| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print June 22, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Medicine, Centre for Gastroenterology, Royal Free and University College Medical School, UCL, London, United Kingdom
1 Correspondence: Department of Medicine, Centre for Gastroenterology, Royal Free and University College Medical School, UCL, Rowland Hill Street, London, NW3 2PF, UK. E-mail: s.keshav{at}medsch.ucl.ac.uk
ABSTRACT
Paneth cells are secretory intestinal epithelial cells located at the base of the crypts of Lieberkühn in the small intestine. They display prominent cytoplasmic granules, containing antibacterial proteins such as lysozyme, secretory phospholipase A2 type IIA, and 
-defensins, which are released into the intestinal lumen in response to a range of stimuli. In this, they resemble circulating leukocytes, which also elaborate and secrete lysozyme and -defensins as part of an antibacterial defense function, and the resemblance is sustained at other levels. The cells also strongly and specifically express the NOD2 gene product, one of an emerging family of critical, intracellular mediators of innate immune responses, which is also highly expressed in peripheral blood mononuclear cells, and they express RNA for tumor necrosis factor , a major myelomonocytic cell-derived cytokine, which has a crucial role in the pathogenesis of diseases such as rheumatoid arthritis and Crohn’s disease (CD). Thus, these cells, which are derived from the pluripotent intestinal epithelial stem-cell lineage, are sessile, resident host-defense cells, which may share with leukocytes the beneficial function of secreting antimicrobial peptides, as well as the potentially harmful capacity for promoting inflammation and tissue damage. Paneth cells are most abundant in the distal small intestine, which is the region most frequently affected by CD, and there is great interest in the potential role of these cells in this condition. This brief review summarizes current knowledge and speculates on how the study of these fascinating cells might be advanced.
Key Words: Crohn’s disease • TNF • defensin • lysozyme • secretory phospholipase A2 • NOD2
INTRODUCTION
In cross-sections of the epithelium of the small intestine, approximately five pyramidal-shaped cells with basally situated nuclei and prominent, large, apical granules are seen at the base of the crypts of Lieberkühn (Fig. 1 ). Their eponym derives from the detailed description by Dr. Josef Paneth of Vienna, published in 1888 [1 ]. The characteristic granules are intensely stained by eosin in routine histopathology sections and can be demonstrated more specifically by specialized stains such as phloxine-tartrazine as well as by immunohistochemistry for their protein products. The first specific product to be localized to Paneth cells was the antibacterial enzyme lysozyme, discovered by Alexander Fleming, and known to be present in circulating leukocytes as well as in a number of exocrine secretions [2 ]. Experimental studies established that the number of Paneth cells in the small intestine was increased as a consequence of luminal bacterial overgrowth, and examination of pathological tissues demonstrated that Paneth cells, which are ordinarily confined to the small intestine and the most proximal parts of the large intestine, appeared throughout the inflamed large intestine [3 , 4 ]. Such cells, appearing in a location from which they are normally absent, are termed metaplastic, and Paneth cell metaplasia is a recognized, histopathological feature of inflammatory bowel disease (IBD) [5 ]. These observations suggest that Paneth cells mediate antibacterial host defense in the intestine and that they may have a role in the pathogenesis of intestinal inflammation. Both possibilities have been supported lately by further evidence, which will be reviewed here. It is also interesting that Paneth cells occupy a position in the crypt epithelium immediately adjacent to the pluripotent intestinal stem cells, and this has prompted speculation that they may have a role in maintaining special properties of the stem cell niche and hence, the overall health of the intestinal crypt [6 ].
|
The study of Paneth cells has been hampered by our inability to isolate or culture these cells and by the fact that existing intestinal epithelial cell lines, which are derived mainly from colorectal cancers, do not exhibit differentiated functions such as abundant expression of -defensins or regulated secretion of antibacterial proteins. However, our view of the potential function of Paneth cells is informed by the increasing number of specific Paneth, cell-restricted protein products, in addition to lysozyme, which has been identified. Some key products are summarized in Table 1
. The major products of Paneth cells are antibacterial proteins, of which the -defensins are entirely specific. Distinct, structurally related -defensin genes are expressed in myelomonocytic cells. Lysozyme is more widely distributed, although in mice, there are two separate genes, one expressed predominantly in myelomonocytic cells (lysozyme M) and the other in Paneth cells (lysozyme P) [7
]. Paneth cells are probably the most abundant source of sPLA2 in the body, although other cells may also synthesize the protein. More comprehensive lists of genes expressed in Paneth cells are available in recently published reviews [26
, 27
].
