You are on page 1of 9

REVIEWS

THERAPEUTICS TARGETING
THE INNATE IMMUNE SYSTEM
Richard J. Ulevitch
Proteins that recognize the components and products of microorganisms have an important role
in innate immunity. Here, I focus on recent advances in our understanding of the function of several
such protein families. In particular, I consider how members of the TLR (Toll-like receptor), NOD
(nucleotide-binding oligomerization domain)-protein and MyD88 (myeloid differentiation primary-
response protein 88) families are providing emerging opportunities for the development of new
therapeutics that modify innate immune responses in ways which benefit the host.

The modulation of immune responses — achieved by leading to the activation of the innate and adaptive
targeting cell-surface receptors or intracellular pathways immune systems. The proteins that are encoded by the
— is one of the main goals in the development of new NOD genes (also known as members of the caterpillar
therapeutics for human immune or inflammatory dis- family) are cytosolic proteins that have a role in various
eases. Traditionally, the focus has been on targets that innate and adaptive immune responses. Of particular
are considered to be part of the adaptive immune sys- interest here are NOD1 and NOD2, which recognize
tem. In this way, major efforts have been made to distinct structures derived from peptidoglycan that are
develop therapeutics for transplantation, autoimmunity not ligands for TLRs10–12.
and cancer. Now, the ever-expanding wealth of informa- In mammals, the main function of the innate
tion about the innate immune system, including the immune system is to detect the presence of invading
identification of cognate ligands of innate immune microorganisms. Importantly, another concept of the
receptors and the elucidation of their downstream sig- function of innate immunity is emerging — that it also
nalling pathways, provides a new set of targets for drug has a role in sterile inflammation. Sterile inflammation
development, which might impact on various human is not driven by the response to infection with micro-
diseases1–9. The innate immune system sits at the inter- organisms but by ligands derived from damaged cells,
section of the pathways of microbial recognition, which are not usually present in the extracellular envi-
inflammation, microbial clearance and cell death, ronment: for example, heat-shock proteins, β-defensins
thereby offering diverse targets for therapeutics (FIG. 1). and oxidized lipids. It can also be speculated that pro-
This review limits its scope to gene families that encode teins modified by oxidation or nitration might be TLR
the key receptors of the innate immune system — the ligands and thereby might promote sterile inflamma-
Toll-like receptor (TLR)- and nucleotide-binding tion. So, in both infection and sterile inflammation, the
oligomerization domain (NOD)-gene families — their innate immune system uses receptors that include
ligands and their downstream signalling pathways. members of the TLR and NOD-protein families9.
Here, the term ligand is used to indicate a molecule Ligands for these receptor families can be divided into
that binds to its cognate receptor, although for TLRs several groups: naturally occurring molecules that are
Department of and NOD proteins proof of the formation of a released from microbial sources, synthetic structures
Immunology, The Scripps ligand–receptor complex is often lacking. TLRs com- based on those of microbial products, small molecules
Research Institute, La Jolla,
California 92037, USA. prise a family of cell-surface and endosomally expressed with no obvious structural relationships to naturally
e-mail: ulevitch@scripps.edu receptors that recognize conserved products unique to occurring ligands and endogenous ligands of host ori-
doi:10.1038/nri1396 microorganisms, such as lipopolysaccharide (LPS), gin. The downstream signalling pathways for the TLRs

512 | JULY 2004 | VOLUME 4 www.nature.com/reviews/immunol


FOCUS ON TLR SIGNALLING

are involved; however, here, only the steps that include


the MyD88-adaptor family are considered. Ultimately, we
Microbial-clearance Apoptotic and necrotic need to define the signalling events and gene-activation
and killing pathways cell-death pathways
patterns that occur downstream of each of the TLRs so
that we can identify appropriate targets for the selective
activation or inhibition of these pathways.
In contrast to the TLRs, NOD1 and NOD2 seem to
signal through pathways that are independent of
Innate immune system
members of the MyD88 family. Instead, after engage-
ment of their cognate ligands, NOD proteins activate
NF-κB through a pathway that requires the kinase
RICK (receptor-interacting serine/threonine kinase; also
Microbial-recognition Sterile-inflammation known as receptor-interacting protein 2, RIP2) 16,17.
pathways pathways RICK and NOD1 or NOD2 interact through the
caspase-recruitment domains (CARDs) present in each
TLR NOD TLR
protein protein. Further details about other proteins involved in
Exogenous Endogenous NOD-protein signalling are unknown at present.
ligands ligands
The central question then is how to use the current
Figure 1 | Innate immunity at the crossroads. The innate information about the signalling pathways of the innate
immune system is at the intersection of several pathways immune system to design new therapeutics. The desired
that influence the balance between health and disease.
These pathways, which are responsible for the recognition
feature of any therapeutic that modifies the innate
of microorganisms and endogenous host-derived ligands, immune response — either as an agonist or antagonist
trigger the clearance and/or killing of microorganisms, as — is, of course, specificity. In this review, examples of
well as apoptotic and necrotic cell-death pathways that agonists and antagonists are provided to illustrate the
depend on pro-inflammatory mediators. NOD, nucleotide- potential utility of drugs targeting innate immunity.
binding oligomerization domain; TLR, Toll-like receptor. Complications that could arise from either blocking or
activating innate immunity are also discussed, as well
as issues that result from targeting innate immune
and NOD proteins show one main difference. TLRs pathways in acute and chronic settings.
couple ligand binding to cell activation through mem-
bers of the myeloid differentiation primary-response TLR family
protein 88 (MyD88)-adaptor family (BOX 1), whereas Over the past five years, there have been remarkable
NOD proteins do not seem to use MyD88-dependent advances in the identification and characterization of
signal-transduction pathways. In the signalling that the TLRs. This information provides the basis for
occurs in an innate immune response, two central considering new ways of modulating the innate
events are the activation of nuclear factor-κB (NF-κB) immune response by designing new therapeutic
and interferon regulatory factor 3 (IRF3) . The NF-κB approaches (FIG. 2).
pathway controls the production of pro-inflammatory The Drosophila studies of Hoffmann and col-
molecules, such as interleukin-1 (IL-1) and tumour- leagues 18–21 provided the basis for the discovery
necrosis factor (TNF), whereas the IRF3 pathway leads of the TLRs, the mammalian homologues of the
to the production of type I interferons (IFN-α and Drosophila protein Toll. The TLR family now consists
IFN-β). In this review, I discuss interesting drug targets of 13 members: although only TLR1 to TLR10 have
that have emerged from recent studies of TLRs and/or been identified in humans. TLR11 does not seem to
NOD proteins, without considering as targets the be present in humans22, and TLR10 is not present in
numerous downstream cytokines and other mediators mice. TLRs are members of the Toll/IL-1R (TIR)-
that are produced following the activation of the innate domain-containing superfamily, and there is consid-
immune system. erable homology between the cytoplasmic domains
Signalling from TLRs involves the Toll/IL-1 receptor of all family members. In particular, a high degree of
(TIR) domain of the receptor interacting with one or similarity exists within the 200 amino-acid region
more members of the MyD88-adaptor family, which that comprises the TIR domain; within this domain
also each contain a TIR domain. These interactions are up to three conserved regions that are crucial for
are known as TIR–TIR domain interactions4,13–15. The the assembly of downstream signalling complexes
engagement of a TLR by its cognate ligand most containing one or more members of the MyD88-
probably facilitates the formation of a hetero- or adaptor family. Other proteins that transduce signals
homodimeric receptor, which then interacts with or otherwise regulate receptor function probably also
cytosolic TIR-domain-containing adaptor proteins. form part of these complexes. Despite the similarity
Numerous studies have investigated the signalling of TIR-domain-containing superfamily members,
pathways immediately downstream of the TIR- there are substantial differences in the extracellular
domain interactions, showing how the molecules regions of these proteins. The extracellular regions of
IRAK1 (IL-1 receptor (IL-1R)-associated kinase 1), the IL-1Rs consist of immunoglobulin-like domains.
IRAK4 and TRAF6 (TNF-receptor-associated factor 6) By contrast, the TLRs have two highly conserved

