (mTORC1 suppression) favors an M2 bias (Harris, et al. 2017; Claude-Taupin, et al. 2018; Liu 2019). These results highlight the potential contribution of infectious dose and autophagy as mediators of macrophage polarization and may be important factors when considering the behavior of low dose persistent bacteria.

Orientia

Orientia tsutsugamushi is an obligate intracellular bacterium that causes the disease scrub typhus. First isolated in 1930 (Seong, Choi and Kim 2001), O. tsutsugamushi has recently reemerged as a pathogen of concern due to disease of livestock in Asia. There is a growing number of human clinical infections showing that the pathogen is now endemic in the Asia Pacific region, with over one billion people at risk (Al Aminet al. 2019; Wangrangsimakulet al. 2019). Unique features, specifically the lack of a peptidoglycan wall in the cell membrane, led to its classification as a non-traditional, Gram-negative bacteria. O. tsutsugamushi can invade endothelial cells, macrophages and polymorphonuclear leukocytes (Rikihisa and Ito 1982; Choet al. 2000) through mechanisms including induced phagocytosis (Urakami, Tsuruhara and Tamura 1983). This occurs when a pathogen possesses a feature that promotes phagocytosis, such as membrane-bound opsonins that facilitate engulfment by an immune phagocyte. Unlike many intracellular bacteria, once inside the macrophage, O. tsutsugamushi induces an M1 macrophage phenotype, causing increased levels of inflammation (Tantibhedhyangkulet al. 2011; Tantibhedhyangkulet al. 2013; Ogawaet al. 2017). This has been suggested as the cause for the inflammation that manifests during scrub typhus disease (Tantibhedhyangkulet al. 2013). Alteration of the acidification process in the phagosome by O. tsutsugamushi has also been shown to affect induction of an immune response in macrophages (Ogawaet al. 2017).

The role of IL-1 receptor signaling is important in the development of a productive host immune response to O. tsutsugamushi (Koo, et al. 2012). A key player in this inflammatory process is the inflammasome, which activates the cysteine protease caspase-1 in the cytosol. Intracellular bacteria activate cytosolic PRR such as NOD that promote inflammasome assembly. Orientia tsutsugamushi escapes the phagosome and replicates in the cytosol through poorly understood mechanisms associated with a loss of the phagosomal membrane (Ewing et al. 1978). Activation of caspase 1 stimulates the production of bioavailable interleukin-1β (IL-1β) in response to O. tsutsugamushi infection (Koo et al. 2012). IL-1β is a mediator of inflammatory responses that, along with NO, are associated with the M1 macrophage phenotype. Surprisingly, it has recently been shown that O. tsutsugamushi propagates at increased levels and survives in the presence of NO in murine macrophages, even following several days of infection (Ogawa et al. 2017). Despite the growing evidence that O. tsutsugamushi promotes macrophage polarization after infection, the bacterial effector molecules that mediate these functions are still unknown.

Rickettsia

The Rickettsia genus is comprised of Gram-negative bacteria that primarily infect endothelial cells and produce a variety of severe clinical manifestations, such as acute fevers and typhus (Fang, Blanton and Walker 2017). These obligate intracellular pathogens are transmitted via arthropods, but often have a vertebrate host (Ismail and McBride 2017). The genus is split into four groups according to their antigenic characteristics: typhus group, spotted fever group (SFG), transitional group and ancestral groups. Among groups within this genus, macrophage polarization has only been described in the context of members of the SFG.

