Targeting Talin1 to reprogram dendritic cell activation landscapes: A mechanistically grounded cell therapy for multiple sclerosis model

Abstract

This study investigated the role of Talin1 in regulating dendritic cell (DC) activation and the neuroprotective benefits of Talin1-knockdown DCs in experimental autoimmune encephalomyelitis (EAE), an animal model of multiple sclerosis (MS). Bone marrow-derived DCs (BMDCs) were transduced with shTalin1 lentiviral vectors in vitro. Their morphological and biochemical profiles, surface molecules expression, cytokines production, capacity to induce T cell responses, as well as regulatory mechanisms, were comprehensively assessed. In vivo, Talin1-knockdown BMDCs loaded with the MOG_35-55 peptide were administered preclinically and therapeutically to EAE mice, with subsequent evaluation of EAE development, inflammatory infiltration, demyelination, and Th/Treg responses. Results demonstrated that Talin1 knockdown significantly inhibited the activation of BMDCs, as evidenced by decreased expression of surface molecules (MHCII, CD80, CD86) and pro-inflammatory cytokines (IL-1β, IL-6, TNF-α), increased expression of anti-inflammatory cytokine (IL-10), differential morphology, ultrastructure, and biochemical characteristics, accompanied by limited ability to stimulate CD4+T cell proliferation and polarization toward Th1 and Th17 subsets. Moreover, RNA-sequencing revealed downregulation of immune/inflammation-related processes and pathways in Talin1-knockdown BMDCs. Mechanistically, inhibition of the TLR4/MyD88/NF-κB pathway in BMDCs contributed to these effects. In vivo, Talin1-knockdown BMDCs significantly delayed the pathogenesis & progression of EAE, alleviated their neurological deficits and pathology, decreased Th1 and Th17 lineage levels, and boosted the abundance of Treg cells. Collectively, these findings indicate that Talin1 orchestrates BMDCs activation, and Talin1-knockdown BMDCs protect against EAE by rebalancing Th1/Th17/Treg dynamics, suggesting a potential approach for the development of precision therapies for MS and other autoimmune disorders.

Introduction

Dendritic cells (DCs), as classical antigen-presenting cells, bridge innate and adaptive immunity, with their phenotype and function closely related to their activation/maturation status [1]. Mature or activated DCs possess a robust capacity to stimulate the activation, proliferation, and differentiation of T cells through the overexpression of a spectrum of surface molecules, including histocompatibility complex class I/II (MHCI/II) and co-stimulatory molecules [2]. In addition to eliciting immune responses, DCs can induce peripheral T-cell tolerance, a crucial mechanism preventing autoimmunity upon encountering self-antigens [3]. These tolerogenic DCs (TolDCs) suppress immune activation by exhibiting low expression of MHC, costimulatory molecules and pro-inflammatory cytokines, while secreting high concentrations of anti-inflammatory cytokines to modify the equilibrium between T helper (Th)1/Th17 ​cells and regulatory T (Treg) cells, positioning TolDCs as a potentially promising cellular therapy for the management of autoimmune diseases and allograft rejection [[3], [4], [5]].

Multiple sclerosis (MS) is a demyelinating disorder of the central nervous system (CNS), characterized by aberrant activation of autologous CD4+T cells, which disrupts the homeostasis between Th1/Th17 and Treg cells, acting as a pivotal player in its pathogenesis [6,7]. Given their established properties in regulating T cell activation and proliferation, DCs are widely recognized as being associated with the pathogenesis and progression of the disease. Correspondingly, an increased number of specialized DCs, which displayed a distinct maturation phenotype and readily promoted the differentiation into the Th17 subset from naïve CD4+T cells, were identified in the cerebrospinal fluid (CSF) of MS patients [8,9]. Conversely, multiple biologically or pharmacologically induced TolDCs (loaded with myelin peptides) had been reported to be neuroprotective in experimental autoimmune encephalomyelitis (EAE) mice, a classical animal model of MS, by modulating peripheral and CNS immunological responses against myelin antigens, thereby providing novel perspectives for efficient and precise treatment for this disorder [[10], [11], [12], [13]]. Therefore, to optimize the application of TolDCs as reliable cellular therapies in MS management, a comprehensive understanding of the underlying factors and intrinsic mechanisms that control the activation of DCs is essential.

Talin1 is a major adaptor member associated with integrin signaling and is indispensable for effective cell migration [14]. Accordingly, a deficiency of Talin1 in immune cells would result in their impaired ability to home to lymph nodes or infiltrate into lesions [15,16]. Notably, our previous research reported that serum soluble Talin1 (sTalin1) concentrations were significantly elevated in MS patients compared with controls, with significantly higher levels observed during acute attacks than in remission [17]. Moreover, sTalin1 levels were highly correlated with the sustained accumulation of disability after MS relapse [17]. Importantly, a recent study reported that Talin1 plays a major role in forming pre-assembled Toll-like receptor (TLR) complexes by direct interaction with myeloid differentiation primary response protein 88 (MyD88), which is essential for rapid and efficient activation of MyD88-dependent TLR signaling pathways (a classical signaling pathway for DC activation) [18]. And Talin1-deficiency caused disrupted inflammatory cytokines secretion [18]. These demonstrations collectively suggest that Talin1 may represent a promising therapeutic for the management of MS and other autoimmune disorders via modulating immune responses associated with DCs.

In the current study, we systematically investigated the effects of Talin1 on multiple phenotypes of DCs, including their morphology, ultrastructure, surface molecules, cytokine secretion, biochemical profiles, and transcriptomic signature, as well as their ability to stimulate CD4⁺ T cell responses in vitro. On this basis, bone marrow-derived dendritic cells (BMDCs) with Talin1 knockdown and loaded with MOG_35-55 peptide were administered to the EAE mice to further investigate their potential immunomodulation roles and possible mechanisms in mitigating EAE. These findings may contribute to the development of highly targeted and efficient therapeutics directed at Talin1 for the management of MS, as well as other autoimmune diseases.

