Abstract
Long noncoding RNAs (lncRNAs) have emerged as critical regulators of inflammation. To further understand the interaction between inflammatory signaling pathways and lncRNAs, we characterized the function of cardiac and apoptosis-related lncRNA (Carlr), an lncRNA expressed in both mouse and human cells of diverse tissues. Carlr expression is increased following NF-κB signaling in macrophages, with concomitant translocation to, and enrichment of, the transcript in the cytoplasm. Knockdown of Carlr results in impaired expression of NF-κB pathway genes and influences the interaction between macrophages and intestinal cells in an inflammatory environment. In human celiac disease patient samples, increased levels of the Carlr transcript were detected in the cytoplasm, alongside elevated expression of NF-κB pathway genes. These findings suggest that increased Carlr expression and/or cytoplasmic localization is required for efficient NF-κB signaling and is associated with the inflamed tissue state observed in human celiac disease.
Introduction
Long noncoding RNAs (lncRNAs) represent a large portion of the noncoding genome and are versatile molecules that regulate diverse cellular processes in many biological pathways, including immunity and inflammation (1, 2). Recent work has demonstrated that lncRNAs are important regulators of lymphocyte function (3), immune cell development (4, 5), and inflammatory signaling (6, 7). lncRNAs do not code for proteins but rather function in numerous other capacities, including as scaffolds for other cellular proteins/RNAs, as direct cofactors of proteins, or as sponges for other RNAs. Although the functions of some lncRNAs are well defined, a clear mechanistic understanding of most lncRNAs remains to be established. Many lncRNAs interact with key signaling pathways to regulate cellular processes. They have also been linked to different diseases such as cancer and intestinal inflammatory diseases such as celiac disease (CeD) (8, 9).
The intestine contains the largest pool of macrophages in the body, which perform many critical functions, including maintenance of mucosal homeostasis and protective immunity, and are implicated in the development of inflammatory intestinal diseases (10). The NF-κB transcription factor functions as a central hub for integrating inflammatory signals, and aberrant activation of the NF-κB pathway in macrophages can lead to the overproduction of proinflammatory cytokines such as TNF-α, IL-1β,and IL-6, leading to mucosal inflammation that contributes to intestinal inflammatory diseases (11). At the same time, induction of NF-κB signaling induces changes in the expression of lncRNAs that have been shown to be important for both cytokine production and subsequent resolution of inflammation (6, 9, 12). To explore the possible involvement of lncRNAs in intestinal inflammatory diseases we screened for lncRNAs whose expression was significantly increased following LPS stimulation in macrophages. We identified one such lncRNA, cardiac and apoptosis-related lncRNA (Carlr), which has recently been characterized in cardiomyocytes (13). Mouse Carlr is a single-exon intergenic lncRNA located between genes Spag6 (BC061194) and Pip4k2a on mouse chromosome 2. Carlr suppresses mitochondrial fission and apoptosis by targeting microRNA-539 and PHB2 in cardiomyocytes. Interestingly, Carlr is known to be expressed in various tissues, including myeloid lineage cells and the intestine (13), and its expression is induced in macrophages and dendritic cells in response to LPS (11, 14, 15). However, microRNA-539 levels do not change in response to LPS stimulation (J.J. Seeley and S. Ghosh, unpublished observations), and LPS-induced genes are not known to be targeted by this microRNA. These findings suggest that Carlr might have other cell type/stimuli–specific functions that remain to be characterized.
In this study, we describe the regulation and function of Carlr upon NF-κB activation in macrophages and intestinal cells of both mouse and human origin. Our results support a role for Carlr in promoting induction of NF-κB–stimulated genes, which may promote the pathogenesis of human inflammatory bowel diseases. These findings highlight the potential for lncRNAs to have divergent, cell type–specific functions and further emphasize their potential importance in human disease.
Materials and Methods
Cell culture, coculture, and in vitro stimulations
Primary bone marrow–derived macrophages were generated by culturing bone marrow from wild-type C57BL/6 mice in 20% L929 cell–conditioned DMEM medium (Life Technologies BRL). For LPS stimulation, cells were seeded in 12-well plates at 105 cells/ml and stimulated with 10 ng/ml Escherichia coli–derived LPS (Sigma-Aldrich). Five and 10 μM BAY-11-7082 were used for the inhibition of NF-κB. The human THP-1 macrophage-like cell line was cultured in RPMI 1640 (Life Technologies BRL) supplemented with 10% FBS and antibiotics. For stimulation, cells were seeded at 105 cells/ml and stimulated with 100 ng/ml LPS. The mouse and human intestinal cell lines C26 and Caco2 (derived from colorectal carcinoma) were cultured in RPMI 1640 and DMEM (Life Technologies BRL), respectively, supplemented with 10% heat-inactivated FCS (Life Technologies BRL) and antibiotics (500 IU/ml penicillin and 100 μg/ml streptomycin; Life Technologies BRL). For coculture experiments, Caco2 cells were plated in six-well plates and allowed to attach overnight. THP-1 cells were added the following day and coculture was maintained in supplemented RPMI 1640 media for the duration of the experiment. After culturing, THP-1 nonadherent cells were removed by three rounds of washing, and the Caco2 adherent cells were detached and used for the RNA or protein extraction.
