TLR9-Dependent Induction of Intestinal α-Defensins by Toxoplasma gondii Free

The intestinal epithelium protects the host against microbial infection not only by forming a physical barrier but also by active participation in host innate defense via cytokine, chemokine, and antimicrobial peptide production (1). Defensins are a group of cationic antimicrobial peptides containing three intramolecular disulfide bonds that function by disrupting the membrane integrity of target organisms. They are categorized as α- and β-defensins depending on the position of their cysteine linkage (2). In mice, α-defensins are exclusively expressed in Paneth cells (PCs), which are epithelial cells located in the crypts of Lieberkhün lining the small intestine (3). PCs are heterogeneously distributed in the small intestine with a higher frequency in the ileum compared with those in the duodenum and jejunum (4).

α-Defensins display a broad spectrum of microbicidal activity against Gram-positive and Gram-negative bacteria and also against fungi, protozoa, and enveloped viruses to protect intestinal multipotent stem cells (5, 6). The primary mechanism by which defensins exhibit antimicrobial activity is the formation of pores in the microbial membrane that distort the integrity and function of the affected microbe (7). Cryptdin (Crp)-mediated cytotoxicity appears to be microbe-specific such that Crp-2 and Crp-3 are cytotoxic to Giardia lamblia (an apicomplexan extracellular parasite) trophozoites, whereas Crp-1 and Crp-6 are not (8). The biological activity of mouse α-defensins (or Crps) requires proteolytic activation. Mouse PCs secrete matured 3.5 kDa Crp as components of secretory granules. Matrix metalloprotease 7 (Mmp7) mediates 8.4 kDa pro-Crp processing intracellularly before secretion (9, 10).

In response to stimulation with a broad range of intestinal pathogens (mostly prokaryotes), PC secretory granules are released in the lumen, and mRNA expression of Crps is upregulated (11). Wehkamp et al. (12) have recently reported that the expression of Crp is regulated by the Wnt/β-catenin signaling pathway and requires the transcription factor Transcription factor-like 2 (also called TCF-4). However, in vivo stimulation of TLR9, with agonistic CpG-oligodeoxynucleotides (CpG-ODNs), also leads to PC degranulation (13). The activation of the TLR9 signaling pathway induces nuclear translocation of IFN regulatory factor 7 and the subsequent transcription of type I IFN-α and IFN-β (14, 15), this mechanism having been studied in plasmacytoid dendritic cells. Type I IFN-α and IFN-β share the same receptor (IFNAR) expressed by epithelial cells and in the lamina propria mainly by dendritic cells. By interacting with its receptor, IFN-α or IFN-β can locally amplify the type I IFN response by either autocrine or paracrine mechanisms (16). It remains unclear whether PCs can release their content directly upon TLR9 activation or whether it requires the subsequent production of type I IFNs. The i.p. treatment with IFN-α can result in the degranulation of rat PCs into the lumen (17). Besides a direct antimicrobial effect, α-defensins may affect microbial pathogenesis by subsequently activating innate immune defenses.

Toxoplasma gondii is an obligate intracellular parasite that is acquired by oral ingestion of either sporozoites from contaminated soil or tissue cysts containing bradyzoites from undercooked meat products, such as pork, mutton, and beef (18–20). We have previously reported that TLR9 is required to initiate the protective immune response following oral infection with T. gondii in mice. When orally infected with T. gondii cysts, B6 TLR9^−/− mice are unable to mount an efficient Th1 IFN-γ–mediated response, leading to increased parasite replication (21). In these studies, we demonstrate that oral challenge with T. gondii upregulates Crp mRNA expression and production by intestinal epithelial cells (IECs) and leads to PC degranulation. Interestingly, this innate immune response mechanism requires TLR9 signaling pathway activation and is mediated by TLR9-dependent type I IFN production. TLR9-dependent production of Crp may in part be responsible for early parasite elimination by this pathway.

Female 8- to 10-wk-old C57BL/6 (B6) wild-type (WT) and B6 Mmp7^−/− from The Jackson Laboratory (Bar Harbor, ME) or TLR9^−/− mice (C57BL/6 background) from Oriental BioService (Kyoto, Japan) were bred and housed under approved conditions at the Animal Research Facility at Dartmouth Medical School. B6 IFNAR^−/− mice were a generous gift from Bernhard Ryffel (University of Orleans and Centre National de la Recherche Scientifique), and B6 germ-free mice were purchased at Institut National de la Recherche Agronomique, Jouy-en-Josas, France.

