Autoimmunity and apoptosis: the Crohn's connection.
B Beutler
DOI: https://doi.org/10.1016/s1074-7613(01)00176-5
IF: 32.4
2001-01-01
Immunity
Abstract:A protean malady of unknown cause, Crohn's disease classically occurs as a relapsing/remitting “regional ileitis” but can affect any part of the gastrointestinal tract, not merely the terminal ileum. On rare occasions, lesions are observed even in the oropharynx or esophagus. Histologically, the lesions are intensely inflammatory, and they may lead to the formation of fistulae between enteric (and sometimes nonenteric) structures. Crohn's disease and ulcerative colitis are the two major forms of inflammatory bowel disease (IBD). Crohn's disease can be distinguished from ulcerative colitis on both clinical and histopathologic grounds. Ulcerative colitis is not associated with fistulae, it affects only the colon, producing characteristic shallow, ulcerative lesions, and it rather frequently presages the development of adenocarcinoma of the colon (whereas Crohn's disease does not). The two diseases can also be distinguished on the basis of their responses to therapy: TNF blockade usually yields an excellent response in Crohn's disease (Van Dullemen et al. 1995Van Dullemen H.M. Van Deventer S.J.H. Hommes D.W. Bijl H.A. Jansen J. Tytgat G.N.J. Woody J. Treatment of Crohn's disease with anti-tumor necrosis factor chimeric monoclonal antibody (cA2).Gastroenterology. 1995; 109: 129-135Abstract Full Text PDF PubMed Scopus (1034) Google Scholar) but rarely or never does so in ulcerative colitis. Crohn's disease has widely been considered an autoimmune disorder, although an infectious cause cannot be excluded. Recurrent hypotheses have held that microbial flora contribute initiators that trigger the disease, or, in some cases, prevent disease. Several knockout mutations produce IBD, and, in at least one case (the TCRα knockout mouse), germ-free animals fail to develop colitis, suggesting the pathogenic involvement of microbial flora (Bhan et al. 2000Bhan A.K. Mizoguchi E. Smith R.N. Mizoguchi A. Spontaneous chronic colitis in TCR alpha-mutant mice; an experimental model of human ulcerative colitis.Int. Rev. Immunol. 2000; 19: 123-138Crossref PubMed Scopus (34) Google Scholar). A caprine mycobacterial model of Crohn's disease has been presented as evidence of an infectious cause (Van Kruiningen et al. 1986Van Kruiningen H.J. Chiodini R.J. Thayer W.R. Coutu J.A. Merkal R.S. Runnels P.L. Experimental disease in infant goats induced by a Mycobacterium isolated from a patient with Crohn's disease. A preliminary report.Dig. Dis. Sci. 1986; 31: 1351-1360Crossref PubMed Scopus (86) Google Scholar). But in the authentic human disease, no microbial pathogen (and indeed, no commensal) has ever been shown to play an essential role. Perhaps the best evidence of involvement of microbes is the therapeutic effect of antibiotics in Crohn's disease: changing the microbial flora does appear to alter clinical course. But this observation is difficult to interpret. First, as already noted, the disease may occur in any part of the gastrointestinal tract, and bacterial flora differ widely from one end to the other. Moreover, a primary inflammatory process might permit the entry of microbes and microbial products into the submucosa or lamina propria, leading to an amplification loop in which inflammation is augmented and more mucosal damage occurs. Antibiotics might mitigate this process. At any rate, there is no concrete proof of a microbial “first cause.” Though mostly sporadic, Crohn's disease is sometimes familial, a fact used to advantage in the assignment of susceptibility loci. The strongest genetic association with Crohn's involves a chromosome 16 locus, designated IBD1 (Curran et al. 1998Curran M.E. Lau K.F. Hampe J. Schreiber S. Bridger S. Macpherson A.J. Cardon L.R. Sakul H. Harris T.J. Stokkers P. et al.Genetic analysis of inflammatory bowel disease in a large European cohort supports linkage to chromosomes 12 and 16.Gastroenterology. 