Methods

Animals

Female C57BL/6 and BALB/c mice (6–8 weeks old, 16–20g) were procured from Vital River Laboratories, Beijing, China. All mice were housed in the Kangtai Medical Laboratory Co. (Hebei, China) animal facility under controlled conditions: a room temperature (RT) of 22 ​± ​2 ​°C, a relative humidity of 40–70 ​%, a 12-h light/12-h dark cycle, and ad libitum access to food and water. The facility was maintained free from specific pathogens. All experimental procedures and protocols adhered to the guidelines of the Animal Care and Use Committee of KangTai Medical Laboratory (Hebei, China, Approval No. MDL2024-04-28-01, MDL2025-06-20-01).

Generation and activation of BMDCs

Murine DCs were generated from bone marrow precursor cells as previously reported [19]. Briefly, the ends of the femur and tibia were aseptically cut, and the marrow was flushed out. Small bone fragments and debris were removed through nylon mesh. Bone marrow-derived mononuclear cells (BMMCs) were obtained by lysing and removing erythrocytes from the above cell suspensions. BMMCs were incubated in Petri dishes at a density of 2 ​× ​10⁶ ​cells/ml in RPMI 1640 medium complemented with 100 IU/mL penicillin and streptomycin (Sigma), 2 ​mM glutamine (Sigma, St. Louis, MO, USA), 10 ​% fetal bovine serum (FBS, Gibco, Invitrogen, Carlsbad, CA, USA), 50 ​ng/mL recombinant mouse granulocyte macrophage colony-stimulating factor (rmGM-CSF; PeproTech, Rocky Hill, NJ, USA), and 50 ​ng/mL rmIL-4 (PeproTech) at 37 ​°C 5 ​% CO₂ to generate BMDCs. Cultures were refreshed on days 2, 4, and 6 by replenishing half the volume of fresh medium complemented with rmGM-CSF and rmIL-4. On day 8, BMDCs were isolated via negative selection using a CD11c Cell Separation Kit according to the manufacturer's instructions (Miltenyi Biotech, Germany). If necessary, 1 ​μg/mL lipopolysaccharide (LPS, Sigma) was incorporated into the culture for indicated times to induce BMDCs activation or maturation.

Talin1 knockdown lentiviral construction and transduction to BMDCs

Lentiviral vectors for shTalin1(shT) and shNC (shN) were purchased from Genechem (Shanghai, China). For lentiviral transduction, BMDCs were seeded in 6-well plates at 8 ​× ​10⁵ ​cells/well and pre-cultured for 24 ​h at 37 ​°C with 5 ​% CO₂. Cells were then transduced with lentiviral vectors at an MOI of 100 in medium supplemented with 2 ​μg/ml Polybrene (Sigma). Transduction was performed at 32 ​°C for 6 ​h, followed by two washes with PBS to remove residual vectors/Polybrene. Washed cells were replenished with fresh complete medium and further cultured for 48 ​h at 37 ​°C/5 ​% CO₂ prior to phenotypic and functional assays. Relevant shRNA (shN, shT1-3) sequence information and verification of Talin1 knockdown by western blotting assays are provided in Supplemental Fig. 1, which confirmed that Talin1 protein was significantly downregulated by shT2.

Quantitative real-time PCR (qRT-PCR) analysis

Total RNA was extracted from BMDCs using Trizol reagent (CWbio, China). Complementary DNA (cDNA) synthesis was performed with 1 ​μg RNA per reaction using the HiScript III 1st Strand cDNA Synthesis Kit (Vazyme, China) according to the manufacturer's instructions. Quantitative real-time PCR (qRT-PCR) was conducted on a QuantStudio 7 Flex system (Thermo Fisher Scientific) using Ultra SYBR Mixture (CWbio) and gene-specific primers (Supplementary Table 1). Amplification conditions consisted of an initial denaturation at 95 ​°C for 10 ​min, followed by 45 cycles of 95 ​°C for 15s and 60 ​°C for 60s. Melting curve analysis (60 ​°C–95 ​°C at 0.3 ​°C/s increment) confirmed amplicon specificity. All reactions were performed in triplicate. Relative expression of IL-1β, IL-6, IL-10, and TNF-α was calculated by the 2^−ΔΔCT method, normalized to GAPDH as the endogenous control [20].

Flow cytometry

BMDCs (2 ​× ​10⁶/well) from different experimental groups were treated in triplicate with LPS (1 ​μg/mL) or dimethyl sulfoxide (DMSO) for 12 ​h. Then, samples were incubated with BD Pharmingen Mouse PerCP-Cy5.5-anti-CD11c, PE-anti-CD80, or FITC-anti-CD86, or APC-anti-MHCII, or isotype controls (BD Biosciences) for 30min (protected from light, RT). After that, cells were rinsed three times with PBS and tested by flow cytometry (BD FACS CaliburII, Biosciences, Franklin Lakes, NJ, USA).

BMDCs from different groups were co-incubated with CD4+T cells for 72 ​h (mentioned in the “Mixed lymphocyte reaction” section). For the proliferation analysis, CD4+T cells (10⁶/ml) were pre-stained with 10 ​μM carboxyfluorescein succinimidyl ester (CFSE, BD Biosciences) at 37 ​°C for 10 ​min, and the reaction was terminated by incubating with pre-cooled 1640 medium at 4 ​°C for 5 ​min. After 72 ​h of co-culture with DCs, non-adherent CFSE-labeled CD4+T cells were reacted by incubating with FITC-anti-CD4 (Thermo Scientific) protected from light, RT for 30min and detected by flow cytometry (BD FACS CaliburII, BD Biosciences). For differentiation into Th cell subsets, non-adherent CD4+T cells were collected and stained with FITC-anti-CD4 or PerCP-Cy5.5-anti-CD4 (Thermo Scientific) at RT for 30min, protected from light. Next, after fixation and permeabilization using a BD Cytofix/Cytoperm kit (BD Biosciences), intracellular staining was performed with PE-anti–IFN–γ, PerCP-Cy5.5-anti-IL-4, or APC-anti-IL-17 (Thermo Scientific) at RT for 1 ​h. The frequency of Th1 (CD4+IFN-γ+), Th2 (CD4+IL-4+) and Th17 (CD4+IL-17+) cells in the population of CD4+T cells was determined with the application of flow cytometry (BD FACS CaliburII, BD Biosciences). For Treg cells differentiation assays, cells were stained with FITC-anti-CD4 or PerCP-Cy5.5-anti-CD4 and PE-anti-CD25 or FITC-anti-CD25, at RT for 30min, protected from light. Then, cells were fixed and permeabilized using the Foxp3 Cytofix/Cytoperm kit (Thermo Scientific). After that, cells were further incubated with APC-anti-Foxp3 or PE-anti-Foxp3 (Thermo Scientific) at RT for 30min, protected from light, and the levels of Treg cells (CD4+CD25+ Foxp3+) were quantified using flow cytometry (BD FACS Calibur II, BD Biosciences).

