NF-kappaB p65-Dependent Transactivation of miRNA Genes following Cryptosporidium parvum Infection Stimulates Epithelial Cell Immune Responses

Cryptosporidium parvum is a protozoan parasite that infects the gastrointestinal epithelium and causes diarrheal disease worldwide. Innate epithelial immune responses are key mediators of the host's defense to C. parvum. MicroRNAs (miRNAs) regulate gene expression at the posttranscriptional level and are involved in regulation of both innate and adaptive immune responses. Using an in vitro model of human cryptosporidiosis, we analyzed C. parvum-induced miRNA expression in biliary epithelial cells (i.e., cholangiocytes). Our results demonstrated differential alterations in the mature miRNA expression profile in cholangiocytes following C. parvum infection or lipopolysaccharide stimulation. Database analysis of C. parvum-upregulated miRNAs revealed potential NF-κB binding sites in the promoter elements of a subset of miRNA genes. We demonstrated that mir-125b-1, mir-21, mir-30b, and mir-23b-27b-24-1 cluster genes were transactivated through promoter binding of the NF-κB p65 subunit following C. parvum infection. In contrast, C. parvum transactivated mir-30c and mir-16 genes in cholangiocytes in a p65-independent manner. Importantly, functional inhibition of selected p65-dependent miRNAs in cholangiocytes increased C. parvum burden. Thus, we have identified a panel of miRNAs regulated through promoter binding of the NF-κB p65 subunit in human cholangiocytes in response to C. parvum infection, a process that may be relevant to the regulation of epithelial anti-microbial defense in general.


Introduction
The protozoan parasite, Cryptosporidium parvum, is a causative agent of human gastrointestinal disease worldwide [1].C. parvum infects the gastrointestinal epithelium to produce a self-limiting diarrhea in immunocompetent individuals but is potentially lifethreatening in immunocompromised persons, especially those with the acquired immunodeficiency syndrome (AIDS) [1,2].Transmission occurs via the fecal-oral route.Humans are infected by ingesting C. parvum oocysts; oocysts then excyst in the gastrointestinal tract releasing infective sporozoites.C. parvum sporozoites can also travel up the biliary tract to infect the epithelial cells lining the biliary tract (i.e.cholangiocytes) [1,3].Mediated by specific ligands on the sporozoite surface and receptors on the host cells, the sporozoite attaches to the apical membrane of epithelial cells and forms a parasitophorous vacuole in which the organism remains intracellular but extracytoplasmic [3].The sporozoite then matures and undergoes further development of its life cycle.With this unique extracytoplasmic niche within epithelial cells preventing a direct infection of other cell types, C. parvum is classified as a ''minimally invasive'' mucosal pathogen [1].
Because of the ''minimally invasive'' nature of C. parvum infection, innate immune responses by epithelial cells are critical to the host's defense against infection.Toll-like receptor (TLR) -and nuclear factor-kappaB (NF-kB) -mediated signaling pathways are important components in epithelial innate immunity to C. parvum infection [4,5].TLRs are transmembrane proteins with highly conserved structural domains [6].Upon engagement of the TLRs by specific ligands, various adaptor molecules including myeloid differentiation factor 88 (MyD88) are selectively recruited to the receptors forming a complex referred to as the ''signalosome'' [6,7].The signalosome then triggers a series of downstream events including activation of the NF-kB [6][7][8].NF-kB subunits bind to the kB sites within the promoters/enhancers of target genes resulting in the transcriptional regulation of multiple genes important to epithelial anti-C.parvum defense [4,5].
MicroRNAs (miRNAs), a newly identified class of endogenous small regulatory RNAs of ,24 nucleotides, are emerging as key mediators of many biological processes and impact gene expression at the posttranscriptional level [9,10].Similar to other RNA molecules, most of miRNAs are initially transcribed as primary transcripts (termed pri-miRNAs) by Poly II and processed by the RNase III Drosha (in the nucleus) and a second RNase III Dicer (in the cytoplasm) to generate mature miRNA molecules [11][12][13].However, molecular mechanisms underlying miRNA gene transcriptional regulation are largely unclear [14].Recent studies on expression of miRNA genes have revealed potential transcriptional regulation by transcription factors, such as NF-kB and C/EBPa [15,16].
While much of our understanding of the cellular processes modulated by miRNAs has come from studies on development and tumorigenesis, the role of miRNAs in immune responses is now being gradually uncovered [17][18][19].The importance of miRNAs in cell-mediated immunity is highlighted by Dicer conditional knockout mice.Specific deletion of dcr-1 in the T cell lineage resulted in impaired T cell development and aberrant T helper cell differentiation and cytokine production [20].In addition, miRNA expression is impacted by cytokines in some model systems.Both interferon (IFN) -a and IFN-b modulate expression of several miRNAs required for their anti-viral responses following infection with hepatitis C virus [21].The TLR4 ligand, lipopolysaccharide (LPS), impacts expression of miR-132, miR-146, and miR-155 in human THP-1 monocytes [15,22].Further characterization of miR-146 revealed that this miRNA may function as a negative regulator of tumor necrosis factor receptor-associated factor 6 and interleukin-1 receptor associated kinase 1 [15].Recent studies also implicate specific miRNAs in controlling various epithelial cell processes such as regulation of cellular differentiation, determination of epithelial cell fate (cell death and proliferation), initiation and regulation of anti-microbial immunity, fine-tuning of inflammatory responses, and activation of specific intracellular signaling pathways [17][18][19]23].Using a non-malignant human cholangiocyte cell line (H69) that expresses multiple TLRs including TLR4 [5], we previously demonstrated that infection of human cholangiocytes by C. parvum in vitro mimics parasitial apical invasion and TLR4/ NF-kB-dependent epithelial responses in vivo [3].Moreover, let-7 regulates TLR4 expression via translational suppression in human cholangiocytes and is involved in epithelial defenses against C. parvum [24].Members of the miR-98/let-7 family also regulate expression of cytokine-inducible SH2-containing protein (CIS) in cholangiocytes following C. parvum infection [25].Together, these findings demonstrate that miRNAs levels in epithelial cells are altered by C. parvum infection and may regulate epithelial anti-C.parvum immune responses.
In this study, we performed an array analysis of miRNA expression in H69 cells following C. parvum infection and LPS stimulation.The analysis revealed significant alterations in miRNA expression in cholangiocytes following C. parvum infection or treatment with LPS.Of those miRNAs upregulated by C. parvum infection, we identified potential NF-kB binding sites in the promoter elements of several miRNA genes.Inhibiting activation of NF-kB p65 blocked C. parvum-induced upregulation of a panel of miRNA genes.Promoter binding and transactivation of the NF-kB p65 subunit of each selected miRNA gene was confirmed by chromatin immunoprecipitation assay and promoter luciferase reporter analysis.Furthermore, functional inhibition of the NF-kB p65-binding miRNAs increased C. parvum burden in cholangiocytes in vitro.These data demonstrate that a panel of miRNAs is regulated through promoter binding of the NF-kB p65 subunit in human cholangiocytes and these miRNAs are involved in epithelial defense in response to C. parvum infection, suggesting a role of miRNAs in regulation of epithelial anti-microbial defense.