|
-DEFENSINS (CRYPTDINS)
In 1989, Ouellette and colleagues [8
] identified the presence of a large family, eventually found to contain at least 16 members, of cysteine-rich peptides expressed in Paneth cells. These peptides, also termed cryptdins, are -defensins, sharing strong structural homology and in vitro and in vivo antibacterial activity with -defensins, previously identified in myelomonocytic cells [28
]. The gene structure of Paneth cell -defensins, comprising two exons, separated by a single intron, is distinct to those expressed in myelomonocytic cells, which have an additional 5' exon [9
]. The identification of -defensins, which have powerful antibacterial properties and may also have other profound cellular effects, has been important for investigating the function of Paneth cells. In mice, maturation of Paneth cell -defensins from precursor prepropeptides to their active form depends on proteolytic cleavage by matrix metalloproteinase 7 (MMP7; matrilysin), which is also expressed specifically in Paneth cells [29
]. Genetic deletion of MMP7 in transgenic mice completely abolishes the production of mature -defensins, and the demonstration that such mice have a measurable defect in intestinal antibacterial defense was an important milestone in establishing a role for Paneth cells in host defense [30
].
Another landmark study, reported by Ayabe and colleagues in 2000 [31
], demonstrated that whole intestinal crypts, which contain Paneth cells, could be induced to secrete -defensins in vitro and that the resulting secretions killed bacteria at concentrations that were readily attained in these experiments. The concentration of -defensins would be much higher in the narrow crypt lumen, supporting the view that these abundant peptides could easily maintain sterility in this microenvironment. In the absence of an immortalized cell line or isolated primary cells, the technique established by Ayabe and colleagues [31
] also pioneered a way in which the regulation of Paneth cell secretion might be studied in vitro in other species, and such experiments using human tissue are under way in our laboratory.
More recently, transgenic expression of a human Paneth cell-specific -defensin [human defensin 5 (HD5)] in mice conferred enhanced resistance to enteric infection with a virulent strain of Salmonella typhimurium [32
]. Thus, by deleting, overexpressing, and measuring their secretion, experimental studies in mice have demonstrated a nonredundant and potentially important antibacterial function for Paneth cell-derived -defensins. In humans, there are only two Paneth cell -defensins, HD5 and HD6, and their maturation depends on an isoform of trypsin rather than on MMP7, and so far, rigorous tests of their physiological function are lacking [33
]. Nonetheless, by analogy with mice, it is expected that they will have a role in enteric antibacterial host defense.
The strong and highly restricted expression of cryptdins in Paneth cells has allowed their promoters to be used in transgenic targeting constructs. This has enabled targeted expression of an attenuated form of the diphtheria toxin and of the viral simian virus 40 T antigen in Paneth cell precursors in transgenic mice [34 ]. These experiments were designed to test the effect of deleting or altering the differentiation of the cell lineage; however, the transgenic mice developed normally and ostensibly exhibited no defect in host defense or in the morphology and function of the intestinal crypt-villus unit. Thus, it seems that at least in the absence of a specific infectious challenge, the loss of Paneth cell antibacterial function can be compensated for by other host defense mechanisms. It will be interesting to see how mice lacking Paneth cells respond to enteric infection with bacteria, which are known to be susceptible to the activity of cryptdins. Similarly, other putative functions of Paneth cells are also unaffected by cell-lineage ablation, suggesting the presence of compensatory mechanisms or the need for more refined analysis and experimental manipulation in mice lacking these cells. For instance, a subtle defect in angiogenic maturation in the intestine has been demonstrated recently in mice with diphtheria toxin-mediated depletion of Paneth cells, which is not apparent in germ-free counterparts lacking the endogenous enteric bacterial flora [35 ].
Although most attention is focused on the antibacterial role of -defensins, there is evidence to suggest that the molecules could support other functions. Related ß-defensins can induce chemotaxis of dendritic cells (DC) in vitro, and murine cryptdin 4 is able to alter the secretion of chloride ions from intestinal epithelial cells in vitro, as well as to induce the expression of proinflammatory genes [36
37
38
]. Thus, the further detailed investigation of -defensin function in Paneth cells may yet provide important new insights into the function of these cells.