NATURE REVIEWS | IMMUNOLOGY VOLUME 4 | JULY 2004 | 5 1 3


REVIEWS

structural features, the most important of which is the their cognate intracellular TLRs also remains to be
leucine-rich repeat (LRR)-containing ectodomain. This determined. This is of particular interest for the syn-
domain is thought to be involved in ligand recognition thetic TLR9 ligand, which is based on non-methylated
and usually comprises 24–29 amino-acid repeats that CpG sequences that are found in bacterial DNA.
contain the canonical repeat XLXXLXLXX, where X The report by Medzhitov, Janeway and Preston-
denotes any amino acid. Both the LRRs and the TIR Hurlburt98 that described a human homologue of
domains could be exploited as drug targets. Drosophila Toll showed that the expression of human
In addition to the conserved structural features Toll (TLR) in human cell lines leads to the activation of
among the TLRs, there seems to be another level of NF-κB and the expression of cytokines. This data led to
organization. The TLRs that are involved in the recogni- the first model for TLR-dependent cell activation, in
tion of microbial products — TLR1, TLR2, TLR4, TLR5, which the expression and multimerization of TLRs is
TLR6 and TLR11 — are displayed on the cell surface. By sufficient to support cell activation. Subsequently, when
contrast, TLR7, TLR8 and TLR9 are localized intra- the cognate ligands for the TLRs were defined, it became
cellularly, and their natural ligands might only be found clear that TLRs associate with other proteins to form a
within acidic compartments, such as phagolysosomes. heteromeric receptor complex. These additional proteins
The natural ligand for TLR9 is bacterial DNA, and might be TLRs that form either homo- or heterodimeric
recent studies using TLR7-deficient mice, as well as complexes or proteins unrelated to the TLR family.
investigations of human TLR8, have shown that these There are several examples of each type of receptor com-
TLRs recognize guanosine and uridine (G+U)-rich plex, including the CD14–MD2–TLR4 complex24,25 and
single-stranded RNA. Whether TLR3 functions at the the TLR2–TLR1 (REFS 26–29), TLR2–TLR6 (REFS 26–29)
cell surface or intracellularly is still unresolved, although and TLR4–TLR5 (REF. 30) complexes. Although het-
some reports indicate that signalling from TLR3 does eromeric complexes have only been identified for TLRs
require endosomal maturation23. These differences that are present at the plasma membrane, it seems
between TLRs indicate that the design of synthetic probable that further components of these complexes
agonists will need to take cellular localization into con- will be identified for both the cell-surface expressed and
sideration. For example, if high-throughput binding intracellular TLRs.
assays are to be used to identify novel TLR ligands, then There is little doubt that the heteromeric nature of the
the pH of the binding solutions will need to reflect the TLR complexes reflects the downstream physiological
potential differences between the plasma membrane pathways. So, can this knowledge be exploited to identify
and intracellular compartments. The precise mecha- new drug targets that do not require the development of
nism by which synthetic ligands enter cells and engage novel agonists or antagonists that directly target the
TLR-ligand-binding sites? To achieve this, agents that
disrupt protein–protein interactions will be required.
Box 1 | MyD88-adaptor family
The figure shows the domain NOD-protein family
structure of the five known MyD88 DD TIR Recently, a family of mammalian proteins that contain
members of the MyD88 a nucleotide-binding oligomerization domain (NOD)
(myeloid differentiation TIRAP TIR and LRRs has been shown to have an important role in
primary-response protein 88) innate immunity, and substantial progress has been
-adaptor family: MyD88, TRIF TIR made in identifying the ligands for the two family
TIRAP (Toll/interleukin-1 members that have known innate immune functions:
receptor (TIR)-domain- TRAM TIR NOD1 and NOD2 (REFS 10–12) (FIG. 3). Members of this
containing adaptor protein; family also have structural features that are found in
also known as MyD88- the large family of R proteins (also known as plant
SARM SAM SAM TIR
adaptor-like protein, disease-resistance proteins)31–33. There are more than
MAL), TRIF (TIR-domain- 20 members of the NOD-protein family, including
containing adaptor protein inducing interferon-β), TRAM (TRIF-related adaptor proteins of known and unknown functions. One com-
molecule) and SARM (sterile α- and armadillo-motif-containing protein)15. Evidence
mon feature of these proteins is the presence of specific
from genetic and biochemical studies has linked signalling from Toll-like receptor 7
structural domains at the amino (N)-terminus,
(TLR7), TLR8 and TLR9 to a pathway that requires only MyD88. By contrast, signalling
including the CARD and pyrin sequences. The latter
from TLR2, after the formation of a heterodimer with either TLR1 or TLR6, is linked to a
allow for unique protein–protein interactions and
pathway that requires a complex of MyD88 and a second adaptor, TIRAP. The activation
of interferon-regulatory factor 3 (IRF3) and the subsequent induction of interferon-β, most probably, the propagation of some downstream
which are elicited by signalling from TLR3 or TLR4, are MyD88-independent and involve signals. The tissue-distribution patterns of NOD pro-
a third adaptor, TRIF. A fourth adaptor, TRAM, is also involved in MyD88-independent teins range from ubiquitous to highly restricted.
signalling through TLR4 (REF. 74–77), and a fifth adaptor, SARM, has been described, but Mutations in several proteins — including NOD2,
its role in TLR signalling is unclear at present. At the most proximal step after ligand MHC class II transactivator (CIITA), neuronal apop-
binding, these proteins of the MyD88-adaptor family provide specificity for the outcome tosis inhibitory protein 5 (NAIP5) and cryopyrin —
of TLR signalling that is initiated following the engagement of different types of ligands are associated with increased disease susceptibilities,
by different TLRs. chronic inflammatory disease and hypersensitivity to
some bacterial infections. The NOD-protein family
DD, death domain; SAM, sterile α-motif.
initiates cellular activation through a multi-protein

