Søren R. Paludan, Institute of Medical Microbiology and Immunology, University of Aarhus, The Bartholin Building, DK-8000 Aarhus C, Denmark. e-mail: email@example.com
The innate immune system constitutes the first line of defense against infections and is also important for initiating the development of an adaptive immune response. The innate immune system recognizes microbial infection through germline-encoded pattern recognition receptors, which are responsible for decoding the microbial fingerprint and activating an appropriate response against the invading pathogen. In this review, we present and discuss current knowledge on how the innate immune system recognizes intracellular pathogens, activates intracellular signaling, induces gene expression, and orchestrates the microbicidal response against pathogens with a habitat within host cells.
Microbial infections activate the host immune system, which aims at eliminating the incoming pathogen, but at the same time may cause harm to the host organism if excessive inflammation is induced (1). The early immune response is termed the innate response, and is activated through germline-encoded receptors, called pattern recognition receptors (PRRs), which recognize molecular patterns conserved through evolution in a wide range of pathogens (2). These molecules, called pathogen-associated molecular patterns (PAMPs), stimulate intracellular signaling, gene expression and hence activation of antimicrobial and inflammatory activities. The innate response therefore exerts a rapid first line of defense against the infection, but at the same time also initiates the process leading to eventual development of an adaptive immune response and establishment of immunological memory.
The role of the innate immune response in host defense against infections has been demonstrated in a wealth of animal studies. Recently, Dr. Casanova has identified humans with rare primary innate immune defects, which are associated with susceptibility to different microbial infections (3–5). These findings underscore the important roles of the innate immune system in early host defense, and also that in-depth knowledge of this part of the immune system may provide novel strategies to prevent and treat infections that cannot be cured by the treatments currently available.
In this review, we discuss how intracellular pathogens are recognized by the innate immune system and how this is relayed into the early cellular host response to infection. Finally, we discuss current knowledge on the mechanisms through which the innate immune system exerts microbicidal activity.
PRINCIPLES OF MICROBIAL RECOGNITON BY THE INNATE IMMUNE SYSTEM
Mammalian organisms can be infected by a number of different classes of microorganisms (viruses, bacteria, protozoa, etc.), which have fundamentally different physiologies, structures and mechanisms of propagation. However, because the innate immune system has only a limited number of PRRs available, recognition must be based on something common to infections with these highly different infectious agents. The current concept is that the two main principles in innate microbial recognition are the detection of PAMPs and aberrant localization of specific classes of molecules (2, 6). PAMPs are microbial molecular structures that are evolutionarily conserved, and hence shared between different microbial species (7–9). In addition, most PAMPs are essential for microbial growth, and are therefore rarely modified by the microorganism as a means to avoid innate recognition. The second principle in innate recognition is the sensing of aberrant location of molecules, notably nucleotide structures (10–13). Microbial infections are associated with the introduction of nucleotides (both RNAs and DNAs) into endosomes and the cytoplasm, which are abnormal localizations for these structures. By expressing nucleotide-sensing PRRs in endosomes and the cytoplasm, the innate immune system thus uses the aberrant location of specific molecules as a means of microbial recognition.
All viruses and a subgroup of bacteria and parasites propagate inside the cells of the host they infect. Viruses enter cells either directly into the cytoplasm or through a pH-dependent process involving the endocytotic pathway (14) (Fig. 1A). Alternatively, viruses may be endocytosed in a manner not leading to productive infection. Depending on the virus, replication takes place in the cytoplasm or the nucleus, and is highly dependent on the involvement of cellular factors. The main viral PAMPs are glycoproteins of the virus particle and virus-derived nucleotide structures, the latter of which are believed to be particularly important for stimulation of innate antiviral defense (15) (Fig. 1A). It has long been speculated (16, 17), and recently formally demonstrated (18) that virus-derived double-stranded (ds) RNA is a viral PAMP recognized by PRRs both in the cytoplasm and in endosomes (Table 1). In addition, the presence of viral DNA or single-stranded (ss) RNA in endosomes is also detected by the innate immune system. Finally, the RNA structure 5′-triphospho-RNA, which is normally not present in the cytoplasm due to the 5′-CAP of cellular mRNA, is a very potent virus-derived PAMP (9, 19).