RNA knockdown
For silencing experiments, macrophages immortalized by inoculation with J2 retrovirus (16) were used for the establishment of stable Carlr knockdown cell lines using the lentiviral pLKO.1-puro vector. Short hairpin RNAs (shRNAs) were cloned in the pLKO.1-puro vector according to the manufacturer’s instructions (Addgene). The same shRNAs were used for transient silencing of mouse intestinal C26 cells. For silencing of human Carlr, two short interfering RNAs were purchased from Integrated DNA Technologies (si3:5′-CAUUUACUUUCCCAUGAUAAAACAG-3′ and si5:5′-AUCACAUUUACUUUCCCAUGAUAAA-3′) and transfected into the THP-1 cells using lipofectamine RNAiMAX (Invitrogen).
Patient samples
CeD was diagnosed according to the European Society of Pediatric Gastroenterology Hematology and Nutrition criteria in effect at the time of patient recruitment, including anti-gliadin, anti-endomysium, and anti-transglutaminase Ab determinations, as well as a confirmatory small bowel biopsy. The study was approved by the Institutional Board (Cruces University Hospital code CEIC-E09/10 and Basque Clinical Trials and Ethics Committee code PI2013072) and analyses were performed after informed consent was obtained from all subjects or their parents. Biopsy specimens from the distal duodenum of each patient were obtained during routine diagnostic endoscopy. None of the patients suffered from any other concomitant immunological disease. None of the controls showed small intestinal inflammation at the time of the biopsy.
RNA extraction, reverse transcription, and qPCR
Total RNA was extracted from cells using Qiagen RNA mini/micro kits. All samples were subjected to DNAse I treatment. A total of 1 μg of RNA was used for reverse transcription (RT) using SSIII enzyme (Invitrogen), and real-time quantitative PCR (qPCR) was carried out using 2× SYBR Green fluorescent dye (Quanta Biosciences). The amplified transcripts were quantified using the comparative cycle threshold method. All experiments were performed in triplicate. TaqMan assays and Universal Master Mix (Applied Biosystems) were used for the quantification of Carlr targets in human samples. All primer sequences are available upon request.
Cellular fractionation
For quantification of Carlr levels in nuclear and cytoplasmic compartments, nuclei were isolated using C1 lysis buffer as described previously (6), and the amount of specific nuclear RNA measured by RT-qPCR was compared with the total amount of the RNA in the whole cell. Nuclear/cytoplasmic protein fractionation was done using hypotonic and hypertonic buffers with Nonidet P-40 detergent.
RNA–protein interaction assay
Carlr was in vitro transcribed using a T7 polymerase after amplification of the region with primers harboring a T7 promoter. RNA was allowed to fold, and 80 ng was mixed with protein lysates and incubated at room temperature for 1 h. After incubation, glycerol was added and the samples were run in a 1% agarose gel with Tris-borate buffer. RNA bands were stained using SYBR Safe.
RNA immunoprecipitation
For RNA immunoprecipitation (RIP) experiments, cells were lysed in RIP buffer (150 mM KCl, 25 mM Tris [pH 7.4], 0.5 mM DTT, 0.5% Nonidet P-40) and homogenized using a syringe. For tissue sample RIP preparation, Santa Cruz Biotechnology protocols were followed. Briefly, tissue was immediately flash frozen and stored at −80°C. Before RIP, tissue was thawed in RIPA buffer (10 mM Tris-Cl [pH 8], 1 mM EDTA, 0.5 mM EGTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS, 140 mM NaCl), disrupted, and homogenized while maintaining temperature at 4°C throughout all procedures. Afterward, samples were incubated on ice for 30 min and centrifuged to obtain cell lysates. For all samples, agarose beads were coupled with anti-p65 and anti-IκBα Abs and incubated with lysates overnight at 4°C. Samples were then washed eight times and bead-bound material was divided for Western blot and qPCR analyses.