Mice were injected i.p. with murine IFN-β purified from transfected Chinese hamster ovary cells, generously provided by Dr. Edward Croze (Bayer Health Care Pharmaceuticals, Wayne, NJ), diluted in PBS (Ca/Mg-free)/0.1% BSA. Mice were infected orally by intragastric gavage with 20 or 35 cysts of the 76K T. gondii strain maintained through passage in CBA/J mice from The Jackson Laboratory. Postinfection, mice were weighed daily. All of the experiments were performed with three to six mice per group analyzed individually and were repeated a minimum of two times; error bars represent the SEM.

Small intestines were gently flushed with 15 ml ice-cold PBS/2 mM EDTA (Ca/Mg-free) and flash-frozen in liquid nitrogen. Total protein extracts were prepared by homogenizing tissue samples in 30% acetic acid and extracted by rotary shaking overnight at 4°C. Extracts were centrifuged at 100,000 × g for 2 h at 4°C, and clarified supernatants were lyophilized after extensive dialysis. Lyophilized samples were dissolved in 5% acetic acid and analyzed by acid/urea PAGE (AU-PAGE) in 12.5% gels at 220 V for 3 h. After AU-PAGE separation, resolved proteins were transferred to 0.22-μm nitrocellulose membranes and blocked with 5% skim milk, and the membranes were incubated sequentially with anti-rabbit mouse Crp-1 (1:500), HRP-conjugated anti-rabbit IgG (1:20,000), and chemiluminescent substrate before exposing the blot to x-ray film. Anti–Crp-1 polyclonal Abs cross-react with mouse Crp-1, Crp-2, Crp-3, and Crp-6 (22).

Small intestines were washed with 15 ml ice-cold PBS/2mM EDTA (Ca/Mg-free). One- to two-centimeter sections from its distal third (this segment was arbitrarily designated as the ileum) were cut and fixed overnight in PBS/10% formalin 10% (Fisher Scientific, Waltham, MA). Each of the tissue sections was then placed in a tissue cassette and embedded in paraffin, and 5-μm thick longitudinal slices were prepared and stained with the phloxine/tartrazine standard protocol. Tartrazine-positive vesicles (orange and pink) contained in pyramidal-shaped cells (PCs) at the bottoms of the crypts were enumerated. Fifteen individual PCs located in at least five different crypts from three mice were used for the enumeration.

Small intestines were washed in PBS/3mM EDTA (Ca/Mg-free), and Peyer’s patches were removed. Intestines were opened longitudinally and cut into 1-cm long fragments. Pieces were incubated in PBS/3 mM EDTA under agitation (10 min, 37°C), and supernatants containing IECs (i.e., enterocytes and PCs) were collected and washed in RPMI 1640 with 5% FBS. This process was repeated twice. DTT (Sigma-Aldrich, St. Louis, MO) was then added to the IEC suspension (1.5 mg per 10 ml). After incubation (15 min, 37°C), cells were washed twice in RPMI 1640 with 5% FBS. Purified IECs were collected at the interface of a 60–30% Percoll gradient (30 min, 300 × g) and washed in RPMI 1640 with 5% FBS. Purified IECs did not contain any CD11c⁺ cells as measured by mRNA expression.

Lamina propria cells were purified as described previously (23). Briefly, small intestines were washed with PBS (Ca/Mg-free), and fat and Peyer’s patches were removed. Small intestines were then opened longitudinally, cut into 1-cm pieces, and washed in PBS/3 mM EDTA (10 min, 37°C). Intestine pieces were then washed in RPMI 1640 with 1 mM EGTA, 1.5 mM MgCl₂, and 5% FBS (10 min, 37°C). Intestine pieces were then washed twice in PBS (Ca/Mg-free) (10 min, 37°C). The tissue was digested in 0.14 Wünsch U/ml Liberase (Roche, Basel, Switzerland) and 5 U/ml DNase I (Sigma-Aldrich) in RPMI 1640 at 37°C for 45 min under agitation. Cell suspensions were washed twice in RPMI 1640 with 2% FBS then laid over Histopaque (density 1.077) and centrifuged (20 min, 300 × g). Cells at the interphase were collected, and washed in RPMI 1640 with 2% FBS before RNA extraction.

