Suppression of NF-κB Activation by Entamoeba histolytica in Intestinal Epithelial Cells Is Mediated by Heat Shock Protein 27*

Little is known about the pathogenesis of Entamoeba histolytica and how epithelial cells respond to the parasite. Herein, we characterized the interactions between E. histolytica and colonic epithelial cells and the role macrophages play in modulating epithelial cell responses. The human colonic epithelial cell lines Caco-2 and T84 were grown either as monoculture or co-cultured in transwell plates with differentiated human THP-1 macrophages for 24 h before stimulation with soluble amebic proteins (SAP). In naive epithelial cells, prolonged stimulation with SAP reduced the levels of heat shock protein (Hsp) 27 and 72. However in THP-1 conditioned intestinal epithelial cells SAP enhanced Hsp27 and Hsp72, which was dependent on the activation of ERK MAP kinase. Hsp synthesis induced by SAP conferred protection against oxidative and apoptotic injuries. Treatment with SAP inhibited NF-κB activation induced by interleukin-1β; specifically, the NF-κB-DNA binding, nuclear translocation of p65 subunit, and phosphorylation of IκB-α were reduced. Gene silencing by small interfering RNA confirmed the role of Hsp27 in suppressing NF-κB activation at IκB kinase (IKK) level. By co-immunoprecipitation studies, we found that Hsp27 interacts with IKK-α and IKK-β, and this association was increased in SAP-treated conditioned epithelial cells. Overexpression of wild type Hsp27 amplified the effects of SAP, whereas a phosphorylation-deficient mutant of Hsp27 abrogated SAP-induced NF-κB inhibition. In conditioned epithelial cells, Hsp27 was phosphorylated at serine 15 after prolonged exposure to SAP. This mechanism may explain the absence of colonic inflammation seen in the majority of individuals infected with E. histolytica.

cell-specific response to amebic infection is poorly understood. Moreover, the majority of research done to unravel the mechanism of amebic colitis has been focused on proinflammatory responses (2, 3) by epithelial cells and a role of NF-B (4) and chemokines such as IL-8 3 (5) as triggering events for inflammation. However, it is noteworthy that only 10% of E. histolytica-infected individuals show symptoms of intestinal inflammation (6), and none of the studies addressed the question of why only a minority of infected individuals develops amebic colitis.
Epithelial cells are the first layer of host defense and have been shown to be the effector cells capable of secreting several mediators in response to pathogens (7,8). Epithelial cells in vivo do not respond in isolation but act in concert with several immune and nonimmune cells present in the lamina propria. To simulate this, several in vitro studies were carried out to assess the epithelial cell responses in the presence of immune cells such as leukocytes and lymphocytes. Under these conditions, a differential response was observed in epithelial cell lines exposed to various pathogenic and nonpathogenic bacteria (9,10). However, the effect of co-culturing with macrophages on epithelial cell responses has not been well characterized.
The universal response to stress has been the induction of a group of highly conserved family of proteins called heat shock proteins (Hsp) and is commonly referred to as heat shock response or stress response. Several pathogens and their products have been shown to induce various Hsp in different cell types including intestinal epithelial cells (11)(12)(13). Hsp serve to protect the cells against several insults such as thermal, toxic, or apoptotic stimuli (12, 14 -16). Epithelial cell induction of Hsp in response to various pathogens such as E. coli, and toxins such as lipopolysaccharide and superantigen have been reported (10,12,14). It has also been shown that epithelial Hsp expression is regulated by cytokines and immune cells such as lymphocytes (13). One emerging concept is that stress response counters the inflammatory response mediated by NF-B, helping to reduce inflammation and promote healing of damaged tissues (12,(17)(18)(19)(20). Because proinflammatory cytokines secreted by immune * This work was supported by grants from the Canadian Institutes of Health Research and the Natural Sciences and Engineering Research Council of Canada. 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  cells have been shown to influence epithelial cell responses (16,21) and macrophages are a major source of these cytokines, we studied the effect of macrophages on epithelial cell response toward E. histolytica. We hypothesized that amebae might be eliciting a protective response whereby inflammation is suppressed in the majority of infected individuals. Thus, we sought to study the stress response induced by ameba components in naïve and macrophage-conditioned colonic epithelial cells and made several interesting and novel observations. For the first time, we showed that macrophage conditioning primes epithelial cells for an augmented Hsp expression in response to amebic components. We identified Hsp27 as the key mediator suppressing NF-B activation by virtue of its association with IB kinase (IKK) complex in intestinal epithelial cells (IEC). We conclude that this could be one of the mechanisms by which colonic inflammation is suppressed in the majority of E. histolyticainfected individuals and that the lack of such protective responses in a susceptible individual could lead to the symptoms associated with amebic colitis.