514 | JULY 2004 | VOLUME 4 www.nature.com/reviews/immunol


FOCUS ON TLR SIGNALLING

complex known as an inflammasome. The composi- Two of the NOD-protein-family members, NOD1
tion and properties of these multi-protein complexes and NOD2, are intracellular sensors for bacterial
are discussed in detail by Martinon and Tschopp34 and products that include subunit structures found in
are not reviewed here. NOD-protein-family members bacterial peptidoglycans10–12. Similar to the TLRs, the
seem to be involved in a number of autoinflammatory LRR domains of NOD1 and NOD2 are thought to
human diseases, such as Blau syndrome, Muckle-Wells contain the ligand-binding sites. The N-terminal
syndrome and familial cold urticaria, thereby indicating domains might also provide additional sites for drug
their importance in inflammation. This probably also interactions. A recent study by Nunez and col-
depends, in part, on members of the inflammatory- leagues35 highlights the complexity of the domains of
caspase family. The blockade of caspases by selective NOD1 and NOD2 that are required for ligand bind-
inhibitors is a well-developed research area and is not ing. The fact that NOD1 and NOD2 recognize lig-
discussed further here. ands derived from bacterial peptidoglycan, are linked
to NF-κB and contain CARDs indicates that these
proteins are involved in pathways that regulate
Block TLR4 agonists:
TIR domain– inflammation and cell death. The cell-death pathways
MPL, RC-529
TIR domain Block receptor– are almost certainly mediated by members of the cas-
interactions ligand or protein– pase family, including those known as inflammatory
protein interactions
CD14
TLR4 caspases34.
MD2
TLR and NOD-protein ligands
Endosome There are at least four broad types of ligand for the
TLR7/8/9
TLRs, and for NOD1 and NOD2: naturally occurring
TIRAP

TRAM
TRIF
MyD88

molecules that are constituents of microorganisms,


synthetic analogues of naturally occurring substances,
small-molecule fully synthetic compounds, and
endogenous components that are released from host
Agonists: cells during processes such as necrosis, apoptosis and
loxoribine
inflammation. Ligands are generally thought of as agon-
MyD88

(TLR7),
resiquimod IRAK1 ists, but several classes of antagonist have been devel-
(TLR7), RIP1 TRAF6
imiquimod oped using the structure of agonists as a template for
(TLR7, TLR8) TRAF6 synthetic analogues. A survey of TLR ligands is shown
CpG ODNs
(TLR9) in TABLE 1. Most of the ligands are derived from micro-
bial sources, and they include polymeric molecules
such as peptidoglycans (which bind TLR2), bacterial
IKK-γ LPS (TLR4), bacterial and viral DNA (TLR9)36,37 and
double-stranded RNA (TLR3). Nonetheless, there are
IKK-α IKK-β many naturally occurring ligands, these are small mole-
cules and include lipoteichoic acid (TLR2), lipoarabi-
nomannan (TLR2) and taxol (TLR4). Large polymeric
molecules are not ideal drug candidates because of
IκB
NF-κB
many factors, including their pharmacodynamic and
p50 p65
pharmacokinetic properties. By contrast, because
Nuclear membrane
small-molecule ligands have been identified for many
of the TLRs, the design of new small molecules with
drug-like properties seems to be within reach. There are
already good examples of synthetic analogues of natu-
NF-κB- rally occurring ligands, such as molecules based on the
binding motif
structure of lipid A, CpG oligodeoxynucleotides (ODNs)
Figure 2 | TLRs as targets for therapy. Toll-like receptor (TLR) signalling is initiated by plasma modelled on bacterial DNA sequences, peptidoglycan
membrane and intracellular (endosomal) TLRs following ligand recognition. TLR4 is shown as subunits containing muramic acid, and unique peptide
a representative membrane receptor and the endosomal TLRs (TLR7, TLR8 and TLR9) are also sequences38. Compounds of the imidazole quinoline
shown. Stimulation of TLRs triggers interaction with the adaptor molecule MyD88 (myeloid
differentiation primary-response protein 88). Other adaptor molecules, TIRAP (Toll/interleukin-1
series, which function as immunostimulants, provide
receptor (TIR)-domain-containing adaptor protein), TRIF (TIR-domain-containing adaptor proof-of-concept that fully synthetic small molecules are
molecule inducing interferon-β) and TRAM (TRIF-related adaptor molecule) are also involved recognized by TLRs.
in TLR4 signalling. The MyD88-dependent signalling pathway leads to the activation of nuclear The ligands for NOD1 and NOD2 are distinct and
factor-κB (NF-κB), which regulates the expression of target genes that encode pro-inflammatory the minimal structural requirements have been defined.
mediators. Several laboratories have developed agonists of TLR4 and the endosomal TLRs that Both are derived from bacterial peptidoglycan: NOD2
can stimulate TLR signalling and should be useful therapeutics for treating various human
recognizes ligands with muramyl-dipeptidic structures,
diseases. The key steps that could be targeted in the development of molecules to inhibit
TLR signalling are also indicated. IκB, inhibitor of NF-κB; IKK, IκB kinase; IRAK, interleukin-1- whereas NOD1 binds to tripeptide structures in
receptor-associated kinase; MPL, monophosphoryl lipid A; ODN, oligodeoxynucleotide; which the terminal amino acid is diaminopimelic acid.
RIP, receptor-interacting protein; TRAF, tumour-necrosis-factor-receptor-associated factor 6. Interestingly, the binding of tripeptides to NOD1