Table 1. PAMPs, PRRs, and cellular locations for innate microbial recognition.
While the majority of bacteria grow outside eukaryotic cells, some bacteria are facultative or even obligate intracellular pathogens [e.g. Listeria monocytogenes (facultative), Mycobacterium tuberculosis (facultative) and Chlamydia trachomatis (obligate)]. The entry pathway for these bacteria is through the endosomal system, and common for all is that endosomal maturation is inhibited before fusion of endosomes with lysosomes, thus preventing bacterial killing (20–22). Bacterial replication therefore takes place in the endosomal compartments or, in the case of L. monocytogenes, in the cytoplasm due to a mechanism of escape from the endosomes (Fig. 1B). Recognition of bacteria in the extracellular space is primarily, but not exclusively, due to sensing of microbial lipid structures of the cell wall (Table 1). Intracellularly, the presence of DNA in endosomes and the cytoplasm is detected by the innate immune system, and finally receptors able to detect bacterial cell wall degradation products are expressed in the cytoplasm.
Among parasites, some protozoa enter cells and create a unique membrane-bounded cytoplasmic compartment, the parasitophorous vacuole, where replication takes place (Fig. 1C). In addition, the protozoa may be taken up through the endosomal route, which does not lead to establishment of infection (22). For recognition of protozoa in the extracellular space, glycophosphatidylinositol membrane anchors constitute a major PAMP (Table 1), and in endosomes, receptors for microbial DNA and dsRNA have been reported to contribute to the innate response. Although the parasitophorous vacuole does indeed interact with the cytoplasm (23), innate recognition of protozoa in the cytoplasm has not been reported. Collectively, microorganisms of a very diverse nature are capable of replicating intracellularly in the infected host. Thus, in order for the innate immune system to recognize intracellular pathogens at all stages of the infection (acquisition, cellular entry, replication and spread), recognition systems are expressed both intra- and extracellularly (Fig. 2).
DETECTION OF INTRACELLULAR PATHOGENS BY THE INNATE IMMUNE SYSTEM
Toll-like receptors (TLRs)
TLRs are membrane-bound PRRs able to recognize PAMPs in the extracellular space and in endosomes (2) (Fig. 2). There are 10 (TLRs 1–10) and 12 (TLRs 1–9 and 11–13) TLR family members in humans and mice, respectively (2). They are transmembrane proteins with a single membrane-spanning domain separating their cytoplasmic signaling domain from their ligand-recognizing receptor. Multiple repeats of a leucine-rich (LRR) motif are found in the ectodomain, where PAMP recognition takes place, whereas the intracellular domain is homologous to the cytoplasmic region of the interleukin (IL)-1 receptor, known as the Toll/IL-1 receptor domain (TIR) domain, and is essential for downstream signaling. Dependent on the TLR, the TIR domain is involved in assembling various intracellular adaptor molecules, which also contain a TIR domain (24). Different combinations of the adaptor molecules give rise to specificity in TLR signaling (24). Downstream signaling and induction of antimicrobial factors will be outlined later in this review.
TLR 3, 7, 8 and 9 are located in the membranes of endosomes and the remaining TLRs are located in the cell-surface membrane (Fig. 2). TLRs located in the cell endosomes recognize different nucleotide species, and are mainly involved in the recognition of nucleic acid from bacteria and viruses, to which they gain access due to the acidic environment, whereas the TLRs on the cell-surface membrane primarily detect bacterial ligands (Table 1).
The TLRs are expressed by a variety of cell types, including epithelial cells, although antigen-presenting cells such as dendritic cells (DCs) and macrophages and other professional immune cells are the ones most prominently expressing this class of PRRs (25).