Western blot
SDS sample buffer was added to cell lysates. Proteins were separated in 10% SDS-PAGE and transferred to polyvinylidene difluoride membranes. Immunoblotting was performed with the following primary Abs: anti–β-tubulin (Sigma-Aldrich), anti-Hdac1, anti-p65, anti-p50 (Santa Cruz Biotechnology), and anti-IκBα (Thermo Scientific). Signals were detected using Pierce ECL Western blotting substrate (Thermo Scientific).
Proliferation assay
Caco2 cells (1.7 × 104) were seeded in a 24-well plate, and THP-1–conditioned medium was added the following morning. Proliferation was measured at days 0, 1, 3, and 5. For crystal violet staining, cells were fixed in 4% paraformaldehyde and stained with 0.1% crystal violet. After staining, cells were washed and 10% acetic acid was added. Absorbance was measured at 590 nm.
Results
Carlr lncRNA expression increases and translocates to the cytoplasm after NF-κB activation in mouse macrophages
To assess the role of lncRNAs in inflammation, we characterized the expression of several lncRNAs in response to LPS stimulation in mouse macrophages. The intergenic lncRNA Carlr located on mouse chromosome 2 (Fig. 1A) was upregulated upon stimulation, resembling what has been described for other LPS-regulated lncRNAs such as linc-Cox2 (6) (Fig. 1B). These results confirmed that LPS induces dynamic changes in multiple lincRNA expression profiles with cell type specificity. Promoter analysis using LASAGNA-Search 2.0 (http://biogrid-lasagna.engr.uconn.edu/lasagna_search/) showed an NF-κB binding site at position −327/328 (from transcription start site) (Fig. 1C). To confirm whether the increase of Carlr in macrophages is due to NF-κB activation, we stimulated cells with LPS in the presence of two different concentrations of the NF-κB inhibitor BAY-11-7082. The increase of Carlr in response to LPS was diminished after NF-κB inhibition in a dose-dependent manner, suggesting that NF-κB is involved in the upregulation of this lncRNA (Fig. 1D). To determine the subcellular localization, and likely site of function of Carlr, cellular fractionation and qPCR analysis were employed. Carlr was found to exhibit a primarily nuclear localization at basal conditions in mouse macrophages (Fig. 1E). However, after LPS stimulation, most Carlr transcript was located in the cytoplasm (Fig. 1E). Interestingly, stimulation of mouse intestinal cells with LPS did not show a change of expression (Supplemental Fig. 1A) or localization of this lncRNA (Fig. 1F). Given the observed expression pattern, we hypothesized that Carlr could be a circular RNA that is transported by a nuclear export system (17); however, attempts to amplify a circular form of Carlr were not successful (data not shown). These results implicate Carlr as an NF-κB–regulated lncRNA that is dynamically expressed following LPS stimulation in mouse macrophages.
(A) Schematic diagram of Carlr intergenic location in mice (mm10). (B) Mouse macrophages stimulated with LPS for indicated time points. Data were normalized to Hprt. Data represent the mean ± SE of three independent experiments. The p values were calculated relative to basal level (based on an unpaired Student t test). (C) Closer view of the Carlr region. Reverse and forward black arrows represent the location of the RT-qPCR primers. The NF-κB binding site was predicted using LASAGNA-Search 2.0. (D) LPS-induced Carlr upregulation is diminished after NF-κB inhibition by BAY-11-7082 (right). RT-qPCR data are represented as the mean ± SE of three independent experiments. TNF levels (left) are shown as a control for the NF-κB inhibition. RT-qPCR analysis on whole-cell and nuclear RNA to determine localization of Carlr in mouse macrophages (E) and intestinal cells (F) is shown. U6 was used as nuclear control and Hprt as cytoplasmic control. Data represent the mean ± SEM of three independent experiments. +p ≤ 0.1, *p < 0.05, **p < 0.01.
(A) Schematic diagram of Carlr intergenic location in mice (mm10). (B) Mouse macrophages stimulated with LPS for indicated time points. Data were normalized to Hprt. Data represent the mean ± SE of three independent experiments. The p values were calculated relative to basal level (based on an unpaired Student t test). (C) Closer view of the Carlr region. Reverse and forward black arrows represent the location of the RT-qPCR primers. The NF-κB binding site was predicted using LASAGNA-Search 2.0. (D) LPS-induced Carlr upregulation is diminished after NF-κB inhibition by BAY-11-7082 (right). RT-qPCR data are represented as the mean ± SE of three independent experiments. TNF levels (left) are shown as a control for the NF-κB inhibition. RT-qPCR analysis on whole-cell and nuclear RNA to determine localization of Carlr in mouse macrophages (E) and intestinal cells (F) is shown. U6 was used as nuclear control and Hprt as cytoplasmic control. Data represent the mean ± SEM of three independent experiments. +p ≤ 0.1, *p < 0.05, **p < 0.01.
