ORIGINAL RESEARCH published: 17 January 2022 doi: 10.3389/fimmu.2021.816930 Genomic Multiple Sclerosis Risk Variants Modulate the Expression of the ANKRD55–IL6ST Gene Region in Immature Dendritic Cells Jorge Mena 1, Iraide Alloza 1, Raquel Tulloch Navarro 1, Ane Aldekoa 1, Javier Dı́ez Garcı́a 2, Ane Villanueva Etxebarria 3,4,5, Cecilia Lindskog 6, Alfredo Antigüedad 7, Sabas Boyero 7, Marı́a del Mar Mendibe-Bilbao 7, Amaya Álvarez de Arcaya 8, José Luis Sánchez Menoyo 9, Edited by: Luciana Midaglia 10, Noelia Villarrubia 11, Sunny Malhotra 10, Xavier Montalban 10, Zsolt Illes, Luisa Marı́a Villar 11, Manuel Comabella 10 and Koen Vandenbroeck 1,12,13* University of Southern Denmark, 1 Inflammation & Biomarkers Group, Biocruces-Bizkaia Health Research Institute, Barakaldo, Spain, 2 Microscopy Facility, Denmark Biocruces-Bizkaia Health Research Institute, Barakaldo, Spain, 3 Kronikgune Institute for Health Services Research, Barakaldo, Reviewed by: Spain, 4 Health Service Research Network on Chronic Diseases Red de Investigación en Servicios de Salud en Enfermedades Tone Berge, Crónicas (REDISSEC), Bizkaia, Spain, 5 Osakidetza-Basque Health Service, Research Unit, Galdakao University Hospital, Oslo and Akershus University College Galdakao, Spain, 6 Department of Immunology, Genetics and Pathology, Rudbeck Laboratory, Uppsala University, Uppsala, of Applied Sciences, Norway Sweden, 7 Department of Neurology, Cruces University Hospital, Osakidetza-Basque Health Service, Biocruces-Bizkaia Health Magdalena Zoledziewska, Research Institute, Barakaldo, Spain, 8 Department of Neurology, Txagorritxu University Hospital, Osakidetza-Basque Health National Research Council (CNR), Italy Service, Bioaraba Health Research Institute, Vitoria-Gasteiz, Spain, 9 Department of Neurology, Galdakao-Usansolo University *Correspondence: Hospital, Osakidetza-Basque Health Service, Biocruces-Bizkaia Health Research Institute, Galdakao, Spain, 10 Servei de Koen Vandenbroeck Neurologia-Neuroimmunologia, Centre d’Esclerosi Múltiple de Catalunya (Cemcat), Institut de Recerca Vall d’Hebron (VHIR), k.vandenbroeck@ikerbasque.org Hospital Universitari Vall d’Hebron, Universitat Autònoma de Barcelona, Barcelona, Spain, 11 Department of Immunology, Hospital Universitario Ramón y Cajal, Instituto Ramón y Cajal de Investigación Sanitaria (IRYCIS), Red Española de Esclerosis Múltiple (REEM), Madrid, Spain, 12 Department of Biochemistry and Molecular Biology, Universidad del Paı´s Vasco (UPV/EHU), Specialty section: Barrio Sarriena, Leioa, Spain, 13 Ikerbasque, Basque Foundation for Science, Bilbao, Spain This article was submitted to Multiple Sclerosis and Neuroimmunology, Intronic single-nucleotide polymorphisms (SNPs) in the ANKRD55 gene are associated a section of the journal Frontiers in Immunology with the risk for multiple sclerosis (MS) and rheumatoid arthritis by genome-wide Received: 17 November 2021 association studies (GWAS). The risk alleles have been linked to higher expression Accepted: 22 December 2021 levels of ANKRD55 and the neighboring IL6ST (gp130) gene in CD4+ T lymphocytes of Published: 17 January 2022 healthy controls. The biological function of ANKRD55, its role in the immune system, and Citation: Mena J, Alloza I, cellular sources of expression other than lymphocytes remain uncharacterized. Here, we Tulloch Navarro R, Aldekoa A, show that monocytes gain capacity to express ANKRD55 during differentiation in Díez García J, Villanueva Etxebarria A, immature monocyte-derived dendritic cells (moDCs) in the presence of interleukin (IL)-4/ Lindskog C, Antigüedad A, Boyero S, Mendibe-Bilbao MdM, granulocyte-macrophage colony-stimulating factor (GM-CSF). ANKRD55 expression Álvarez de Arcaya A, levels are further enhanced by retinoic acid agonist AM580 but downregulated Sánchez Menoyo JL, Midaglia L, Villarrubia N, Malhotra S, Montalban X, following maturation with interferon (IFN)-g and lipopolysaccharide (LPS). ANKRD55 Villar LM, Comabella M and was detected in the nucleus of moDC in nuclear speckles. We also analyzed the Vandenbroeck K (2022) Genomic adjacent IL6ST, IL31RA, and SLC38A9 genes. Of note, in healthy controls, MS risk Multiple Sclerosis Risk Variants Modulate the Expression of the SNP genotype influenced ANKRD55 and IL6ST expression in immature moDC in opposite ANKRD55–IL6ST Gene Region in directions to that in CD4+ T cells. This effect was stronger for a partially correlated SNP, Immature Dendritic Cells. Front. Immunol. 12:816930. rs13186299, that is located, similar to the main MS risk SNPs, in an ANKRD55 intron. doi: 10.3389/fimmu.2021.816930 Upon analysis in MS patients, the main GWAS MS risk SNP rs7731626 was associated Frontiers in Immunology | www.frontiersin.org 1 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC with ANKRD55 expression levels in CD4+ T cells. MoDC-specific ANKRD55 and IL6ST mRNA levels showed significant differences according to the clinical form of the disease, but, in contrast to healthy controls, were not influenced by genotype. We also measured serum sgp130 levels, which were found to be higher in homozygotes of the protective allele of rs7731626. Our study characterizes ANKRD55 expression in moDC and indicates monocyte-to-dendritic cell (Mo–DC) differentiation as a process potentially influenced by MS risk SNPs. Keywords: ANKRD55, IL6ST, sgp130, multiple sclerosis, autoimmune INTRODUCTION highly correlated among them, i.e., rs6859219 (Chr 5: 56142753), rs10065637, and rs71624119 (D’ > 0.98; R2 > 0.84 in European The ANKRD55 gene, located on chromosome 5q11.2, contains population), but more modestly to rs7731626 (R2 < 0.53), were single-nucleotide polymorphisms (SNPs) that emerged from found to be associated with MS (8–10), RA (11–14), Crohn’s genome-wide association studies (GWAS) as risk variants for disease, and inflammatory bowel disease (15, 16). The intronic disorders with predominantly autoimmune or chronic location of all these SNPs points to allelic variation in regulatory inflammatory etiopathogenesis. Among these SNPs, rs7731626 elements as potential mechanism for transcriptional modulation (Chr 5: 56148856) is the variant displaying the strongest of ANKRD55 and/or adjacent genes. Of these, several are association with both rheumatoid arthritis [RA; P < 2 × 10-22; immunologically relevant (IL6ST and the related receptor ref (1, 2)] and multiple sclerosis [MS; P = 4 × 10-15; ref (3)]. This subunit IL31RA). variant is also associated with pediatric autoimmune diseases As ANKRD55’s proximate neighbor, IL6ST is especially (P = 1 × 10-10; ref (4). Other SNPs in ANKRD55, located like striking due to its well-documented role in inflammation and rs7731626 in intronic areas next to exon 6 (Figure 1), have, as autoimmune disease (Figure 1) (17). The IL6ST gene codes for well, been associated with the risk for autoimmune disease with the interleukin (IL)-6 receptor b unit, also known as glycoprotein genome-wide significance levels. Members of a trio of SNPs 130 (gp130), which is part of the functional receptor of the FIGURE 1 | Autoimmune risk SNPs located in the ANKRD55 gene area. Plot represents individual LD r2 values of SNPs (ocher circles) located in the delineated genomic interval with the main autoimmune risk SNP rs7731626 (blue circle). SNPs independently associated with autoimmune diseases with genome-wide significance are indicated (rs71624119, rs10065637, rs6859219). Genes located in this interval are represented in the lower part of the figure. Combined recombination rate is provided (black lines). SNPs rs28722705 and rs13186299 have been associated robustly in GWAS with monocyte count (5, 6) and monocyte percentage of white cells (7), respectively. Image information was produced by means of ldlink-nci-nih-gov. Frontiers in Immunology | www.frontiersin.org 2 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC pleiotropic cytokine IL-6 and is ubiquitously expressed. The the MS and RA lead risk variant rs7731626 as the SNP with the gp130 chain forms also part of the receptor of other cytokines most significant cis-eQTL effect for the expression of ANKRD55 1 with demonstrated roles in autoimmune disease including IL-27 in spleen and whole blood (GTEx Analysis Release V8) . + and IL-35 (17). Biological deregulation of IL-6 is functionally However, rs7731626 colocalized with CD4 T-cell specific associated with chronic inflammation and autoimmunity (17). eQTLs for expression of both ANKRD55 and IL6ST. This IL-6 can signal through two different pathways, i.e., classical suggests that this locus has a pleiotropic mechanistic effect on signaling or trans-signaling, which mediate anti-inflammatory or transcription of the two genes in this specific cell subset (34). pro-inflammatory response, respectively (17, 18). Distinct Risk for autoimmune disease conferred by rs7731626 is soluble forms of gp130 (sgp130) generated through alternative correlated with increased mRNA expression of ANKRD55 and splicing or alternative polyadenylation, act as natural specific IL6ST and with reduced cis CpG methylation in CD4+ T cells (34, inhibitors of IL-6 trans-signaling; these include sgp130 (19), 35). Using naive CD4+ T-cell capture Hi-C data, a significant sgp130-RAPS (20), and sgp130-E10 (21). In MS, IL-6 chromatin interaction between the intronic area around blockage is thought to limit immune-mediated tissue damage. rs7731626 and the promoter of IL6ST was observed, The IL-6 receptor blocking humanized monoclonal antibody potentially clarifying this coregulation in 3D space (31, 34, 36). (mAb) tocilizumab inhibits experimental autoimmune On the whole, current evidence explains genetic risk for MS and encephalomyelitis (EAE) (22) and is effective in