A total of 2 μg RNeasy-purified (Qiagen, Valencia, CA) mRNA was reverse-transcribed using a GeneAmp RNA PCR kit (Applied Biosystems, Foster City, CA). A total of 200 ng cDNA was amplified using the 2× SYBR Green PCR master mix (Applied Biosystems) on an iCycler (Bio-Rad, Hercules, CA). Relative mRNA expression were normalized to β-actin and was expressed using the ΔΔCt method where relative mRNA expression is 2^{−(expΔCt − controlΔCt)} × 1000. See Table I for the sequences of primers used for this study.

A total of 1 μg genomic DNA was prepared from the small intestine or mesenteric lymph node using a DNeasy tissue kit (Qiagen). The T. gondii B1 gene was amplified using the 2× SYBR Green PCR master mix (Applied Biosystems) on an iCycler. The primers used for QB1 were: forward, 5′-GGAACTGCATCCGTTCATGAG-3′; reverse, 5′-TCTTTAAAGCGTTCGTGGTC-3′. A standard curve for parasite equivalents was generated using a plasmid as described previously (24).

Groups were compared using the unpaired t test in Prism statistical software (GraphPad, San Diego, CA). Values of *p < 0.005 or **p < 0.001 were considered significant.

Quantitative expression of mRNA for a panel of Crps (from 2–5) by purified IECs from B6 WT mice was measured by two-step SYBR green quantitative real-time PCR following oral infection with T. gondii (Table I). All of the Crp mRNAs were upregulated in response to T. gondii infection; however, the observed effect was most apparent in Crp-3 and Crp-5 mRNA expression. The expression of these two Crps increased above constitutive level within 24 h after oral infection and maintained three to five times higher than those naive mice at least until day 5 postinfection (Fig. 1A). Increased production of Crps following oral infection by T. gondii was confirmed at the protein level by AU-PAGE immunoblots using a pan–Crp-1, Crp-2, Crp-3, and Crp-6 polyclonal Ab (Fig. 1B). Mmp7, required for Crp activation before secretion, was constitutively expressed at the mRNA level and increased to day 5 postinfection (Fig. 1C) when the experiment was terminated.

Phloxine/tartrazine is a stain used for demonstrating acidophilic inclusion bodies. In the small intestine, PC tartrazine-positive granules appear red on a yellow background. At days 1 and 3 after oral challenge with T. gondii, B6 WT mice showed reduced frequencies of secretory granules, preferentially located at the apical sides of the cells, indicating PC degranulation (Fig. 2A, 2B). Five days postinfection, the secretory granule content of PCs was slightly higher than that observed in control (D0) mice. However, there was no quantitative shift in the frequency of pyramidal-shaped cells at the bottoms of the crypts in response to infection (five to eight per crypt), indicating that PCs did not proliferate (Fig. 2A). Confirmation was provided using anti-Crps–related sequence 4c Abs, an antimicrobial peptide contained specifically in PCs (29, 30, data not shown). These observations suggested that oral infection with T. gondii induces Crp production and secretion by PCs.

Rumio et al. (13) have shown that activation of the TLR9 signaling pathway with agonistic CpG-ODNs leads to PC degranulation in mice. We have previously observed that TLR9 expression, by both parenchymatous cells and hematopoietic cells, is required for Th1 polarization of the immune system postinfection with T. gondii (21). In the current study, TLR9 expression and the TLR9 signaling pathway activation marker IFN-β within the small intestine of B6 WT mice housed either in normal or germ-free conditions were measured by real-time PCR. TLR9 and IFN-β mRNA expression were modulated during the first week following oral challenge with T. gondii in B6 WT mice housed in both normal and germ-free conditions (Supplemental Fig. 1). Although B6 WT mice housed in germ-free facilities exhibited a reduced level of TLR9 mRNA expression in their small intestine as compared with that in mice raised in normal conditions, these data indicated that T. gondii can directly activate the TLR9 signaling pathway independent of commensal gut microflora.

To investigate whether the activation of the TLR9 signaling pathway by T. gondii is linked with the release of Crps by PCs, IECs were purified from infected B6 WT or TLR9^−/− mice to measure the production of Crp-3 and Crp-5. IECs from B6 TLR9^−/− mice failed to modulate Crp mRNA expression in response to T. gondii infection, whereas IECs from B6 WT mice demonstrated increased expression of Crp-3/-5 mRNA (Figs. 1, 3) through day 5 postinfection when the study was terminated. This indicates that TLR9 is required for Crp mRNA modulation in response to T. gondii infection in mice.