C. parvum infection induces alterations in miRNA expression in cholangiocytes in vitro
To globally assess miRNA expression in epithelial cells following C. parvum infection, we performed a microarray analysis of mature miRNA expression in H69 cells [26].The miRCURY TM LNA human microRNAs assays (version 8.1; Exiqon; Vedbaek, Denmark) covers a total of up to 600 known human mature miRNAs and were used as previously described [27].The quality of the RNA was verified using an Agilent 2100 Bioanalyzer (Figure S1).A total of 383 mature miRNAs were detected in the uninfected H69 cells.Of the miRNAs expressed, miR-23b, miR-30b, miR-30c, and miR-125b expression were significantly increased in H69 cells after exposure to live C. parvum infection for 12 h (p, = 0.05; Figure 1A and Table S1).Five additional miRNAs (miR-15b, miR-16, miR-27b, miR-24, and miR-21) showed a tendency to increase (0.05,p, = 0.20) (Figure 1A).A total of 19 miRNAs were significantly downregulated (p, = 0.05) and 30 additional miRNAs showed a tendency to decrease (0.05,p, = 0.20) following C. parvum infection (Figure 1A and Table S1).Sham-infected control cells (H69 cells exposed to heat-inactivated C. parvum oocysts after incubation at 65uC for 30 min) displayed a similar miRNA expression profile as noninfected control samples (Table S1).Microarray analysis of mature miRNAs was also performed on H69 cells treated with LPS (1 mg/ ml for 8 h).Interestingly, most of the miRNAs upregulated by C. parvum also displayed an increased expression in cells treated by LPS (Figure 1B and Table S1).Nevertheless, increased expression of additional 13 miRNAs was identified in LPS-treated cells but not in cells exposed to C. parvum.A total of 31 miRNAs showed a decreased expression in LPS-treated cells and 10 of them were also downregulated by C. parvum (Figure 1B and Table S1).No LPS contamination in the C. parvum preparation was detected using the Limulus Amebocyte Lysate (LAL) test kit (Bio-Whittaker) (data not shown).All microarray data were described in accordance with MIAME guidelines and deposited at ArrayExpress (accession number: E-MEXP-2050 and E-MEXP-2052).
Real-time PCR analysis using primers and probes for mature miRNAs (Ambion) was performed to assess the kinetics of selected miRNAs in H69 cells following C. parvum infection.Increased expression of miR-125b, miR-21, miR-23b, miR-30b and miR-16 was detected in H69 cells following C. parvum infection for 12 h to 24 h, but not in the early time points (2 h to 8 h) (Figure 2A).Increased expression of miR-125b, miR-16, miR-23b, miR-21 and miR-30b, as well as decreased expression of miR-98, was further

Author Summary
MicroRNAs (miRNAs) are newly identified small non-coding RNAs that regulate gene expression at the posttranscriptional level.While much of our understanding of the cellular processes modulated by miRNAs has come from studies on development and tumorigenesis, the role of miRNAs in immune responses is now being gradually uncovered.Nevertheless, whether miRNA-mediated posttranscriptional gene regulation is involved in the fine-tuning of epithelial cell immune responses against pathogen infection remains undefined.Cryptosporidium parvum is a protozoan parasite that infects gastrointestinal epithelium.TLR/NF-kB-mediated innate immune responses by epithelial cells are critical to the host's defense to infection.Using an in vitro model of human cryptosporidiosis, we show here differential alterations in the miRNA expression profile in biliary epithelial cells following C. parvum infection.Promoter binding of NF-kB p65 subunit accounts for the upregulation of a panel of miRNA genes in cells infected by C. parvum.Importantly, functional inhibition of several NF-kB p65-dependent miRNAs in epithelial cells increases C. parvum infection burden.Our findings suggest that host epithelial cells activate NF-kB signaling to regulate miRNA expression in response to C. parvum infection.Moreover, NF-kB-mediated miRNA expression is involved in epithelial anti-microbial defense.Our study provides new insights into epithelial cell immunoregulation.confirmed in cells following C. parvum infection for 12 h by Northern blot (Figure 2B).An increased expression of the precursors for miR-125b, miR-16, miR-21 and miR-23b was also detected in cells following C. parvum infection by Northern blot (Figure 2B).No positive signal for the above human miRNAs was detected in C. parvum RNA using the probes or primers for miRNA real-time PCR (data not shown) and Northern blot (Figure 2C), demonstrating the specificity of these probes for human miRNAs.Downregulation of selected miRNAs induced by C. parvum, including miR-98, miR-320 and miR-424, was further confirmed by bead-based multiplexed miRNA expression assay using the FlexmiR TM Select kit (Figure 2D).For those miRNAs that did not show significant alterations in cells following C. parvum infection as revealed by the microarray analysis, we selected miR-326 for bead-based multiplexed analysis and no change was detected in C. parvum infected cells (Figure 2D), further confirming the accuracy of the array data.
Database analysis of upregulated miRNAs in cholangiocytes following C. parvum infection reveals potential NF-kB binding sites in their promoter elements Differential alterations in the mature miRNA expression profile of C. parvum-infected H69 cells suggest that miRNA gene expression is finely controlled in epithelial cells in response to C. parvum infection.One potential mechanism for selectively altering miRNA levels is through activation of distinct intracellular signaling pathways and nuclear transcription factors [15,16].This mechanism is consistent with our previous data demonstrating that C. parvum infection activates the NF-kB pathway in cholangiocytes through microbial recognition of TLR4 and TLR2 [28].We hypothesized that activation of the NF-kB pathway is involved in the transcription of select miRNAs upregulated by C. parvum.Based on TFSEARCH (http://www.cbrc.jp/research/db/TFSEARCH.html)and MOTIF (http://motif.genome.jp/)database searches [29,30], many of these miRNA genes have putative NF-kB binding sites in their potential promoter elements [31][32][33][34] (Table 1).Several miRNAs upregulated in H69 following C. parvum infection are cluster miRNAs; e.g., miR-23b, miR-27b and miR-24 are from the mir-23b-27b-24-1 gene cluster and miR-15b and miR-16 from the mir-15b-16-2 cluster [31,32].The promoters of the mir-125b-1 and mir-30b genes have not been characterized and it is unknown whether they have potential NF-kB binding sites.Transactivation of most NF-kB-dependent genes requires NF-kB p65 binding to the promoter [8] and nuclear translocation of p65 was demonstrated following C. parvum infection of cholangiocytes [28].Coupled with the results showing some similar changes in miRNA expression in H69 cells treated with LPS (which activates TLR4/NF-kB signaling in H69 cells), we then focused on determining whether p65 binds to the promoter and transactivates the miRNA genes upregulated by C. parvum infection.