Co-culturing of Epithelial Cells with Macrophages-T84 or Caco-2 cells between 10 and 30 passages grown in either regular or transwell plates for 7-10 days to achieve confluency were used. For siRNA experiments, subconfluent (40 -50%) Caco-2 cells grown for 2-3 days were used. Human monocyte-like cell line THP-1 was differentiated with phorbol 12-myristate 13-acetate for 3 days and quiescence for 24 h before use. Transwells with epithelial cells were kept in culture plates containing 2 ϫ 10 6 macrophages for 24 h (30h for siRNA studies), removed from co-culture, and immediately used for experiments. Epithelial cells were kept under low serum condition (5% fetal bovine serum) during co-culturing and subsequent treatments and without antibiotics for siRNA studies.
Western Blot-Cellular extracts from amebic protein-treated epithelial cells were prepared by scraping into the sample buffer containing SDS and mercaptoethanol and boiled for 10 min, and equal volumes were separated in 12% SDS-polyacrylamide gels and transferred onto nitrocellulose membrane (Bio-Rad). For nonreducing gel run, sample buffer without mercaptoethanol was used. The membranes were blocked in 3.5% skim milk-TBS-T (20 mM Tris-HCl, pH 7.5, 500 mM NaCl. 0.1% Tween 20) at 4°C overnight, incubated with primary antibodies in 1% skim milk-TBS-T at 4°C overnight, washed three times with TBS-T, and incubated with horseradish peroxidase-conjugated secondary antibody in skim milk-TBST overnight at 4°C. After three washes each in TBS-T and TBS-3T, the blot was developed with the enhanced chemiluminescence system (Amersham Biosciences) according to the manufacturer's instructions.
Electrophoretic Mobility Shift Assay-Nuclear extracts were collected using the NE-PER kit (Pierce) and protein quantified by BCA assay. Annealed double-stranded heat shock element oligonucleotide Ϫ107 to Ϫ83 of the human Hsp70 gene (5Ј-GAT CTC GGC TGG AAT ATT CCC GAC CTG GCA GCC GA-3Ј) (Sigma Genosys) or NF-B consensus oligonucleotide (Santa Cruz Biotechnology) was labeled with [ 32 P]ATP, using T4 polynucleotide kinase (Invitrogen). Unlabeled nucleotides were removed using Sephadex G-25 columns. The binding reaction consists of a 20-liter total volume of 0.5 ng of DNA probe, 5 g of nuclear extract, 1 g of poly(dI-dC) in the binding buffer (12 mM HEPES, 60 mM KCl, 4 mM MgCl 2 , 1 mM EDTA, 1 mM of dithiothreitol, and 12% glycerol, pH 8.0), and incubation for 30 min at room temperature. DNA-protein complexes were resolved by electrophoresis on 6% polyacrylamide gels at 4°C in TBE buffer (90 mM Tris borate, 2 mM EDTA). The gels were subsequently dried and autoradiographed with intensifying screens at Ϫ70°C.
Neutral Red Assay-Neutral red (Sigma) was reconstituted in serum-free medium and added to cells at 1.14 mM concentration. After 2 h of incubation, the medium was removed, and the cells were washed twice with phosphate-buffered saline; finally, the incorporated neutral red was released from the cells by incubation for 15 min at room temperature in the presence of 2 ml of the extraction buffer containing acetic acid (1%, v/v) and ethanol (50%, v/v). To measure the dye taken up, the cell lysis products were centrifuged, and the supernatants were spectrophotometrically measured at 540 nm.
Co-immunoprecipitation-Cells were lysed in 1% CHAPS buffer (10 mM Tris-HCl, pH 7.4, 150 mM NaCl with 1% CHAPS) with protease inhibitor mixture (Roche Applied Science). 500 g of cell extracts were incubated overnight with 5 g of agarose-conjugated anti-Hsp antibodies (Santa Cruz Biotechnology). Precipitate was washed thrice with lysis buffer, dissolved in 2ϫ Laemmli buffer, boiled, and separated by SDS-PAGE, transferred to nitrocellulose membrane, and detected by Western blot analysis using ECL (Amersham Biosciences).
