Cathelicidin Mediates Innate Intestinal Defense against Colonization with Epithelial Adherent Bacterial Pathogens
http://www.100md.com
免疫学杂志 2005年第8期
Abstract
Cathelicidin-related antimicrobial peptide (mCRAMP), the sole murine cathelicidin, is encoded by the gene Cnlp. We show that mCRAMP expression in the intestinal tract is largely restricted to surface epithelial cells in the colon. Synthetic mCRAMP had antimicrobial activity against the murine enteric pathogen Citrobacter rodentium, which like the related clinically important human pathogens enteropathogenic Escherichia coli and enterohemorrhagic E. coli, adheres to the apical membrane of intestinal epithelial cells. Colon epithelial cell extracts from Cnlp+/+ mice had significantly greater antimicrobial activity against C. rodentium than those of mutant Cnlp–/– mice that lack mCRAMP. Cnlp–/– mice developed significantly greater colon surface and crypt epithelial cell colonization, surface epithelial cell damage, and systemic dissemination of infection than Cnlp+/+ mice after oral infection with C. rodentium. Moreover, Cnlp+/+ mice were protected from oral infections with C. rodentium inocula that infected the majority of Cnlp–/– mice. These results establish cathelicidin as an important component of innate antimicrobial defense in the colon.
Introduction
Cathelicidins are evolutionarily conserved antimicrobial peptides present in all mammals. They are synthesized in precursor form with a conserved N-terminal signal sequence, a cathelin domain, and a C-terminal domain that is cleaved to release a peptide with antimicrobial activity (1, 2). Murine cathelicidin-related antimicrobial peptide (mCRAMP)4 is the murine ortholog of the sole and structurally related human cathelicidin, LL-37/hCAP18 (3). In contrast to other antimicrobial peptides in the human intestinal tract, such as -defensins that are restricted to small intestinal Paneth cells (4) or -defensins that are expressed by crypt, villous, and surface epithelial cells throughout the small and large intestine (5), the distribution of cathelicidin in the human intestinal tract is limited to surface epithelial cells in the colon and stomach (6, 7).
The murine enteric bacterial pathogen Citrobacter rodentium has important virulence features in common with human enteropathogenic Escherichia coli (EPEC) (8, 9, 10, 11, 12). Within 1 week of oral infection, C. rodentium colonize the murine colon and like EPEC and enterohemorrhagic E. coli (EHEC) in humans, C. rodentium adheres to intestinal epithelial cells where it produces the characteristic attaching and effacing lesion (10, 13, 14). Normally, few if any bacteria invade the mucosa and reach extraintestinal sites (15, 16). At later times, infection results in mucosal erosions, increased epithelial cell proliferation, and crypt hyperplasia (17, 18). Little is known about early innate mucosal defenses against this pathogen, although acquired immunity is needed to control and eliminate C. rodentium (10, 16, 19, 20, 21).
In light of the selective expression of cathelicidin in the human colon, we hypothesized that cathelicidin produced by colon surface epithelial cells may be important in host innate defense against pathogenic enteric bacteria for which lifestyle involves adherence to those cells. Using genetically mutant mice that lack cathelicidin and with C. rodentium as a model murine epithelial adherent pathogen, we have explored the functional importance of cathelicidin in early intestinal innate antimicrobial defense.
Materials and Methods
Mice
mCRAMP-deficient (Cnlp–/–) mice were generated as described (22). Cnlp+/+ littermates were used as controls. Cnlp–/– mice were healthy and fertile and did not manifest differences in intestinal mucosal architecture from Cnlp+/+ littermates used as controls. All studies were approved by the University of California at San Diego Animal Welfare Committee.
Bacteria
The following bacteria were used: C. rodentium (21); enteropathogenic E. coli serotype O111:NM, eae+ isolate DEC12f (CDC B170; American Type Culture Collection No. 43887; American Type Culture Collection, Manassas, VA); enterohemorrhagic E. coli O157:H7 strain 86-24 (gift of M. Donnenberg).
To determine total bacterial numbers in the colon and feces of Cnlp+/+ and Cnlp–/– mice, colon or feces were homogenized in TS broth. For aerobic culture, samples were plated onto tryptic soy blood agar containing 5% sheep blood and then incubated at 37°C overnight. For anaerobic culture, samples were plated onto Columbia blood agar with 5% sheep blood and then incubated at 37°C for 48 h in anaerobic culture system (Oxide Atmosphere Generation System for Anaerobic Bacteriology and AnaeroGen) after which CFU were determined. Gram staining was performed on 40 colonies from each sample.
Infections
Mice were infected orally with C. rodentium. Bacterial numbers in feces or homogenized colon were determined by plating on MacConkey agar (21). The detection limit of the CFU assay was 103 colonies per g of feces or per colon and <10 colonies per organ in spleen and mesenteric lymph nodes. Identity of representative bacterial colonies was verified by PCR (21).
Histological analysis and immunohistochemistry
Colons processed as Swiss rolls (21) were fixed (1% ZnSO4, 10% formalin) and embedded in paraffin. Sections (5 μm) were treated with 0.3% H2O2 in PBS for 10 min at room temperature. For mCRAMP staining, sections were incubated with goat serum (2%) and BSA (2%) for 1 h at room temperature, followed by incubation overnight at 4°C with 2 μg/ml rabbit anti-mCRAMP IgG (22, 23) or 2 μg/ml preimmune IgG from the same rabbit. HRP-labeled goat anti-rabbit IgG (1/400 dilution; Jackson ImmunoResearch Laboratories) was used as secondary Ab. 3,3'-Diaminobenzidine (Sigma-Aldrich) was used for visualization. Sections were counterstained with hematoxylin. For C. rodentium staining, paraffin-embedded or frozen sections fixed in 4% paraformaldehyde were incubated overnight with rabbit anti-C. rodentium IgG (1/1000; gift of Dr. B. Vallance) or an identical concentration of control rabbit IgG, and stained as above, or with Cy3-labeled goat anti-rabbit IgG (Jackson ImmunoResearch Laboratories) for confocal microscopy. Alexa 488-conjugated phalloidin was used for F-actin staining (Molecular Probes).
RNA extraction and RT-PCR
Total cellular RNA extraction and qualitative and real time RT-PCR were performed as described previously (6, 7) using the primers and conditions indicated in Table I.
Table I. Oligonucleotide primers
Isolation of colon epithelial cells
Colon epithelium was isolated as described before (24) and contained surface and crypt epithelium with few, if any, contaminating lamina propria cells. For protein lysates, epithelial cells were homogenized in 100 mM phosphate buffer containing 0.1% Triton X-100 and 0.5% protease inhibitor mixture III (Calbiochem), and centrifuged at 16,000 x g for 20 min. The protein content of the supernatants was measured by Bradford assay (Bio-Rad).
Antimicrobial assays
Aliquots containing 2500 CFU of bacteria were suspended in 25 μl of PBS (pH 7.4) with or without titrated concentrations of synthetic mCRAMP peptide (SynPep), 5 μg/μl epithelial cell lysate from Cnlp+/+ or Cnlp–/– mice, or 5 μg/μl BSA. After 2 h at 37°C, bacterial suspensions were plated on Luria-Bertani agar and incubated overnight at 37°C; then CFU were determined. EC50 is the mCRAMP concentration that decreases CFU by 50% relative to the peptide-free control. For inhibition zone assays (25), C. rodentium (107/ml) were added to 1.5% agarose containing tryptone (5 mg/ml) and plated in a 1-mm layer in 10-cm plastic dishes. Colon epithelial cell lysates (10 μl) were added to 3-mm-diameter wells in the agar, plates were incubated overnight at 37°C, and the diameter of the bacteria-free zone was determined.
Statistical analysis
Comparisons between groups used unpaired t tests or rank sum tests.