Parallel to TLR3 and 7/8, two cytosolic PRRs named RIG-I and melanoma differentiation-associated gene 5 (MDA5) detect intracellular RNA species, associated with virus infection (Fig. 2), and initiate activation of downstream signaling and induction of cytokines (15). RIG-I and MDA5 are homologous interferon (IFN)-inducible proteins containing two amino-terminal caspase activation and recruitment domains (CARDs), a carboxy-terminal DExD/H-Box RNA helicase domain and a C-terminal regulatory domain (RD). The helicase domain and the RD interact with specific RNA species and the CARDs are responsible for downstream signaling (12, 26). Recently, it has been shown that RIG-I and MDA5 bind short ssRNA and long ssRNA, respectively (27, 28). In addition, RIG-I has been found to detect 5′ triphosphate RNA. dsRNAs are not found in the cytoplasm normally, but accumulate during most viral infections (9, 18, 19). The majority of viral RNAs are uncapped, bearing 5′ triphosphates, whereas host RNAs have 7-methyl-guanosine caps.
A differential role has been determined for RIG-I and MDA5 in viral recognition, possibly reflecting the differences in the ligand specificity of the two PRRs. A wealth of mainly negative RNA viruses have been shown to be detected by RIG-I, whereas only picorna and noroviruses have been found to require MDA5 for recognition (29–32). Two virus families, reovirus and flavivirus, initiate an antiviral response though both RIG-I and MDA5 (33). To add further complexity to the picture, it was recently reported that MDA5 plays a role in the recognition of defective interfering virus particles, which is a by-product of replication by many viruses (34).
Besides RIG-I and MDA5, the family of RLRs also includes a third member, LGP2, which lacks the CARD domains and may act as a negative regulator molecule, possibly by forming heterodimeric complexes with RIG-I and MDA5, although the precise mechanism by which it works is still poorly understood (35–37).
Nucleotide-binding and oligomerization domain (NOD)-like receptors (NLRs)
In addition to the already mentioned PRRs, a group called NLRs has been identified as PRRs. The family consists of 23 and 34 members in humans and mice, respectively, among which are the NODs and Nalps [NACHT, LRR and pyrin domain (PYD) containing] (38). Members of the NLR superfamily have an overall structure consisting of an N-terminal effector domain, responsible for downstream signaling, a central nucleotide-binding NACHT/NAD domain, mediating self-oligomerization upon ligand binding, and a C-terminal LRR region similar to the one found in the TLRs, which are involved in the recognition of specific PAMPs (38).
NOD1 and NOD2 are the best-characterized NLRs and have been shown to take part in the detection of intracellular bacteria (39–41), recognizing different muropeptides derived from the bacterial cell wall (Table 1). The effector domains of the NODs have two CARD domains mediating activation of signaling pathways by homotypic interactions with an adaptor protein, receptor-interacting protein (RIP) 2 (42). NOD1 and 2 have only been linked to bacterial recognition, and it remains unknown whether these PRRs are as restricted in ligand binding as the current literature suggests, or alternatively, whether they could also recognize PAMPs derived from other classes of microbes.
Another prominent NLR group involved in innate immune defense is the Nalps. As the name implies, Nalps are characterized by having an N-terminal effector pyrin domain (PYD), besides the standard LRR and NACHT/NAD domains (43, 44). Among the 14 members of Nalps in humans, Nalp1 and Nalp3 are able to form the inflammasome complexes. This multiprotein complex activates caspase-1, which induces maturation of the proinflammatory cytokines, IL-1β and IL-18, to their biologically active forms (43, 44). In the inflammasome, Nalp interacts with the adapter apoptosis-associated speck-like protein containing a CARD (ASC), through PYD–PYD binding, and ASC bridges to caspase-1 via homotypic CARD–CARD interactions (43, 44).
