Non-coding RNA LINC01010 regulates macrophage polarization and innate immune functions by modulating NFκB signaling pathway

Abstract

Innate immune response is regulated by tissue resident cells or infiltrated immune cells such as macrophages (Mφ) that play critical role in normal tissue development, homeostasis, and repair of damaged tissue. However, the epigenetic mechanisms that regulate Mφ plasticity and innate immune functions are not well understood. Long noncoding RNA (lncRNA) are among the most abundant class of transcriptome but their function in myeloid cell biology is less explored. In this study, we deciphered the regulatory role of previously uncharacterized lncRNAs in Mφ polarization and innate immunity responses. Two LncRNAs showed notable changes in their levels during M1 and M2 Mφ differentiation. Our findings indicate that LINC01010 expression increased and AC007032 expression decreased significantly. LINC01010 exhibit myeloid cell-specificity, while AC007032.1 is ubiquitous and expressed in both myeloid and lymphoid (T cells, B cells and NK cells). Expression of these lncRNAs is dysregulated in periodontal disease (PD), a microbial biofilm-induced immune disease, and responsive to LPS from different oral and non-oral bacteria. Knockdown of LINC01010 but not AC007032.1 reduced the surface expression of differentiation markers CD206 and CD68, and M1Mφ polarization markers MHCII and CD32. Furthermore, LINC01010 RNAi attenuated bacterial phagocytosis, antigen processing and cytokine secretion suggesting its key function in innate immunity. Mechanistically, LINC01010 knockdown Mφ treated with E. coli LPS exhibit significantly reduced expression of multiple NFκB pathway genes. Together, our data highlight function role of a PD-associated lncRNA LINC01010 in shaping macrophage differentiation, polarization, and innate immune activation.

Graphical Abstract

Introduction

The ability of immune cells to stimulate adept immune responses is crucial for clearing pathogens and maintenance of immune homeostasis. Myeloid inflammatory cells, including monocytes, macrophages (Mφ), and dendritic cells (DC) are key activators of innate immune responses by recognition of pathogens and bridge adaptive arm of immunity (1–3). The pathogens interacting with specific receptors (viz toll-like receptors; TLRs) located at these immune cells activate a plethora of downstream signals, which induces the differentiation, polarization, and their imperative innate immune functions (4–6). Identification of endogenous regulatory molecules that control differentiation and function of these multifunctional immune cells is necessary to understand the immune homeostasis. Various classes of non-coding RNAs are acknowledged as key regulatory factors in diverse biological pathways including differentiation, apoptosis, immunity, and inflammation, etc (7, 8). The non-coding transcripts with size >200 nucleotides are an emerging class of regulatory RNA defined as lncRNAs (9–12). Unlike protein coding genes, lncRNAs exhibit high tissue specificity, poor sequence conservation, and relatively lower expression. In addition, lncRNAs are present in both the nuclear and cytoplasmic compartments wherein, they can control transcriptional, post-transcriptional, and translational output of the genome (9–14). Not surprisingly, aberrations in the expression of lncRNA have been linked to the occurrence of diseases such cancer, neurological disorders, autoimmune, etc., (15–22).

The role of lncRNAs in regulating Mφ differentiation, polarization towards M1-like or M2-like phenotype and their associated innate immune functions remains less explored. M1-like Mφ exhibit a pro-inflammatory phenotype, while M2-like Mφ are involved in anti-inflammatory or reparative in function (23, 24). Few studies have highlighted the lncRNA-mediated regulation of Mφ polarization. For instance, lncRNA RP11–361F15.2 is expressed in osteosarcoma and promotes CPEB4-mediated tumorigenesis and M2-like polarization of tumor-associated macrophages through miR-30c-5p (25). LncRNA ANCR inhibits the M1-like macrophage polarization by targeting FOXO1 expression, which promotes the invasion and migration in gastric cancer (26). Also, lncRNA COX2 prevents the immune evasion and metastasis in hepatocellular carcinoma by altering M1/M2 macrophage polarization. Two different mouse cancer hepatic cell line Hepal-6 and HepG2 were incubated with RAW264.7 M1 and RAW264.7 M2 Mφ. The expression of lncRNA Cox-2 was predominantly expressed in RAW264.7 M1 compared with M2 Mφ. The inhibition of lncRNA Cox-2 by RNAi promotes the polarization of M2 Mφ along with an increase of proliferation and inhibition of apoptosis in hepatocellular carcinoma (27). Functional significance of lncRNA-mediated pathways regulating the differentiation, M1/M2 polarization and is still largely unexplored.

We have previously identified hundreds of differentially expressed lncRNAs during M1 and M2 Mφ differentiation including known and novel sequences. In this study, we characterized hitherto unknown biological functions of LINC01010, a cell-type specific lncRNA, in Mφ biology. LINC01010 expression is responsive to periodontal disease and exhibit time-dependent variation in response to oral and non-oral bacteria. Our results show novel biological functions of a Mφ-enriched lncRNA LINC01010 that regulates differentiation, polarization, and innate immune activity by modulating NFκB pathways related genes.