1998; 115: 1066-1071Abstract Full Text Full Text PDF PubMed Scopus (153) Google Scholar). Loci on chromosomes 1p, 3q, 4q, and 12q have also been described (Cho et al. 1998Cho J.H. Nicolae D.L. Gold L.H. Fields C.T. LaBuda M.C. Rohal P.M. Pickles M.R. Qin L. Fu Y. Mann J.S. et al.Identification of novel susceptibility loci for inflammatory bowel disease on chromosomes 1p, 3q, and 4q evidence for epistasis between 1p and IBD1.Proc. Natl. Acad. Sci. USA. 1998; 95: 7502-7507Crossref PubMed Scopus (336) Google Scholar, Curran et al. 1998Curran M.E. Lau K.F. Hampe J. Schreiber S. Bridger S. Macpherson A.J. Cardon L.R. Sakul H. Harris T.J. Stokkers P. et al.Genetic analysis of inflammatory bowel disease in a large European cohort supports linkage to chromosomes 12 and 16.Gastroenterology. 1998; 115: 1066-1071Abstract Full Text Full Text PDF PubMed Scopus (153) Google Scholar). In a recent issue of Nature, Hugot et al. 2001Hugot J.P. Chamaillard M. Zouali H. Lesage S. Cezard J.P. Belaiche J. Almer S. Tysk C. O'Morain C.A. Gassull M. et al.Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn's disease.Nature. 2001; 411: 599-603Crossref PubMed Scopus (4480) Google Scholar and Ogura et al. 2001aOgura Y. Bonen D.K. Inohara N. Nicolae D.L. Chen F.F. Ramos R. Britton H. Moran T. Karaliuskas R. Duerr R.H. et al.A frameshift mutation in NOD2 associated with susceptibility to Crohn's disease.Nature. 2001; 411 (a): 603-606Crossref PubMed Scopus (3986) Google Scholar have reported that mutations of the gene encoding NOD2 are strongly associated—and almost certainly etiologically linked—to the development of Crohn's disease. While not all Crohn's disease can be explained by the NOD2 mutations found so far, the bulk of the chromosome 16 association with Crohn's disease is explained by the findings. In fact, pooling all variant alleles from the study of Hugot et al., it would appear that 201/468 patients and only 15/103 controls displayed coding changes, a figure that is overwhelmingly significant in itself and takes no account of deep intronic mutations, promoter mutations, large deletions, and the like. We shall come to the function of NOD2 presently. But first, let us consider precisely what was done. Hugot et al. used an unbiased genetic search based first on linkage disequilibrium mapping, then on pedigree disequilibrium testing, and finally on very extensive sequence analysis of the best candidate gene. This is to say that they built their case in a rigorous and painstaking manner. They ultimately identified a multitude of mutations (more than 30 in all), the great majority of which were observed only in Crohn's patients. This is a key point in the establishment of a causal relationship, since an isolated mutation might be dismissed as a nonetiologic marker in linkage disequilibrium with the authentic defect. Ogura et al. appear to have proceeded in a more intuitive fashion, declaring NOD2 as a candidate locus from the very outset. They identified three mutations in NOD2, one of which was the relatively common distally placed frameshift error also found by Hugot et al. The clear causality of the NOD2 mutations in familial Crohn's disease illustrates the resolving power of pure genetic methods, applied to a complex disease in the postgenomic era. Not only is the solidity of the core discovery beyond question, but it is doubtful that the discovery could have been made in the absence of genomic data—at least any time soon. A point not much emphasized in either study, but nonetheless true, is the inference that some cases of Crohn's disease are probably solely attributable to NOD2 mutations. It is not the case, as might earlier have been thought, that a chance association of a large collection of susceptibility alleles at different loci are required for disease to develop. Rather, at least in the case of the rare (private) mutations that were observed by Hugot et al., an overwhelming correlation is observed with disease, as much so as with a monogenic disorder of rather high penetrance. This speaks to the type of variants responsible for polygenic disease in general. For at least some diseases, isolated rare mutations at several unlinked (but perhaps functionally related) loci can have determinative importance. Inevitably, victory over such a difficult and important problem yields a sense of elation. The work of Hugot et al. and Ogura et al. have placed the notion that Crohn's disease has a discrete genetic etiology on firmer footing than ever before. Yet, where interpretation is concerned, the matter