otherwise RA conferred by ANKRD55 SNPs in terms of a cis CpG refractory neuromyelitis optica spectrum disorder (NMOSD; methylation-dependent mechanism to increase the expression ref (23, 24). Satralizumab, a second-generation anti-IL6R mAb, of two genes, ANKRD55 and IL6ST, in CD4+ T lymphocytic cells. has been approved for the treatment of NMOSD (25). Here, we show that CD14+ monocytes, which express only The precise biological function of ANKRD55, on the other negligible levels of ANKRD55 mRNA (30), gain capacity to hand, is not yet known. Recent interactome and knockdown produce higher levels of ANKRD55 during differentiation into studies have provided some initial though heterogeneous immature monocyte-derived dendritic cells (moDCs) in the indications. On the basis of integrated mass spectrometry presence of IL-4 and granulocyte-macrophage colony- protein interaction datasets, ANKRD55 was first identified as a stimulating factor (GM-CSF). ANKRD55 levels are further novel member of the intraflagellar transport machinery (26). enhanced by a retinoic acid agonist but are suppressed by Correspondingly, morpholino knockdown of the gene appeared interferon (IFN)-g/lipopolysaccharide (LPS) treatment. We to disrupt ciliogenesis (26). In HEK293 cells, recombinant analyzed also neighboring genes including IL6ST, IL31RA, and ANKRD55 interacts in the cytosol preferentially with 14-3-3 SLC38A9. We demonstrate that in immature moDC, ANKRD55 isoforms and ATP- or nucleotide-binding proteins (27). It and IL6ST expression appears to be modulated by MS risk SNPs appeared also capable of entering the nucleus of HEK293 cells with an opposite allelic effect compared to that observed in CD4+ where it is found preferentially associated with RNA-binding T lymphocytes. proteins and proteins involved in sumoylation (27). In preadipocytes, silencing of ANKRD55 by the lentiviral clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) system MATERIALS AND METHODS increased both their proliferation rate and lipolysis (28). A direct functional relevance of ANKRD55 to autoimmune Patients and Controls processes has not been demonstrated. Buffy coats and serum from healthy donors and MS patients’ ANKRD55 was first reported as highly expressed in CD4+ whole blood were collected upon written informed consent in effector memory cells (29). Peripheral blood CD4 + T accordance with the local ethics committees from Bilbao [Comité lymphocytes, but not CD8+, CD19+, CD56+, or CD14+ subsets, É tico de Investigació n Clı́nica de Euskadi (CEIC-E)], Barcelona express high levels of ANKRD55 mRNA, and the risk allele of (Comité É tico de Investigació n Clı́nica Hospital Universitari Vall rs6859219 is associated with higher levels of ANKRD55 mRNA D’Hebron), and Madrid (Comité de É tica de la Investigació n del in the CD4+ subset (30). This variant is associated with DNA Hospital Ramó n y Cajal) correspondingly. All patients were methylation status at CpG sites mapping to ANKRD55-IL6ST under no treatment or placebo at the time of sample collection. intergenic and ANKRD55 intronic areas. Increased methylation A total of 128 biological samples of MS patients from 3 different at these sites seen with the protective allele correlates with Spanish cohorts (IIS Biocruces Bizkaia MS Cohort, Barakaldo; decreased expression of both ANKRD55 and IL6ST mRNA in Hospital Universitari Vall d’Hebron MS Cohort, Barcelona; and CD4+ T cells (31). The highly correlated rs71624119 variant was Hospital Universitario Ramó n y Cajal MS Cohort, Madrid) were confirmed to colocalize to a cis-expression quantitative trait locus included in the study [54 primary progressive (PP), 70 relapsing- (eQTL) in CD4+ T cells that significantly affected the expression remitting (RR), and 4 secondary progressive (SP)] in addition to of ANKRD55. Evidence for shared effect of this SNP with an 103 buffy coats from healthy control samples provided by the eQTL for IL6ST was weaker but still passed the false discovery Biobanco Vasco. Patients’ demographic and clinical data are rate (FDR) threshold (32). In a study in peripheral blood summarized in Supplementary Table S1. samples, both rs10065637 and rs6589219 colocalized with eQTLs for the expression of ANKRD55 but not of IL6ST (33). Similarly, the Genotype-Tissue Expression (GTEx) Project lists 1 https://www.gtexportal.org/home/gene/ANKRD55 Frontiers in Immunology | www.frontiersin.org 3 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC Peripheral Blood Mononuclear Cell Monocyte-Derived Dendritic Cell Subpopulations Differentiation and Maturation Peripheral blood mononuclear cells (PBMCs) were isolated from CD14+ monocytes were isolated from fresh PBMCs by positive buffy coats using a Ficoll®-Paque density gradient (Cytiva, Ref. selection using CD14 MACS MicroBeads (Miltenyi Biotec, Ref. 17-1440-03) as indicated by the manufacturer. Fresh PBMCs 130-050-201) and cultured at a density of 106 cells/ml in moDC were used for all experiments on monocytes and moDC. differentiation medium (containing IL-4/GM-CSF; Miltenyi Comparative analysis of gene expression by qPCR was Biotec, Ref. 130-094-812) for 6 days, with fresh medium added performed in CD4+, CD8+, CD14+, CD19+, and CD56+ cells on day 3. Maturation of moDCs was induced using combinations purified from PBMC that had been stored in liquid nitrogen. of different agents, including LPS (0.2–2 µg/ml, as indicated) CD4+ T-helper lymphocytes, CD8+ T cytotoxic lymphocytes, from Salmonella (Sigma-Aldrich, Ref. L6143), 0.5 pg/ml IFN-g CD19+ B lymphocytes, CD14+ monocytes, and CD56+ natural (Peprotech, Ref. AF-300-02), 1 mg/ml polyinosinic-polycytidylic killer (NK) cells were separated by positive selection using CD4 acid (PolyI:C; Sigma-Aldrich, Ref. P1530), and 1 mg/ml (Ref. 130-045-101), CD8 (Ref. 130-045-201), CD19 (Ref. 130- oligodeoxyribonucleotide CpG (Miltenyi Biotec, Ref. 130-100- 050-301), CD14 (Ref. 130-050-201), and CD56 (Ref. 130-050- 243). To induce tolerogenicity, DCs were treated during the 401) human MicroBeads (all from Miltenyi Biotec) differentiation process with 100 nM retinoic acid receptor-a correspondingly, following the manufacturer’s instructions, as agonist AM580 (Sigma-Aldrich, Ref. A8843), 10 nM vitamin D3 described before (30). (STEMCELL Technologies, Ref. 72412), 1 mM prostaglandin E2 (PGE2) (Sigma-Aldrich, Ref. P0409), 20 ng/ml IL-10 (Peprotech, Isolation of Myeloid Dendritic Cells and Ref. AF-200-10), 1 mM dexamethasone (Sigma-Aldrich, Ref. Plasmacytoid Dendritic Cells D4902), and 10 ng/ml rapamycin (STEMCELL Technologies, Human myeloid DCs (mDCs) and plasmacytoid DCs (pDCs) Ref. 73362), and all were added on day 0, and again on day 3, of were isolated from healthy donors’ fresh blood. PBMCs were the moDC differentiation period. For evaluation of maturation, obtained from 30 ml of blood by Ficoll®-Paque density gradient cells were analyzed with Mo–DC Differentiation Inspector using SepMate™-50 tubes (STEMCELL Technologies, Ref. antibody cocktail [Miltenyi Biotec, Ref. 130-093-567; anti- 85450). CD14+ monocytes were depleted from PBMCs using human CD14-FITC antibody (clone Tük4, isotype: mouse CD14 human MicroBeads (Miltenyi Biotec, Ref. 130-050-201). IgG2a), anti-human CD83-APC (clone: HB15, isotype: mouse Remaining PBMCs were separated into two equal parts and IgG1), and anti-human CD209-PE (clone DCN-47.5.4, isotype: processed with EasySep™ Human Myeloid DC Enrichment Kit mouse IgG1)] or isotype control cocktail following (STEMCELL Technologies, Ref. 19061) for mDC isolation or manufacturer’s instructions and analyzed on a MACSQuant® EasySep ™ Human Plasmacytoid DC Enrichment Kit flow cytometer (Miltenyi Biotec). (STEMCELL Technologies, Ref. 19062) for pDC isolation following the manufacturer’s protocol. To verify the purity of DNA Extraction, RNA Isolation, and each subpopulation, cells were analyzed by flow cytometry cDNA Synthesis (MACSQuant® flow cytometer, Miltenyi Biotec) using anti- Genomic DNA (gDNA) was obtained from PBMCs with human CD1c (BCDA-1)-APC antibody (Miltenyi Biotec, Ref. PureLink™ Genomic DNA purification kit (Invitrogen, Ref. 130-110-595) for mDC and CD303 (BDCA-2)-APC antibody K182001) and used for genotyping analyses. Total RNA was (Miltenyi Biotec, Ref. 130-114-177) for pDC, as shown in extracted from cells using TRI Reagent® (Sigma-Aldrich, Ref. Supplementary Figure S1. Isotype control human IgG1-PE T9424), NucleoSpin RNA Kit (Macherey-Nagel, Ref. 740955), or antibody (Miltenyi Biotec, Ref. 130-113-428) was used as RNeasy Mini Kit (Qiagen, Ref. 74104) following each negative control for CD14 and isotype control human IgG1- manufacturer’s instructions, treated with DNAse (Sigma- APC as negative control for CD1c and CD303 (Miltenyi Biotec, Aldrich Ref. AMPD1 or Qiagen Ref. 79254) and measured by Ref. 130-113-434). Nanodrop® 2000 spectrophotometer (Thermo Fisher Scientific) to determine the concentration. cDNA was synthesized from 100 Sorting of Classical, Intermediate, and to 600 ng of total RNA using High-Capacity cDNA Reverse Non-Classical Monocytes Transcription Kit (Applied Biosystems, Ref. 4368814). Monocytes were obtained from human PBMCs with Pan Monocyte Isolation kit (Miltenyi Biotec, Ref. 130-096-537) Primer Design, qPCR, Drop Digital following the manufacturer’s protocol, labeled with anti- PCR, and Genotyping human CD14-PE antibody (Miltenyi Biotec, Ref. 130-110- qPCR was performed using 5–30 ng cDNA, specific primer pairs 519) and CD16-FITC (Miltenyi Biotec, Ref. 130-113-392), (Supplementary Table S2) and Fast SYBR® Green Master Mix and sorted in a FACSJazz™ cell sorter (BD Biosciences). (Applied Biosystems, Ref. 4385612) or SsoAdvanced™ Universal Sorter strategy consisted of a size-complexity selection of SYBR® Green Supermix (Bio-Rad, Ref. 1725272), following monocytes followed by discrimination of three monocyte manufacturer’s instructions, in a 7300, 7500 Fast or 7900HT subpopulations: classical (CD14 hi /CD16 - ), intermediate Fast Real-Time PCR System (Applied Biosystems). IL6ST (CD14 hi /CD16 + ), and non-classical (CD14 + /CD16 + ), as primers are best coverage primers that co-amplify the main described in a previous article (37). splice variants of IL6ST by annealing to shared transcribed Frontiers in Immunology | www.frontiersin.org 4 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC sequences (isoforms 206, 201, 205, 209, 204, 212, 203). ACTB ROUT test and removed, and Shapiro–Wilk test was and/or GAPDH was used as reference genes for normalization of performed to determine if the variables were normally mRNA expression level, and data were analyzed applying the 2-DCt distributed. Mann–Whitney test, Wilcoxon signed-rank test, method. Samples were examined in triplicate for each condition, unpaired t-test, or paired t-test were performed to analyze the and no-template controls were included. The threshold was set statistical difference between two groups. Kruskal–Wallis test, within the linear phase of the logarithmic amplification plot. In Friedman test (followed by Dunn’s multiple comparison test), or specific experiments as stated in the text, absolute gene transcript ANOVA test were used to compare differences between three or quantification was performed by droplet digital PCR (ddPCR) from more groups. Correlation analyses were determined using 350 pg or 30 ng of cDNA using ddPCR Supermix for Probes (No Pearson correlation coefficient. Values of p ≤ 0.05 were dUTP) (Bio-Rad, Ref. 1863023) and specific primer pairs considered significant (*), p ≤ 0.01 very significant (**), p ≤ (Supplementary Table S2) in a QX200 Droplet Digital PCR 0.001 highly significant (***), and p ≤ 0.0001 extremely System (Bio-Rad) according to manufacturer’s protocol. gDNA significant (****). For analysis of group-by-time interaction of Samples were genotyped for SNPs rs6859219, rs7731626, and the 2-Dct values of IL6ST, IL31RA, and SLC38A9 gene expression rs13186299 using commercial TaqMan® probes (Thermo Fisher with those of ANKRD55, longitudinal data analysis was Scientific, Ref. 4351379) and TaqPath™ ProAmp™ master mix performed in moDC samples for four treatment conditions (Applied Biosystems, Ref. A30865) in a 7500 Fast Real-Time PCR (immature control, immature AM580-treated, IFN-g/LPS- System (Applied Biosystems) following manufacturer’s instructions. matured control, and IFN-g/LPS-matured AM580-treated) using generalized linear mixed univariable models. Random Immunofluorescence intercepts and unstructured variance-covariance matrix were CD14+ monocytes were plated on coverslips coated with poly-D- used. Quantification of immunofluorescent images of lysine (Sigma-Aldrich, Ref. P0899) and differentiated for 6 days ANKRD55 in the nucleus of moDC was performed as into moDC as mentioned above. Cells were fixed and described before (39). permeabilized with ice-cold methanol for 5 min, blocked with 1% bovine serum albumin (BSA; Sigma-Aldrich, Ref. A9418) in phosphate buffered saline (PBS) for 30 min, stained with defined primary antibodies for 1 h at room temperature (RT), incubated RESULTS with corresponding fluorescent secondary antibodies for 1 h at RT protected from light, stained with 1 mg/ml 4′,6-diamidino-2- Expression of ANKRD55, IL6ST, IL31RA, phenylindole (DAPI) (Sigma-Aldrich, Ref. D9542) for 5 min at DDX4, SLC38A9, and Soluble gp130 RT, and mounted on glass slides using Fluoromount-G Isoforms in Peripheral Blood Mononuclear (Invitrogen, Ref. 00-4958-02). Between each step, coverslips Cell Subpopulations were washed 3 times with PBS. Primary antibodies used were MS risk alleles are associated with enhanced expression of anti-ANKRD55 rabbit polyclonal (Human Protein Atlas, Ref. ANKRD55 in CD4+ T cells (30, 32), but genotypic effects on HPA 061649; 1:250), anti-ALYREF mouse monoclonal (Thermo the expression of the flanking genes in PBMC subsets have not Fisher Scientific, Ref. MA1-26754; 1:500), anti-HNRNPC mouse been reported. To identify cellular sources of transcripts in monoclonal (Sigma-Aldrich, Ref. AMAB91010; 1:100), and anti- PBMC, five subpopulations were isolated from healthy donors: human CD209. Secondary antibodies used were AlexaFluor 488- CD4+ T, CD8+ T, CD19+ B lymphocytes, CD14+ monocytes, and conjugated goat anti-mouse IgG H&L (Abcam, Cat. No. CD56+ NK cells. qPCR analysis was performed to assess mRNA ab150113; 1:1,000) and AlexaFluor 594-conjugated goat anti- levels of ANKRD55 and neighboring genes including the rabbit [F(ab’)2] (Thermo Fisher, Ref. A-11072; 1:500). Images isoforms sgp130, sgp130-RAPS, and sgp130-E10. Similar to were maximal intensity projections of serial Z-optical sections ANKRD55, IL6ST was most highly expressed in CD4+ T-helper acquired using a Zeiss LSM 880 Airyscan microscope with a ×63 lymphocytes, but in contrast to it, it was also detectable in the immersion objective (NA: 1.4), and pictures were analyzed with other four subpopulations at a lower level of expression custom-made macros in Fiji software (38). (Supplementary Figure S2). IL31RA was uniquely detected in CD14+ monocytes, and DDX4 was undetectable in any of the ELISA PBMC subpopulations. SLC38A9 was highly expressed in B Levels of sgp130 were measured in serum samples of MS patients ly mphocy tes but also detec table in the re main ing using commercial ELISA (R&D, Ref. DY228) following subpopulations. The three soluble gp130 isoforms were highly manufacturer’s instructions. Samples were diluted (1:100 and expressed in CD4+ T helper lymphocytes but less so in the other 1:200) and measured in duplicate. Absorbance was read in a subpopulations studied, similar to the patterns seen for IL6ST Varioskan Flash (Thermo Fisher Scientific), and data were (Supplementary Figure S2). The influence of homozygosity for analyzed using GraphPad Prism v.7 software. MS risk SNPs rs6859219 and rs7731626 and of the correlated SNP rs13186299 (explained below) on the expression in healthy Statistical Analysis control (HC) CD4+ T cells of ANKRD55, IL6ST, and the three Data were analyzed with GraphPad Prism v.7 software and are sgp130 isoforms is shown in Supplementary Figure S3. presented as the mean ± standard error of the mean (SEM), Homozygotes for the risk alleles (CC, GG, and GG, unless otherwise indicated. Outliers were identified applying respectively) expressed significantly higher levels of ANKRD55 Frontiers in Immunology | www.frontiersin.org 5 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC than those for the protective alleles, and directionally similar but coding isoforms ANKRD55-201 (ENSEMBL nomenclature; full- weaker trends were seen for all other gene transcripts. In the length isoform; protein encoded Mr = 68.4 kDa) and ANKRD55- non-CD4+ subsets, the rs13186299 G risk allele was significantly 202 (shorter protein-coding isoform; Mr = 36.9 kDa) and the associated with lower IL6ST expression in CD56+ cells (p < 0.01) non-coding processed transcript ANKRD55-204, which were (Supplementary Figure S4). originally designated isoforms 001, 005, and 007, respectively (30). To ascertain more robustly the effect of AM580 on ANKRD55 Is Expressed in Immature ANKRD55 expression in moDC, we performed qPCR using Monocyte-Derived Dendritic Cells and ANKRD55 isoform-specific primer pairs each amplifying a Induced by Retinoic Acid Receptor unique ANKRD55 splice variant, as described before (30). This Alpha Agonist AM580 analysis showed that immature moDCs express the three splice We decided to study ANKRD55 gene expression in monocyte- variants, and that AM580 significantly enhanced the expression derived DC prompted by some earlier findings. Martin et al. (40) levels of all three. Fold upregulation of transcripts 201, 202, and demonstrated, and we confirmed (37), high expression of the 204 by AM580 amounted to, respectively, 2.8 (p < 0.01 compared IL22RA2 gene in immature moDC but not in monocytes or to untreated), 2.2 (p < 0.05), and 2.9-fold (p < 0.01) (Figure 2J). mature moDC. IL22RA2 is located at an MS risk locus (3) and is Circulating monocytes can be distinguished based on relative poorly or not expressed in PBMC subsets. We hypothesized that its expression levels of CD14 and CD16 surface proteins, namely, unique expression pattern perhaps indicates a role for immature classical (CD14hi/CD16-), intermediate (CD14hi/CD16+), and non- moDC in MS pathogenesis, and that other MS risk genes may be classical (CD14+/CD16+) subsets. These are thought to contribute expressed during the monocyte-to-DC (Mo–DC) differentiation differentially to immune surveillance of the central nervous system stage. Thus, purified CD14+ monocytes from healthy donors were (CNS) and to inflammatory responses in MS (43, 44). We assessed cultivated in Mo–DC differentiation medium containing IL-4 and ANKRD55 isoforms and IL6ST and IL31RA gene expression in GM-CSF for 6 days to induce formation of immature moDCs. these monocyte subsets separated by fluorescence-activated cell Proper differentiation was assessed by CD209 (DC-SIGN) sorting (FACS) as well as in the corresponding immature DCs immunofluorescence (IF) microscopy and was associated with obtained by cultivation of each subset in Mo–DC differentiation intense CD209 staining observed in >95% of cells. Cells were medium for 6 days (Figure 2K). Very low but detectable levels of harvested for gene expression analysis by qPCR (Figures 2A–D) ANKRD55 isoforms 201 and 204, though not of 202, were detected on days 0 (monocyte stage), 1, 2, 4, and 6 throughout this period. in all three monocyte subsets. The highest expression level was The ANKRD55 and IL6ST, IL31RA, and SLC38A9 genes were found in the classical subset, in line with Schmiedel et al. (45) who included in the analysis to assess differences in patterns of gene reported 0.4 transcripts per million (TPM) of ANKRD55 in classical expression. ANKRD55 gene expression started to rise from day 2 vs. 0.1 TPM in non-classical monocytes (45). Differentiation of the onward (p < 0.01 for comparison of levels on day 6 to those on day monocyte subsets into immature DC strongly enhanced expression 0). In contrast, expression levels of IL6ST did not vary significantly, levels of all three ANKRD55 transcripts with highest levels for while those of IL31RA decreased (p < 0.01 for comparison of levels ANKRD55-201 (11.5-fold upregulation) and ANKRD55-204 (15.2- at day 4 vs. 0). SLC38A9 expression levels peaked at day 1 (p < 0.01). fold upregulation) found in the classical monocyte-derived DC Since expression of DDX4 was not detected, this gene was not subset. IL6ST was expressed at much higher levels than further studied. The effects of a series of tolerogenic compounds ANKRD55 in the monocyte subsets but was less strongly known to modulate gene expression in moDCs (41) including increased in the moDC (1.4-fold upregulation in classical AM580, an analog of retinoic acid that acts as a selective retinoic monocyte DC subset), while IL31RA was downregulated in each acid receptor alpha (RARa) agonist, vitamin D3, IL-10, PGE2, moDC subset. dexamethasone, and rapamycin were tested by addition to Mo–DC medium over the course of the differentiation period. Of these, ANKRD55 Expression Is Downregulated in AM580 appeared to enhance mRNA levels both of ANKRD55 (3.5- IFN-g/LPS-Matured Monocyte-Derived fold) and of IL31RA (379-fold), while the other compounds either Dendritic Cells inhibited the expression or had more modest effects on mRNA We next assessed the effect of moDC maturation on the expression levels of any gene (Figures 2E–H). A qPCR time course analysis of of ANKRD55 and neighboring genes by treating the immature cells the effect of AM580 on each of these genes compared to untreated on day 6 for 6 h with Toll-like receptor (TLR)9 agonist CpG, or cells was performed (gray curves in Figures 2A–D). In the presence combinations of (TLR)4 agonist LPS and TLR3 agonist poly(I:C), or of AM580, moDC produced slightly higher levels of ANKRD55 and of IFN-g and LPS (Figure 3). Percentage of mature cells assessed by IL6ST. Since retinoic acid is essential for GM-CSF-induced flow cytometry of CD83 and CD209 (DC-SIGN) marker expression ALDH1A2 expression in DC (42), bioactivity of AM580 was measured over 4 independent experiments was (average ± STDEV) ascertained through upregulation of ALDH1A2 (p < 0.01 79.0% ± 8.9% with LPS + IFN-g, 62.8% ± 12.4% with LPS + poly(I: compared to untreated; Figure 2I). RARA and RARG were C), and 23.75% ± 6.4% with CpG, while spontaneous maturation expressed in the moDC, with the former significantly affected (control) amounted to 9.25% ± 5.1%. Expression levels of the three (p < 0.01) only following treatment with IFN-g/LPS. ANKRD55 isoforms were measured and found to be significantly Previously, we had found that CD4+ T lymphocytes coexpress downregulated in moDC matured with IFN-g/LPS (p < 0.05 vs. three distinct ANKRD55 mRNA splice variants, i.e., protein- control; Figures 3A, B), and similar weaker trends were seen for the Frontiers in Immunology | www.frontiersin.org 6 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC A B C D E F G H I J K FIGURE 2 | Expression of ANKRD55 and neighboring genes in immature moDC. (A–D) Monocytes were cultivated for 6 days in IL-4/GM-CSF medium for differentiation into immature moDC in the absence (colored curves) or presence (gray curves) of AM-580, and expression levels of the indicated genes were measured by qPCR at the start of the cultivation (day 0) and on days 1, 2, 4, and 6. Mean ± SEM of 3 independent measurements; Friedman test (followed by Dunn’s multiple comparison test) for comparison of data points in each curve with day 0 and Wilcoxon test for comparison of both curves (Control vs. AM580). (E–H) Effect of tolerogenic compounds on gene expression in immature moDC by qPCR. Concentrations of compounds are provided in the Materials and Methods. Mean ± SEM of maximum 5 independent measurements; Friedman test (followed by Dunn’s multiple comparison test). (I) Analysis of ALDH1A2, RARa, and RARg expression by qPCR. Mean ± SEM; n = 6; Friedman test (followed by Dunn’s multiple comparison test). (J) Expression levels of three discrete ANKRD55 splice variants 201, 202, and 204 in immature moDC. Mean ± SEM; n = 10; Wilcoxon test. (K) Classical (Cla; CD14hi/CD16-), intermediate (Int; CD14hi/CD16+), and non- classical (NC; CD14+/CD16+) monocytes were isolated and separately cultivated for 6 days to differentiate into moDC. Gene expression levels were measured in the original monocyte subsets (M) and derived moDC (D). Mean ± SEM; n = 3; Mann–Whitney test for comparison between monocytes and moDC for each gene and subset. *p ≤ 0.05, **p ≤ 0.01. other two conditions. In fact, extent of suppression of ANKRD55 costaining with CD209, a membrane marker that is absent in expression appeared proportional to the degree of moDC monocytes but appears strongly in moDC (Figures 4A, B). Ratio maturation. A time course analysis was performed to assess nuclear ANKRD55/DAPI in moDC was significantly higher than ANKRD55 gene expression and that of its neighboring genes that in monocytes (p < 0.001). We also compared immature, following initiation of maturation of moDCs with IFN-g/LPS in mature, and AM580-treated moDC. In a quantitative analysis of the presence or absence of AM580. Maturation of moDC induced IF images, IFN-g/LPS treatment significantly reduced nuclear rapid downregulation of ANKRD55 and SLC38A9 (p < 0.05 at 3 h) ANKRD55 levels (p < 0.0001), while AM580 had no effect gene expression with the lowest levels recorded from 3 to 12 h (Figure 4C). Given a nuclear pattern of irregular spots, we following addition of LPS (Figure 3), and this effect was seen in verified whether ANKRD55 could be associated with nuclear either untreated or AM580-pretreated cells. IFN-g/LPS had little speckles, nuclear irregularly shaped dynamic structures (46). As effect on IL6ST gene expression but induced high IL31RA shown in Figures 4D, E, ANKRD55 colocalized consistently with expression levels in the untreated moDC (p < 0.01 at 12 h and ALYREF and HNRNPC, two confirmed constituents of nuclear p < 0.05 at 24 h after LPS addition for comparison with levels at speckles (46). day 5). We further performed absolute quantification by ddPCR of We used IF microscopy to detect intracellular location of transcript levels in natural circulating conventional (cDC) and ANKRD55. ANKRD55 was borderline visible in the nuclei of plasmacytoid (pDC) dendritic cells, CD14+, and CD4+ cells, and monocytes but stained clearly in the cytosol and in the nuclei in untreated and treated moDC (Figure 5). As measured in 350 pg of immature moDC. Differentiation of moDC was confirmed via cDNA, ANKRD55 transcript copy numbers per ml (average ± Frontiers in Immunology | www.frontiersin.org 7 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC A B C D E F FIGURE 3 | Effect of maturation of moDC on the expression of ANKRD55, IL6ST, IL31RA, and SLC38A9. (A) Comparison of the effect of maturation by CpG, poly(I: C)/LPS and IFN-g/LPS on expression levels of the main ANKRD55 splice variants by qPCR. Mean ± SEM; n = 3; Friedman test (followed by Dunn’s multiple comparison test). (B) Flow cytometry to assess the maturation percentage induced by the indicated stimuli based on the expression of CD209 (expressed in moDC) and CD83 (expressed in mature moDC only). Shown is a representative experiment out of 4 performed. (C–F) MoDCs cultivated for 5 days in MoDC medium were matured by addition of IFN-g and 24 h later LPS. Gene expression was analyzed by qPCR coinciding with time of IFN-g (day 5) and LPS (0 h) addition and 3, 6, 12, 24, and 48 h later. The experiment was also performed in cells differentiated in the presence of AM-580 (gray curves). Mean ± SEM of 3 independent measurements; Friedman test (followed by Dunn’s multiple comparison test) for comparison of data points in each curve with day 5 and Wilcoxon test for comparison of both curves (control vs. AM580). *p ≤ 0.05, **p ≤ 0.01. STDEV) in pDC (4.3 ± 2.8) and mDC (1.4 ± 1.5) were slightly showed significant group-by-time interactions with ANKRD55 in all higher than those in CD14+ monocytes (1 ± 1.15) (Figure 5A). In tested conditions. Group-by-time interactions, however, were not contrast, IL6ST, IL31RA, and SLC38A9 were expressed at much statistically significant when comparing ANKRD55 with IL6ST or higher levels in CD14+ cells than those in pDC or mDC (Figure 5E). with SLC38A9 in both control and AM580-administered groups of AM580 enhanced basal ANKRD55 transcript levels in immature the immature sample and with SLC39A9 in both control and moDC at day 6 from 88.8 ± 46.2 to 208.3 ± 63.8 copies/ml, and IFN- AM580-administered groups of the mature sample. Therefore, in g/LPS downregulated day-5 immature moDC ANKRD55 levels those cases, the similarity between the analyzed groups in terms of from 89.3 ± 93.3 to 19.5 ± 14.9 copies/ml, both measured in 30 ng the trend among the evaluated times cannot be rejected. Subsequent of cDNA (Figures 5C, D). In comparison, CD4+ T cells expressed genetic analysis was performed focusing on ANKRD55 and IL6ST in much higher levels of ANKRD55 (1,643 ± 1,741 copies/ml) also the immature untreated moDC. measured in 30 ng cDNA (Figure 5B). A longitudinal data analysis was performed in order to Multiple Sclerosis Risk Single-Nucleotide determine whether the curves representing ANKRD55 and IL6ST, Polymorphisms Modulate the Expression IL31RA, and SLC38A9 mRNA expression levels (Figures 2, 3) are of ANKRD55 and IL6ST in Immature statistically dissimilar within each treatment condition. The p-values Monocyte-Derived Dendritic Cells corresponding to the group-by-time interaction term’s significance To assess whether ANKRD55 MS risk genotypes regulate ANKRD55 in the adjusted generalized linear mixed univariable models are and IL6ST expression in immature moDCs, we performed qPCR in collected in Table 1. The results show that IL31RA gene expression day-6 immature moDCs generated from healthy subjects who were Frontiers in Immunology | www.frontiersin.org 8 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC A B C D E FIGURE 4 | Localization of ANKRD55 in (A) monocytes (1.024 × 1.024 pixels = 0.131 microns/pixel) and (B–E) moDC by immunofluorescence microscopy. Rightmost graph represents mean ± SEM in monocytes and immature moDC (n = 5 cellular ROIs/group; unpaired t-test). (B) Colocalization of ANKRD55 with CD209, a membrane marker specific for moDC (2.048 × 2.048 pixels = 0.065 microns/pixel). (C) Effect of AM580 and IFN-g/LPS treatment on nuclear ANKRD55 signal in immunofluorescence (1,808 × 1,808 pixels = 0,074 microns/pixel). The diagram on the right provides quantitative analysis of nuclear ANKRD55 immunofluorescence [mean ± SEM; n ≥ 59 cellular ROIs/condition; Kruskal–Wallis test (followed by Dunn’s multiple comparison test)]. (D) Colocalization of ANKRD55 with ALYREF (1.024 × 1.024 pixels = 0.131 microns/pixel). (E) Colocalization of ANKRD55 with HNRNPC (1.024 × 1.024 pixels = 0.131 microns/pixel). ***p ≤ 0.001, ****p ≤ 0.0001. Frontiers in Immunology | www.frontiersin.org 9 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC A B C D E F FIGURE 5 | Digital droplet PCR (ddPCR) quantification of gene transcript copy numbers in pDC, mDC, CD14+ monocytes, PBMC fraction minus monocytes (F-), and moDC. Analysis was performed in 350 pg cDNA (A, E, F) or 30 ng of cDNA (B–D). (C) ddPCR quantification of ANKRD55 transcript copy numbers in moDC treated with AM-580. (D) ANKRD55 copy numbers in moDC matured or not with IFN-g/LPS measured at 6 h after addition of stimulus. (F) Transcript levels of markers specific for mDC (CD1c) and pDC (IL3RA, CLEC4C). genotyped for MS risk SNPs rs6859219 and rs7731626. Compared p < 0.01, respectively) (Figures 6A, B). The mRNA levels of sgp130 to the risk alleles, the protective A alleles of both SNPs were and sgp130-RAPS in immature untreated moDC appeared, as well, associated with higher mRNA expression of ANKRD55 (p < 0.01 to be higher in carriers of the A alleles (Supplementary Figure S5). and p < 0.05, respectively) and of IL6ST (not significant trend and Notably, in CD4+ T lymphocytes, the opposite directional effect is TABLE 1 | p-Values corresponding to the group-by-time interaction term’s significance in the adjusted generalized linear mixed univariable models. Compared evolutions Immature Mature Control AM580 Control AM580 ANKRD55 vs. IL6ST 0.5772 0.2973 0.0016 0.0142 ANKRD55 vs. IL31RA 0.0001 0.0010 0.0304 0.0005 ANKRD55 vs. SLC38A9 0.7150 0.8674 0.4231 0.0972 Frontiers in Immunology | www.frontiersin.org 10 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC A B C D E F FIGURE 6 | (A–C) Modulation of ANKRD55 and IL6ST gene expression in immature moDC of healthy donors by cis genomic MS risk variants (mean ± SEM; n ≥ 28; Mann–Whitney test); (A) rs6859219, (B) rs7731626, and (C) the correlated ANKRD55-intronic SNP rs13186299. (D) Assessment of ANKRD55 and IL6ST expression levels according to three-SNP marker combinations (mean ± SEM; n ≥ 28; one-way ANOVA test). (E) Haplotype distribution in the European population of the three SNPs. Image information was produced by means of ldlink.nci.nih.gov. (F) Genomic variants rs13186299 (left) and rs7731626 (right) are the most significant single- tissue eQTLs for IL6ST and ANKRD55 in whole blood, respectively, based on information collected from the GTEx portal (gtexportal.org/home/). The risk genotype of each of both SNPs is GG, and the protective allele is C for rs13186299 and A for rs7731626. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. n.s., not significant. observed with the A alleles associated with lower levels of ANKRD55 risk and monocyte count traits. The Variant-to-Gene pipeline compared to the risk alleles (27, 28, 31; also, in Supplementary (genetics.opentargets.org) that integrates QTL data, Promoter Figure S3). The opposite effects in moDC could be due to linkage Capture Hi-C, in silico functional predictions, and variant disequilibrium (LD) with distinct genomic variants correlated to distance to gene prioritizes ANKRD55 and IL6ST as the genes specific phenotypic traits of myeloid or monocytic cells. We functionally implicated by all four variants. In the GTEx portal, performed a search for GWAS variants in LD with these SNPs rs13186299 is recorded as the most significant single-tissue eQTL and identified two candidate SNPs meeting these criteria, for IL6ST expression in whole blood (p = 4.4 × 10-11; normalized rs28722705 (Chr 5: 56158115) and its proxy rs13186299 (Chr 5: effect size 0.17) (Figure 6F). 56159818) (D’ = 1; R2 = 1), which emerged with genome-wide Reanalysis of the expression data in the sample cohort now significance from GWAS on monocyte count (p ≤ 5 × 10-27; 5, 6) genotyped for rs13186299 (Figures 6C, D) revealed a highly and monocyte percentage of white cells (p = 7 × 10-22; 7), significant association of mRNA levels of ANKRD55 and IL6ST in respectively. Both are intron variants located in ANKRD55 immature moDC with genotype (p < 0.001 and p < 0.0001, (Figure 1) that occur in partial LD with both ANKRD55 MS risk respectively), identifying this SNP as exerting a higher impact SNPs (D’ = 0.62–0.98; R2 = 0.22–0.26). Interestingly, in the PheWAS upon ANKRD55 and IL6ST expression in moDC than either MS database (that integrates studies by the FinnGen, UK Biobank, and risk SNP. Expression levels showed roughly similar variations GWAS Catalog consortia; genetics.opentargets.org), MS risk SNPs according to marker combination (exception made for rs7731626 and rs6859219 are reported to be associated with combination 3), with the rs13186299 GG genotype always monocyte count with genome-wide significance (p = 1.8 × 10-14 segregating with low levels of either gene product. In the and p = 5.2 × 10-19, respectively), suggesting colocalization of MS European population, the MS risk SNP rs7731626 G allele is Frontiers in Immunology | www.frontiersin.org 11 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC uniquely found in a single haplotype together with the rs13186299 genotype was significantly associated with mRNA expression levels G allele (Figure 6E). Therefore, homozygotes for the rs7731626 G of ANKRD55 (p < 0.01) but not with that of IL6ST, while the other allele, corresponding to 38% of European rs7731626 sample two SNPs were not associated with expression levels of any gene genotypes, are predicted to display genetically determined low (Figures 7A, C). In immature moDCs, no association between levels of both genes in moDC. Levels of ANKRD55 splice variants genotype of any SNP and expression level of ANKRD55 or IL6ST induced by AM580 in immature moDC compared to untreated cells was observed (Figures 7B, D). It is worth noting that expression were significantly higher in homozygous carriers of MS risk alleles levels of both genes in moDC of rs13186299 GG homozygotes were but not significantly different in homozygotes of the protective higher in MS patients (Figure 7B) than in healthy controls alleles, though the scarcity of the latter genotype restricted the power (Figure 6C) (trending at p = 0.12 for ANKRD55 and significant of the test in this group (Supplementary Table S3). for IL6ST at p < 0.0001 upon comparison of both groups), while those of C carriers were not significantly different (p > 0.05 for ANKRD55 and IL6ST Gene Expression comparison of both groups). This observation is suggestive for Patterns in CD4+ T Cells and Monocyte- allele-restricted preconditioning of CD14+ monocytes in MS Derived Dendritic Cells of Multiple patients. Expression levels of the three sgp130 isoforms, i.e., Sclerosis Patients sgp130, sgp130-RAPS, and sgp130-E10, were also measured by We next set out to verify whether these genetic influences on qPCR but did not vary with SNP genotype, though two weaker ANKRD55 and IL6ST gene expression are measurable in CD4+ T associations (p < 0.05; CD4+ sgp130-RAPS for rs13186299 and lymphocytes and immature moDCs derived from CD14+ moDC sgp130-E10 for rs7731626) were observed (Figures 7C, D). monocytes freshly isolated from venous blood samples from We calculated Pearson correlation coefficients between untreated MS patients (clinical and demographic data of MS individual ANKRD55 and IL6ST mRNA expression levels in patients in Supplementary Table S1). In CD4+ T cells, rs7731626 either cell subset. In immature moDC, expression levels of IL6ST B A C D E F FIGURE 7 | Expression of ANKRD55, IL6ST, and sgp130 isoforms in CD4+ T cells (A, C) and immature moDC (B, D) of MS patients. Panels (A, B) represent the effects of rs7731626 genotype in CD4+ T cells (mean ± SEM; n = 40, Cohort 2; Kruskal–Wallis test) and of rs13186299 genotype in moDC (mean ± SEM; n = 59, Cohorts 1 and 2; Mann–Whitney test), respectively. (C, D) Plot p-values for the three SNPs’ effects on the expression of ANKRD55, IL6ST, sgp130, sgp130-RAPS, and sgp130-E10 in CD4+ T cells and moDC (Mann–Whitney or Kruskal–Wallis tests). (E, F) represent Pearson correlation of ANKRD55 and IL6ST 2-Dct values in CD4+ T cells and in moDC of MS patients from panels (A, B), respectively. **p ≤ 0.01, ****p ≤ 0.0001. n.s., not significant. Frontiers in Immunology | www.frontiersin.org 12 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC and ANKRD55 were moderately positively correlated with moDC compared to the protective alleles. Though we did not Pearson’s r = 0.45 (p < 0.0001), while in CD4+ T cells, their perform a formal eQTL study, further analysis revealed that these expression levels were more strongly correlated with Pearson’s allelic effects in moDC are arguably due to LD of the MS risk SNPs r = 0.76 (p < 0.0001) (Figures 7E, F). We considered clinical with another nearby intronic SNP in ANKRD55, rs13186299. This course as a variable. In moDC, increased ANKRD55 gene SNP has not arisen from GWAS in MS but from a GWAS on expression was observed in PP MS patients compared to HC monocyte percentage of white blood cells (7) and is the main (p < 0.05), as well as decreased IL6ST expression in both PP MS GTEx eQTL SNP for IL6ST expression in whole blood. Tissue- and RR+SP MS groups (p < 0.01; Figures 8A, B). Finally, we specific opposite eQTL effects have been observed for GWAS quantified serum levels of sgp130 by ELISA in MS patients. SNPs associated with complex traits including MS, with a fraction Sgp130 levels did not differ in PP vs. RR+SP patients (Figure 8C) of these showing discordant effects in untreated CD4+ T cells and but showed significant variation according to rs7731626 monocytes (45, 47, 48). Around 30% of the 200 GWAS autosomal genotype. Higher levels of sgp130 were observed in risk variants contained within the current MS genomic map can be homozygotes of the protective allele (p < 0.05), and similar tagged to cis-QTL effects on gene expression in either naive CD4+ trends were seen with both other SNPs (Figure 8D). T cells or monocytes (3). Five SNPs passed the FDR threshold for expression of ANKRD55 in CD4+ T cells and PBMCs but not in monocytes or brain (3). Our work now links ANKRD55 MS risk DISCUSSION SNPs already associated with a constitutive regulatory cis effect on ANKRD55 and IL6ST gene expression in CD4+ T cells now to a In the present study, we demonstrate that the ANKRD55 gene is distinct regulatory element mediating an opposite IL-4/GM-CSF- induced in monocytes during differentiation into immature induced cis effect in monocytes. moDC in the presence of IL-4/GM-CSF. Expression of MoDCs, used widely as an in vitro DC model, are in fact a highly ANKRD55 in these cells is enhanced by a retinoic acid agonist plastic cell type capable of acquiring a multitude of discrete and suppressed by maturation with IFN-g/LPS treatment, while functions. They are closely related to and share transcriptomic expression of IL6ST, the adjacent gene, appears less affected by gene signatures with naturally occurring inflammatory DCs that either of these treatments. Expression levels of both genes in emerge from monocytes under inflammatory conditions (49–52) immature moDC are similarly modulated by the MS risk SNPs and that do not develop via a common DC precursor in contrast to rs6859219 and rs7731626, with opposite directional effect to that pDC and cDC. Inflammatory DCs show considerable phenotypic seen in T lymphocytes: MS risk alleles are associated with higher heterogeneity, and they function mainly at a site of inflammation levels of ANKRD55 and IL6ST in CD4+ T lymphocytes of healthy rather than migrating to lymph nodes. One such subset, SLAN+ DC, controls (29–32) but, as shown here, with lower levels in immature represents a subfraction of CD14+/CD16+ monocytes with A B C D FIGURE 8 | Differential (A) ANKRD55 and (B) IL6ST gene expression levels in immature moDC according to clinical course of MS, PP, or RR+SP, as measured by qPCR. Mean ± SEM; n = 80, Cohorts 1 and 2, Kruskal–Wallis test (followed by Dunn’s multiple comparison test). (C) Levels of serum sgp130 protein measured by ELISA in PP MS and RR+SP MS. Mean ± SEM; n = 112, Cohorts 1, 2, and 3, Mann–Whitney test (n.s.). (D) Serum sgp130 levels in MS patients from panel (C) stratified for rs6859219, rs7731626, or rs13186299 genotype. Kruskal–Wallis test (followed by Dunn’s multiple comparison test). *p ≤ 0.05, **p ≤ 0.01. n.s., not significant. Frontiers in Immunology | www.frontiersin.org 13 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC pronounced pro-inflammatory activities that has been shown to genomic cis-variants affecting cell-specific sgp130 isoform splicing accumulate in highly inflammatory brain lesions in MS patients or stability, and/or posttranslational processing and trafficking of (53). More recently, single-cell RNA sequencing studies identified sgp130, that are correlated to the main MS risk SNP. A missense the DC subset DC4 that is marked by high CD16, shares signatures variant (Gly148Arg) in exon 5 of the IL6ST gene, rs2228044, that with monocytes, and may be related to SLAN+ cells (51, 52). has been associated with higher serum sgp130 levels (59) is not in Though their role in MS is still poorly understood, inflammatory LD with rs7731626 (R2 = 0.0008 in the European population), DCs are known to stimulate CD4+ T cells to produce IL-17 and ruling out a mechanistic correlation. In light of these current induce Th17 differentiation from naive CD4+ T cells (49). These observations, and considering serum sgp130 appears to be a activities may be mediated through IL-6 trans-signaling into T cells polygenic trait influenced by genomic variants acting in trans that abrogates de novo induction of autoimmune inhibitory (60), decreased circulating sgp130 levels constitute likely a regulatory T cells (Tregs) (54) and promotes differentiation of secondary effect of the major causative link to MS risk conferred pathogenic Th17 cells (55). Th17 cells are crucial contributors to by rs7731626, which thus far is understood as upregulation of both CNS inflammation in MS (56). These effects can be modulated by ANKRD55 and IL6ST expression in CD4+ T lymphocytes (34, 36). sgp130, which inhibits IL-6 trans-signaling by binding to the IL-6/ Though the expression of both these genes has been reported sIL-6R complex, thus blocking its interaction with gp130 cell surface to be affected to a similar extent in HC by rs7731626 (34), we receptors (57). show that in MS patients, rs7731626 is significantly associated Given this context, as part of this study, we analyzed the with ANKRD55 expression in the CD4+ T cells of MS patients expression of three secreted gp130 isoforms (sgp130, sgp130- (Figure 7A). Roostaei et al. (35) reported a cis-eQTL effect for RAPS, and sgp130-E10) produced by the IL6ST gene by qPCR this SNP on ANKRD55 expression in both naive and memory and of circulating sgp130 by ELISA. In CD4+ T cells of HC, MS risk CD4+ T cells of MS patients. In immature moDC of MS patients, alleles when compared to protective alleles are associated with ANKRD55 and IL6ST expression was independent of the tested higher sgp130 isoform mRNA levels, while in moDC, they are SNPs, in contrast to HC, whose expression levels of either gene associated with lower levels. Serum levels of sgp130 were higher in were significantly associated with rs7731626 and rs13186299. MS patients, homozygote of the protective allele of MS risk SNPs, This discrepancy may be the result of preconditioning of blood reaching significance for rs7731626 (Figure 8D), and this is monocytes by factors enriched in serum of MS patients, prior to concordant with a recent publication showing significantly their isolation and differentiation into moDC, and was only seen increased serum sgp130 levels in MS carriers of the protective in rs13186299 GG homozygotes. Though we did not perform a allele of the correlated MS risk SNP rs71624119 (58; Figure 1). In functional analysis, we assessed whether clinical course could be the latter study, during fingolimod treatment, serum levels of sgp130 a priming factor affecting moDC expression of both genes. In the increased in all patients but to a higher degree in patients who are PP form of MS, but not in RR or SP MS, blood monocytes are homozygous for the risk allele (58). Intrinsically, findings from our characterized by an IL-1b signature consequential to NLRP3 study and the study by Bedri et al. (58) are compatible with the inflammasome overactivation (61). Analysis of immature moDC trans-signaling model of higher protection against MS risk through generated from circulating monocytes of PP, RR, and SP MS increased sgp130-mediated downregulation of IL-6/sIL6R- patients revealed enhanced expression of ANKRD55 uniquely in dependent pro-inflammatory activities and highlight these MS the PP MS group, while IL6ST was decreased in all MS patients risk SNPs as potential determinants of circulating sgp130. regardless of clinical course. These effects were independent of However, this pattern is incongruent with the rs7731626 genotype genotype, which therefore in HC but less so in MS determines effect on sgp130 mRNA levels in CD4 + T lymphocytes moDC coregulation of ANKRD55 and IL6ST. (Supplementary Figure S3), which shows a pattern in In this study, ANKRD55 was observed in both cytosolic and homozygotes that is opposed to that of serum sgp130 levels nuclear compartments of moDC. We found that nuclear (Figure 8D). One possible explanation is that rs7731626, ANKRD55 colocalized with proteins typically contained within probably via LD with eQTL-linked genomic variants, such as nuclear speckles. Nuclear speckles were originally thought to be rs13186299, may affect serum sgp130 production by other PBMC sites for storage and modification of RNA splicing factors but are subsets in opposite allelic direction to that in CD4+ T lymphocytes. now recognized as hubs for integrating all the nuclear gene DICE database2 indicates opposite genotype effects for rs7731626 expression regulation steps (46). Previously, using a distinct on IL6ST transcription in B cells, classical and non-classical antibody to ANKRD55 and different permeabilization methods monocytes, and NK cells when compared to CD4+ T cells. Still, (27), we could not demonstrate clearly its nuclear localization in among PBMC, CD4+ T lymphocytes, the most abundant PBMC HEK293 cells by IF but demonstrated its association with the subset, are comparatively also the highest producers of IL6ST and mitotic spindle. However, the enrichment of RNA-binding sgp130 mRNA (Protein Atlas; Supplementary Figure S2). The proteins in the HEK293 nuclear interactome of ANKRD55 genotype-serum sgp130 ELISA data in this study that contrasts with (27) is compatible with a nuclear speckles-based location seen this could refer therefore in fact to a relatively limited contribution here, as some proteins we identified in that interactome are of CD4+ T lymphocytes to circulating levels of sgp130 or to established NS proteins including the RNA helicases DDX3X, additional sgp130-producer contributors from vessel walls or non- DDX21, nuclear RNA export factor NXF1, RNA-binding protein PBMC subsets. It may also point to as yet undisclosed effects of RBM14, and RNA stabilizer ELAVL1 (27, 46). 2 https://dice-database.org Frontiers in Immunology | www.frontiersin.org 14 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC In conclusion, this study reports coregulation of ANKRD55 AUTHOR CONTRIBUTIONS and IL6ST in immature moDCs by genomic variants associated with MS risk. Though little is known about ANKRD55’s JM, IA, RTN, AAl, JDG, AVE, SM, and NV contributed to the biological relevance to autoimmunity, its expression in acquisition and analysis of the experimental data. CL provided immature moDC and upregulation by retinoic acid and assistance and antibodies for immunofluorescence microscopy. downregulation by TLR maturation factors suggest a specific AAn, SB, MM-B, AAA, JLSM, LM, XM, LMV, and MC provided role associated with the differentiation stage of monocyte biological samples and clinical data of MS patients. The article was precursor to DCs. Our study highlights the importance of written by JLSM and KV. All authors edited the article. All authors performing more extensive genome-wide QTL studies in contributed to the article and approved the submitted version. intermediate or induced states of immune cells in order to gather an integrated understanding of the multifarious expression of MS risk conferred by associated genomic variants. FUNDING This research was supported by grants to KV from MINECO DATA AVAILABILITY STATEMENT (SAF2016-74891-R), Instituto de Salud Carlos III (FIS-PI20/00123), Gobierno Vasco RIS3 (Ref. 2019222043), and Red Española de The original contributions presented in the study are included in Esclerosis Mú ltiple (REEM; RD16/0015/0005). NV and LMV were the article/Supplementary Material. Further inquiries can be supported by ISCIII (FIS-PI18/00572) and REEM (RD16/0015/0001). directed to the corresponding author. RTN is a recipient of a fellowship from the Secretarı́a Nacional de Ciencia y Tecnologı́a e Innovació n (SENACYT; Convocatoria Doctorado de Investigació n Ronda III, 2018; Ref. BIDP-III-2018-12) of the Gobierno Nacional, Repú blica de Panamá . ETHICS STATEMENT The studies involving human participants were reviewed and approved by the Comité É tico de Investigació n Clı́nica de Euskadi, SUPPLEMENTARY MATERIAL the Comité É tico de Investigació n Clínica Hospital Universitari Vall D’Hebron, and the Comité É tico de la Investigació n del Hospital The Supplementary Material for this article can be found online at: Ramó n y Cajal. The patients/participants provided their written https://www.frontiersin.org/articles/10.3389/fimmu.2021.816930/ informed consent to participate in this study. full#supplementary-material 10. International Multiple Sclerosis Genetics Consortium (MSGC), Beecham AH, REFERENCES Patsopoulos NA, Xifara DK, Davis MF, Kemppinen A, et al. Analysis of 1. Okada Y, Wu D, Trynka G, Raj T, Terao C, Ikari K, et al. Genetics of Immune-Related Loci Identifies 48 New Susceptibility Variants for Multiple Rheumatoid Arthritis Contributes to Biology and Drug Discovery. Nature Sclerosis. Nat Genet (2013) 45(11):1353–60. doi: 10.1038/ng.2770 (2014) 506(7488):376–81. doi: 10.1038/nature12873 11. Eyre S, Bowes J, Diogo D, Lee A, Barton A, Martin P, et al. High-Density 2. Laufer VA, Tiwari HK, Reynolds RJ, Danila MI, Wang J, Edberg JC, et al. Genetic Mapping Identifies New Susceptibility Loci for Rheumatoid Arthritis. Genetic Influences on Susceptibility to Rheumatoid Arthritis in African- Nat Genet (2012) 44:1336–40. doi: 10.1038/ng.2462 Americans. Hum Mol Genet (2019) 28(5):858–74. doi: 10.1093/hmg/ddy395 12. Kim K, Bang SY, Lee HS, Cho SK, Choi CB, Sung YK, et al. High-Density 3. International Multiple Sclerosis Genetics Consortium (MSGC). Multiple Sclerosis Genotyping of Immune Loci in Koreans and Europeans Identifies Eight New Genomic Map Implicates Peripheral Immune Cells and Microglia in Rheumatoid Arthritis Risk Loci. Ann Rheum Dis (2015) 74(3):e13. Susceptibility. Science (2019) 365(6460):eaav7188. doi: 10.1126/science.aav7188 doi: 10.1136/annrheumdis-2013-204749 4. Li YR, Li J, Zhao SD, Bradfield JP, Mentch FD, Maggadottir SM, et al. Meta- 13. Má rquez A, Kerick M, Zhernakova A, Gutierrez-Achury J, Chen WM, Analysis of Shared Genetic Architecture Across Ten Pediatric Autoimmune Onengut-Gumuscu S, et al. Meta-Analysis of Immunochip Data of Four Diseases. Nat Med (2015) 21(9):1018–27. doi: 10.1038/nm.3933 Autoimmune Diseases Reveals Novel Single-Disease and Cross-Phenotype 5. Astle WJ, Elding H, Jiang T, Allen D, Ruklisa D, Mann AL, et al. The Allelic Associations. Genome Med (2018) 10(1):97. doi: 10.1186/s13073-018-0604-8 Landscape of Human Blood Cell Trait Variation and Links to Common 14. Stahl EA, Raychaudhuri S, Remmers EF, Xie G, Eyre S, Thomson BP, et al. Complex Disease. Cell (2016) 167:1415–29. doi: 10.1016/j.cell.2016.10.042 Genome-Wide Association Study Meta-Analysis Identifies Seven New 6. Chen MH, Raffield LM, Mousas A, Sakaue S, Huffman JE, Moscati A, et al. Rheumatoid Arthritis Risk Loci. Nat Genet (2010) 42:508–14. doi: 10.1038/ng.582 Trans-Ethnic and Ancestry-Specific Blood-Cell Genetics in 746,667 15. Jostins L, Ripke S, Weersma RK, Duerr RH, McGovern DP, Hui KY, et al. Host- Individuals From 5 Global Populations. Cell (2020) 182(5):1198–213. Microbe Interactions Have Shaped the Genetic Architecture of Inflammatory doi: 10.1016/j.cell.2020.06.045 Bowel Disease. Nature (2012) 491(7422):119–24. doi: 10.1038/nature11582 7. Vuckovic D, Bao EL, Akbari P, Lareau CA, Mousas A, Jiang T, et al. The 16. Liu JZ, van Sommeren S, Huang H, Ng SC, Alberts R, Takahashi A, et al. Polygenic and Monogenic Basis of Blood Traits and Diseases. Cell (2020) 182 Association Analyses Identify 38 Susceptibility Loci for Inflammatory Bowel (5):1214–31. doi: 10.1016/j.cell.2020.08.008 Disease and Highlight Shared Genetic Risk Across Populations. Nat Genet 8. Alloza I, Otaegui D, Lopez de Lapuente A, Antigüedad A, Varadé J, Nú ñez C, (2015) 47(9):979–86. doi: 10.1038/ng.3359 et al. ANKRD55 and DHCR7 are Novel Multiple Sclerosis Risk Loci. Genes 17. Rose-John S. Interleukin-6 Family of Cytokines. Cold Spring Harb Perspect Immun (2012) 13:253–7. doi: 10.1038/gene.2011.81 Biol (2018) 10(2):a028415. doi: 10.1101/cshperspect.a028415 9. Lill CM, Schjeide BMM, Graetz C, Liu T, Damotte V, Akkad DA, et al. Genome- 18. Scheller J, Chalaris A, Schmidt-Arras D, Rose-John S. The Pro- and Anti- Wide Significant Association of ANKRD55 Rs6859219 and Multiple Sclerosis Inflammatory Properties of the Cytokine Interleukin-6. Biochim Biopys Acta Risk. J Med Genet (2013) 50:140–3. doi: 10.1136/jmedgenet-2012-101411 (2011) 1813(5):878–88. doi: 10.1016/j.bbamcr.2011.01.034 Frontiers in Immunology | www.frontiersin.org 15 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC 19. Jostock T, Müllberg J, Ozbek S, Atreya R, Blinn G, Voltz N, et al. Soluble Gp130 is 38. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, the Natural Inhibitor of Soluble Interleukin-6 Receptor Transsignaling Responses. et al. Fiji: An Open-Source Platform for Biological-Image Analysis. Nat Eur J Biochem (2001) 268(1):160–7. doi: 10.1046/j.1432-1327.2001.01867.x Methods (2012) 9:676–82. doi: 10.1038/nmeth.2019 20. Tanaka M, Kishimura M, Ozaki F, Hashimoto H, Okubo M, Murakami M, 39. Shihan MH, Novo SG, Le Marchand SJ, Wang Y, Duncan MK. A Simple et al. Cloning of Novel Soluble Gp130 and Detection of its Neutralizing Method for Quantitating Confocal Fluorescent Images. Biochem Biophys Rep Autoantibodies in Rheumatoid Arthritis. J Clin Invest (2000) 106(1):137–44. (2021) 25:100916. doi: 10.1016/j.bbrep.2021.100916 doi: 10.1172/JCI7479 40. Martin JCJ, Bé riou G, Heslan M, Chauvin C, Utriainen L, Aumeunier A, et al. 21. Sommer J, Garbers C, Wolf J, Trad A, Moll JM, Sack M, et al. Alternative Interleukin-22 Binding Protein (IL-22BP) is Constitutively Expressed by a Intronic Poyladenylation Generates the Interleukin-6 Trans-Signaling Subset of Conventional Dendritic Cells and is Strongly Induced by Retinoic Inhibitor Sgp130-E10. J Biol Chem (2014) 289(32):22140–50. doi: 10.1074/ Acid. Mucosal Immunol (2014) 7:101–13. doi: 10.1038/mi.2013.28 jbc.M114.560938 41. Schinnerling K, Garcı́a-Gonzá lez P, Aguilló n JC. Gene Expression Profiling of 22. Brod SA, Bauer VL. Ingested (Oral) Tocilizumab Inhibits EAE. Cytokine Human Monocyte-Derived Dendritic Cells – Searching for Molecular (2014) 68(2):86–93. doi: 10.1016/j.cyto.2014.04.003 Regulators of Tolerogenicity. Front Immunol (2015) 6:528. doi: 10.3389/ 23. Ayzenberg I, Kleiter I, Schröder A, Hellwig K, Chan A, Yamaura T, et al. fimmu.2015.00528 Interleukin 6 Receptor Blockade in Patients With Neuromyelitis Optica 42. Iwata M, Yokota A. Retinoic Acid Production by Intestinal Dendritic Cells. Nonresponsive to Anti-CD20 Therapy. JAMA Neurol (2013) 70(3):394–7. Vitam Horm (2011) 86:127–52. doi: 10.1016/B978-0-12-386960-9.00006-X doi: 10.1001/jamaneurol.2013.1246 43. Chuluundorj D, Harding SA, Abernethy D, La Flamme AC. Expansion and 24. Kieseier BC, Stüve O, Dehmel T, Goebels N, Leussink VI, Mausberg AK, et al. Preferential Activation of the CD14(+)CD16(+) Monocyte Subset During Disease Amelioration With Tocilizumab in a Treatment-Resistant Patient Multiple Sclerosis. Immunol Cell Biol (2014) 92:509–17. doi: 10.1038/icb.2014.15 With Neuromyelitis Optica: Implication for Cellular Immune Responses. 44. Waschbisch A, Schröder S, Schraudner D, Sammet K, Weksler B, Melms A, et al. JAMA Neurol (2013) 70(3):390–3. doi: 10.1001/jamaneurol.2013.668 Pivotal Role for CD16+ Monocytes in Immune Surveillance of the Central 25. Sellner J, Sitte HH and Rommer PS. Targeting Interleukin-6 to Treat Nervous System. J Immunol (2016) 196:1558–67. doi: 10.4049/jimmunol.1501960 Neuromyelitis Optica Spectrum Disorders: Implications From Immunology, 45. Schmiedel BJ, Singh D, Madrigal A, Valdovino-Gonzalez AG, White BM, the FCRn Pathway and Clinical Experience. Expert Rev Neurother (2020) 20 Zapardiel-Gonzalo J, et al. Impact of Genetic Polymorphisms on Human (5):509–16. doi: 10.1080/14737175.2020.1757434 Immune Cell Gene Expression. Cell (2018) 175(6):1701–15. doi: 10.1016/ 26. Drew K, Lee C, Huizar RL, Tu F, Borgeson B, McWhite C, et al. Integration of j.cell.2018.10.022 Over 9,000 Mass Spectrometry Experiments Builds a Global Map of Human 46. Galganski L, Urbanek MO, Kryzosial WJ. Nuclear Speckles: Molecular Protein Complexes. Mol Syst Biol (2017) 13(6):932. doi: 10.15252/msb.20167490 Organization, Biological Function and Role in Disease. Nucleic Acids Res 27. Ugidos N, Mena J, Baquero S, Alloza I, Azkargorta M, Elortza F, et al. (2017) 45:10350–68. doi: 10.1093/nar/gkx759 Interactome of the Autoimmune Risk Protein ANKRD55. Front Immunol 47. Mizuno A, Okada Y. Biological Characterization of Expression Quantitative (2019) 10:2067. doi: 10.3389/fimmu.2019.02067 Trait Loci (eQTLs) Showing Tissue-Specific Opposite Directional Effects. Eur 28. Chen Z, Yu H, Shi X, Warren CR, Lotta LA, Friesen M, et al. Functional J Hum Genet (2019) 27:1745–56. doi: 10.1038/s41431-019-0468-4 Screening of Candidate Causal Genes for Insulin Resistance in Human 48. Raj T, Rothamel K, Mostafavi S, Ye C, Lee MN, Replogle JM, et al. Polarization Preadipocytes and Adipocytes. Circ Res (2020) 126(3):330–46. doi: 10.1161/ of the Effects of Autoimmune and Neurodegenerative Risk Alleles in CIRCRESAHA.119.315246 Leukocytes. Science (2014) 344:519–23. doi: 10.1126/science.1249547 29. Hu X, Kim H, Stahl E, Plenge R, Daly M, Raychaudhuri S. Integrating 49. Segura E, Touzot M, Bohineust A, Cappuccio A, Chiocchia G, Hosmalin A, Autoimmune Risk Loci With Gene-Expression Data Identifies Specific et al. Human Inflammatory Dendritic Cells Induce Th17 Cell Differentiation. Pathogenic Immune Cell Subsets. Am J Hum Genet (2011) 89(4):496–506. Immunity (2013) 38:336–48. doi: 10.1016/j.immuni.2012.10.018 doi: 10.1016/j.ajhg.2011.09.002 50. Schlitzer A, McGovern N, Ginhoux F. Dendritic Cells and Monocyte-Derived 30. Lopez de Lapuente A, Feliú A, Ugidos N, Mecha M, Mena J, Astobiza I, et al. Cells: Two Complementary and Integrated Functional Systems. Semin Cell Novel Insights Into the Multiple Sclerosis Risk Gene ANKRD55. J Immunol Dev Biol (2015) 41:9–22. doi: 10.1016/j.semcdb.2015.03.011 (2016) 196(11):4553–65. doi: 10.4049/jimmunol.1501205 51. Collin M, Bigley V. Human Dendritic Cell Subsets: An Update. Immunology 31. Clark AD, Nair N, Anderson AE, Thalayasingam N, Naamane N, Skelton AJ, (2018) 154:3–20. doi: 10.1111/imm.12888 et al. Lymphocyte DNA Methylation Mediates Genetic Risk at Shared 52. Villani AC, Satija R, Reynolds G, Sarkizova S, Shekhar K, Fletcher J, et al. Immune-Mediated Disease Loci. J Allergy Clin Immunol (2020) 145 Single-Cell RNA-Seq Reveals New Types of Human Blood Dendritic Cells, (5):1438–51. doi: 10.1016/j.jaci.2019.12.910 Monocytes, and Progenitors. Science (2017) 356(6335):eaah4573. doi: 32. Chun S, Casparino A, Patsopoulos NA, Croteau-Chonka DC, Raby BA, De 10.1126/science.aah4573 Jager PL, et al. Limited Statistical Evidence for Shared Genetic Effects of 53. Thomas K, Dietze K, Wehner R, Metz I, Tumani HJ, Schultheisz T, et al. eQTLs and Autoimmune-Disease-Associated Loci in Three Major Immune- Accumulation and Therapeutic Modulation of 6-Sulfo LacNAc+ Dendritic Cell Types. Nat Genet (2017) 49(4):600–5. doi: 10.1038/ng.3795 Cells in Multiple Sclerosis. Neurol Neuroimmunol Neuroinflamm (2014) 1(3): 33. Westra HJ, Peters MJ, Esko T, Yaghootkar H, Schurman C, Kettunen J, et al. e33. doi: 10.1212/NXI.0000000000000033 Systematic Idenification of Trans eQTLs as Putative Drivers of Known Disease 54. Dominitzki S, Fantini MC, Neufert C, Nikolaev A, Galle PR, Scheller J, et al. Associations. Nat Genet (2013) 45(10):1238–43. doi: 10.1038/ng.2756 Cutting Edge: Trans-Signaling via the Soluble IL-6R Abrogates the Induction 34. Kundu K, Mann AL, Tardaguila M, Watt S, Ponsting H, Vasquez L, et al. of FoxP3 in Naive CD4+CD25 T Cells. J Immunol (2007) 179:2041–5. doi: Genetic Associations at Regulatory Phenotypes Improve Fine-Mapping of 10.4049/jimmunol.179.4.2041 Causal Variants for Twelve Immune-Mediated Diseases. bioRxiv (2020). 55. Bettelli E, Carrier Y, Gao W, Korn T, Strom TB, Oukka M, et al. Reciprocal doi: 10.1101/2020.01.15.907436 Developmental Pathways for the Generation of Pathogenic Effector TH17 and 35. Roostaei T, Klein HU, Ma Y, Felsky D, Kivisäkk P, Connor SM, et al. Proximal and Regulatory T Cells. Nature (2006) 441:235–8. doi: 10.1038/nature04753 Distal Effects of Genetic Susceptibility to Multiple Sclerosis on the T Cell 56. Kebir H, Kreymborg K, Ifergan I, Dodelet-Devillers A, Cayrol R, Bernard M, et al. Epigenome. Nat Commun (2021) 12(1):7078. doi: 10.1038/s41467-021-27427-w Human TH17 Lymphocytes Promote Blood-Brain Barrier Disruption and Central 36. Ló pez-Isac E, Smith SL, Marion MC, Wood A, Sudman M, Yarwood A, et al. Nervous System Inflammation. Nat Med (2007) 13:1173–5. doi: 10.1038/nm1651 Combined Genetic Analysis of Juvenile Idiopathic Arthritis Clinical Subtypes 57. Narazaki M, Yasukawa K, Saito T, Ohsugi Y, Fukui H, Koishihara Y, et al. Soluble Identifies Novel Risk Loci, Target Genes and Key Regulatory Mechanisms. Forms of the Interleukin-6 Signal-Transducing Receptor Component Gp130 in Ann Rheum Dis (2020) 80:321–8. doi: 10.1136/annrheumdis-2020-218481 Human Serum Possessing a Potential to Inhibit Signals Through Membrane- 37. Gó mez-Ferná ndez P, Urtasun A, Paton AW, Paton JC, Borrego F, Dersh D, Anchored Gp130. Blood (1993) 82:1120–6. doi: 10.1182/blood.V82.4.1120.1120 et al. Long Interleukin-22 Binding Protein Isoform-1 is an Intracellular 58. Bedri SK, Fink K, Manouchehrinia A, Lundström W, Kockum I, Olsson T, Activator of the Unfolded Protein Response. Front Immunol (2018) 9:2934. et al. Multiple Sclerosis Treatment Effects on Plasma Cytokine Receptor doi: 10.3389/fimmu.2018.02934 Levels. Clin Immunol (2018) 187:15–25. doi: 10.1016/j.clim.2017.08.023 Frontiers in Immunology | www.frontiersin.org 16 January 2022 | Volume 12 | Article 816930 Mena et al. ANKRD55–IL6ST and DC 59. Wonnerth A, Katsaros KM, Krychtiuk KA, Spedi WS, Kaun C, Thaler K, et al. Publisher’s Note: All claims expressed in this article are solely those of the authors Glycoprotein 130 Polymorphism Predicts Soluble Glycoprotein 130 Levels. and do not necessarily represent those of their affiliated organizations, or those of Metabolism (2014) 63(5):647–53. doi: 10.1016/j.metabol.2014.02.005 the publisher, the editors and the reviewers. Any product that may be evaluated in 60. Bonomi A, Veglia F, Baldassarre D, Strawbridge RJ, Golabkesh Z, Sennblad B, this article, or claim that may be made by its manufacturer, is not guaranteed or et al. Analysis of the Genetic Variants Associated With Circulating Levels of endorsed by the publisher. Sgp130. Results From the IMPROVE Study. Genes Immun (2020) 21:100–8. doi: 10.1038/s41435-019-0090-z Copyright © 2022 Mena, Alloza, Tulloch Navarro, Aldekoa, Díez García, Villanueva 61. Malhotra S, Costa C, Eixarch H, Keller CW, Amman L, Má rtinez-Banaclocha Etxebarria, Lindskog, Antigüedad, Boyero, Mendibe-Bilbao, Álvarez de Arcaya, Sań chez H, et al. NLRP3 in Flammasome as Prognostic Factor and Therapeutic Target Menoyo, Midaglia, Villarrubia, Malhotra, Montalban, Villar, Comabella and in Primary Progressive Multiple Sclerosis Patients. Brain (2020) 143:1414–30. Vandenbroeck. This is an open-access article distributed under the terms of the doi: 10.1093/brain/awaa084 Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) Conflict of Interest: The authors declare that the research was conducted in the are credited and that the original publication in this journal is cited, in accordance with absence of any commercial or financial relationships that could be construed as a accepted academic practice. No use, distribution or reproduction is permitted which does potential conflict of interest. not comply with these terms. Frontiers in Immunology | www.frontiersin.org 17 January 2022 | Volume 12 | Article 816930