There Are Several Classes Of Pattern
Recognition Receptors
PRRs on phagocytic cells recognize and are activated by PAMPs
Because the ability to
discriminate friend from foe is of paramount importance for any self‐respecting
phagocyte, macrophages are fairly bristling with receptors capable of recognizing diverse PAMPs. Many of the PRRs
are also expressed on DCs, NK cells, neutrophils and mast cells, as well as
cells of the adaptive immune system. Several of these PRRs are lectin‐like and
bind multivalently with considerable specificity to exposed microbial surface
sugars with their characteristic rigid three‐dimensional geometric
configurations. They do not bind appreciably to the array of galactose or
sialic acid groups that are commonly the penultimate and ultimate sugars that
decorate mammalian surface polysaccharides, so providing the molecular basis
for discriminating between self and nonself microbial cells. Other PRRs detect
nucleic acids derived from bacterial and viral genomes by virtue of modifications
not commonly found within vertebrate nucleic acids or conformations not
normally found in the cytoplasm (e.g.,
double‐stranded RNA).
PRRs are a diverse group of
receptors that can be subdivided into at least five distinct families (TLRs,
CTLRs, NLRs, RLRs, and scavenger receptors) based upon structural features.
Another class of sensors has also emerged in recent years, the cytosolic DNA
sensors (CDSs), which contains a structurally diverse set of cytosolic
DNA‐sensing receptors that are pre- dominantly involved in detecting
intracellular bacteria and viruses. Multiple receptors also exist in each
class, with the result that in excess of 50 distinct PRRs may be expressed by a
phagocyte at any given time.
Cell‐associated PRRs decode the nature of infection
As noted earlier, there are
several classes of cell‐associated PRRs, some of which are plasma
membrane‐associated (e.g., many of the Toll‐like receptors as well as the
C‐type lectin receptors and scavenger receptors), some of which face the
luminal space of endosomes (TLR3, 7, 8, 9) and some of which are cytoplasmic
(NOD‐like receptors, RIG‐I‐like receptors, cytoplasmic DNA sensors). In general
terms, each PRR is specific for a distinct PAMP and, combined with the
different cellular compartments that PRRs reside in, this conveys considerable
information concerning the nature of the pathogen and whether it is
extracellular, has been captured through phagocytosis (i.e., is within
endosomes) or has invaded the cytoplasm. This information helps to tailor the
response towards what will be most effective for the particular class of
pathogen by influencing the nature of the cytokines that are produced by the
responding cell.
Engagement of several categories of PRR simultaneously may be required
for effective immune responses
Although this is an area of
ongoing research, combinatorial PRR signaling is probably very
important for the initiation of effective immune responses. Thus, the
triggering of a single type of PRR, in a DC for example, may not be fully
effective for the initiation of a robust adaptive immune response, as this
could indicate either a low level of infection, or that the DC is at a
considerable distance from the site of infection (and has simply encountered a
few stray PAMPs that have been released owing to lysis of the infectious
agent). However, phagocytosis of a single bacterium by a DC is likely to
stimulate multiple categories of PRR simultaneously, leading to
synergistic activation of several signal transduction pathways, thereby signifying
that a robust response is warranted. Furthermore, it is likely that engagement
of different combinations of PRRs underpins the different types of immune
response that are required to successfully contain different types of
infection: intracellular, extracellular, large parasite, yeast, bacterial,
viral, etc.
As we shall see throughout this book,
delivery of two (or more) different signals in tandem is a common theme
in immune reactions and can lead to very different outcomes compared
with delivery of either signal on its own. We will now look at the various PRR
families in more detail.
Toll‐like receptors (TLRs)
A major subset of the PRRs belong
to the Toll‐like receptor (TLR) family, named on the basis of
their similarity to the Toll receptor in the fruit fly, Drosophila. The
history of the discovery of the TLR family is interesting, as it perfectly
illustrates the serendipitous nature of scientific discovery and illustrates
how very important findings can originate in the most unlikely places.
