Entamoeba histolytica Induces Intestinal Cathelicidins but Is Resistant to Cathelicidin-Mediated Killing

Abstract

The enteric protozoan parasite Entamoeba histolytica is the cause of potentially fatal amebic colitis and liver abscesses. E. histolytica trophozoites colonize the colon, where they induce inflammation, penetrate the mucosa, and disrupt the host immune system. The early establishment of E. histolytica in the colon occurs in the presence of antimicrobial human (LL-37) and murine (CRAMP [cathelin-related antimicrobial peptide]) cathelicidins, essential components of the mammalian innate defense system in the intestine. Studying this early step in the pathogenesis of amebic colitis, we demonstrate that E. histolytica trophozoites or their released proteinases, including cysteine proteinase 1 (EhCP1), induce intestinal cathelicidins in human intestinal epithelial cell lines and in a mouse model of amebic colitis. Despite induction, E. histolytica trophozoites were found to be resistant to killing by these antimicrobial peptides, and LL-37 and CRAMP were rapidly cleaved by released amebic cysteine proteases. The cathelicidin fragments however, did maintain their antimicrobial activity against bacteria. Degradation of intestinal cathelicidins is a novel function of E. histolytica cysteine proteinases in the evasion of the innate immune system in the bowel. Thus, early intestinal epithelial colonization of invasive trophozoites involves a complex interplay in which the ultimate outcome of infection depends in part on the balance between degradation of cathelicidins by amebic released cysteine proteinases and upregulation of proinflammatory mediators which trigger the inflammatory response.

INTRODUCTION

The organism Entamoeba histolytica is a protozoan parasite that causes amebic colitis and liver abscesses through water- and food-borne infection. Approximately 10% of the world's population is infected with Entamoeba, and it is a major cause of death from parasitic infections (29). Colonization of the intestinal tract by E. histolytica follows binding of the amebic surface Gal/GalNAc adherence lectin to epithelial mucin oligosaccharides, with subsequent degradation of the mucin polymer network, extracellular matrix proteins, and components of the innate host defense by released cysteine proteinases (17, 20, 21, 27, 28). This early establishment of E. histolytica triggers an inflammatory response, which plays a role in the ultimate outcome of infection (4, 13).

Cathelicidins are small cationic antimicrobial peptides of the mammalian innate immune system with broad activity against bacteria (6, 10, 11, 15) and protozoa (7, 9, 19). LL-37 is the only cathelicidin described in humans (8) and CRAMP (cathelin-related antimicrobial peptide) is the cathelicidin found in mice (6). Both LL-37 and CRAMP have related structure, function, and distribution in epithelial cells, including the intestine of humans and mice, and are part of the defense against microbial epithelial infections (32). For example, expression of LL-37 mRNA and protein was increased by Helicobacter pylori in gastric epithelial cells (11), and CRAMP protected mice from colonic colonization with Citrobacter rodentium (15) and cutaneous infection with group A Streptococcus (25). On the other hand, virulent strains of Streptococcus pyogenes and Shigella spp. inactivated or downregulated LL-37 expression (5, 16). The role of innate cathelicidins in the defense from intestinal parasitic infections such as amebiasis is unknown.

To explore the role of intestinal antimicrobial peptides as part of the innate defense against amebiasis, we investigated the interactions of human (LL-37) and murine (CRAMP) cathelicidins and E. histolytica trophozoites in vitro and in vivo. This study shows that trophozoites or their released proteinases upregulate synthesis of cathelicidin mRNA and protein in human intestinal epithelial cell lines and in mice. Native cysteine proteinases and recombinant cysteine proteinase 1 of E. histolytica (rEhCP1) degrade LL-37 and CRAMP, although the fragments maintain their antimicrobial activity against bacteria. In contrast, E. histolytica trophozoites are resistant to killing by both intact and cleaved antimicrobial cathelicidins.

