|Home | About | Journals | Submit | Contact Us | Français|
Biliary cryptosporidiosis is associated with acquired immunodeficiency syndrome (AIDS) cholangiopathy and occurs almost exclusively in adult patients with AIDS. Infection of biliary epithelial cells (cholangiocytes) with Cryptosporidium parvum induces Toll-like receptor (TLR) 4 expression and stimulates a TLR-dependent response against infection. Here, we tested whether human immunodeficiency virus type 1 (HIV-1) Tat affects TLR expression and, hence, anti–C. parvum defense responses. Using an in vitro model of human biliary cryptosporidiosis, we found that recombinant Tat protein increased TLR4 mRNA expression in both uninfected and C. parvum–infected cholangiocytes. Conversely, Tat decreased TLR4 protein levels and suppressed C. parvum–induced TLR4 protein expression. Using actinomycin to inhibit transcription, we found that Tat increased the half-life of TLR4 mRNA from ~25 to 60 min, and RNA gel-shift assays demonstrated direct binding of Tat to TLR4 mRNA. In vitro transcription/translation studies suggested that Tat does not affect transcription but does decrease TLR4 translation. Importantly, more parasites were found in Tat-treated cells than in control cells 48h after infection. These findings suggest that Tat inhibits cholangiocyte TLR4protein expression through translational inhibition. These events appear to diminish the ability of cholangiocytes to initsiate an innate immune response to C. parvum. We suggest that these findings may contribute to the unusual susceptibility of HIV-infected individuals to biliary cryptosporidiosis.
Cryptosporidium parvum is an enteric pathogen and a common cause of gastroenteritis in humans. In immunocompetent individuals, infection is limited to the intestine and causes acute, self-limited disease. However, in those who are immunosuppressed, cryptosporidiosis may cause potentially fatal complications, including bile duct damage [1–5]. Although biliary cryptosporidiosis may occur in young men with the rare X-linked hyper-IgM syndrome [6, 7], among adults, it has been reported only in patients with AIDS [1–5]. AIDS cholangiopathy is a group of biliary disorders (including secondary sclerosing cholangitis) [8, 9] that are caused by opportunistic infection of the biliary tree, most commonly with C. parvum. Although the clinical features of this syndrome are well documented, it is not clear why individuals infected with HIV are almost uniquely susceptible to biliary cryptosporidiosis, in contrast to individuals immunocompromised via other mechanisms.
Both innate and adaptive immunity are implicated in the resolution of and resistance to C. parvum infection [1, 10]. We previously demonstrated that C. parvum infection induces human β-defensin (hBD) 2 expression in cultured human biliary epithelial cells (cholangiocytes) via Toll-like receptor (TLR)–associated activation of NF-κB . Furthermore, we demonstrated that C. parvum infection increases TLR2 and TLR4 protein expression in cholangiocytes .
HIV-1 infects hepatocytes, Kupffer cells, and endothelial cells in the liver [13, 14], but infection of cholangiocytes has not been reported. However, HIV-derived soluble peptides released from infected cells can affect uninfected cells [15–17]. One of these peptides, Tat (HIV-encoded transactivator of transcription), is taken up by many cell types, including epithelial cells , and it has been shown to inhibit salmonella invasion of the intestinal epithelial cell line HT-29  and to induce ion secretion in Caco-2 cells; it may, therefore, contribute to HIV-1–associated intestinal disease . In HIV-infected cells, Tat promotes transcription of the viral genome by interacting with the transactivation response (TAR) element present on the 5′ region of all viral mRNAs. Furthermore, it has been suggested that HIV-1 prevents RNA silencing through Tat interactions with the RNAse III molecule Dicer .