Previous observations from our laboratory show that type I IFN-α and IFN-β are increased in TLR9-expressing mouse IECs (21, data not shown) (Supplemental Figs. 1, 2). Mice were treated i.p. with exogenous mouse IFN-β (mIFN-β) to determine whether Crp modulation is directly linked to TLR9 ligation or is a downstream effect due to type I IFN production resulting from TLR9 activation. To address this question, B6 WT mice were treated daily with mIFN-β (5000 U per day per mouse) starting the day prior to infection. Crp mRNA expression by purified IECs was measured at serial time points postinfection (Fig. 4). A significant increase in Crp-3/-5 mRNA expression was observed following mIFN-β treatment of uninfected mice compared with that in untreated uninfected mice. During the early stage of infection (days 1–3), the Crp mRNA upregulation was identical whether mice were treated with mIFN-β or infected with T. gondii. At day 5, there was no substantial difference in Crp mRNA expression between mIFN-β–treated and infected mice, suggesting that parasite-driven type I IFN production is sufficient for Crp mRNA upregulation.

To confirm the direct effect of type I IFN upon Crp mRNA upregulation by IECs independently of TLR9 engagement, B6 TLR9^−/− mice were treated i.p. with a high dose of mIFN-β (5000 U per day per mouse). Mice lacking TLR9 were unable to produce type I IFN and were unable to modulate Crp mRNA expression in response to T. gondii infection (Fig. 4, Supplemental Fig. 3). Nevertheless, IFN-β paracrine activity remains functional in this model. As measured by real-time PCR, B6 TLR9^−/− mice treated with mIFN-β demonstrated a 30-fold increase in IFN-β mRNA expression in the small intestine compared with that in the untreated group. This was correlated with an upregulation in Crp-3 and Crp-5 expression by IECs (Fig. 4). At day 5 posttreatment, despite a decrease in overall type I IFN production by T. gondii infection in TLR9^−/− mice, levels of IFN-β were maintained when mice were supplemented with IFN-β (Supplemental Fig. 3). Similarly to B6 WT mice (Fig. 4), when type I IFN expression was maintained at a high level of expression, B6 TLR9^−/− mice kept producing Crp-3 and Crp-5 mRNA as late as day 5 after the beginning of the treatment with mIFN-β (Fig. 5). These observations confirm the ability of type I IFNs produced upon TLR9 signaling pathway activation to trigger Crp mRNA upregulation.

Previous studies have shown that TLR9 agonists induce IFN-β production, which stimulates its own production via autocrine- or paracrine-dependent mechanisms. Mice lacking IFNAR expression (B6 IFNAR^−/−) are unresponsive to type I IFN (15). As shown in Fig. 6A, B6 IFNAR^−/− mice orally infected with 35 cysts of T. gondii were unable to sustain IFN-β production in the small intestine after 24 h.

To further characterize the requirement of type I IFN signal to induce Crp release by PCs following TLR9 signaling pathway activation by T. gondii, Crp-3/-5 mRNA expression by purified IECs was measured and PC degranulation analysis in B6 IFNAR^−/− mice was performed. After oral challenge with T. gondii, mice lacking a functional type I IFN receptor did not upregulate Crp-3 and Crp-5 mRNA production (Fig. 6B).

To investigate whether Crps have an anti-T. gondii activity, parasite burden was measured by real-time PCR in the intestine from infected Mmp7-deficient mice (B6 Mmp7^−/−) that cannot process immature Crp into biologically active Crp (9). Postinfection, the lack of intracellular mature Crp in B6 Mmp7^−/− mice was correlated with an increase in parasite number during the early phase of the infection. At day 3 postinfection, the parasite burdens in the small intestines from B6 Mmp7^−/− mice were, respectively, 5- and 7-fold higher compared with that in B6 WT mice. However, this effect was transitory. A week after the infection, the parasite burden in the small intestine from B6 Mmp7^−/− mice was identical to that of control mice (Fig. 7A). Th1 cytokines, in particular IFN-γ, are essential to parasite clearance in the infected murine host (31). To address whether the higher parasite burden observed in B6 Mmp7^−/− mice during the early phase of infection is due to a defect in stimulation of an efficient Th1 immune response, IFN-γ, TNF-α, and T-bet mRNA expression, the hallmark of a Th1-like immune response, were measured in these mice (Fig. 7B and data not shown). During the early phase of the infection, IFN-γ mRNA expression by purified lamina propria cells of B6 Mmp7^−/− mice was significantly lower than that observed in B6 WT mice.