remains perplexing. In neither report do the authors draw mechanistic interpretations that lie beyond the realm of honest debate. As to the question of first cause, the debate has been sharpened rather than settled. NOD2 (and its closest homolog NOD1) are remarkable proteins in that they each contain N-terminal caspase recruitment domain (CARD) motifs, a single, more centrally placed nucleotide (ATP or GTP) binding motif, and still more distally placed leucine-rich repeats (LRRs; all of them of the ribonuclease inhibitor type) (Figure 1). More distant relatives bearing at least partially similar domain structure include CLAN (a.k.a. CARD12 or IPAF [Poyet et al. 2001Poyet, J.L., Srinivasula, S.M., Tnani, M., Razmara, M., Fernandes-Alnemri, T., and Alnemri, E.S. (2001). Identification of Ipaf, a human caspase-1 activating protein related to Apaf-1. J. Biol. Chem.Google Scholar, Geddes et al. 2001Geddes B.J. Wang L. Huang W.J. Lavellee M. Manji G.A. Brown M. Jurman M. Cao J. Morgenstern J. Merriam S. et al.Human card12 is a novel ced4/apaf-1 family member that induces apoptosis.Biochem. Biophys. Res. Commun. 2001; 284: 77-82Crossref PubMed Scopus (92) Google Scholar]) and Apaf-1. CARD motifs are functionally and structurally related to death domains (DDs) and are famously (though not universally) involved in pathways leading to programmed cell death. Their function is discussed in detail below. Nucleotide binding motifs exist in many proteins of diverse function. In the proteins under discussion here, they are believed to comprise a “self-oligomerization” motif. LRRs, which are also very common and exist in proteins of diverse function, are here regarded as “ligand binding” or “activation” motifs. LRRs are, in the minds of some, linked to the Toll-like receptor (TLR) family. Both Hugot et al. and Ogura et al. have emphasized the possibility of a functional connection between the LRRs in the NOD2 protein and those in the TLRs, which are widely regarded as the primary sensors of the innate immune system (Beutler 2000Beutler B. Tlr4 central component of the sole mammalian LPS sensor.Curr. Opin. Immunol. 2000; 12: 20-26Crossref PubMed Scopus (626) Google Scholar, Aderem and Ulevitch 2000Aderem A. Ulevitch R.J. Toll-like receptors in the induction of the innate immune response.Nature. 2000; 406: 782-787Crossref PubMed Scopus (2521) Google Scholar). It is here that a few errors of fact have been made, and corrections are in order. When Hugot et al. comment that “susceptibility to spontaneous inflammatory bowel disease (IBD) in mice has been associated with mutations in Toll-like receptor 4 (TLR4)—a member of a family of NF-κB activators that is known to bind LPS through its LRR domains [authors' references 19 and 20],” they are, in the main argument, incorrect. So far as anyone knows, the Tlr4 mutation in C3H/HeJ mice (corresponding to the P712H mutation) has nothing at all to do with the colitis observed by Sundberg et al. C3H/HeJ mice do not develop spontaneous colitis, nor do C57BL/10ScCr mice, nor any other mice with defects or deletions of the Tlr4 locus. Rather, the C3H/HeJ-Bir substrain appears to have acquired other mutations that cause colitis, unmapped (and essentially unstudied) to the present day (Sundberg et al. 1994Sundberg J.P. Elson C.O. Bedigian H. Birkenmeier E.H. Spontaneous, heritable colitis in a new substrain of C3H/HeJ mice.Gastroenterology. 1994; 107: 1726-1735PubMed Google Scholar). No dependence upon a TLR4 mutation has been established. Hugot et al. are correct in their observation that TLR4 binds LPS in order to trigger cell activation (Poltorak et al. 2000Poltorak A. Ricciardi-Castagnoli P. Citterio A. Beutler B. Physical contact between LPS and Tlr4 revealed by genetic complementation.Proc. Natl. Acad. Sci. USA. 2000; 97: 2163-2167Crossref PubMed Scopus (387) Google Scholar, da Silva et al. 2001da Silva C.J. Soldau K. Christen U. Tobias P.S. Ulevitch R.J. Lipopolysaccharide is in close proximity to each of the proteins in its membrane receptor complex transfer from CD14 to TLR4 and MD-2.J. Biol. Chem. 2001; 276: 21129-21135Crossref PubMed Scopus (531) Google Scholar). However, the role of the LRR domains in binding is not yet established. It is entirely conceivable that LPS never has physical contact with