Nature Immunology
2, 9 - 10 (2001)
doi:10.1038/83222
Toll we meet again . . .Bruce Beutler
& Alexander PoltorakDepartment of Immunology, The Scripps Research Institute
, 10550 N. Torrey Pines Road, La Jolla,
CA 92037, USA. (bruce@scripps.edu)
Remodeling of chromatin, distinct from activation of NF- B,
is a newly identified function of Toll-like receptors in mammals. In Drosophila
, Toll receptors play a role in development as well as immunity; yet in
mammals, they have an immunological role only. Further clues to the evolutionary
development of the Toll receptors are emerging.More than half-a-billion years ago, when the progenitors of both insects
and mammals lived on earth, a system for innate immunity was already well
entrenched among the Bilateria. To recognize pathogens, primitive organisms
relied upon a system of receptors, each endowed with a cytoplasmic feature
that, far in the future, would be known as the "TIR", for Toll−interleukin
1 (IL-1) receptor domain. Mutational pressure and natural selection ultimately
tore the genome of our common ancestor asunder. Some proteins represented
in mammals no longer exist in flies, whereas some proteins in flies have no
counterparts in mammals1,
2. Yet the TIR domain has persisted
in each line and, to an amazing degree, shows evidence of a common immunological
function.
In this issue of Nature Immunology, Weinmann et al.3 have identified one more function of the TIR domain. In brief, the
authors have shown that remodeling of chromatin structure in macrophages is
one of the endpoints of TIR domain signal transduction, an endpoint that is
split from activation of nuclear factor− B (NF-B), the
response studied by many workers in the mammalian Toll-like receptor (TLR)
field. Although Weinmann et al. chose to examine accessibility of the
IL-12 p40 promoter for their study, the centrality of nucleosome remodeling
to broader developmental processes will surely stimulate a close examination
of other genes. One might initially think that their findings cast mammalian
TLR function in a role more similar to that in Drosophila. But on further
consideration, the work may actually mark the discovery of a TLR function
that is quite unique to mammals.
The prototypic Toll receptor of Drosophila was first identified
in connection with an early and essential developmental event4.
Toll is required for ventral polarization of the Drosophila embryo
and some mutations of Toll cause dorsalization of the embryo. A total of 12
Drosophila genes have been identified in the "dorsal group".
The other genes encode proteins that are required either to produce a functional
ligand for Toll or proteins that transduce the Toll signal to the level of
Dorsal, which—like mammalian NF-B—is a member of the Rel
superfamily of transcription factors. The relative simplicity of the Rel family
in Drosophila (only three representatives, Dorsal, Dif and Relish,
exist in flies) and the fact that antimicrobial peptide genes in Drosophila
bear promoter motifs that suggest Rel family member interactions, prompted
an investigation of the role played by Toll in Drosophila immunity.
Hence, it was discovered that Toll coordinates Drosophila antifungal
and antibacterial responses5,
6.
Early on, it was noted that the IL-1 receptor had cytoplasmic-domain homology
to Toll7. Later, a collection of mammalian orphan receptors
resembling Toll were identified in EST databases and one member of the family
(dubbed h-Toll, later called TLR4) was shown to be capable of activating NF-B
in mammalian cells8,
9. However, the function of these receptors
remained unknown until positional cloning work revealed that mouse Tlr4
encodes the transducing subunit of the mammalian endotoxin receptor10. Subsequently, knockout studies revealed that another member of
the family (TLR2) acts as a receptor for bacterial lipopeptides11.
A third member of the family (TLR9) transduces the signal initiated by unmethylated
DNA12. It seems that each TLR recognizes, at most, a small number
of molecules of microbial origin, which each act to ignite the innate immune
response.
Yet, to date, no developmental function has been ascribed to TLRs in mammals.