Rickettsia conorii, the causative agent of Mediterranean spotted fever, induces an M2-like phenotype in macrophages. Using quantitative proteomics, it was revealed that processes such as glycolysis and the pentose phosphate pathway (PPP) are reduced. As a result, the production of products, such as pyruvate, ribose, nucleotides and ROS from the PPP, is decreased (Curto et al. 2019). Together, these outcomes limit pro-inflammatory signaling and the resulting polarization to the M1 phenotype. In contrast to these effects, activation of the TCA cycle has also been observed. R. montanensis, a non-pathogenic strain in humans, displays different effects on metabolic processes in the absence of development of an M2 bias (Curto et al. 2019). Similar effects on pyruvate production as on glycolysis were noted; however, no changes in the levels of the enzymes for fatty acid β-oxidation, glutaminolysis, the TCA cycle, the respiratory complex or mitochondrial transporters were observed (Curto et al. 2019). These metabolic changes limit inflammatory responses and promote development of the M2 phenotype. Comparative proteomics with other bacteria that induce M2 phenotypes are needed to determine if this phenomenon is restricted to R. montanensis or reflects a more commonly used strategy of bacterial pathogenesis. These proteomic studies are important to further our understanding of the role of cellular metabolism to promote an immune polarization bias in the in the context of intracellular bacterial infection.

Burkholderia

Burkholderia pseudomallei is an intracellular Gram-negative saprophyte and the causative agent of melioidosis. This infectious disease causes over 89,000 deaths per year and has mortality rates of ∼40% despite treatment. The bacteria can infect a variety of cell types, including neutrophils, macrophages, dendritic cells, fibroblasts and keratinocytes, but shows a preference for macrophages (Whiteley et al. 2017). Burkholderia pseudomallei employs an impressive array of virulence mechanisms to overcome the human immune response. Like several other intracellular pathogens, B. pseudomallei utilizes a T3SS to inject effector proteins into the host immune cell and subvert its defenses while establishing colonization.

Survival of Bpm in macrophages has been firmly established, and the mechanistic basis for survival has been extensively investigated (Jones, Beveridge and Woods 1996). Burkholderia pseudomallei modulated host cell autophagy, similar to other intracellular pathogens (Cullinane et al. 2008; Devenish and Lai 2015; Krakauer 2018). Another mechanism is the alteration of the Lipid A region of bacterial lipopolysaccharide (LPS) during various growth conditions, which provides evidence for the modulation of LPS during chronic infections (Norris et al. 2018). One study showed that the acquisition of nitrate in an anaerobic environment is an important factor for B. pseudomallei survival (Pinweha et al. 2018). The ability to modulate the host iron balance has been to shown to improve intracellular survival (Schmidt et al. 2018). Modulation of the mitogen-activated protein kinases occurs with in vitro and in vivo B. pseudomallei infection, although sufficient data has not been accumulated to characterize the complete mechanism (D'Elia et al. 2017). Lack of superoxide dismutase C results in lower B. pseudomallei survival in vitro in both human and murine models (Vanaporn et al. 2011). Production of a proteasome inhibitor increases the survival and growth of B. pseudomallei (Wagley et al. 2017). Sphingosine-1-phosphate is also required for B. pseudomallei survival (Custódio et al. 2016) as well as the presence of functional type 3 and 6 secretion systems (T3SS and T6SS, respectively).

While there are many protein, lipid and carbohydrate molecules from the host and the pathogen that play a role in B. pseudomallei survival, currently there are no known studies that fully characterize the polarization of macrophages after B. pseudomallei infection. However, studies have shown that the decreased production of NO and 8-iso-PGF_2α, a bioactive product of free radicals that induces lipid peroxidation, is found in macrophages in vitro and from patients with melioidosis. This could suggest that B. pseudomallei is directing the macrophages to an M2-like phenotype (Nathan, Qvist and Puthucheary 2005) (Utaisincharoen et al. 2001).

Listeria

The Listeria genus contains 20 species that are facultative anaerobic intracellular bacteria. Of these species, Listeria monocytogenes is a major human pathogen and the causative agent of listeriosis. The infection is primarily acquired via ingestion of contaminated food, and can cause gastroenteritis, as well as perinatal and systemic infections (Drevets and Bronze 2008). Listeria monocytogenes enters the responding macrophages through bacterial surface proteins or phagocytosis and escapes the phagosome into the cytoplasm. Here it is free to replicate inside the cell, while avoiding external humoral immune mediators (Shaughnessy and Swanson 2007). Using host cell components, it promotes actin polymerization, which it utilizes to travel from cell to cell (Cheng et al. 2018).