NATURE REVIEWS | IMMUNOLOGY VOLUME 4 | JULY 2004 | 5 1 5


REVIEWS

Agonists/
alter the prevailing T helper 2 (TH2)-cell-type response
antagonists in sensitized individuals. Therefore, longer-lasting
treatments afforded by manipulating the innate
Ligand-
recognition immune system would provide a great advance.
domain Immunotherapy can result in decreased symptoms,
reduced medication use and enhanced quality of life. It
NOD NOD Potential can also prevent the transition from allergic disease to
protein inhibitory
site
asthma. So, immunotherapy might have therapeutic as
CARD well as preventive effects. However, the present design
Inhibit of treatments is limited by our lack of understanding of
CARD–CARD the essential mechanisms that are involved in triggering
interactions
chronic inflammatory disease. For example, in asthma,
we do not have a mechanistic explanation for the selec-
RICK tive activation of TH2 cells39–41. Nonetheless, it is clear
Kinase domain that one aim of treating asthma with immunomodula-
tors would be to try to change the TH-cell balance in asth-
NF-κB activation and matic individuals: that is, from the dominant TH2-type
caspase activation inflammation to a TH1-type response. Central to this
approach is the use of synthetic TLR ligands. In partic-
ular, there has been a sustained effort to use oligonu-
Secretion of pro-inflammatory
cytokines (TNF and IL-1β) cleotides that contain non-methylated CpG motifs to
shift the balance to a TH1-type response40. This strategy
stems from the ‘hygiene hypothesis’, which states that
Immune responses
the marked reduction in infection in children in mod-
ern society produces decreased TH1-type responses and
Figure 3 | NOD1 and NOD2 as targets for therapy. concomitant TH2-type inflammation. Approaches to
Signalling pathways mediated by the nucleotide-binding treat patients using TH1-type cytokines have failed and,
oligomerization domain (NOD)-protein-family members in some cases, have caused serious side effects. There
NOD1 and NOD2 are dependent on the receptor-interacting are numerous preclinical studies showing that admin-
serine/threonine kinase (RICK) to initiate downstream istration of the TLR9 ligand CpG ODN induces the
signalling. RICK interacts with NOD1 and NOD2 through
production of TH1-type cytokines, blocks the produc-
homophilic caspase-recruitment domain (CARD)–CARD
interactions. Key steps that could be targeted for the
tion of TH2 cytokines and prevents the symptoms of
development of therapeutics are shown. IL, interleukin; asthma. Treatment with CpG ODNs seems to reduce
NF-κB, nuclear factor-κB; TNF, tumour-necrosis factor. disease, without the serious side effects produced fol-
lowing the administration of TH1-type cytokines.
However, at present, there is too little attention paid to
does not require the presence of a muramic acid the potential side effects of chronic administration of
backbone. Neither NOD1 nor NOD2 seem to recognize immunostimulatory molecules; potential drawbacks
polymeric peptidoglycan structures, regardless of the of this type of therapy are discussed later.
bacterial source.
Adjuvants/immunostimulation. Another area that
Targeting innate immunity in human disease could benefit from targeting the innate immune
Targeting the innate immune system could provide response is the development of new adjuvants for vac-
therapeutic benefit in both acute and chronic human cines. Vaccines are crucial for the management and pre-
diseases. Efforts are also being directed towards using vention of infectious disease. Although our endogenous
TLR ligands in vaccines. Examples of each of these defences mount effective responses to microorganisms,
applications are discussed here. it seems that enhanced innate immunity could be bene-
ficial in specific settings. Previously, such enhancement
Asthma. Considerable effort is going into the design of of immune responses has been achieved using complex
new approaches to treat asthma. The prevalence of mixtures of microbial products, which most probably
allergic diseases, including asthma, has increased function through the innate immune system to enhance
markedly in the past three decades. Unfortunately, adaptive immune responses42,43. In addition, we now
there are few new drugs that offer cures for these prob- know that several types of TLR ligand are efficacious as
lems, or even provide control or treatment. We still rely vaccine adjuvants44,45. This has been demonstrated for
on glucocorticoids and bronchodilators, used in com- synthetic TLR7 agonists (imiquimod and resiquimod),
bination and in various forms of administration, as the TLR9 agonists (CpG ODNs) and TLR4 agonists (lipid A
standard practice. Although they are effective, these analogues), thereby indicating the merits of identifying
drugs need to be administered at least twice daily and synthetic TLR agonists. Here, I discuss a few important
are not without long-term effects, especially in the examples that involve TLR4, TLR7 and TLR9 (TABLE 2).
increasingly young patient population. Furthermore, There are now considerable data indicating that the use
corticosteroids that block inflammatory cells do not of synthetic TLR4 agonists can be beneficial. Some of

516 | JULY 2004 | VOLUME 4 www.nature.com/reviews/immunol


FOCUS ON TLR SIGNALLING

the most compelling data have been obtained using caused by human papillomavirus51,53. The imidazole
monophosphoryl lipid A (MPL)46–48. This compound is quinolines are the first class of TLR agonists to be
a chemically modified derivative of the lipid A moiety of used for several clinical applications in humans.
LPS. MPL is considerably less toxic than LPS but has a Another compound, loxoribine (7-allyl-7,8-dihydro-
similar immunostimulatory activity. MPL has been 8-oxo-guanosine), has antitumour activity and func-
used extensively in clinical trials, in formulations of tions through a TLR7– MyD88-dependent pathway,
both prophylactic and therapeutic vaccines that have which was demonstrated using TLR7- and MyD88-
targeted various diseases, including cancer, infectious deficient mice as a source of cells54–56. The activity of
disease and allergies. It seems to have a strong safety loxoribine and resiquimod depend on endosomal
profile and good efficacy. A set of related compounds maturation and presumably acidification. In this
that are also lipid A mimetics, aminoalkylglucosaminide regard, TLR7 and TLR8 are similar to TLR9 and seem
4-phosphates (AGPs), have been found to be TLR4 to form a subgroup within the TLR family, which
ligands and some molecules from this new group of functions in endosomal/lysosomal compartments.
immunostimulants will undoubtedly move into clinical TLR7 also recognizes bropirimine, an immuno-
trials. In particular, one derivative, RC-529, has been stimulant known to have antitumour effects in super-
used in a hepatitis B vaccine trial, replacing the standard ficial transitional-cell carcinoma of the bladder.
alum-based adjuvant. In nearly all of those vaccinated Bropirimine might have direct effects on tumour cells
with an MPL adjuvant, protective antibody levels were rather than functioning through inducible mediators;
reached after only two doses of the vaccine, in contrast however, more study is required to understand its
to the current protocol in which three doses are mechanism of action.
required to obtain similar levels of protection.
Recently, single-stranded RNA has been shown to be Infection. Last, one might consider targeting innate
a ligand for mouse TLR7 and human TLR8 (REFS 49,50). immune pathways in the acute setting, where there is
However, it is well known that several small-molecule local or systemic infection. Here, blunting the pro-
immunostimulants are also ligands for TLR7 and inflammatory cascade might be achieved using recep-
TLR8. The prototypical compound, resiquimod, tor antagonists based on inhibitory antibodies or
induces IFN-α and other cytokines. Resiquimod is antagonistic ligand mimetics. Another approach could
one of a class of imidazole quinolines and has been be to induce protection by activating TLRs and/or
the subject of several clinical investigations. It has NOD1/2 to elicit nonspecific resistance effects and
been shown to be an immune-response modifier that enhance host resistance to bacterial or viral infection.
enhances cutaneous immune responses: it is effective Workers at Corixa Corporation have extended the pre-
against external genital warts and other diseases vious studies of AGPs (lipid A mimetics) and showed

Table 1 | Toll-like receptor ligands


Receptor Naturally occurring Synthetic analogues Fully synthetic small molecules
Exogenous ligands
TLR1 N.D. Triacyl lipopeptides –
TLR2 Lipoproteins/lipopeptides Di- and triacyl lipopeptides –
Peptidoglycan
Lipoteichoic acid
Lipoarabinomannan
Atypical lipopolysaccharide
TLR3 Double-stranded RNA Poly I:C –
TLR4 Lipopolysaccharide LPS/lipid A mimetics, such as MPL Synthetic lipid A, E5564
Taxol*
HSP60 (Chlamydia pneumoniae)
TLR5 Bacterial flagellin Discontinuous 13-amino-acid peptide –
TLR6 N.D. Diacyl lipopeptides –
TLR7 (G+U)-rich single-stranded RNA* Oligonucleotides Imidazole quinolines (imiquimod, resiquimod)
Guanosine nucleotides (loxoribine)
TLR8 (G+U)-rich single-stranded RNA‡ – Imidazole quinolines (imiquimod)
TLR9 Bacterial DNA CpG oligodeoxynucleotides –
Viral DNA
Other DNA with low content of
non-methylated CpG sequences
Endogenous ligands
TLR2 HSP70 – –
TLR4 HSP60 – –
Oligosaccharides of hyaluronic acid
*Ligand for mouse TLR only. ‡Ligand for human TLR only. HSP, heat-shock protein; LPS, lipopolysaccharide; MPL, monophosphoryl lipid A; N.D., not determined; Poly I:C,
polyinosinic–polycytidylic acid; TLR, Toll-like receptor.