MATERIALS AND METHODS

E. histolytica trophozoites and released proteinases.

E. histolytica trophozoites (strain HM1: IMSS) were grown axenically at 37°C in trypsin-yeast-iron medium supplemented with Diamond vitamins and 15% adult bovine serum. Amebic released proteinases were isolated from trophozoites of E. histolytica (2 × 10⁶/ml) in the mid-logarithmic growth phase in Dulbecco's phosphate-buffered saline (PBS) (Invitrogen, Grand Island, NY) with HEPES (10 mM), l-cysteine (0.1%), and ascorbic acid (0.02%) (pH 7.2), which maintained >95% viability (by trypan blue exclusion) as previously described (26).

Recombinant cysteine proteinases and activity assays.

Recombinant cysteine proteinase 1 of E. histolytica (rEhCP1) was expressed in Escherichia coli as a thioredoxin fusion protein (amino terminus) with a six-residue histidine tail (carboxy terminus) and refolded to an active enzyme with a pH optimum of 6.0 as previously described (20). The proteinase activity of E. histolytica trophozoites, released proteinases, and rEhCP1 was determined by measuring the release of the fluorescent leaving group, 4-amino-7-methylcoumarin (AMC), from the synthetic peptide substrate Z-Arg-Arg-AMC (Bachem) (pH 6.0) in a Fluoroskan-Ascent fluorometer (Labsystems) and expressed as relative fluorescent units (RFU) (20). Reaction specificity was determined by preincubating proteinases with the vinyl sulfone cysteine proteinase inhibitor, WRR483 (20 μM), for 25 min at room temperature (RT) (20).

Coculture of human colonic epithelial cells and E. histolytica.

HT-29 human colonic epithelial adenocarcinoma cells (American Type Culture Collection; HTB-38) were maintained in Dulbecco's modified Eagle's medium (DMEM) (Cellgro Mediatech) with 10% fetal bovine serum (Benchmark Gemini) and penicillin (100 U/ml) and streptomycin (100 μg/ml) (HyClone Thermo) in a humidified environment of 95% air and 5% CO₂. Cells were grown in 6-well tissue culture plates (Costar) until nearly confluent and then differentiated with n-butyrate (2 mM) for 3 days, which results in upregulation of LL-37 mRNA and protein (10). Differentiated cells were incubated with DMEM alone or with E. histolytica trophozoites (4 × 10⁵/well) for up to 120 min, or rEhCP1 or released proteases from the same number of trophozoites (4 × 10⁵/well) for 30 to 45 min. The experiments were repeated three times.

E. histolytica infection of mice.

Male C3H/HeJ mice (6 weeks old from The Jackson Laboratory) were maintained under specific-pathogen-free conditions. Mice were pretreated with dexamethasone (10 mg/kg) given intraperitoneally daily for 4 days prior to surgery. The cecum was externalized under general anesthesia and injected with 2 × 10⁶ cecum-passed E. histolytica trophozoites. Pretreatment with dexamethasone for 4 days increases the infection rate with E. histolytica to 97% at 7 days postchallenge (12); otherwise, only 60% of untreated mice become infected. Likewise, partial resistance of C3H/HeJ mice to amebic infection was abrogated with dexamethasone through an innate, lymphocyte-independent mechanism (1). Groups of 4 or 5 mice were sacrificed at 1, 3, and 7 days after infection, and the entire cecum was immediately frozen in liquid nitrogen. Frozen cecum tissue was later homogenized in liquid nitrogen for quantitative PCR (qPCR) studies or cut in cryo-sections (7 μm) for confocal microscopy. Cecal tissues were also fixed in paraformaldehyde, embedded in paraffin, cut (5 μm), mounted on slides, and stained with hematoxylin and eosin (H&E) for histological examination. Disease severity in the cecum after infection with E. histolytica for 1, 3, and 7 days was quantified by histopathological scoring (scale, 0 to 5) for each section and averaged for each mouse as previously described (14). All animal studies were reviewed and approved by the University of California, San Diego, Institutional Animal Care and Use Committee.

Quantification of cathelicidin mRNA.

