1. Introduction
microRNAs (miRNAs) represent a class of small non-coding RNAs of ~22 nucleotides in length that repress gene expression on a post-transcriptional level by targeting the 3′ UTRs of cellular mRNA leading to its degradation or inhibition of translation [1]. miRNAs were implicated in a wide range of physiological as well as pathological processes, including inflammatory response, apoptosis, growth and cancer, neurodegenerative and cardiovascular diseases [2]. Increasing evidence suggests an important role of miRNAs in the immune response against infectious agents [3–5]. Previous work focused on and revealed direct anti-viral activity of miRNAs through repression of viral mRNA production [6]. Conversely, viral miRNAs were found to antagonize the host mRNA leading to a suppression of the anti-viral response [7].
Recently, a role of miRNAs in the response against bacterial pathogens has been proposed. miRNAs were shown to be effective against Pseudomonas syringae infection in plants [8]. Similar to viruses, P. syringae was found to secrete proteins that bind host miRNA and subsequently modulate immune response [8]. Furthermore, Rao and colleagues described the presence miRNAs expressed by pathogenic Pseudomonas aeruginosa strains which were isolated from adult patients with cystic fibrosis [9]. Xiao et al. uncovered a Helicobacter pylori-dependent induction of miR-146b and miR-155 in gastric epithelial cells with subsequent inhibition of IL-8, a central cytokine in the chemotaxis of leukocytes [10]. Further investigation revealed that miRNAs control major inflammatory pathways, such as the TLR-mediated activation of the NF-kB pathway [10]. While P. syringae and H. pylori remain extracellular during infection, a recent study showed altered immune response of mice deficient in miR-155 to the facultative intracellular pathogen Salmonella [5]. Schulte et al. uncovered the regulation of IL-6 and IL-10 by miRNAs of the let-7 family and miR-155 induction by secreted effector proteins of Salmonella rather than the invading pathogen [5].
In this study, we observed differential regulation of miRNAs and associated target transcripts in epithelial cells following infection with Listeria monocytogenes. L. monocytogenes is a Gram-positive, facultative intracellular bacterium that has been used widely for the elucidation of immune processes in a variety of hosts and tissues. L. monocytogenes facilitates its entry into non-phagocytic cells, such as epithelial Caco-2 cells, via surface bound and secreted effector proteins known as internalins. Internalized Listeriae are able to escape from the hostile phagocytic vacuole using the effector protein listeriolysin (LLO), a secreted toxin that is essential for the pathophysiology and intracellular survival of L. monocytogenes.
Using defined mutants that variously lack individual virulence factors, this study provides evidence that the ability and extent of Listeria induced regulation of host miRNAs strongly depends on cellular localization, on secreted and membrane-bound proteins of the pathogen.
2. Materials and Methods
2.1. Bacterial Strains and Growth Conditions
L. monocytogenes EGD-e [11] and its isogenic deletion mutants Δhly [12] and ΔinlAB [13] were used in this study. Bacteria were grown in BHI broth overnight at 37 °C with shaking at 180 rpm. Overnight cultures were diluted into 1:50, grown to mid-exponential phase (OD600nm = 1.0) and used for further experiments.
2.2. Eukaryotic Cell Culture
Human epithelial cells (Caco-2) were cultured in MEM with 10% fetal calf serum (FCS) and 5% non-essential amino acids, respectively. Cells were maintained at 37 °C in 5% CO2.
2.3. LLO Purification
LLO is expressed and purified from a recombinant L. innocua 6a strain harboring the hly gene [14]. Briefly supernatant fluids were concentrated using a Millipore filtration apparatus followed by batch absorption onto Q-sepharose (Pharmacia, Freiburg, Germany) and pre-equilibrated with loading buffer (50 mM NaH2PO4, pH 6.2). The non-absorbed fraction was centrifuged and desalted by transferring through a super loop to a HiPrep 26/10 desalting column (Pharmacia, Freiburg, Germany) where loading buffer (50 mM NaH2PO4, pH 6.2) was used to elute the desalted fraction. This fraction was subsequently filtered through a Millipore filter (0.22 μm) and loaded onto a Resource-S column previously equilibrated with 50 mM NaH2PO4, pH 6.2. The pure toxin eluted reproducibly from the column at 0.21 to 0.28 M NaCl using elution buffer (50 mM NaH2PO4 1M NaCl, pH 5.6). Fractions were collected and individually tested for hemolytic activity. Yields of the toxins range from 1 to 5 mg/L supernatant with a hemolytic activity (HU) of 20,000 HU/mg purified protein. One hemolytic unit (HU) is expressed as the amount of toxin required to lyse 50% of a 1% suspension of sheep erythrocytes. The toxin showed a high purity as seen using SDS-PAGE analysis, was efficiently recognized with LLO-specific antibodies, and exhibited hemolytic activity on sheep erythrocytes at both pH 6.0 and pH 7.4 respectively.