Knockdown of Carlr alters the NF-κB pathway in mouse cells
To gather more information about the function of Carlr, we analyzed a panel of 92 NF-κB target genes to correlate the increased expression and cytoplasmic export of this lncRNA with the differential expression of genes in this pathway. After LPS stimulation, genes were grouped by their temporal expression pattern using STEM, a software application designed specifically to identify temporal expression profiles and the genes associated with these profiles (18). We observed that Carlr is coexpressed with a group of NF-κB target genes (including Tnfaip3, Ptgs2, Nfkb1, and Nfkb2 or Il1a and Il1b) (Supplemental Fig. 1B), suggesting that it might function as a modulator similar to the NF-κB interacting lncRNA (19). To address the role of Carlr in NF-κB signaling and downstream pathways, we silenced this lncRNA using two different target sequences in mouse macrophages. We first confirmed that Carlr expression was significantly reduced in the cells transfected with these shRNAs (Fig. 2A). When silencing Carlr we observed significantly decreased expression of 4 (Nfkb2, Ptgs2, Il1a, and Il1b) out of 10 of the analyzed coexpressed NF-κB targets in both macrophage knockdown cell lines (Fig. 2B). To address whether Carlr silencing affects only NF-κB–dependent gene expression or is also involved in nuclear NF-κB activity, we measured the amount of NF-κB p65 subunit (a component of the most abundant NF-κB complex) in the sh-Carlr macrophages. Both silenced cell lines had less nuclear p65 than did the cells transduced with the empty vector (Fig. 2C). Silencing of Carlr also led to a decrease in phosphorylation of IκBα, which results in less degradation of IκBα and consequently less nuclear translocation of NF-κB (Supplemental Fig. 1C). Taken together, these results suggest that Carlr is involved in the NF-κB signaling cascade in mouse macrophages.
(A) Carlr levels are significantly reduced in both sh1 and sh2 cell lines. Empty vector was used as a control. Data represent the mean ± SEM of four independent experiments. *p < 0.05, **p < 0.01. (B) RT-qPCR analysis was performed on the sh1 and sh2 cell lines to determine the effect of Carlr knockdown on the correlated genes. Genes with statistically significant decreased expression in both cell lines are shown. Data are represented as the mean ± SEM of four experiments. *p ≤ 0.05, **p < 0.01, based on an unpaired Student t test. (C) p65 levels were decreased in the nucleus of sh1 and sh2 cell lines. β-Tubulin and Hdac1 were used as loading controls. Western blot signal was quantified using ImageJ software. *p ≤ 0.05, **p < 0.01.
(A) Carlr levels are significantly reduced in both sh1 and sh2 cell lines. Empty vector was used as a control. Data represent the mean ± SEM of four independent experiments. *p < 0.05, **p < 0.01. (B) RT-qPCR analysis was performed on the sh1 and sh2 cell lines to determine the effect of Carlr knockdown on the correlated genes. Genes with statistically significant decreased expression in both cell lines are shown. Data are represented as the mean ± SEM of four experiments. *p ≤ 0.05, **p < 0.01, based on an unpaired Student t test. (C) p65 levels were decreased in the nucleus of sh1 and sh2 cell lines. β-Tubulin and Hdac1 were used as loading controls. Western blot signal was quantified using ImageJ software. *p ≤ 0.05, **p < 0.01.
Carlr is expressed in human tissues
These observations led us to examine whether a homologous lncRNA for mouse Carlr is transcribed from the human genome. Alignment and conservation analysis using different databases and analysis tools showed that the mouse and human genomic regions exhibit significant sequence homology (Fig. 3A). Moreover, data from the Epigenome Roadmap project (20) shows RNA sequencing signals in this region in several cell/tissue types (Supplemental Fig. 2A). We designed qPCR primers and evaluated the expression of human Carlr using a pool of RNAs derived from 12 human tissues. We observed that this region is transcribed from the human genome in most of the tissues analyzed, showing higher expression in stomach, intestine, and lung (Fig. 3B). When RT-qPCR primers located in a conserved region upstream from the Carlr region were used, no transcript amplification was observed (Supplemental Fig. 2B). We then analyzed the expression of Carlr in the human monocytic cell lines U937 and THP-1. Carlr was found to be expressed in these human cell lines, and its expression was also induced in response to LPS stimulation as seen in mouse macrophages (Fig. 3C, Supplemental Fig. 2C). To evaluate whether the regulation of Carlr is conserved across species, we assessed the localization of Carlr in the human monocytic cell lines. Localization of Carlr in THP-1 and U937 cells showed a predominantly cytoplasmic location (Fig. 3C, Supplemental Fig. 2D). Interestingly, these cell lines have been described as having constitutively active NF-κB (21), which would therefore explain the cytoplasmic location of Carlr prior to LPS stimulation.