Overexpression of Wild Type and Phosphorylation-deficient Mutant Hsp27 in M-IEC-Plasmids (pcDNA3.1) encoding wild type hamster Hsp25 and phosphorylation deficient-mutant Hsp25 (AA) in which Ser 15 and Ser 90 are replaced by alanine were kindly provided by Dr. T. Tomako (McGill University, Montreal, Canada). Caco-2 cells were transfected with empty plasmid (vector), wild type, and AA by FuGENE reagent (Promega) overnight according to the manufacturer's instructions. A transfection efficiency of 50 -60% was observed using a green fluorescent protein control. The cells were recovered for 24 h while simultaneously co-cultured with differentiated THP-1, and then at the end of 24 h, conditioned IEC was treated with SAP and IL-1␤ as described before.
In Vitro Kinase Assay-Cells were incubated with cell lysis buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 25 mM ␤-glycerophosphate, 2 mM EDTA, 2 mM pyrophosphate, 1 mM orthovanadate, and 1% Triton X-100, 1 mM dithiothreitol, 1 mM NaF with protease inhibitors) followed by the addition of anti-IKK-␣ antibody. Following overnight end-to-end rotation of tubes at 4°C, immunoprecipitates were washed three times with lysis buffer and once with kinase buffer (20 mM Tris, pH 7.5, 1 mM MnCl 2 , 10 mM MgCl 2 , 20 mM ␤-glycerophosphate, 0.1 mM sodium orthovanadate, 2 mM NaF, and 1 mM dithiothreitol). Immunoprecipitates were finally resuspended in 20.0 l of kinase buffer containing 5 Ci of [␥-32 P]ATP and incubated at 30°C for 30 min. 1 g of glutathione S-transferase-IB␣ (Santa Cruz Biotechnology) was used as substrate. The reaction was terminated by boiling with SDS sample buffer for 5 min. Finally, the protein was resolved on 10% SDS-PAGE, the gel was dried, and the radioactive bands were visualized by autoradiography. The cell lysates were also checked for IKK-␣ expression for normalization. Statistical Analysis-The blots were scanned, and the densitometric values were obtained by using the Image J program. Statistical analysis to check significance was done with Student's t test using Prism software. The graphs plotted were from two to three independent experiments, and the error bars in all of the graphs represent the means Ϯ S.D.

Differential Induction of Hsp27 and Hsp72 in Naive and Macrophage-conditioned Epithelial Cells-Because the induction of
Hsp is a universal stress response against various insults and has been widely shown to have an anti-inflammatory effect, we checked the expression of Hsp representing four families. Although SAP treatment suppressed Hsp27 and Hsp72 in naïve IEC (Fig. 1A), the same two Hsp were significantly overexpressed in macrophage-conditioned Caco-2 and T84 colonic epithelial cells (Fig. 1B). Expression levels of Hsp60 and 90 remained unaltered in both naïve and conditioned IEC (not shown). Surprisingly, even though both secretory components and SAP elicited this stress response, SAP was consistently found to be twice as potent as secretory components (Fig. 1C) and was therefore used for all further studies. It should be noted that co-culturing alone did not affect the basal Hsp expression significantly (Fig. 1B). Also, treatment of IEC with 24 h THP medium was sufficient to prime epithelial cells for ameba-induced Hsp72 induction (Fig. 2) without the need for co-culturing. This confirms that macrophage secretions can alter epithelial cell responses to pathogens.
Ameba-induced Hsp Expression Is ERK MAP Kinasedependent-MAPK play important roles in regulating both inflammatory and stress responses (12) and Hsp have been shown to be regulated by different MAPK in intestinal epithelial cells (16,25). Hence, we checked whether ameba-induced expression of Hsp27 and 72 are regulated by MAPK. As shown in Fig. 3A, when the conditioned epithelial cells was pretreated with PD98059, an ERK MAPK inhibitor, SAP-induced Hsp expression was significantly inhibited. Correspondingly we  observed the activation of ERK MAPK by amebic proteins as early as 30 min (Fig. 3B).