MATERIALS AND METHODS

Primary human monocyte isolation and differentiation

Freshly prepared buffy coats were collected from healthy donors (Sylvan N. Goldman, Oklahoma Blood Institute, Oklahoma City, OK, USA) and CD14+ monocytes were obtained by density gradient centrifugation and magnetic bead isolation. In brief, PBMCs were purified by use of Ficoll Paque (GE Healthcare, Piscataway, NJ, USA)-based density centrifugation. PBMCs were incubated with magnetically labeled CD14 beads (Miltenyi Biotec, Cologne, Germany), according to the manufacturer’s instructions. Monocyte purity and viability were >95%, as determined by flow cytometry (Supplementary Figure S1a). Monocytes were plated at a density of 2X10⁶/ml in DMEM, supplemented with penicillin (100 U/ml), streptomycin (100 mg/ml), and gentamicin (50 mg/ml) and after 2h substituted with media containing 10% heat-inactivated FBS (Life Technologies, Grand Island, NY, USA) and rhGM-CSF or rhM-CSF (both 50 ng/ml; PeproTech, Rocky Hill, NJ, USA) for M1 and M2 macrophage, respectively. Cells were harvested at 18h, day 3, 5, and 7 for total RNA isolation (28–32). At day 7, cells were harvested, and differentiation confirmed by flow cytometric analysis of CD68 and CD206 (Supplementary Figure S1b).

Immune cell sorting

Isolated PBMCs from human buffy coats were resuspended in FACS staining buffer (PBS-1%BSA) in 15 mL tubes. Cells were stained with anti-CD3 (FITC anti-human CD3 Antibody; clone OKT3; Bio Legend), anti-CD19 (PE anti-human CD19; clone HIB19; Bio Legend), anti CD8 (Alexa Flour 700 anti-CD8 Antibody; clone RPA-T8; BD Pharmingen^™) and anti CD56 (APC anti-CD56 Antibody; clone 5.1 H11; Bio Legend) antibodies to sort T cells, B cells and NK cells, respectively. After 45 mins of incubation at 4°C cells were centrifuged at 1500 rpm for 5 min to remove unbound antibodies and washed three times with FACS staining buffer. In parallel, PBMCs were stained with single color staining controls to set-up the flow-sorter. Typically, 25×10⁶/ml were used for the flow sorting per tube. Prior to flow-sorting, cells were passed through 70μ filter to remove cell aggregates from and prevent capillary clogging during the flow-sorting. Each cell type was collected in complete RPMI media supplemented with 10% FBS in flow tubes. First CD3+ and CD8+ cells were sorted and then CD56+ cells were sorted from CD3 negative fraction to obtain the NK cells without NKT fraction. A detailed gating strategy for lymphoid cells sorting and their purity is provided in Supplementary Figure S2.

Study population and sample collection

The present investigation was an observational, cross-sectional study approved by the Institutional Review Board and the Ethics Research Committee at the University of Illinois Chicago, College of Dentistry (IRB Protocol# 2015–1093; 33). The study was conducted according to the ethical principles of the Helsinki Declaration. All participants were informed of the aims of the study and signed the informed written consent form prior to entering the study. The present study comprised of 21 individuals from a multiethnic group, of both sexes, and divided into two groups: periodontally healthy (Control, Healthy (H); n = 8) or chronic periodontitis (Experimental, Diseased (D); n = 13). For the diseased group, a biological sample (gingival tissue including the gingival epithelium, col area and underlying connective tissue) were obtained at baseline. For the control group (H), samples were derived from healthy gingival tissues normally discarded during routine crown lengthening procedures. Crown lengthening is a common surgical procedure that removes a collar of gingival tissue to reestablish proper tooth length and width for placement of a tooth filling, fabrication of a crown, or to address esthetic concerns due to excess gingival tissues. Inclusion criteria included male and female patients ages 18–65 years and in good systemic health. Exclusion criteria included chronic disease (diabetes, hepatitis, renal failure, clotting disorders, HIV, etc.), smokers, antibiotic therapy for any medical or dental condition within a month before screening, and subjects taking medications known to affect periodontal status (e.g., phenytoin, calcium channel blockers, cyclosporine).

Total RNA extraction, cDNA synthesis and qPCR analysis

Cells were washed three times with PBS, and 700 μl of TriZol reagent (Invitrogen, CA, USA) was added to 24-wells culture plate in each condition. Total RNA was isolated from 18 h, day 3, day 5, and day 7 differentiated cells using miRNeasy micro kit (Qiagen, Gaithersburg, MD, US). A total of 250 ng RNA was used to synthesize cDNA, which was synthesized from high-capacity cDNA Reverse transcription kit (ThermoFisher Scientific, Grand Island, NY, USA). The expression levels of AC007032.1 LINC01010, and beta-actin genes were analyzed by qPCR reaction using SYBR Green Gene Expression Master Mix (Applied Biosystems, USA) in a StepOne 7500 thermocycler (Applied Biosystems, USA). The Ct values of three replicates were analyzed to calculate fold change using the 2^−ΔΔCt method. The expression levels of LINC01010 and AC007032.1, and β-actin genes were also analyzed by RT-qPCR in monocytes-derived M1 and M2 Mφ challenged with Porphyromonas gingivalis-LPS (Pg-LPS) (1μg ng/ml), Aggregatibacter actinomycetemcomitans (Aa-LPS) (100 ng/ml) or E. coli LPS (100 ng/ml) for 4 and 24 h (32). Cells were harvested and treated with Qiazol for isolating of total RNA from each sample using the miRNeasy Micro Kit (Qiagen), according to manufacturer’s instructions. First-strand cDNA was synthesized from 500 ng total RNA using the high-capacity cDNA Reverse Transcription kit (ThermoFisher Scientific, Grand Island, NY, USA).