Even knock-out of the gene that encodes MyD88—which is believed to transduce
signals through all of the TLRs as well as the IL-1 and IL-18 receptors—has
no reported effect upon development. In addition, a human patient who was
shown to be coresistant to IL-1 and endotoxin (and who, by implication, probably
fails to respond to signals initiated by any of the TLRs) was reported to
be developmentally normal, though profoundly immunodeficient13.
It is certain that many of the Drosophila genes that are of key
importance to the ventralizing signal transduced by Toll were diverted to
other functions in mammals or expunged entirely from the mammalian genome.
Witness the upstream components of the Dorsal group. Gd, the first proteolytic
enzyme in the perivitelline cascade that leads to Toll activation, most closely
resembles mammalian prothrombin. Snake and Easter, the next components of
the cascade, vaguely resemble kallikrein and the complement component C1s.
Each of these mammalian proteins might arguably be involved in the inflammatory
response but not in development per se. And so far as is known, Spätzle,
the proximal ligand for Toll in both developmental and innate immunity, is
not represented in the mammalian genome at all. Rather, there is abundant
evidence that the mammalian TLRs have direct contact with microbial products
and require such contact to detect infection14,
15.
What, then, has become of the developmental functions of mammalian TLRs?
Ordinarily, developmental processes are among the most conserved in nature.
Yet Drosophila are absolutely dependent upon Toll and at least one
of its paralogs for development, whereas mammals are seemingly not dependent
upon any of the TLRs. Although the core of the Toll signaling pathway is conserved
between insects and mammals and although some of the target genes activated
in insects and mammals are functionally similar, others (those related to
early embryonic development) are apparently not. In addition, in Drosophila
, Toll is known to be required for the execution of a genetic program
that leads to the development of ventral structures but is not known to cause
activation or repression of genes through changes in chromatin structure.
Weinmann et al.3 find that the mammalian Toll homolog
TLR4 can generate a signal that is capable of affecting chromatin structure.
It achieves this without the need for an intact Rel gene. Hence, the known
Toll-related developmental pathway of Drosophila, which proceeds through
Dorsal, is not mimicked. Something quite different appears to be happening.
It is known that lipopolysaccharide (LPS), through TLR4, is capable of activating
the mitogen-activated protein kinase (MAPK) pathway, the phosphatidyl inositol-3
kinase (PI3K) pathway, the "stress-activated" (SAP or Jun) kinase
pathway and the p38 pathway. Through TLR4, LPS can also activate IL-1 receptor−associated
kinase (IRAK) and tumor necrosis factor receptor−associated factor (TRAF-6),
which are believed to lie upstream of IB phosphorylation. In addition,
knockout mutations of MyD88 do not entirely abolish either signaling to the
level of NF-B translocation or signaling to the level of p38 kinase.
Hence, there is ample room to contemplate events that are independent of NF-B
activation and its resultant induction of many genes within macrophages. Perhaps
one or more of the known kinase activation cascades lies upstream of the chromatin
modification that is observed; perhaps an entirely new pathway is responsible.
Returning to the question of what has become of the developmental function
of mammalian TLRs: might some Drosophila Toll paralogs have developmental
effects of which we remain ignorant? Might signals from these receptors effect
fundamental changes in chromatin structure? Certain evolutionary considerations
are important in this regard, as it must be appreciated that Drosophila
have seized upon one ancient line of TLR proteins for most of their needs,
whereas mammals have adopted another line for all of theirs16.
At least three basic lines of TIR domain−bearing protein seem to
have predated the divergence of Drosophila and mammals. Mammals have
retained only two of these lines, whereas all three are represented in
Drosophila. Hence, the "lost" developmental functions of the
mammalian TLRs may have been subserved by the "lost" line, to
which eight of the nine Drosophila TLRs belong (Fig.
1). Within the conserved TIR domain, Drosophila Toll-9 (cg5528)
is more similar to the mammalian TLRs than to any of its Drosophila
paralogs. It is therefore likely that this protein, rather than other
Drosophila Toll receptor, associates with Drosophila MyD88 (cg2078).