2.4. Infection Assays and LLO Treatment
Caco-2 cells were maintained in 6-well plates following at conditions described above. Bacteria at MOI 10 were added to the monolayer of cells. One hour of post infection, followed by washing with 1 × PBS, cells were supplemented with fresh media containing 20 μg/mL gentamycin to remove extracellular bacteria. After one hour of gentamycin treatment cells were lysed using a mixture of RLT lysis buffer and 1% β-mercaptoethanol and used for RNA isolation.
LLO at different concentrations (25 ng/mL and 50 ng/mL), was preactivated with dithiothreitol before administration to Caco-2 cells. Following incubation with LLO for one hour, cells were lysed with RLT lysis buffer and 1% β-mercaptoethanol.
2.5. RNA Isolation
RNA was isolated from cell lysate samples using the Qiagen miRNeasy Kit. Briefly, cell lysate samples were transferred to the QIA Shredder column and centrifuged at 13,200 rpm. An equal amount of 70% of ethanol was added to the eluted sample and mixed thoroughly. These samples were passed through a nucleic acid binding column which is supplied by the miRNeasy Kit (Qiagen). The DNA present on the column was digested using RNase-free DNase (Qiagen) for 30 min at RT and RNA was eluted by RNase free water. The quantity of isolated RNA was measured with NanoDrop analyzer (NanoDrop Technology, Rockland, MA, USA) and quality was assessed by running the samples on Nano-chips for 2100 Bioanalyzer (Agilent, Böblingen, Germany).
2.6. miRNA Microarray
For this analysis we used the biochip “Geniom Biochip MPEA homo sapiens & mus musculus” (febit, Heidelberg, Germany). The probes are designed as the reverse complements of all major mature miRNAs and the mature sequences as published in the current Sanger miRBase release (version 14.0 September 2009, see http://microrna.sanger.ac.uk/sequences/index.shtml) for homo sapiens & mus musculus. Techniqual and procedural details are described in detail in supplementary material.
2.7. Reverse Transcription Reaction and Quantitative Real-Time PCR Analysis
First strand cDNA was generated for mRNA by using SuperScript II reverse transcriptase (Invitrogen) and miScript reverse transcription kit (Qiagen) for miRNAs using 1 μg of RNA for each reaction.
Quantitative real-time PCR analysis was performed by using AB Prism 7900 HT system. All forward and reverse primers used for PCR were purchased from Qiagen. We used RNUA1 as internal controls for miRNA expression normalization and HPRT for target mRNA expression normalization.
The reaction mixture volume of 25 μL for mRNA quantitative real-time PCR was applied using 100 ng cDNA for each reaction. For miRNA quantitative real-time PCR analysis 3 ng of cDNA per 50 μL reaction set-up was used. For each primer the efficiency was calculated by standard curve which was generated by using different concentrations of genomic DNA in real time PCR. The expression level of mRNA and miRNA was calculated by normalizing its quantity to the respective expression of the internal control in Caco-2 epithelial cells. Threshold cycle values (CT) of the tested transcripts were determined and normalized expression of each target gene was given as the ΔCT between the log2 transformed CT of the target gene and the log2 transformed CT of the internal control. Log2 transformed gene expression levels (ΔCT) of each target transcript were expressed as log2 differences from control (=log2 ΔΔCT method). Data was acquired and analyzed with the SDS 2.3 and RQ-Manager 1.2, respectively.
2.8. Statistical Data Analysis of Infection Experiments
All infection and toxin experiments were performed for a minimum of three times. Significant differences between two values were compared with a paired Student’s t-test. Values were considered significantly different when the p value < 0.05.
3. Results
3.1. <italic>L. monocytogenes</italic> Differentially Induces miRNAs Dependent on Cellular and Subcellular Localization
Based on miRNA expression analysis using microarrays, we selected a subset of miRNA candidates that were differentially deregulated following wild type infection of epithelial Caco-2 cells. We focused on miRNAs that have a biologically validated role in vitro or in vivo. These miRNAs were validated using qRT-PCR. Relative expression levels obtained by both techniques showed a robust correlation (Figure S1).
In addition to the wild-type infection, Caco-2 cells were infected with two isogenic mutant strains or incubated with purified listeriolysin (LLO). The hly mutant strain is unable to produce LLO and remains in the phagocytic vacuole after host cell infection. ΔinlAB remains in the extracellular space because of the inability to induce bacterial uptake into epithelial cells.