After incubation with E. histolytica (trophozoites, released proteases, or rEhCP1), HT-29 cells were washed with PBS, detached with 0.05% trypsin-EDTA (Gibco) for 10 min at 37°C, and pelleted by centrifugation (5 min, 400 × g, 4°C). Total RNA was purified using Nucleospin columns (Nucleospin RNA II; Macherey-Nagel), and 5 μg of total RNA was reverse transcribed to cDNA (Applied Biosystem and Turbo DNA-free Ambion). Real-time qPCR was performed using a StepOnePlus real-time PCR system (Applied Biosystem). Each reaction mixture contained 2 μl of cDNA, 2× SYBR Mesa green (Eurogentec), and primers (200 nM) for LL-37 (sense, 5′-ACA CAG CAG TCA CCA GAG GAT-3′; antisense, 5′-GAA GAA ATC ACC CAG CAG GGC-3′), human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (sense, 5′-CAT GTT CGT CAT GGG TGT GAA-3′; antisense, 5′-AGT GAT GGC ATG GAC TGT GGT-3′), CRAMP (sense, 5′-CTT CAA CCA GCA GTC CCT AGA CA-3′; antisense, 5′-TCC AGG TCC AGG AGA CGG TA-3′), murine GAPDH (sense, 5′-TGT GAT GGG TGT GAA CCA CGA GAA-3′; antisense, 5′-AGT GAT GGC ATG GAC TGT GGT CAT-3′), and murine villin, an actin binding protein expressed mainly in the brush border of the epithelium (sense, 5′-AGG TGG TGG CTG CCT CTT CCA G-3′; antisense, 5′-CGG GAG TGG TGA TGT TGA GAG AGC C-3′). The reaction mixtures were incubated at 50°C for 2 min followed by 95°C for 10 min and amplified by 15 s of denaturation at 95°C followed by 1 min of annealing at 66°C for a total of 40 cycles. Values of human LL-37 mRNA were corrected relative to the housekeeping gene coding for human GAPDH. Values of murine CRAMP mRNA were corrected relative to the murine housekeeping genes coding for GAPDH or villin. Data were expressed as fold changes (mean ± standard error [SE]) of cathelicidin mRNA between nonstimulated and stimulated cells or animals (10) (StepOne Software v2.1; Applied Biosystem).

Immunoblots.

Confluent HT-29 cells incubated with E. histolytica were washed with PBS and lysed for 5 min at room temperature with 100 μl/well of cell lysis buffer (10× cell signaling buffer with 1% sodium dodecyl sulfate, 100 μM E-64 [Roche Applied Science], 1 mM phenylmethylsulfonyl fluoride [Sigma], 25× O-Complete [Roche Applied Science], and 4 mM Pefabloc SC [Roche Applied Science]. The lysate was sonicated (3× for 2 s each), and the supernatant was collected after centrifugation (10 min, 13,000 × g, 4°C). Protein concentration was measured by microbicinchinonic acid (micro-BCA) protein assay (Thermo Fisher Scientific).

Proteins from HT-29 cells (30 μg/lane) or homogenized murine cecal mucosa (100 μg/lane) were size separated by SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membranes (0.45-μm pore) (Immobilon-P; Millipore). Membranes were blocked (120 min, room temperature) with 5% nonfat dry milk in TBS-T (150 mM NaCl, 10 mM Tris base, 0.05% Tween 20 [pH 7.4]) and probed (16 h, 4°C) with polyclonal rabbit antibodies to LL-37 (1:10,000 in TBS-T) (8) or CRAMP (1:2,500 in TBS-T) (6) and monoclonal mouse anti-β-actin (Sigma) (1:10,000 in TBS-T). After washing with TBS-T, membranes were incubated (60 min, room temperature) with horseradish peroxidase (HRP)-conjugated anti-rabbit antibodies (HRP-goat anti-rabbit IgG [H + L] (ZyMax, Zymed Laboratories) (1:10,000 in TBS-T) or anti-mouse antibodies (HRP-horse anti-mouse IgG (H&L) (Cell Signaling Technology) (1:10,000 in TBS-T). Immune-reactive bands were detected by enhanced chemiluminescence (Supersignal West Femto, Thermo Fisher Scientific) and exposure to X-ray film (5 to 45 s) (Eastman Kodak). Protein bands were quantified using Image Quant 5.0 (Molecular Dynamics).