Listeria monocytogenes infection of macrophages induces an M1-like polarization. It has also been shown to promote the change from an M2 to an M1 phenotype in a pre-polarized macrophage. This was shown consistently during studies using the L. monocytogenes platform as a therapeutic to activate tumor- associated macrophages (Rai et al. 2012; Lizotte et al. 2014; Xu et al. 2020). The mechanism of macrophage polarization upon L. monocytogenes infection was revealed to occur via the Notch signaling pathway as also observed in Ehrlichia infections (Xu et al. 2012). RBP-J is an important nuclear transducer of Notch signaling, and it is responsible for amplifying the Toll-Like Receptor 4 (TLR4) induced expression of mediators of classically activated M1 macrophages. An RBP-J dependent activation of the transcription factor IRF8 was further identified as an important signaling event that regulates the pro-inflammatory milieu and M1 polarization following L. monocytogenes infection (Xu et al.2012).

CONCLUSION

Here, we summarized the current understanding of macrophage polarization in immunity to, and pathogenesis of, important intracellular bacterial pathogens. The current state of knowledge for the role of M1 and M2 bias in bacterial disease is limited to date, ranging from preliminary identification of some cytokine mediators that could influence bias, to partially developed mechanistic models of M1/M2 regulation at different stages of disease. Our greatest gap in knowledge is an incomplete understanding of the molecular mechanisms by which bacterial pathogens influence macrophage polarization. The role of polarization to direct protective or pathogenic immune responses to several bacterial pathogens, and the immune regulatory determinants of polarization, also remain to be fully described. Research directed to address these gaps at the bacterial/host interface are needed to improve our understanding of macrophage–pathogen interactions and inform development of clinical interventions that shape innate and adaptive immune responses through modulation of macrophage polarization.

Contrary to previous assumptions based on metabolic needs of the pathogen, no clear pattern of polarization has been observed between strictly obligate and facultative intracellular pathogens. As summarized in Table 1, polarization is not determined by association with the intracellular localization within the macrophage, and the use of a pathogen-adapted vacuole/endosome does not appear to result in a consistent macrophage polarization phenotype. Another important point to emphasize is the relationship of pathogenic versus non-pathogenic effects on polarization in response to exposure to genomically similar species. Ehrlichia muris and IOE cause the opposite effect in macrophages, as observed with similar comparisons of virulent Rickettsial and Mycobacterial species with non-virulent counterparts. Comparisons within and among these species suggest a pattern toward induction of M2 bias by virulent isolates, although a greatly expanded analysis is needed. An alternate approach to understand polarization may come through analysis of host metabolic changes during disease progression. As shown following Rickettsia infections, a shift in metabolic preferences determines M1 and M2 polarization (Curto et al. 2019). As demonstrated with Mycobacteria, metabolic changes that occur as the infection progresses from an acute to a chronic state may shift macrophages from M1 to M2 phenotypes (Shi et al. 2019). Expanded studies in well-defined models of acute and chronic disease are needed to determine the role of metabolic shifts that differentially polarize macrophages for disease progression.

Table 1.

Polarization of Intracellular Bacterial Pathogens.