Splenocytes were stimulated with MOG_35-55 peptide and leukocyte-stimulated cocktails, as described in the “MOG-specific reactivity assay” section. For Th cell detection, cells were stained with Paciffic-Blue-anti-CD4 (BD Biosciences) at RT for 30min, protected from light. After washing with PBS, cells underwent fixation and permeabilization (BD Cytofix/Cytoperm kit, BD Biosciences). Intracellular staining was performed with antibodies, including PE-Cy7-anti–IFN–γ, or PE-anti-IL-4, or FITC-anti-IL-17(BD Biosciences) at RT for 1 ​h, protected from light. We next examined the frequency of Th1 (CD4+IFN-γ+), Th2 (CD4+IL-4+) and Th17 (CD4+IL-17+) cells with the application of flow cytometry (BD FACS Calibur II, BD Biosciences). Regarding Treg cell assays, the cells were incubated with Paciffic-Blue-anti-CD4 and APC-anti-CD25 at RT for 30min, protected from light. Next, cell fixation and permeabilization were conducted using the Foxp3 Cytofix/Cytoperm kit (BD Biosciences), and intracellular staining was performed with PE-anti-Foxp3 (BD Biosciences). The proportion of Treg (CD4+CD25+Foxp3+) cells was detected by flow cytometry (BD FACS Calibur II, BD Biosciences). CellQuest and Flowjo V10 software were applied to analyze the flow cytometry data.

Electron microscopy

For transmission electron microscopy (TEM) imaging, BMDCs (10⁶ ​cells/mL) were initially fixed with 2.5 ​% glutaraldehyde in 0.1 ​M cacodylate buffer (pH 7.4) for 2 ​h at 4 ​°C, followed by post-fixation with 1 ​% osmium tetroxide for 1 ​h. Samples were dehydrated through a graded ethanol series and embedded in Epon 812 resin. Ultrathin sections (70 ​nm) were cut using a Leica UC7 ultramicrotome, stained with uranyl acetate and lead citrate, and examined under a Hitachi HT7800 TEM operated at 80 ​kV.

For scanning electron microscopy (SEM) analysis, BMDCs (10⁶ ​cells/mL) adhered to pre-coated coverslips were fixed with glutaraldehyde (3 ​%) at 4 ​°C for 90 ​min, followed by post-fixation with osmium tetroxide (1 ​%) for 20 ​min. After dehydration through an ethanol gradient, specimens were subjected to critical-point drying CO₂, sputter-coated with gold-palladium, and imaged using a Hitachi SU8000 SEM at an accelerating voltage of 3.0 ​kV.

Raman spectroscopy and data analysis

BMDCs from 4 groups (shN, shT, shN ​+ ​LPS, shT ​+ ​LPS; n ​= ​3 for each group) were washed with 1 ​mL deionized water 3 times. 0.5-2μL cell suspensions were pipetted onto an aluminum Raman slide (Shanghai D-band Co.,75 ​mm∗25 ​mm) and allowed to air dry. Raman spectroscopy detection was preceded by instrument calibration with a silicon wafer, which was calibrated to a signal of 520.7 ​cm⁻¹. Raman spectroscopy data collection was conducted using a Witec microscopic Raman spectrometer (alpha300R) equipped with 63x water immersive objective lens (Zeiss W Plan Apochromat 63×, N.A ​= ​1×), 100× air objective lens (Zeiss EC EPIPLAN, N.A. ​= ​0.75), a piezoelectric stage (UHTS 300, WITec, GmbH), a green solid-state excitation laser (λ ​= ​532 ​nm, 32 ​mW, WITec, GmbH), and an imaging spectrometer (Newton, Andor Technology Ltd. UK) with a grating of 600 groove/mm and a thermoelectrically cooled (60 ​°C) charged-coupled detector. Optical connection to the objective lens was achieved via a 10 ​μm diameter single-mode silicon fiber optic cable with the following parameter settings: spectral range 191–3945 ​cm⁻¹, laser power: 24mw, and a laser excitation spot size of 350 ​nm. The excitation laser intensity was maintained between each scan. Following quality control to eliminate unqualified spectra, a sum of 150 single spectra (50 ​cells/sample, 100 points/cell, 0.1s/point) were collected for each group. “The MASS” and “caret” packages of R 3.6.3 software were applied to conduct Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) analysis and plots. “Rtsne” package of R 3.6.3 software was applied to conduct t-Distributed Stochastic Neighbor Embedding (t-SNE) analysis. The intercepted plots of PCA, LDA and t-SNE were analyzed by using the PCA, lda and Rtsne function, respectively. Besides, three machine learning models, including Support Vector Machine (SVM), Neural Networks (NN), and K-Nearest Neighbors (KNN), were constructed using the “caret” package in R software (v3.6.3) to assess the discriminative ability of Raman spectral profiles. A library was built using 80 ​% of the overall data from each set of spectra, and the remaining 20 ​% of the data was incorporated into the library for comparison, ultimately contributing to the final accuracy of the samples. Meanwhile, receiver operating characteristic (ROC) curves were employed to assess the discriminatory power of Raman spectra between different groups.