Lipopolysaccharide (LPS, also called endotoxin), a major component of the cell
walls of Gram‐ negative bacteria, was long known to provoke strong immune
responses in animals and is a good example of a classical PAMP. Indeed, LPS is
one of the major contributors to septic shock, the severe immune reaction that
results when a bacterial infection reaches the bloodstream, and which is often
fatal. For these reasons, immunologists tried to identify the LPS receptor in
human and mouse for many years, largely without success. However, a major
breakthrough came when the Toll receptor was found to be involved in sensing
microbial infection in adult fruit flies. This in itself was quite a surprise
because the Toll receptor had already been identified, many years before, as a
major regulator of dorsal– ventral patterning (i.e., specifying which surface
of the fly is the back and which is the underside) during early embryonic
development of Drosophila. A curious fact that emerged was that the
intracellular domain of Drosophila Toll contained a motif, now known as
the Toll/IL‐1 receptor (TIR) signaling motif, that was very similar to the
cytoplasmic signaling domain identified in the IL‐1 receptor, a molecule that
was already well known to be involved in immune signaling in mammals. Putting
two and two together, this led to the identification of the whole TLR family in
mammals, as these receptors all possess a TIR domain within their cytoplasmic
regions.
A series of TLRs have now been
identified (there are 10 distinct TLRs in humans), all of which act as sensors
for PAMPs (Figure 1.16). TLR ligands include peptidoglycan, lipoproteins,
mycobacterial lipoarabinomannan, yeast zymosan, flagellin, microbial DNA,
microbial RNAs, as well as other pathogen‐derived ligands (Table 1.1). Although
many TLRs are displayed on the cell surface, some, such as TLR3 and TLR7/8/9
that are responsive to intracellular viral RNA and unmethylated bacterial DNA,
are located in endosomes and become engaged upon encounter with phagocytosed
material (Figure 1.16). Engagement of TLRs with their respective ligands drives
activation of nuclear factor kB (NFkB) and several members of the
interferon‐regulated factor (IRF) family of transcription factors, depending on
the specific TLR. Combinatorial activation of TLRs is also possible, for
example TLR2 is capable of responding to a wide diversity of PAMPs and
typically functions within heterodimeric TLR2/TLR1 or TLR2/TLR6 complexes
(Table 1.1).
All TLRs have the same basic
structural features, with multiple N‐terminal leucine‐rich repeats (LRRs)
arranged in a horseshoe or crescent‐shaped solenoid structure that acts as the
PAMP‐binding domain (Figure 1.17). Upon binding of a PAMP, TLRs transduce signals into the cell
via their TIR domains, which recruit adaptor proteins within the cytoplasm
(such as MyD88) that possess similar TIR motifs. These adaptors propagate the
signal downstream, culminating in activation of
NFκB and interferon regulatory family (IRF) transcription factors, which
regulate the transcription of a whole battery of inflammatory cytokines and
chemokines (Figure 1.16 and Figure 1.18). As we will discuss later in this
chapter, the IRF transcription factors control the expression of, among other
things, type I interferons. The latter cytokines are especially important in
defense against viral infections as they can induce the expression of a series
of proteins that can interfere with viral mRNA translation and viral
replication, as well as induce the degradation of viral RNA genomes.
C‐type lectin receptors (CTLRs)
Phagocytes also display another
set of PRRs, the cell‐bound C‐type (calcium‐dependent) lectins,
of which the macrophage mannose receptor is an example. Other members of this
diverse and large family include Dectin‐1, Dectin‐2, Mincle, DC‐ SIGN, Clec9a,
and numerous others. These transmembrane proteins possess multiple carbohydrate
recognition domains whose engagement with their cognate microbial PAMPs generates
intracellular activation signals through a variety of signaling pathways.
However, some C‐type lectin receptors (CTLRs) do not trigger robust
transcriptional responses and function primarily as phagocytic receptors. The
CTLR family is highly diverse and the ligands for many receptors in this
category are the subject of ongoing research. But it can be said that members
of the CTLR family broadly serve as sensors for extracellular fungal species.