Cellular localization of cathelicidins.

For intracellular detection, HT-29 cells were grown on chamber slides (Lab-Tek II, Nalgene, Nunc), differentiated with butyrate, and incubated with E. histolytica trophozoites (4 × 10⁵ in 200 μl) or medium alone for 60 min at 37°C as described above. Cryo-sections were obtained from the cecal mucosa of mice infected with E. histolytica for 3 days. Samples were washed with PBS, fixed (15 min, room temperature) with 4% paraformaldehyde, permeabilized with PBS plus 0.25% Triton X-100 (PBS-T) (10 min at room temperature), and blocked (60 min, room temperature) with PBS-T plus 10% goat serum. Avidin and biotin were applied as per the manufacturer's instructions (Vector). After further washing with PBS, HT-29 cells were incubated (16 h, 4°C) with polyclonal rabbit antibodies to LL-37 (8) and mouse cecal mucosa with polyclonal rabbit antibodies to CRAMP (6) and mouse IgM to CD15 for neutrophils (Abcam) (1:750 in PBS-T plus 1% bovine serum albumin [BSA]). Following washes with PBS-T, the cells were incubated (60 min, room temperature) with biotinylated goat anti-rabbit IgG (Vector) (1:250 in PBS-T). After additional washes with PBS-T, cells were incubated (60 min, room temperature) with Alexa Fluor 488-conjugated goat anti-mouse IgM (1:250 in PBS-T) (Invitrogen), Alexa Fluor 555-conjugated streptavidin (1:500 in PBS-T) (Invitrogen), and To-Pro-3 (1:750) (Invitrogen) antibodies, washed with PBS-T, mounted with Prolong Gold antifade reagent with DAPI (4-,6-diamidino-2-phenylindole) (Invitrogen), and analyzed using a confocal laser scanning microscope (True Confocal Scanner SP5; Leica).

Cleavage of cathelicidins by amebic cysteine proteinases.

Synthetic LL-37 (LLGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTES) and CRAMP (GLLRKGGEKIGEKLKKIGQKIKNFFQKLVPQPEQ) peptides (50 μM) (Bachem) were incubated with rEhCP1 (50 and 200 RFU) for 60 min at 37°C in a buffer consisting of 0.4 M citric acid, 0.4 M Na₂HPO₄, 0.5 mM EDTA, and 1 mM dithiothreitol (DTT) (pH 6.0). The rEhCP1 was incubated alone or pretreated with either of the cysteine proteinase inhibitors WRR483 (20 μM) or E-64 (20 μM) for 20 min at room temperature. The products were identified by 20% SDS-PAGE followed by Western blotting with anti-LL37 and anti-CRAMP antibodies as described above.

Killing of E. histolytica by cathelicidins.

Trophozoites of E. histolytica (10⁵/well), alone or preincubated with either WRR483 (20 μM) or E-64 (20 μM) for 30 min at room temperature, were incubated in triplicate with 5 to 50 μM synthetic LL-37 or CRAMP in 96-well plates (Maxisorp surface; Nunc) for 24 h in an anaerobic environment (Mitsubishi Gas Chemical) (75 μl/well). These concentrations of LL-37 and CRAMP encompass the typical range of cathelicidin killing of bacteria of 8 to 32 μM (6, 11, 15) and reflect physiological levels, since LL-37 concentrations in body fluids, including saliva and ascitic fluid, oscillate between 0 to 6 μM (3). The number of viable amoebae was determined by quantification of total cellular ATP. Briefly, lysis buffer (BacTiter-Glo substrate; Promega) was added (75 μl/well), the mixture was incubated (5 min, room temperature), and luminescence was recorded in a SpectraMax M-5 microplate reader. Measurement of ATP from a standard curve with known number of trophozoites was used to calculate the number of amoebae. The number of viable trophozoites was corroborated by counting in a hematocytometer (by trypan blue exclusion).