Taken together, these data illustrate that after T. gondii oral infection in mice, the TLR9 signaling pathway is activated, resulting in type I IFN production in the small intestine. Sustained intestinal production of type I IFNs allows for Crp release by PCs in the lumen in response to T. gondii oral challenge. This early TLR9 engagement may participate in the clearance of the parasite directly or indirectly by promoting a Th1 response.

Besides being a physical barrier, the intestinal epithelium is an important compartment involved in innate immune defense against invasive pathogens. In most circumstances, the intracellular parasite T. gondii gains entry to the host via oral infection from contaminated soil or meat products. From its entry in the small intestine, the parasite is disseminated throughout the host to all organs, in particular muscle and brain. The gut epithelial barrier is therefore a strategic place to prevent or at least to limit parasite invasion. In the current study, we observed that Crp is an early innate response molecule that is upregulated in the intestine in response to TLR9 activation by the parasite and appears independent of the gut microflora.

After T. gondii oral infection in mice, the TLR9 signaling pathway is activated, resulting in type I IFN production in the small intestine. This sustained intestinal production of type I IFN is a key mediator for Crp release by PCs in the lumen in response to T. gondii oral challenge. As shown in B6 Mmp7^−/− mice, Crp release participates directly or, more likely, indirectly in the early clearance of the parasite. We observed that mouse α-defensins (Crps), mostly Crp-3 and Crp-5, are produced by IECs and that PCs secrete granules in response to oral infection of B6 WT mice. The increase in Crp production (at the mRNA and protein levels) can be secondary to stimulation of the PCs by the parasite, or rather quantitative amplification of the PCs. Oral infection of B6 WT mice with T. gondii cysts leads to massive recruitment of neutrophils, macrophages, dendritic cells, and lymphocytes in the lamina propria and to the production of proinflammatory cytokines that might enhance PC differentiation (32–34). During the early phase of T. gondii infection (days 1–3), Mmp7 mRNA expression, linked to the production of Mmp7 necessary for Crp maturation, is maintained at a constant level, and the frequency of PCs (pyramidal-shaped cells containing tartrazine-positive vesicles enumerated in tissue sections) is identical to that observed in naive mice. These findings would indicate that the increased production and release of Crp are linked to an increase in PC activity in response to parasite exposure.

α-Defensins require proteolytic processing and are signal-induced–released to be active in the intestinal lumen. LPS lipoteichoic acid, lipid A, muramyl dipeptide, and CpG-ODNs elicit Crp secretion, suggesting a potential role for innate pattern recognition molecules, such as TLRs and nucleotide-binding oligomerization domain proteins, in the regulation of Crp release (11, 35). A previous report and unpublished observations from our laboratory have shown that TLR9, but not TLR2 and/or TLR4, is required for an efficient antimicrobial Th1 immune response induction following oral T. gondii infection in mice (21). In germ-free conditions, the current study shows a reduction of the TLR9 signaling pathway activation level compared with that in B6 WT mice housed in standard conditions. TLR signaling can occur by either endogenous molecules (e.g., IL-1, DNA, or RNA) or microbial signature molecules (e.g., pathogen-associated molecular patterns) (36, 37). In mice lacking gut microflora, constitutive expression of TLR9-dependent factors is therefore attenuated but not silenced. After oral challenge with T. gondii cysts, the activation pattern of the TLR9 signaling pathway was independent of colonization with commensal microflora. T. gondii-infected, germ-free mice initiated a normal Th1 response (data not shown), suggesting that T. gondii alone can specifically induce TLR9 signaling pathway activation in the absence of normal commensal microbial signaling molecules.

TLRs are widely expressed in mouse small intestine by both hematopoietic and nonhematopoietic cell subsets. For instance, enterocytes and PCs express TLR9, as well as APCs within the lamina propria (38–41). Rumio et al. (13) have reported that in B6 WT mice treated with TLR9 agonist (CpG-ODNs) PCs release antimicrobial peptides in the lumen. PCs are heterogeneously distributed throughout the small intestine and located at the bottom of Lieberkhün crypts. Interestingly, PC degranulation in T. gondii-infected B6 WT mice was observed in the majority of crypts. The lack of functional TLR9 impaired Crp production and release by PCs, implying a link between TLR9 activation and Crp production. Due to the broad PC degranulation throughout the entire rather than targeted portions of the small intestine and the low number of parasites used in the study, we hypothesized that this process was dependent upon TLR9-associated factors rather than the TLR9 signaling pathway itself.