Differential expression of primary transcripts of C. parvum-upregulated mature miRNAs in H69 cells
We then analyzed the kinetics of alterations of the primary transcripts (pri-miRNAs) for select mature miRNAs upregulated by C. parvum as listed in Table 1.H69 cells were exposed to C. parvum for various periods of time and pri-miRNAs of interests were quantified by real-time PCR (primers listed in Table S2).Expression of pri-miR-125b-1, pri-miR-21, pri-miR-23b-27b-24-1, pri-miR-30b, pri-miR-30c-1, pri-miR-15a-16-1, and pri-miR-15b-16-2 showed a time-dependent increase in cells following C. parvum infection, with a peak at 8 h or 12 h after exposure to the parasite (Figure 3).In contrast, no significant increase of pri-miR-125b-2 and pri-miR-30c-2 was detected in cells after exposure to C. parvum infection (Figure 3), suggesting a differential expression of the primary transcripts of C. parvum-upregulated miRNAs.

Promoter binding of NF-kB p65 subunit is required for the transcription of select miRNA genes induced by C. parvum in H69 cells
To test whether NF-kB p65 subunit is involved in C. parvuminduced transactivation of pri-miR-125b-1, we exposed H69 cells to C. parvum infection in the presence of SC-514, an IKK2 inhibitor that prevents p65-associated transcriptional activation of the NF-kB pathway [35].SC-514 blocked the C. parvum-induced increase of pri-miR-125b-1 (Figure 4A).To further test the potential transactivation of mir-125b-1 gene by p65 subunit, rapid amplification of 59 complementary DNA ends (59-RACE) PCR was used to identify the 59 end of pri-miR-125b-1.Primers were designed to amplify pri-miR-125b-1 based on the sequence obtained from the Sanger miRNA Registry (Table S2).Database analysis revealed two potential p65 binding sites in the upstream Table 1.Analysis of C. parvum-upregulated miRNAs in cholangiocytes reveals potential transactivation of their genes by NF-kB.sequence of mir-125b-1 (Figure 4B).Increased binding of p65 to the binding site at 21059, but not the putative binding site at 22455, in the promoter element of mir-125b-1 gene (Figure 4B) was demonstrated by chromatin immunoprecipitation (ChIP) analysis using specific primers for each putative binding site (Table S2).C. parvum-induced transactivation of the mir-125b-1 gene by p65 was further confirmed by using luciferase reporter gene constructs that spanned the mir-125b-1 promoter (Figure 4C).C. parvum infection increased luciferase activity in cells transfected with the luciferase constructs that encompassed the binding site for p65 at 21059 in the promoter region of mir-125b-1 gene.A mutant of the p65 binding site at 21059 blocked C. parvuminduced luciferase activity.In addition, SC-514 significantly inhibited the increase of luciferase activity induced by C. parvum infection (Figure 4C).Moreover, p65-associated transactivation of the mir-125b-1 promoter was also confirmed by the upregulation of luciferase activity after p65 overexpression in the cells (Figure 4D).As an additional control, we analyzed IL-8 transactivation, a p65dependent process induced by C. parvum in epithelial cells [36].NF-kB p65-dependent increase of IL-8 mRNA expression and binding of p65 to the promoter of IL-8 gene in cells exposed to C. parvum were confirmed (Figure S2).Together, these data demonstrate that p65 binding to the promoter element of the mir-125b-1 gene mediates mir-125b upregulation in H69 cells in response to C. parvum infection.The dynamics of p65 nuclear translocation were confirmed by Western blot analysis of p65 in the nuclear extracts from H69 cells following C. parvum infection (Protocol S1 and Figure S3), correlated to the kinetics of C. parvum-induced expression of pri-miRNAs in cells (Figure 3).Consistent with previous results, maximal p65 translocation was observed at 8 h after exposure to C. parvum [5].