Some examples of ligands for CTLRs include β‐glucans (which binds Dectin‐1),
mannose (which binds Dectin‐2), and α NOD‐like receptors (NLRs)
Turning now to the sensing of infectious
agents that have succeeded in gaining access to the interior of a cell,
microbial products can be recognized by the so‐called NOD‐like receptors
(NLRs). Unlike TLRs and CTLRs that reside within the plasma membrane or
intracellular membrane compartments, NLRs are soluble proteins that reside in
the cytoplasm, where they also act as receptors for PAMPs. Although a diverse
family of receptors (Figure 1.19), NLRs typically contain an N‐terminal
protein–protein interaction motif that enables these proteins to recruit
proteases or kinases upon activation, followed by a central oligomerization
domain and multiple C‐terminal leucine‐rich repeats (LRRs) that act as the
sensor for pathogen products (Figure 1.19). The NLRs can be subdivided into
four subfamilies on the basis of the motifs present at their N‐termini. NLRs are
thought to exist in an autoinhibited state with their N‐terminal domains folded
back upon their C‐terminal LRRs, a conformation that prevents the N‐terminal
region from interacting with its binding partners in the cytoplasm. Activation
of these receptors is most likely triggered through direct binding of a PAMP to
the C‐terminal LRRs which has the effect of disrupting the interaction between
the N‐ and C‐ termini of the NLR and permits oligomerization into a complex
that is now capable of recruiting either an NFkB‐activating kinase (such as
RIP‐2) or members of the caspase family of proteases that can proteolytically
process and activate the IL‐1β precursor into the mature, biologically active
cytokine.
A very well‐studied NLR complex,
called the inflammasome, is assembled from NLRP3 in response to
LPS in combination with bacterial virulence factors, and is important for the
production of IL‐1β as well as IL‐18. However, for full activation of the
inflammasome and liberation of IL‐1β, a second signal in the form of a
membrane‐damaging bacterial toxin (which
can also be mimicked by a variety of noxious agents) is required. This second
signal appears to permit the efflux of K+ ions from the cytosol, which permits
full assembly of the inflammasome, caspase‐1 activation, and processing of
IL‐1β and IL‐18 downstream (Figure 1.20).
RIG‐I‐like helicase receptors (RLRs)
The RIG‐I‐like helicases are a
relatively recently discovered family that act as intracellular sensors for
viral‐derived RNA (Figure 1.21). Similar to the NLRs, RIG‐I‐like helicase receptors
(RLRs) are found in the cytoplasm and are activated in response to
double‐stranded RNA and are capable of directing the activation of NFkB and
IRF3/4 that cooperatively induce antiviral type I interferons (IFNα and β).
RIG‐I (retinoic acid‐ inducible gene I) and the related MDA‐5 (also called
Helicard) protein can directly bind to different forms of viral RNA (either
unmodified 5′‐triphosphate ssRNA or dsRNA, respectively) in the cytoplasm,
followed by propagation of their signals via MAVS (mitochondrial‐associated
viral sensor), again leading to activation of IRFs and NFkB (Figure 1.22).
Cytosolic DNA sensors
A number of proteins belonging to
different families are capable of sensing cytosolic DNA or cyclic
dinucleotides. Host cell DNA is normally sequestered safely in the nuclear or
mitochondrial compartments and cannot trigger these sensors, except under
pathological conditions that involve release of mitochondrial DNA into the
cytosol, for example. However, bacterial or viral DNA can trigger the
activation of the AIM2 or IFI16 DNA sensors and
this can lead to assembly of a complex involving the Pyrin‐domain‐containing
adaptor (ASC), leading to activation of caspase‐1 and IL‐1β Activation of the AIM2 inflammasome can
also lead to death of the cell. IFI16 can also recognize cytosolic DNA and can
either propagate signaling by forming a complex with ASC and caspase‐1, similar
to the AIM2 inflammasone, or via STING, which is discussed below. Two
additional DNA‐sensing path- ways have also been discovered very recently and
both make use of STING (stimulator of interferon genes) a
molecule that can either directly bind to cytoplasmic DNA or can respond to
cyclic GAMP, a molecule that is generated by an upstream enzyme called cGAS,
which detects cytoplasmic DNA and synthesizes cGAMP in response (Figure 1.23).