Transient siRNA transfection

Transient transfection of siRNA was performed using Lipofectamine 2000 (Invitrogen-Life Technologies Corporation, Carlsbad, CA, USA), as per manufacturer’s instructions. siRNAs were used at final concentration of 100 nM (28–30). As a transfection positive control, siGLO Red Transfection Indicator is a fluorescent oligonucleotide duplex labeled with DY-547 (ThermoFisher Scientific, Grand Island, NY, US). Two siRNAs were designed for lncRNA LINC01010 and AC007032.1 using IDT siRNA tool (Integrated DNA Technologies Coralville, IA). After 36 h of transfection, total RNA was isolated using miRNeasy kit (Qiagen, Gaithersburg, MD, US). A total of 250 ng RNA was reverse transcribed using high-capacity cDNA Reverse transcription kit (ThermoFisher Scientific, Grand Island, NY, USA). The expression levels of LINC01010 and AC007032.1, and β-actin genes were analyzed by RT-qPCR using SYBR Green Gene Expression Master Mix (Applied Biosystems, USA) in a StepOne 7500 thermocycler (Applied Biosystems, USA). The Ct values of three replicates were analyzed to calculate fold change using the 2^−ΔΔCt method.

Flow cytometry

Cells were harvested and washed in ice-cold PBS supplemented with 1% (v/v) FBS and 0.08% sodium azide. Cellular debris were excluded based on size (forward scatter [FSC]) and granularity (side scatter [SSC]). The FSC/SSC gate for Monocyte comprised ~60%, total events. Couplets were excluded based on SSC versus FSC and SSC versus pulse width measurements. M1Mφ and M2Mφ were stained for cell surface markers with FITC, PE, and APC conjugated antibodies. For polarization analysis, human antibodies for CD32-PE, HLA-DR-FITC (both M1 marker), CD163-APC, CD209-PE (M2 marker) were purchased from BD Pharmingen (San Diego, CA, USA) or BioLegend (San Diego, CA, USA). Unstained and isotype control (BD Pharmingen (San Diego, CA, USA) were used as controls. Samples were analyzed using a BD Accuri^™ C6 flow cytometer (BD Biosciences, San Jose, CA. Further analysis was performed using FlowJo software (Tree Star, Ashland, OR). Cells were gated according to their forward scatter (FSC) and side scatter (SSC) properties including the larger cells with high granularity and excluding the small-sized debris with a low SSC and FSC shown at the bottom left corner of the dot plot.

Viability assays

Cell viability was determined by use of the CellTiter 96 AQueous Cell Proliferation Assay Kit (Promega, Madison, WI, USA). In brief, Mφ were plated at 400,000/well in 96-well plates and transfected as described above and assays performed after 36 h, according to the manufacturer’s instructions.

Bacterial phagocytosis assay

Monocytes-derived M1 and M2 macrophages (400,000/well, 96-well plate) were transfected with siRNAs targeting LINC01010, AC007032.1 or control siRNA on day 6. Transient transfection of siRNAs (at a final concentration of 100 nM) was performed using Lipofectamine 2000 (Invitrogen-Life Technologies Corporation, Carlsbad, CA, USA), as per manufacturer’s instructions. Phagocytosis assay was performed with pHRodo Rhodamine conjugated E. coli (Invitrogen, Carlsbad, CA) 36 h post-transfection, according to the manufacturer’s protocol (28–30). pHrodo dye conjugates are non-fluorescent outside the cell, but fluoresce in the acidic environment within phagosomes, specifically detecting intracellular phagocytosis. Briefly, the labeled E. coli bioparticles were resuspended in Live Imaging Buffer (Life Technologies) at a final concentration of 1 mg/ml and homogenized by sonication for 2 min and resuspended in culture media. Cells were incubated with labeled E. coli for 2 h at 37 °C. To remove loosely bound bacteria on the cell surface, cells were washed thrice with 1X PBS. Finally, random images were captured by EVOS microscopy system for each donor (n=3) (ThermoFisher Scientific, Grand Island, NY, US) and the rhodamine expression was analyzed by flow cytometry.

Cytokine analysis

Supernatants were collected from M1Mφ and M2Mφ transfected with LINC01010, AC007032.1, and control siRNA after challenge with pHRodo Rhodamine conjugated E. coli exposure for 4h. The cytokine/chemokine levels were analyzed by multiplex assays. Multiplex analysis of IL-6, IL-8, IL-1β and TNF-α was performed using Milliplex MAP Human Cytokine/Chemokine Magnetic Bead Panel (Millipore, Billerica, MA, USA). Data were collected on Bio-Plex Flow cytometer (Bio-Rad, Hercules, CA, USA).

Antigen uptake and processing assay

Monocytes-derived M1 and M2 macrophages (400,000/well, 96-well plate) were transfected by si-AC007032.1, si-LINC01010 or control siRNA. After 36 h post-transfection, cells were treated with DQ^™-conjugated Ova (1mg/ml, Molecular Probes, Grand Island, NY) in respective complete media for 2 h at 37 °C. DQ^™-Ova consist of Ova that are heavily conjugated with BODIPY, resulting in self-quenching. Upon proteolytic degradation of DQ-Ova to single dye-labeled peptides, bright green fluorescence is observed in the endosomes. After assay incubation, cells were washed thrice with PBS, fixed with 2% PFA and Bodipy expression was analyzed by flow cytometry. Also, random images were captured by EVOS microscopy system for each donor (n=3) (ThermoFisher Scientific, Grand Island, NY, US) to show the BODIPY florescence levels.