NATURE REVIEWS | IMMUNOLOGY VOLUME 4 | JULY 2004 | 5 1 7


REVIEWS

Table 2 | Toll-like receptors as drug targets in clinical trials non-selectively block IKK activity using non-steroidal
anti-inflammatory drugs. At present, the most
Receptor Ligand Application References
progress towards developing selective inhibitors is in
TLR4 MPL adjuvant Adjuvant allergy treatment 78,79
the development of selective small-molecule inhibitors
MPL adjuvant Vaccines 46,80,81 of IKK-β. These might have applications in controlling
Ribi 529 Vaccines 81 events downstream of TLR activation in chronic dis-
E5564 Endotoxaemia, liver disease 82–84 ease. In particular, it is possible to envisage the use of
TLR7 Loxoribine Cancer 85 such inhibitors to treat chronic inflammatory diseases,
Resiquimod Increased cutaneous immune
in which sterile inflammation that results from TLR
activation is important. The use of other selective
responses 86–89
kinase inhibitors could also potentially control TLR-
Imiquimod Basal-cell carcinoma 90–92 driven inflammation. The kinase targets might include
TLR9 CpG ODNs Vaccine adjuvant 93–96 IRAK4, RICK and the mitogen-activated protein
CpG ODNs Melanoma 97 kinases; this topic is not considered further here.
CpG ODNs Non-Hodgkin lymphoma 97 The current progress in defining the specific roles
MPL, monophosphoryl lipid A; ODN, oligodeoxynucleotide; TLR, Toll-like receptor.
of members of the MyD88-adaptor family in TLR sig-
nalling offers new possibilities for the selective block-
ade of pathways downstream of the TLRs. We now
that they provide remarkable protection in Listeria or require the development of new therapeutics to target
influenza-virus challenge models, in which there is specific protein–protein interactions. This has proven
no administration of the microorganism itself or to be a formidable barrier, with only limited progress
additional antigen57. being made. Screening of conventional chemical
Considerable data indicate a role for the innate libraries comprising small-molecule drug-like struc-
immune system in antiviral responses. The work of tures has not provided compounds able to block
Finberg and colleagues58 supports the contention that protein–protein interactions. In part, this results from
viral proteins might directly activate TLRs, such as the typical flatness of the interface that characterizes
TLR4, thereby resulting in an antiviral response. protein–protein interactions. Large surface areas are
Subsequently, it was shown that stimulation of TLR3 involved, usually 750–1500 Å2, and there are no deep
and TLR4 results in the induction of a subset of genes, cavities present that could function as small-molecule-
including those that encode type I interferons, which binding sites. Many researchers, including Wells and
inhibit viral replication59. Signalling from TLR3 and his co-workers70 have defined ‘hot spots’ that represent
TLR4 has been shown to involve both the NF-κB and the most important regions for high-affinity binding.
IRF3 pathways. More recently, a detailed understanding As we accumulate knowledge about each type of mole-
of these pathways has emerged from the efforts of cule, the TLRs, NOD-protein-family members and
Beutler and others60–65. In particular, it has been shown adaptor proteins should become more tractable tar-
that pathways linked to TLR3 and TLR9 contribute to gets. There are several strong examples that support
innate responses to systemic viral infection. Although the contention that progress is being made. One is in
neither pathway alone has a dominant role, mutations the unrelated area of p53 interactions but the other
affecting both result in markedly decreased antiviral provides data about the targeting of TIR domains. In
responses. Therefore, targeting the TLR3, TLR4 and the former, Vassilev et al.71 have discovered a group of
TLR9 pathways allows for an enhanced response to cer- synthetic compounds, known as nutlins, that specifi-
tain viral infections. There are numerous small-molecule cally block p53–MDM2 interactions. This finding fol-
agonists that have the potential to enhance resistance to lowed a series of elegant studies by others, which were
viral infection. A new strategy involving covalent linkage based on designing peptidic inhibitors and then iden-
of antigen to CpG ODNs promises to provide an tifying the specific binding site of p53 that mediates
effective way of improving the immunomodulatory MDM2 binding. This work could well provide a model
activities of CpG, by enhancing antigen uptake and for future efforts in the design of drugs to target innate
stimulation of antigen-presenting cells66,67. This con- immune pathways.
cept has been extended to clinical trials with the use of In fact, Rebek and colleagues72 have provided the
allergen–CpG ODN conjugates68. much needed proof-of-concept that TIR domains
could be targets for selective inhibitors. Using the
Targeting intracellular pathways TIR-domain interactions between MyD88 and the
The important role of NF-κB in TLR signalling is now type 1 IL-1R (IL-1R1) as a model, they synthesized
well established. The NF-κB-signalling pathways are and characterized a low-molecular-weight MyD88
one of the main targets for new drug discovery and mimic, hydrocinnamoyl-l-valyl-pyrrolidine. This
development69. Numerous studies have shown that compound was developed by modelling a tripeptide
involvement of the IKK complex (inhibitor of NF-κB sequence of the BB-loop of the TIR domain —
(IκB)-kinase complex), a downstream element in (F/Y)-(V/L/I)-(P/G). The compound seems to be
TLR-signalling pathways, is essential. The two kinases selective in both cell-based and in vivo systems.
present in the IKK complex, IKK-α and IKK-β, are Development of this low-molecular-weight inhibitor
potential targets, and recent studies have been able to of the TIR-domain interaction between MyD88 and

518 | JULY 2004 | VOLUME 4 www.nature.com/reviews/immunol


FOCUS ON TLR SIGNALLING

IL-1R1 indicates that the design of further selective functions and the potentially harmful effects that
inhibitors of TIR-domain-containing proteins is result from TLR activation remains a serious challenge
within reach. This promises to yield an entirely new for those designing new therapeutics.
generation of potential anti-inflammatory compounds
to modulate the innate immune system. Conclusions
We now have detailed knowledge of the ligands, receptors
Complications with therapeutics and intracellular-signalling pathways that control innate
As therapeutics, CpG ODNs might require administra- immune responses during infection and inflammation,
tion over long time periods. The effect of chronic and this provides new approaches for highly selective
administration has not yet been addressed in detail. therapeutics. New information about the role of NOD2
However, a recent publication reported the conse- in human disease exemplifies how genetic approaches
quences of chronic CpG ODN administration73: mice will increase our understanding of disease pathogenesis
given daily injections of CpG ODNs developed consid- and how best to target individual proteins or pathways.
erable pathology in their lymphoid organs, with The potential to modulate the innate immune system is
changes in both structure and function. Specifically, the supported by the first generation small-molecule ligands
report showed destruction of lymphoid-follicle archi- of TLR7 and TLR8, the imidazole quinolines. This class of
tecture and suppression of follicular dendritic cells and drug was developed for topical use, targeting papilloma-
germinal-centre B lymphocytes. Furthermore, after virus-induced warts, without understanding its exact
three weeks, multi-focal liver necrosis and haemor- mechanism of action. Furthermore, the ongoing trials
rhagic ascites developed. These effects were dependent based on targeting TLR9 with CpG ODNs will enhance
on both CpG ODNs and TLR9. Additional studies will our knowledge of the effects of blocking or enhancing
be required to determine whether the administration selective pathways of innate immunity. It should now be
other TLR ligands induces similar effects and how the possible to use our combined knowledge of the TLR- and
CpG-ODN-induced injury correlates with dosage and NOD-protein-family members, together with the
its more-beneficial effects. recently developed tools for drug discovery, to identify
The use of TLR antagonists might have applica- new therapeutic molecules with selective actions. Our
tions in acute settings, in which dampening the pro- progress will only be limited by our ability to express the
inflammatory response of the innate immune system key proteins of the innate immune system for use in
is beneficial. In particular, TLR antagonists could be high-throughput screening assays, and our ability to
used to treat septic shock. However, the important develop small-molecule libraries containing chemical
role of the innate immune system in host defence scaffolds that are directed towards the hot spots of pro-
against microorganisms indicates that only limited tein–protein interactions. Hopefully, the fruits of
uses of antagonists will be found. Certainly, their use ongoing basic research are within reach of the drug-
in any chronic settings will require a high degree of development community, and we will soon see how
selectivity among the various TLRs. Determining our newly found knowledge can be used to better treat
how to maintain the balance between host-defence human disease.