With this information as background as well as our recent studies demonstrating that TLR4 is regulated by the let-7 family of microRNAs , we proposed that Tat, by interfering with Dicer function, might promote overexpression of TLR4 and alter the normal cholangiocyte defense response against C. parvum. What our experiments actually showed was that recombinant Tat protein increased TLR4 mRNA expression in cultured human cholangiocytes but decreased TLR4 protein expression, suggesting the involvement of a posttranscriptional gene regulatory pathway. Supporting this interpretation was our finding that Tat binds directly to TLR4 mRNA, presumably to a Tat-binding motif similar to the HIV TAR sequence, a process that likely inhibits TLR4 mRNA translation. The consequence of these events appears to be a diminished host defense response, because Tat protein diminishes hBD2 expression and hampers clearance of C. parvum in an in vitro model of biliary cryptosporidiosis. Thus, we have identified a novel HIV-1 Tat–mediated regulatory pathway for TLR4 expression in cholangiocytes, which we propose may contribute to the unusual susceptibility of HIV-infected individuals to biliary cryptosporidiosis.
C. parvum oocysts were purchased from a commercial source (Bunchgrass Farms). Before cell culture infection, oocysts were treated with 1% sodium hypochlorite on ice for 20 min, excysted in 0.75% and 0.25% trypsin at 37°C, washed, and centrifuged at 720 g for 5 min. Sporozoites were tested for lipopolysaccharide contamination using the Limulus Amoebocyte Lysate Pyrogent Plus Single Test Kit (Cambrex BioScience), as reported elsewhere .
The human cholangiocyte cell line H69 is an SV40-transformed cell line derived from a normal liver . Recombinant full-length HIV-1 Tat protein (86 aa) was purchased from Immunodiagnostics, diluted in saline citrate buffer, and tested for lipopolysaccharide contamination as described above. The concentrations used here are consistent with those used in other studies addressing the physiological roles played by Tat found in tissue  and serum  of patients.
Quantitative RT-PCR (LightCycler) was performed using FastStart DNA Master SYBR Green mix (Roche). RNA was isolated using TRI reagent (Sigma-Aldrich) and reverse transcribed using an RT kit (SuperScript III; Invitrogen). Standards for each target transcript and the 18s normalizing control were generated by PCR amplification. Spectrophotometry was then used to assess the concentration of each amplicon. Copy number was determined on the basis of amplicon molecular weight. A standard curve was used to extrapolate quantitative information. The primers used for specific amplification of TLR4, TLR2, myeloid differentiation factor (MyD) 88, and hBD2 have been described elsewhere .
H69 cells were exposed to C. parvum in medium with or without 100–200 ng/mL Tat protein and incubated overnight, and quantitative immunoblots were performed as described elsewhere . Briefly, samples were separated by SDS-PAGE, transferred to nitrocellulose, incubated with 2 μg/mL primary monoclonal antibody IMG-5031A, developed against aa 100–200 of human TLR4 (Imgenex) and 0.2 μg/mL horseradish peroxidase (HRP)–conjugated secondary antibody, and revealed by electrochemiluminescence (Amersham). Actin was blotted to confirm equal loading. Immunoreactive areas were analyzed using an imaging densitometer (model GS-700) and Molecular Analyst software (Bio-Rad Laboratories).
Medium containing 5 μg/mL actinomycin D was added to H69 cells. After a 10-min incubation, recombinant Tat (50–100 ng/mL) was added. RNA was isolated at 0, 45, 90, and 120 min. Quantitative RT-PCR was performed. A model of exponential decay was used to calculate the approximate half-life of each transcript. TLR4 protein degradation rate was assessed in the presence or absence of recombinant Tat. H69 cells were treated with cycloheximide (20 μg/mL) in the presence or absence of Tat (100 ng/mL). Lysates were harvested at 0, 6, 18, and 24 h. Quantitative Western blots were performed.
In vitro transcription/translation was performed using the TNT Quick Coupled Transcription/Translation System (Promega). Briefly, 2 μg of pcDNA3.1-TLR4 plasmid (provided by M. F. Smith, University of Virginia, Charlottesville) was added to the master mix. Transcend Biotin-lysyl-tRNA (Promega) and methionine were added to the reaction and incubated at 30°C for 90 min in the absence or presence of Tat (25–50 ng/mL). The translational products were separated by SDS-PAGE, transferred to nitrocellulose, and detected using the Transcend Non-Radioactive Translational Detection System (Promega). Briefly, membranes were blocked in TBS plus 0.5% Tween 20, incubated in streptavidin-HRP conjugate (1:10,000), washed, incubated in chemiluminscent substrate, and exposed to film. Semiquantitative RT-PCR for TLR4 was also performed on the transcription/translation product.