sPLA2
Another important antibacterial gene product, which is expressed specifically in Paneth cells is the type IIA sPLA2, which is structurally related to bee and snake venom enzymes rather than to intracellular signaling enzymes found in most mammalian cells [11
, 39
]. The Paneth cell enzyme (sPLA2) is a prominent secretory product, which is also found in the circulation, where it might originate from hepatocytes, synthesized as part of the acute-phase response [13
]. This is compatible with the observation that the circulating concentration of sPLA2 is regulated by systemic inflammation [40
]. Furthermore, in addition to its role as an antibacterial protein, sPLA2 has proinflammatory and proatherogenic effects [41
, 42
]. Curiously, expression of the sPLA2 gene, which is naturally deficient in C57/Bl6 mice, suppresses the development of intestinal adenomas in the APCmin/+ mouse model, which mimics the human disease familial adenomatous polyposis coli [43
]. Thus, sPLA2 is a multifunctional protein product of Paneth cells, which has antibacterial, proinflammatory, and possibly anti-oncogenic properties. Unlike lysozyme and the -defensins, however, it is unclear if circulating leukocytes synthesize or secrete sPLA2, although protein or RNA has been detected in hepatocytes, chondrocytes, and megakaryocytes [12
13
14
15
].
TNF-
An intriguing observation, which we first reported in mice, is that Paneth cells constitutively express the gene for TNF- [16
]. At that time, although specific expression of the TNF- RNA could readily be shown in Paneth cells and not in other epithelial cells or in lamina propria leukocytes, it was not possible to demonstrate expression of the protein. A potential explanation of this dissociation between RNA and protein expression is suggested by the well-known, post-transcriptional regulation of mRNA stability and translation, which is exerted by an adenosine-uridine (AU)-rich element in the 3'-untranslated region of the TNF- gene [44
]. Subsequent to our discovery of the expression of TNF- in murine Paneth cells, a number of reports have confirmed this in other species including in humans in the disease necrotizing enterocolitis, which affects neonates [17
, 18
]. We have now also demonstrated specific expression of TNF- RNA in adult human intestine, where it appears to be induced rather than constitutive [19
]. The function of TNF- derived from Paneth cells is unknown, although a recent report suggests that expression of TNF- is induced when crypts and Paneth cells are damaged and that the cytokine plays a role in the reconstitution of the crypt cell population [45
].
The prominent, apical localization of secretory granules in Paneth cells suggests that their protein products are released into the lumen; however, it is possible that products such as TNF- are released basolaterally by an alternative secretory pathway, and this view is more plausible in light of the important cell-to-cell signaling role performed by TNF-, which is, of course, an important cytokine in innate and acquired immunity and in inflammation, and the localization of this gene product to Paneth cells strongly suggests that they participate in these processes, which are central to conditions such as CD, one of the two main forms of IBD, which affects 
150 per 100,000 of the population in the United States and United Kingdom [46
]. Experimental proof of this supposition has yet to be provided, although it is relevant to note that the therapeutic administration of antibodies to TNF-, such as Infliximab, is powerfully effective in treating CD, and the mode of action of these antibodies remains poorly defined [47
].
Furthermore, deletion (
) of the AU-rich element (ARE) of the TNF- gene in transgenic mice results in the development of severe inflammatory lesions in the small intestine and not other parts of the intestine of the TNFARE mouse line [48
]. The anatomical localization and microscopic pathology of these lesions resemble CD in human subjects to a remarkable extent. TNFARE mice do not exhibit a generalized or widespread inflammatory tendency; in addition to the CD-like lesions of the small intestine, they also develop a severe, deforming polyarthritis, which resembles rheumatoid arthritis, and other major organs, such as the lungs, liver, and skin among others, are unaffected. The potential relevance of this mouse model to human disease is highlighted further by the clinical observation that anti-TNF antibodies are most widely used to treat rheumatoid arthritis and CD.
NOD2
Interest in the potential role of Paneth cells in the pathogenesis of IBD has been boosted recently by the observation that these cells strongly and specifically express the NOD2 gene (Fig. 2
) [19
, 22
]. NOD2 is a recently described member of the NOD family of intracellular signaling molecules, which are central to the control of inflammation [49
]. Related proteins include pyrin and cryopyin and other components of the putative inflammasome, and inherited mutations in this large family of proteins are associated with many different inflammatory conditions such as Muckle-Wells syndrome and familial Mediterranean fever. In a landmark series of papers published in 2001 [50
51
52
], mutations in the NOD2 gene were shown to be associated specifically with CD. Subsequently, it has become clear that although CD can affect different parts of the intestine, mutations in the NOD2 gene are only associated with an increased risk of CD affecting the small intestine and in particular, the terminal ileum [53
]. Ileal and ileo-colonic CD accounts for 80% of all cases of CD, and heterozygous carriage of mutations in the NOD2 gene increases risk approximately threefold and homozygous carriage, 40 fold compared with wild-type.
|
The cellular function of the NOD2 gene product is unknown, although it is now apparent that cellular responses to muramyl dipeptide (MDP), a component of bacterial peptidoglycan, are reduced markedly by mutations in the NOD2 gene [54 ]. Thus, NOD2 may be regarded as an intracellular counterpart of the larger group of pathogen-associated molecular pattern receptors, such as the Toll-like receptors. This view is supported further by the demonstration that the related NOD1 protein also mediates cellular responses to a constituent of bacterial peptidoglycan—MDP with a mesodiamiopimelic (meso-DAP) residue attached in place of leucine. Leucine-MDP and meso-DAP MDP are found, respectively, in gram-positive and gram-negative bacteria, suggesting that various NOD proteins may allow cells to respond to and distinguish between different bacterial determinants [55 ].