Ligands recognized by the inflammasome have a diverse origin, because bacterial, viral and cellular danger-associated molecular patterns (DAMPs) have been shown to be detected by this complex. Among the PAMPs able to activate the inflammasome are nucleic acids derived from bacteria, viruses and mammalian cells (45–47).
Cytoplasmic DNA sensors
Several studies have indicated the existence of one or more cytosolic DNA PRRs (48–52), thus demonstrating that DNA, in addition to being recognized in endosomes by TLR9 (11), is also an intracellular PAMP. Cytosolic dsDNA (synthetic or naturally derived from either pathogen or host cells) induces a type I IFN response independent of both TLRs and RLRs (51).
Recently, the group of Dr. Taniguchi found an IFN-inducible protein, which they named DNA-dependent activator of IFN regulatory factors (IRFs) (DAI) (13). They demonstrated that DAI is a novel PRR, interacting with DNA and activating the IRF-family transcription factors for type I IFN induction. DAI contains two binding domains for left-handed Z-form DNA in the N-terminal region and a centrally located region, presumably also having B-DNA-binding abilities (13). The C-terminal region of DAI is essential for activation of downstream signaling pathways, upon binding to DNA (53). Since we are still in the early phase of characterization of DAI, no clear picture has emerged on how this DNA sensor is activated. DAI binds DNA in a length-dependent manner independent of the sequence and whether it is in the B- or the Z-form, although B-DNA is a far more potent activator of the receptor than DNA in the Z-form (53).
DAI is the first identified cytosolic DNA sensor, but experiments indicate the existence of one or more redundant DNA receptors, because cytosolic DNA-induced IFN-β can only partially be inhibited by short interfering RNA against DAI (53).
The biological role of cytosolic DNA recognition in microbial infections is still largely unknown, but initial results indicate possible involvement in the recognition of infections with both viruses and intracellular bacteria (52, 53). Specifically, cytosolic DNA recognition seems to be involved in innate detection of adenovirus and herpes simplex virus (HSV), and at least for HSV, DAI seems to be partially responsible for this recognition, since knockdown of DAI reduces HSV-1-mediated ifnb mRNA induction (13). It seems that the field of cytosolic DNA recognition remains largely unexplored, including full knowledge on the repertoire of receptors involved.
Altogether, intracellular pathogens can be recognized through different classes of PRRs in different intracellular and extracellular locations owing to, due to exposure of PAMPs to these receptors of the innate immune system. As will be described in the next section, PAMP engagement triggers intracellular signaling and expression of gene products involved in the innate antimicrobial response.
PRR SIGNALING IN RESPONSE TO PATHOGEN RECOGNITION
A common characteristic shared by all families of PRRs is their ability to activate the three major signaling pathways: mitogen-activated protein kinases (MAPKs), IRFs and nuclear factor (NF)-κB (Fig. 3). The MAPK pathway activates the transcription factor activator protien 1 (AP-1), and together with NF-κB, contributes to induction of proinflammatory cytokines (54). IRF-3 and -7 are essential for expression of type I IFN and IFN-stimulated genes, whereas IRF-5 participates in the activation of both type I IFNs and proinflammatory mediators (55, 56).
To illustrate the principles in TLR signaling, we will describe the mechanism of signaling downstream of TLR4. The MAPKs cascade is initially activated by the cytoplasmic TIR domain of TLR4, which mediates through the four adaptor proteins, namely, myeloid differentiation primary response protein 88 (MyD88), TIR-domain-containing adaptor inducing IFN-β (TRIF), MyD88-adaptor-like (Mal) and TRIF-related adaptor molecule (57–61). Following activation, the adaptor proteins transmit the signal via the kinases IL-1 receptor-associated kinase (IRAK)-4, IRAK-1/2, and RIP1, which, along with tumor necrosis factor (TNF) receptor-associated factor (TRAF) 6, activate transforming growth factor β-activated kinase (TAK) 1 in association with TAB2/3, through a mechanism dependent on the E3 ubiquitin ligase activity of the TRAF molecule (54, 62–64). TAK1 activates IκB kinase (IKK)-α/β to release NF-κB from its inhibitory subunit IκB, as well as MAPKs (64). In addition, TLR4 has been shown to activate IRF-3/7 in a TRIF- and TRAF3-dependent manner (65, 66). Like TLR4, TLR2 and TLR3 utilize a Mal- and TRIF adaptor-dependent pathway, respectively (59, 67).