PCR array

Monocytes-derived M2 macrophages transfected with si-LINC01010, si-AC007032.1 or si-control were challenged with E. coli LPS (100 ng/ml). After 24 h post-treatment, cells were harvested and total RNA was isolated using miRNeasy Micro Kit (Qiagen), according to manufacturer’s instructions. First-strand cDNA was synthesized from 500 ng total RNA using the High-Capacity cDNA Reverse Transcription kit (ThermoFisher Scientific, Grand Island, NY, USA). A PCR array plate (96 well), containing 88 primer sets directed against NFκB signaling pathways-related genes and 8 housekeeping primer sets (Real Time Primers, LLC Elkins Park, PA). Real-time PCR was performed using a StepOne 7500 thermocycler (Applied Biosystems, Carlsbad, CA). PCR array data were analyzed using the Qiagen GeneGlobe Data Analysis Center (https://www.qiagen.com/us/shop/genesand-pathways/data-analysis-center-overview-page/). Expression levels were normalized with respect to beta 2 microglobubin (B2M) as housekeeping because it demonstrated the most consistent levels across all transfections. Next, the fold change was calculated with respect to the control siRNA. Real-time PCR was also carried out in three independent donors.

Statistical Analysis

All the data were analyzed and plotted using GraphPad Prism (La Jolla, USA). The results are represented as SD or SEM from three independent replicates. P-values were calculated using Students t-test, and P <0.05 was considered significant. *P < 0.05, **P < 0.01, ***P < 0.001.

RESULTS

LINC01010 and AC007032.1 are differentially expressed during monocyte-to-macrophage differentiation and exhibit cell-type specific expression

We recently performed global lncRNAs profiling of differentiating monocyte derived M1 and M2 Mφ and identified many differentially expressed known and novel lncRNAs (Valverde et al., Submitted; Gene Expression Omnibus Accession Number GSE192642). We selected two differentially expressed lncRNAs LINC01010 and AC007032.5 that were up- and down-regulated, respectively, in both M1 and M2 Mφ (Fig. 1A, B). LINC01010, barely detectable in monocytes, increases progressively during M1Mφ differentiation and peaks at day 7, while in M2Mφ its levels peak at 18 h and maintained at later time points. On the contrary, AC007032.5 expression exhibit striking similarity between M1 and M2 Mφ with marked reduction at 18 h and decrease further until day 7. We confirmed LINC01010 and AC007032.5 expression in monocytes and differentiating M1/M2Mφ in an independent cohort (n=4/group) by RT-qPCR and observed that their expression corroborates with RNAseq profiling (Fig. 1C, D).

Fig 1. LncRNAs LINC01010 and AC007032.1 are differentially expressed during macrophage differentiation and exhibit cell restricted expression.

LncRNA expression profiling was performed in differentiating M1 and M2 Mφ (18h, day 3, 5 and 7) using RNA-seq analysis. Line graphs showing FPKM values of (A) LINC01010 and (B) AC007032.1 in M1 and M2 Mφ. The accuracy of the screening was based on fold change (a cut off between −1.25 and 1.25) and significance (p≤0.05 using ANOVA) to identify the differentially upregulated and downregulated lncRNA. Validation of RNAseq data by RT-qPCR in an independent cohort. Total RNA isolated from a separate cohort of monocytes and differentiating M1 and M2 Mφ (n=4) was examined for the expression of (C) LINC01010 and (D) AC007032.1 by RT-qPCR. Actin expression was used as housekeeping gene. Ct values of three replicates were analyzed to calculate fold change using the 2^−ΔΔCt method. Expression profiles of (E) LINC01010 and (F) AC007032.1 in different myeloid and lymphoid cell types. CD14+ monocytes, day 7 differentiated M1 and M2 Mφ (myeloid cells), CD3+ T cells, CD19+ B cells and CD56+ NK cells were isolated from healthy human donors. Expression of lncRNAs LINC01010 and AC007032.1 was quantified by RT-qPCR. Actin expression was used as housekeeping gene. Fold change was calculated using 2^−ΔΔCt method. Data are means ± SEM of four independent donors. Student’s t-test was conducted to calculate p-values (*p < 0.05, **p < 0.01, ***p < 0.001).

To examine cell-type specificity of candidate lncRNAs, we quantified the comparative expression of these lncRNAs in myeloid (monocyte, M1 and M2 Mφ) and lymphoid cell (CD3+ T cells, CD19+ B cells and CD56+ NK cells) types. LINC01010 was predominantly expressed in M1 and M2 Mφ, while its levels were barely detectable in any of the lymphoid cells examined (Fig. 1C). Conversely, lncRNA AC007032.5 was detected in both myeloid and lymphoid cells; however, its expression was markedly lower in lymphoid compartment compared to myeloid cells (Fig. 1D). These results show that LINC01010 and AC007032.5 exhibit myeloid-cell type specificity.

Expression of LINC01010 and AC007032.1 is dysregulated in periodontal disease and responsive to different TLR ligands

Periodontal disease (PD) is microbial biofilm-mediated inflammatory disease caused due to destruction of tissues supporting tooth and affects more than 750 million people worldwide (34–36). Overt infiltration and inflammatory activity of myeloid cells, in particular Mφ, is associated with PD progression and severity. P. gingivalis (Pg) and A. actinomycetemcomitans (Aa) are two key etiological bacteria involved in periodontal pathogenesis (37–39). Information on lncRNA involvement in PD pathogenesis and the role of periodontal bacteria in lncRNA expressional perturbation remains poorly understudied (40–42). Therefore, we examined the expression of candidate lncRNAs is impaired in PD. Compared to periodontally healthy gingiva, we observed significantly higher expression of LINC01010, while AC007032.1 levels were reduced (Fig. 2D, E).