RNA sequencing

Total RNA was extracted with TRIzol reagent (Invitrogen, CA, USA) according to the manufacturer's protocol. The purity and quantification of RNA were assayed by a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). The integrity of RNA was examined by an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Samples of acceptable purity, quantity and completeness were used in the construction of the library. Ribosomal RNA was eliminated with the application of the Ribo-off rRNA Depletion Kit (vazyme, Nanjing, China), and the libraries were constructed by the VAHTS Universal V6 RNA-seq Library Prep Kit (vazyme). The libraries were sequenced on the Illumina Novaseq 6000 platform, yielding 150 bp of paired-end reads. Approximately 549.23 ​M raw reads were created for each sample. Raw reads in fastq format are initially subjected to fastp processing to remove low-quality reads and achieve clean reads [21]. Around 79.99 ​GB of clean reads were then retained for subsequent analysis. The valid data volume of each sample ranged from 12.2 to 14.67 ​GB, the Q30 bases ranged from 96.86 ​% to 97.35 ​%, and the average GC content was 51.1 ​%. The clean reads are mapped to the reference genome utilizing HISAT2. The number of reads for each gene was obtained by HTSeq-count, followed by calculation of Fragments Per Kilobase of transcript per Million mapped reads (FPKM) for each gene [22,23]. PCA analysis was performed by using R 3.2.0 software to assess the biological reproducibility of the samples. Differential expression analysis was conducted on DESeq2 [24]. A q-value of less than 0.05, foldchange >2 or foldchange <0.5 was the threshold for a statistically differentially expressed gene (DEG). Hierarchical clustering analysis of DEGs was conducted in R 3.2.0 software to visualize the patterns of gene expression in separate groups and samples. Based on the hypergeometric distribution, enrichment analysis of GO, KEGG pathways, Reactome and WikiPathways of DEGs was conducted individually with R 3.2.0 software to screen the significantly enriched terms [25,26]. The histograms and bubble plots of the significantly enriched terms were plotted using R 3.2.0 software. The RNA-sequencing and analysis were performed by OE Biotech Co., Ltd. (Shanghai, China). The data have been deposited at the China National Center for bioinformation (CNCB) at https://ngdc.cncb.ac.cn/gsa/s/nT1bg65z (GSA No.: CRA024173).

Western blotting

Western blotting was performed as previously described [19]. BMDCs were lysed in RIPA buffer supplemented with 1 ​× ​phosphatase inhibitor cocktail (Roche) and 1 ​mM PMSF. The mouse spinal cord and cervical lymph nodes were homogenized in ice-cold RIPA buffer, centrifuged at 12,000g for 15 ​min, and the supernatants were collected. Protein concentrations were determined by using BCA assay (Novagen). Lysates were resolved on 12 ​% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels and transferred to polyvinylidene difluoride (PVDF) membranes. After blocking with 5 ​% BSA/TBST for 1 ​h at RT, membranes were incubated overnight at 4 ​°C with primary antibodies diluted in 2 ​% BSA/TBST: Nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB)-p65 (p65, 1:1000, Cell Signaling Technology (CST)-8242, Danvers, MA, USA), Phospho–NF–κB p65 (p-p65, Ser536, 1:500, CST-3033, Inhibitor of NF-κB (IκB, 1:500, CST-9242), Phospho-IκB (p-IκB, Ser36, 1:500, Abcam, Waltham, MA, USA), Interleukin-1 receptor-associated kinase 4 (IRAK4, 1:500, Abcam), Phospho-IRAK4 (p-IRAK4, Tyr345, 1:500, Abcam), Tumor necrosis factor receptor-associated factor 6 (TRAF6, 1:500, Abcam), Glyceraldehyde-3-phosphate dehydrogenase (GAPDH, 1:10,000, Real-ab, Tianjin, China), and Talin1 (1:1000, Abcam). After TBST washes, blots were probed with HRP-conjugated anti-rabbit IgG (1:2000, CST-7074) for 1 ​h at RT. The density of protein bands was analyzed by using Image J software and expressed as relative values [20].

Co-immunoprecipitation (co-IP)

LPS-stimulated BMDCs (1 ​μg/ml, 1 ​h) were harvested by cell scraping (VWR), lysed in Triton X-100 lysis buffer (Sigma) supplemented with protease/phosphatase inhibitors, and pre-cleared with 20 ​μL Protein A magnetic beads (Sigma) at 4 ​°C for 30min. For immunoprecipitation, pre-cleared lysates were incubated overnight at 4 ​°C with: TLR4 (1ug/100μL, Invitrogen, California, USA) or K63-Ubiquitin (1ug/100μL, Abcam) antibodies or species-matched IgG (as a negative control). Antibody complexes were captured by adding 30 ​μL Protein A magnetic beads (4 ​°C, 3 ​h). Beads were washed three times with lysis buffer, then boiled in 2 ​× ​Laemmli buffer (Bio-Rad). Samples were resolved by SDS-PAGE and immunoblotted using the following antibodies: TLR4 (1:500, Invitrogen), MyD88 (1:500, Novus Biologicals, Centennial, USA), IRAK4(1:500, Abcam), or TRAF6 (1:500, Novus Biologicals). Input controls represented 10 ​% of the total lysate protein.

Confocal microscopy

BMDCs were seeded onto μ-slide IV^0.4 (Ibidi) for 30min, then fixed at RT for 15 ​min with BD Cytofix (BD Bioscience). Fixed cells were permeabilized with 0.5 ​% Triton X-100 (Sigma) for 10min. After three washes with BD Perm/Wash buffer (BD Bioscience), cells were blocked with 5 ​% goat serum (Sigma) for 1 ​h at RT. Primary antibodies were incubated at 4 ​°C overnight: anti-TLR4 (1:200, Invitrogen), anti-MyD88 (1:200, Novus Biologicals), anti-IRAK4 (1:200, Abcam). Following PBS washes, species-matched secondary antibodies (1:200, Jackson ImmunoResearch) were applied at RT for 1 ​h. After stringent PBS washes, nuclei were counterstained with DAPI (5 ​μg/ml, Invitrogen) at RT for 10min. Imaging was performed on an LSM 880 with Airyscan (Zeiss). Z-stacks were acquired using identical settings across groups. Images were processed in ImageJ software, and colocalization analysis was conducted via the Colocalization Finder plugin to obtain the Pearson correlation coefficient (Colocalization index, r) [27].

Mixed lymphocyte reaction & antigen-specific T cell response

The mixed lymphocyte reaction was performed as previously described [19]. Briefly, CD4+T cells were isolated and purified from BALB/c mouse splenic mononuclear cells using magnetic beads and a specific kit for negative selection according to the manufacturer's instructions (Miltenyi Biotech). BMDCs were generated and stimulated with LPS as previously described [19]. Cells were further treated with mitomycin C (30 ​mg/L, Sangon Biotech, Shanghai, China) for 30 ​min and washed with PBS twice before harvesting. Then, BMDCs from different groups were incubated in triplicate with CD4+T cells at a ratio (BMDCs:CD4+T) of 1:5 for 72 ​h. CD4+T cells alone were used as a negative control (NC). The proliferation, differentiation (into Th1, Th2, Th17 and Treg cells) and cytokine secretion of CD4+T cells were assayed by CFSE staining, flow cytometry and cytometric bead array, respectively (detailed in “Flow cytometry” and “Cytometric Bead Array” sections).