E. histolytica Induction of Hsp72 Is Independent of Gal-lectin and Cysteine Proteinase-The three well recognized virulent factors in E. histolytica are the surface adhesin Gal/GalNac lectin, cysteine proteinases, and amebapore. The pathogenic effects of the Gal-lectin and cysteine proteinases have been extensively documented (26), and it was of interest to determine whether the up-regulation of Hsp72 is mediated by these virulent factors. As shown in Fig. 4, pretreatment of SAP with either galactose or E-64 failed to inhibit Hsp72 expression, indicating that Hsp induction by SAP is independent of Gal-lectin and cysteine proteinases. In contrast, significant abrogation of Hsp72 induction was noted by boiled SAP, suggesting that a yet to be characterized protein moiety might be responsible for this effect.
Amebic Components Induce Hsp72 via HSF-1-We then studied the mechanism of up-regulation of Hsp72 by amebic components. Quercetin is a flavanoid compound and a known inhibitor of Hsp synthesis by inhibiting HSF-1 (27,28), and it was of interest to determine whether it could inhibit SAP-mediated Hsp72 induction. As shown in Fig. 5A, quercetin (25 M) significantly inhibited Hsp72 expression. Quercetins at high concentrations are known to inhibit protein synthesis. Thus, to confirm that quercetin inhibits SAP induced Hsp72 expression by inhibiting HSF-1 and not by inhibiting Hsp protein synthesis, we treated the cells with equimolar concentrations of cycloheximide, an inhibitor of protein synthesis. Under these conditions cycloheximide treatment failed to inhibit SAP induced Hsp72 expression, confirming that amebic components indeed induce Hsp synthesis by activating HSF-1. To conclusively prove the role of HSF-1 in ameba-induced Hsp synthesis, we silenced the HSF-1 gene by siRNA (Fig. 5B) and observed inhibition of Hsp72 expression by SAP in macrophagetreated epithelial cells (Fig. 5C). Consistent with its role in regulating Hsp72 expression, HSF-1 was activated by amebic proteins as shown by Western blot analysis (Fig. 5D). We found that amebic components were able to induce trimerization of HSF-1 in M-IEC at a time point of 3 h. There was a corresponding increase in the DNA binding activity of HSF-1, which was inhibited by pretreatment with quercetin (Fig.  5E). These studies confirm that amebic components activate HSF-,1which in turn induces Hsp gene expression.
Heat Shock Response by Amebic Components Inhibits NF-B Activation Induced by IL-1␤-Because NF-B was shown to be the key mediator of colonic inflammation in amebic infection (4) and stress response is increasingly being shown to inhibit this molecule (12,(17)(18)(19)(20), we studied the activation of NF-B following amebic protein treatment. For this we treated M-IEC with SAP for 12 h and then stimulated with IL-1␤, a prototypical NF-B activator. As shown, pretreatment with amebic proteins inhibited NF-B-DNA binding activity (Fig. 6A), nuclear translocation of NF-B p65 subunit (Fig. 6B), and IB-␣ phosphorylation (Fig. 6C) induced by IL-1␤ in macrophage-conditioned colonic epithelial cells.