Results
Expression of mCRAMP in murine intestine
mCRAMP mRNA was constitutively expressed in the intestinal tract, where it is largely restricted to the colon, with little expression in the small intestine and no expression in the stomach or liver (Fig. 1A). By real time RT-PCR, the colon had >25-fold higher levels of mCRAMP mRNA than the small intestine (not shown). mCRAMP protein was selectively produced by surface colon epithelial cells with little, if any, mCRAMP immunostaining of epithelial cells that line the deeper crypts or cells in the lamina propria (Fig. 1B). No mCRAMP was immunostaining was detected in the small intestine (Fig. 1B). These data show that mCRAMP distribution in murine colon parallels that of the cathelicidin LL-37/hCAP18 in the human colon (6).
FIGURE 1. mCRAMP expression. A, Tissues of Cnlp+/+ mice, as indicated, were analyzed for mCRAMP and -actin mRNA transcripts by RT-PCR. B, Normal colon and small intestine (Int) from Cnlp+/+ mice was immunostained with mCRAMP-specific Ab (left) or control preimmune rabbit IgG (right). Original magnification, x200 for colon and x400 for small intestine.
mCRAMP has antimicrobial activity against C. rodentium in vitro
To evaluate in vitro antimicrobial activity of mCRAMP against C. rodentium, bacteria were incubated with synthetic mCRAMP. mCRAMP inhibited C. rodentium growth with an EC50 of 1.8 μM at physiological pH and salt concentrations and was similarly effective against two related human pathogens, EPEC and EHEC (Fig. 2A). To determine the antimicrobial activity of mCRAMP produced by colon epithelial cells, colon epithelial cell extracts from Cnlp+/+ and Cnlp–/– mice were assayed for antimicrobial activity against C. rodentium. Extracts from Cnlp+/+ mice had significantly greater antimicrobial activity against C. rodentium than those from Cnlp–/– mice (Fig. 2, B and C).
FIGURE 2. Antimicrobial activity of mCRAMP against C. rodentium in vitro. A, C. rodentium, EPEC, or EHEC was incubated without or with synthetic mCRAMP, after which CFU were determined. The EC50 of mCRAMP is shown for each bacterial strain. Values are mean ± SEM from three repeated experiments. B, Inhibition zone assay for C. rodentium. Left, Increased zone of clearing after addition of 0 or 4 μg/μl epithelial cell lysate from Clnp+/+ compared with Clnp–/– mice. Right, Titrated quantities of colonic epithelial cell lysate from Clnp+/+ () or Clnp–/– () mice were added. Values are mean ± SEM of three repeated experiments. *, p < 0.01 with Cnlp+/+ mice. C, C. rodentium were incubated in PBS containing BSA or epithelial cell lysate from Clnp+/+ or Clnp–/– mice, each at 5 μg/μl. Values are mean ± SEM of three repeated experiments. *, p < 0.01 BSA vs Clnp+/+; **, p < 0.05 Clnp+/+ vs Clnp–/–.
Increased colon colonization and systemic infection with C. rodentium in mCRAMP-deficient mice
To determine whether colon epithelial cell mCRAMP decreases colon colonization with C. rodentium, Cnlp+/+ and Cnlp–/– mice initially were infected with C. rodentium using an oral inoculum that caused infection in 100% of Cnlp+/+ mice (5 x 108 bacteria/mouse). This resulted in maximal colon colonization with C. rodentium within 1 week. Cnlp–/– mice developed significantly higher fecal counts of C. rodentium (8-fold) than did Cnlp+/+ littermates (Fig. 3A) and, at the peak of colonic colonization, had 10- to 30-fold higher numbers of C. rodentium in the spleen and mesenteric lymph nodes (Fig. 3, B and C). However, by day 14 after infection, significant differences in fecal bacterial counts between Cnlp+/+ and Cnlp–/– mice were no longer apparent (Fig. 3A). Infection was cleared by both Cnlp+/+ and Cnlp–/– mice when tested at 21 and 28 days after infection with undetectable numbers of C. rodentium in the feces of either Cnlp+/+ or Cnlp–/– mice. These data indicate that the presence of mCRAMP is important during early colonization of the colon epithelium.
FIGURE 3. Increased intestinal colonization and systemic infection of mCRAMP-deficient mice. Cnlp+/+ and Cnlp–/– mice were infected orally with C. rodentium at 5 x 108/mouse (A–C) or 2.5 x 107/mouse (D and E). Fecal counts of C. rodentium were determined at the indicated times after infection, and counts of C. rodentium in the spleen (B), mesenteric lymph nodes (MLN) (C), and colon (E) were assayed on day 7 after infection. Data points are from individual mice. Dotted lines indicate detection limit of the CFU assay. Horizontal bars are geometric means. *, p < 0.05; **, p < 0.01. N.S., not significant.
We next assessed whether the presence of mCRAMP is sufficient to prevent colon colonization with C. rodentium. For these experiments, Cnlp+/+ and Cnlp–/– mice were infected with a lower inoculum of C. rodentium (2.5 x 107 bacteria/mouse). As shown in Fig. 3, D and E, an inoculum that does not result in established infection in any of the wild-type mice resulted in the infection of a majority of Cnlp–/– mice. Total numbers of commensal colonic bacteria in Cnlp+/+ and Cnlp–/– mice were similar (Fig. 4). Thus, differences in the number of endogenous bacteria in the colon did not explain differences in susceptibility of Cnlp+/+ and Cnlp–/– mice to surface colonization with C. rodentium. These studies establish mCRAMP as an important, nonredundant component of antimicrobial defense in the colon. As shown in Fig. 5, the histological changes accompanying C. rodentium infection, with the development of marked crypt hyperplasia by 14 days after infection, are consistent with those known to occur during this infection (17, 18).
FIGURE 4. Similar bacterial commensal counts in the colon and feces of Cnlp+/+ and Cnlp–/– mice. A, Anerobic and aerobic bacteria in the colon and feces of Cnlp+/+ and Cnlp–/– mice (n = 3 each). There were no significant differences between Cnlp+/+ and Cnlp–/– mice. B and C, Gram staining was performed on 40 colonies from each sample from the colon or feces. There was no significant difference in the distribution of Gram-positive and Gram-negative bacteria between Cnlp+/+ and Cnlp–/– mice. (+), Gram positive; (–), Gram negative.
FIGURE 5. Histology of uninfected and C. rodentium-infected colon. Cnlp+/+ and Cnlp–/– mice were left uninfected or infected orally with C. rodentium (5 x 108/mouse). Sections of colon from uninfected mice and those infected for 7 and 14 days were stained with H&E. Original magnification, x200.
C. rodentium colonization of surface epithelium
We next investigated colonization of colon epithelial cells and mucosal injury over the course of C. rodentium infection. At 7 days after infection of Cnlp+/+ mice (5 x 108 C. rodentium/mouse), bacteria resided in close proximity to mCRAMP-expressing surface epithelial cells, and notably few bacteria were found in association with epithelial cells deeper within the colon crypts (Fig. 6Aa). By comparison, there was no epithelial surface colonization in Cnlp+/+ mice infected with the lower bacterial inoculum (2.5 x 107; Fig. 6Ac). At both inocula, C. rodentium colonization of surface epithelium was significantly greater in Cnlp–/– than in Cnlp+/+ mice and at the higher inocula was accompanied by surface epithelial cell damage. Consistent with greater colonization, numerous bacteria were associated with epithelial cells in the upper crypt region of Cnlp–/– mice (Fig. 6Ae). The number of upper colon crypts colonized per section were 11.2 ± 2.0, n = 6, in Cnlp+/+ mice and 44.3 ± 12.2, n = 7, in Cnlp–/– mice (mean ± SEM, p < 0.05).
FIGURE 6. C. rodentium and mCRAMP immunostaining of infected colon. A, Cnlp+/+ (a–d) or Cnlp–/– mice (e-h) were infected orally with 5 x 108 (a, b, e, and f) or 2.5 x 107 (c, d, g, and h) C. rodentium. Colon obtained 7 days postinfection was immunostained for C. rodentium (a, c, e, and g) or with control Ab (b, d, f, and h). B, Colon from Cnlp–/– mice (a) and Cnlp+/+ mice (b) infected with 5 x 108 bacteria were stained for C. rodentium (red) and actin (green) 7 days postinfection. Sections were analyzed by confocal microscopy. Yellow staining (arrows) shows colocalization of C. rodentium and actin at the apical surface. C, Colons from Cnlp+/+ mice infected with 5 x 108 C. rodentium for 7 days were stained with anti-mCRAMP (a) or control Ab (b). Original magnification, x200 (A and C) and x400 (B).