Antimicrobial activity of cathelicidins.

E. coli cells (ATCC 25922) were washed with PBS and diluted to 5 × 10⁵ CFU/ml in 20% SOB (2% tryptone, 0.05% NaCl, 0.5% yeast extract, and 1% MgCl₂ plus 50 mM Na₂HCO₃ and 1 mM NaH₂PO₄ [pH 7.4]). The bacterial suspension was incubated with synthetic LL-37 (100 μM) and CRAMP peptides (32 μM) alone or preincubated with rEhCP1 (1.5 μM) for 0, 45, and 90 min at 37°C. Dilutions were plated directly onto a LB agar plate, and numbers of surviving CFU were determined after incubation overnight at 37°C.

Statistical analysis.

Statistical analysis of differences between groups was done by unpaired t test or analysis of variance (ANOVA) and Bonferroni posttests (GraphPad Prism 4.0). Graphs represent two to three independent experiments, and error bars represent means ± standard error. Differences in values were considered significant when P was <0.05.

RESULTS

E. histolytica induces intestinal cathelicidin expression.

We first investigated whether expression of colonic epithelial cathelicidins is upregulated by E. histolytica. For in vivo studies, we used a murine model of amebic colitis that shows significant histopathological alterations of the cecal mucosa after infection with E. histolytica. Microscopically, submucosal infiltration of uniformly distributed or aggregated lymphocytes and plasma cells was observed in the cecum 1 day postinfection (inflammation score of 3.1 ± 0.6 in a 0-to-5 scale) (14) (Fig. 1A). Epithelial ulcerations with numerous amoebae were observed at 3 and 7 days that involved up to 20% of the luminal circumference and extended into the crypts (3 ± 0.5 in a 0-to-5 scale) (Fig. 1B, C, and D). In addition, luminal neutrophils were detected and crypt hyperplasia was found in nonulcerated areas of the cecum. The mouse cathelicidin, CRAMP, was significantly induced by E. histolytica infection in this model of amebic colitis. Levels of cecal CRAMP mRNA increased >4-fold by 3 days and >100-fold at 7 days (Fig. 2A) (P < 0.01). Increased intestinal expression of CRAMP was also detected by immunoblotting of cecal mucosal extracts of E. histolytica-infected mice after 1 and 7 days, as evidence by CRAMP bands in the full-length (18-kDa) and C-terminal (10-kDa) forms of the peptide (Fig. 2B). Similar levels of β-actin, used as a loading control, were found in both infected and uninfected mice (Fig. 2B). Immunostaining revealed that increased CRAMP protein expression after E. histolytica infection mostly occurred in cecal epithelial cells rather than in CD15⁺ neutrophils (Fig. 1E and F).

FIG 1.

E. histolytica induces colonic epithelial cathelicidins. (A to F) In a mouse model of amebic colitis, cecum infected with E. histolytica showed submucosal infiltration with lymphocytes and plasma cells (arrowhead) at 1 day after infection (A) and submucosal inflammation plus ulcerated epithelium (arrowhead) with the presence of numerous trophozoites (arrows) at 3 and 7 days after infection (B and C) compared with controls (D). Murine CRAMP protein was mostly localized in epithelial cells (red; arrow) and not in neutrophils (green; arrowhead) in the cecum from mice infected with E. histolytica for 3 days (E). A few neutrophils were found in the cecum of control mice (F). (G and H) In human colonic HT-29 epithelial cells, LL-37 protein (green; arrow) was increased after incubation with E. histolytica trophozoites (G) but not with medium alone (H). Nuclei are stained by To-Pro-3 (blue).

FIG 2.