1. Yamamoto, M., Takeda, K. & Akira, S. TIR domain- Provides the first definition of the structural 23. Nishiya, T. & DeFranco, A. L. Ligand-regulated chimeric
containing adaptors define the specificity of TLR signaling. requirements of the NOD1 ligand. receptor approach reveals distinctive subcellular localization
Mol. Immunol. 40, 861–868 (2004). 13. O’Neill, L. A. et al. Mal and MyD88: adapter proteins and signaling properties of the Toll-like receptors. J. Biol.
2. Beutler, B., Hoebe, K., Du, X. & Ulevitch, R. J. How we involved in signal transduction by Toll-like receptors. Chem. 279, 19008–19017 (2004).
detect microbes and respond to them: the Toll-like receptors J. Endotoxin Res. 9, 55–59 (2003). 24. Miyake, K. Endotoxin recognition molecules, Toll-like
and their transducers. J. Leukoc. Biol. 74, 479–485 (2003). 14. Martin, M. U. & Wesche, H. Summary and comparison of receptor 4–MD-2. Semin. Immunol. 16, 11–16 (2004).
3. Underhill, D. M. Toll-like receptors: networking for success. the signaling mechanisms of the Toll/interleukin-1 receptor 25. Akashi, S. et al. Lipopolysaccharide interaction with
Eur. J. Immunol. 33, 1767–1775 (2003). family. Biochim. Biophys. Acta 1592, 265–268 (2002). cell surface Toll-like receptor 4–MD-2: higher affinity than
4. O’Neill, L. A. The role of MyD88-like adapters in Toll-like 15. O’Neill, L. A., Fitzgerald, K. A. & Bowie, A. G. The Toll–IL-1 that with MD-2 or CD14. J. Exp. Med. 198, 1035–1042
receptor signal transduction. Biochem. Soc. Trans. 31, receptor adaptor family grows to five members. Trends (2003).
643–647 (2003). Immunol. 24, 286–290 (2003). 26. Wetzler, L. M. The role of Toll-like receptor 2 in microbial
5. Dunne, A. & O’Neill, L. A. The interleukin-1 receptor/Toll–like 16. Ogura, Y. et al. Nod2, a Nod1/Apaf-1 family member that disease and immunity. Vaccine 21, S55–S60 (2003).
receptor superfamily: signal transduction during inflammation is restricted to monocytes and activates NF-κB. J. Biol. 27. Kirschning, C. J. & Schumann, R. R. TLR2: cellular sensor
and host defense. Sci. STKE [online] 2003, re3 (2003). Chem. 276, 4812–4818 (2001). for microbial and endogenous molecular patterns. Curr.
6. Takeda, K., Kaisho, T. & Akira, S. Toll-like receptors. Annu. 17. Kobayashi, K. et al. RICK/Rip2/CARDIAK mediates Top. Microbiol. Immunol. 270, 121–144 (2002).
Rev. Immunol. 21, 335–376 (2003). signalling for receptors of the innate and adaptive immune 28. Hajjar, A. M. et al. Cutting edge: Functional interactions
7. Chamaillard, M.,Girardin, S. E., Viala, J. & Philpott, D. J. systems. Nature 416, 194–199 (2002). between Toll-like receptor (TLR) 2 and TLR1 or TLR6 in
Nods, Nalps and Naip: intracellular regulators of bacterial- Provides important evidence implicating RICK in both response to phenol-soluble modulin. J. Immunol. 166,
induced inflammation. Cell. Microbiol. 5, 581–592 (2003). TLR- and NOD-protein signalling pathways. 15–19 (2001).
8. Inohara, N. & Nunez, G. NODs: intracellular proteins involved 18. Hoffmann, J. A. The immune response of Drosophila. 29. Ozinsky, A. et al. The repertoire for pattern recognition
in inflammation and apoptosis. Nature Rev. Immunol. Nature 426, 33–38 (2003). of pathogens by the innate immune system is defined by
5, 371–382 (2003). 19. Gobert, V. et al. Dual activation of the Drosophila Toll cooperation between Toll-like receptors. Proc. Natl Acad.
9. Akira, S. & Takeda, K. Toll-like receptor signalling. Nature pathway by two pattern recognition receptors. Science Sci. USA 97, 13766–13771 (2000).
Rev. Immunol. 4, 499–511 (2004). 302, 2126–2130 (2003). Provides key data showing how TLRs comprise
10. Girardin, S. E. et al. Peptidoglycan molecular requirements 20. Imler, J. L. & Hoffmann, J. A. Toll signaling: the TIReless heteromeric subunits that determine ligand
allowing detection by Nod1 and Nod2. J. Biol. Chem. 278, quest for specificity. Nature Immunol. 4, 105–106 (2003). specificity.
41702–41708 (2003). 21. Gottar, M. et al. The Drosophila immune response against 30. Mizel, S. B., Honko, A. N., Moors, M. A., Smith, P. S.
11. Chamaillard, M. et al. An essential role for NOD1 in host Gram-negative bacteria is mediated by a peptidoglycan & West, A. P. Induction of macrophage nitric oxide
recognition of bacterial peptidoglycan containing recognition protein. Nature 416, 640–644 (2002). production by Gram-negative flagellin involves signaling
diaminopimelic acid. Nature Immunol. 4, 702–707 (2003). Provides a new paradigm that describes different arms via heteromeric Toll-like receptor 5/Toll-like receptor 4
12. Girardin, S. E. et al. Nod1 detects a unique muropeptide of the innate immune response in insects. complexes. J. Immunol. 170, 6217–6223 (2003).
from Gram-negative bacterial peptidoglycan. Science 22. Zhang, D. et al. A Toll-like receptor that prevents infection by 31. Schulze-Lefert, P. Plant immunity: the origami of receptor
300, 1584–1587 (2003). uropathogenic bacteria. Science 303, 1522–1526 (2004). activation. Curr. Biol. 14, R22–R24 (2004).