(A) Conservation analysis of Carlr region performed using VISTA (36). Mouse sequence is shown on the x-axis and percentage similarity to the corresponding human sequence is shown on the y-axis. The graphical plot is based on sliding-window analysis of the underlying genomic alignment. A 100-bp sliding window at 40-bp nucleotide increments was used. Pink shading indicates conserved noncoding genomic region (top). The image at the bottom represents pairwise alignment of the mouse and human genomes corresponding to Carlr region. Genome alignment was produced by blastz using WashU EpiGenome Browser. (B) Human Carlr expression (relative to stomach) measured by RT-qPCR in RNA pool of different tissues purchased from Clontech (human total RNA master panel II) (mean ± SD; two independent RT-qPCRs). (C) Expression of Carlr in basal conditions and after LPS stimulation (top) and localization of Carlr (bottom) in the human monocytic cell line THP-1. Data represent the mean ± SEM of three independent experiments. (D) Expression (left) and cytoplasmic amount (right) of Carlr in the human intestinal cell line Caco2 at basal conditions and after incubation with LPS and with LPS-stimulated THP-1 cells. +p ≤ 0.1.
(A) Conservation analysis of Carlr region performed using VISTA (36). Mouse sequence is shown on the x-axis and percentage similarity to the corresponding human sequence is shown on the y-axis. The graphical plot is based on sliding-window analysis of the underlying genomic alignment. A 100-bp sliding window at 40-bp nucleotide increments was used. Pink shading indicates conserved noncoding genomic region (top). The image at the bottom represents pairwise alignment of the mouse and human genomes corresponding to Carlr region. Genome alignment was produced by blastz using WashU EpiGenome Browser. (B) Human Carlr expression (relative to stomach) measured by RT-qPCR in RNA pool of different tissues purchased from Clontech (human total RNA master panel II) (mean ± SD; two independent RT-qPCRs). (C) Expression of Carlr in basal conditions and after LPS stimulation (top) and localization of Carlr (bottom) in the human monocytic cell line THP-1. Data represent the mean ± SEM of three independent experiments. (D) Expression (left) and cytoplasmic amount (right) of Carlr in the human intestinal cell line Caco2 at basal conditions and after incubation with LPS and with LPS-stimulated THP-1 cells. +p ≤ 0.1.
Enterocytes exist in close association with tissue macrophages that are located within the lamina propria close to the epithelial monolayer. Macrophages located in the intestinal epithelium interact with epithelial cells via soluble factors as cytokines. An increase in inflammatory cytokines can cause alterations in the epithelial cells, leading to mucosal inflammation. In response to LPS, macrophages become activated, releasing proinflammatory cytokines that can disrupt the adjacent enterocyte monolayer and lead to inflammatory disease (22, 23). To evaluate the role of Carlr in the macrophage–enterocyte interaction, we assessed the localization of Carlr in the intestinal cell line Caco2 at steady-state and in the presence of LPS and activated THP-1 cells. This cell line is widely used as a model for human intestinal inflammatory diseases and is able to produce NF-κB–regulated cytokines such as IL-6 (24–26). Caco2 cells showed nuclear localization of Carlr at basal conditions; however, when coculturing the intestinal cells with LPS or with activated THP-1 cells, resembling the intestinal inflammatory environment, we saw that Carlr expression was increased and at the same time translocated to the cytoplasm, mainly in the coculture condition (Fig. 3D). These results indicate that LPS-induced Carlr translocation to the cytoplasm is a sign of inflammation of human macrophages and intestinal cells that should be present in the intestinal epithelia of inflammatory patients.