Infection with wild-type bacteria leads to significantly increased expression of miR-146b, miR-16 and miR-155 expression in Caco-2 cells compared to non-infected cells (Figure 1).
As previously described for Salmonella [5], we also observed a significant downregulation of let-7a1 (Figure 1), a member of the let-7 family that is implicated in immune response and cancer development. We further observed a strong downregulation of miR-145 (Figure 1). A recent study demonstrated that blocking miR-145 led to a strong anti-inflammatory and reduced airway hyper responsiveness comparable to the effects obtained following glucocorticoid treatment [15].
Compared to wild-type infection both mutant strains induced significant deregulation of miR-146b and miR-16 (Figure 1). The expression differed with respect to the directionality of regulation for these miRNAs; while upregulated following wild-type infection, the expression of both miRNAs was decreased following infection with both mutant strains. Furthermore, we observed significant downregulation of let-7a1 by both mutant strains without significant differences compared to expression following wild-type infection (Figure 1).
There was no significant difference in expression of miR-146b, miR-16, let-7a1 and miR-145 between the Δhly and ΔinlAB strains (Figure 1).
3.2. Wild-Type and LLO-Deficient Bacteria Induce miR-155, While the ΔinlAB Mutant Strain Suppresses miR-155 Expression
miR-155 is one of the best characterized miRNAs and is involved in innate immune response to a variety of pathogens, including but not limited to H. pylori, P. syringae and Salmonella. We show that L. monocytogenes induces strong miR-155 expression in Caco-2 cells (Figure 1). Strikingly, infection with Δhly also provoked a comparable induction of miR-155. In contrast, ΔinlAB not only lacked the ability to induce of miR-155, but significantly downregulated miR-155 compared to wild-type infection and control (Figure 1).
3.3. Purified LLO Induces the Expression of miR-146b, miR-16 and miR-155 in Caco-2 Cells
In a further step, we sought to investigate the regulation of the above studied miRNAs after incubation with purified LLO. Expression of three miRNAs, miR-146b, miR-16 and miR-155 was significantly increased in infected cells compared to non-infected controls (Figure 2A).
Strikingly, miR-146b displayed an inverted expression pattern compared to Δhly infection indicating that expression of miR-146b is directly connected to the presence of LLO. While unchanged after Δhly infection compared to control, miR-16 is seen upregulated after LLO incubation emphasizing the importance of this effector protein in miRNA regulation induced by L. monocytogenes. In contrast, induction of miR-155 expression was comparable in both settings, following infection with Δhly strains as well as LLO incubation. To quantify the effect of higher doses of LLO on the magnitude of miR-155 induction we used a higher toxin dose. We observed no significant changes in miR-155 expression between both LLO toxin concentrations (Figure 2B).
In contrast to the changes seen in miR-155 expression, miR-145 and let-7a1 expression showed no significant deregulation in Caco-2 cells incubated with LLO (Figure 2A).
3.4. Deregulation of mRNAs That Are Targeted by miRNAs
To estimate the correlation between miRNA deregulation and downstream effects on target mRNA of particular miRNAs we measured mRNA expression levels of important targets of these miRNAs. These include major inflammatory cytokines and interleukins such as IL-6, IL-8, TNF-α, IFN-β (Figure 3 and Table 1).
In concordance with miRNA deregulation, there are significant changes of target mRNA levels in Caco-2 cells infected with wild-type and Δhly or ΔinlAB mutant strains compared to control cells.
4. Discussion
In this study we demonstrate for the first time that L. monocytogenes mediates differential deregulation of miRNAs in the human epithelial cell line Caco-2. Using wild-type bacteria, two isogenic mutants Δhly and ΔinlAB, and purified toxin we show that listeriolysin and internalins are involved in miRNA expression and regulation of the putative target transcripts. miRNA microarrays were used to screen and select a subset of miRNA candidates that were significantly deregulated and have biologically validated roles in host response to external stimuli. These miRNAs, including miR- 16, miR-145, mir146, miR-155 and let-7a1 were further investigated.