Functional inhibition of selected p65-dependent miRNAs in cholangiocytes increases C. parvum infection burden
To test whether miRNAs are involved in cholangiocyte defense responses against C. parvum infection, we assessed parasite burden over time in cultured cholangiocytes transfected with various anti-miRs thereby inhibiting function of specific C. parvum-upregulated miRNAs.Anti-miRs (anti-miR TM miRNA inhibitors) are commercially available, chemically modified single stranded nucleic Expression of pri-miRNAs corresponding to C. parvum-upregulated mature miRNAs was quantified by real-time PCR.These showed significant upregulation in cells following C. parvum infection were presented as ''+'' and those without significant increase were presented as ''2''.Effects of NF-kB inhibitor SC-514 on C. parvuminduced upregulation of pri-miRNAs were also assessed by real-time PCR and presented as ''+'' (if the inhibitory effect was significant) and ''2'' (if not significant).
Putative promoter element for each miRNA gene was either based on previous studies as referred in Table 1  acids designed to specifically bind to and inhibit endogenous miRNAs [21].Cells were transfected with specific anti-miRs (30 nM, Ambion) or a mixture of anti-miRs to miR-125b, miR-23b and miR-30b (a total of 30 nM with 10 nM for each), and then exposed to C. parvum.Following incubation with a constant number of C. parvum sporozoites for 2 h to allow sufficient host-cell attachment and cellular invasion [3,24], cells were washed with culture medium to remove non-attached and non-internalized parasites.Cells were then cultured for an additional 2 h or 22 h.Parasite burden was assessed in the samples using a real-time PCR approach as we previously reported [24].The parasite burden following exposure to C. parvum for 2 h was similar in all cultures, including those transfected with the siRNA to Drosha or the specific anti-miRs (Figure 5A and 5C), suggesting that those miRNAs do not affect initial parasite host cell attachment and cellular invasion.Additionally, SC-514 treatment did not impact parasite burden at this time point (Figure 5A).Consistent with our previous studies [24], a significant increase in parasite burden was identified in SC-514-treated H69 cells 24 h after initial infection (Figure 5B).Cells transfected with the siRNA to Drosha displayed an increased parasite burden as compared to control cells (Figure 5B).Interestingly, we also detected a significantly higher parasite burden 24 h after initial infection in cells treated with the anti-miRs to miR-125b, miR-23b, and miR-30b, as well as a mixture of three anti-miRs, compared with that in control cells (Figure 5D); anti-miRs to miR-16 and miR-21 did not impact infection burden (Figure 5D).Increase of parasite burden 24 h after initial infection in H69 cells treated with SC-514 or select anti-miRs was further confirmed by immunofluorescent microscopy (Figure 5E).The targets of a majority of known miRNAs are still yet to be identified.C. parvum-responsive miRNAs may regulate the expression of proteins of various functions related to epithelial anti-C.parvum defense.Using computational analyses as previously reported [19,[37][38][39], we identified a variety of potential targets of C. parvum-responsive miRNAs selected on the basis of their known involvement in immune related responses (Table S3).

Discussion
There is emerging evidence that miRNAs play a critical role in the regulation of both innate and adaptive immunity [17][18][19].A better understanding miRNA expression changes in epithelial cells following C. parvum infection will provide new insights in miRNAassociated epithelial defense to C. parvum.Using an in vitro model of human biliary cryptosporidiosis, we report significant alterations in miRNA expression profiles in epithelial cells following C. parvum infection.Our analysis of miRNAs upregulated by C. parvum in H69 cells revealed that mir-125b-1, mir-23b-27b-24-1, mir-21, and mir-30b genes are transactivated via potential promoter binding of the NF-kB p65 subunit.These data provide several insights relevant to miRNA expression regulation in cholangiocytes following C. parvum infection.First, similar to the regulation of miRNA genes in other cells [16,40,41], promoter binding of transcription factors regulates miRNA genes in epithelial cells in response to C. parvum infection.Therefore, transcription factor-mediated miRNA expression and subsequent gene regulation at the posttranscriptional level through miRNA targeting may be an important element of host responses against C. parvum infection.Since similar alterations in miRNA expression profile were identified in LPS-treated cells, this observation may also be relevant to cellular gene regulation in general.Second, transactivation of miRNA genes that produce the same mature miRNA can be differentially controlled.Specifically, both mir-125b-1 and mir-125b-2 genes can produce mature miR-125b, but only transactivation of mir-125b-1 gene was detected in cells following C. parvum infection.Indeed, differential activation of genes for the same mature miRNA molecule has been previously reported [31].Finally, transactivation of genes of cluster miRNAs or as introns in other gene alleles may be controlled by the same promoter element.Of note, miR-23b, miR-27b and miR-24 are cluster gene miRNAs and co-transcribed with a host gene, C9orf3 [31].C. parvum infection upregulates expression of the mature forms of these three miRNAs, as well as pri-miR-23b-27b-24-1 and the host gene transcript, C9orf3.Our data are consistent with recent studies demonstrating transcriptional control of genes that code cluster miRNAs or that encode both miRNAs and other host transcripts [31,32].
The NF-kB family of transcription factors consists of five members, p50, p52, p65 (RelA), c-Rel, and RelB [8].The transcription activation domain (TAD) necessary for the positive regulation of gene expression is present only in p65, c-Rel, and RelB [8].Thus, promoter binding of p65, c-Rel and RelB is usually associated with gene transactivation [8,[42][43][44].Because they lack TADs, p50 and p52 may repress transcription unless they associate with a TAD-containing NF-kB family member or another protein capable of coactivator recruitment [8,45,46].Increased nuclear translocation of p65 and p50 was previously reported in human cholangiocytes following C. parvum infection [28].In this study, we demonstrated that promoter binding of the NF-kB p65 subunit is required for transactivation of the mir-125b-1, mir-23b-27b-24-1, mir-21 and mir-30b genes in cells following C. parvum infection.Although transactivation of mir-30c-1 and mir-15b-16-2 genes was observed in C. parvum-infected cells and potential NF-kB binding sites were identified in their promoter elements, inhibition of p65 activation failed to inhibit transactivation of either mir-30c-1 or mir-15b-16-2 in H69 cells following C. parvum-infection.In addition, miR-146b, miR-155, and miR-9 have been reported to be NF-kB-dependent miRNAs in monocytes or neutrophils [15,47,48].Although miR-146b and miR-155 are expressed in cholangiocytes, no upregulation of either miR-146b or miR-155 was detected in H69 cells following C. parvum infection.Given the complexity and variability in the gene structure for each miRNA, it is obvious that multiple mechanisms are involved in the transcriptional regulation of human miRNA genes [32,33,49].Therefore, transcription of miRNA genes is expected to be a dynamic process in response to the constant alterations in intracellular signals.miRNA expression thus reflects the final integrated result of multiple interrelated signals on miRNA transcription.In this regard, other transcription factors, such as AP-1, c-myc, C/EBPa, may also be involved in the transcriptional regulation of miRNA genes in epithelial cells in response to C. parvum infection.Future studies will focus on whether nuclear translocation of p50 is involved in the C. parvuminduced down-regulation of miRNA expression.
miRNAs have been identified in both mammalian and nonmammalian cells including virus and parasites [9,14,50,51].Expression of miRNAs in C. parvum has not yet been examined and whether C. parvum-derived miRNAs can be localized in infected host cells is unknown.Nevertheless, the probes used in the microarray analysis in this study are human-miRNA specific with minimal cross-interaction with known miRNAs from other species.Cells of sham-infection control (host cells plus heat-inactivated C. parvum oocysts) displayed a similar expression profile of human miRNAs compared with non-infected control cells as assessed by microarray analysis.Finally, by Northern blot and real-time PCR, no positive signal was detected in C. parvum RNA alone using the probes/primers for selected human miRNAs, confirming host-cell specificity of detected miRNAs.The TLR/NF-kB signaling is critical to innate epithelial immune defenses to microbial infection including parasites [4,52].We previously demonstrated that TLR4 and TLR2 are involved in cholangiocyte immune response to C. parvum infection via activation of NF-kB [5].Here, we expanded our previous studies by demonstrating that miRNAs may also regulate TLR/ NF-kB-mediated epithelial anti-C.parvum defense.We indentified a panel of miRNA genes that are transactivated via p65 promoter binding in cholangiocytes in response to C. parvum infection.Transfection of cells with anti-miRs to miR-125b, miR-23b or miR-30b, but not anti-miRs to miR-16 or miR-21, significantly increased parasite burden in cholangiocytes.The molecular mechanisms by which C. parvum-responsive miRNAs modulate epithelial anti-C.parvum defense are largely unclear.Previous studies demonstrated that let-7 regulates TLR4 expression and is involved in epithelial defense against C. parvum [24].Various immune related genes are identified as potential targets for these C. parvum-responsive miRNAs using computational analyses.The concept that a pathogen encodes mRNAs targeted by host miRNAs has recently emerged as an important mechanism of host anti-viral defense [21].Likewise, it is of interest to test the possibility that host cell miRNAs target the internalized parasite mRNAs and silence genes of the pathogen.The direct C. parvumhost cell cytoplasmic tunnel-connection [53] could mediate exchange of molecules, including miRNAs, between the host cells and internalized parasite.Further investigation should test whether p65 promoter binding transactivates LPS-responsive miRNA genes.This also raises the possibility that transactivation of miRNA genes through promoter binding of NF-kB subunits may be involved in host anti-microbial responses in general.
In summary, this first miRNA profiling in cholangiocytes in response to C. parvum infection in vitro revealed significant alterations in cellular miRNA expression.The mechanism by which C. parvum induces upregulation of a panel of miRNAs in cholangiocytes involves transactivation of miRNA genes through promoter binding of the NF-kB p65 subunit.In addition, functional inhibition of the upregulated miRNAs increases C. parvum infection burden in cholangiocytes in vitro thereby implicating these miRNAs in host cell defense to the parasite.These data demonstrate a key role for miRNAs in epithelial immune responses against C. parvum infection and may provide new insights into general mechanisms of the regulation of epithelial anti-microbial immunity.