the LRRs at all. Likewise, the comment of Ogura et al. that “genetic variation in the LRRs of TLR4 accounts for inter-individual differences in bronchial responsiveness to aerosolized LPS” is unfounded. The truth is that the mutations cited—whatever their phenotypic effect—do not reside within the LRRs but between them (at residues 299 and 399 of the polypeptide chain). Ogura and colleagues actually make some fairly hazardous turns, both in their experiments and in their conclusions. Transfecting HEK293 cells with NOD2 expression plasmids and then exposing the cells to exceedingly high concentrations of LPS (10 μg/ml), these authors maintain that NOD2 creates sensitivity to LPS and offer that it does so by acting as an “intracellular receptor” for LPS. In the case of NOD1, the authors have earlier pointed to other evidence in support of an intracellular LPS receptor (Inohara et al. 2001Inohara N. Ogura Y. Chen F.F. Muto A. Nunez G. Human nod1 confers responsiveness to bacterial lipopolysaccharides.J. Biol. Chem. 2001; 276: 2551-2554Crossref PubMed Scopus (436) Google Scholar) and have explained that some cells, such as those of the intestinal mucosa, would be ill-served by responding to LPS in the external milieu. True enough. But extremely strong evidence would be required to sustain the hypothesis of a cytoplasmic sensor for LPS, and, so far, only a few publications may be brought to bear on the matter. For several decades, LPS has been taken as the quintessential microbial activator of innate immune responses. Genetic studies of LPS sensing led directly to the present concept that TLRs are the principal sensors of the innate immune system (Poltorak et al. 1998aPoltorak A. He X. Smirnova I. Liu M.-Y. Van Huffel C. Du X. Birdwell D. Alejos E. Silva M. Galanos C. et al.Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice mutations in Tlr4 gene.Science. 1998; 282 (a): 2085-2088Crossref PubMed Scopus (6157) Google Scholar, Poltorak et al. 1998bPoltorak A. Smirnova I. He X.L. Liu M.Y. Van Huffel C. McNally O. Birdwell D. Alejos E. Silva M. Du X. et al.Genetic and physical mapping of the Lps locus- identification of the toll-4 receptor as a candidate gene in the critical region.Blood Cells Mol. Dis. 1998; 24 (b): 340-355Crossref PubMed Scopus (286) Google Scholar). What is true for LPS may not be true for all microbial molecules. Some may have intracellular sensors and others may not. But for the present, since it has been proposed that the NOD1 and NOD2 molecules may act as intracellular LPS sensors, we will confine our discussion to LPS. A substantial body of evidence now holds that there is only one conduit for LPS signal transduction in mammalian cells, at least when LPS is presented from a source external to the cell. This evidence does not preclude the existence of multiple LPS concentrating mechanisms, nor does it cast doubt on the highly ramified character of the LPS signaling pathway. It may also be the case that LPS is engulfed within a phagosome prior to activation of the primary sensor. However, leaving aside whatever happens above the membrane and whatever happens beneath, it has been understood for over 30 years that in the absence of a single, critical protein (formerly known as the product of the Lps locus and now identified as TLR4 [Poltorak et al. 1998aPoltorak A. He X. Smirnova I. Liu M.-Y. Van Huffel C. Du X. Birdwell D. Alejos E. Silva M. Galanos C. et al.Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice mutations in Tlr4 gene.Science. 1998; 282 (a): 2085-2088Crossref PubMed Scopus (6157) Google Scholar]), nothing much happens when LPS—externally presented—comes in contact with mammalian cells (Heppner and Weiss 1965Heppner G. Weiss D.W. High susceptibility of strain A mice to endotoxin and endotoxin-red blood cell mixtures.J. Bacteriol. 1965; 90: 696-703PubMed Google Scholar, Sultzer 1968Sultzer B.M. Genetic control of leucocyte responses to endotoxin.Nature. 1968; 219: 1253-1254Crossref PubMed Scopus (253) Google Scholar, Coutinho and Meo 1978Coutinho A. Meo T. Genetic basis for unresponsiveness to lipopolysaccharide in C57BL/10Cr mice.Immunogenetics. 