Carlr is involved in the signaling interaction between macrophages and intestinal cells during inflammation
As interactions between intestinal cells and macrophages induced cytoplasmic translocation of Carlr in the intestinal cells, we sought to evaluate the role of Carlr in the interplay between human macrophages and intestinal cells in an inflammatory setting. To assess the involvement of Carlr in this interaction, we silenced this lncRNA in THP-1 cells using two different short interfering RNAs (Supplemental Fig. 3A). Upon LPS stimulation of THP-1 cells with Carlr knockdown, we observed that two of the putative Carl-regulated genes whose expression correlates with Carlr, IL1B, and PTGS2 were not induced following silencing (Fig. 4A). Both of these genes are important players in inflammatory intestinal diseases, and their activation induces secretion of PG and the IL-1β inflammatory cytokine by THP-1 cells, soluble factors that modulate interactions between epithelial and immune cells. To further analyze the relationship between THP-1 and intestinal cells, we used a coculture model of intestinal inflammation (27). After coculturing LPS-stimulated THP-1 cells with intestinal Caco2 cells, we observed an increase in proliferation of the intestinal cells (Fig. 4B) that was diminished when Carlr was previously silenced in THP-1 cells (Fig. 4C). This decrease in the proliferation of Caco2 cells could be mediated by the lower levels of IL-1β and PTGS2 observed in the Carlr knockdown cells. To further evaluate the effect of silencing Carlr on the interaction of macrophage and intestinal cells, we quantified the amount of NF-κB nuclear translocation in Caco2 cells. THP-1 cells release inflammatory cytokines, inducing an inflammatory cascade in intestinal cells with subsequent translocation of NF-κB to the nucleus. This inflammatory sequence resembles what is observed in intestinal inflammatory diseases where NF-κB activity is increased (28, 29). When Carlr was silenced in THP-1 cells, we observed that translocation of NF-κB was diminished in intestinal cells following coculture (Fig. 4D), consistent with the decrease in proliferation and lower induction of IL1B and PTGS2 in THP-1 cells. Taken together, these results suggest that Carlr is an important player in the signaling between the macrophages and intestinal cells in an inflammatory environment.
(A) LPS induction of IL1B and PTGS2 genes is inhibited in THP-1 cells silenced for Carlr. Data represent mean ± SEM of three independent experiments. (B) Crystal violet staining of Caco2 cells incubated under different conditions for 48 h in 24-well plates. (C) Proliferation of Caco2 cell after culturing with conditioned media of LPS-stimulated THP-1 cells with and without Carlr silencing. Data represent the mean ± SEM of three independent experiments. (D) p50 and p65 translocation to the nucleus in Caco2 cultured alone and with THP-1 cells under different conditions. +p ≤ 0.1, *p < 0.05, **p < 0.01.
(A) LPS induction of IL1B and PTGS2 genes is inhibited in THP-1 cells silenced for Carlr. Data represent mean ± SEM of three independent experiments. (B) Crystal violet staining of Caco2 cells incubated under different conditions for 48 h in 24-well plates. (C) Proliferation of Caco2 cell after culturing with conditioned media of LPS-stimulated THP-1 cells with and without Carlr silencing. Data represent the mean ± SEM of three independent experiments. (D) p50 and p65 translocation to the nucleus in Caco2 cultured alone and with THP-1 cells under different conditions. +p ≤ 0.1, *p < 0.05, **p < 0.01.
Carlr interacts with the active form of NF-κB
To further characterize Carlr function, we evaluated whether this lncRNA is able to interact with protein complexes using a native agarose gel EMSA. We observed changes in the migration pattern of in vitro–transcribed Carlr upon mixing with protein lysates from THP-1 cells, indicating that it is able to interact with protein complexes (Fig. 5A). As we are interested in the NF-κB pathway and previously saw that Carlr modulates NF-κB signaling, we performed a computational prediction using RPISeq that predicts RNA protein interaction using datasets extracted from the Protein–RNA Interface Database (30). This analysis suggested that human Carlr preferentially interacts with p65 NF-κB, rather than p50 or IκBα (Fig. 5B). To experimentally evaluate the ability of Carlr to interact with these members of the NF-κB family, we performed RIP experiments in human THP-1 cells. p65 Ab immunoprecipitation showed that Carlr bound to the p65 member of the NF-κB complex (Fig. 5C). Interestingly, although we were able to coimmunoprecipitate the inactive NF-κB complex, that is, NF-κB bound to IκBα, we did not detect significant enrichment of Carlr, suggesting that Carlr interacts with NF-κB only after it is released from the inhibitory complex. Taken together, these data suggest that Carlr is able to interact with protein complexes, and that it binds NF-κB p65 after it is released from the inhibitory complex, before translocation to the nucleus.