C. parvum and human cholangiocyte cell line
C. parvum oocysts of the Iowa strain were purchased from a commercial source (Bunch Grass Farm, Deary, ID).H69 cells (a gift of Dr. D. Jefferson, Tufts University) are SV40 transformed normal human cholangiocytes originally derived from liver harvested for transplant.These cholangiocytes continue to express biliary epithelial cell markers, including cytokeratin 19, gamma glutamyl transpeptidase and ion transporters consistent with biliary function and have been extensively characterized [26].

In vitro infection model and infection assay
An in vitro model of human biliary cryptosporidiosis using H69 cells was employed in these studies.Before infecting cells, C. parvum oocysts were treated with 1% sodium hypochlorite on ice for 20 min followed by extensive washing with DMEM-F12 medium.Oocysts were then added to the cell culture to release sporozoites to infect cells [54].Infection was performed in culture medium (DMEM-F12 with 100 U/ml penicillin and 100 mg/ml strepto-mycin) containing viable C. parvum oocysts (oocysts with host cells in a 5:1 ratio).Inactivated organisms (treated at 65uC for 30 min) were used for sham infection controls.All experiments were performed in triplicate.For the inhibition experiments, SC-514 (Calbiochem) was added to the medium.Cells were pre-treated with SC-514 for 1 h prior to C. parvum infection.SC-514 was used at a concentration of 100 mM, which showed no cytotoxic effects on H69 cells or on C. parvum sporozoites, in these studies.
Real-time PCR and immunofluorescent microscopy were used to assay C. parvum infection as previously reported [24].Briefly, primers specific for C. parvum 18s ribosomal RNA (forward: 59-TAGAGATTGGAGGTTGTTCCT-39 and reverse: 59-CTCCA-CCAACTAAGAACGGCC-39) were used to amplify the cDNA specific to the parasite.Primers specific for human plus C. parvum 18s were used to determine total 18s cDNA [24].Data were expressed as copies of C. parvum 18s vs total 18s.For immunofluorescent microscopy, cells were fixed with 2% paraformaldehyde and incubated with a polyclonal antibody against C. parvum (a gift from Dr. Guan Zhu, Texas A&M University) followed by antirabbit FITC-conjugated secondary antibody (Molecular Probes) and co-staining with 49, 6-diamidino-2-phenylindole (DAPI, 5 mM) to stain cell nuclei.Labeled cells were assessed by confocal laser scanning microscopy.

miRCURY TM LNA array analysis of miRNAs
The Exiqon (Vedbaek, Denmark) miRCURY LNA microRNA arrays and service to process the samples were used [27].Briefly, H69 cells were grown to 80% confluence and exposed to C. parvum oocysts for 12 h or LPS (1 mg/ml) for 8 h.Total RNAs from H69 cells or C. parvum oocysts were prepared with the mirVana TM miRNA Isolation Kit according to the manufacturer's instruction (Ambion).The quality of isolated RNAs was verified by an Agilent 2100 Bioanalyzer profile (Figure S1).A mixture of equal amounts of total RNAs from the control and C. parvum-infected cells were used as the reference pool.A total of 2 mg RNA from each sample was then labeled with the Hy5 TM fluorescent label and the reference pool labeled with Hy3 TM using the miRCURY TM LNA Array labeling kit (Exiqon).The labeled samples and reference pool were then mixed pair-wise and hybridized to the miRCURY TM LNA array containing capture probes targeting all human miRNAs listed in the miRBASE version 8.1 (Exiqon).After hybridization, the slides were scanned and quantified signals normalized by Exiqon using the global Lowess (Locally Weighted Scatterplot Smoothing) regression algorithm.Normalized Hy5/Hy3 ratios were used for further analysis as previously reported [55][56][57].