NATURE REVIEWS | IMMUNOLOGY VOLUME 4 | JULY 2004 | 5 1 9


REVIEWS

32. Nimchuk, Z., Eulgem, T., Holt, B. F. & Dangl, J. L. Among the first reports to indicate a role for the innate 81. Vernacchio, L. et al. Effect of monophosphoryl lipid A
Recognition and response in the plant immune system. immune system in antiviral responses. (MPL) on T-helper cells when administered as an
Annu. Rev. Genet. 37, 579–609 (2003). 59. Doyle, S. et al. IRF3 mediates a TLR3/TLR4-specific antiviral adjuvant with pneumocococcal-CRM197 conjugate
33. Martin, G. B., Bogdanove, A. J. & Sessa, G. Understanding gene program. Immunity 17, 251–263 (2002). vaccine in healthy toddlers. Vaccine 20, 3658–3667
the functions of plant disease resistance proteins. Annu. 60. Hoebe, K. & Beutler, B. LPS, dsRNA and the interferon (2002).
Rev. Plant Biol. 54, 23–61 (2003). bridge to adaptive immune responses: Trif, Tram, and 82. Lynn, M. et al. Extended in vivo pharmacodynamic
34. Martinon, F. & Tschopp, J. Inflammatory caspases; linking other TIR adaptor proteins. J. Endotoxin Res. 10, 130–136 activity of E5564 in normal volunteers with experimental
an intracellular innate immune system to autoinflammatory (2004). endotoxemia. J. Pharmacol. Exp. Ther. 308, 175–181
diseases. Cell 117, 561–574 (2004). 61. Tabeta, K. et al. Toll-like receptors 9 and 3 as essential (2004).
35. Tanabe, T. et al. Regulatory regions and critical residues components of innate immune defense against mouse 83. Wong, Y. N. et al. Safety, pharmacokinetics, and
of NOD2 involved in muramyl dipeptide recognition. EMBO cytomegalovirus infection. Proc. Natl Acad. Sci. USA 101, pharmacodynamics of E5564, a lipid A antagonist, during
J. 23, 1587–1597 (2004). 3516–3521 (2004). an ascending single-dose clinical study. J. Clin. Pharmacol.
36. Lund, J., Sato, A., Akira, S., Medzhitov, R. & Iwasaki, A. 62. Hoebe, K. et al. Upregulation of costimulatory molecules 43, 735–742 (2003).
Toll-like receptor 9-mediated recognition of Herpes simplex induced by lipopolysaccharide and double-stranded RNA 84. Liang, E. et al. Pharmacokinetics of E5564, a
virus-2 by plasmacytoid dendritic cells. J. Exp. Med. 198, occurs by Trif-dependent and Trif-independent pathways. lipopolysaccharide antagonist, in patients with impaired
513–520 (2003). Nature Immunol. 4, 1223–1229 (2003). hepatic function. J. Clin. Pharmacol. 43, 1361–1369
37. Krug, A. et al. Herpes simplex virus type 1 activates murine 63. Beutler, B. et al. Lps2 and signal transduction in sepsis: (2003).
natural interferon producing cells through Toll-like receptor 9. at the intersection of host responses to bacteria and viruses. 85. Agarwala, S. S., Kirkwood, J. M. & Bryant, J. Phase 1,
Blood 103, 1433–1437 (2004). Scand. J. Infect. Dis. 35, 563–567 (2003). randomized, double-blind trial of 7-allyl-8-oxoguanosine
38. Kaisho, T. & Akira, S. Regulation of dendritic cell function 64. Honda, K. et al. Selective contribution of IFN-α/β signaling (loxoribine) in advanced cancer. Cytokines Cell. Mol.
through Toll-like receptors. Curr. Mol. Med. 3, 373–385 to the maturation of dendritic cells induced by double- Ther. 6, 171–176 (2000).
(2003). stranded RNA or viral infection. Proc. Natl Acad. Sci. USA 86. Sauder, D. N., Smith, M. H., Senta-McMillian, T.,
39. El Biaze, M. et al. T cell activation, from atopy to asthma: 100, 10872–10877 (2003). Soria, I. & Meng T. C. Randomized, single-blind,
more a paradox than a paradigm. Allergy 58, 844–853 65. Matsumoto, M., Funami, K., Oshiumi, H. & Seya, T. Toll-like placebo-controlled study of topical application of
(2003). receptor 3: a link between Toll-like receptor, interferon and the immune response modulator resiquimod in healthy
40. Hussain, I. & Kline, J. N. CpG oligodeoxynucleotides: viruses. Microbiol. Immunol. 48, 147–154 (2004). adults. Antimicrob. Agents Chemother. 47, 3846–3852
a novel therapeutic approach for atopic disorders. 66. Heit, A. et al. Cutting edge: Toll-like receptor 9 expression (2003).
Curr. Drug Targets Inflamm. Allergy 2, 199–205 (2003). is not required for CpG DNA-aided cross-presentation of 87. Jones, T. Resiquimod 3M. Curr. Opin. Investig. Drugs
41. Kuchroo, V. K., Umetsu, D. T., DeKruyff, R. H. & DNA-conjugated antigens but essential for cross-priming 4, 214–218 (2003).
Freeman, G. J. The TIM gene family: emerging roles of CD8 T cells. J. Immunol. 170, 2802–2805 (2003). 88. Imiquimod (Aldara) for actinic keratoses. Med. Lett. Drugs
in immunity and disease. Nature Rev. Immunol. 67. Heit, A. et al. CpG-DNA aided cross-priming by cross- Ther. 46, 42–44 (2004).
3, 454–462 (2003). presenting B cells. J. Immunol. 172, 1501–1507 (2004). 89. Prinz, B. M. et al. Treatment of Bowen’s disease with
42. Melief, C. J. Strategies for immunotherapy of cancer. 68. Speigelberg, H. L., Horner, A. A., Takabayashi, K. & Raz, E. imiquimod 5% cream in transplant recipients.
Adv. Immunol. 75, 235–282 (2000). Allergen-immunostimulatory oligodeoxynucleotide Transplantation 77, 790–791 (2004).
43. Hess, J., Schaible, U., Raupach, B. & Kaufmann, S. H. conjugate: a novel allergoid for immunotherapy. Curr. Opin. 90. Kreuter, A. et al. Treatment of anal intraepithelial
Exploiting the immune system: toward new vaccines against Allergy Clin. Immunol. 2, 547–551 (2002). neoplasia in patients with acquired HIV with imiquimod
intracellular bacteria. Adv. Immunol. 75, 1–88 (2000). 69. Karin, M., Yamamoto, Y. & Wang, Q. M. The IKK NF-κB 5% cream. J. Am. Acad. Dermatol. 50, 980–981 (2004).
44. Kaisho, T. & Akira, S. Toll-like receptors as adjuvant system: a treasure trove for drug development. Nature 91. Huber, A. et al. Topical imiquimod treatment for nodular
receptors. Biochim. Biophys. Acta 1589, 1–13 (2002). Rev. Drug Discov. 3, 17–26 (2004). basal cell carcinomas: an open-label series. Dermatol.
45. Jiang, Z. H. & Koganty, R. R. Synthetic vaccines: the role 70. Arkin, M. R. & Wells, J. A. Small-molecule inhibitors of Surg. 30, 429–430 (2004).