Ameba-induced Stress Response Inhibits IKK Activity via Hsp27-Because IB is phosphorylated by IKK, we checked IKK activity in an in vitro kinase assay. As shown in Fig. 6D, amebic pretreatment suppressed IL-1␤-induced IKK activity. Moreover, when HSF-1 gene was silenced by siRNA, this suppression was abrogated, suggesting that stress response is involved in this suppression. Thus, to precisely identify which Hsp mediates this effect, we silenced Hsp27 and 72 and checked the IKK activity. As shown in Fig. 7A, silencing Hsp27 significantly restored IKK activity, whereas Hsp72 did not have a significant effect on this suppression. We observed a significant suppression in the expression of both Hsp27 (not shown) and Hsp72 (Fig. 7B) (20,29). Hence, for the first time we checked the interaction between Hsp and IKK in intestinal epithelial cells and found that Hsp27, 60, and 90 associate with IKK-␣ (Fig. 8A), and treatment with SAP increased the association of Hsp27 with IKK-␣ by 2-fold (Fig. 8B). As expected, siRNA of Hsp27 reduced this association. Results from Figs. 7A and 8B together confirm that Hsp27 negatively regulates IKK activity. All of the Hsp tested, Hsp27, 60, 72, and 90 interacted with IKK-␤ in Caco-2 IEC (Fig. 8A) and again Hsp27 interaction with IKK-␤ was also enhanced by SAP treatment (data not shown).    A, conditioned IEC were transfected with vector (Vec), wild type (WT), or mutant Hsp25 plasmids during co-culture as described, and whole cell lysates were subjected to SDS-PAGE and probed with antibodies against hamster Hsp25 and actin. B, M-IEC overexpressing hamster Hsp25 plasmids were treated with SAP for 12 h, and lysates were subjected to immunoprecipitation using anti-hamster Hsp25 or control rabbit IgG antibody followed by Western blot using IKK-␣ antibody. Cell lysates from the same cells were probed with IKK-␣. C, conditioned IEC transfected with vector, wild type, or mutant Hsp25 plasmids were pretreated or not with SAP and IL-1␤, and whole cell lysates were checked for phospho-IB-␣ by Western blot as described. Bottom panel, 10 g of nuclear extracts from cells treated similarly were probed with anti-p65 antibody. D, M-IEC treated with SAP for different time periods and cell lysates probed with Ser(P) 15 -Hsp27 and actin. Representative blots from at least two different experiments were shown. IB, immunoblot.

Phosphorylation of Hsp27 Is Required for Interaction with
and Inhibition of IKK-To confirm that Hsp27 mediated IKK inhibition in IEC and to fully understand the mechanism involved, we overexpressed wild type and phosphorylation-deficient mutant of hamster Hsp25 (AA) in conditioned epithelial cells (Fig. 9A). Transfected M-IEC was treated with SAP and checked for interaction between Hsp25 and IKK-␣. As shown in Fig. 9B, although overexpression of wild type Hsp25 enhanced SAP-induced association with IKK-␣, Hsp25AA significantly inhibited this interaction. Consistent with the negative regulation of IKK by Hsp27, this decreased interaction between IKK-␣ and Hsp27 abrogated the inhibitory effect of SAP on IKK activity, seen as restored IB phosphorylation and NF-B p65 nuclear translocation (Fig. 9C). Consistently, we observed that SAP induced a delayed but significant phosphorylation of Ser 15 Hsp27 in M-IEC (Fig. 9D).

Ameba-induced Stress Response in Conditioned Epithelial Cells Has Cytoprotective Function-Because
Hsp also have cytoprotective abilities in different cells, we checked whether amebic proteins can protect IEC against injuries caused by diverse agents. As shown in Fig. 10A, pretreatment with SAP significantly increased cell viability following oxidative injury by hydrogen peroxide and reduced caspase-3 cleavage induced by the apoptotic agent, Fas L (Fig. 10B).

DISCUSSION
In this study we show that SAP have protective effects mediated by Hsp on intestinal epithelial cells. Interestingly, the Hsp were induced only in epithelial cells that have been exposed to macrophage secretions for 24 h but not in the naive IEC. We made two novel observations: first, this is the first report on the protective effect of amebic proteins, and second, macrophage secretions can prime epithelial cells to elicit stress responses.
Co-culture systems have been popular because of their better simulation of in vivo situation wherein epithelial and immune cells exist in close proximity and respond to each other components. Several studies reported a differential epithelial response, mostly with respect to cytokines, following exposure to immune cells. Two recent studies reported the induction of Hsp25 and Hsp72 in mouse intestinal epithelial cells co-cultured with lymphocytes, and IL-2 was found to be the mediator (13,16). We made the novel observation that intestinal epithelial cells express Hsp in response to amebic proteins following exposure to macrophage secretions.