Fig 2. LINC01010 and AC007032.1 exhibit aberrant expression in inflamed gingiva and are responsive to TLR ligation.

Histograms showing expression of lncRNA LINC01010 and AC007032.1 in periodontally healthy and diseased gingival biopsies. Total RNA was isolated from healthy and diseased tissues (n=8 subjects/group) and the expression of (A) LINC01010 and (B) AC007032.1 was quantified by RT-qPCR. M1 and M2 Mφ were challenged with (C) A. actinomycetemcomitans LPS (100 ng/ml), (D) P. gingivalis LPS (1μg/ml] and (E) E. coli LPS (100 ng/ml) derived liposaccharide for 4 and 24 h. Time-kinetics of LINC01010 and AC007032.1 expression was examined by RT-qPCR. Actin expression was used as housekeeping gene. Ct values of three replicates were analyzed to calculate fold change using the 2^−ΔΔCt method. Data are means ± SEM of three independent donors. Student’s t-test was conducted to calculate p-values (*p < 0.05, **p < 0.01, ***p < 0.001).

We next asked whether these lncRNAs are important in innate immune regulation. LncRNA expression was quantified in response to lipopolysaccharide (LPS) from Pg (a TLR2/4/6 ligand), Aa (a TLR4 ligand), two key etiological bacterial pathogens causing periodontal disease, and E. coli (a TLR4 ligand), a non-oral bacterium (42). M1 and M2 Mφ were challenged with Pg-LPS (1 μg/ml), Aa-LPS (100 ng/ml) and E. coli LPS (100 ng/ml) for 4 and 24 h and the expression of candidate lncRNAs were examined by RT-qPCR. We observed that LINC01010 expression was significantly upregulated in Aa-LPS challenged M1Mφ at 4- and 24h; however, no significant differences were observed in M2Mφ compared to untreated Mφ (Fig. 2A; upper panel). M1Mφ challenged with Pg-LPS exhibit significant upregulation of LINC01010 at 24 h. We also observed significant upregulation of LINC01010 in M2Mφ at both times compared to untreated Mφ (Fig. 2B; upper panel). Expression of LINC01010 showed time-dependent increase in E. coli LPS treated M1 and M2 Mφ (Fig. 2C; upper panel). On the contrary, the expression of AC007032.1 in M1Mφ and M2Mφ challenged with Aa LPS, Pg LPS and E. coli LPS showed significant reduction at 4 and 24 h compared to untreated Mφ (Fig. 2A–C; lower panel). Together, these results show that LINC01010 and AC007032.1 exhibit altered expression in PD and respond to different PAMPs suggesting their role in Mφ polarization and innate immune regulation.

Knockdown of LINC01010 impairs macrophage differentiation and polarization

Next, we asked whether the differentially expressed LINC01010 and AC007032.1 affect macrophage differentiation and polarization. To examine the functional impact of lncRNAs, we designed two siRNAs (A and B) for each of these lncRNAs and transfected M2Mφ with either siRNA alone or in combination. Compared to si-Control, all the siRNAs tested showed significant knockdown of target lncRNAs (Fig. 3A). Based on the robust RNAi of target lncRNAs, we selected si-LINC01010B and si-AC007032.1B for our subsequent experiments. Compared to si-control, we noticed similar cell viability indicating that lncRNA targeting siRNAs does not induce apoptosis in cells within the time examined (Supplementary Figure S3).

Fig 3. Knockdown of LINC01010.1 impair differentiation and polarization of macrophages.

(A) Screening of siRNAs of candidates of lncRNAs by qPCR. M2 macrophages were transfected with siRNA A, B or a combination of both targeting LINC01010 or AC007032.1 (at a final concentration of 100 nM). The expression of candidate lncRNAs were normalized by actin as housekeeping. Data are means ± SEM of three independent donors. The Ct values of three replicates were analyzed to calculate fold change using the 2^−ΔΔCt method. Student’s t-test was conducted to calculate p-values. *p < 0.05, **p < 0.01, ***p < 0.001. (B) Knockdown of LINC01010 impair differentiation markers in M2 macrophages. Cells were transfected with si-LINC01010 or si-AC007032.1 and differentiation markers CD206 and CD68 were analyzed by flow cytometry. Histograms showing CD206 and CD68 positive cells in M2 macrophages transfected with si-lncRNA compared to control. Same isotype controls were used for both the siRNAs. Data are means ± SEM of three independent donors. Student’s t-test was conducted to calculate p-values. *p < 0.05, **p < 0.01, ***p < 0.001, **** p < 0.0001. (C) Transient silencing of LINC01010 suppress M1 and M2 macrophage polarization. M1 and M2 macrophages were transfected with si-LINC01010 and si-AC007032.1 and the surface expression of polarization markers were quantified by flow cytometry in M1 and M2 macrophages. Histograms showing percentage positive cells expressing M1 (HLA-DR and CD32) and M2 (CD209 and CD163) polarization markers in lncRNA knockdown cells. Same isotype controls were used for both the siRNAs. Data are means ± SEM of three independent donors. Student’s t-test was conducted to calculate p-values. *p < 0.05, **p < 0.01.