Although surface epithelial cells of Cnlp+/+ mice were colonized with fewer bacteria than mCRAMP-deficient mice, the interaction between bacteria and the epithelial apical surface appeared similar in both as assessed by confocal microscopy and colocalization of bacteria and F-actin (Fig. 6B, a and b). As in uninfected Cnlp+/+ mice (Fig. 1B), mCRAMP expression 7 days after infection with C. rodentium was limited to surface colon epithelial cells and was not found in cells in the lamina propria (Fig. 6Ca). There was little increase in inflammatory cells in the colon lamina propria 7 days after infection, although some inflammatory cells infiltrated the surface epithelium, and this was similar in Cnlp+/+ and Cnlp–/– mice (Figs. 5 and 6C).
Decreased epithelial mCRAMP expression at later times after infection
Two weeks after infection with C. rodentium (5 x 108 bacteria/mouse), the colon mucosa manifested crypt hyperplasia, a moderate to severe inflammatory cell infiltrate, and surface epithelial erosions. At this time, colon epithelial cells from Cnlp+/+ mice lacked mCRAMP, as assessed by immunostaining (Fig. 7, A and B). Moreover, the extent of colon pathology was similar in infected wild-type and mCRAMP-deficient mice (Figs. 5 and 7A), with the only difference being that scattered cells with characteristics of neutrophils stained for mCRAMP in Cnlp+/+ (Fig. 7C) but not mCRAMP-deficient mice. mCRAMP mRNA transcripts were significantly decreased in isolated epithelial cells from Cnlp+/+ mice (Fig. 7G) whereas expression of the neutrophil chemoattractant MIP-2 was up-regulated in epithelial cells and whole colon tissue of infected compared with uninfected Cnlp+/+ mice (Fig. 7H).
FIGURE 7. Colon mucosa 14 days postinfection. Colon from Cnlp+/+ mice infected with 5 x 108 C. rodentium 14 days earlier was stained for mCRAMP (A–C) or with control preimmune rabbit IgG (D–F). Original magnification, x200. Boxed insets in A and D are magnified in B, C and E, F, respectively. B, Lack of mCRAMP staining of surface epithelium; C, mCRAMP staining of neutrophils in the lamina propria, as indicated by arrows. G and H, Relative mRNA levels for mCRAMP (left) and MIP-2 (right) in whole colon and isolated epithelial cells of C. rodentium-infected Cnlp+/+ mice 14 days after infection determined by real time RT-PCR. Data are fold change on day 14 infected () relative to uninfected controls (). Values are mean ± SEM of three repeated experiments. *, p < 0.01 compared with uninfected control. N.S., not significant.
Discussion
Our findings establish the functional importance of epithelial cathelicidin as a component of early innate antimicrobial defense in the colon. Moreover, these data provide evidence that a cathelicidin produced by colon epithelial surface cells has a nonredundant function in host defense against a clinically important class of foodborne pathogens that colonize the intestinal tract. At lower inocula, epithelial mCRAMP prevented, and at higher inocula significantly decreased, epithelial and luminal colonization, epithelial cell damage, and systemic infection with a bacterial pathogen the lifestyle and pathogenicity of which are characterized by adherence to intestinal epithelial cells, rather than epithelial cell invasion. The phenotype of the mCRAMP-deficient mice is novel and differs from that of mice lacking IL-12p40 or TNFR55 where the major effect is a marked delay in bacterial eradication and a significant increase in bacterial numbers 2 wk after infection (26, 27). The latter data are consistent with the known importance of acquired B cell- and T cell-mediated immunity in clearing C. rodentium (26, 27). Although mice lacking mCRAMP were more susceptible to initial infection, reflecting the failure of innate defenses to protect against epithelial colonization, once infected, unlike IL-12p40 or TNFR55-deficient mice (26, 27), mCRAMP-deficient mice could clear infection as rapidly as wild-type mice, showing that acquired immunity was not compromised. Decreased epithelial mCRAMP expression at this later stage of infection likely reflects increased epithelial cell proliferation and the replacement of more differentiated surface epithelial cells with less mature cells migrating from an actively proliferating crypt compartment (6).
mCRAMP and human LL-37/hCAP18 manifest the same restricted distribution to surface epithelial cells in the colon and have similar killing activities against enteric pathogens in vitro (Ref. 6) and M. Iimura, unpublished data). Taken together, these strongly suggest that the murine model used herein has relevance for human enteric infections. This notion gains support from the observation mCRAMP-deficient mice have decreased resistance to cutaneous infections with group A streptococci (28) and from clinical studies in atopic dermatitis patients that suggest that human cathelicidin is important in preventing skin infections (29).
The concentrations of mCRAMP required for in vitro bacterial killing raise the question as to whether cathelicidin concentrations, or concentrations of other classes of antimicrobial peptides that can mediate antimicrobial activity in vitro, are attained in vivo. Low micromolar concentrations of mCRAMP were required for in vitro killing of C. rodentium, EHEC, and EPEC herein, and up to 2- to 4-fold higher concentrations were required for in vitro killing of several other enteric pathogens by human LL-37/hCAP18 (6, 7) or mCRAMP (M. Imura, unpublished data). In the case of the -defensins, a different class of antimicrobial peptides that are produced by small intestinal Paneth cells, in vitro killing of several enteric pathogens occurred over a similar concentration range (4). Recent studies indicate that these and higher concentrations of antimicrobial peptides are produced in the intestinal microenvironment in vivo. For example, human cathelicidin in gastric juice of Helicobacter pylori-infected individuals reached concentrations up to 15 μM, yet effectively killed H. pylori in vitro at 10-fold lower concentrations (7), and murine -defensin concentrations that exceed those required for microbicidal activity in vitro were estimated to be present in the microenvironment of small intestinal crypts (4). Of note, relatively small differences (2- to 4-fold) in the amount of cathelicidin required to kill group B Streptococcus in vitro were paralleled in vivo by a significant difference in the extent of skin ulceration caused by those bacteria (28).
We considered the possibility that a decrease in epithelial mCRAMP might result in increased bacterial numbers in the colon and that, in turn, might decrease the ability of C. rodentium to colonize the colon. However, that was not the case. Nonetheless, we cannot exclude the possibility that there are qualitative differences in the commensal flora of Cnlp+/+ and Cnlp–/– mice that might influence the colonization potential of C. rodentium. We favor the notion that differences in the magnitude of surface colonization in Cnlp+/+ and Cnlp–/– mice are due to a direct effect of mCRAMP on C. rodentium, although the effects of mCRAMP on other innate functions (30) including those of colon surface epithelial cells may also play a role.
Like its human counterpart LL-37/hCAP18, mCRAMP was produced by colon surface epithelium, but not by epithelial cells that line the small intestinal villi and crypts (6). This firmly establishes a paradigm whereby the distribution of different antimicrobial defense molecules is regionally restricted in the intestine, and within a given region, restricted to cell types that differ in either lineage or differentiation state. Such differences support the notion that different antimicrobial defense molecules are host adapted to have different functional roles in various intestinal sites. Whereas Paneth cell -defensins may protect adjacent stem cells in the small intestinal crypts from enteric pathogens (4, 31), the selective expression of mCRAMP by surface epithelial cells in the colon could function to effectively protect against colon and epithelial cell colonization with an important class of epithelial adherent foodborne pathogen.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by National Institutes of Health Grants DK59860, DK35108, and AI56075 and by a grant from the Cystic Fibrosis Foundation.
2 Current address: Laboratory of Epithelial Immunobiology, RIKEN Research Center for Allergy and Immunology, Yokohama, Japan.
3 Address correspondence and reprint requests to Dr. Martin F. Kagnoff, Laboratory of Mucosal Immunology, Departments of Medicine and Pediatrics, Mail Code 0623D, University of California, San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0623. E-mail address: mkagnoff{at}ucsd.edu
4 Abbreviations used in this paper: mCRAMP, mouse cathelicidin-related antimicrobial peptide; EPEC, enteropathogenic Escherichia coli; EHEC, enterohemorrhagic E. coli.