Notably, basal Hsp expression remains unchanged in macrophage-conditioned IEC, suggesting that epithelial cells are only primed rather than activated by macrophage secretions. The cross-talk between macrophages and IEC could be bi-directional in the sense that macrophages also respond to epithelial secretions such as monocyte chemotactic protein-1 and transforming growth factor-␤ (30). However, amebic proteins could induce Hsp72 in epithelial cells that have been incubated with macrophage secretions without co-culturing, suggesting that epithelial cell modification of macrophages is not required to elicit a differential epithelial response. Because differentiated THP-1 cells are shown to constitutively produce a variety of cytokines such as TNF-␣, IL-1␤, IL-6, IL-8, etc. (31), and IL-1␤ induce Hsp expression (32), we checked the role of these cytokines in priming the IEC. Adding neutralizing antibodies against TNF-␣ and IL-1␤ in the macrophage medium did not significantly inhibited ameba-induced Hsp expression (data not shown). Previous studies (33) have shown that co-culture with activated peripheral blood mononuclear cells altered T84 epithelial cell physiology and that exogenously added TNF-␣ or interferon-␥ did not mimic the changes induced by immune cells. This suggests that mediators in immune cell secretions have an effect (both adverse and beneficial) on epithelial responses, but their identities are not known.
IEC are known to express Hsp in response to diverse insults such as pathogens, pathogen-specific molecules, and chemicals, and immune cells can alter this response. Accordingly, we checked the expression of four Hsp belonging to different families and found specific up-regulation of Hsp27 and 72 but not Hsp60 and 90 (data not shown) by SAP. Our observation of involvement of ERK MAP kinase in ameba-induced Hsp expression is consistent with previous reports (16,34,35) and extends the current thinking that MAP kinases play an important role in regulating both inflammatory and stress pathways. After establishing that amebic proteins induce Hsp in macrophage primed epithelial cells by activating HSF-1, we proceeded to check the functional significance of this phenomenon. Because our objective was to understand how colonic inflammation was absent in the majority of infected individuals and given the evidence that Hsp suppress NF-B activation, we critically analyzed whether and how Hsp induced by amebic proteins inhibits IL-1␤-mediated NF-B activation. We chose IL-1␤ for two reasons: first, is it is a prototypical NF-B inducer, and second, it was shown to play a role in ameba-induced colitis (2,36). Moreover, inhibiting NF-B significantly reduces amebic colitis in a mouse model of intestinal amebiasis (4). NF-B is a ubiquitous transcription factor that regulates a number of genes involved in inflammation and immune response (37). Activation of this transcription factor is critically regulated at multiple steps. Recently, the inhibitory effects of Hsp on NF-B activation are increasingly being demonstrated in different cell systems. Hsp72 has been found to associate with the p65 subunit of NF-B and inhibits the nuclear transport of the latter in FIGURE 10. Amebic proteins confer protection against oxidative and apoptotic injuries. A, M-IEC were treated with SAP (100 g/ml) for 12 h and then incubated with 10 mM of hydrogen peroxide for further 12 h, and cell viability was checked by neutral red assay as described. **, p Ͻ 0.01 compared with control; *, p Ͻ 0.05 compared with H 2 O 2 alone. NS, not significant; Ctl, control. B, M-IEC (T84 cells) were exposed to SAP (100 g/ml) for 12 h before treating with 1.0 g/ml of Fas L for 12 h. Cleavage of Pro caspase-3 was examined by Western blot. The experiment was repeated three times, and one representative blot is shown.
T-cells (38), and Hsp27 has been shown to be a ubiquitin-binding protein regulating the degradation of IB expression, thereby indirectly influencing NF-B activation (39). Recent studies show that the IKK complex is the potential target for Hsp inhibition of NF-B pathway (17,18,40). In particular, one study showed that Hsp27 interacts with IKK complex and negatively regulates its activation by TNF-␣ (20) in HeLa cells. Because SAP increased Hsp72 expression and inhibited IL-1␤induced NF-B p65 nuclear translocation, we checked for their interaction in IEC but failed to see any association (data not shown). We found that IB phosphorylation was also inhibited, which was a direct result of reduced IKK activity by SAP treatment. siRNA technology has rapidly become a revolutionary tool for efficient silencing of gene expression in a variety of experimental settings (41). We exploited this powerful system to silence Hsp genes to understand their role in ameba mediated suppression of IKK activation and found that silencing HSF-1 or Hsp27 but not 72 resulted in significant abrogation of this inhibition. Because different Hsp interact with and regulate IKK complex, we checked whether amebae-induced Hsp interacted with the IKK subunits in intestinal epithelial cells. For the first time, we reported an association between IKK and Hsp in IEC. Hsp27, 60, and 90 constitutively interact with IKK-␣, and the interaction of Hsp27 with IKK-␣ was strongly enhanced following SAP treatment. Although Hsp27 and Hsp90 have been previously shown to associate with the IKK complex, this is the first report of the interaction between Hsp60 and IKK.