(A) Nondenaturing agarose gel for in vitro transcription of Carlr incubated with different protein lysates shows that Carlr is able to interact with protein complexes. (B) Interaction prediction of Carlr with NF-κB members p65 and p50 and with the inhibitory protein IκBα using RPISeq tool (30); tubulin is used as an unrelated protein. (C) Levels of Carlr bound to NF-κB and IκBα after RIP performed in THP-1 cell lysates. Data are represented as mean ± SEM of three independent immunoprecipitations. Representative Western blot of the immunoprecipitation (left). *p = 0.05 (right). cyt, cytoplasm; nuc, nucleus; n.s., not significant.
(A) Nondenaturing agarose gel for in vitro transcription of Carlr incubated with different protein lysates shows that Carlr is able to interact with protein complexes. (B) Interaction prediction of Carlr with NF-κB members p65 and p50 and with the inhibitory protein IκBα using RPISeq tool (30); tubulin is used as an unrelated protein. (C) Levels of Carlr bound to NF-κB and IκBα after RIP performed in THP-1 cell lysates. Data are represented as mean ± SEM of three independent immunoprecipitations. Representative Western blot of the immunoprecipitation (left). *p = 0.05 (right). cyt, cytoplasm; nuc, nucleus; n.s., not significant.
Carlr is associated with CeD pathogenesis
Based on its role in the NF-κB pathway and its expression in the human gastrointestinal tract, macrophages, and intestinal cells, we assessed the potential involvement of Carlr in CeD. CeD is an immune-mediated enteropathy that presents with constitutive NF-κB activation (31). To assess the role of lncRNAs in inflammatory bowel diseases, and specifically in CeD, we evaluated the expression of the putative Carlr targets in intestinal biopsies of celiac patients and control individuals. Three of the putative Carlr-regulated targets, that is, TNFAIP3, IL1B, and PTGS2, showed statistically significant, higher expression (p ≤ 0.05) in celiac patients compared with control samples (Fig. 6A). These results suggest that human Carlr could be involved in the dysregulation of these genes during CeD pathogenesis. Next, we measured the expression of Carlr in intestinal biopsies and, contrary to our expectations, we saw that Carlr levels were decreased (p = 0.018) in celiac patients (Supplemental Fig. 3C). As NF-κB is constitutively active in the celiac mucosa (28, 31), we hypothesized that this signaling pathway could make Carlr exclusively cytoplasmic in celiac patients, enabling its function as an inducer of certain disease-characteristic NF-κB pathway genes. To test this hypothesis, we isolated the nuclei of intestinal biopsy cells and quantified the fraction of Carlr in the nucleus compared with the whole cell. As predicted, celiac biopsies showed elevated cytoplasmic localization of Carlr, whereas in control biopsies Carlr was mainly located in the nucleus (Fig. 6B). Moreover, as previously seen in vitro, Carlr was found to interact with active NF-κB complex in both control (noninflamed) and celiac patient (inflamed) intestinal biopsies (Fig. 6C) but could not bind the inactive complex (Supplemental Fig. 3D). Taken together, these findings indicate that elevated Carlr cytoplasmic activity is associated with disease-characteristic elevated NF-κB signaling in CeD tissue.
(A) CeD patient intestinal biopsies have significantly higher levels of the Carlr targets measured by RT-qPCR. RPLPO was used as housekeeping gene. White circles show the medians; box limits indicate the 25th and 75th percentiles as determined by R software; whiskers extend 1.5-fold the interquartile range from the 25th and 75th percentiles; polygons represent density estimates of data and extend to extreme values. **p < 0.05, *p = 0.05; unpaired Student t test. (B) Subcellular localization of Carlr in cells isolated from biopsies of five CeD patients and three controls. RPLPO was used as a control; p = 0.023, unpaired Student t test. (C) Levels of Carlr bound to NF-κB after RIP performed in lysates from three control (noninflammed) and three CeD patients (inflamed) (right). Representative Western blot of the immunoprecipitation (left).
(A) CeD patient intestinal biopsies have significantly higher levels of the Carlr targets measured by RT-qPCR. RPLPO was used as housekeeping gene. White circles show the medians; box limits indicate the 25th and 75th percentiles as determined by R software; whiskers extend 1.5-fold the interquartile range from the 25th and 75th percentiles; polygons represent density estimates of data and extend to extreme values. **p < 0.05, *p = 0.05; unpaired Student t test. (B) Subcellular localization of Carlr in cells isolated from biopsies of five CeD patients and three controls. RPLPO was used as a control; p = 0.023, unpaired Student t test. (C) Levels of Carlr bound to NF-κB after RIP performed in lysates from three control (noninflammed) and three CeD patients (inflamed) (right). Representative Western blot of the immunoprecipitation (left).