Organism	Species	M1 or M2 macrophage phenotype	Proposed mechanism	Cell model	Obligative vs Facultative	Replication Location	References
Burkholderia	B. pseudomallei	Unknown	Unknown	–	Facultative	Cytoplasm	(Whiteley et al. 2017)
Mycobacterium	M. tuberculosis	M1 acute, M2 chronic	Various from host. Bacterial: ESAT-6, DnaK, serine proteases, Mtbhsp60	Human, Murine	Obligative	Vacuole and cytoplasm	(Parveen et al. 2013; Lopes et al. 2014; Huang et al. 2015; Lim et al. 2016; García-González et al. 2019)
	M. leprae	M1 tuberculoid, M2 lepromatous	Host JAG-1 for M1 Bacterial PGL1 for M2	Human	Obligative	Cytoplasm	(Salgame et al. 1991; Yamamura et al. 1992; Montoya et al. 2009; Fallows et al. 2016)
Orientia	O. tsutsugamushi	M1	Unknown	Murine	Obligative	Cytoplasm	(Tantibhedhyangkul et al. 2011; Tantibhedhyangkul et al. 2013; Ogawa et al. 2017)
Ehrlichia	E. muris	M2	mTORC1 pathway suppression	Murine	Obligative	Vacuole	(Paddock and Childs 2003; Munderloh et al. 2009; Luo et al. 2011; Haloul et al. 2019)
	Ixodes ovatus ehrlichia	M1	mTORC1 pathway activation	Murine	Obligative	Vacuole	(Haloul et al. 2019)
Salmonella	S. Typhimurium	M2	T3SS effectors SarA/SteE induce IL-10 production	Human,	Facultative	Vacuole	(Goldberg et al. 2018; Jaslow et al. 2018; Stapels et al. 2018)
Chlamydiae	C. pneumoniae	M2	Unknown	Murine	Obligative	Vacuole	(Kane et al. 1999; Gracey et al. 2013; Buchacher et al. 2015; Nadella et al. 2017)
	C. muridarum	No polarization	Unknown	Murine	Obligative	Vacuole	(Gracey et al. 2013)
	C. trachomatis	Unknown	Unknown	Human	Obligative	Vacuole	(Kane et al. 1999)
Coxiella	C. burnetii	M1 acute, M2 chronic	Unknown	Human	Obligative	Vacuole	(Benoit et al. 2008; Amara, Bechah and Mege 2012; Dragan, Kurten and Voth 2019)
Legionella	L. pneumophila	M1	Two proposed mechanisms	Murine	Facultative	Vacuole	(Kusakaet al. 2018; Nanjo et al. 2018)
Rickettsia	R. conorii	M2	Unknown	Human	Obligative	Cytoplasm	(Curto et al. 2019)
	R. montanensis	M1	Unknown	Human, Monkey	Obligative	Cytoplasm	(Curto et al. 2019)
Borrelia	B. burgdorferi	M2	Polarization specific to tissue location	Murine	Obligative	Unknown	(Lasky, Olson and Brown 2015)
Listeria	L. monocytogenes	M1	Notch-RBP-J signaling pathway	Murine BMDMs	Facultative	Cytoplasm	(Xu et al. 2012)
Shigella	S. dysenteriae	M1	Porin activates TLR2 and 6	Murine	Facultative	Cytoplasm	(Biswas et al. 2007)
Brucella	B. abortus	M0-Impaired polarization	LC3-dependant Autophagy	Human PBMCs	Facultative	Vacuole	(Wang et al. 2017)

The use of different experimental systems should also be carefully considered for interpretation and comparisons of results. The animal models, as well as the species origin and phenotype of cells and cell lines differs among the various investigations. The baseline phenotype (i.e. M0 or M1) and pre-polarization or polarization conditions also differed among several studies, as did the approaches for infection. Polarization of macrophages can be influenced by many factors, including cytokines, chemokines, transcription factors, SOCS proteins and amino acid metabolism amongst others. Investigators working with macrophages or macrophage cell lines should carefully consider the way in which differentiating cytokines, activation conditions, and how cells are processed and maintained may affect polarization. Murray et al., began to catalogue how various stimuli influence macrophage function to standardize methods in the field (Murray et al. 2014). Carefully recording and publishing these methods and procedure details enhances the field as it moves towards a comprehensive understanding. The choice of approaches and conditions could impact the interpretation of plasticity, especially as compared to the more straightforward changes in polarization state that occur due to initial exposure to bacterial stimuli.

Long term, understanding the mechanisms by which macrophage polarization develops or changes during infection may be key to the identification of new targets for intervention to combat disease due to intracellular bacterial pathogens. Due to the roles of M1 and M2 bias for antimicrobial activity and immune regulation, respectively, identifying mechanisms of polarization is also important for combatting antimicrobial resistance and immune mediated pathologies.

Conflict of interest. None declared.