Received for publication December 13, 2004. Accepted for publication February 11, 2005.
References
Murakami, M., T. Ohtake, R. A. Dorschner, R. L. Gallo. 2002. Cathelicidin antimicrobial peptides are expressed in salivary glands and saliva. J. Dent. Res. 81:845.
Zanetti, M.. 2004. Cathelicidins, multifunctional peptides of the innate immunity. J. Leukocyte Biol. 75:39.
Pestonjamasp, V. K., K. H. Huttner, R. L. Gallo. 2001. Processing site and gene structure for the murine antimicrobial peptide CRAMP. Peptides 22:1643
Ayabe, T., D. P. Satchell, C. L. Wilson, W. C. Parks, M. E. Selsted, A. J. Ouellette. 2000. Secretion of microbicidal -defensins by intestinal Paneth cells in response to bacteria. Nat. Immunol. 1:113
O’Neil, D. A., E. M. Porter, D. Elewaut, G. M. Anderson, L. Eckmann, T. Ganz, M. F. Kagnoff. 1999. Expression and regulation of the human -defensins hBD-1 and hBD-2 in intestinal epithelium. J. Immunol. 163:6718.
Hase, K., L. Eckmann, J. D. Leopard, N. Varki, M. F. Kagnoff. 2002. Cell differentiation is a key determinant of cathelicidin LL-37/human cationic antimicrobial protein 18 expression by human colon epithelium. Infect. Immun. 70:953.
Hase, K., M. Murakami, M. Iimura, S. P. Cole, Y. Horibe, T. Ohtake, M. Obonyo, R. L. Gallo, L. Eckmann, M. F. Kagnoff. 2003. Expression of LL-37 by human gastric epithelial cells as a potential host defense mechanism against Helicobacter pylori. Gastroenterology 125:1613.
Schauer, D. B., S. Falkow. 1993. Attaching and effacing locus of a Citrobacter freundii biotype that causes transmissible murine colonic hyperplasia. Infect. Immun. 61:2486
Schauer, D. B., S. Falkow. 1993. The eae gene of Citrobacter freundii biotype 4280 is necessary for colonization in transmissible murine colonic hyperplasia. Infect. Immun. 61:4654.
Frankel, G., A. D. Phillips, M. Novakova, H. Field, D. C. Candy, D. B. Schauer, G. Douce, G. Dougan. 1996. Intimin from enteropathogenic Escherichia coli restores murine virulence to a Citrobacter rodentium eaeA mutant: induction of an immunoglobulin A response to intimin and EspB. Infect. Immun. 64:5315.
Luperchio, S. A., J. V. Newman, C. A. Dangler, M. D. Schrenzel, D. J. Brenner, A. G. Steigerwalt, D. B. Schauer. 2000. Citrobacter rodentium, the causative agent of transmissible murine colonic hyperplasia, exhibits clonality: synonymy of C. rodentium and mouse-pathogenic Escherichia coli. J. Clin. Microbiol. 38:4343.
Luperchio, S. A., D. B. Schauer. 2001. Molecular pathogenesis of Citrobacter rodentium and transmissible murine colonic hyperplasia. Microb. Infect. 3:333.
Vallance, B. A., C. Chan, M. L. Robertson, B. B. Finlay. 2002. Enteropathogenic and enterohemorrhagic Escherichia coli infections: emerging themes in pathogenesis and prevention. Can. J. Gastroenterol. 16:771.
Deng, W., B. A. Vallance, Y. Li, J. L. Puente, B. B. Finlay. 2003. Citrobacter rodentium translocated intimin receptor (Tir) is an essential virulence factor needed for actin condensation, intestinal colonization and colonic hyperplasia in mice. Mol. Microbiol. 48:95.
Barthold, S. W., G. L. Coleman, P. N. Bhatt, G. W. Osbaldiston, A. M. Jonas. 1976. The etiology of transmissible murine colonic hyperplasia. Lab. Anim. Sci. 26:889.
Maggio-Price, L., K. L. Nicholson, K. M. Kline, T. Birkebak, I. Suzuki, D. L. Wilson, D. Schauer, P. J. Fink. 1998. Diminished reproduction, failure to thrive, and altered immunologic function in a colony of T-cell receptor transgenic mice: possible role of Citrobacter rodentium. Lab. Anim. Sci. 48:145
Barthold, S. W.. 1980. The microbiology of transmissible murine colonic hyperplasia. Lab. Anim. Sci. 30:167.
Johnson, E., S. W. Barthold. 1979. The ultrastructure of transmissible murine colonic hyperplasia. Am. J. Pathol. 97:291.[
Higgins, L. M., G. Frankel, G. Douce, G. Dougan, T. T. MacDonald. 1999. Citrobacter rodentium infection in mice elicits a mucosal Th1 cytokine response and lesions similar to those in murine inflammatory bowel disease. Infect. Immun. 67:3031.
Ghaem-Maghami, M., C. P. Simmons, S. Daniell, M. Pizza, D. Lewis, G. Frankel, G. Dougan. 2001. Intimin-specific immune responses prevent bacterial colonization by the attaching-effacing pathogen Citrobacter rodentium. Infect. Immun. 69:5597.
Maaser, C., M. P. Housley, M. Iimura, J. R. Smith, B. A. Vallance, B. B. Finlay, J. R. Schreiber, N. M. Varki, M. F. Kagnoff, L. Eckmann. 2004. Clearance of Citrobacter rodentium requires B cells but not secretory immunoglobulin A (IgA) or IgM antibodies. Infect. Immun. 72:3315
Gallo, R. L., K. J. Kim, M. Bernfield, C. A. Kozak, M. Zanetti, L. Merluzzi, R. Gennaro. 1997. Identification of CRAMP, a cathelin-related antimicrobial peptide expressed in the embryonic and adult mouse. J. Biol. Chem. 272:13088.
Dorschner, R. A., V. K. Pestonjamasp, S. Tamakuwala, T. Ohtake, J. Rudisill, V. Nizet, B. Agerberth, G. H. Gudmundsson, R. L. Gallo. 2001. Cutaneous injury induces the release of cathelicidin anti-microbial peptides active against group A Streptococcus. J. Invest. Dermatol. 117:91.
Iimura, M., T. Nakamura, S. Shinozaki, B. Iizuka, Y. Inoue, S. Suzuki, N. Hayashi. 2000. Bax is downregulated in inflamed colonic mucosa of ulcerative colitis. Gut 47:228.
Frohm, M., H. Gunne, A. C. Bergman, B. Agerberth, T. Bergman, A. Boman, S. Liden, H. Jornvall, H. G. Boman. 1996. Biochemical and antibacterial analysis of human wound and blister fluid. Eur. J. Biochem. 237:86
Simmons, C. P., N. S. Goncalves, M. Ghaem-Maghami, M. Bajaj-Elliott, S. Clare, B. Neves, G. Frankel, G. Dougan, T. T. MacDonald. 2002. Impaired resistance and enhanced pathology during infection with a noninvasive, attaching-effacing enteric bacterial pathogen, Citrobacter rodentium, in mice lacking IL-12 or IFN-. J. Immunol. 168:1804.
Ong, P. Y., T. Ohtake, C. Brandt, I. Strickland, M. Boguniewicz, T. Ganz, R. L. Gallo, D. Y. Leung. 2002. Endogenous antimicrobial peptides and skin infections in atopic dermatitis. N. Engl. J. Med. 347:1151.
Scott, M. G., D. J. Davidson, M. R. Gold, D. Bowdish, R. E. Hancock. 2002. The human antimicrobial peptide LL-37 is a multifunctional modulator of innate immune responses. J. Immunol. 169:3883.