Bead-based multiplex sandwich immunoassays
A bead-based multiplex sandwich immunoassay, read with a Luminex 200 system (Luminex), was used to measure the concentrations of selected miRNAs as previously reported [57].Briefly, total cellular RNAs are isolated using the mirVana TM miRNA Isolation Kit (Ambion).An amount of 0.5 mg of total RNAs was used for Biotin-labeling using the FlexmiR MicroRNA Labeling Kit for selected miRNAs (Luminex).Signals for miRNAs were recorded and standardized to the standard beads according to the manufacturer's instructions (Luminex).

Real-time PCR
For real-time PCR analysis of mature miRNAs, total RNAs were extracted using the mirVana TM miRNA Isolation kit (Ambion).An amount of 0.05 mg total RNAs was reverse-transcribed by using the Taqman MicroRNA Reverse Transcription Kit (Applied Biosystems).Comparative real-time PCR was performed in triplicate using Taqman Universal PCR Master Mix (Applied Biosystems) on the Applied Biosystems 7500 FAST real-time PCR System.Mature miRNA-specific primers and probes were obtained from Applied Biosystems.Normalization was performed by using RNU6B primers and probes.Relative expression was calculated by using the comparative CT method [56].
For analysis of pri-miRNAs, total RNA was isolated from cells with Trizol reagent (Ambion).RNAs were treated with DNAfree TM Kit (Ambion) to remove any remaining DNA.Comparative real-time PCR was performed by using the SYBR Green PCR Master Mix (Applied Biosystems).Specific primers for pri-miRNAs were listed in Table S2.All reactions were run in triplicate.The Ct values were analyzed using the comparative Ct (DDCt) method and the amount of target was obtained by normalizing to the endogenous reference (GAPDH) and relative to the control (non-treated cells) [58].

Northern blot
Total RNAs harvested as above were run on a 15% Tris/ Borate/EDTA (90 mM Tris/64.6 mM boric acid/2.5 mM EDTA, pH 8.3)-urea gel (Invitrogen) and transferred to a Nytran nylon transfer membrane (Ambion).LNA DIG-probes for selected miRNAs (Exiqon) were hybridized using UltraHyb reagents (Ambion) according to the manufacturer's instructions with blotted snRNA RNU6B as a control.

59-RACE PCR
59-RACE PCR was utilized to identify 59 end of miRNA primary transcripts to localize the start sites of mir-125b-1, mir-30b and mir-30d.Primer sequences are listed in Table S2.The SMART TM RACE cDNA Amplification Kit (Clontech) was used for the analysis.Total RNA was isolated for H69 cells treated with a Drosha siRNA (Santa Cruz biotechnology) as previously reported [32].

ChIP
ChIP analysis was performed with a commercially available ChIP Assay Kit (Upstate Biotechnologies) in accordance with the manufacturer's instructions.In brief, 1610 6 H69 cells were cultured in 15-cm culture dishes and exposed to C. parvum in the presence or absence of SC-514 for 8 h.The chromatin fraction was immunoprecipitated for overnight at 4uC using anti-NF-kB p65 (Upstate Biotechnologies).PCR amplification was performed in a total volume of 25 ml with specific primers.The forward and reverse primers used for each gene were listed in Table S2.

Luciferase reporter constructs and luciferase assay
Promoters of miRNAs were amplified by PCR from human genomic DNA.PCR primers were listed in Table S2.The PCR products were separated by agarose gel electrophoresis, and the DNA fragments then isolated and cloned in the restriction enzyme digested pGL3 Basic Vector (Promega) using T4 DNA ligase (Fisher scientific).All constructs were confirmed by sequencing.Mutations were introduced into the NF-kB binding sites using the QuikChange site-directed mutagenesis kit (Stratagene).H69 cells were transfected with each reporter construct for 24 h and then exposed to C. parvum oocysts for 8 h in the presence or absence of SC-514 followed by assessment of luciferase activity.Luciferase activities were then measured and normalized to the control b-gal level.The luciferase activity of each construct was compared with that of the promoterless pGL3 basic vector.

Supporting Information
Table S1 miRNA expression profile in cholangiocytes following C. parvum infection and LPS stimulation.Data represent the mean6SE of the log 2 (Hy5/Hy3) ratios from non-infected cell cultures (n = 3), C. parvum infected cultures (n = 3), LPS treated cultures (n = 3), and one cell culture exposed to heated-inactivated C. parvum (Sham) by using the miRCURY TM LNA Array (Version 8.1). a , p, = 0.05; b , 0.05,p, = 0.20, compared with noninfected cells; NA = not detectable.Found at: doi:10.1371/journal.ppat.1000681.s001(0.09 MB PDF) Table S2 Primers used for PCR and construct generating.Listed in this table are all the primers used in this study for the real-time PCR and RACE PCR, as well as those for ChIP analysis and construct generating. a Restriction enzyme sites were indicated by lowercase letters.Found at: doi:10.1371/journal.ppat.1000681.s002(0.02 MB PDF) Table S3 Prediction of immune-related target genes of C. parvum-responsive miRNAs.Prediction of immune-related target genes for C. parvum-responsive miRNAs was performed with the computerlized predictive algorithms as previously reported [19,[37][38][39].Some of the predicted targets have been experimentally confirmed [24,25,27,59,60] and the corresponding miRNAs are in red font. .parvum-induced increase of pri-miR-30b, suggesting p65-dependent miR-30b expression.(B) We performed 59-RACE PCR to identify the 59end of pri-miR-30d and identified a potential p65 binding site at 2472 of its upstream sequence.H69 cells were transfected with the luciferase gene reporter construct covering the potential p65 binding site within the putative promoter of mir-30d and then exposed to C. parvum.These results support that pri-miR-30b and pri-miR-30d are not transcribed from the same gene in human cholangiocytes, inconsistent with previous results suggesting that pri-miR-30b and pri-miR-30d may be transcribed from the same gene on chr8 [31,40].(C) To clarify how p65 is involved in the transactivation of miR-30b gene transactivation, we performed 59-RACE PCR but failed to amplify the corresponding sequence (data not shown).Nevertheless, database analysis revealed one potential binding site for NF-kB in the upstream sequence of miR-30b precursor.ChIP analysis detected an increased binding of p65 to this region in cells following C. parvum infection.(D) Luciferase reporter gene analysis demonstrated a significant increase in luciferase reporter activity in cells following C. parvum infection or overexpressed with p65.*, p,0.05 vs. the non-infected control (in A and D) or empty pCMV vector control (in E); #, p,0.05 vs. C. parvum infected cells (in A and D).Found at: doi:10.1371/journal.ppat.1000681.s009(0.73 MB TIF) Figure S7 p65-independent expression of miR-30c and miR-16 in cholangiocytes in response to C. parvum infection.(A and B) miR-30c is transcribed from two genes, mir-30c-1 and mir-30c-2, localized on chr1 and chr6, respectively [31].Real-time PCR analysis revealed an increase of pri-miR-30c-1(A), but not pri-miR-30c-2 (B), in H69 cells following C. parvum infection.Treatment of cells with SC-514 failed to block C. parvum-induced expression of pri-miR-30c-1 (A).(C and D) miR-16 is transcribed from two genes, mir-15a-16-1 and mir-15b-16-2 localized on chr13 and chr3 and clustered with miR-15a and miR-15b, respectively [31].Increased expression of pri-miR-15a-16-1 (at 12 h; C) and pri-miR-15b-16-2 (at 8 h and 12 h; D) was detected in H69 cells after C. parvum infection.Treatment of cells with SC-514 failed to block either pri-miR-15a-16-1 (C) or pri-miR-15b-16-2 (D).*, p,0.05 vs. the non-infected control.