For assessing antigen-specific T cell responses, BMDCs were loaded with MOG_35-55 peptide (20μg/ml) and stimulated with LPS (1ug/ml) for 24 ​h. A portion of BMDCs was further treated with mitomycin C (30 ​mg/L, Sangon Biotech) for 30 ​min and washed with PBS. Spleen CD4+T cells were obtained by negative selection from EAE mice 20 days after immunization with MOG_35-55. Finally, BMDCs were incubated with CD4+T cells at a ratio (BMDCs:CD4+T) of 1:5 for 72 ​h. The proliferation and differentiation of CD4+T cells were assayed by CFSE staining and flow cytometry (detailed in the “Flow cytometry” section).

Induction, treatment, and evaluation of EAE

Female C57BL/6 mice, 8–10 weeks of age, received subcutaneous injections at 4 points on the back with 250μg/mouse of myelin oligodendrocyte glycoprotein peptide MOG_35-55 (China Peptides (QYAOBIO), Shanghai, China) emulsified in Freund's complete adjuvant (1:1) (CFA, Sigma), containing heat-inactivated Mycobacterium tuberculosis H37Ra (4 ​mg/mL, Difco Laboratories, Detroit, MI, USA). On days 0 and 2, mice were further injected intraperitoneally with pertussis toxin (500ng/mouse, Listlabs, Campbell, California, USA). The mice were randomized and administered with vehicle or BMDCs and evaluated daily for weight and clinical signs by two blinded observers, according to a well-established neurological scoring system ranging from 0 to 15 points [28]. This scoring system is calculated as the sum of the tail and limb scores: For the tail, 0 ​= ​no sign; 1 ​= ​partial paralysis; 2 ​= ​complete paralysis. The limbs are evaluated separately: 0 ​= ​no sign; 1 ​= ​weak or altered gait; 2 ​= ​partial paralysis; 3 ​= ​complete paralysis. Accordingly, an animal with complete paralysis in all four limbs would receive a score of 14, and mortality was assigned a score of 15. For interventions in experimental animals, different groups of BMDCs were loaded with MOG_35-55 peptide (20μg/ml) and stimulated with LPS (1ug/ml) for 24 ​h. The cells or vehicle saline were administered into EAE mice through the tail vein on days 8, 12, and 16 (preclinical intervention) or days 12 and 16 (therapeutic intervention) post-immunization [29], and the mice were sacrificed on day 20 post-immunization for subsequent assessments.

Histopathology, immunohistochemistry and immunofluorescence assays

On day 20 post-immunization, mice (n ​= ​5–6 per group) were perfused with 4 ​% (w/v) paraformaldehyde, and spinal cord tissues were harvested and paraffin-embedded. The spinal cord lumbar enlargement sections (4μm/each section) of individual mice were processed with hematoxylin & eosin (HE) and Luxol Fast Blue (LFB) staining and semi-quantitatively assessed for inflammation and demyelination according to a standard protocol described previously [30]. Five to six white matter regions in each spinal cord section were randomly selected by two observers blinded to the grouping information, and the mean score for these regions was used to represent the degree of inflammation or demyelination of each section. The inflammatory infiltration score was graded as follows: 0, normal; 1, cellular infiltrates partially around meninges and blood vessels; 2, 1–10 lymphocytes infiltrate in a region; 3, 11–100 lymphocytes infiltrate in a region; 4, over 100 lymphocytes infiltrate in a region [19,30]. Demyelination was scored as: 0, normal; 1, rare areas of sporadic myelin sheath loss; 2, a few areas of demyelination; 3, massive myelin sheath loss [30].

Immunohistochemistry (IHC) staining was performed and analyzed as described previously [31]. Briefly, sections were deparaffinized, rehydrated and subjected to treatment with 3 ​% (v/v) H₂O₂ in methanol for 30 ​min to block endogenous peroxidase activity, followed by antigen retrieval. Then, the sections were incubated overnight with the following primary antibodies: anti–IFN–γ (1:50, BOSTER bio, Wuhan, Hubei, China), anti-IL-17(1:50, BOSTER bio), and anti-IL-10 (1:50, BOSTER bio) antibodies at 4 ​°C. The sections were further incubated with biotinylated secondary antibody (1:200, ABLAB, Henan, China) in combination with avidin-peroxidase (1:1000; Sigma). Finally, the tissue sections were examined under a light microscope (Olympus BX-61) and semi-quantitative analysis was performed by multiplying the staining intensity with the percentage of positive staining cells using ImageJ software. The percentage of positive staining cells was categorized into five grades: 0 for <5 ​%; 1 for 6 ​%–25 ​%; 2 for 26 ​%–50 ​%; 3 for 51 ​%–75 ​%; 4 for >75 ​% [31]. Staining intensity was determined by four degrees: 0, negative; 1, weak; 2, moderate; 3, strong [31]. Each section was observed for five to six high magnification ( ​× ​400) fields in the white matter of the spinal cord to obtain the average scores.

Spinal cord sections from each group were deparaffinized, rehydrated, and antigen-retrieved. Sections were further blocked with 3 ​% BSA/PBS and stained with anti-CD11c (1:400, CST) or anti-CD3 (1:200, CST) antibodies and intracellular anti-IL-10 (1:200, Abcam), followed by fluorescent secondary antibody. Signals were visualized and assessed by using a fluorescence microscope (Olympus 1000, Japan) and ImageJ software.

MOG-specific reactivity assay

To assess MOG_35-55-specific splenocyte reactivity, 2 ​× ​10⁵ fresh splenocytes from each group were restimulated with MOG_35-55 peptide (20 ​μg/mL) in 96-well plates at 37 ​°C 5 ​% CO₂ for 48 ​h and treated with leukocyte-stimulated cocktails (2 ​μL/10⁶ ​cells, BD Biosciences) at 37 ​°C under 5 ​% CO₂ for 6 ​h. After that, the cells were stained with relevant antibodies and detected by flow cytometry (BD FACS CaliburII, BD Biosciences) to assess the proportions of Th1, Th2, Th17, and Treg cells (detailed in the “Flow cytometry” section). The levels of IFN-γ, IL-4, IL-17, and IL-10 in the cell supernatants were detected by cytometric bead array (detailed in the “Cytometric Bead Array” section).