Wilson, C. L., A. J. Ouellette, D. P. Satchell, T. Ayabe, Y. S. Lopez-Boado, J. L. Stratman, S. J. Hultgren, L. M. Matrisian, W. C. Parks. 1999. Regulation of intestinal -defensin activation by the metalloproteinase matrilysin in innate host defense. Science 286:113.(Mitsutoshi Iimura, Richar)
Cathelicidin-related antimicrobial peptide (mCRAMP), the sole murine cathelicidin, is encoded by the gene Cnlp. We show that mCRAMP expression in the intestinal tract is largely restricted to surface epithelial cells in the colon. Synthetic mCRAMP had antimicrobial activity against the murine enteric pathogen Citrobacter rodentium, which like the related clinically important human pathogens enteropathogenic Escherichia coli and enterohemorrhagic E. coli, adheres to the apical membrane of intestinal epithelial cells. Colon epithelial cell extracts from Cnlp+/+ mice had significantly greater antimicrobial activity against C. rodentium than those of mutant Cnlp–/– mice that lack mCRAMP. Cnlp–/– mice developed significantly greater colon surface and crypt epithelial cell colonization, surface epithelial cell damage, and systemic dissemination of infection than Cnlp+/+ mice after oral infection with C. rodentium. Moreover, Cnlp+/+ mice were protected from oral infections with C. rodentium inocula that infected the majority of Cnlp–/– mice. These results establish cathelicidin as an important component of innate antimicrobial defense in the colon.
Introduction
Cathelicidins are evolutionarily conserved antimicrobial peptides present in all mammals. They are synthesized in precursor form with a conserved N-terminal signal sequence, a cathelin domain, and a C-terminal domain that is cleaved to release a peptide with antimicrobial activity (1, 2). Murine cathelicidin-related antimicrobial peptide (mCRAMP)4 is the murine ortholog of the sole and structurally related human cathelicidin, LL-37/hCAP18 (3). In contrast to other antimicrobial peptides in the human intestinal tract, such as -defensins that are restricted to small intestinal Paneth cells (4) or -defensins that are expressed by crypt, villous, and surface epithelial cells throughout the small and large intestine (5), the distribution of cathelicidin in the human intestinal tract is limited to surface epithelial cells in the colon and stomach (6, 7).
The murine enteric bacterial pathogen Citrobacter rodentium has important virulence features in common with human enteropathogenic Escherichia coli (EPEC) (8, 9, 10, 11, 12). Within 1 week of oral infection, C. rodentium colonize the murine colon and like EPEC and enterohemorrhagic E. coli (EHEC) in humans, C. rodentium adheres to intestinal epithelial cells where it produces the characteristic attaching and effacing lesion (10, 13, 14). Normally, few if any bacteria invade the mucosa and reach extraintestinal sites (15, 16). At later times, infection results in mucosal erosions, increased epithelial cell proliferation, and crypt hyperplasia (17, 18). Little is known about early innate mucosal defenses against this pathogen, although acquired immunity is needed to control and eliminate C. rodentium (10, 16, 19, 20, 21).
In light of the selective expression of cathelicidin in the human colon, we hypothesized that cathelicidin produced by colon surface epithelial cells may be important in host innate defense against pathogenic enteric bacteria for which lifestyle involves adherence to those cells. Using genetically mutant mice that lack cathelicidin and with C. rodentium as a model murine epithelial adherent pathogen, we have explored the functional importance of cathelicidin in early intestinal innate antimicrobial defense.
Materials and Methods
Mice
mCRAMP-deficient (Cnlp–/–) mice were generated as described (22). Cnlp+/+ littermates were used as controls. Cnlp–/– mice were healthy and fertile and did not manifest differences in intestinal mucosal architecture from Cnlp+/+ littermates used as controls. All studies were approved by the University of California at San Diego Animal Welfare Committee.
Bacteria
The following bacteria were used: C. rodentium (21); enteropathogenic E. coli serotype O111:NM, eae+ isolate DEC12f (CDC B170; American Type Culture Collection No. 43887; American Type Culture Collection, Manassas, VA); enterohemorrhagic E. coli O157:H7 strain 86-24 (gift of M. Donnenberg).
To determine total bacterial numbers in the colon and feces of Cnlp+/+ and Cnlp–/– mice, colon or feces were homogenized in TS broth. For aerobic culture, samples were plated onto tryptic soy blood agar containing 5% sheep blood and then incubated at 37°C overnight. For anaerobic culture, samples were plated onto Columbia blood agar with 5% sheep blood and then incubated at 37°C for 48 h in anaerobic culture system (Oxide Atmosphere Generation System for Anaerobic Bacteriology and AnaeroGen) after which CFU were determined. Gram staining was performed on 40 colonies from each sample.
Infections
Mice were infected orally with C. rodentium. Bacterial numbers in feces or homogenized colon were determined by plating on MacConkey agar (21). The detection limit of the CFU assay was 103 colonies per g of feces or per colon and <10 colonies per organ in spleen and mesenteric lymph nodes. Identity of representative bacterial colonies was verified by PCR (21).
Histological analysis and immunohistochemistry
Colons processed as Swiss rolls (21) were fixed (1% ZnSO4, 10% formalin) and embedded in paraffin. Sections (5 μm) were treated with 0.3% H2O2 in PBS for 10 min at room temperature. For mCRAMP staining, sections were incubated with goat serum (2%) and BSA (2%) for 1 h at room temperature, followed by incubation overnight at 4°C with 2 μg/ml rabbit anti-mCRAMP IgG (22, 23) or 2 μg/ml preimmune IgG from the same rabbit. HRP-labeled goat anti-rabbit IgG (1/400 dilution; Jackson ImmunoResearch Laboratories) was used as secondary Ab. 3,3'-Diaminobenzidine (Sigma-Aldrich) was used for visualization. Sections were counterstained with hematoxylin. For C. rodentium staining, paraffin-embedded or frozen sections fixed in 4% paraformaldehyde were incubated overnight with rabbit anti-C. rodentium IgG (1/1000; gift of Dr. B. Vallance) or an identical concentration of control rabbit IgG, and stained as above, or with Cy3-labeled goat anti-rabbit IgG (Jackson ImmunoResearch Laboratories) for confocal microscopy. Alexa 488-conjugated phalloidin was used for F-actin staining (Molecular Probes).
RNA extraction and RT-PCR
Total cellular RNA extraction and qualitative and real time RT-PCR were performed as described previously (6, 7) using the primers and conditions indicated in Table I.
Table I. Oligonucleotide primers
Isolation of colon epithelial cells
Colon epithelium was isolated as described before (24) and contained surface and crypt epithelium with few, if any, contaminating lamina propria cells. For protein lysates, epithelial cells were homogenized in 100 mM phosphate buffer containing 0.1% Triton X-100 and 0.5% protease inhibitor mixture III (Calbiochem), and centrifuged at 16,000 x g for 20 min. The protein content of the supernatants was measured by Bradford assay (Bio-Rad).
Antimicrobial assays
Aliquots containing 2500 CFU of bacteria were suspended in 25 μl of PBS (pH 7.4) with or without titrated concentrations of synthetic mCRAMP peptide (SynPep), 5 μg/μl epithelial cell lysate from Cnlp+/+ or Cnlp–/– mice, or 5 μg/μl BSA. After 2 h at 37°C, bacterial suspensions were plated on Luria-Bertani agar and incubated overnight at 37°C; then CFU were determined. EC50 is the mCRAMP concentration that decreases CFU by 50% relative to the peptide-free control. For inhibition zone assays (25), C. rodentium (107/ml) were added to 1.5% agarose containing tryptone (5 mg/ml) and plated in a 1-mm layer in 10-cm plastic dishes. Colon epithelial cell lysates (10 μl) were added to 3-mm-diameter wells in the agar, plates were incubated overnight at 37°C, and the diameter of the bacteria-free zone was determined.
Statistical analysis
Comparisons between groups used unpaired t tests or rank sum tests.
Results
Expression of mCRAMP in murine intestine
mCRAMP mRNA was constitutively expressed in the intestinal tract, where it is largely restricted to the colon, with little expression in the small intestine and no expression in the stomach or liver (Fig. 1A). By real time RT-PCR, the colon had >25-fold higher levels of mCRAMP mRNA than the small intestine (not shown). mCRAMP protein was selectively produced by surface colon epithelial cells with little, if any, mCRAMP immunostaining of epithelial cells that line the deeper crypts or cells in the lamina propria (Fig. 1B). No mCRAMP was immunostaining was detected in the small intestine (Fig. 1B). These data show that mCRAMP distribution in murine colon parallels that of the cathelicidin LL-37/hCAP18 in the human colon (6).