Except for TRIF-dependent TLR4 triggering of the IRF pathway, TLR3/7/9 are unique among the TLRs, both in their endosomal location and in their ability to activate IRF-3/7 and thus stimulating type I IFN induction (24). As like in TLR4 signaling, IRAKs and TRAF3/6 are used by TLR7/9 (65, 66, 68, 69). In addition, IKK-α has been found to be vital for TLR7/9-induced type I IFN production, whereas TBK1 and IKK-ɛ are not essential for this process (70). IRF7, an essential transcription factor in IFN-α induction, is constitutively expressed in a subset of DCs, called plasmacytoid DCs (pDCs), and this, together with a high expression of TLR7/9, makes these cells a major source of IFN-α (71). Finally, IRF5 is activated primarily in pDCs through TLR7/9 in a MyD88-dependent pathway, with essential roles for TRAF6 and IRAK1 (55, 56, 72).
RIG-I and MDA5 activate the transcription factors NF-κB and IRF-3/7 (12, 73). The RLRs associate with mitochondrial antiviral signaling (MAVS) (74) [also called IPS-1 (75), VISA (76) and Cardif (77)] via the CARDs in a homotypic interaction (75). MAVS is located in the outer mitochondrial membrane, a position that is essential for its function as an adaptor molecule in RLR signaling (77, 78). From MAVS the signal is transmitted through a complex called the TRADDosome, consisting of FADD, RIP1 and TRADD (78, 79). In addition, MAVS-triggered IFN-α/β induction takes place in a TRAF3-dependent mode (80). The effect of TRAF3 is mediated by a kinase complex consisting of the kinases IKK-ɛ and TBK1, scaffolded by TANK, NAP1 or SINTBAD (65, 66, 81–85). As previously outlined, the RLRs also activate the NF-κB cascade downstream of MAVS. This activation remains incompletely described but involves TRADD, FADD, RIP1, and caspase 8/10 (79, 86, 87).
The cytosolic DNA receptor DAI has been shown to activate NF-κB as well as IRF-3. The precise mechanism by which this takes place is still largely unknown, although it has been reported that DNA-activated DAI assembles with TBK1 and IRF3 (13, 53).
Altogether, PRRs activate a number of signaling pathways in response to pathogen recognition. Different PRRs activate overlapping yet distinct signals and thus biological responses. It should be noted that the specific mechanism of action of a given signal pathway displays some degree of cell-type specificity, as recently illustrated for the role TRADD in TLR signaling in fibroblasts vs macrophages (88, 89).
INDUCTION OF INFLAMMATORY AND ANTIMICROBIAL MEDIATORS – REGULATION AT THE LEVELS OF TRANSCRIPTION AND BEYOND
The purpose of the signal transduction activated by innate recognition of microbes is to activate a cellular response capable of eliminating the pathogen and also to trigger appropriate resolution and tissue repair mechanisms (1). This is a very complex process, which is regulated at several molecular levels in the cells. As described above, activation of transcription factors represents an important end point in many signal transduction pathways activated in response to microbial recognition. The transcription factors in turn bind to specific DNA sequences in gene promoters and activate transcription. Importantly, many promoters harbor binding sites for several transcription factors, which consequently act in synergy to stimulate gene expression (Fig. 4A). This allows stimuli from different receptors to converge at the level of transcription. The IFN-β promoter represents one such promoter, which is activated by concerted action of different transcription factors (90). This promoter contains binding sites for NF-κB, IRFs and AP-1, and all transcription factors are required to activate transcription from the IFN-β promoter (90, 91). Importantly, while NF-κB is required for initiating the assembly of the complex of transcriptions factors and related factors on the promoter, IRF-3 is the main amplifier giving rise to high output transcription (91).