E. histolytica induces expression of CRAMP in cecum of mice. (A) Ceca of mice intracecally infected with E. histolytica (Infected) or inoculated with medium alone (Uninfected [control]) for 1, 3, and 7 days were homogenized, and total RNA was extracted and reverse transcribed to cDNA. CRAMP mRNA was quantified by reverse transcription-PCR (RT-PCR), and values were corrected relative to villin and are expressed as fold change (log₁₀) from expression in control mice of the same strain. *, P < 0.05 compared with control group; **, P < 0.05 among days of infection. (B) CRAMP was detected by immunoblots in cecal mucosa of mice uninfected or infected with E. histolytica for 1 and 7 days. *, full-length cathelicidin protein (18 kDa); **, C-terminal cathelicidin peptide (5 kDa).

In support of the findings in the mouse infection model, the human cathelicidin, LL-37, was also upregulated by E. histolytica of human colon epithelial cell cultures. Thus, coincubation of human HT-29 cells with trophozoites (4 × 10⁵/well) increased levels of LL-37 mRNA after 30 min of incubation by 1.5- to 2.5-fold. In parallel, LL-37 protein expression increased markedly compared with that of noninfected cells (Fig. 1G and H). Both LL-37 and CRAMP proteins were localized primarily in the cytoplasm of the epithelial cells, although some immunoreactive protein was detected outside cells. Active released proteinases from the same number of E. histolytica trophozoites (4 × 10⁵/well) and rEhCP1 alone (9 to 36 ng/μl) induced a similar increase of LL-37 mRNA in HT-29 cells within 30 to 45 min (2- to 3.3-fold). Colonic HT-29 cells exposed to intact trophozoites and released proteinases of E. histolytica remained viable during the 2-h experiment. These results suggest that E. histolytica upregulates production of intestinal epithelial cathelicidins at both the mRNA and protein levels.

E. histolytica trophozoites are resistant to LL-37-mediated killing.

Since cathelicidins are induced by E. histolytica and have broad antimicrobial activity against bacteria (6, 10, 11, 15) and parasites (7, 9, 19), we investigated if LL-37 or CRAMP had activity against E. histolytica. Surprisingly, we found LL-37 or CRAMP did not kill E. histolytica trophozoites (10⁶/ml) at concentrations as high as 50 μM and incubation times as long as 24 h, as measured with an ATP-based assay for amebic viability (Fig. 3) (P > 0.05). Likewise, lower numbers of E. histolytica trophozoites (10⁴/ml) were not affected by LL-37 or CRAMP (up to 50 μM) (data not shown). Similar results were found by direct counts of trophozoites with trypan blue viability staining.

FIG 3.

Human LL-37 and murine CRAMP do not kill E. histolytica. Trophozoites (10⁶/ml), either alone (Control) or preincubated with cysteine proteinase inhibitor WRR483 (20 μM) or E-64 (20 μM), were incubated with 5 to 50 μM LL-37 (A) or CRAMP (B) for 24 h. The numbers of viable trophozoites were quantified by measuring the levels of cellular ATP and are expressed as percentage relative to control cells.

Released cysteine proteases of E. histolytica degrade LL-37 and CRAMP.

Based on our observation that LL-37 did not kill amebic trophozoites, we tested the hypothesis that LL-37 was inactivated by released E. histolytica cysteine proteinases, which can cleave multiple host proteins, including cellular matrix proteins, MUC2 mucin, complement, immunoglobulins, and anaphylatoxins (17, 20, 21, 27, 28). We found that the full-length LL-37 protein (18 kDa) derived from human colonic cells was partially degraded after 30 min of incubation with E. histolytica trophozoites (30% degraded), while the C-terminal peptide of LL-37 (5 kDa) required 2 h to detect a significant effect (44% degraded) (Fig. 4A). Likewise, ameba-released proteases degraded LL-37 from human HT-29 cells after 60 min of incubation, and this degradation was suppressed by the cysteine proteinase inhibitor WRR483 (20) (Fig. 4B). Synthetic LL-37 and CRAMP were also digested by a single major released cysteine proteinase, recombinant cysteine proteinase rEhCP1 (20), in a concentration-dependent manner, which was inhibited by both cysteine proteinase inhibitors WRR483 and E-64 (Fig. 4C). To determine if the resistance of E. histolytica to LL-37 was due to degradation of cathelicidins by amebic proteinases, trophozoites were preincubated with WRR483 or E-64 before exposure to LL-37 or CRAMP for 24 h. Preincubation of trophozoites with WRR483 or E-64 slowed amebic growth moderately, but this effect was independent of the cathelicidins (Fig. 3). These results indicate E. histolytica trophozoites are naturally resistant to cathelicidins whether they are proteolytically cleaved or not.