Type I IFN-α and IFN-β are produced upon TLR9 signaling pathway activation and trigger PC degranulation when injected i.p. in rats (17). The lack of Crp release by PCs in TLR9^−/− infected mice is associated with the abrogation of type I IFN production in our model. When B6 WT or TLR9^−/− mice were treated with mIFN-β, Crp-3/-5 production and PC degranulation were restored. These data indicate that PCs are functionally active in TLR9^−/− mice and that type I IFN signal is sufficient to induce the release of secretory granules. TLR9 signaling pathway engagement induces IFN-β production, which acts as signaling molecule to further activate IFN-α and IFN-β production (15). Unlike B6 WT mice, IFNAR^−/− mice were only able to initiate immediate but transient production IFN-α and IFN-β in the small intestine following T. gondii oral challenge. Type I IFNs have been described to promote protozoan parasite killing and also to enhance IFN-γ production by NK cells (42, 43), a crucial feature for the clearance of T. gondii (44). Intestinal parasite burden was dramatically increased in B6 IFNAR^−/− mice (data not shown), showing that the abrogation of the TLR9-dependent type I IFN amplification loop in this model allows a higher rate of T. gondii proliferation in the small intestine. Moreover, the lack of functional type I IFN signaling appears to be associated with an abrogation of Crp-3/-5 production and release by PCs, indicating that TLR9-dependant type I IFN production and signaling are mediating PC activation in our model.

The microbicidal activity of defensins, like other cationic antimicrobial peptides, is thought to involve disruption of membrane integrity. Crp-4 and Crp-5 are the most cationic peptides and have the greatest combined bacteriostatic and bactericidal activity against Escherichia. coli (45). Crps are also able to kill protozoa, such as the apicomplexan parasite Giardia (8). Intestinal parasite load in infected B6 Mmp7^−/− mice, measured shortly postinfection (day 3), is decreased as compared with that in their WT counterparts. In B6 Mmp7^−/− mice, the production of mature Crps is impaired. We observed a substantial impact of Mmp7 in this toxoplasma model, indicating that antimicrobial activity of Crps can occur in the lumen prior to invasion into the epithelial barrier. However, upregulation of Crp-3 and Crp-5 production by the epithelial cells begins 24 h postinfection, a time when most of the parasites have already invaded the epithelial cells and cannot be affected by the release of Crps into the lumen. In addition to their microbicidal activity, human α-defensins (human neutrophil peptides) also display signaling function. For example, human neutrophil peptide 3 promotes chloride efflux in vitro (46, 47) and can also induce in vitro the secretion of NF-κB–dependent chemokines, such as CXCL8 or CCL2 (48) from T84 human epithelial cells. Chemokines play a major role for in vivo initiation of a protective Th1 immune response to clear the parasite (49, 50). In infected B6 Mmp7^−/− mice, the early production (days 1 and 3) of NF-κB–dependant chemokines, such as CCL2, CCL3, and CCL5, by IECs is slightly lower than that in B6 WT mice (data not shown). This might explain a defect in attraction of scavenger immune cells and explain the increase in parasite load within the gut of B6 Mmp7^−/− mice. In addition, the lack of Crp production in B6 Mmp7^−/− mice was associated to an overall defect in IFN-γ production. The lower production of IFN-γ in B6 Mmp7^−/− mice may account for the increased parasite burden following infection. Thus, Crps may be important in the innate immune response against T. gondii by combining microbicidal activity and immune cell recruitment.

In the current study, we have observed that T. gondii can elicit a TLR9-mediated immune response, resulting in type I IFN expression and secretion in the gut-associated lymphoid tissue. In turn, type I IFNs act as signaling molecules, amplifying their own production over time and throughout the entire small intestine, leading to the release of Crps by the PCs.

We are very thankful to Dr. Andre J. Ouellette and Michael T. Shanahan (Department of Pathology and Laboratory Medicine, University of California at Irvine) for their technical support and advice to study PCs and Crps at the protein and histological levels.

Disclosures The authors have no financial conflicts of interest.