Figure 1 .
Figure 1.Expression profiling of mature miRNAs in cholangiocytes following C. parvum infection and LPS stimulation.(A) miRNA expression profile in H69 cells following C. parvum infection.The left panel shows a heat-map of selected miRNAs that showed changes in expression in H69 cells following C. parvum infection.The horizontal axis indicates samples of non-infected cells (n = 3; Control-1, -2, and -3) and cells after exposure to live C. parvum for 12 h (n = 3, C. parvum-1, -2, and -3).The right panel shows expression of miRNAs in H69 cells following C. parvum infection.Cellular levels of miRNAs were presented as the log 2 (Hy5/Hy3) ratios which passed the filtering criteria variation across the samples.p values are from the t' test.hsa = Homo sapiens.(B) Comparison of miRNA expression patterns in H69 cells following C. parvum infection for 12 h and LPS stimulation for 8 h.Graphics indicate those miRNAs showing an increased or decreased expression (including those significant change when p, = 0.05 and those with a tendency to change when 0.05,p, = 0.20) in cells after treatment with LPS (n = 3) or exposure to C. parvum (n = 3).A complete description of miRNA expression profiles in cells was listed in Table S1.doi:10.1371/journal.ppat.1000681.g001

Figure 2 .
Figure 2. Altered expression of selected miRNAs confirmed by real-time PCR, Northern blot and Luminex bead analyses.(A) Alterations of selected miRNA expression in cells after exposure to C. parvum for various periods of time as assessed by real-time PCR.The amount of mature miRNAs was obtained by normalizing to the level of snRNA RNU6B in the samples.Data are expressed as the amount of mature miRNAs in the infected samples relative to the control uninfected samples and representative of three independent experiments.(B) Alterations of selected miRNA expression in cells after exposure to C. parvum for 12 h as determined by Northern blot.snRNA RNU6B was used as a control to ensure equal loading.Representative Northern blots (C.parvum infected cells vs non-infected control) from three independent experiments are shown.(C) Total RNA isolated from C. parvum oocysts was also blotted to demonstrate the specificity of the probes.(D) Alterations of selected miRNA expression in cells after exposure to C. parvum for 12 h as assessed by bead-based miRNA Luminex analysis.The amount of mature miRNAs was obtained by normalizing the samples to the positive control beads provided by the company (Luminex).Data are representative of three independent experiments.*, p,0.05 vs. the non-infected control.doi:10.1371/journal.ppat.1000681.g002

Figure 3 .
Figure 3. Differential expression of primary transcripts of C. parvum-upregulated mature miRNAs in H69 cells.H69 cells were exposed to C. parvum for 2 h to 24 h and primary transcripts (pri-miRNAs) of select miRNAs were quantified by real-time PCR.The amount of pri-miRNAs was obtained by normalizing to the level of GAPDH in the samples.Data are expressed as the amount of pri-miRNAs in the infected samples relative to the control uninfected samples and representative of three independent experiments.*, p,0.05 vs. the non-infected control.doi:10.1371/journal.ppat.1000681.g003

Figure 4 .
Figure 4. Promoter binding of p65 transactivates miR-125b-1 gene to increase miR-125b expression following C. parvum infection.(A) p65-dependent upregulation of pri-miR-125b-1 in cholangiocytes following C. parvum infection.Data are presented as the relative expression level of pri-miR-125b-1 in H69 cells following C. parvum infection in the presence or absence of SC-514 as assessed by real-time PCR.(B) C. parvum increases promoter element binding of p65 to mir-125b-1 gene.The schematic diagram shows two potential binding sites in the putative promoter element of mir-125b-1 gene.ChIP analysis revealed increased binding of p65 to the binding site at 21059, but not at 22455, of mir-125b-1 promoter element in cells following infection.Representative ChIP gels are shown in the upper panel and densitometry analysis of the gels in the lower panel.(C) H69 cells were transfected with various luciferase reporter constructs spanning the potential p65 binding sites of the mir-125b-1 promoter.The transfected cells were exposed to C. parvum in the presence or absence of SC-514.Luciferase activity was measured and presented as the ratio of the activity of the test construct with the control luciferase reporter construct.Six reporter constructs containing the mutants of the two potential NF-kB binding sites were also utilized for the analysis as indicated.(D) H69 cells were co-transfected with the pCMV-p65 to overexpress p65 and the luciferase reporter construct containing the mir-125b-1 promoter for 24 h followed by measurement of luciferase activity.*, p,0.05 vs. the non-infected control (in A and C) or empty pCMV vector control (in D); # , p,0.05 vs. C. parvum infected cells (in A and C).doi:10.1371/journal.ppat.1000681.g004 or identified by RACE-PCR in this study (indicated by asterisks).Potential NF-kB binding site(s) in the promoter region was identified by the TFSEARCH (http://www.cbrc.jp/research/db/TFSEARCH.html) and MOTIF (http:// motif.genome.jp/).Positive promoter binding of p65 subunit to the predicted NF-kB binding site was confirmed by ChIP analysis and marked as ''+''; otherwise marked as ''2'' if no p65 binding was detected.C. parvum-induced transactivation of miRNA gene by p65 was further confirmed by using luciferase reporter gene constructs that spanned the promoter region of each individual gene.If C. parvum infection increased luciferase activity in cells transfected with the luciferase constructs containing the binding site for p65 and this induction was blocked by SC-514, it was presented as ''+''; otherwise presented as ''2''.NS = not selected for further ChIP analysis or luciferase reporter assay in this study.In such cases, these genes are most likely not regulated by p65 promoter binding because their transactivation was not induced by C. parvum infection or the induced transactivation was not inhibited by SC-514 as assessed by real-time PCR.Refer to Figures S4, S5, S6 and S7 for details.doi:10.1371/journal.ppat.1000681.t002