FIG 4.

Cathelicidins are cleaved by E. histolytica and their released amebic cysteine proteases. Confluent human colonic epithelial HT-29 cells were incubated with E. histolytica trophozoites (4 × 10⁵/well) (A) or their released proteases (B), either alone or preincubated with inhibitor WRR483 (20 μM). (C) Synthetic LL-37 and CRAMP (50 μM) were incubated with rEhCP1 (50 and 200 RFU), either alone or preincubated with cysteine proteinase inhibitors E-64 (20 μM) and WRR483 (20 μM). Products were separated by SDS-PAGE, transferred to PVDF membranes, and stained with rabbit anti-LL-37 or anti-CRAMP antibodies. *, full-length cathelicidin protein (18 kDa); **, C-terminal cathelicidin peptide (5 kDa).

LL-37 and CRAMP have antibacterial activity even after cleavage by amebic cysteine proteinases.

We next investigated if cathelicidins maintain antibacterial activity after cleavage by amebic proteinases. The antibacterial activity of cathelicidins after cleavage by active rEhCP1 was investigated in a bioassay using E. coli. Both LL-37 and CRAMP, with or without rEhCP1, killed E. coli by 90 min (Fig. 5A and B) (P < 0.05). The growth rate of E. coli was decreased by rEhCP1 alone by 90 min (P < 0.05) but still remained significantly higher than in E. coli cells treated with cathelicidins (P < 0.05). These results suggest that cathelicidins retain antimicrobial activity against bacteria after cleavage by amebic cysteine proteases.

FIG 5.

Cathelicidins LL-37 and CRAMP kill E. coli even after incubation with cysteine proteinases of E. histolytica. E. coli was incubated with LL-37 (A) and CRAMP (B), either alone or preincubated with rEhCP1 for 24 h, and CFU were determined. *, P < 0.05 compared with groups of E. coli cells exposed to either LL-37 or CRAMP.

DISCUSSION

Cathelicidins are naturally occurring peptides with broad antimicrobial activity (6, 7, 9–11, 15, 19), although the role of cathelicidins produced by intestinal epithelium in defense against E. histolytica was not known. We show here that E. histolytica upregulates intestinal murine cathelicidin, CRAMP, in a murine model of amebic colitis. Likewise, whole E. histolytica trophozoites and released proteinases induce human intestinal cathelicidin LL-37 at the mRNA and protein levels. Furthermore, a single cysteine proteinase of E. histolytica, rEhCP1, was sufficient to induce human intestinal cathelicidin LL-37. EhCP1 is biologically significant as it is representative of four major released amebic cysteine proteinases (EhCP1, -2, -3, and -5) with similar substrate specificities for arginine in the P₂ position and is specifically targeted by cysteine inhibitor WRR483 (20). It is possible that other cysteine proteinases of E. histolytica such as EhCP5, which induces an NF-κB-mediated proinflammatory response in colonic cells (13), may also be involved in cathelicidin induction. Increase of LL-37 mRNA expression (3- to 4-fold) was also reported in the human colonic adenocarcinoma cell line (HCA-7) coincubated with Salmonella enterica serovar Dublin or enteroinvasive Escherichia coli O29 for 60 min (10). The observed early upregulation of colonic LL-37 mRNA by E. histolytica after 30 to 45 min of coincubation is in line with a previous report in which E. histolytica induced interleukin-8 (IL-8) mRNA in HT-29 cells within 1 h, with a maximum at 2 h (4). The less pronounced expression of cathelicidins in vitro by HT-29 cells cocultured with E. histolytica compared with in vivo murine infection suggests that cathelicidins may be costimulated by factors present in the host, such as 1,25-dihydroxyvitamin D (30).