To evaluate if the lncRNAs impact the differentiation process, M2Mφ transfected with si-LINC01010, si-AC007032.1 or si-control were examined for the expression of differentiation markers CD206 and CD68 by flow cytometry. Our results show that the expression of CD206 (~40%) and CD68 (~70%) were significantly reduced in si-LINC01010 compared to si-control (Fig. 3B). However, AC007032.1 knockdown did not show significant changes in the expression of either CD206 or CD68 in M2Mφ.

Next, we examined the impact of candidate lncRNA knockdown on Mφ polarization. siLINC010101B-transfected M1Mφ showed significant reduction in M1 polarization markers, HLA-DR (~20%) and CD32 (~33%) (Fig. 3C; Upper panel), while no significant changes were observed in si-AC007032.1B transfected cells (Fig. 3C; Upper panel). In M2Mφ, the expression of CD209 was significantly reduced (~60%) in LINC01010 knockdown, while CD163 levels remain unchanged compared to si-control (Fig. 3C; lower panel). On the contrary, the expression of CD209 and CD163 were unaffected in si-AC007032.1 transfected cells (Fig. 3C; lower panel). Taken together, our results indicate that LINC01010 play an important role in macrophage differentiation and polarization and its knockdown exhibit more pronounced impact on M1-like phenotype.

LINC01010 knockdown exhibit differential impact on the phagocytosis and antigen processing in polarized macrophages

Macrophages are key phagocytes and after invading pathogens endocytosis, they process and present antigenic epitopes to T cells to activate adaptive immunity (44–49). However, the role of lncRNAs in this regard remains understudied. Therefore, we examined the role of LINC01010 and AC007032.1 in regulating phagocytosis and antigen processing in polarized macrophages. M1Mφ and M2Mφ were transfected with si-LINC01010 and si-AC007032.1 and treated with rhodamine-labeled E. coli to assess their impact on phagocytosis. Knockdown of LINC01010 but not AC007032.1 attenuated E. coli phagocytosis in M1 and M2 Mφ as observed by reduced rhodamine signals under florescence imaging (Fig. 4A; upper panel). We also quantified bacterial uptake by flow cytometry and observed significant reduction (~30% lower geo. MFI) in E. coli phagocytosis by LINC01010 RNAi (Fig. 4B; lower panel). Interestingly, we noticed that reduction in bacterial phagocytosis was more pronounced in M2Mφ in LINC01010 RNAi suggesting a differential biological function of this lncRNA in polarized Mφ.

Fig 4. LINC01010 RNAi attenuates phagocytosis by macrophages.

M1 and M2 macrophages were transfected with control siRNA, si-LINC01010 or si-AC007032.1 and challenged with rhodamine labeled E. coli bioparticles for 4 h to assess phagocytosis. Representative immunofluorescence images showing rhodamine (red) signal indicating phagocytosis in lncRNA knockdown in (A) M1 and (B) M2 macrophages. Final magnification: X40, scale bar corresponds to 100 μm. Quantitative assessment of bacterial phagocytosis by (C) M1 and (D) M2 macrophages after 4 h of incubation with rhodamine labeled E. coli bioparticles and analyzed by flow cytometry. Overlay histograms showing signal intensity in lncRNA RNAi compared to control (upper panel). Bar graphs showing percent geo. MFI values normalized to control (lower panel). Data are means ± SEM of three independent donors. Student’s t-test was conducted to calculate p-values. *p < 0.05, **p < 0.01, ***p < 0.001, **** p < 0.0001.

Next, we assessed the impact of lncRNAs on the uptake and processing of a soluble antigen, ovalbumin. M1Mφ and M2Mφ transfected with si-LINC0101, si-AC007032.1 or si-control were treated with DQ-Ovalbumin after 36 h post-transfection. LINC01010 RNAi showed a marked reduction in ovalbumin processing (green BODIPY signals) in both M1 and M2 Mφ. However, no significant changes in antigen processing were observed in M1Mφ or M2Mφ transfected with si-AC007032.1 (Fig. 5A, B). Flow cytometric analysis showed a significant reduction in BODIPY geo. MFI in M1 (~35%) and M2 (~20%) Mφ in si-LINC01010 transfected (Fig 5C, D). Our results indicate that lncRNA LINC01010 may have an important role in macrophage innate immune response by regulating phagocytosis and antigen processing.

Figure 5. Knockdown of LINC010101 suppress antigen processing by polarized macrophages.

M1 and M2 macrophages were transfected with control, si-LINC01010 or si-AC007032.1 siRNAs and incubated with Bodipy-conjugated ovalbumin (DQ-Ova) for 4h. Representative immunofluorescence images show ovalbumin processing as reflected by green signals in (A) M1 and (B) M2 Mφ. Final magnification: X40, scale bar corresponds to 100 μm. Macrophages were harvested after 4 h of DQ-Ova incubation and analyzed by flow cytometry. Overlay of histograms of lncRNAs or control siRNA transfected (C) M1 and (D) M2 macrophages showing differences in BODIPY dye signals (upper panel). Histograms showing percent of geometrical MFI values in si-lncRNA-transfected M1 and M2 macrophages (lower panel). Data are means ± SEM of three independent donors. Student’s t-test was conducted to calculate p-values. *p < 0.05, **p < 0.01, ***p < 0.001, **** p < 0.0001.