1978; 7: 17-24Crossref PubMed Scopus (94) Google Scholar). Hence, so far as is known, cells derived from C3H/HeJ and C57BL/10ScCr mice are indifferent to LPS present in the surrounding the medium. LPS is known to be internalized by cells if presented to them in aggregated form (Weel et al. 1989Weel J.F. Hopman C.T. van Putten J.P. Stable expression of lipooligosaccharide antigens during attachment, internalization, and intracellular processing of Neisseria gonorrhoeae in infected epithelial cells.Infect. Immun. 1989; 57: 3395-3402PubMed Google Scholar, Warner et al. 1985Warner S.J. Savage N. Mitchell D. Characteristics of lipopolysaccharide interaction with human peripheral- blood monocytes.Biochem. J. 1985; 232: 379-383PubMed Google Scholar, Kriegsmann et al. 1993Kriegsmann J. Gay S. Brauer R. Endocytosis of lipopolysaccharide in mouse macrophages.Cell Mol. Biol. 1993; 39: 791-800PubMed Google Scholar), and at least some of it reaches the cytosol (Diaz-Laviada et al. 1991Diaz-Laviada I. Ainaga J. Portoles M.T. Carrascosa J.L. Muncio A.M. Pagani R. Binding studies and localization of Escherichia coli lipopolysaccharide in cultured hepatocytes by an immunocolloidal-gold technique.Histochem. J. 1991; 23: 221-228Crossref PubMed Scopus (16) Google Scholar, Municio et al. 1990Municio A.M. Abarca S. Carrascosa J.L. Garcia R. Diaz-Laviada I. Ainaga M.J. Portoles M.T. Pagani R. Risco C. Bosch M.A. Immunocytochemical localization of bacterial lipopolysaccharide with colloidal-gold probes in different target cells.Adv. Exp. Med. Biol. 1990; 256: 199-202Crossref PubMed Scopus (8) Google Scholar). Nonetheless, TLR4 mutations forbid responses to LPS. This would suggest that any “internal receptor” for LPS would require a very exclusive route of presentation. Sansonetti and his colleagues (Philpott et al. 2000Philpott D.J. Yamaoka S. Israel A. Sansonetti P.J. Invasive Shigella flexneri activates NF-kappa B through a lipopolysaccharide-dependent innate intracellular response and leads to IL-8 expression in epithelial cells.J. Immunol. 2000; 165: 903-914PubMed Google Scholar) have used microinjection to introduce LPS directly into the cytosol of epithelial cells and have demonstrated the occurrence of NF-κB activation under these conditions. As their reasoning goes, many epithelial cells that fail to respond to LPS (like those in the gut) nonetheless require a means of detecting intracellular bacteria should they gain entry to the cytosol. Although their studies were performed using HeLa cells, which might well express an TLR4, the cells nonetheless fail to respond to LPS presented externally by activating NF-κB, and, therefore, some form of detection must occur when microinjection of LPS is performed. The data are interesting but leave certain questions unanswered. It is difficult to say just what class of molecules is sensed. Is the detection unique to LPS, or are a wide variety of microbial molecules capable of triggering a response through the same pathway (i.e., is LPS one of many molecules that fulfill the necessary structural rules)? Does access of Gram-negative bacteria to the cytosol occur in a viable cell, and if so, does it assure the presence of free LPS in the cytosol? Moreover, the identity of the receptor is not addressed by these studies. Ogura et al. now propose that NOD1 and NOD2 may represent sensing molecules for LPS—perhaps those detected by Philpott et al. A certain lapse of logic is evident in that NOD2 expression is purportedly macrophage specific, and macrophages (unlike epithelial cells) are certainly capable of sensing LPS in their external environment. They are not (like epithelial cells) compelled to live in contact with enormous concentrations of LPS. But leaving such arguments aside, the case for assigning a sensing function to NOD1 and/or NOD2 remains weak. Unfortunately, the authors have chosen an inadequate means to test the LPS sensing function of the candidates. The approach is essentially the same as that used by their forebears in the TLR field, who once regarded TLR2 as the LPS receptor (Yang et al. 1998Yang R.-B. Mark M.R. Gray A. Huang A. Xie M.H. Zhang M. Goddard A. Wood W.I. Gurney A.L. Godowski P.J. Toll-like receptor-2 mediates lipopolysaccharide-induced cellular signalling.Nature. 