E. histolytica-induced increased mRNA expression was accompanied by increased production of cathelicidin proteins, primarily localized in the cytoplasm of intestinal epithelial cells rather than in neutrophils. This corresponds to the natural location of mature LL-37 on the surface and upper crypts of human colonic epithelium (10) and CRAMP in the murine colonic epithelium (15). We also detected LL-37 protein outside cells, suggesting secretion by colonic epithelial cells. This is consistent with a previous report in which precursor protein and LL-37 peptide were detected in both gastric epithelial cells and secreted gastric fluids of patients infected with H. pylori (11). Thus, intestinal epithelium-derived cathelicidins are induced by E. histolytica at both the mRNA and protein levels as part of the innate immune response to amebic infection.

The observed resistance of E. histolytica to cathelicidins contrasts with their antimicrobial activity against a broad spectrum of other pathogens. Human LL-37 and mouse CRAMP effectively kill multiple bacteria, including Pseudomonas aeruginosa, Citrobacter rodentium, Staphylococcus aureus, Streptococcus pyogenes, Salmonella spp., Escherichia coli, and Helicobacter pylori (6, 10, 11, 15). Much less is known about the effect of antimicrobial peptides on parasites, although cathelicidins have been increasingly explored as antiparasitic compounds. Cathelicidins from other animal species, including SMAP-29 BMAPs, Bac7, and PG-1, have shown some antiparasitic activity against Trypanosoma brucei, Leishmania donovani, and Cryptosporidium parvum (7, 9, 19). Since cathelicidin killing relies on disruption of the integrity of microbial membranes (24), the composition of the surface membrane of E. histolytica likely determines this innate resistance to cathelicidins. The membrane of E. histolytica trophozoites contains galactose/N-acetyl galactosamine-inhibitable lectin (Gal/GalNac-lectin), lipophosphoglycan (LPG), and lipophosphopeptidogyclan (LPPG) (22), which participate in adhesion and induction of a proinflammatory response but may also protect from the lytic action of proteins such as cathelicidins or complement components (2, 23). Another strategy of resistance could involve the highly charged LPPG glycoconjugates of E. histolytica (22), which might neutralize cationic defensins by altering the cell surface charge, like the anionic capsule polysaccharides of Pseudomonas aeruginosa and Escherichia coli (18). The exact mechanism of resistance remains to be determined in future studies.

We demonstrated that cathelicidins are proteolytically cleaved by E. histolytica. Trophozoites, released proteases, or rEhCP1 of E. histolytica rapidly degraded the full-length (18 kDa) and the C-terminal portion (5 kDa) of LL-37 in human colonic cells. Both LL-37 and CRAMP still killed E. coli even after cleavage by amebic cysteine proteinases. This suggests that proteolytic processing of cathelicidins at epithelial surfaces by E. histolytica may generate smaller peptides with antimicrobial activity. Host proteinases such as kallikreins of the skin processed LL-37 to multiple biologically active smaller peptides by cleaving its C-terminal portion preferentially at arginine (31), the preferred peptide in the P₂ position of EhCP1 (20).

In conclusion, we show for the first time that E. histolytica induces intestinal cathelicidins, although trophozoites are resistant to killing by these antimicrobial peptides. Moreover, cathelicidins are rapidly cleaved by released amebic cysteine proteases. These studies emphasize the complex interplay between invasive trophozoites and their released cysteine proteinases in the early colonization of the intestinal epithelium. The ultimate outcome of infection depends on the balance between degradation of host innate defenses, such as cathelicidins, and upregulation of proinflammatory mediators, which trigger the inflammatory response.