The 5′ region of TLR4 was sub-cloned from pcDNA3.1-TLR4 into the EGFP-N2 vector (Clontech) by means of the HindIII and EcoR1 restriction sites. PCR mutagenesis was performed, resulting in a 12-bp deletion. The primers used were as follows: forward, 5′-CCCAAGCTTCCTCTCACCCTTTAGCCCA-3′; mutated reverse, 5′-GTAAACTTGATAGTCCAGAAAAGGCTCTCTGGATGGGGTTTCCTGTCAATA-3′; mutated forward, 5′-TATTGACAGGAAACCCCATCCAGAGAGCCTTTTCTGGACTATCAAGTTTAC-3′; and reverse, 5′-CCGGAATTCTTCAATGGTCAAATTGCACAG-3′. PCR amplification was performed using the 2 amplicons as template and primer, resulting in an 862-bp fragment that was subcloned into EGFP-N2. TLR4-EGFP and the TLR4-EGFP mutant were transfected into H69 cells by means of Fugene HD reagent (Roche). Transfected cells were incubated in the presence or absence of 50 ng/mL recombinant Tat protein for 24 h. Fluorescent intensities were quantified using a FLX 800 micro-plate fluorescence reader (BioTek Instruments).
Three PCR-generated oligos with an incorporated T7 promoter were created using the TLR4-EGFP constructs as template. The primers used were as follows: for TLR4 downstream control, 5′-TAATACGACTCACTATAGGGAGTTTACAGAAGCTGGTG-3′ (forward) and 5′-TGGATAAGATTGTGAGCCAC-3′ (reverse); for TLR4 TAR-like sequence, 5′-TAATACGACTCACTATAGGGAGCCTAAGCCACCTCTCTAC-3′ (forward) and 5′-TTGTCTCCACAGCCACCAG-3′ (reverse). The TLR4-EGFP mutant was used as template to generate the mutant amplicon. In vitro transcription was performed using a MAXIscript kit (Ambion), in accordance with the manufacturer’s instructions. The radiolabeled RNA was incubated at 85°C for 5 min in binding buffer (Pierce) and then cooled to room temperature. Recombinant Tat protein was added (50 ng/mL) and incubated for 15 min at 30°C. For super-shift assays, HIV-1 Tat antibody (Immunodiagnostics) was added and incubated for an additional 20 min. Complimentary 78mer single-stranded oligonucleotides containing the HIV-1 TAR sequence and a T7 promoter were synthesized and annealed (sense, 5′-TAATACGACTCACTATAGGGAGGTCTCTCTGGTTAGACCAGATCTGAGCCTGGGAGCTCTCTGGCTAACTAGGGAACC-3′; antisense, 5′-GGTTCCCTAGTTAGCCAGAGAGCTCCCAGGCTCAGATCTGGTCTAACCAGAGAGACCTCCCTATAGTGAGTCGTATTA-3′). The unlabeled transcription product was used as a competimer. The binding assay reaction was gel electrophoresed in 1 × TBE (Tris, boric acid, and EDTA) buffer and exposed to film for detection of the labeled RNA.
H69 cells were incubated for 12 h in the presence or absence of Tat (50–100 ng/mL). C. parvum sporozoites were added to the cell culture. After a 6-h incubation, cells were fixed in paraformaldehyde and processed for TLR4 immunofluorescence by means of a monoclonal antibody against TLR4 (IMG 417A; Imgenex) and a polyclonal antibody against C. parvum (anti-CP2), as reported elsewhere . Slides were analyzed on a Zeiss LSM510 confocal microscope using identical parameters. Fluorescent intensity was determined using Zeiss LSM510 software. Briefly, individual infection sites (a minimum of 50 for each experimental condition) were traced, and the mean value of pixel intensity was determined. The mean intensity was then determined for the population of infection sites. For infectivity assays, infected cell cultures were incubated for 2, 24, and 48 h; fixed in paraformaldehyde; and processed for immunofluorescence using anti-CP2 antibody. The infection percentage was determined by counting the number of infection sites per the total number of cells in 20 fields at ×400 magnification.