FIGURE 1. mCRAMP expression. A, Tissues of Cnlp+/+ mice, as indicated, were analyzed for mCRAMP and -actin mRNA transcripts by RT-PCR. B, Normal colon and small intestine (Int) from Cnlp+/+ mice was immunostained with mCRAMP-specific Ab (left) or control preimmune rabbit IgG (right). Original magnification, x200 for colon and x400 for small intestine.
mCRAMP has antimicrobial activity against C. rodentium in vitro
To evaluate in vitro antimicrobial activity of mCRAMP against C. rodentium, bacteria were incubated with synthetic mCRAMP. mCRAMP inhibited C. rodentium growth with an EC50 of 1.8 μM at physiological pH and salt concentrations and was similarly effective against two related human pathogens, EPEC and EHEC (Fig. 2A). To determine the antimicrobial activity of mCRAMP produced by colon epithelial cells, colon epithelial cell extracts from Cnlp+/+ and Cnlp–/– mice were assayed for antimicrobial activity against C. rodentium. Extracts from Cnlp+/+ mice had significantly greater antimicrobial activity against C. rodentium than those from Cnlp–/– mice (Fig. 2, B and C).
FIGURE 2. Antimicrobial activity of mCRAMP against C. rodentium in vitro. A, C. rodentium, EPEC, or EHEC was incubated without or with synthetic mCRAMP, after which CFU were determined. The EC50 of mCRAMP is shown for each bacterial strain. Values are mean ± SEM from three repeated experiments. B, Inhibition zone assay for C. rodentium. Left, Increased zone of clearing after addition of 0 or 4 μg/μl epithelial cell lysate from Clnp+/+ compared with Clnp–/– mice. Right, Titrated quantities of colonic epithelial cell lysate from Clnp+/+ () or Clnp–/– () mice were added. Values are mean ± SEM of three repeated experiments. *, p < 0.01 with Cnlp+/+ mice. C, C. rodentium were incubated in PBS containing BSA or epithelial cell lysate from Clnp+/+ or Clnp–/– mice, each at 5 μg/μl. Values are mean ± SEM of three repeated experiments. *, p < 0.01 BSA vs Clnp+/+; **, p < 0.05 Clnp+/+ vs Clnp–/–.
Increased colon colonization and systemic infection with C. rodentium in mCRAMP-deficient mice
To determine whether colon epithelial cell mCRAMP decreases colon colonization with C. rodentium, Cnlp+/+ and Cnlp–/– mice initially were infected with C. rodentium using an oral inoculum that caused infection in 100% of Cnlp+/+ mice (5 x 108 bacteria/mouse). This resulted in maximal colon colonization with C. rodentium within 1 week. Cnlp–/– mice developed significantly higher fecal counts of C. rodentium (8-fold) than did Cnlp+/+ littermates (Fig. 3A) and, at the peak of colonic colonization, had 10- to 30-fold higher numbers of C. rodentium in the spleen and mesenteric lymph nodes (Fig. 3, B and C). However, by day 14 after infection, significant differences in fecal bacterial counts between Cnlp+/+ and Cnlp–/– mice were no longer apparent (Fig. 3A). Infection was cleared by both Cnlp+/+ and Cnlp–/– mice when tested at 21 and 28 days after infection with undetectable numbers of C. rodentium in the feces of either Cnlp+/+ or Cnlp–/– mice. These data indicate that the presence of mCRAMP is important during early colonization of the colon epithelium.
FIGURE 3. Increased intestinal colonization and systemic infection of mCRAMP-deficient mice. Cnlp+/+ and Cnlp–/– mice were infected orally with C. rodentium at 5 x 108/mouse (A–C) or 2.5 x 107/mouse (D and E). Fecal counts of C. rodentium were determined at the indicated times after infection, and counts of C. rodentium in the spleen (B), mesenteric lymph nodes (MLN) (C), and colon (E) were assayed on day 7 after infection. Data points are from individual mice. Dotted lines indicate detection limit of the CFU assay. Horizontal bars are geometric means. *, p < 0.05; **, p < 0.01. N.S., not significant.
We next assessed whether the presence of mCRAMP is sufficient to prevent colon colonization with C. rodentium. For these experiments, Cnlp+/+ and Cnlp–/– mice were infected with a lower inoculum of C. rodentium (2.5 x 107 bacteria/mouse). As shown in Fig. 3, D and E, an inoculum that does not result in established infection in any of the wild-type mice resulted in the infection of a majority of Cnlp–/– mice. Total numbers of commensal colonic bacteria in Cnlp+/+ and Cnlp–/– mice were similar (Fig. 4). Thus, differences in the number of endogenous bacteria in the colon did not explain differences in susceptibility of Cnlp+/+ and Cnlp–/– mice to surface colonization with C. rodentium. These studies establish mCRAMP as an important, nonredundant component of antimicrobial defense in the colon. As shown in Fig. 5, the histological changes accompanying C. rodentium infection, with the development of marked crypt hyperplasia by 14 days after infection, are consistent with those known to occur during this infection (17, 18).
FIGURE 4. Similar bacterial commensal counts in the colon and feces of Cnlp+/+ and Cnlp–/– mice. A, Anerobic and aerobic bacteria in the colon and feces of Cnlp+/+ and Cnlp–/– mice (n = 3 each). There were no significant differences between Cnlp+/+ and Cnlp–/– mice. B and C, Gram staining was performed on 40 colonies from each sample from the colon or feces. There was no significant difference in the distribution of Gram-positive and Gram-negative bacteria between Cnlp+/+ and Cnlp–/– mice. (+), Gram positive; (–), Gram negative.
FIGURE 5. Histology of uninfected and C. rodentium-infected colon. Cnlp+/+ and Cnlp–/– mice were left uninfected or infected orally with C. rodentium (5 x 108/mouse). Sections of colon from uninfected mice and those infected for 7 and 14 days were stained with H&E. Original magnification, x200.
C. rodentium colonization of surface epithelium
We next investigated colonization of colon epithelial cells and mucosal injury over the course of C. rodentium infection. At 7 days after infection of Cnlp+/+ mice (5 x 108 C. rodentium/mouse), bacteria resided in close proximity to mCRAMP-expressing surface epithelial cells, and notably few bacteria were found in association with epithelial cells deeper within the colon crypts (Fig. 6Aa). By comparison, there was no epithelial surface colonization in Cnlp+/+ mice infected with the lower bacterial inoculum (2.5 x 107; Fig. 6Ac). At both inocula, C. rodentium colonization of surface epithelium was significantly greater in Cnlp–/– than in Cnlp+/+ mice and at the higher inocula was accompanied by surface epithelial cell damage. Consistent with greater colonization, numerous bacteria were associated with epithelial cells in the upper crypt region of Cnlp–/– mice (Fig. 6Ae). The number of upper colon crypts colonized per section were 11.2 ± 2.0, n = 6, in Cnlp+/+ mice and 44.3 ± 12.2, n = 7, in Cnlp–/– mice (mean ± SEM, p < 0.05).
FIGURE 6. C. rodentium and mCRAMP immunostaining of infected colon. A, Cnlp+/+ (a–d) or Cnlp–/– mice (e-h) were infected orally with 5 x 108 (a, b, e, and f) or 2.5 x 107 (c, d, g, and h) C. rodentium. Colon obtained 7 days postinfection was immunostained for C. rodentium (a, c, e, and g) or with control Ab (b, d, f, and h). B, Colon from Cnlp–/– mice (a) and Cnlp+/+ mice (b) infected with 5 x 108 bacteria were stained for C. rodentium (red) and actin (green) 7 days postinfection. Sections were analyzed by confocal microscopy. Yellow staining (arrows) shows colocalization of C. rodentium and actin at the apical surface. C, Colons from Cnlp+/+ mice infected with 5 x 108 C. rodentium for 7 days were stained with anti-mCRAMP (a) or control Ab (b). Original magnification, x200 (A and C) and x400 (B).