Our observation that all Hsp tested interacted with IKK-␤ is surprising but could be attributed to the use of mild detergent (CHAPS) in the cell lysis buffer for co-immunoprecipitation. We found that Hsp27 association with IKK-␤ was also increased by amebic protein treatment (not shown). A previous report (20) showed that Hsp27 association with IKK-␤ but not with IKK-␣ increased in response to TNF-␣. The reason(s) for this selective interaction is unclear.
Because post-translational modifications of Hsp27 are known to regulate its biological activity (20), we tested the role of phosphorylation of Hsp27 in its ability to inhibit NF-B activation. Overexpression of the phosphorylation-deficient mutant of hamster Hsp25 (at serines 15 and 90) in IEC resulted in significant reversal of NF-B inhibition by amebic proteins. There are two sites phosphorylated in hamster Hsp27 (Ser 15 and Ser 90 ) and three in human (Ser 15 , Ser 78 , and Ser 82 ). Ser 78 -Ser 82 of human Hsp27 probably works in tandem as the equivalent of the unique Ser 90 in hamster Hsp25. We also observed that SAP induces significant phosphorylation of Ser 15 (Fig. 9D) but not Ser 72 (data not shown) after prolonged treatment in M-IEC. Because Hsp27 phosphorylation depends on p38 MAP kinase (20), this could have resulted from a delayed activation of p38 MAP kinase, as we did not find either p38 (data not shown) or Hsp27 phosphorylation in the early time periods of 1 h. Nonetheless it is clear that overexpression of Hsp27 amplified the SAP effects, whereas blocking the Hsp27 phosphorylation significantly abrogated the amebic inhibitory effect on NF-B activation. These studies confirm and further strengthen the concept that post-translational modifications, particularly, the phosphorylation of Hsp27 plays an important role in NF-B regulation via inhibiting IKK activity. Extending our observa-tions on the protective effects of ameba-induced stress responses, we also showed that prior treatment with amebic proteins protected the epithelial cells against hydrogen peroxide and Fas L treatment. This is not surprising in view of the potent apoptotic inhibitory effects of Hsp72 (42,43). In fact, both Hsp27 and 72 are powerful chaperones and inhibit key effectors of the apoptotic machinery including apoptosome, caspase activation complex, and apoptosis-inducing factor (44). Consistently, we also observed increased cell survival following apoptotic stimuli in M-IEC pretreated with amebic proteins. Although this observation supports the anticipated protective effects of stress response, its functional significance with respect to ameba pathogenesis is difficult to surmise. Contactdependent apoptosis induction by E. histolytica was reported in FIGURE 11. Schematic model for suppression of amebic colitis. Soluble amebic proteins activate the heat shock transcription factor-1 to form trimers from its inactive monomers. Activated HSF-1 induces the expression of Hsp27 and Hsp72, which perform different functions. Phospho-Hsp27 associates with IKK and inhibits its activity, thereby suppressing NF-B activation induced by the proinflammatory cytokines released in response to amebic infection. Suppression of NF-B favors increased apoptosis. However, Hsp72, which apparently does not play a role in NF-B signaling in IEC, might function to promote cell survival through its potent anti-apoptotic activity. Together the stress response induced by amebic proteins functions to inhibit intestinal inflammation and promote cell survival.
immune cells such as neutrophils, T-cells, and erythrocytes (23,45,46). It was also shown that E. histolytica preferentially ingests apoptotic Jurkat T-cells. But hitherto no data are present on epithelial cell apoptosis by the parasite. Assuming a similar phenomenon for epithelial cells, it is reasonable to argue that stress response by the host is an efficient way to circumvent parasite-induced cell death. At present, it is not known what soluble amebic molecule(s) is responsible for the induction of Hsp. In summary, we have shown that amebic proteins can inhibit NF-B activation and promote cell survival via stress protein expression in macrophage primed intestinal epithelial cells (Fig. 11). These studies for the first time, demonstrate a potential mechanism by which intestinal inflammation induced by E. histolytica might be inhibited in majority of infected individuals and suggests that amebic colitis could result from the lack of such protective responses to suppress pro-inflammatory cytokine induction in a minority of susceptible individuals.