It might be expected that it elicits a response similar to that of the mammalian
TLRs. Yet to date, the phenotype of a Toll-9−deficient or a MyD88−deficient
Drosophila mutant remains unknown.
 | | |
The gene examined by Weinmann et al. (which encodes the IL-12 p40
subunit) did not exist 500 million years ago and, so far, is recognized only
in mammals. IL-12 is a cytokine that chiefly serves adaptive immunity and,
specifically, development of the T helper subset 1 response. Adaptive immunity
is known only in the subphylum Vertebrata. It is possible, then, that
the observations made pertain to a subset of genes that have post-developmental
functions, and one might imagine that the ability of TLR4 signals to de-repress
chromatin was acquired only recently. Although Weinmann et al. refer
to "TLR" signals, they have only studied signals that emanate
from TLR4. It remains to be seen whether other TLRs are similarly capable
of eliciting such a response.
The global importance of this study is hard to gauge, again because only
one gene was analyzed. C3H/HeJ mice certainly seem developmentally intact,
as do C57BL/10ScCr mice, although both strains fail to signal via TLR4.
Yet, the tonic influence of LPS and other microbial molecules upon immune
development is something that is poorly understood. Gnotobiotic animals are
well known to have remarkably hypoplastic lymphoid tissues. Though many interpretations
might be attached to this observation, it is possible that a number of genes
required for normal lymphoid development are dependent upon the direct interface
with the microbial world that is embodied by the TLRs and respond to such
contact in the most fundamental way: through changes in accessibility.
REFERENCES
-
Adams, M. D. et al. Science 287, 2185−2195 (2000). | Article | PubMed | ISI |
-
Rubin, G. M. et al. Science 287, 2204−2215 (2000). | Article | PubMed | ISI | ChemPort |
-
Weinman, A. S. et al. Nature Immunol. 2, 51−57 (2001). | Article | PubMed | ISI | ChemPort |
-
Belvin, M. P. & Anderson, K. V. Annu. Rev. Cell Dev. Biol. 12, 393−416 (1996). | Article | PubMed | ISI | ChemPort |
-
Lemaitre, B., Nicolas, E., Michaut, L., Reichhart, J. M. & Hoffmann, J. A. Cell 86, 973−983 (1996). | Article | PubMed | ISI | ChemPort |
-
Rutschmann, S. et al. Nature Immunol. 1 342−347 (2000). | Article | PubMed | ISI | ChemPort |
-
Gay, N. J. & Keith, F. J. Nature 351, 355−356 (1991). | Article | PubMed | ISI | ChemPort |
-
Medzhitov, R., Preston-Hurlburt, P. & Janeway, C. A. Jr Nature 388, 394−397 (1997). | Article | PubMed | ISI | ChemPort |
-
Rock, F. L., Hardiman, G., Timans, J. C., Kastelein, R. A. & Bazan, J. F. Proc. Natl Acad. Sci. USA 95, 588−593 (1998). | Article | PubMed | ChemPort |
-
Poltorak, A. et al. Science 282, 2085−2088 (1998). | Article | PubMed | ISI | ChemPort |
-
Takeuchi, O. et al. Immunity 11, 443−451 (1999). | Article | PubMed | ISI | ChemPort |
-
Hemmi, H. et al. Nature 408, 740−745 (2000). | Article | PubMed | ISI | ChemPort |
-
Kuhns, D. B., Long Priel, D. A. & Gallin, J. I. J. Immunol. 158, 3959−3964 (1997). | PubMed | ISI | ChemPort |
-
Poltorak, A., Ricciardi-Castagnoli, P., Citterio, A. & Beutler, B. Proc. Natl Acad. Sci. USA 97, 2163−2167 (2000). | Article | PubMed | ChemPort |
-
Lien, E. et al. J. Clin. Invest. 105, 497−504 (2000). | PubMed | ISI | ChemPort |
-
Du, X., Poltorak, A., Wei, Y. & Beutler, B. Eur. Cytok. Netw. 11, 362−371 (2000). | ISI | ChemPort |
|