1998; 395: 284-288Crossref PubMed Scopus (1072) Google Scholar, Kirschning et al. 1998Kirschning C.J. Wesche H. Merrill A.T. Rothe M. Human toll-like receptor 2 confers responsiveness to bacterial lipopolysaccharide.J. Exp. Med. 1998; 188: 2091-2097Crossref PubMed Scopus (643) Google Scholar), an idea now all but universally rejected. The notion of TLR2-based LPS signaling was derived from the response that was elicited when HEK293 cells were transfected to express TLR2 at high copy number and then exposed to high concentrations of LPS. Never consistent with the view that there must be a single sensing pathway for LPS—at least when it is presented from the outside of the cell—the phenomenon was ultimately ascribed to contaminants present in commercial LPS preparations (Hirschfeld et al. 2000Hirschfeld M. Ma Y. Weis J.H. Vogel S.N. Weis J.J. Cutting edge repurification of lipopolysaccharide eliminates signaling through both human and murine toll-like receptor 2.J. Immunol. 2000; 165: 618-622PubMed Google Scholar). There is no reason to think otherwise in the present case. Just as LPS was shown to bind to NOD1 (Inohara et al. 2001Inohara N. Ogura Y. Chen F.F. Muto A. Nunez G. Human nod1 confers responsiveness to bacterial lipopolysaccharides.J. Biol. Chem. 2001; 276: 2551-2554Crossref PubMed Scopus (436) Google Scholar), Yang and colleagues showed that LPS could bind to TLR2. But in neither case was binding shown to be of high affinity or specificity. The overall lesson is clear: when analyzing the LPS response pathway, hard genetic data (i.e., natural or targeted mutations) should be coupled with the use of a highly purified (or better still, chemically synthesized) agonist. So far, there is no “strong” evidence to support a direct role for NOD1 or NOD2 in LPS sensing, notwithstanding the transfection-based studies that the authors have presented. Strong evidence of the thesis that NOD proteins act as cytoplasmic receptors would entail (1) a well-controlled demonstration that chemically synthesized LPS is sensed when injected into the cytoplasm of cells from mice lacking TLR4 and (2) demonstration that mice lacking NOD proteins cannot respond to the same stimulus. Of course, one cannot ignore the possibility that the NOD proteins contribute to LPS signaling at some point downstream from the receptor (and there is no fundamental argument against such a possibility). But so far, no compelling data have been presented in support of this scenario either. Another question of interpretation is immediately raised. If the NOD proteins do not initiate signals in response to stimulation by an external agent (i.e., LPS), they most likely act as signaling intermediates. Inohara et al. 1999Inohara N. Koseki T. Del Peso L. Hu Y. Yee C. Chen S. Carrio R. Merino J. Liu D. Ni J. Nunez G. Nod1, an Apaf-1-like activator of caspase-9 and nuclear factor-kappaB.J. Biol. Chem. 1999; 274: 14560-14567Crossref PubMed Scopus (598) Google Scholar and Ogura et al. 2001aOgura Y. Bonen D.K. Inohara N. Nicolae D.L. Chen F.F. Ramos R. Britton H. Moran T. Karaliuskas R. Duerr R.H. et al.A frameshift mutation in NOD2 associated with susceptibility to Crohn's disease.Nature. 2001; 411 (a): 603-606Crossref PubMed Scopus (3986) Google Scholar, Ogura et al. 2001bOgura Y. Inohara N. Benito A. Chen F.F. Yamaoka S. Nunez G. Nod2, a Nod1/Apaf-1 family member that is restricted to monocytes and activates NF-kappaB.J. Biol. Chem. 2001; 276 (b): 4812-4818Crossref PubMed Scopus (1119) Google Scholar have shown that NOD1 and NOD2 can activate NF-κB via RIP2, which establishes a physical connection between the two molecules. However, the “direction” of signaling is less clear. The fact that RIP2 is able to initiate NF-κB translocation when engaged by NOD2 says nothing of the origin of the signal. NOD2 might be activated to engage RIP2, or RIP2 might be activated to engage NOD2. The issue is discussed further below, but, to begin with, some phylogenetic perspective might be of use, since the relative position to which the NOD proteins have been assigned in the signaling pathway has been based partly upon inferred relationship to other proteins, and the credibility of these inferences is therefore an important issue. At least in part, the hypothesis that LPS directly binds to NOD2 seems to have grown from the fact that NOD2 has LRR motifs. In this characteristic, it shares a structural motif with TLR4, which indisputably does bind LPS (Poltorak et al. 2000Poltorak A. Ricciardi-Castagnoli P. Citterio A. Beutler B. Physical contact between LPS and Tlr4 revealed by genetic complementation.Proc. Natl. Acad. Sci. USA. 2000; 97: 2163-2167Crossref PubMed Scopus (387) Google Scholar, Lien et al. 2000Lien E. Means T.K. Heine H. Yoshimura A. Kusumoto S. Fukase K. Fenton M.J. Oikawa M. Qureshi N. Monks B. et al.Toll-like receptor 4 imparts ligand-specific recognition of bacterial lipopolysaccharide.J. Clin. Invest. 