Figure 5 .
Figure 5. Functional inhibition of selected p65-dependent miRNAs in cholangiocytes increases C. parvum infection burden in vitro.(A) A similar number of parasites was detected in cells transfected with Drosha siRNA or treated with SC-514 after initial exposure to C. parvum for 2 h as quantified by real-time PCR.(B) Transfection of cells with Drosha siRNA or treated with SC-514 increased C. parvum infection burden in cholangiocytes in vitro 24 h after initial exposure to the parasite.(C) Effects of anti-miRs on C. parvum after initial exposure to C. parvum for 2 h.(D) Transfection of cells with anti-miRs on C. parvum infection burden in cholangiocytes 24 h after initial exposure to the parasite.*, p,0.05 vs. nontreated cells or cells transfected with a control siRNA (in B) or non-specific control anti-miR (in D). (E) Effects of anti-miRs or SC-514 on C. parvum burden in cholangiocytes in vitro 24 h after initial exposure to the parasite as assessed by immunofluorescent microscopy.C. parvum parasites were stained in green and nuclei in blue.Bars = 5 mm.doi:10.1371/journal.ppat.1000681.g005 Figure S1 Quality control of RNA.Total RNAs from cells were prepared with the mirVana TM miRNA Isolation Kit according to the manufacturer's instructions (Ambion).The quality of the isolated RNAs was verified by examining the Agilent 2100 Bioanalyzer profile of the sample.Representative RNA profiles from non-infected H69 cells (A), cells exposed to live (B) and heatinactivated C. parvum oocysts (C) are shown.Found at: doi:10.1371/journal.ppat.1000681.s004(1.62 MB TIF) Figure S2 Promoter binding of p65 transactivates the IL-8 gene in cholangiocytes in response to C. parvum infection.(A) p65dependent upregulation of IL-8 mRNA in cholangiocytes following C. parvum infection.Bars represent the levels of IL-8 mRNA in cells following C. parvum infection in the presence or absence of SC-514 as assessed by real-time PCR.(B) A schematic diagram shows the structure of IL-8 gene.ChIP analysis demonstrated increased binding of p65 to the binding site at IL-8 promoter in cells following infection.*, p,0.05 vs. non-infected cells; #, p,0.05 vs. C. parvum infected cells.Found at: doi:10.1371/journal.ppat.1000681.s005(0.28 MB TIF) Figure S3 Nuclear translocation of p65 in cholangiocytes cells induced by C. parvum.Cells were exposed to C. parvum for various periods of time and nuclear extracts obtained as described in Protocol S1.The NF-kB p65 subunit was detected by Western blot.Actin was used as a loading control.Representative Western blots are shown.Found at: doi:10.1371/journal.ppat.1000681.s006(0.61 MB TIF) Figure S4 Promoter binding of p65 transactivates mir-21 gene to increase miR-21 expression in biliary epithelial cells in response to C. parvum infection.(A) p65-dependent upregulation of pri-miR-21 in cholangiocytes following C. parvum infection.A schematic diagram illustrates the structure of the mir-21 gene that encodes human miR-21.Bars represent the expression levels of pri-miR-21 in cells following C. parvum infection in the presence or absence of SC-514 as assessed by real-time PCR.(B) C. parvum increases promoter binding of p65 to the mir-21 gene.ChIP analysis revealed increased binding of p65 to the promoter binding site at +1395, but not at +1167 in H69 cells following infection.(C) H69 cells were transfected with various luciferase reporter constructs covering the potential binding sites of the mir-21 promoter and then exposed to C. parvum in the presence or absence of SC-514.A FigureS7p65-independent expression of miR-30c and miR-16 in cholangiocytes in response to C. parvum infection.(A and B) miR-30c is transcribed from two genes, mir-30c-1 and mir-30c-2, localized on chr1 and chr6, respectively[31].Real-time PCR analysis revealed an increase of pri-miR-30c-1(A), but not pri-miR-30c-2 (B), in H69 cells following C. parvum infection.Treatment of cells with SC-514 failed to block C. parvum-induced expression of pri-miR-30c-1 (A).(C and D) miR-16 is transcribed from two genes, mir-15a-16-1 and mir-15b-16-2 localized on chr13 and chr3 and clustered with miR-15a and miR-15b, respectively[31].Increased expression of pri-miR-15a-16-1 (at 12 h; C) and pri-miR-15b-16-2 (at 8 h and 12 h; D) was detected in H69 cells after C. parvum infection.Treatment of cells with SC-514 failed to block either pri-miR-15a-16-1 (C) or pri-miR-15b-16-2 (D).*, p,0.05 vs. the non-infected control.Found at: doi:10.1371/journal.ppat.1000681.s010(0.58 MB TIF) Protocol S1 Nuclear translocation of p65.Found at: doi:10.1371/journal.ppat.1000681.s011(0.01 MB PDF)

Table 2 .
Promoter binding of NF-kB p65 subunit in C. parvum-induced transactivation of miRNA genes in H69 cells.