All values are given as means ± SEs. Means of groups were compared using Student’s unpaired t test or analysis of variance, as appropriate.
TLR4 mRNA expression was increased in cells treated overnight with 100 ng/mL Tat (figure 1A), whereas expression of TLR2 and the adaptor protein MyD88 was not significantly altered (figure 1A). Similarly, treatment of the intestinal epithelialcell line HT-29 with 100ng/mL Tat protein resulted in a 3-fold increase in TLR4 mRNA expression, compared with that in untreated cells (data not shown). We further found that cholangiocyte TLR4mRNA expression was dose responsive to Tat protein (figure 1B). We also confirmed, using quantitative RT-PCR, our previous observation that TLR4mRNA expression is unaffected in C. parvum–infected cultured cells (figure 1C). However, when infections were performed in the presence of Tat, we again observed a corresponding increase in TLR4 mRNA expression, suggesting that Tat may regulate the level of TLR4 mRNA by either selectively increasing transcription or stabilizing the TLR4 transcript. To address whether Tat induced TLR4 mRNA stability, we approximated the mRNA half-life using actinomycin D to block transcription and a model of exponential decay. TLR4 mRNA in non–Tat-treated cholangiocytes turned over rapidly, with a half-life of ~25 min; however, in the presence of Tat, TLR4 mRNA exhibited a half-life >2 times that in untreated cells (figure 1D). Tat treatment did not affect the β-actin or TLR2 mRNA half-life in this same assay (data not shown).
In contrast to the observed increased expression of TLR4 mRNA, Tat treatment decreased TLR4 protein expression, as determined by quantitative Western blots of both H69 and HT-29 cell lysates (figure 2A and 2B). Tat treatment did not affect TLR2 protein expression in H69 cells (figure 2A), and, in support of previous observations , this protein was not detected in HT-29 cells (data not shown). We next addressed TLR4 protein expression in infected cells in the presence or absence of Tat. C. parvum infection, as expected, increased TLR4 expression >2-fold; however, in the presence of Tat, we observed suppression of the C. parvum–induced TLR4 up-regulation (figure 2C). Expression of TLR4 was also observed by immunofluorescent confocal microscopy (figure 2D). TLR4 localized to infection sites; however, in the presence of Tat, the level of TLR4 protein is diminished ~2-fold, as assessed by fluorescent intensity quantification (figure 2D). We next addressed the rate of TLR4 protein degradation in the presence or absence of Tat. The half-life of TLR4 protein in cholangiocytes was estimated to be ~15 h in both untreated and Tat-treated cell cultures (figure 3A). Consequently, we addressed the potential effect of Tat on TLR4 translation using in vitro transcription and translation. In the presence and absence of Tat, TLR4 transcription readily occurred in our in vitro system, as demonstrated by RT-PCR (figure 3B). However, when Tat was included in the reaction at 50 ng/mL, TLR4 protein production was diminished (figure 3C), suggesting that Tat does not interfere with the transcription of TLR4 mRNA but rather inhibits TLR4 protein expression.
We next asked whether Tat regulates TLR4 expression through Tat interactions with TLR4 mRNA. Tat binds to the TAR sequence of nascent RNA transcripts and can increase the expression of viral and host transcripts through this interaction. Using an in silico search of the TLR4 primary transcript, we identified a sequence with ~80% similarity over 25 bp comprising the TAR stem-loop sequence (figure 4A). Furthermore, a TLR4 sequence spanning 150 nt, which included the region with sequence similarities to the TAR sequence, was analyzed in silico using mfold software (version 3.2) [28, 29]. The mfold software predicted a secondary structure, with the lowest ΔG (−50.2 kcal/mol) contained in a stem-loop structure with similarities to the TAR stem loop, including a nearly identical loop sequence (figure 4B). Notably, sequence similarity to the TAR sequence was not identified in the TLR2 transcript.