miR-16 is required for the rapid degradation of inflammatory mediators that contain AU-rich sequences, such as TNF-α, IL-6 and IL-8. Interestingly, miR-16 was previously reported to be upregulated in NIH 3T3 cells infected with murine gammaherpesvirus 68, a virus closely related to Epstein-Barr virus (EBV) and Kaposi’s sarcoma associated herpesvirus (KSHV) [21]. Activation of miR-16 gene was also observed in cholangiocytes in a p65-independent manner by Cryptosporidium parvum, a protozoan parasite that infects the gastrointestinal epithelium [22]. We observed a significant upregulation of miR-16 by wild-type bacteria and purified LLO, while absence of hly and inlAB resulted in significantly decreased expression of miR-16. Other studies have shown that miR-16 expression is stable among a variety of cell lines and expression is not altered by a variety of immune modulators. The observed toxin mediated induction of miR-16 and subsequent targeting of inflammatory mediators may therefore represent a targeted miRNA mediated mechanism of immunmodulation triggered by L. monocytogenes rather than an unspecific host cell response to infection [17,23,24].
miRNA expression profiling in human macrophages has shown that miR-146 and miR-155 are endotoxin-responsive genes that are involved in several immune and inflammatory pathways [25,26]. A recent study revealed that miR-146b upregulation leads to inhibition of H. pylori induced inflammatory response in human gastric epithelial cells. miR-146b was shown to inhibit IL-8 expression, possibly through interleukin-1 receptor-associated kinase 1 (IRAK1) and TNF receptor-associated factor 6 (TRAF6), two major adaptor molecules in TLR receptor signaling and NF-kB activation [27]. Thus miR-146b is a potent target to aim in order to manipulate host response. We show that miR-146b is mainly induced in a LLO-dependent manner during infection with L. monocytogenes and emphasize the central role of LLO the regulation of host miRNA. Caco-2 cells express TLR2 and TLR4 [28], two cell surface receptors that are targeted by listerial virulence factors including LLO. Thus, we suggest that Listeria induced miR-146b induction and subsequent target gene interaction may be triggered by LLO via a TLR-mediated pathway.
miR-155 has an established regulatory role in several pathways of innate and adaptive immune response [26]. Our results show that wild-type bacteria and purified LLO at two different doses induce miR-155 expression to a similar extent. However, upregulation of miR-155 also occurs following incubation with the LLO deficient mutant strain indicating that this induction is also triggered through a vacuole-dependent pathway. This process is possibly mediated by MyD88, since vacuolar signaling and subsequent expression regulation in listerial infection is entirely dependent on this adaptor molecule [29]. MyD88 also integrates TLR-signaling triggered by extracellular stimuli, such as LLO incubation. We conclude that miR-155 induction may be triggered through both LLO-dependent and an LLO-independent vacuolar mediated pathway. Both routes may merge in a common pathway that results in a comparable miR-155 induction as observed in this study.
Interestingly, the expression of miR-155 was strongly reduced following infection with ΔinlAB compared to wild-type bacteria or Δhly. Thus, we suggest a new functional role for internalins in the regulation of miR-155 that subsequently results in increased degradation of the pro-inflammatory response mediated by TNF-α.
A recent study investigated the role of miR-145 in the inflammatory response in human colonic tissue of patients with ulcerative colitis [30]. miR-145 was strongly upregulated in inflamed colon segments of affected subjects who are at increased risk to develop colon cancer. A further study demonstrated that blocking miR-145 led to a strong anti-inflammatory response and reduced airway hyper responsiveness [15]. Thus, downregulation of miR-145 by L. monocytogenes as observed in this study may serve as a further mechanism of diminishing host immune response and facilitate survival of the pathogen. Furthermore, miR-145 was predicted to target IFN-β [18], a type I interferon that exhibits inflammatory and anti-inflammatory effects upon infection with L. monocytogenes. In line with miR-145 downregulation, IFN-β was strongly upregulated upon infection of Caco-2 cells indicating a possible contribution of miR-145 in its regulation, although it did not reach statistical significance.
Previous reports implicated miR-145 in the release of intestinal mucus components such as mucin (e.g., MUC1 or MUC2) that mediate an exocytosis mechanism leading to decreased uptake of L. monocytogenes into epithelial cells. L. monocytogenes was shown to counteract this mechanism via binding MUC2 by InlB, InlC and InlJ [31]. It is known that miR-145 controls the suppression of MUC1 causing a reduction of β-catenin, as well as the oncogenic cadherin 11 [32]. Thus, downregulation of miR-145 by the host cell results in decreased bacterial uptake. Overall miR-145 has a complex role in response to infection with L. monocytogenes and warrants further study.
Recently, downregulation of let-7 family members was identified as control major regulators of inflammation, including IL-6 and IL-10 in macrophages and HeLa cells upon infection with Salmonella [5]. We observed a similar regulation in Caco-2 cells following Listeria infection suggesting an analogous role of this host miRNA in Gram-positive and Gram-negative pathogens.