At the level of transcription, additional layers of specificity can be conferred by the specific sequences in the cis elements in gene promoters. This is illustrated by the IFN-stimulated response elements (ISREs), which are present in IFN-stimulated genes (ISGs). Careful analyses of ISGs have demonstrated that ISREs, although all containing a core sequence motif (5′-GAAANNGAAAG/CT/C-3′), differ in their ability to drive transcription through different IRF family members (92). One set of ISREs recruit only the IFN-activated transcription factor ISGF3, composed of STAT1, STAT2 and IRF-9, and hence activates transcription selectively after stimulation with type I IFN (Fig. 4B). Another group of ISREs binds IRF-3, but does not bind to ISGF3 (e.g. the IFN-β promoter), and is therefore able to stimulate transcription immediately after innate viral recognition, but not as part of the host response induced by IFNs. A third class of ISREs displays less specificity, and allows both ISGF3 and IRF-3 to activate transcription. Although we still lack a full understanding of the relationship between DNA sequence and transcription factor binding in the ISRE element, including the involvement of sequences without the core transcription factor-binding site, it is clear that this phenomenon has a strong impact on the transcriptional response to infection (92).
Besides transcriptional regulation of the innate immune response, it is becoming apparent that cells activate and regulate this response at additional levels. This includes control of mRNA stability and translation (93, 94), and also posttranslational processes including maturation of proteins and control of protein stability (95). With respect to regulation at the RNA level, the MAPK p38 has been demonstrated to play an important role (96, 97). As described above, p38 is activated by most PRRs upon ligand engagement and leads to downstream activation of the dimeric transcription factor AP-1. In addition, p38 stimulates activation of a family of kinases, the MAPK-activated protein kinases (MKs) (96), which affect mRNA function in various ways (Fig. 4C). MK2 and 3 are related kinases that interact with the p38α isoform and phosphorylate proteins involved in conferring stabilization and translation of mRNA through the AU-rich elements in the 3′-untranslated region of the mRNA (97, 98). MAPK interacting kinase (MNK) 1 and 2 are also activated by p38, and stimulate translation of mRNAs (96, 99). At the mechanistic level, MNK works by phosphorylating the cap-binding protein eIF4E of the mRNA translation initiation machinery (99).
Finally, gene expression is regulated at the protein level. For instance, ubiquitin-dependent proteasome-mediated degradation of intracellular proteins is an important mechanism for controlling protein stability (100). Also, protease-mediated processing of proteins is utilized as a means to induce maturation of proteins. This is seen for the cytokines IL-1β and IL-18, which are produced as intracellular proforms (95), and cleaved to the mature form by the inflammasome as described above (45–47, 101) (Fig. 4D). The inflammasome is activated in response to cytoplasmic PAMPs and DAMPs and is constituted of a member of the NLR family, which is assumed to bind the PAMP/DAMP, an adaptor protein and caspase 1 (45–47, 101) (Fig. 4D). Once processed into mature cytokines, IL-1β and IL-18 are secreted out of the cell and exert biological activity.
Altogether, orchestration of the cellular response to microbial infection is a complex process, where signals activated by distinct PAMPs–PRRs act in synergy to promote and fine-tune the innate response toward the invading pathogen. The aim of this response is to control the microbe without generating excessive tissue damage and at the same time provide an optimal stimulus for the developing adaptive immune response. To achieve this, gene expression is regulated at multiple levels, including gene transcription, mRNA stability, mRNA translation, as well as protein modification, processing and stability.