The N-terminus of TLR4, which included the region of TLR4 with sequence similarity to TAR, was cloned into the EGFP vector. When Tat was included in the medium, a general decrease in EGFP expression was detected in the TLR4-EGFP–transfected cells, as determined by fluorimetry. Conversely, a 12-bp deletion from the TLR4 sequence that disrupts the predicted loop structure inhibited Tat-induced fluorescence reduction, suggesting a sequence specific–inhibition or secondary structure dependent–inhibition of TLR4-EGFP translation (figure 5).
To directly assess whether Tat could bind to this mRNA structure in vitro, a 113-bp fragment of TLR4 that included the sequence with the TAR-like loop structure was amplified by PCR. Using in vitro transcription and RNA gel-shift assays, we determined that Tat protein binds to this RNA fragment (figure 6A). Preincubation with a competimer (nonradiolabeled HIV TAR sequence) inhibited this shift. Additionally, a PCR-generated deletion of 12 bp that disrupts the predicted loop structure of the TLR4 transcript (figure 4B) inhibited the binding of Tat to TLR4 mRNA—hence, no shift was observed (figure 6A). Additionally, no shift was observed when the assay was performed with a region of TLR4 mRNA downstream of the predicted TAR-like loop structure (figure 6A). Binding of Tat to TLR4 RNA was further confirmed by a supershift assay using a Tat-specific monoclonal antibody (figure 6B).
We previously demonstrated a TLR-dependent increase in the transcription of the antimicrobial peptide hBD2 in response to C. parvum cholangiocyte infection . We therefore asked whether expression of hBD2 in C. parvum–infected cells differs in the presence and absence of Tat. C. parvum infection induced an increase in the transcription of hBD2, as expected (figure 7A). Conversely, when infections were performed in the presence of Tat, a significant decrease in hBD2 mRNA expression was observed 24 h after infection (figure 7A). Additionally, we previously demonstrated that inhibition of TLR signaling cascades diminishes the ability of cholangiocytes to clear C. parvum infection in vitro . We therefore asked whether Tat treatment of cholangiocytes diminishes the clearing of parasites in vitro. At 2 h after infection, the percentage of infected cells was similar among untreated and Tat-treated cells (figure 7B). However, by 48 h after infection, the number of parasites in cells treated with Tat at 100 ng/mL was 3-fold higher than that in untreated cells (figure 7B), suggesting that Tat treatment diminished the cholangiocyte defense against C. parvum infection.
The results of the present study provide the first evidence that HIV-1 Tat protein diminishes cholangiocyte recognition of and response to C. parvum through direct interaction with and translational suppression of a host cellular transcript, TLR4. Using a human immortalized but nonmalignant cholangiocyte cell line, we have shown that (1) Tat-treated cholangiocytes exhibit diminished TLR4 protein expression in uninfected and C. parvum–infected cells despite increased mRNA expression; (2) Tat inhibits the translation of and interacts directly with TLR4 mRNA; (3) Tat-treated cholangiocytes have diminished production of the antimicrobial peptide hBD2 in response to C. parvum infection; and (4) Tat-treatment diminishes the clearance of C. parvum by cholangiocytes. We propose that our findings provide an additional mechanistic explanation for the unusual susceptibility of HIV-infected individuals to biliary cryptosporidiosis.
HIV-1 Tat protein is essential for viral replication and acts as a transactivator of the HIV long terminal repeat by enhancing the processivity of RNA polymerase II elongation complexes . Through a similar mechanism, Tat transactivates many endogenous cellular genes, including those for cytokines and protooncogenes . Additionally, Tat can be released as a soluble factor, can be readily taken up by uninfected bystander cells, and modulates host cell protein activity through direct interactions . In the context of transactivation, Tat binds to a stem-loop structure on all nascent viral transcripts, the TAR region. Interactions between Tat and this domain require the uridine-rich bulge, whereas Tat protein also binds directly to the cyclin T1 domain of the positive transcription elongation factor complex pTEFb  and induces loop-specific binding of pTEFb to TAR RNA [34, 35]. Our RNA gel-shift analysis demonstrated that Tat not only binds to the TLR4 TAR-like region but that this interaction could be competitively inhibited with HIV-1 TAR RNA.