As the NOD2 protein is expressed highly and specifically in Paneth cells, it is natural to ask what consequences flow from this and what the effect of carriage of a mutant protein is on cellular function. Both questions still await definitive answers. However, recent observations suggest that carriage of mutant alleles of the NOD2 gene can affect the function of Paneth cells and that this may have an impact on CD. For example, in a study of -defensin gene expression in CD-derived intestinal tissue, Wehkamp and colleagues [56
] demonstrated reduced expression of the HD5 and HD6 genes in inflamed tissue compared with normal tissue from patients with CD. This reduced expression of HD5 and HD6 was more profound in individuals who also carried mutations in the NOD2 gene. More recently, transgenic mice lacking NOD2 protein have been shown to have profoundly reduced expression of specific -defensin genes and to exhibit a defect in the clearance of Listeria monocytogenes administered via the oral route [57
]. It is interesting that clearance of bacteria administered intraperitoneally or intravenously was unaffected by the absence of the NOD2 protein, suggesting a specific intestinal effect, which is compatible with a role mediated by Paneth cells rather than ubiquitously distributed circulating cells such as monocytes. The mice also have defective responses to MDP in terms of the activation of monocytes and antibody production after immunization using MDP as an adjuvant. However, the mice do not develop overt intestinal inflammation, which limits their use for determining how the NOD2 gene actually relates to the pathogenesis of CD.
HOW DO MUTATIONS IN THE NOD2 GENE PREDISPOSE TO CROHN’S DISEASE?
Our increasing knowledge about the gene expression profile of Paneth cells has provided useful clues to understanding the role of these cells in antibacterial host defense and now, potentially, in intestinal immunity and inflammation. A unifying hypothesis that could account for the current data regarding Paneth cells, -defensins, the NOD2 gene, and CD is that inherited mutations in the NOD2 gene alter the expression of -defensins, reduce antibacterial defenses in the ileum, and allow bacteria and their products to accumulate and thus, initiate and propagate an uncontrolled, inflammatory reaction. Evidence supporting the hypothesis that excessive bacterial colonization of the ileum is linked causally to CD includes the identification of particular strains of adhesive-invasive Escherichia coli, which are peculiar to CD-affected ileum [58
]. However, the case that reduced, local antibacterial defenses are an important or critical element in the development of CD remains to be proved. Counter-arguments, which need to be addressed, include the lack of inflammation in mice that lack Paneth cells altogether or lack all -defensins as a consequence of the deletion of the MMP7 gene. Furthermore, in clinical practice, antibiotics, including powerful, broad-spectrum regimens administered for many months, have limited efficacy, and immunosuppressive treatments such as the use of corticosteroids, thiopurines, and anti-TNF- antibodies usually have a rapid and dramatically beneficial effect on disease activity [59
, 60
]. Thus, at least in established, active disease, excessive inflammation, rather than reduced antibacterial defense, appears to drive the pathological process, and an important, outstanding question for investigators in this field is to determine how Paneth cells might respond to and influence the inflammatory response. Dysregulated production of TNF- protein, which is of central importance in CD, could be a critical factor. Other possibilities include possible proinflammatory activities of the -defensins themselves, some of which have been shown to alter the function of epithelial cells. Related molecules, the ß-defensins, have also been implicated in immunoregulatory functions, such as the chemotactic recruitment of DC, and the potential for pleotropic effects of these versatile and abundant peptides cannot be excluded. Another multifunctional protein product of Paneth cells, sPLA2, may also promote inflammation locally and systemically, and its role in CD has yet to be fully elucidated. Recently, clinical trials have reported encouraging results from treating CD with recombinant GM-CSF [21
], which, interestingly, has also been detected at the mRNA and protein level in Paneth cells [20
]. Thus, a number of different strands of evidence converge on the Paneth cell as a potentially key mediator in localized ileal inflammation.