Cytometric bead array

Cytokine levels were detected using the LEGENDplex Mouse inflammation Panel, which assesses IL-β, IL-6, TNF-α, IL-10, IFN-γ, IL-4, and IL-17 (Biolegend, California, USA) following the manufacturer's instructions. Briefly, 25 ​μL of cell supernatant samples, assay buffer, and mixed beads were added to each well (totaling 75 ​μL). The plates were then incubated for 2 ​h. After two washes performed with a vacuum filtration unit, 25 ​μL of detection antibodies were added and incubated for 1 ​h. Without an intervening wash step, 25 ​μL of SA-PE was added, followed by a 30-min incubation. Finally, wells were washed twice using the vacuum filtration unit, 150 ​μL of 1X Wash Buffer was added, and samples were read on a flow cytometer (BD FACS Calibur II, BD Biosciences).

Statistical analysis

Data were expressed as mean ​± ​standard deviation (SD). Differences between two groups were analyzed by the Mann-Whitney U test or two-tailed Student's t-test, and differences among three or more groups were analyzed by one-way ANOVA with Turkey's multiple comparisons. Neurological scores at different time points of the EAE model were analyzed by two-way ANOVA, and the likelihood of the onset of EAE was analyzed by the Log-rank (Mantel-Cox) test. A P-value <0.05 was considered statistically significant. Statistical analysis and plotting were performed by using GraphPad Prism 9.5.1 software.

Results

Talin1 regulates BMDCs activation upon LPS stimulation in vitro

Talin1 knockdown alters morphological, ultrastructural characteristics, and surface molecular expression of LPS-treated BMDCs

DC activation and maturation are characterized by a variety of phenotypic alterations, including morphological and ultrastructural changes, as well as the upregulation of MHCII and co-stimulatory molecules (e.g., CD80, CD86). To identify the roles of Talin1 on the phenotypic activation of DCs, mouse BMDCs transduced with control (shN) or Talin1 knockdown (shT) lentiviral vectors were treated with or without LPS (1 ​μg/mL) and the above parameters were examined. Prior to this, the efficacy of candidate shT (shT1-3) lentiviral vectors in reducing Talin1 expression in BMDCs was evaluated, leading to the selection of shT2 as the superior knockdown lentiviral tool for subsequent studies (Supplemental Fig. 1).

SEM analysis (Fig. 1a–d) showed that in the absence of LPS stimulation, BMDCs from both shN and shT groups exhibited sparse short protrusions, consistent with the morphological features of immature/inactivated BMDCs. Upon LPS treatment, cells in the shN group (shN ​+ ​L) developed abundant long branched protrusions, characteristic of mature/activated BMDCs, whereas the shT group cells (shT ​+ ​L) displayed only a few of short protrusions (similar to those of immature or semi-mature BMDCs). TEM analysis (Fig. 1e–h) revealed that without LPS stimulation, the cytoplasm of BMDCs in both shN and shT groups contained abundant phagosomes and small amounts of organelles such as rough endoplasmic reticulum and mitochondria. Under LPS-treated conditions, BMDCs in the shN group (shN ​+ ​L) exhibited a significant reduction in the abundance of phagosomes, along with abundant organelles, including well-developed rough endoplasmic reticulum, mitochondria, and Golgi apparatus. In contrast, BMDCs in the shT group (shT ​+ ​L) exhibited significantly decreased number of phagosomes, while no increase in the abundance of organelles was observed. In summary, Talin1 knockdown did not affect the morphology and ultrastructure of BMDCs without LPS stimulation, but it significantly altered the morphology and ultrastructure of LPS-treated BMDCs, inducing a shift toward an immature/semi-mature phenotypic state.

Fig. 1.

Talin1 knockdown alters the morphological, surface molecules and cytokine characteristics during BMDCs activation. BMDCs transduced with control (shN) or Talin1-knockdown (shT) lentiviral vectors were treated with control medium (shN vs. shT) or LPS (1 ​μg/mL, shN ​+ ​L vs. shT ​+ ​L) for 12 ​h. a-d, Scanning Electron Microscopy (SEM) analysis of BMDCs in different groups. e-h, Transmission electron microscopy (TEM) analysis of BMDCs with different treatments. The red thin arrow indicates phagosomes, and the red heavy arrow indicates Golgi apparatus; the blue thin arrow indicates mitochondria, and the blue heavy arrow indicates endoplasmic reticulum. i, Flow cytometry analysis of surface makers (CD80, CD86, MHCII) of BMDCs (n ​= ​6/group). j-m, Relative transcriptional levels of IL-1β, IL-6, TNF-α, and IL-10 in BMDCs (n ​= ​3/group). n-q, Concentrations of IL-1β, IL-6, TNF-α, and IL-10 in the supernatants of BMDCs (n ​= ​3/group). Data are expressed as the mean ​± ​SD. ns, no statistical difference; ∗P ​< ​0.05, ∗∗P ​< ​0.01, ∗∗∗P ​< ​0.001, ∗∗∗∗P ​< ​0.0001.

Flow cytometric analysis of surface molecular expression showed that in the absence of LPS treatment, CD80 levels in BMDCs from the shT group were significantly lower than that from the shN group (P ​= ​0.01), while after LPS stimulation, levels of CD80 (P ​< ​0.0001), CD86 (P ​< ​0.0001), and MHCII (P ​= ​0.002) in the shT group (shT ​+ ​L) were all reduced compared with those in the shN group (shN ​+ ​L) (Fig. 1i). Since BMDCs stimulated by LPS produce higher levels of pro-inflammatory cytokines, levels of these indicators were detected by qRT-PCR and cytometric bead array. Our results showed that knockdown of Talin1 attenuated LPS-induced upregulation of IL-1β, IL-6, and TNF-α, and increased the expression of IL-10 at both transcriptional and protein levels (Fig. 1 j-q). Taken together, these findings suggest that Talin1 knockdown predisposes BMDCs to an inactivated/immature pattern of surface molecules and cytokines upon LPS stimulation.