In response to STING activation, type I IFNs are generated which have potent
antiviral properties.
Scavenger receptors
Scavenger receptors represent yet
a further class of phagocytic receptors that recognize a variety of anionic
polymers and acetylated low‐density proteins. The role of the CD14 scavenger
molecule in the handling of Gram‐negative LPS (lipopolysaccharide endotoxin)
merits some attention, as failure to do so can result in septic shock. The biologically
reactive lipid A moiety of LPS is recognized by a plasma LPS‐binding protein,
and the complex that is captured by the CD14 scavenger molecule on the
phagocytic cell then activates TLR4. However, unlike the PRRs discussed above,
engagement of scavenger receptors are typically insufficient on their own to
initiate cytokine activation cascades.
PRR engagement results in cell activation and proinflammatory cytokine
production
Upon encountering ligands of any
of the aforementioned PRRs, the end result is a switch in cell behavior from a
quiescent state to an activated one. Activated macrophages and neutrophils are
capable of phagocytosing particles that engage their PRRs and, as we have seen
from our discussion of the various classes of PRRs, upon engagement of the
latter they also release a range of cytokines and chemokines that amplify the
immune response further (see Figure 1.12). As the reader will no doubt have
noticed, engagement of many of the above PPRs results in a signal transduction
cascade culminating in activation of NFkB, a transcription factor that controls
the expression of numerous immunologically important molecules such as
cytokines and chemokines. In resting cells, NFκB is sequestered in the cytoplasm
by its inhibitor IkB, which masks a nuclear localization signal on the former.
Upon binding of a PAMP to its cognate PRR, NFκB is liberated from IκB because
of the actions of a kinase that phosphorylates IκB and promotes its
destruction. NFκB is now free to translocate to the nucleus, seek out its
target genes, and initiate transcription (see Figure 1.18).
Some of the most important
inflammatory mediators synthesized and released in response to PRR engagement
include the antiviral interferons (also called type I interferons),
the small protein cytokines IL‐1β, IL‐6, IL‐12, and tumor necrosis factor α
(TNFα), which activate other cells through binding to specific receptors, and
chemokines, such as IL‐8, which represent a subset of chemoattractant
cytokines. Collectively, these molecules amplify the immune response further
and have effects on the local blood capillaries that permit extravasation of
neutrophils, which come rushing into the tissue to assist the macrophage in
dealing with the situation (see Figure 1.15).
Dying cells also release molecules capable of engaging PRRs
As we have mentioned earlier,
cells undergoing necrosis (but not apoptosis) are also capable of releasing
molecules (i.e., DAMPs) that are capable of engaging PRRs (see Figure 1.3). The
identity of these molecules is only slowly emerging, but includes HMGB1, members of the S100
calcium‐binding protein family, HSP60 and the classical cytokines IL‐1α and
IL‐33. Certain DAMPs appear to be able to bind to members of the TLR family
(i.e., HMGB1 has been suggested to signal via TLR4), while others such as IL‐1α
and IL‐33 bind to specific cell surface
receptors that possess similar intracellular signaling motifs to the TLR
receptors.
DAMPs are involved in amplifying
immune responses to infectious agents that provoke cell death and also play a
role in the phenomenon of sterile injury, where an immune
response occurs in the absence of any discernable
infectious agent (e.g., the bruising that occurs in response to a compression
injury that does not breach the skin barrier represents an innate immune
response). Indeed, Polly Matzinger has proposed that robust immune responses
are only seen when nonself is detected in combination with tissue damage (i.e.,
a source of DAMPs). The thinking here is that the immune system does not need
to respond if an infectious agent is not causing any harm. Thus, PAMPs and
DAMPs may act synergistically to provoke more robust and effective immune
responses than would occur in response to either alone.