Although surface epithelial cells of Cnlp+/+ mice were colonized with fewer bacteria than mCRAMP-deficient mice, the interaction between bacteria and the epithelial apical surface appeared similar in both as assessed by confocal microscopy and colocalization of bacteria and F-actin (Fig. 6B, a and b). As in uninfected Cnlp+/+ mice (Fig. 1B), mCRAMP expression 7 days after infection with C. rodentium was limited to surface colon epithelial cells and was not found in cells in the lamina propria (Fig. 6Ca). There was little increase in inflammatory cells in the colon lamina propria 7 days after infection, although some inflammatory cells infiltrated the surface epithelium, and this was similar in Cnlp+/+ and Cnlp–/– mice (Figs. 5 and 6C).
Decreased epithelial mCRAMP expression at later times after infection
Two weeks after infection with C. rodentium (5 x 108 bacteria/mouse), the colon mucosa manifested crypt hyperplasia, a moderate to severe inflammatory cell infiltrate, and surface epithelial erosions. At this time, colon epithelial cells from Cnlp+/+ mice lacked mCRAMP, as assessed by immunostaining (Fig. 7, A and B). Moreover, the extent of colon pathology was similar in infected wild-type and mCRAMP-deficient mice (Figs. 5 and 7A), with the only difference being that scattered cells with characteristics of neutrophils stained for mCRAMP in Cnlp+/+ (Fig. 7C) but not mCRAMP-deficient mice. mCRAMP mRNA transcripts were significantly decreased in isolated epithelial cells from Cnlp+/+ mice (Fig. 7G) whereas expression of the neutrophil chemoattractant MIP-2 was up-regulated in epithelial cells and whole colon tissue of infected compared with uninfected Cnlp+/+ mice (Fig. 7H).
FIGURE 7. Colon mucosa 14 days postinfection. Colon from Cnlp+/+ mice infected with 5 x 108 C. rodentium 14 days earlier was stained for mCRAMP (A–C) or with control preimmune rabbit IgG (D–F). Original magnification, x200. Boxed insets in A and D are magnified in B, C and E, F, respectively. B, Lack of mCRAMP staining of surface epithelium; C, mCRAMP staining of neutrophils in the lamina propria, as indicated by arrows. G and H, Relative mRNA levels for mCRAMP (left) and MIP-2 (right) in whole colon and isolated epithelial cells of C. rodentium-infected Cnlp+/+ mice 14 days after infection determined by real time RT-PCR. Data are fold change on day 14 infected () relative to uninfected controls (). Values are mean ± SEM of three repeated experiments. *, p < 0.01 compared with uninfected control. N.S., not significant.
Discussion
Our findings establish the functional importance of epithelial cathelicidin as a component of early innate antimicrobial defense in the colon. Moreover, these data provide evidence that a cathelicidin produced by colon epithelial surface cells has a nonredundant function in host defense against a clinically important class of foodborne pathogens that colonize the intestinal tract. At lower inocula, epithelial mCRAMP prevented, and at higher inocula significantly decreased, epithelial and luminal colonization, epithelial cell damage, and systemic infection with a bacterial pathogen the lifestyle and pathogenicity of which are characterized by adherence to intestinal epithelial cells, rather than epithelial cell invasion. The phenotype of the mCRAMP-deficient mice is novel and differs from that of mice lacking IL-12p40 or TNFR55 where the major effect is a marked delay in bacterial eradication and a significant increase in bacterial numbers 2 wk after infection (26, 27). The latter data are consistent with the known importance of acquired B cell- and T cell-mediated immunity in clearing C. rodentium (26, 27). Although mice lacking mCRAMP were more susceptible to initial infection, reflecting the failure of innate defenses to protect against epithelial colonization, once infected, unlike IL-12p40 or TNFR55-deficient mice (26, 27), mCRAMP-deficient mice could clear infection as rapidly as wild-type mice, showing that acquired immunity was not compromised. Decreased epithelial mCRAMP expression at this later stage of infection likely reflects increased epithelial cell proliferation and the replacement of more differentiated surface epithelial cells with less mature cells migrating from an actively proliferating crypt compartment (6).
mCRAMP and human LL-37/hCAP18 manifest the same restricted distribution to surface epithelial cells in the colon and have similar killing activities against enteric pathogens in vitro (Ref. 6) and M. Iimura, unpublished data). Taken together, these strongly suggest that the murine model used herein has relevance for human enteric infections. This notion gains support from the observation mCRAMP-deficient mice have decreased resistance to cutaneous infections with group A streptococci (28) and from clinical studies in atopic dermatitis patients that suggest that human cathelicidin is important in preventing skin infections (29).
The concentrations of mCRAMP required for in vitro bacterial killing raise the question as to whether cathelicidin concentrations, or concentrations of other classes of antimicrobial peptides that can mediate antimicrobial activity in vitro, are attained in vivo. Low micromolar concentrations of mCRAMP were required for in vitro killing of C. rodentium, EHEC, and EPEC herein, and up to 2- to 4-fold higher concentrations were required for in vitro killing of several other enteric pathogens by human LL-37/hCAP18 (6, 7) or mCRAMP (M. Imura, unpublished data). In the case of the -defensins, a different class of antimicrobial peptides that are produced by small intestinal Paneth cells, in vitro killing of several enteric pathogens occurred over a similar concentration range (4). Recent studies indicate that these and higher concentrations of antimicrobial peptides are produced in the intestinal microenvironment in vivo. For example, human cathelicidin in gastric juice of Helicobacter pylori-infected individuals reached concentrations up to 15 μM, yet effectively killed H. pylori in vitro at 10-fold lower concentrations (7), and murine -defensin concentrations that exceed those required for microbicidal activity in vitro were estimated to be present in the microenvironment of small intestinal crypts (4). Of note, relatively small differences (2- to 4-fold) in the amount of cathelicidin required to kill group B Streptococcus in vitro were paralleled in vivo by a significant difference in the extent of skin ulceration caused by those bacteria (28).
We considered the possibility that a decrease in epithelial mCRAMP might result in increased bacterial numbers in the colon and that, in turn, might decrease the ability of C. rodentium to colonize the colon. However, that was not the case. Nonetheless, we cannot exclude the possibility that there are qualitative differences in the commensal flora of Cnlp+/+ and Cnlp–/– mice that might influence the colonization potential of C. rodentium. We favor the notion that differences in the magnitude of surface colonization in Cnlp+/+ and Cnlp–/– mice are due to a direct effect of mCRAMP on C. rodentium, although the effects of mCRAMP on other innate functions (30) including those of colon surface epithelial cells may also play a role.
Like its human counterpart LL-37/hCAP18, mCRAMP was produced by colon surface epithelium, but not by epithelial cells that line the small intestinal villi and crypts (6). This firmly establishes a paradigm whereby the distribution of different antimicrobial defense molecules is regionally restricted in the intestine, and within a given region, restricted to cell types that differ in either lineage or differentiation state. Such differences support the notion that different antimicrobial defense molecules are host adapted to have different functional roles in various intestinal sites. Whereas Paneth cell -defensins may protect adjacent stem cells in the small intestinal crypts from enteric pathogens (4, 31), the selective expression of mCRAMP by surface epithelial cells in the colon could function to effectively protect against colon and epithelial cell colonization with an important class of epithelial adherent foodborne pathogen.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by National Institutes of Health Grants DK59860, DK35108, and AI56075 and by a grant from the Cystic Fibrosis Foundation.
2 Current address: Laboratory of Epithelial Immunobiology, RIKEN Research Center for Allergy and Immunology, Yokohama, Japan.
3 Address correspondence and reprint requests to Dr. Martin F. Kagnoff, Laboratory of Mucosal Immunology, Departments of Medicine and Pediatrics, Mail Code 0623D, University of California, San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0623. E-mail address: mkagnoff{at}ucsd.edu
4 Abbreviations used in this paper: mCRAMP, mouse cathelicidin-related antimicrobial peptide; EPEC, enteropathogenic Escherichia coli; EHEC, enterohemorrhagic E. coli.