2000; 105: 497-504Crossref PubMed Scopus (670) Google Scholar, da Silva et al. 2001da Silva C.J. Soldau K. Christen U. Tobias P.S. Ulevitch R.J. Lipopolysaccharide is in close proximity to each of the proteins in its membrane receptor complex transfer from CD14 to TLR4 and MD-2.J. Biol. Chem. 2001; 276: 21129-21135Crossref PubMed Scopus (531) Google Scholar). However, it must be observed that the densely packed LRR array present in each of the NOD proteins differs markedly from the haphazard array of LRRs in the TLR ectodomains, which are often widely separated from one another by zones of indeterminate structure (Figure 1). The LRRs in TLR protein ectodomains (the “typical” type) are structurally distinguishable from those in defensive proteins of plants and from those in the NODs (the “Ribonuclease Inhibitor” type). In addition, the distribution of the LRRs within the NOD proteins is strongly suggestive of the LRR array in ribonuclease inhibitor (RI), the 3D structure of which has been solved at a high resolution (Kobe and Deisenhofer 1993Kobe B. Deisenhofer J. Crystal structure of porcine ribonuclease inhibitor, a protein with leucine-rich repeats.Nature. 1993; 366: 751-756Crossref PubMed Scopus (518) Google Scholar). Indeed, simple modeling exercises show that, like the RI array, the NOD1 arrays yield a “horseshoe” shape, albeit a partial one. A first guess as to the function of such a motif array would be that, like RI (which tightly associates with RNase A according to crystallographically defined interactions [Kobe and Deisenhofer 1995Kobe B. Deisenhofer J. A structural basis of the interactions between leucine-rich repeats and protein ligands.Nature. 1995; 374: 183-186Crossref PubMed Scopus (554) Google Scholar]), the NODs engage yet-unidentified protein ligands rather than LPS. Ogura and colleagues suggest that the NOD1 and NOD2 proteins are homologs of defensive proteins in plants that contain nucleotide binding domains and LRRs. The assertion is a tenuous one. The case for homology between plant and animal proteins that bear TIR motifs is quite secure because TIRs are rare motifs, very large compared with LRRs, and, across vast evolutionary distances, they exhibit a defensive function. But given the relative abundance and functional diversity of nucleotide binding domains and LRRs in mammalian proteins, a claim of common descent is much harder to support. Ogura et al. also point to the similarity between Apaf-1/Ced4/Dark and the two NOD proteins. Here, a case for common ancestry is even harder to support. Apaf-1 has no LRRs at all, only a CARD domain and WD40 motifs. The LRR and WD40 motifs are rarely found within the same protein, and they certainly did not arise from a common ancestral sequence during eukaryotic evolution, both motifs being found in diverse plant and animal representatives. However, a case for common function (that is, convergent evolution) can be made. The WD40 motifs and LRRs might both subserve activation. The former is believed to form a “propeller” assembly, while the latter forms a “horseshoe.” The WD40 motifs repress activation of the protein in the absence of cytochrome C. However, akin to receptors that function as molecular “switches,” cytochrome C modifies the conformation of the WD40 array and permits self-oligomerization to occur via the nucleotide binding domain. This permits transproteolytic activation of caspase-9 as a result of induced proximity. Caspase-9, which is attached to Apaf-1 through a CARD interaction, then causes more distal events in the apoptotic cascade (Li et al. 1997Li P. Nijhawan D. Budihardjo I. Srinivasula S.M. Ahmad M. Alnemri E.S. Wang X. Cytochrome c and dATP-dependent formation of Apaf-1/caspase-9 complex initiates an apoptotic protease cascade.Cell. 1997; 91: 479-489Abstract Full Tex