of adjuvants in immune targeting. Curr. Med. Chem. 10, protein–protein interactions: progressing towards the 92. Herbert, W. C. Imiquimod and the treatment of cutaneous
1423–1439 (2003). dream. Nature Rev. Drug Discov. 3, 301–317 (2004). T-cell proliferative diseases: at the threshold. Skinmed.
46. Evans, J. T. et al. Enhancement of antigen-specific immunity 71. Vassilev, L. T. et al. In vivo activation of the p53 pathway 2, 273–274 (2003).
via the TLR4 ligands MPL adjuvant and Ribi. 529. Expert by small-molecule antagonists of MDM2. Science 303, 93. Leifer, C. A., Verthelyi, D. & Klinman, D. M. Heterogeneity
Rev. Vaccines 2, 219–229 (2003). 844–848 (2004). in the human response to immunostimulatory CpG
47. Persing, D. H. Taking Toll: lipid A mimetics as adjuvants and 72. Bartfai, T. et al. A low molecular weight mimic of the Toll/IL-1 oligodeoxynucleotides. J. Immunother. 26, 313–319
immunomodulators. Trends Microbiol. 10, S32–S37 (2002). receptor/resistance domain inhibits IL-1 receptor-mediated (2003).
48. Stover, A. G. et al. Structure-activity relationship of synthetic responses. Proc. Natl Acad. Sci. USA 100, 7971-7976 94. Tulic, M. K. Amb a 1-immunostimulatory
Toll-like receptor 4 agonists. J. Biol. Chem. 279, 4440–4449 (2003). oligodeoxynucleotide conjugate immunotherapy
(2004). Provides the first evidence indicating that it is decreases the nasal inflammatory response. J. Allergy
49. Heil, F. et al. Species-specific recognition of single-stranded possible to target TIR–TIR domain interactions Clin. Immunol. 113, 234–241 (2004).
RNA via Toll-like receptor 7 and 8. Science 303, 1526–1529 with a small molecule. 95. Bohle, B. AIC. Dynavax. Curr. Opin. Investig. Drugs
(2004). 73. Heikenwalder, M. et al. Lymphoid follicle destruction and 4, 603–607 (2003).
Identifies a naturally occurring ligand for TLR 7 and immunosuppression after repeated CpG 96. Immunostimulatory DNA–Dynavax. AIC, Amb a 1
TLR 8. oligodeoxynucleotide administration. Nature Med. 10, immunostimulatory conjugate, HBV-ISS, ISS 1018, ISS
50. Diebold, S. S., Kaisho, T., Hemmi, H., Akira, S. & 187–192 (2004). DNA, ISS DNA–dynavax, ISS1, ISS2. Drugs R D 3, 193–196
Reis e Sousa, C. Innate antiviral responses by means 74. Hoebe, K. et al. Identification of Lps2 as a key transducer (2002).
of TLR7-mediated recognition of single-stranded RNA. of MyD88-independent TIR signalling. Nature 424, 743–748 97. Paul, S. Technology evaluation: CpG-7909, Coley.
Science 303, 1529–1531 (2004). (2003). Curr. Opin. Mol. Ther. 5, 553–559 (2003).
Identifies a naturally occurring ligand for TLR7. This seminal paper describes the use of forward 98. Medzhitov, R., Preston-Hurlburt, P. & Janeway, C. A. Jr.
51. Skinner R. B. Jr. Imiquimod. Dermatol. Clin. 21, 291–300 genetics to decipher TLR-signalling pathways and A human homologue of the Drosophila Toll protein signals
(2003). identifies Lps2 as a gene encoding TRIF. activation of adaptive immunity. Nature 388, 394–397
52. Hengge, U. R. & Cusini, M. Topical immunomodulators for 75. Kawai, T., Adachi, O., Ogawa, T., Takeda, K. & Akira, S. (1997).
the treatment of external genital warts, cutaneous warts and Unresponsiveness of MyD88-deficient mice to endotoxin.
molluscum contagiosum. Br. J. Dermatol. 149 (Suppl. 66), Immunity 11, 115–122 (1999). Acknowledgements
15–19 (2003). 76. Yamamoto, M. et al. TRAM is specifically involved in the I thank J. C. Mathison for comments and editorial support and
53. Bernard, H. U. Established and potential strategies against Toll-like receptor 4-mediated MyD88-independent signaling P. Rutledge for administrative support. This work was supported
papillomavirus infections. J. Antimicrob. Chemother. 53, pathway. Nature Immunol. 4, 1144–1150 (2003). by the National Institutes of Health, The Charles Dana Foundation
137–139 (2004). 77. Fitzgerald, K. A. et al. LPS–TLR4 signaling to IRF-3/7 and The Novartis Foundation.
54. Akira, S. & Hemmi, H. Recognition of pathogen-associated and NF-κB involves the Toll adapters TRAM and TRIF.
molecular patterns by TLR family. Immunol. Lett. 85, 85–95 J. Exp. Med. 198, 1043–1055 (2003). Competing interests statement
(2003). 78. Drachenberg, K. J., Heinzkill, M., Urban, E. & Woroniecki, S. R. The author declares competing financial interests: See Web version
55. Heil, F. et al. The Toll-like receptor 7 (TLR7)-specific stimulus Efficacy and tolerability of short-term specific immunotherapy for details.
loxoribine uncovers a strong relationship within the TLR7, 8 with pollen allergoids adjuvanted by monophosphoryl lipid A
and 9 subfamily. Eur. J. Immunol. 33, 2987–2997 (2003). (MPL) for children and adolescents. Allergol. Immunopathol.
56. Hemmi, H. et al. Small anti-viral compounds activate (Madr.) 31, 270–277 (2003). Online links
immune cells via the TLR7 MyD88-dependent signaling 79. Mothes, N. et al. Allergen-specific immunotherapy with
pathway. Nature Immunol. 3, 196–200 (2002). a monophosphoryl lipid A-adjuvanted vaccine: reduced DATABASES
57. Baldridge, J. R. et al. Immunostimulatory activity of seasonally boosted immunoglobulin E production and The following terms in this article are linked online to:
aminoalkyl glucosaminide 4-phosphates (AGPs): induction inhibition of basophil histamine release by therapy-induced Entrez Gene:
of protective innate immune responses by RC-524 and blocking antibodies. Clin. Exp. Allergy 33, 1198–1208 http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene
RC-529. J. Endotoxin Res. 8, 453–458 (2002). (2003). CD14 | IKK-α | IKK-β | IL-1R1 | IRAK1 | IRAK4 | IRF3 | MD2 |
58. Kurt-Jones, E. A. et al. Pattern recognition receptors TLR4 80. Bienzle, U. et al. Immunization with an adjuvant hepatitis B MyD88 | NOD1 | NOD2 | RICK | SARM | TIRAP | TLR1 | TLR2 |
and CD14 mediate response to respiratory syncytial virus. vaccine after liver transplantation for hepatitis B-related TLR3 | TLR4 | TLR5 | TLR6 | TLR7 | TLR8 | TLR9 | TLR11 | TRAF6 |
Nature Immunol. 1, 398–401 (2000). disease. Hepatology 38, 811–819 (2003). TRAM | TRIF

520 | JULY 2004 | VOLUME 4 www.nature.com/reviews/immunol

You might also like