A number of cellular proteins specifically interact with double-stranded RNA (dsRNA) transcripts. Most cellular transcripts comprise a dsRNA structure at some point in their existence and as a consequence most likely interact with dsRNA-binding proteins, a process likely to confer critical regulatory events that control gene expression. It is noted that the predicted site of Tat-TLR4 interaction resides in the coding sequence, whereas translational regulatory factors interact predominantly, yet not exclusively [36, 37], with regulatory elements present in the 5′ or 3′ untranslated region. Additional studies are required to more precisely identify the critical molecular characteristics required for interaction between Tat and TLR4 mRNA and to identify cellular factors that may contribute to translation repression.
TLRs are key to epithelial innate immunity through detection of invading pathogens and subsequent activation of signaling pathways leading to the release of cytokines or chemokines and antimicrobial peptides [38–42]. Our previous data demonstrated that cholangiocytes express all 10 known TLRs and that 2 of these, TLR2 and TLR4, are involved in cholangiocyte immune responses to C. parvum infection via activation of NF-κB and subsequent secretion of antimicrobial peptides, such as hBD2 . Expression of TLRs by epithelia is tightly regulated to ensure an appropriate response on pathogen recognition. Therefore, the observed decrease in cholangiocyte TLR4 protein expression after Tat treatment may have profound effects on the capacity of host cells to mount an appropriate, efficient immune response to a secondary infection. Indeed, although Tat can promote cytokine secretion in a TLR-independent manner, we determined that the antimicrobial peptide hBD2 is down-regulated after treatment with Tat, which corresponds to a decrease in TLR4 expression. Furthermore, cholangiocytes in culture are readily infected with C. parvum in the presence or absence of Tat, yet parasite numbers are diminished over the course of 48 h in the absence of Tat, whereas the number of parasites increases over the same time frame in the presence of Tat. Therefore, this result suggests that, in addition to the requirement of CD4+T cells and corresponding interferon-γ production for the clearance of C. parvum , cholangiocyte-specific factors that are affected by Tat treatment are involved in the clearance of the parasite.
Highly active antiretroviral therapy (HAART) has decreased the incidence of biliary cryptosporidiosis  and AIDS cholangiopathy and has dramatically improved survival rates in individuals with AIDS cholangiopathy . However, C. parvum biliary infection and AIDS cholangiopathy remain significant problems in individuals who lack access to or fail to respond to HAART. We have previously demonstrated that C. parvum is cytopathic to cultured uninfected bystander cholangiocytes through a paracrine Fas/FasL-dependent apoptotic mechanism  and that HIV-1 Tat protein sensitizes cultured cholangiocytes to C. parvum–induced Fas/FasL-dependent apoptotic cell death . Thus, Tat not only sensitizes cholangiocytes, a cell type refractive to HIV-1 infection, to the cytopathic effect of biliary cryptosporidiosis, but our present results suggest that Tat also inhibits the response of cholangiocytes to this opportunistic pathogen. Therefore, this finding provides an additional explanation as to why HAART is associated with a decreased incidence of biliary cryptosporidiosis and AIDS cholangiopathy.
In summary, using an in vitro model of biliary cryptosporidiosis, we have demonstrated that Tat interacts with TLR4 mRNA and inhibits TLR4 translation. Thus, HIV-1 Tat protein diminishes cholangiocyte expression of an important pathogen-recognition receptor, a process that may enhance the susceptibility of individuals with AIDS to C. parvum biliary infection as well as be relevant to other opportunistic infections in patients with AIDS in general, particularly gram-negative bacterial infections of the biliary tract. It will be of interest to extend these investigations to the potential interactions between Tat and other cellular mRNAs with putative TAR-like domains and the specific cellular processes required for Tat-induced translational suppression.
Financial support: National Institutes of Health (grants DK57993 to N.F.L., DK76922 to S.P.O., A1062261 to A.D.B., and A2071321 to X.-M.C.); Mayo Foundation.
Potential conflicts of interest: none reported.
Presented in part: American Society for Cell Biology Conference, Washington, DC, 1–5 December 2007 (abstract 532).