LINC01010 knockdown impairs innate cytokine response in macrophages by perturbing NFκB pathway

Antigen recognition by Mφ triggers activation of signaling cascade that induce innate inflammatory response (50–52). Because we observed impairment in bacterial and antigen processing by LINC010101 knockdown, we decided to examine the impact of M1Mφ and M2Mφ on the secretion of pro-inflammatory cytokines. Supernatants were collected from M1Mφ and M2Mφ transfected with si-LINC010101 and challenged with E. coli for 4 and 24 h. A multiplex analysis of pro-inflammatory cytokines (IL-1β, IL-6, IL-8 and TNF-α) was performed to assess whether LINC01010 impair TLR signaling. In general, most of the cytokine examined showed marked reduction in M1Mφ and M2Mφ transfected with si-LINC01010 (Fig. 6 and Table 1). The secretion of IL-6 and TNF-α showed significant reduction (p<0.05) in si-LINC01010-transfected M1 and M2 Mφ (~90%) at both early (4 h) and late (24 h) time points compared to control siRNA (Fig. 6A). IL-1β levels were reduced (~90%) at late time point in si-LINC01010 transfected M1 and M2 Mφ, while IL-8 only showed significant downregulation (80%) in M2Mφ transfected at 4h compared to control siRNA (Fig. 6A). These results suggest that LINC01010 expression is critical in potentiating innate immune activity in Mφ.

Fig 6. LINC01010 RNAi modulate cytokine response by downregulating various NFκB pathway related genes.

M1 and M2 macrophages were transfected with si-control or si-LINC01010 and challenged with E. coli after 36 h. Supernatants were collected at 4 and 24 h to examine secreted cytokines using multiplex bead array. (A) Histograms showing levels of four pro-inflammatory cytokines IL-1β, IL-6, IL-8 and TNF-α. Data are expressed as mean ± SEM of four independent donors. Student’s t-test was conducted to calculate p-values *p < 0.05, **p < 0.01, ***p < 0.001. (B) Macrophages were transfected with control or si-LINC01010 and challenged with E. coli. After 24 h post-challenge, total RNA was isolated using RNeasy Kit (Qiagen) and reverse transcribed. The expression of NFκB pathway genes was assessed by PCR array containing 88 gene primers. Heatmap showing differentially expressed genes in LINC01010 knockdown cells (n=3/group) compared to control siRNAs. Data was analyzed using the Qiagen GeneGlobe Data Analysis Center (Source: https://www.qiagen.com/us/shop/genesand- pathways/data-analysis-center-overview-page/). B2M was used as housekeeping to normalize the expression levels. Student’s t-test was conducted to calculate p-values (*p < 0.05).

Table 1.

Secreted cytokine levels in M1 and M2 macrophages transfected with control or si-LINC01010 siRNAs and challenged with E.coli for 4h and 24h.

Cytokine response against E. coli LPS is mediated by TLR4 ligation and subsequent activation of NFκB pathway, involving activation of multiple genes. To examine the impact of LINC01010 on NFκB signaling, expression of genes involved in NFκB pathway were quantified in Mφ transfected with si-LINC010101 and challenged with E. coli LPS. Compared to si-control, PCR array data identified ten downregulated genes in M2Mφ transfected with si-LINC010101 (Fig. 6B). These include multiple genes involved in innate immune activation (IRAK1; fold change ≤ −2.23, SELP; fold change ≤ −2.23), cytokines production (CCL2; fold change ≤ −6.31, AGT; fold change ≤ −2.23, IL-10; fold change ≤ −2.23, TNFRSF10B; fold change ≤ −2.23, TNFSF10; fold change ≤ −2.20) and TLR expression (TLR2; fold change ≤ −2.43). Intriguingly, we did not observe any upregulated genes in our dataset. These results show that LINC010101 RNAi impair genes predominantly involved in pro-inflammatory pathways and further suggest an important role in innate immunity.

DISCUSSION

Differentiation of macrophages towards M1 (pro-inflammatory: pathogen recognition, phagocytosis and clearance) and M2 (anti-inflammatory: tissue repair and resolution of overt inflammation) phenotype, and their respective functions depends on coordinated transcriptional and post-transcriptional gene regulation (53). Of late, advances in genomics characterized the important roles of lncRNAs in gene regulation via interacting with DNA, RNA and regulating the three-dimensional nuclear organization (54). Although growing number of reports strongly suggest a critical role of lncRNA expressional variation in immune cell development and phenotype acquisition (55–57). Whether lncRNA expression is needed to initiate the primary monocyte to macrophage M1/M2 differentiation and innate immune responses remain elusive. Therefore, we performed a comprehensive time-kinetics of lncRNA expression profiles during monocyte to M1 and M2 macrophage differentiation (18h, 3d, 5d and 7d) by RNA-seq and identified a large repertoire of lncRNAs including LINC01010 and AC007032.1 (Valverde et al., submitted). Significant upregulation and downregulation of LINC01010 and AC007032.1 during initial time points (18h to 3d; both in RNA-seq and RT-qPCR datasets) strongly suggested them as initial responders towards macrophage lineage commitment. The enrichment of these lncRNAs in myeloid cells rather than cell of lymphoid origin strongly suggested this observation. This observation was further validated by exploring the role if these lncRNAs in chronic periodontitis, a disease marked by high bacterial burden and infiltration of macrophages. Significantly high expression of LINC01010 and concomitant downregulation of AC007032.1 in gingiva from PD subjects strongly reflected the infiltration of macrophages in PD and its potential use as biomarker to monitor disease and resolution post-therapy.