PANETH CELL PROLIFERATION, DIFFERENTIATION, AND SECRETION
Although we still have an incomplete understanding of how proliferation, differentiation, and secretion of Paneth cells are controlled, recent reports provide important clues (Fig. 3 ). Activation of T lymphocytes in the lamina propria can induce proliferation of Paneth cells, although the exact mechanism is unknown [61 ]. Furthermore, when Paneth cells are damaged or destroyed, powerful mechanisms seem to ensure that they are replaced from undifferentiated precursor cells within a few hours [62 ]. It is possible, although unproven, that similar mechanisms result in the appearance of metaplastic Paneth cells in IBD.
|
-defensins [63
, 64
]. This important insight should prompt further investigation into the regulation of Paneth cell function by wnt proteins and their antagonists, potentially released by adjacent epithelial cells as well as cells in the lamina propria such as myofibroblasts, endothelial cells, lymphocytes, and macrophages [25
]. The secretory responses of Paneth cells have been studied in vivo using elaborate light- and electron-microscopical determination of changes in the appearance of the apical granules [65 , 66 ]. These experiments demonstrated that the cells secrete in response to muscarinic cholinergic agonists, and Ayabe et al. [31 ] confirmed this observation by using isolated, intact crypts in vitro. Furthermore, Ayabe et al. [31 ] demonstrated that murine Paneth cells also secrete granule contents at concentrations that are bactericidal in vitro, in response to microbial products including LPS, peptidoglycan, and MDP. Protozoal and fungal components, however, did not promote secretion. It will be interesting to know exactly how human Paneth cells respond to various microbial determinants in comparison with their murine counterparts.
The expression of TNF- in Paneth cells offers the prospect of determining if local production of this key cytokine, rather than systemic or leukocyte-derived production, is critical in the development of CD. Genetic deletion of the AU-rich element in mouse TNF-, which leads to overproduction of the protein in macrophages and fibroblasts, may similarly increase protein production in Paneth cells. This question could be settled by further investigation in the original TNFARE mouse line and by developing Paneth cell-restricted expression of the TNFARE gene, which we are currently pursuing.
The association of mutations in the NOD2 gene and the development of CD are major advances in understanding the pathogenesis of this disease, and it is likely that further experimental investigation in mice lacking the NOD2 gene will ultimately yield insight into how mutations relate to disease. As mice lacking NOD2 do not seem to develop CD spontaneously, crossing of such mice against a background with increased susceptibility to CD, such as an attenuated form of the TNFARE mouse, or the Samp1/Yit strain, which develops ileal inflammation spontaneously, may be informative [67
].
SUMMARY
Paneth cells comprise a significant proportion of the cell mass in the crypts of the small intestine, and there is compelling evidence to suggest that they have an important antibacterial role. They are also implicated in the pathogenesis of an important idiopathic, inflammatory disease of the intestine, and therefore, like leukocytes, they may display a dual aspect—a protective, antibacterial function and a potentially harmful, proinflammatory effect (Fig. 4 ). The emergence of new molecular tools and recent insight into the regulation of secretion, differentiation, and proliferation of these cells may allow this hypothesis to be tested experimentally. Detailed investigation of the function of Paneth cells is justified in part, as it could lead to highly targeted therapies for diseases such as CD, as well as for the more fundamental reason that these fascinating cells have remained enigmatic for too long.
|
Received October 3, 2005; revised January 7, 2006; accepted March 13, 2006.
REFERENCES
Histochem. J. 26,142-151[CrossRef][Medline] transcripts in normal bowel and in necrotizing enterocolitis. TNF gene expression by Paneth cells, intestinal eosinophils, and macrophages Am. J. Pathol. 142,1858-1865[Abstract]
B J. Biol. Chem. 276,4812-4818-defensin activation by the metalloproteinase matrilysin in innate host defense Science 286,113-117-defensins by intestinal Paneth cells in response to bacteria Nat. Immunol. 1,113-118[Medline] in intestinal epithelial cell proliferation following Paneth cell destruction Scand. J. Gastroenterol. 37,154-160[CrossRef][Medline] as a key cytokine in the inflammatory processes of Crohn’s disease—the mechanisms of action of infliximab Aliment. Pharmacol. Ther. 13(Suppl. 4),3-8-defensin expression Gut 53,1658-1664This article has been cited by other articles:
|
|
 |
|
  M.-I. Fernandez, B. Regnault, C. Mulet, M. Tanguy, P. Jay, P. J. Sansonetti, and T. Pedron Maturation of Paneth Cells Induces the Refractory State of Newborn Mice to Shigella Infection J. Immunol., April 1, 2008; 180(7): 4924 - 4930. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