Discussion
Recent studies have implicated an increasing number of lncRNA molecules in inflammation and immunity. In this study, we characterize the role of the lncRNA Carlr in the NF-κB pathway in the context of intestinal inflammation. Mouse Carlr was first identified in cardiomyocytes, where it exhibits decreased expression in response to anoxia (13). In the present study, we found that Carlr expression is increased in mouse and human macrophages in response to LPS stimulation and that this increase is dependent on NF-κB activation, highlighting the cell type and stimuli specificity of lncRNA expression. Indeed, we found that Carlr is translocated to the cytoplasm after NF-κB activation in mouse and human macrophages. Several lncRNAs have been described to reside within, or to be dynamically shuttled, to the cytoplasm where they regulate protein localization, mRNA translation, and stability (32), so we wanted to further analyze the function of this lncRNA and its link with inflammation.
In mammalian gut tissue, enterocytes are closely associated with different types of immune cells, including tissue macrophages that are located in the lamina propia (23). For instance, macrophages exposed to LPS secrete cytokines and growth factors that can directly, or indirectly, affect epithelial function (22). Therefore, macrophages play an important role in immune and inflammatory events of the intestinal mucosa. Interestingly, human Caco2 intestinal cells cultivated with activated macrophages also show a translocation of Carlr to the cytoplasm, suggesting that the cytoplasmic location of this lncRNA could be a hallmark of intestinal inflammation.
To further characterize the function of Carlr, we assessed its role in NF-κB pathway gene expression. We observed that downregulation of Carlr decreases the levels of some NF-κB pathway genes and also influences the translocation of the transcription factor to the nucleus, suggesting that it contributes to the perpetuation of the signaling cascade. Knockdown of Carlr in human THP-1 cells reduces the levels of IL1B and PTGS2 gene expression induced by LPS. These two genes are important regulators of intestinal inflammation; they are present at high levels in the inflammatory microenvironment. PTGS2 has been linked with inflammation-associated cancer development and proliferation, whereas IL-1β is able to induce the NF-κB pathway, two major events in the development of intestinal inflammatory diseases (33, 34). Interestingly, intestinal cells cultured in conditioned media from Carlr knockdown macrophages show less proliferation and a decrease in NF-κB translocation to the nucleus, suggesting that Carlr is important in the interaction between macrophages and intestinal cells that occurs during the development of inflammatory diseases.
RNA EMSA and immunoprecipitation experiments revealed that Carlr lncRNA is able to interact with proteins, specifically the activated NF-κB transcription factor complex. We hypothesize that the cytoplasmic form of Carlr could recognize IκBα-free NF-κB and facilitate its transport to the nucleus, thus augmenting the signaling pathway. However, further studies are needed to decipher how and under what specific condition Carlr modulates NF-κB activation.
Finally, we wanted to assess the relevance of Carlr to human inflammation and immunity. CeD is an immune-mediated enteropathy in which NF-κB is constitutively active (31), and it is a useful disease model to assess the potential involvement of this lncRNA in inflammatory diseases of the gut. When we explored Carlr expression in healthy and diseased human tissue, we surprisingly found that its expression is not increased in the patient tissue. However, further analyses revealed that in celiac patients most Carlr is located in the cytoplasm, whereas in control individuals it is mainly found in the nucleus. Experiments done in vitro showed that long-term LPS stimulation does not induce such an increase of Carlr, suggesting that after induction and translocation to the cytoplasm, the levels of this lncRNA remain stable, thus explaining the lack of overexpression in the disease tissue. The phenotype of Carlr cytoplasmic localization in celiac patients resembles our results with macrophage–intestinal cell cocultures, suggesting that although Carlr expression is not elevated in CeD patients, it may still mediate an inflammatory function. Additionally, we found that the Carlr targets IL1B and PTGS2 are overexpressed in CeD patient biopsies, in line with the increased NF-κB activation and proliferation described in this disease (28, 35). Moreover, Carlr also interacts with activated NF-κB in intestinal biopsies, supporting the hypothesis that it may play a role in the development of inflammation.
Taken together, these results implicate Carlr as a potential novel player of the NF-κB inflammatory pathway and a clinically intriguing candidate associated with the pathogenesis of immune-mediated diseases.
Footnotes
This work was supported by National Institutes of Health Grant R01 DK102180 and by a Juan de la Cierva reincorporation fellowship from the Spanish Ministry of Economy and Competitiveness.
The online version of this article contains supplemental material.
References
Disclosures
The authors have no financial conflicts of interest.