MECHANISMS OF INNATE DEFENSE AGAINST INTRACELLULAR PATHOGENS
The goal of the innate immune response is to mount a first line of defense without causing excessive tissue damage and at the same time facilitate and stimulate development of a specific and lasting adaptive immune response (1). To achieve this, several mechanisms are involved, and here we will focus the discussion on the innate microbicidal response.
In order to control microbial replication and spread, the innate immune system utilizes at least two principal strategies: (i) production of molecules that directly exert or stimulate antimicrobial activity (e.g. defensins and type I IFNs) and (ii) mobilization and activation of leukocytes of the innate immune system (e.g. macrophages).
Type I IFNs are expressed by most cell types in response to virus infections and secreted into the extracellular milieu, where they can act in an auto- or paracrine fashion through binding to the IFNAR receptor (102). IFN stimulation induces a broad and very characteristic gene expression profile (103, 104), which, through a number of mechanisms, contributes to inhibition of viral replication (105). Recently, the role of type I IFNs in other microbial infections has been studied, and it has been reported that replication of other intracellular pathogens such as the malaria parasite and also Toxoplasma gondii is reduced by type I IFN (106, 107). This is in contrast to the reported elevated resistance of IFNAR−/− mice to infection with the intracellular bacteria Listeria and Chlamydia (108, 109). Therefore, the role of type I IFNs in infection with nonviral intracellular pathogens remains to be fully clarified.
Recently, a novel class of IFNs has been identified, and termed type III IFNs or IFN-λ (110, 111). The type III IFNs share many functions with type I IFN, but unlike the type I IFNs, which exert antiviral activity toward all cell types, the type III IFNs target primarily epithelial cells (112), and consequently play an important role in innate antiviral defense at epithelial surfaces (113), which constitute a major portal of entry for viral infections.
As mentioned above, another important mechanism of innate antimicrobial defense is through activation of leukocytes with phagocytic activity (114, 115). Macrophages and neutrophils are cells specialized in phagocytosing opsonized objects from the extracellular environment. The mechanisms of microbicidal activity of the phagocytosed microorganisms include acidification of phagolysosomes, proteolytic enzymes and free radical formation (116).
Collectively, recognition of microbial pathogens by the innate immune system stimulates antimicrobial activities, which, based on PRR-mediated decoding of the nature of the pathogen, target specific classes of microbes. For instance, type I IFNs is primarily induced by virus-recognizing PRRs, and potently activate antiviral defense. It should also be mentioned that the antimicrobial strategies described above do exert a large degree of cross-talk. For instance, type I IFNs stimulate leukocyte recruitment and activation, which in turn amplify IFN function due to synergistic action with other cytokines of the innate system (117).
Since the discovery of the first human TLR in 1997 (118), our understanding of how the innate immune system recognizes microbial infections and activates the first line of defense against infections has been revolutionized. In this review, we have focused on the molecular and cellular mechanisms governing the initial response to infections with intracellular pathogens. In addition to what has been discussed in this article, it should be emphasized that additional layers of complexity exist, because microbes are often recognized through cell type-specific sets of PRRs (29, 119), which may be translated into cell type-specific responses. Moreover, the interactions between specialized cell types and the local environment at sites of infection may also give rise to a high degree of organ specificity in the innate host response to infections (120). Finally, infections in living organisms may be associated with tissue damage leading to the release of host-derived DAMPs, which are also recognized by PRRs and hence contribute to activation of the innate immune response. One of the major future challenges in innate immunity research is to understand how different PAMPs and DAMPs coordinately activate host responses to infection, and how this contributes to microbial elimination and tissue repair, and may also give rise to inflammatory diseases.
S. B. R. and L. S. R. were supported by fellowships from the Faculty of Health Science, Aarhus University. This work was supported by research grants from the Danish Medical Research Council (grant no. 271-06-0438), the Lundbeck Foundation, Elvira og Rasmus Riisforts almenvelgørende Fond, Aarhus University Research Foundation, Augustinus Fonden and Kathrine og Vigo Skovgaards Fond.