Received for publication December 13, 2004. Accepted for publication February 11, 2005.
References
Murakami, M., T. Ohtake, R. A. Dorschner, R. L. Gallo. 2002. Cathelicidin antimicrobial peptides are expressed in salivary glands and saliva. J. Dent. Res. 81:845.
Zanetti, M.. 2004. Cathelicidins, multifunctional peptides of the innate immunity. J. Leukocyte Biol. 75:39.
Pestonjamasp, V. K., K. H. Huttner, R. L. Gallo. 2001. Processing site and gene structure for the murine antimicrobial peptide CRAMP. Peptides 22:1643
Ayabe, T., D. P. Satchell, C. L. Wilson, W. C. Parks, M. E. Selsted, A. J. Ouellette. 2000. Secretion of microbicidal -defensins by intestinal Paneth cells in response to bacteria. Nat. Immunol. 1:113
O’Neil, D. A., E. M. Porter, D. Elewaut, G. M. Anderson, L. Eckmann, T. Ganz, M. F. Kagnoff. 1999. Expression and regulation of the human -defensins hBD-1 and hBD-2 in intestinal epithelium. J. Immunol. 163:6718.
Hase, K., L. Eckmann, J. D. Leopard, N. Varki, M. F. Kagnoff. 2002. Cell differentiation is a key determinant of cathelicidin LL-37/human cationic antimicrobial protein 18 expression by human colon epithelium. Infect. Immun. 70:953.
Hase, K., M. Murakami, M. Iimura, S. P. Cole, Y. Horibe, T. Ohtake, M. Obonyo, R. L. Gallo, L. Eckmann, M. F. Kagnoff. 2003. Expression of LL-37 by human gastric epithelial cells as a potential host defense mechanism against Helicobacter pylori. Gastroenterology 125:1613.
Schauer, D. B., S. Falkow. 1993. Attaching and effacing locus of a Citrobacter freundii biotype that causes transmissible murine colonic hyperplasia. Infect. Immun. 61:2486
Schauer, D. B., S. Falkow. 1993. The eae gene of Citrobacter freundii biotype 4280 is necessary for colonization in transmissible murine colonic hyperplasia. Infect. Immun. 61:4654.
Frankel, G., A. D. Phillips, M. Novakova, H. Field, D. C. Candy, D. B. Schauer, G. Douce, G. Dougan. 1996. Intimin from enteropathogenic Escherichia coli restores murine virulence to a Citrobacter rodentium eaeA mutant: induction of an immunoglobulin A response to intimin and EspB. Infect. Immun. 64:5315.
Luperchio, S. A., J. V. Newman, C. A. Dangler, M. D. Schrenzel, D. J. Brenner, A. G. Steigerwalt, D. B. Schauer. 2000. Citrobacter rodentium, the causative agent of transmissible murine colonic hyperplasia, exhibits clonality: synonymy of C. rodentium and mouse-pathogenic Escherichia coli. J. Clin. Microbiol. 38:4343.
Luperchio, S. A., D. B. Schauer. 2001. Molecular pathogenesis of Citrobacter rodentium and transmissible murine colonic hyperplasia. Microb. Infect. 3:333.
Vallance, B. A., C. Chan, M. L. Robertson, B. B. Finlay. 2002. Enteropathogenic and enterohemorrhagic Escherichia coli infections: emerging themes in pathogenesis and prevention. Can. J. Gastroenterol. 16:771.
Deng, W., B. A. Vallance, Y. Li, J. L. Puente, B. B. Finlay. 2003. Citrobacter rodentium translocated intimin receptor (Tir) is an essential virulence factor needed for actin condensation, intestinal colonization and colonic hyperplasia in mice. Mol. Microbiol. 48:95.
Barthold, S. W., G. L. Coleman, P. N. Bhatt, G. W. Osbaldiston, A. M. Jonas. 1976. The etiology of transmissible murine colonic hyperplasia. Lab. Anim. Sci. 26:889.
Maggio-Price, L., K. L. Nicholson, K. M. Kline, T. Birkebak, I. Suzuki, D. L. Wilson, D. Schauer, P. J. Fink. 1998. Diminished reproduction, failure to thrive, and altered immunologic function in a colony of T-cell receptor transgenic mice: possible role of Citrobacter rodentium. Lab. Anim. Sci. 48:145
Barthold, S. W.. 1980. The microbiology of transmissible murine colonic hyperplasia. Lab. Anim. Sci. 30:167.
Johnson, E., S. W. Barthold. 1979. The ultrastructure of transmissible murine colonic hyperplasia. Am. J. Pathol. 97:291.[
Higgins, L. M., G. Frankel, G. Douce, G. Dougan, T. T. MacDonald. 1999. Citrobacter rodentium infection in mice elicits a mucosal Th1 cytokine response and lesions similar to those in murine inflammatory bowel disease. Infect. Immun. 67:3031.
Ghaem-Maghami, M., C. P. Simmons, S. Daniell, M. Pizza, D. Lewis, G. Frankel, G. Dougan. 2001. Intimin-specific immune responses prevent bacterial colonization by the attaching-effacing pathogen Citrobacter rodentium. Infect. Immun. 69:5597.
Maaser, C., M. P. Housley, M. Iimura, J. R. Smith, B. A. Vallance, B. B. Finlay, J. R. Schreiber, N. M. Varki, M. F. Kagnoff, L. Eckmann. 2004. Clearance of Citrobacter rodentium requires B cells but not secretory immunoglobulin A (IgA) or IgM antibodies. Infect. Immun. 72:3315
Gallo, R. L., K. J. Kim, M. Bernfield, C. A. Kozak, M. Zanetti, L. Merluzzi, R. Gennaro. 1997. Identification of CRAMP, a cathelin-related antimicrobial peptide expressed in the embryonic and adult mouse. J. Biol. Chem. 272:13088.
Dorschner, R. A., V. K. Pestonjamasp, S. Tamakuwala, T. Ohtake, J. Rudisill, V. Nizet, B. Agerberth, G. H. Gudmundsson, R. L. Gallo. 2001. Cutaneous injury induces the release of cathelicidin anti-microbial peptides active against group A Streptococcus. J. Invest. Dermatol. 117:91.
Iimura, M., T. Nakamura, S. Shinozaki, B. Iizuka, Y. Inoue, S. Suzuki, N. Hayashi. 2000. Bax is downregulated in inflamed colonic mucosa of ulcerative colitis. Gut 47:228.
Frohm, M., H. Gunne, A. C. Bergman, B. Agerberth, T. Bergman, A. Boman, S. Liden, H. Jornvall, H. G. Boman. 1996. Biochemical and antibacterial analysis of human wound and blister fluid. Eur. J. Biochem. 237:86
Simmons, C. P., N. S. Goncalves, M. Ghaem-Maghami, M. Bajaj-Elliott, S. Clare, B. Neves, G. Frankel, G. Dougan, T. T. MacDonald. 2002. Impaired resistance and enhanced pathology during infection with a noninvasive, attaching-effacing enteric bacterial pathogen, Citrobacter rodentium, in mice lacking IL-12 or IFN-. J. Immunol. 168:1804.
Ong, P. Y., T. Ohtake, C. Brandt, I. Strickland, M. Boguniewicz, T. Ganz, R. L. Gallo, D. Y. Leung. 2002. Endogenous antimicrobial peptides and skin infections in atopic dermatitis. N. Engl. J. Med. 347:1151.
Scott, M. G., D. J. Davidson, M. R. Gold, D. Bowdish, R. E. Hancock. 2002. The human antimicrobial peptide LL-37 is a multifunctional modulator of innate immune responses. J. Immunol. 169:3883.
Wilson, C. L., A. J. Ouellette, D. P. Satchell, T. Ayabe, Y. S. Lopez-Boado, J. L. Stratman, S. J. Hultgren, L. M. Matrisian, W. C. Parks. 1999. Regulation of intestinal -defensin activation by the metalloproteinase matrilysin in innate host defense. Science 286:113.(Mitsutoshi Iimura, Richar)