Talin1 knockdown alters the Raman spectral profiles of BMDCs post-LPS stimulation

Single-cell Raman spectroscopy is a powerful optical method for non-invasively generating chemical fingerprints of samples and has been successfully applied to multidimensional characterization and differentiation of a wide range of immune cells. Here, Raman spectroscopy was employed to identify chemometric differences among various groups of BMDCs. Visualization of PCA, LDA, and t-SNE showed that the shT and shN groups could not be clearly differentiated under LPS-untreated conditions. In contrast, after LPS stimulation, a clear distinction emerged between these two groups, as evidenced by the clustering of single cells within each group and their separation from the other group (Fig. 2 a-f, Supplemental Fig. 2).

Fig. 2.

Talin1 knockdown changes the biochemical fingerprint profiles of BMDCs detected by Raman spectroscopy with machine learning. BMDCs (n ​= ​3/group) transduced with control (shN) or Talin1-knockdown (shT) lentiviral vectors were treated with control medium (shN vs. shT) or LPS (1 ​μg/mL, shN ​+ ​L vs. shT ​+ ​L) for 12 ​h. a-b, Sample visualized PCA, t-SNE, LDA, and mean values plots between shN and shT groups (shN vs. shT). c-d, Sample visualized PCA, t-SNE, LDA, and mean values plots between groups of shN ​+ ​L and shT ​+ ​L (shN ​+ ​L vs. shT ​+ ​L). e-f, Sample visualized PCA, t-SNE, LDA, and mean values plots among shN, shT, shN ​+ ​L, and shT ​+ ​L groups (shN vs. shT vs. shN ​+ ​L vs. shT ​+ ​L). The purple, green, red, and blue points in the plots represent samples from shN, shT, shN ​+ ​L, and shT ​+ ​L groups, respectively. g-h, Differentiation of biochemical characteristics between groups of shN and shT (shN vs. shT) was assessed by the SVM, NN, and KNN models, as well as ROC curve. i-j, Discriminability of biochemical profiles between groups of shN ​+ ​L and shT ​+ ​L (shN ​+ ​L vs. shT ​+ ​L) was evaluated by using the SVM, NN, and KNN models, as well as the ROC curve. k, The biochemical distinction was identified among shN, shT, shN ​+ ​L, and shT ​+ ​L groups by machine learning models of SVM, NN, and KNN. PCA, Principal Component Analysis; LDA, Linear Discriminant Analysis; t-SNE, t-Distributed Stochastic Neighbor Embedding. SVM, Support Vector Machine; NN, Neural Networks; KNN, K-Nearest Neighbors.

We further assessed the distinctions among BMDC groups based on their Raman spectral profiles using three machine learning models (SVM, NN, and KNN). These models were configured to perform either a four-class classification task, which simultaneously distinguishes four groups of BMDCs (shN, shT, shN ​+ ​L, and shT ​+ ​L), or a binary classification task, which distinguishes two groups of BMDCs (shN vs. shT and shN ​+ ​L vs. shT ​+ ​L). Accordingly, the Raman dataset, consisting of a total of 150 spectra for each group, was split into a training set and a testing set at a spectral ratio of 80:20. Spectra from the same cell were exclusively assigned to either the training or the testing set. The SVM, NN, and KNN models achieved an overall accuracy of 98 ​%, 98 ​%, and 94 ​% (for the shN ​+ ​L vs. shT ​+ ​L binary classification), and 66 ​%, 60 ​%, and 63 ​% (for the shN vs. shT binary classification), respectively, when classifying the testing set into two groups (Fig. 2 g,i). Moreover, the three models achieved an overall accuracy of 82 ​%, 67 ​%, and 73 ​%, respectively in categorizing the test set into four groups (Fig. 2k). Within the four-class classification, spectra belonging to the shN ​+ ​L group were predicted with the highest accuracy in all three models at 100 ​%, 94 ​%, and 100 ​%, respectively. The overall low prediction accuracy for this four-class classification task was due to the fact that the two groups, shN and shT, could not be clearly distinguished from each other (Fig. 2k). These data indicated that the Raman spectra of shN ​+ ​L and shT ​+ ​L were significantly different and could be effectively distinguished by machine learning modeling, unlike the Raman profiles of shN and shT. ROC curve analysis yielded similar results to machine learning (Fig. 2 h,j). Collectively, these findings suggest that Talin1 knockdown exerts a prominent effect on the biochemical properties of LPS-treated BMDCs, but it has no significant effect on the relevant profiles of BMDCs unstimulated by LPS.

Talin1 knockdown suppresses LPS-induced BMDCs activation or maturation through downregulating the TLR4/MyD88/NF-κB signaling pathway

Talin1 significantly regulates multiple immune-related pathways during BMDCs activation

To further explore the potential regulatory mechanisms of Talin1 on DC activation, we performed RNA-sequencing on shN ​+ ​L and shT ​+ ​L groups. The samples were first categorized into different intervals according to their expression values (FPKM), and the number of genes expressed within each interval was calculated. The results indicated a consistent distribution of gene expression between these 2 groups (Fig. 3a). A sample-to sample distance heatmap was obtained based on gene expressions, revealing high correlation among samples within the same group and low correlation between samples from different groups (Fig. 3b). PCA also revealed a clear separation between shN ​+ ​L and shT ​+ ​L groups (Fig. 3c). Differential expression analysis identified a total of 560 genes were significantly up-regulated and 877 genes were down-regulated between the shN ​+ ​L and shT ​+ ​L groups (Fig. 3d). The distribution of differential genes and relevant cluster analysis are presented in Fig. 3e–f. We further performed various enrichment analysis (GO, KEGG, Reactome, Wikipathway) for differentially expressed genes (up-regulated, down-regulated, and total), which demonstrated that multiple immunity/inflammatory processes and pathways were differentially expressed between shN ​+ ​L and shT ​+ ​L groups (Fig. 3 g-i, m-o; Supplemental Fig. 3). Besides, differential genes from three GO biological processes (including positive regulation of cell migration, cytokine activity, and inflammatory response), closely associated with the activation of DCs were clustered, which showed that the expression of the relevant genes in group shN ​+ ​L was overall higher than those in group shT ​+ ​L, indicating a compromised activation landscape in BMDCs upon Talin1 knockdown (Fig. 3 j-l). To more precisely target immunology-related pathways influenced by Talin1 knockdown, we conducted KEGG enrichment analysis of the downregulated genes to identify specific categories related to “signal transduction” and “immune system” processes. The data revealed that genes significantly down-regulated in the shT ​+ ​L group were enriched in the TNF, NF-κB and JAK-STAT signaling pathways within the “signal transduction” category and in the Hematopoietic cell lineage, Antigen processing and presentation, and NOD/Toll-like receptor-related pathways within the “immune system” category (Fig. 3 p-q). Collectively, our findings suggest that Talin1 knockdown yields an adverse effect on processes or pathways associated with the activation or maturation of BMDCs.