Expression of diverse TLRs on differentiated macrophages confer them to detect and respond to various Pathogen-associated molecular patterns (PAMPs). We examined the impact of different TLR agonists from oral [Pg LPS (TLR2/6) and Aa-LPS (TLR4 agonist)] and non-oral pathogens E. coli LPS (TLR4 agonist) on candidate lncRNA expression and noticed upregulation of LINC01010 and downregulation of AC007032.1 suggesting that their expression level may determine the activation state of M1 and M2 Mφ. Previous studies have demonstrated that lncRNAs Rel (58) and FIRRE (59) respond during LPS-induced TLR activation of macrophage in terms of their expression dynamics. Together, these results suggest that TLR-responsive, myeloid-cell specific lncRNAs can be used as potential diagnostic and prognostic gene target to monitor periodontal disease.

Studies on macrophage differentiation that precedes M1/M2 Mφ polarization are scarce, so far. In a study, Yang et al., reported that lncRNA NTT expression is associated with CD68 expression (25). Recent studies from our lab enlightened the putative involvement of lncRNAs (RN7SK, GAS5 and MALAT1) in macrophage differentiation (41, 60). Based on RNAi studies, we noticed that LINC01010 knockdown, but not AC007032.1 regulates the expression of both differentiation markers CD206 and CD68. We characterized a novel functional lncRNA LINC01010 that regulate macrophage differentiation.

Immature macrophages, under the influence of extracellular milieu, acquire either M1-like or M2-like phenotype. Reduced surface expression of both M1 (HLA-DR and CD32) and M2 (CD209) specific markers in macrophages treated with si-LINC01010 strongly suggests the involvement of this lncRNA in macrophage polarization. Earlier studies also suggest the involvement of lncRNA expression in M1 and M2 macrophage polarization. The overexpression of lncRNA TCONS_00019715 was reported in IFNγ and LPS induced M1Mφ polarization (61). A study by Munteanu et al. underscored the role of lncRNA FENDRR in M1Mφ polarization induced with IFNγ (62). Also, overexpression of lncRNA GAS5 was reported to impair disease resolution in murine model of experimental autoimmune encephalomyelitis (EAE) (63). Interestingly, Ahmad et al. reported the upregulation of lncRNA MALAT1 in both M1 and M2 Mφ, but the knockdown of MALAT1 significantly increased the expression of M2 surface markers rather than M1 (41). The role of lncRNAs LINC01010 identified is apparently more significant because this lncRNA is required throughout the monocyte to macrophage differentiation and polarization events.

Immuno-surveillance via macrophages lies in apposite antigen uptake/processing, and phagocytosis of harmful disease pathogens (64). The involvement of lncRNAs in the regulation of these processes has not been studied exclusively, as only a handful of reports are available. Overexpression of lncRNA Dnm3os was shown to promote inflammation and phagocytosis of macrophages (65). A study by Hung et al., underscores the necessity of plaque-enriched lncRNA in regulating phagocytosis of monocyte-derived macrophages (66). Previous studies from our laboratory have demonstrated the involvement of GAS5, RN7SK and MALAT1 in regulating phagocytosis and, antigen uptake and processing in both M1 and M2Mφ (41, 60). In this study, we demonstrated that knockdown of lncRNA LINC01010 attenuated both phagocytosis of E. coli as well us ovalbumin processing both in M1 as well as M2 macrophages. Our results demonstrate regulatory function of a myeloid-cell specific lncRNAs LINC01010 in shaping innate immune responses. Cytokine secretion from macrophages, after phagocytosis and antigen presentation, provides signal 3 during the course of T cells activation (67). Both M1 and M2 macrophages in this study exhibited significantly reduced secretion of pro-inflammatory cytokines: IL-1β, IL-6, IL-8 and TNFα in LINC01010 knockdown Mφ. This observation also corroborates with our results above demonstrating diminished phagocytosis of E. coli particles. Earlier reports also support the effect of lncRNAs expression on the gene expression/secretion of pro-inflammatory cytokines. In a study, lncRNA GAS5 overexpression induced M1Mφ phenotype was associated with the induction of nitric oxide synthase (iNOS), IL-1β, and TNF-α (68). Further, lncRNA GAS5 also reported to induce proinflammatory cytokines (IL-6, IL-1β, and TNF-α) in ox-LDL-induced THP-1 macrophages (68). Also, primary macrophages and BMDMs treated with LPS reported to upregulate lncRNA HOTAIR, which mechanistically favors nuclear translocation of NF-κB to produce pro-inflammatory cytokine genes (69). Other study shows that treatment of THP1 cells by IFN-γ induces the lncRNA FENDRR and favors M1 macrophage polarization associated with significantly increased secretion of cytokines including IL-1β, TNF-α, and CXCL10. Dramatic increase in STAT1 phosphorylation during this course turned out be one of plausible mechanism towards pro-inflammatory cytokine induction (70). However, the exact mechanism by which lncRNAs LINC01010 regulate the pro-inflammatory cytokine secretion needs further investigation.

In conclusion, we have characterized novel biological function of macrophage-enriched lncRNA LINC01010 in macrophage differentiation and polarization. Knockdown of LINC01010 impairs key innate immune functions of macrophages viz antigen uptake, phagocytosis, and cytokine secretion. Mechanistically, LINC01010 regulate cytokine signaling by suppressing multiple genes involved in NFκB pathway suggesting its role in potentiating macrophage innate immune response.

Data and materials availability:

All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials.