Dendritic cells (DC), the classic antigen-presenting cells of the immune system, switch from an adhesive, phagocytic phenotype in tissues, to a mature, nonadhesive phenotype that enables migration to lymph nodes to activate T cells and initiate antitumor responses. Monocyte-derived DCs are used in cancer immunotherapy, but their clinical efficacy is limited. Here, we show that cultured bone marrow–derived DCs (BM-DC) expressing dysfunctional β2-integrin adhesion receptors displayed enhanced tumor rejection capabilities in B16.OVA and B16-F10 melanoma models. This was associated with an increased CD8+ T-cell response. BM-DCs expressing dysfunctional β2-integrins or manipulated to disrupt integrin adhesion or integrin/actin/nuclear linkages displayed spontaneous maturation in ex vivo cultures (increased costimulatory marker expression, IL12 production, and 3D migration capabilities). This spontaneous maturation was associated with an altered DC epigenetic/transcriptional profile, including a global increase in chromatin accessibility and H3K4me3/H3K27me3 histone methylation. Genome-wide analyses showed that H3K4me3 methylation was increased on DC maturation genes, such as CD86, Il12, Ccr7, and Fscn1, and revealed a role for a transcription factor network involving Ikaros and RelA in the integrin-regulated phenotype of DCs. Manipulation of the integrin-regulated epigenetic landscape in wild-type ex vivo–cultured BM-DCs enhanced their functionality in tumor rejection in vivo. Thus, β2-integrin–mediated adhesion to the extracellular environment plays an important role in restricting DC maturation and antitumor responses through regulation of the cellular epigenetic and transcriptional landscape. Targeting β2-integrins could therefore be a new strategy to improve the performance of current DC-based cancer immunotherapies.

Dendritic cells (DC) are professional antigen-presenting cells. After capturing antigen in peripheral tissues, they mature and migrate to lymph nodes. In lymph nodes, DCs present antigen to T cells resulting in T-cell activation. DCs therefore act at the crossroad of innate and adaptive immunity and are crucial for antitumor responses, which makes them promising anticancer agents. The only DC-based cancer therapy, sipuleucel-T (Provenge), was approved by the FDA for prostate cancer in 2010. Other DC-based immunotherapeutics have been tested in patients with for example malignant melanoma, prostate cancer, malignant glioma, and renal cell cancer (1). DC-based immunotherapeutics show low toxicity and induce immune responses in at least half of the treated patients. However, their clinical efficacy is less impressive, for a variety of reasons, including problems in the ability of injected DCs to migrate to lymph nodes (only about 1% of injected cells migrate) and in their T-cell activation and polarization capacity, which may not be optimal for antitumor responses (1). Therefore, although these therapies are promising, further development to optimize their functionality in patients is required.

DCs are broadly classified into conventional (or classic) cDCs and plasmacytoid pDCs, which have specialized functions in the immune system. Migratory DCs are DCs that have left their tissue of origin and migrated to lymph nodes to activate T cells. Broadly speaking, CD103+ migratory cDC1 are essential for CD8+ T-cell activation, whereas CD11b+ migratory cDC2 are the most efficient at driving CD4+ T-cell activation. DCs also migrate to lymph nodes in steady state, which is important for the induction of tolerance. Based on their transcriptional profile, granulocyte–macrophage colony-stimulating factor (GM-CSF)–cultured murine monocyte-derived DCs (moDC) resemble in vivo migratory DCs, making them a useful in vitro tool to investigate migratory DC function (2), and resemble the human moDCs (derived in culture with GM-CSF and IL4) that are used in the majority of immunotherapeutics tested in patients (3, 4). The migratory phenotype of DCs is associated with increased expression of Ccr7, the chemokine receptor for CCL19/CCL21, which guides DCs to lymph nodes. Migratory DCs also express many other genes associated with cell migration, such as Cd74 (5), Arc/Arg3.1 (6), and Fscn1 (7). However, the mechanisms mediating the transcriptional switch between the antigen capture phenotype of immature DCs and the migratory phenotype of mature DCs remain to be unraveled.

We have previously reported that the mature, migratory phenotype of DCs is restricted by β2-integrins (8). β2-integrins are a family of heterodimeric adhesion receptors that share the common β2 chain, which is also known as CD18. The four members are αLβ2 (or CD11a/CD18), αMβ2 (or CD11b/CD18), αXβ2 (or CD11c/CD18), and αDβ2 (or CD11d/CD18). β2-integrins are expressed on leukocytes and are involved in many essential immunologic processes that require cell adhesion to other cells, such as leukocyte adhesion under flow, immunologic synapse formation and phagocytosis. β2-integrin–mediated adhesion is controlled by cytoplasmic factors such as talin and kindlin-3, which regulate integrin activation and ligand binding. A mutation of three threonines of the β2-integrin cytoplasmic domain [TTT/AAA β2-integrin knock-in (KI)] renders all β2-integrins inactive by significantly reducing the binding of kindlin-3 (9). We have previously shown that murine bone marrow–derived DCs (BM-DC) expressing inactive integrins become reprogrammed to a mature, migratory phenotype, characterized by major transcriptional changes, increased surface expression of CD40, CD80, CD86, and CCR7, as well as increased cytokine production (IL10, IL12; ref. 8). These results highlight β2-integrins as essential regulatory elements of the phenotypic switch of DCs to the mature, migratory phenotype, but the mechanisms mediating this cellular reprogramming have remained elusive.

In this study, using murine BM-DCs as a model system, we have discovered a β2-integrin adhesion-regulated mechanism that controls the DC phenotypic switch from the antigen capture phenotype to the mature, migratory phenotype. We found that β2-integrins regulate the DC epigenetic state (histone methylation) and transcriptional landscape through transcription factors (TF) such as Ikaros and RelA. Furthermore, integrins restricted the expression of DC costimulatory molecules and cytokines, the cell's 3D migration speed and DC-mediated tumor rejection in vivo. External targeting of this integrin-regulated transcriptional program of DCs led to increased DC-mediated tumor rejection in two murine melanoma models in vivo. Our results highlight a new function of β2-integrin–mediated immune cell adhesion in regulating DC epigenetic and transcriptional programming, offering a potential target to optimize DC-based immunotherapy approaches in the future.

Mice and bone marrow collection

Integrin TTT/AAA β2-integrin KI mice have been previously described (9). Wild-type (WT) littermates and male and female C57BL/6N mice purchased from Charles River and Scanbur were used as controls for Integrin TTT/AAA β2-integrin KI mice. C57BL/6-Lmnaflox/flox mice were from Jackson (Jackson Laboratory Stock No 026284) and were crossed with C57BL/6-LysM-CRE (Jackson Laboratory Stock No 004781) to generate mice from which lamin A/C knockout (KO) moDCs could be generated; littermates were used as controls. Bone marrow from CCR7−/− mice was kindly provided by Professor Reinhold Foerster (Hannover Medical School, Hannover, Germany). Experiments were performed using male and female mice between ages 8 and 39 weeks. Animals were housed under conventional conditions in groups of up to five animals per cage with free access to food and water. Bone marrow was collected from euthanized animals and used for BM-DC cultures immediately, except for Lamin and CCR7−/− for which the isolated bones were shipped to Finland, and then bone marrow was isolated. For tumor experiments, C57BL/6J female mice were purchased from Scanbur. Mice were 8 weeks old when the experiments started. The experiments were performed according to the Finnish Act on Animal Experimentation (62/2006) and approved by the Finnish National Animal Experiment Board.

DC culture

DCs were generated by culturing bone marrow for 9–10 days in 10 ng/mL GM-CSF (PeproTech, cat. #AF-315-03) in DC media, i.e., RPMI (Lonza, cat. #BE12-167F) +10% FCS, 100 U/mL penicillin–streptomycin (penicillin, Orion, cat. #465161; streptomycin, Thermo Fisher Scientific, cat. #D7253—100 g), and 2 mmol/L L-glutamine (Thermo Fisher Scientific, cat. #BP379-100), at 37°C in a humidified atmosphere of 5% CO2. Media were added/changed on days 3, 6, and 8. In some experiments, cells were treated with cytochalasin D (Cyto D) 10 μg/mL (Tocris, cat. #1233), tranylcypromine hydrochloride (TCP) 5 μmol/L (Tocris, cat. #3852), lenalidomide 10 μmol/L (Sigma, cat. #SML2283), 8-hydroxy-5-quinolinecarboxylic acid (IOX1) 5 μmol/L (Cayman, cat. #11572), lipopolysaccharide (LPS) 100 ng/mL (Sigma, cat. #L6529), bafilomycin 12.5 nmol/L (Sigma, cat. #B1793), or MG132 2.5 μmol/L (Sigma, cat. #M7449), all overnight. For some experiments, DCs were seeded on fibronectin (10 μg/mL, Calbiochem, cat. #86088-83-7) or coated coverslips overnight. For DC incubation in suspension, DCs were detached from plates by incubating cells in 5 mmol/L EDTA (Sigma, cat. #E5134—1 kg) for 12 minutes and subsequently washing the cells off by pipetting up and down multiple times with 2% FCS/PBS, resuspended in DC media (described above) to 1 × 106/mL, and left in 50 mL conical centrifuge tubes (Falcon) with loose lid overnight.

Tumor models and cell lines

The murine melanoma cell line B16.F10 was purchased from the American Type Culture Collection (ATCC), and the ovalbumin-expressing murine melanoma cell line B16.OVA was provided by Prof. Richard Vile (Mayo Clinic, Rochester, MN). Both cell lines were genotyped in the beginning of the year. B16.OVA were used at a passage of 25 and B16F10 at a passage of 5. Mycoplasma testing was routinely done after 1 week post culturing by analyzing the supernatant of the cells with MycoAlertTM Mycoplasma Detection Kit (Lonza, cat. #LT07). Both cell lines were cultured in RPMI-1640 medium (Lonza, cat. #12167F) containing 10% FBS, 2 mmol/L L-glutamine (Thermo Fisher Scientific, cat. #BP379–100), 1% penicillin/streptavidin (penicillin, Orion cat. #465161; streptomycin, Thermo Fisher Scientific, cat. #D7253–100 g). B16.OVA was cultured under Geneticin (Thermo Fisher Scientific, cat. #10131035) selection at 37°C in a humidified atmosphere of 5% CO2. HL-60 transfected with a plasmid to overexpress lamin A/C along with controls transfected with ZsGreen1 retroviral vector were generated as previously described (10). HL-60 were provided by Ronen Alon in 2020, authentication and mycoplasma testing had been done in Ronen Alon's lab in 2017 and the cells stored in liquid nitrogen until shipping. HL-60 were used at passage 8. Cells were maintained in RPMI +10% FCS, 20 mmol/L HEPES, 100 U/mL penicillin–streptomycin, and 2 mmol/L L-glutamine (same vendors as DC culture), with media changes every three to four days. Two days prior to experiments, cells were differentiated into myeloid cells according to a previously published protocol, using 270 ng/mL A23187 (Chemcruz, cat. #sc-3591; ref. 11, 12).

Polyacrylamide gels with varying stiffness

In some experiments, cells were placed on polyacrylamide gels with different stiffness (Young's Moduli: 0.87 kPa, 11, 90; ref. 13) overnight. Gels were coated with iC3b (6 μg/mL; Calbiochem, cat. #204863-250UG).

ELISA

The level of IL12 in WT and KI BM-DC supernatants was assessed with mouse IL12/IL23 p40 allele-specific DuoSet ELISA kit according to the manufacturer's instructions (R&D, cat. #DY499). Briefly, a 96-well microplate was first coated with capture antibody overnight. The next day, the wells were washed and blocked with reagent diluent [1% BSA (Biowest, cat. #P6154) in PBS (Lonza, cat. #17-516F)] for a minimum of one hour. The plate was then washed, and standards and samples were added. Following a 2-hour incubation, the plate was washed, and the detection antibody was added to the wells. After 2 hours, the plate was washed and streptavidin-HRP was added for 20 minutes. After washing, substrate solution (R&D, cat. #DY499) was added to the wells. Following visible color development (or latest after 20 minutes), sulphuric acid (Acros Organics, cat. #124645001; diluted 1:4 in water) was added to stop the reaction. The optical density was immediately determined at 450 nm and background at 540 and 570 nm wave lengths using EnSpire fluorometer (PerkinElmer). Background readings were subtracted from the readings at 450 nm and the sample IL12 concentration was determined based on the standard curve. All steps were performed at room temperature. All washing steps were performed by washing the plate first twice with 0.05% Tween (Fisher BioReagents, cat. #BP337) in PBS and then once with PBS only. The plate was covered from light during incubation with streptavidin-HRP or substrate. All reagents were included in the kit DY499 unless stated otherwise.

Western blotting

Cells were lysed in M-PER lysis buffer (Thermo Fisher Scientific, cat. #78501) in the presence of phosphatase and protease inhibitors (Thermo Fisher Scientific, cat. #A32959), and lysates were analyzed by Western blotting. Primary antibodies against H3K4me3 (Cell Signaling Technology, cat. #9751S) and against H3K27me3 were from Cell Signaling Technology (cat. #9733S).

Immunofluorescence staining and microscopy

DCs (1 × 106) were seeded onto coverslips that were uncoated or coated with either fibronectin (10 μg/mL; Calbiochem, cat. #86088–83–7) or fibrinogen (10 μg/mL; Sigma, cat. #F-3879) on day 9, left overnight in RPMI (Lonza, cat. #BE12-167F) +10% FCS, 100 U/mL penicillin–streptomycin (penicillin, Orion, cat. #465161; streptomycin, Thermo Fisher Scientific, cat. #D7253—100 g) and 2 mmol/L L-glutamine (Thermo Fisher Scientific, cat. #BP379–100) and then fixed with 1% PFA. DCs incubated in suspension were left to adhere for 20 minutes before PFA fixation. Cells were permeabilized with 0.2% Triton × in PBS for 5 minutes and stained for 1 hour at room temperature with the following antibodies diluted in 1% FBS/PBS according to the suppliers' recommendations: H3K27me3 (Cell Signaling Technology, cat. #9733S), H3K4me2 (Cell Signaling Technology, cat. #9725T), H3K4me3 (Cell Signaling Technology, cat. #9751S), RNA polymerase II (phospho S2, Abcam cat. # ab5095), RelA (Santa Cruz Biotechnology, cat. #SC-8008), lamin A/C (Abcam, cat. #ab108922), and Lamin B (Santa Cruz Biotechnology, B-10, cat. #sc-374015). Cells were subsequently stained with an Alexa 594–conjugated anti-rabbit IgG (Thermo Fisher Scientific, cat. #A-21207) or Alexa 555–conjugated anti-mouse IgG (Thermo Fisher Scientific, cat. #A21422). Corrected total cell fluorescence (CTCF) values were calculated based on measurements of area, intensity, and background with ImageJ as previously described (14). Cells (25–100) per condition were measured for each animal. Nuclei were stained with DAPI (Sigma, cat. #D9542).

All slides were imaged using a Leica SP5 II (Leica Microsystems) LAS AF Lite Software using a sequential scan, with 561 laser (10% laser power): 580–752 nm, smart gain 900 V, smart offset 0.0% for scan 1. Scan 2 was done with 405 Laser (10% laser power): 415–575 nm, smart gain 750 V, smart offset −1.0%. Z-stacks were taken with the following parameters: max QD405/488/561/635 mirror, pinhole 111.39 μm, zoom: 1.00; objective 63X, z-volume 10.127 μm, 0.25 μm step size, line average 3, format 512 × 512.

DC flow cytometry

The following conjugated antibodies were used for flow cytometry of DCs and HL-60, analysis included at least 10,000 whole cells (see Supplementary Fig. S1 for gating strategies and Supplementary Fig. S2 for representative histograms; company, catalog numbers and clones given in brackets): CD11a-PE (BioLegend, cat. #101008, clone 2D7), CD18-FITC (BD Biosciences, cat. #553292, clone C71/16), CD11c-PE-Cy7 (eBioscience, cat. #25-0114-81, clone N418), CD11b-APC (BioLegend, cat. #17-0112-82, clone M1/70), MHCII-APC-eFluor780 (eBioscience, cat. #47-5321-82, clone M5/114.15.2), CD80-APC (cat. #17-0801-81, eBioscience, clone 16-10A1), CCR7-PE (cat. #120106, BioLegend, clone 4B12), CD40-PE (cat. #124609, BioLegend, clone 3/23), CD86-FITC (cat. #11-0862-85, BD Bioscience, clone GL1). Fc block (cat. #14-0161-86, eBioscience, clone 93) was included in all stains. The following isotype controls were used: FITC, BD Pharmingen, cat. #556652; PE, BD Pharmingen, cat. #556653; PE-Cy7, eBioscience, cat. #25-4321-81; APC, eBioscience, cat. #17-4888-81; APC-eFluor780, eBioscience, cat. #47-4321-80). For analysis of DC and macrophage proportions, cells were first resuspended in FACS buffer containing anti-CD16/32 (BD Pharmingen, cat. #553142, clone 2.4G2) to block Fc receptors. Cell-surface staining was then performed using the following antibodies: MHC class (II I-A/I-E) APC-eFluor780 (eBiocience, cat. #47-5321-82, clone M5/114.15), CD11c AlexaFluor700 (eBioscience, cat. #56-0114-80, clone N418), CD11b FITC (eBioscience, cat. #11-0112-82, clone M1/70), CD115 APC (eBioscience, cat. #17-1152-80, clone AFS98), CD135 PE (eBioscience, cat. #12-1351-81, clone A2F10) and MerTK BV421 (BioLegend, cat. #151510, clone 2B10C42). FMO controls were used for each fluorochrome and in addition, the following isotype controls were used: Armenian hamster IgG AlexaFluor700 (eBioscience, cat. #56-4888-80, clone eBio299Arm), rat IgG2a kappa APC (eBioscience, cat. #17-4321-81, clone eBR2a), rat IgG2a kappa PE (eBioscience, cat. #12-4321-80, clone eBR2a), rat IgG2a kappa BV421 (BioLegend, cat. #400535, clone RTK2758). Propidium iodide (from Sigma, cat. #P4864, or Merck, cat. #P4864Merck) was used to detect dead cells.

Acquisition was performed on an LSRFortessa flow cytometer (Becton Dickinson). Data were analyzed using FlowJo software (TreeStar).

Cell migration assays

3D migration assays were performed with μ-Slide Chemotaxis 3D (Ibidi, cat. #80326) imaging slides according to the manufacturer's protocol. Briefly, DCs were mixed into a bovine collagen I (CellSystems, cat. #5005-B) mix and injected into the slide's thin imaging strip. After 45 minutes of collagen polymerization, one of two chambers flanking the imaging strip was filled with DC media mentioned above, the other with media containing mCCL19 (R&D Systems, cat. #440-M3-025). One slide was used for each condition. DCs were imaged using the 3I Marianas imaging system (3I Intelligent Imaging Innovations) by utilizing multipoint imaging. Controls were always imaged before treated or KI/KO DCs. A 10×/0.30 EC Plan-Neofluar Ph1 WD = 5.2 M27 objective was used, the dish was placed in a heated sample chamber (+37°C), in controlled 5% CO2 atmosphere. Cells were imaged using bright-field microscopy. β2-integrin KI versus WT leukocyte migration was compared for 4 hours by cell tracking every 2 minutes (for all other analyses 1 minute). For calculating the mean migration speed the velocities of only migratory DCs (defined as DCs that migrated faster than 1 μm/minute) were incorporated.

Chromatin immunoprecipitation assays

For chromatin immunoprecipitation (ChIP), β2-integrin KI, WT, WT TCP, and WT TCP + suspension DCs were fixed in 1% formaldehyde/PBS for 10 minutes at room temperature, crosslinking was stopped by adding glycine (Sigma, cat. #50046) to a final concentration of 0.125 mol/L for 5 minutes, followed by scraping adherent cells off the plates. Cell pellets of 10 × 106 DCs per condition were lysed in 300 μL of RIPA buffer (Pierce, cat. #89900) and sonicated with Bioruptor (Diagenode; number of cycles = 15, power = HIGH, ON = 30 seconds, OFF = 30 seconds). At least 3 × 106 DCs were used per one IP. IPs were carried out with 5 μg antibody, H3K4me2 (Cell Signaling Technology, cat. #9725T, C64G9), H3K4me3 (Cell Signaling Technology, cat. #9751S, C42D8) overnight at 4°C in a rotating wheel. The immuno-complexes were collected with 50 μL of blocked protein A sepharose (GE Healthcare, cat. #17-0780-01) at 4°C for two hours with rotation. The beads were pelleted by centrifugation at 4°C for one minute at 500 × g and washed sequentially for five minutes on rotation with 1 mL of the following buffers: low-salt wash buffer (RIPA; 10 mmol/L Tris–HCl (pH 8.0), 0.1% SDS, 1% Triton X-100, 1 mmol/L EDTA, 140 mmol/L NaCl, 0.1% sodium deoxycholate), high-salt wash buffer (10 mmol/L Tris–HCl (pH 8.1), 0.1% SDS, 1% Triton X-100, 1 mmol/L EDTA, 500 mmol/L NaCl, 0,1% sodium deoxycholate) and LiCl wash buffer (10 mmol/L Tris–HCl (pH 8.1), 0.25 mmol/L LiCl, 0.5% IGEPAL CA-630, 0,5% sodium deoxycholate, 1 mmol/L EDTA). Finally, the beads were washed twice with 1 mL of TE buffer [10 mmol/L Tris–HCl (pH 8.0), 1 mmol/L EDTA]. Chromatin was eluted in 150 μL of 1% SDS in TE buffer. The crosslinking was reversed by adding NaCl to a final concentration of 200 mmol/L and incubating at 65°C overnight. The eluate was treated with Proteinase K (Sigma, cat. #P6556), and the DNA was recovered by extraction with phenol/chloroform/isoamylalcohol (25/24/1, Sigma; cat. #77617) and precipitated with 0.1 volume of 3 mol/L sodium acetate (pH 5.2) and two volumes of ethanol using glycogen as a carrier.

ChIP libraries were prepared for Illumina NextSeq 500 using NEBNext ChIP-seq DNA Sample Prep Master Mix Set for Illumina (NEB, cat. #E6240) and NEBNext Multiplex Oligos for Illumina (Index Primers Set 1; NEB, cat. #E7335) according to the manufacturer's protocols. Sequencing was performed with NextSeq500 at Biomedicum Functional Genomics Unit (FuGU), University of Helsinki. Sequencing was performed in duplicate. ChIP-sequencing (ChIP-seq) data sets were aligned using Bowtie2 [using ChIPster software (15)] to version mm10 (GRC m38) of the mouse genome with the default settings. To visualize and present ChIP-seq data, we used Integrative Genomics Viewer (IGV; ref. 16) and EaSeq (http://easeq.net; ref. 17). Fastq files are deposited in NCBI's Gene Expression Omnibus (GEO) database with BioProject ID PRJNA599281 and BioSample accession number SAMN14687879.

Assay for transposase-accessible chromatin sequencing

Assay for transposase-accessible chromatin sequencing (ATAC-seq) was performed according to the Omni-ATAC protocol (18), which is a modified version of the protocol published in the original Buenrostro and colleagues papers (19, 20). In the Omni-ATAC method, the composition of the cell lysis buffer and the transposition reaction buffer is modified to yield a higher number of peaks and reduce mitochondrial reads. Briefly, 50,000 DCs were used for each reaction/sample and lysed for 3 minutes on ice using 50 μL cold ATAC-resuspension buffer (10 mmol/L Tris–HCl, pH: 7.5, 10 mmol/L NaCl, 3 mmol/L MgCl2 in sterile water) containing 0.1% NP40, 0.1% Tween-20 and 0.01% Digitonin. The lysates were then washed with 1 mL cold ATAC-resuspension buffer containing only 0.1% Tween-20. Transposition mix (50 μL for each sample) was prepared using 25 μL 2× transposition buffer (Illumina, cat. #15027866), 2.5 μL Tn5 transposase enzyme (Illumina, cat. #15027865), 16.5 μL PBS, 0.5 μL 1% digitonin, 0.5 μL 10% Tween-20 and 5 μL sterile nuclease-free H2O. Pellet nuclei were resuspended in 50 μL transposition mix and incubated at 37°C for 30 minutes on a thermomixer with mixing. These reactions were cleaned up using a PCR cleanup and concentrator kit (Thermo Fisher Scientific GeneJET PCR purification kit, cat. #K0701) and eluted in 20 μL elution buffer. The eluted DNA was PCR amplified for 10 cycles using the following 50 μL PCR mix: 2.5 μL of each of the i5 and i7 index sequences (Illumina Nextera DNA CD Indexes, cat. #20015881), 25 μL NEBNext Ultra II Q5 Master Mix (cat. #M0544S) and 20 μL transposed samples. The PCR products were purified and size selected using AMPure XP beads (Beckman Coulter, cat. #A63880), the libraries were quantified using KAPA library quantification kit (KAPAbiosystems, cat. #07960140001). The size of the pooled library was examined by TapeStation. Paired-end (2 × 42 bp) sequencing was performed on Illumina NextSeq 500 at the Biomedicum Functional Genomics Unit (FuGU), University of Helsinki.

Paired reads were aligned to the mouse reference genome mm10 using STAR. PCR duplicates were removed using Picard followed by removal of mitochondrial reads. Peak-calling was performed using MACS2 with parameters –nomodel –shift -100 –extsize 200 –broad. Footprinting analysis on ATAC-seq data was performed using Transcription factor Occupancy prediction By Investigation of ATAC-seq Signal (TOBIAS; ref. 21). TOBIAS is a comprehensive computational framework that is capable of handling Tn5 insertion bias correction, calculating footprint scores within regulatory regions and estimating bound and unbound TF binding sites. The Tn5 transposase inherent insertion bias was corrected using TOBIAS ATACorrect tool and footprinting scoring was performed using TOBIAS ScoreBigwig. TF binding across the genome of β2-integrin KI DCs and WT DCs was predicted by TOBIAS BINDetect. Fastq files are deposited at NCBI's GEO database with BioProject ID PRJNA599281 and BioSample accession number SAMN13742355.

RNA sequencing

The RNA-sequencing (RNA-seq) library preparation and sequencing protocol was published previously (8). Here, the fastq raw files for all the three KI and three WT control samples were reanalyzed using the method below. After quality control performed by FastQC, sequencing reads fastq files were then aligned with HISAT2 alignment tool against the mouse reference genome (Mus musculus GRCm38.95). Transcripts were then counted per gene using a python package HTSeq. Differential expression analysis statistics was the performed by DESeq2. All genes having P <0.05 were considered as differentially expressed genes. Log2 fold change (log2FC) values, calculated by DESeq2, were indicator of upregulated (having positive log2FC values) and downregulated genes (having negative log2FC values).

Ikaros-regulated genes are from Baratin and colleagues (22) and PU.1-activated genes are listed from Boutboul and colleagues (23). Ikaros-regulated genes were compared with all upregulated genes in KI DCs identified by our RNA-seq analysis as described above. PU.1-regulated genes were compared with only those genes in our RNA-seq analysis that were at least 2-fold upregulated in KI DCs. Common intersect genes were identified using R. GO-analyses of the intersect genes were performed by metascape (24). RelA-bound genes (based on TLR stimulation) in BM-DCs were listed in Oh and colleagues (25) and were compared with all upregulated genes in the KI DCs RNA-seq analysis.

TRRUST analysis

Transcriptional Regulatory Relationships Unraveled by Sequence-based Text mining (TRRUST) is a manually curated database of human and mouse transcriptional regulatory networks. It uses sentence-based text mining approach. Currently, TRRUST contains 8,444 TF-target regulatory relationships of 800 human TFs, and 6,552 TF-target regulatory relationships for 828 mouse TFs. TRRUST version 2 was used here to identify active TFs associated with our genes of interest. The input gene list for this analysis was the common intersect 132 gene list of “Ikaros activated” and “upregulated in the KI.”

Quantitative real-time PCR

Total RNA was isolated from DCs with Nucleospin RNA kit (Macherey-Nagel, cat. #740955) and converted into cDNA using the High-Capacity cDNA Reverse Transcription kit (Thermo Fisher Scientific, cat. #4368814) according to the manufacturer's protocols. Quantitative real-time PCR (qRT-PCR) was performed using TaqMan chemistry. Briefly, the cDNA was amplified in 11 μL volume containing TaqMan Fast Advanced Master Mix (Thermo Fisher Scientific, cat. #4444557) and TaqMan primers/probes (CCR7 Mm01301785_m1, Thermo Fisher Scientific). Each sample was run in triplicate, 18S rRNA (Thermo Fisher Scientific, cat. #4333760T) was used as a reference gene, and no template control was included in the assay. Reactions were run with CFX96 Touch Real-Time PCR Detection System (Bio-Rad) and data analyzed with CFX Maestro (Bio-Rad) and finalized with Excel (Microsoft). Relative units were calculated by using the comparative CT method.

Tumor experiments

Tumors were injected subcutaneously in the right flank (B16.OVA 3 × 105 tumor cells/mouse and B16-F10 1.5 × 105 tumor cells/mouse) and grown until palpable. Mice were injected peritumorally with either 1 million β2-integrin KI DCs, WT DCs, or PBS solution (MOCK). Eight mice per each condition were used. Tumor growth was followed with a caliper every second day. Caliper measurements were done on both sides of the tumor and the volume was calculated by the following formula: V = (W × 2 × L), where W is tumor width and L is tumor length.

Immunophenotyping

To analyze the immune response systematically, blood, tumor and lymph node samples were immunophenotyped with flow cytometry. Peripheral blood (PB) was collected from the tail vein of mice before DC treatment and 4 days after. Red blood cells were lysed with ACK buffer (0.15 mol/L NH4Cl, 0.01 mol/L KHCO3, and 0.1 mmol/L EDTA). Tumors, spleens, and draining lymph nodes were collected at the endpoint and cryopreserved at −80°C before assessing immune cell content. See Supplementary Fig. S1 for gating strategies and Supplementary Fig. S2 for representative dot plots.

All samples were first resuspended in FACS buffer (2% FCS in PBS) containing anti-CD16/32 (BD Biosciences, cat. #553142, either clone 93 or 2.4G2) to block Fc receptors. The following antibodies were used for the phenotyping of lymphocytes (company, catalog numbers and clones given in parentheses): CD45-PE-Cy7 (eBioscience, clone 30-F11, cat. #25-0451-82) or CD45 BV650 (BD Biosciences, cat. #563410, clone 30-F11), CD3e AlexaFluor488 (eBioscience, cat. #53-0031-82, clone 145-2C11), CD4-PerCP (BD, cat. #561090, clone RM4-5) or CD4 PE-Cy7 (BioLegend, cat. #100528; clone RM4-5), CD8a-APC-Cy7 (BioLegend, cat. #100713, clone 53-6.7), and CD69-APC (BioLegend, cat. #104513, clone H1.2F3). Analysis of tumor-specific T-cell responses was done on samples taken on day 30 of the B16 OVA experiment: TruStain FcX anti-mouse CD16/32 (BioLegend, cat. #101320), CD8-FITC (ProImmune, cat. #A502-3B-E), CD3e-PE (BD Pharmingen, cat. #550353), and CD19-PerCP (BioLegend, cat. #115531). For intracellular staining (Tregs), cells were first surface stained and then fixed and permeabilized with Foxp3/Transcription Factor Staining Buffer Set (eBioscience, cat. #00-5523-00) and stained with Foxp3 APC (eBiocience, cat. #17-5773-80, clone FJK-16s). SIINFEKL epitope-specific T cells were studied using APC-labeled H-2Kb/SIINFEKL pentamer (ProImmune, cat. F093-84B214 E). Samples were acquired with LSRFortessa (BD) and analyzed with FlowJo (BD).

Statistical analysis

The Student two-tailed t test and the Mann–Whitney test were used to calculate statistical significance in GraphPad Prism 6. For migration assays, Mann–Whitney tests were used. For tumor growth experiments, two-way ANOVA was used. All P values are shown as *, <0.05; **, <0.01; ***, <0.005; ****, <0.0001.

Integrin inactivation increases DC maturation, migration, and DC-mediated antitumor responses in vivo

We have previously shown that β2-integrins restrict the mature, migratory phenotype of DCs, and DC-mediated T-cell activation and T helper cell (Th1) polarization in vivo (8, 26). Here, we show that the β2-integrin TTT/AAA KI mutation, which inhibits the interaction between the integrin and its cytoplasmic regulator kindlin-3, resulting in expressed but inactive β2-family integrins, specifically resulted in significantly increased BM-DC surface expression of the costimulatory molecules CD40, CD80, CD86, and the chemokine receptor CCR7 (Fig. 1A), but it did not affect macrophage/DC ratios in GM-CSF cultures (Supplementary Fig. S3). Additionally, β2-integrin KI DCs produced increased amounts of IL12 (p40; Fig. 1B) and migrated significantly faster than WT DCs in a 3D collagen matrix in the presence of CCL19 (Fig. 1C). LPS-induced activation of WT DCs led to significantly higher levels of CD40, CD80, and CD86 expression and IL12 production than were seen for nonactivated β2-integrin KI DC (Supplementary Figs. S2H and S4A), although migration speed was higher for β2-integrin KI DCs than for LPS-stimulated WT DCs (Supplementary Fig. S4B).

To further examine DC functionality in vivo, we performed tumor rejection experiments with subcutaneously injected B16-OVA melanoma cells. After palpable tumors were established, unstimulated WT or β2-integrin KI DCs were injected peritumorally. Both β2-integrin KI and WT DCs initiated an antitumor response that controlled tumor growth; however, β2-integrin KI DCs initiated a superior tumor rejection response (Fig. 1D). There was a significant increase in OVA-specific CD8+ T cells infiltrating tumors in mice injected with β2-integrin KI DCs compared with mice injected with WT DCs or MOCK DCs (Fig. 1E). Additionally, mice injected with β2-integrin KI DCs had significantly increased OVA-specific CD8+ T cells in the spleen and draining lymph nodes, demonstrating the systemic response to the tumor (OVA) protein (Fig. 1E).

Subsequent analysis of immune cell populations in B16-F10 tumors (Fig. 1F and G) revealed significantly more activated CD8+ T cells in the tumors of mice treated with β2-integrin KI DCs compared with mice treated with WT DCs (Fig. 1H and I). There were similar amounts of CD4+ T cells in the tumors of mice treated with β2-integrin KI DCs and those treated with MOCK (Supplementary Fig. S5A). However, analysis of pooled tumor-draining lymph nodes showed an increase of CD4+ T cells in mice that received β2-integrin KI DCs (Supplementary Fig. S5B). In addition, there were significantly fewer CD4+ T cells present in the blood stream in mice injected with β2-integrin KI DCs compared with mice that received WT DCs (Supplementary Fig. S5C). Taken together, these results indicate that active β2-integrins restrict DC maturation and migration, and that DCs expressing inactive integrins induce increased tumor suppression in vivo, through an increased CD8+ T-cell response.

Inactivating β2-integrins leads to an altered DC epigenetic state

We set out to investigate the molecular mechanism regulating DC maturation and migration downstream of β2-integrins. To identify regulatory pathways that could be responsible for the large gene-expression changes in β2-integrin KI DCs compared with WT DCs that we previously reported (8), we analyzed histone methylation in WT and KI DCs. H3K4me3 is associated with active gene expression, whereas H3K27me3 is associated with repressed genes and heterochromatin formation; bivalent promoters carrying both marks usually characterize “poised” genes, ready to be switched on in specific cell types (27). We therefore analyzed global levels of both methylation marks in WT and β2-integrin KI DCs using immunofluorescence. We found that β2-integrin KI DCs had significantly increased intensity of both H3K4me3 and H3K27me3 compared with WT DCs (Fig. 2AF). In β2-integrin KI DCs H3K27me3 was present at the periphery of the nucleus (Fig. 2E), which is characteristic for heterochromatin in most mammalian cells (28). Histone methylation changes in β2-integrin KI DCs were confirmed by Western blotting (Fig. 2G). Also, TLR stimulation of DCs (with LPS) led to increased histone methylation (Supplementary Fig. S6). We also found increased phosphorylated Pol II staining in these cells, indicating increased gene transcription (Fig. 2H).

Disrupting β2-integrin–mediated cell adhesion or the actin cytoskeleton results in DC maturation

We have shown that β2-integrins regulate the epigenetic state, phenotype, and antitumor function of DCs. To investigate the role of cell adhesion in this process, we disrupted β2-integrin–mediated DC adhesion by detaching WT DCs from culture plates and incubating them in suspension overnight. This procedure resulted in upregulation of the costimulatory molecules CD40, CD80, and CD86 as well as increased IL12 production (Fig. 3A and B; ref. 8). We found that these phenotypic changes were accompanied by increased H3K4me3 and H3K27me3 levels (Fig. 3C). Furthermore, we found that 3D migration speed in the presence of CCL19 was significantly increased for the first hour of the 3D migration assay (Fig. 3D), but this was not sustained over the whole assay period (presumably because detached cells start adhering again to their environment).

In addition to β2-integrins (CD11a, CD11b, and CD11c), murine BM-DCs also express β1-integrins and β3-integrins (29). However, adhesion of β2-integrin KI DCs neither to the β1-integrin ligand fibronectin nor to the β3-integrin ligand fibrinogen led to a reduction in histone methylation in these cells, showing that DC epigenetic state is regulated specifically by β2-integrins, and not generally by integrin-mediated cell adhesion (Fig. 3E and F).

Integrins link to the actin cytoskeleton within cells, and we have previously reported that integrin/actin links are disrupted in β2-integrin KI DCs (8). Disrupting the actin cytoskeleton in WT DCs using cytochalasin D resulted in similar upregulation of costimulatory molecule expression as is seen in β2-integrin KI DCs (Figs. 3G and 1A; ref. 8). Cytochalasin D treatment also resulted in significantly increased levels of H3K4me3 and H3K27me3 in DCs (Fig. 3H), indicating that β2-integrin/actin links regulate the epigenetic status of DCs. Overall, these results confirm that β2-integrin–mediated cell adhesion and β2-integrin/actin links regulate DC epigenetic state, maturation, and migration.

Β2-integrin/actin/lamin A/C cross-talk restricts histone methylation in DCs

Actin mechanically links integrin adhesion sites to the nucleus via the linker of nucleoskeleton and cytoskeleton (LINC) complex, which binds to the nuclear envelope composed of lamins (30). Lamins directly bind to DNA, specifically regions enriched with H3K9me2/3 and H3K27me3 (31). Furthermore, lamin A/C levels are regulated by mechanical stress, with cells on softer matrices having lower levels (32). We thus hypothesized that mechanical cross-talk between β2-integrins/kindlin-3/actin and nuclear lamins may physically regulate histone methylation state in DCs. Indeed, immunofluorescence staining of lamin A/C revealed a significant decrease of lamin A/C protein level in β2-integrin KI DCs compared with WT DCs (Fig. 3I); lamin B levels were unchanged (Fig. 3J). In addition, placing DCs on softer matrices coated with integrin ligands (and thereby reducing integrin-mediated mechanical stress in these cells) led to increased H3K4me3 levels (Fig. 3K). Lamin A/C was not regulated at the transcriptional level in β2-integrin KI BM-DCs (Fig. 3L). However, treating cells with the lysosomal inhibitor bafilomycin, but not the proteosomal inhibitor MG132, led to upregulation of lamin A/C protein levels in cells (Fig. 3M), showing that DCs respond to changes in mechanical stress by lysosomal degradation of lamin, as has previously been reported in other cell types (33).

To directly investigate the role of this integrin/actin/lamin link in regulating the epigenetic status of DCs, we utilized lamin A/C KO DCs. These cells displayed significantly increased H3K4me3 and H3K27me3 levels (Fig. 3N). However, lamin A/C KO DCs did not display increased costimulatory marker expression (Supplementary Fig. S7A), elevated IL12 (p40) production (Supplementary Fig. S7B), or increased 3D migration speed (Supplementary Fig. S7C). These results indicate that although lamin A/C absence in DCs results in changes of DC epigenetic status, lamin A/C absence alone is not enough to regulate the DC maturation status downstream of integrins when adhesion remains unchanged.

We hypothesized that by bypassing integrin-mediated mechanical signaling altogether, by increasing lamin levels in nonadhesive cells, it should be possible to modulate gene expression through this lamin-mediated epigenetic mechanism. Indeed, overexpressing lamin A/C in myeloid differentiated HL-60 cells led to significantly reduced CD80 and CCR7 expression in these cells (Fig. 3O and P), confirming that modulating lamin A/C levels can regulate expression of particular costimulatory molecules in myeloid cells. However, as lamin A/C deletion in DCs on its own was not sufficient to induce DC phenotypical changes, altered gene expression and function in adhesive DCs may involve the recruitment of specific (integrin/actin-regulated) TFs downstream of integrin outside-in signaling.

Altered epigenetic landscape and chromatin accessibility in DCs expressing dysfunctional β2-integrins

To further investigate β2-integrin–mediated gene regulation on a global scale in an unbiased manner, we performed H3K4me3 ChIP-seq using WT and β2-integrin KI DCs. As is shown in Fig. 4A, the mutation of the β2-integrin resulted in a global increase of methylation peak width compared with WT DCs. There was a significant overlap of H3K4me-tagged genes and genes that displayed increased expression in β2-integrin KI DCs (analyzed by RNA-seq analysis); 235 genes consistently displayed higher H3K4me3 around the transcription start site (TSS) in β2-integrin KI DCs than in WT DCs, and of these, 91 displayed increased RNA expression in β2-integrin KI cells (Fig. 4B). Among these were genes associated with DC maturation and migration, including Cd40, Cd80, Cd86, Il12b, Ccr7, and Fscn1 (Fig. 4CH). H3K4me3 active histone methylation marks are therefore increased on a broad range of genes associated with the mature, migratory phenotype of β2-integrin KI DCs.

An Ikaros-regulated gene network drives the phenotype of β2-integrin KI migratory DCs

An altered epigenetic landscape in cells is associated with altered chromatin accessibility, enabling gene-expression changes. To further investigate the role of β2-integrin–mediated chromatin regulation, we utilized ATAC-seq to map open chromatin regions in WT and β2-integrin KI DCs. Inactivating β2-integrins led to large-scale chromatin remodeling in DCs (Supplementary Fig. S8). Most ATAC-seq–enriched peaks in the β2-integrin KI DCs originated from promoter and intragenic regions of genes and they altogether scored around 70%, whereas intergenic peaks were also observed but at a lower portion (approximately 30% regions). Overall, ATAC-seq peaks around TSS and immediate downstream sequences were higher in the β2-integrin KI DCs (Supplementary Fig. S8).

We next sought to identify putative TF binding sites in the accessible chromatin regions that may be responsible for driving gene-expression changes associated with the mature, migratory phenotype of DCs with dysfunctional β2-integrins. Analysis of ATAC-seq peaks using TOBIAS identified several families of TFs enriched in β2-integrin KI DCs. The top 5 scoring TFs based on binding occupancy in β2-integrin KI DCs are shown in Table 1. Among these, Ikaros (encoded by the Ikzf1 gene) binding sites were the most frequent, PU.1 (encoded by the Spi1 gene) also scored highly. Ikaros is important for pDC development (23, 34), and overexpression of PU.1 directs stem and lymphoid progenitor cells toward the myeloid lineage (35). PU.1 is also essential for defining the cDC lineage (35), and Ikaros/PU.1 interactions are involved in regulation of myeloid genes (36). Ikaros-activated gene targets that overlap with β2-integrin KI upregulated gene-expression profiles are shown in Fig. 5A. Genes activated by Ikaros and upregulated in β2-integrin KI DCs included important genes involved in DC-mediated regulation of T-cell activation and DC migration, such as Ccr7, Cd86, and Il12b (Supplementary Table S1). Next, we compared PU.1-regulated genes with the RNA-seq upregulated genes in β2-integrin KI DCs. We found that 170 genes upregulated in the β2-integrin KI DCs were also activated by PU.1, and 100 of them were at least 2-fold upregulated (Fig. 5B). This list of 100 genes contained important genes involved in DC migratory and inflammatory phenotypes (Supplementary Table S1). Overall, our data support the presence of an Ikaros/PU.1-regulated gene network that controls the transcriptional changes in mature, migratory DCs deficient in β2-integrin adhesiveness.

As the Ikaros TF cluster was the top-scoring TOBIAS, we next investigated the role of Ikaros in regulating the β2-integrin KI DC phenotype. We found that lenalidomide (a cancer drug that leads to selective proteasome-induced degradation of Ikaros TFs; ref. 37) significantly reduced IL12 (p40) production in β2-integrin KI DCs (Fig. 5C). In addition, lenalidomide treatment led to significantly reduced β2-integrin KI DC migration speed in vitro (Fig. 5D). These results support the conclusion that Ikaros-family TFs play a major role in regulating the mature migratory phenotype of DCs downstream of β2-integrin adhesion.

A mechanically regulated Ikaros/NF-κB network in DCs

Our TOBIAS analysis identified an Ikaros TF network regulated by β2-integrins in DCs. However, RNA-seq analysis (Supplementary Table S2) indicated that Ikzf1 mRNA levels were not upregulated in β2 integrin KI cells (although mRNA levels of Ikzf3 and Ikzf4 were upregulated), raising the question of how loss of β2-integrin–mediated adhesion might activate the Ikaros TF network in β2-integrin KI DCs. TF network analysis showed connections between Ikaros and several other TFs including the NF-κB family members NFKB1, REL and RelA (Supplementary Fig. S9). NF-κB signaling has previously been implicated as critical for the homeostatic DC migratory phenotype, although the upstream regulators of this pathway have not been identified (22). We have previously shown LPS-induced NF-κB signaling to be upregulated in β2-integrin KI DCs (8). Furthermore, RelA works together with Ikaros to regulate late gene expression in LPS-stimulated macrophages (25), with Ikaros leading to opening up of chromatin allowing for increased RelA binding at the same sites (25). We therefore compared our β2-integrin KI DC RNA-seq data set with validated RelA targets (ChIP-seq) from TLR-matured DCs (38). The results show that genes upregulated in the β2-integrin KI data set were highly enriched for validated RelA binding sites (1076 sites; Fig. 5E). To further elucidate TF–target interactions, we used TRRUST (39). TRRUST analysis of overlapping Ikaros-activated and β2-integrin KI upregulated genes indicated that the top TF regulating this group of genes was NF-κB, with RelA in third place (Supplementary Table S3). TRRUST database also lists REL, RELA, and NFκB1 among the TFs that share common targets with Ikaros. We therefore investigated whether NF-κB signaling was regulated by β2-integrin adhesion. Indeed, we found that RelA nuclear localization was significantly upregulated in β2-integrin KI DCs and nonadherent WT DCs detached from their adhesive substrate (Fig. 5F and G). Furthermore, in WT DCs placed on soft matrices (Fig. 5H and I), RelA displayed significantly increased nuclear localization compared with cells on stiffer matrix, which correlates with increased H3K4me3 staining in cells cultured on softer matrices (Fig. 4). In addition, on very stiff surfaces, lamin A/C KO DCs displayed significantly increased RelA nuclear localization compared with WT DCs (Fig. 5I; 90 kPa gels), indicating that a loss of lamin A/C facilitates nuclear translocation of RelA particularly on stiffer substrates. Together, these results show that RelA translocation into the nucleus is regulated at least in part by mechanotransduction through the integrin/lamin axis, and identify RelA as one of the TFs in the Ikaros TF network that accounts for the β2-integrin KI DC phenotype.

Taken together, our results show that β2-integrin–mediated adhesion is functionally coupled to an Ikaros/RelA-regulated transcriptional network. This network regulates the epigenetic status, gene-expression landscape and mature, migratory phenotype of DCs, allowing DCs to respond to and functionally adapt to their tissue environment through β2-integrin–mediated cues.

Targeting integrin-regulated epigenetic changes increases DC migration and antitumor responses in vivo

Our result show that β2-integrins regulate epigenetic status and DC phenotype, and that blocking cell adhesion by cell detachment from culture plates leads to a transient increase in DC maturation, migration, and H3K4me3 histone methylation. We reasoned that this phenotype may be transient (as seen in the cell migration assay), as cell reattachment may reverse the phenotype. We further reasoned that it may be possible to stabilize the adhesion-regulated epigenetic profile of DCs by directly targeting histone methylation in detached WT DCs. Therefore, we targeted H3K4 methylation directly by inhibiting lysine-specific demethylase 1 (LSD1) using TCP. We expected that this treatment should increase H3K4me2/3 levels, based on previous studies (40). TCP treatment indeed resulted in increased H3K4me2/H3K4me3 histone methylation on Ccr7 (Fig. 6A), but not on other genes that were upregulated in β2-integrin KI DCs (e.g., Cd86). Correspondingly, Ccr7 mRNA level were significantly increased following combination treatment of TCP + suspension (Fig. 6B).

We next investigated whether TCP treatment stabilized the mature, migratory phenotype of detached WT DCs. TCP treatment alone did not result in increased costimulatory molecule expression; however, combining TCP treatment with suspension resulted in increased CD86 expression in DCs (Fig. 6C). Likewise, IL12 (p40) production was not increased in response to TCP treatment alone but was increased in response to combination treatment (Fig. 6D). However, TCP treatment alone was sufficient to increase the 3D migration speed of DCs in response to the CCR7 ligand CCL19 (Fig. 6E). Similar results were achieved with another histone demethylase inhibitor, IOX (Fig. 6F), further strengthening our findings. Furthermore, combination-treated CCR7−/− DCs displayed reduced migration speed (Fig. 6G), showing that migration was CCR7-dependent.

Finally, we investigated whether targeting the β2-integrin–regulated epigenetic changes in DCs would affect DC-mediated tumor rejection in vivo (Fig. 6H). Injection of TCP + suspension-treated WT DCs into tumor-bearing mice resulted in a reduction of tumor growth that was very similar to that mediated by β2-integrin KI DC-treated mice, with CD4+ T-cell recruitment into tumors, and reduced Treg numbers in tumors (Supplementary Fig. S10). Experiments with TCP + suspension-treated CCR7−/− DCs showed that this increased tumor rejection was completely dependent on the chemokine receptor, presumably because it is required for DC trafficking from the tumor vicinity to the lymph nodes to induce T-cell activation (Fig. 6I).

In conclusion, we have shown that β2-integrin–dependent adhesions connect to the nucleus via actin and the main nuclear lamina adaptor lamin A/C, thereby regulating the epigenetic landscape in DCs. Furthermore, integrins mechanically regulate RelA nuclear translocation, an Ikaros/NF-kB TF network, chromatin accessibility, and the gene-expression profile of DCs. β2-integrin–mediated mechanosignaling thereby restricts excessive immune responses, primarily by suppressing the mature, migratory phenotype of DCs and DC-mediated T-cell activation, resulting in attenuated antitumor responses in vivo.

Many cell types respond not only to biochemical cues but also to mechanical cues, such as tissue stiffness, shear flow and stretching, and these (and other) cues can influence gene-expression programs and cell fate. Immune cells such as DCs constantly traffic between tissues (peripheral tissues, lymphatics, lymph nodes) with different mechanical properties, which could potentially influence their functions. However, it is currently poorly understood how cell adhesion and mechanical cues sensed by distinct tissue leukocytes, and in particular DCs, shape immune cell gene-expression profiles and, thus, their cellular functions, although it has been previously reported that disrupting physical interactions of DCs can strongly induce their activation (41).

Integrins play major roles in mechanotransduction because they mediate adhesion to other cells and extracellular ligands, sense stiffness, and transmit mechanical information into cells through signaling pathways, but also through direct mechanical links through their intracellular tails, via actin to the cell nucleus and chromatin (through the LINC (Sun/nesprin) complex and nuclear lamins (42)). This is linked to gene tethering to the nuclear envelope, which correlates with tissue-specific gene suppression (43). When cells such as fibroblasts are cultured on stiff matrices, they display increased integrin-mediated cell adhesion, and the tension is transmitted through the integrin cytoplasmic domains (“tails”), through the cytoskeleton to the nucleus, where it can affect directly on chromatin organization. In contrast, cells on softer matrices display reduced adhesion and downregulation of lamin A/C, which can also have an impact on chromatin structure. β2-integrins, which are adhesion receptors abundantly expressed by all immune cells, therefore potentially have the ability to mechanically connect the extracellular environment of the cell to the nucleus to directly influence gene expression through several mechanisms, although these processes have not been extensively studied in immune cells.

In addition to their important roles in immune cell trafficking and activation, β2-integrins have been reported to restrict myeloid cell signaling and responses in many different settings. β2-integrins can inhibit TLR signaling in macrophages, promote tolerance, and suppress inflammation, consistent with homeostatic and regulatory roles of these receptors in immunity and inflammation (26, 44–50). Furthermore, we have previously shown that β2-integrins play a key role in regulating DC gene expression, and that β2-integrins inhibit DC maturation, migration to lymph nodes, and DC-mediated Th1 polarization in vivo (8, 26). In addition, we have discovered a β2-integrin–actin–MRTF-A/SRF pathway that regulates cell adhesion, traction force generation, and cytoskeletal gene expression in DCs (51). However, this pathway regulates only a small subset of the genes affected by the β2-integrin KI mutation and is not involved in regulating DC 3D migration. Therefore, exactly how β2-integrin–mediated adhesion restricts immune cell function has remained poorly understood.

Here we show that β2-integrin–mediated adhesion to the extracellular environment regulates chromatin structure and the epigenetic landscape of DCs (e.g., histone methylation), and therefore DC gene transcription, maturation, migration, and function in vitro and in vivo. These conclusions are based on our observations that (i) disrupting DC adhesion by transferring them to suspension; (ii) interrupting the cytoskeletal associations of their β2-integrins; (iii) interfering with their actin cytoskeleton; (iv) deleting lamin A/C; and (v) culturing DCs on softer matrices, drive similar global changes in histone methylation, providing evidence of mechanical regulation of the DC epigenetic landscape by β2-integrin–mediated adhesion.

We found that open chromatin regions in β2-integrin KI DCs are enriched with binding sites for Ikaros and other TFs such as PU.1. Our motif discovery analysis (known and de novo) identified Ikaros as the most significant TF in β2-integrin KI-enriched ATAC-seq peaks, suggesting that these TFs directly contribute to the mature migratory phenotype of these DCs. Indeed, our functional assays identified Ikaros as a main regulator of the β2-integrin KI DC phenotype. Ikaros/RelA are known to cooperate in regulation of gene transcription in myeloid cells (25), and our results further implicate RelA as an β2-integrin–regulated factor in this Ikaros transcriptional network. We therefore hypothesize that β2-integrin–mediated regulation of NF-κB signaling provides at least one link between integrins and the Ikaros gene network that regulates DC phenotype and function. However, other mechanisms involved in β2-integrin–mediated regulation of the Ikaros gene network in these cells should also be considered, for example, direct regulation of Ikaros complex through β2-integrin–regulated signaling pathways.

Based on our results, we propose that β2-integrin–mediated adhesion to the extracellular environment restricts DC programming to the mature, migratory phenotype by regulating the DC epigenetic/chromatin landscape. β2-Integrin–mediated adhesion regulates DC programming through β2-integrin/actin/chromatin cross-talk, affecting histone methylation status and chromatin accessibility. In addition, β2-integrins regulate TF networks including Ikaros and RelA, further affecting gene transcription programs in these cells. We propose that β2-integrin–mediated regulation can be used to fine-tune gene-expression programs in these cells in response to the mechanical properties of the environment, in particular the stiffness of the interstitial space, which is controlled by collagen composition, density, and crosslinking (52). Furthermore, after detaching from their environment in peripheral tissues, DCs interact with other cells and structures, such as lymphatic endothelial cells, and travel through the lymphatic system to lymph nodes. Such mechanical interactions could potentially confer DCs with a “mechanical memory” of the environments they engage and travel through. This memory of their integrated interactions with the extracellular matrix and with other cells, together with the biochemical cues they encounter, will determine DC location, function, and survival.

However, in addition to conferring cells with an ability to adhere to the environment and sense extracellular mechanical properties, β2-integrins have the ability to be switched on, when adhesion is needed, and switched off, when it is not. Therefore, we propose that β2-integrins may also provide cells with a mechanism of “erasing” their mechanical memory, for example, when a more pressing, overriding chemotactic or other stimulatory signal is encountered (such as during bacterial infection). Indeed, DCs downregulate β2-integrins and other cell adhesion receptors during maturation in both homeostatic and inflammatory settings (53, 54), and the loss of β2-integrin adhesion in these cells contributes to the transcriptional changes associated with these switches. These large-scale transcriptional changes associated with epigenetic programming of the mature phenotype of DCs, dramatically enhance T-cell activation and tumor infiltration of activated T cells in vivo.

A striking reflection of this concept of β2-integrin–dependent DC memory is the phenotype of adhesion-deficient DCs, which become reprogrammed to the mature, migratory phenotype, and induce better DC-mediated tumor rejection in two different melanoma models. Furthermore, by pharmacologically targeting this β2-integrin–regulated epigenetic reprograming in DCs, we could confer enhanced tumor-rejecting capabilities to WT DCs. Our study demonstrates that a central β2-integrin–mediated mechanism can be externally targeted to optimize DC- and T-cell mediated tumor control in vivo, although its relevance in human DCs requires further studies. Nevertheless, these and other approaches to directly target β2-integrins and their cytoskeletal adaptors may be potentially promising means to optimize immunotherapy approaches in the future.

C. Guenther reports grants from the Magnus Ehrnrooth Foundation during the conduct of the study. S.C. Fagerholm reports grants from Academy of Finland, University of Helsinki, Magnus Ehrnrooth Foundation, E-Rare/Academy of Finland, Liv och Hälsa Foundation, Sigrid Juselius Foundation, and Swedish Cultural Foundation during the conduct of the study. No disclosures were reported by the other authors.

C. Guenther: Data curation, formal analysis, funding acquisition, investigation, visualization, writing–original draft, writing–review and editing. I. Faisal: Data curation, formal analysis, investigation, methodology, writing–original draft, writing–review and editing. M. Fusciello: Formal analysis, investigation, writing–review and editing. M. Sokolova: Formal analysis, investigation. H. Harjunpää: Formal analysis, investigation, writing–review and editing. M. Ilander: Formal analysis, investigation. R. Tallberg: Investigation. M.K. Vartiainen: Funding acquisition, project administration, writing–review and editing. R. Alon: Resources, writing–review and editing. J.-M. Gonzalez-Granado: Resources, writing–review and editing. V. Cerullo: Conceptualization, resources, supervision, writing–review and editing. S.C. Fagerholm: Conceptualization, resources, supervision, funding acquisition, validation, project administration, writing–review and editing.

This study was funded by the Academy of Finland, by project grants to S.C. Fagerholm, as well as under the framework of E-Rare-3, the Sigrid Juselius Foundation, the University of Helsinki (HiLIFE), Liv och Hälsa Foundation (all to S.C. Fagerholm), and the Magnus Ehrnrooth Foundation (to C. Guenther and S.C. Fagerholm). ERA-Net for Research on Rare Diseases E-RARE (the LADOMICS consortium) covers funding for both R. Alon and S.C. Fagerholm. Furthermore, J. Gonzalez-Granado is funded by Instituto de Salud Carlos III (ISCIII) (PI17/01395; PI20/00306; SNSI3 program) with cofunding from the European Regional Development Fund (ERDF) “A way to build Europe.” V. Cerullo acknowledges the European Research Council under the Horizon 2020 framework, ERC-Consolidator Grant (agreement no. 681219), Jane and Aatos Erkko Foundation (project no. 4705796), HiLIFE Fellow (project no. 797011004), Finnish Cancer Foundations (project no. 4706116), and Magnus Ehrnrooth Foundation (project no. 4706235). M.K. Vartiainen is funded by Academy of Finland, Jane and Aatos Erkko, Sigrid Juselius, and Finnish Cancer foundations, as well as Helsinki Institute of Life Science. Virginia Zorita [Centro Nacional de Investigaciones Cardiovasculares Carlos III (CNIC), 28029 Madrid, Spain] is acknowledged for her help with tissue processing. Sara W. Feigelson (Weizmann Institute of Science, Israel) is acknowledged for her help with shipments of the HL-60 cell variants. Reinhold Förster is acknowledged for kindly providing bone marrow from CCR7−/− mice.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Anguille
S
,
Smits
EL
,
Lion
E
,
van Tendeloo
VF
,
Berneman
ZN
. 
Clinical use of dendritic cells for
cancer therapy
.
Lancet Oncol
2014
;
15
:
e257
67
.
2.
Helft
J
,
Böttcher
J
,
Chakravarty
P
,
Zelenay
S
,
Huotari
J
,
Schraml
BU
, et al
GM-CSF mouse bone marrow cultures comprise a heterogeneous population of CD11c+MHCII+ macrophages and dendritic cells
.
Immunity
2015
;
42
:
1197
211
.
3.
Sallusto
F
,
Lanzavecchia
A
. 
Efficient presentation of soluble antigen by cultured human dendritic cells is maintained by granulocyte/macrophage colony-stimulating factor plus interleukin 4 and downregulated by tumor necrosis factor alpha
.
J Exp Med
1994
;
179
:
1109
18
.
4.
Huber
A
,
Dammeijer
F
,
Aerts
JGJV
,
Vroman
H
. 
Current state of dendritic cell-based immunotherapy: opportunities for in vitro antigen loading of different DC subsets?
Front Immunol
2018
;
9
:
2804
.
5.
Faure-André
G
,
Vargas
P
,
Yuseff
M-I
,
Heuzé
M
,
Diaz
J
,
Lankar
D
, et al
Regulation of dendritic cell migration by CD74, the MHC class II-associated invariant chain
.
Science
2008
;
322
:
1705
10
.
6.
Ufer
F
,
Vargas
P
,
Engler
JB
,
Tintelnot
J
,
Schattling
B
,
Winkler
H
, et al
Arc/Arg3.1 governs inflammatory dendritic cell migration from the skin and thereby controls T cell activation
.
Sci Immunol
2016
;
1
:
eaaf8665
.
7.
Yamakita
Y
,
Matsumura
F
,
Lipscomb
MW
,
Chou
P-c
,
Werlen
G
,
Burkhardt
JK
, et al
Fascin1 promotes cell migration of mature dendritic cells
.
J Immunol
2011
;
186
:
2850
9
.
8.
Morrison
VL
,
James
MJ
,
Grzes
K
,
Cook
P
,
Glass
DG
,
Savinko
T
, et al
Loss of beta2-integrin-mediated cytoskeletal linkage reprograms dendritic cells to a mature migratory phenotype
.
Nat Commun
2014
;
5
:
5359
.
9.
Morrison
VL
,
MacPherson
M
,
Savinko
T
,
Lek
HS
,
Prescott
A
,
Fagerholm
SC
. 
The β2 integrin–kindlin-3 interaction is essential for T-cell homing but dispensable for T-cell activation in vivo
.
Blood
2013
;
122
:
1428
36
.
10.
Yadav
SK
,
Feigelson
SW
,
Roncato
F
,
Antman-Passig
M
,
Shefi
O
,
Lammerding
J
, et al
Elevated nuclear lamin A is permissive for granulocyte transendothelial migration but not for motility through collagen I barriers
.
J Leukoc Biol
2018
;
104
:
239
51
.
11.
Guo
Q
,
Zhang
L
,
Li
F
,
Jiang
G
. 
The plasticity and potential of leukemia cell lines to differentiate into dendritic cells
.
Oncol Lett
2012
;
4
:
595
600
.
12.
Li
Q
,
Ozer
H
,
Lindner
I
,
Lee
KP
,
Kharfan-Dabaja
MA
. 
Protein kinase C blockade inhibits differentiation of myeloid blasts into dendritic cells by calcium ionophore A23187
Int J Hematol
2005
;
81
:
131
7
.
13.
Dembo
M
,
Wang
YL
. 
Stresses at the cell-to-substrate interface during locomotion of fibroblasts
.
Biophys J
1999
;
76
:
2307
16
.
14.
Abashidze
A
,
Gold
V
,
Anavi
Y
,
Greenspan
H
,
Weil
M
. 
Involvement of IKAP in peripheral target innervation and in specific JNK and NGF signaling in developing PNS neurons
.
PLoS One
2014
;
9
:
e113428
.
15.
Kallio
MA
,
Tuimala
JT
,
Hupponen
T
,
Klemelä
P
,
Gentile
M
,
Scheinin
I
, et al
Chipster: user-friendly analysis software for microarray and other high-throughput data
.
BMC Genomics
2011
;
12
:
507
.
16.
Robinson
JT
,
Thorvaldsdóttir
H
,
Winckler
W
,
Guttman
M
,
Lander
ES
,
Getz
G
, et al
Integrative Genomics Viewer
.
Nat Biotechnol
2011
;
29
:
24
6
.
17.
Lerdrup
M
,
Johansen
JV
,
Agrawal-Singh
S
,
Hansen
K
. 
An interactive environment for agile analysis and visualization of ChIP-sequencing data
.
Nat Struct Mol Biol
2016
;
23
:
349
57
.
18.
Corces
MR
,
Trevino
AE
,
Hamilton
EG
,
Greenside
PG
,
Sinnott-Armstrong
NA
,
Vesuna
S
, et al
An improved ATAC-seq protocol reduces background and enables interrogation of frozen tissues
.
Nat Methods
2017
;
14
:
959
62
.
19.
Buenrostro
J
,
Wu
B
,
Chang
H
,
Greenleaf
W
. 
ATAC-seq: a method for assaying chromatin accessibility genome-wide
.
Curr Protoc Mol Biol
2015
;
109
:
21.29.1
21.29.9
.
20.
Buenrostro
JD
,
Giresi
PG
,
Zaba
LC
,
Chang
HY
,
Greenleaf
WJ
. 
Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position
.
Nat Methods
2013
;
10
:
1213
8
.
21.
Bentsen
M
,
Goymann
P
,
Schultheis
H
,
Klee
K
,
Petrova
A
,
Wiegandt
R
, et al
ATAC-seq footprinting unravels kinetics of transcription factor binding during zygotic genome activation
.
Nat Commun
2020
;
11
:
4267
.
22.
Baratin
M
,
Foray
C
,
Demaria
O
,
Habbeddine
M
,
Pollet
E
,
Maurizio
J
, et al
Homeostatic NF-κB signaling in steady-state migratory dendritic cells regulates immune homeostasis and tolerance
.
Immunity
2015
;
42
:
627
39
.
23.
Boutboul
D
,
Kuehn
HS
,
Wyngaert
ZV
,
Niemela
JE
,
Callebaut
I
,
Stoddard
J
, et al
Dominant-negative IKZF1 mutations cause a T, B, and myeloid cell combined immunodeficiency
J Clin Invest
2018
;
128
:
3071
87
.
24.
Zhou
Y
,
Zhou
B
,
Pache
L
,
Chang
M
,
Khodabakhshi
AH
,
Tanaseichuk
O
, et al
Metascape provides a biologist-oriented resource for the analysis of systems-level datasets
.
Nat Commun
2019
;
10
:
1523
.
25.
Oh
K-S
,
Gottschalk
RA
,
Lounsbury
NW
,
Sun
J
,
Dorrington
MG
,
Baek
S
, et al
Dual roles for ikaros in regulation of macrophage chromatin state and inflammatory gene expression
.
J Immmunol
2018
;
201
:
757
71
.
26.
Savinko
TS
,
Morrison
VL
,
Uotila
LM
,
Wolff
CHJ
,
Alenius
HT
,
Fagerholm
SC
. 
Functional Beta2-integrins restrict skin inflammation in vivo
.
J Invest Dermatol
2015
;
135
:
2249
57
.
27.
Bernstein
BE
,
Mikkelsen
TS
,
Xie
X
,
Kamal
M
,
Huebert
DJ
,
Cuff
J
, et al
A bivalent chromatin structure marks key developmental genes in embryonic stem cells
.
Cell
2006
;
125
:
315
26
.
28.
Solovei
I
,
Wang
AS
,
Thanisch
K
,
Schmidt
CS
,
Krebs
S
,
Zwerger
M
, et al
LBR and lamin A/C sequentially tether peripheral heterochromatin and inversely regulate differentiation
.
Cell
2013
;
152
:
584
98
.
29.
Gawden-Bone
C
,
West
MA
,
Morrison
VL
,
Edgar
AJ
,
Millan
SJM
,
Dill
BD
, et al
A crucial role for beta2 integrins in podosome formation, dynamics and Toll-like-receptor-signaled disassembly in dendritic cells
.
J Cell Sci
2014
;
127
:
4213
24
.
30.
Harr
JC
,
Gonzalez-Sandoval
A
,
Gasser
SM
. 
Histones and histone modifications in perinuclear chromatin anchoring: from yeast to man
.
EMBO Rep
2016
;
17
:
139
55
.
31.
van Steensel
B
,
Belmont
AS
. 
Lamina-associated domains: links with chromosome architecture, heterochromatin, and gene repression
.
Cell
2017
;
169
:
780
91
.
32.
Swift
J
,
Ivanovska
IL
,
Buxboim
A
,
Harada
T
,
Dingal
PCDP
,
Pinter
J
, et al
Nuclear lamin-A scales with tissue stiffness and enhances matrix-directed differentiation
.
Science
2013
;
341
:
1240104
.
33.
Borroni
AP
,
Emanuelli
A
,
Shah
PA
,
Ilić
N
,
Apel-Sarid
L
,
Paolini
B
, et al
Smurf2 regulates stability and the autophagic–lysosomal turnover of lamin A and its disease-associated form progerin
.
Aging Cell
2018
;
17
:
e12732
.
34.
Cytlak
U
,
Resteu
A
,
Bogaert
D
,
Kuehn
HS
,
Altmann
T
,
Gennery
A
, et al
Ikaros family zinc finger 1 regulates dendritic cell development and function in humans
.
Nat Commun
2018
;
9
:
1239
.
35.
Chopin
M
,
Lun
AT
,
Zhan
Y
,
Schreuder
J
,
Coughlan
H
,
D'Amico
A
, et al
Transcription factor PU.1 promotes conventional dendritic cell identity and function via induction of transcriptional regulator DC-SCRIPT
.
Immunity
2019
;
50
:
77
90
.
36.
Zarnegar
MA
,
Rothenberg
EV
. 
Ikaros represses and activates PU.1 cell-type-specifically through the multifunctional Sfpi1 URE and a myeloid specific enhancer
Oncogene
2013
;
31
:
4647
54
.
37.
Krönke
J
,
Udeshi
ND
,
Narla
A
,
Grauman
P
,
Hurst
SN
,
McConkey
M
, et al
Lenalidomide causes selective degradation of IKZF1 and IKZF3 in multiple myeloma cells
Science
2014
;
343
:
301
5
.
38.
Garber
M
,
Yosef
N
,
Goren
A
,
Raychowdhury
R
,
Thielke
A
,
Guttman
M
, et al
A high-throughput chromatin immunoprecipitation approach reveals principles of dynamic gene regulation in mammals
.
Mol Cell
2012
;
47
:
810
22
.
39.
Han
H
,
Cho
JW
,
Lee
S
,
Yun
A
,
Kim
H
,
Bae
D
, et al
TRRUST v2: an expanded reference database of human and mouse transcriptional regulatory interactions
.
Nucleic Acids Res
2018
;
46
:
D380
D6
.
40.
Kidder
BL
,
Hu
G
,
Zhao
K
. 
KDM5B focuses H3K4 methylation near promoters and enhancers during embryonic stem cell self-renewal and differentiation
.
Genome Biol
2014
;
15
:
R32
.
41.
Gallucci
S
,
Lolkema
M
,
Matzinger
P
. 
Natural adjuvants: endogenous activators of dendritic cells
.
Nat Med
1999
;
5
:
1249
55
.
42.
Cho
S
,
Irianto
J
,
Discher
DE
. 
Mechanosensing by the nucleus: from pathways to scaling relationships
.
J Cell Biol
2017
;
216
:
305
15
.
43.
Naetar
N
,
Ferraioli
S
,
Foisner
R
. 
Lamins in the nuclear interior - life outside the lamina
.
J Cell Sci
2017
;
130
:
2087
96
.
44.
Varga
G
,
Balkow
S
,
Wild
MK
,
Stadtbaeumer
A
,
Krummen
M
,
Rothoeft
T
, et al
Active MAC-1 (CD11b/CD18) on DCs inhibits full T-cell activation
.
Blood
2007
;
109
:
661
9
.
45.
Ehirchiou
D
,
Xiong
Y
,
Xu
G
,
Chen
W
,
Shi
Y
,
Zhang
L
. 
CD11b facilitates the development of peripheral tolerance by suppressing Th17 differentiation
.
J Exp Med
2007
;
204
:
1519
24
.
46.
Han
C
,
Jin
J
,
Xu
S
,
Liu
H
,
Li
N
,
Cao
X
. 
Integrin CD11b negatively regulates TLR-triggered inflammatory responses by activating Syk and promoting degradation of MyD88 and TRIF via Cbl-b
.
Nat Immunol
2010
;
11
:
734
42
.
47.
Leon
F
,
Contractor
N
,
Fuss
I
,
Marth
T
,
Lahey
E
,
Iwaki
S
, et al
Antibodies to complement receptor 3 treat established inflammation in murine models of colitis and a novel model of psoriasiform dermatitis
.
J Immunol
2006
;
177
:
6974
82
.
48.
Haasken
S
,
Auger
JL
,
Binstadt
BA
. 
Absence of beta2 integrins impairs regulatory T cells and exacerbates CD4+ T cell-dependent autoimmune carditis
.
J Immunol
2011
;
187
:
2702
10
.
49.
D'Agata
ID
,
Paradis
K
,
Chad
Z
,
Bonny
Y
,
Seidman
E
. 
Leucocyte adhesion deficiency presenting as a chronic ileocolitis
.
Gut
1996
;
39
:
605
8
.
50.
Cohen
SJ
,
Gurevich
I
,
Feigelson
SW
,
Petrovich
E
,
Moser
M
,
Shakhar
G
, et al
The integrin co-activator Kindlin-3 is not required for lymphocyte diapedesis
.
Blood
2013
;
122
:
2609
17
.
51.
Guenther
C
,
Faisal
I
,
Uotila
LM
,
Asens
ML
,
Harjunpää
H
,
Savinko
T
, et al
A β2-integrin/MRTF-A/SRF pathway regulates dendritic cell gene expression, adhesion, and traction force generation
.
Front Immunol
2019
;
10
:
1138
.
52.
Feit
H
,
Kawai
M
,
Mostafapour
S
. 
The role of collagen crosslinking in the increased stiffness of avian dystrophic muscle
.
Muscle Nerve
1989
;
12
:
486
92
.
53.
Winzler
C
,
Rovere
P
,
Rescigno
M
,
Granucci
F
,
Penna
G
,
Adorini
L
, et al
Maturation stages of mouse dendritic cells in growth factor–dependent long-term cultures
.
J Exp Med
1997
;
185
:
317
28
.
54.
Miller
JC
,
Brown
BD
,
Shay
T
,
Gautier
EL
,
Jojic
V
,
Cohain
A
, et al
Deciphering the transcriptional network of the dendritic cell lineage
.
Nat Immunol
2012
;
13
:
888
99
.