Fig. 3.

Talin1 knockdown affects the transcriptional profiles during BMDC activation, determined by RNA sequencing with enrichment analysis. BMDCs (n ​= ​3/group) transduced with control or Talin1-knockdown lentiviral vectors were treated with LPS (1 ​μg/mL, LPS vs. shTalin1-LPS) for 12 ​h. a, Distribution of gene expression for each sample. b, Heatmap of sample−to−sample distances between groups of LPS and shTalin1-LPS. c, Visualized PCA plot of samples. d, Histogram of differentially expressed genes. e, Volcano plot of differential expression genes (q ​< ​0.05, ILog₂FCI >1). Gray dots represent non-significantly different genes, and red and blue dots represent significantly up- and down-regulated genes, respectively. f, Differential genes grouping clustering plot (q ​< ​0.05, ILog₂FCI >1). Red color indicates relatively high expression of genes, blue color indicates relatively low expression of genes. g-i, Top 30 GO terms of downregulated, upregulated and total differential genes between LPS and shTalin1-LPS groups. j-l, Clustering plots of differential genes in the GO terms of “positive regulation of cell migration”, “cytokine activity”, and “inflammatory response”. m-o, Top 20 KEGG enrichment terms of downregulated, upregulated and total differential genes between LPS and shTalin1-LPS groups. p-q, Bubble plots of KEGG pathway enrichment analysis showing significantly downregulated genes in “Environmental information processing-signal transduction” and “Organismal systems-immune system” categories. PCA, Principal Component Analysis.

Talin1 regulates BMDCs activation via the TLR4/MyD88/NF-κB pathway

TLR4/MyD88/NF-κB pathway is a classical signaling pathway crucial for DC activation. Talin1 has recently been reported to be involved in forming a preassembled TLR complex. Notably, our RNA-sequencing data revealed that both the Toll-like receptor-related pathway and NF-κB signaling pathway were significantly enriched with down-regulated genes. To this end, the role of Talin1 on the expression and interaction of key molecules related to the TLR4/MyD88/NF-κB pathway in BMDCs were evaluated. The data revealed that Talin1 knockdown consistently impaired the TLR4/MyD88/NF-κB pathway activation upon LPS stimulation, as determined by the declined levels (ratios) of p-p65 (p-p65/p65), p-IκB (p-IκB/-IκB), p-IRAK4 (p-IRAK4/IRAK4), as well as compromised IκB degradation and TRAF6 K63-ubiquitination (Fig. 4 a-b, Supplemental Fig. 4). Interestingly, p-p65, p-IκB and p-IRAK4 were significantly upregulated after 1 ​h of LPS stimulation, whereas the levels of the above molecules were only weakly increased or even decreased after 12 ​h of LPS stimulation. Since TLR4/MyD88/IRAK4 interactions are necessary for activation of the TLR4/MyD88/NF-κB pathway, TLR4/MyD88 and TLR4/IRAK4 interactions were further examined using co-immunoprecipitation and confocal microscopy. Relevant data indicated that Talin1 knockdown yielded no marked effect on the binding and co-localization of TLR4/MyD88 (P ​= ​0.1014), but significantly reduced the interaction of TLR4/IRAK4 (P ​= ​0.0012) after 1 ​h of LPS stimulation (Fig. 4 c-e). In summary, Talin1 knockdown suppresses the TLR4/MyD88/NF-κB pathway activation by downregulating the interaction of TLR4/IRAK4 and phosphorylation/ubiquitination levels of related molecules, ultimately impairing the activation of the investigated BMDCs.

Fig. 4.

Talin1 knockdown compromises the key molecules activation or interaction of TLR4/MyD88/NF-κB pathway. BMDCs transduced with control (shN) or Talin1-knockdown (shT) lentiviral vectors were treated with LPS (1 ​μg/mL) for indicated times. a, Talin1 knockdown declined (in trend) the p-p65 level and IκB degradation, and significantly reduced expressions of p-IκB and p-IRAK4 in BMDCs detected by western blotting. b-c, Co-immunoprecipitation (co-IP) of K63-Ub or TLR4, followed by immunoblot analysis for TRAF6 or TLR4/MyD88/IRAK4. 10 ​% of the cell lysates used for co-IP were run as input controls. The data revealed that Talin1 knockdown remarkably reduced TRAF6 ubiquitination (K63-Ub) and inhibited IRAK4 recruitment to TLR4 without affecting MyD88 recruitment to TLR4 in BMDCs treated with LPS for 1 ​h. d-e, Confocal images and colocalization analysis of TLR4/MyD88 and TLR4/IRAK4 in BMDCs with different treatments (n ​= ​6–7 ​cells/group). The data showed that Talin1 knockdown significantly inhibited TLR4/IRAK4 colocalization without affecting TLR4/MyD88 colocalization. Data are expressed as the mean ​± ​SD. ∗P ​< ​0.05, ∗∗P ​< ​0.01.

Talin1 knockdown impairs BMDC-mediated activation of allogeneic and antigen-specific CD4+T cell responses

Since DC represents a major promoter of T cell proliferation and differentiation, we sought to determine whether Talin1 affects DC-mediated T cell responses in vitro. BMDCs from C57BL/6 mice treated with or without LPS were co-cultured with splenic CD4+T cells from BALB/C mice at a ratio of 1:5 for 72 ​h. Consequently, there was no significant difference in the proliferation and differentiation of CD4+T cells co-cultured with BMDCs from the shN and shT groups. In contrast, co-culture with BMDCs from the shT ​+ ​L group significantly inhibited the proliferation of CD4+T cells (P ​< ​0.0001), decreased the proportions of Th1 (P ​= ​0.004) and Th17 ​cells (P ​= ​0.0377), and downregulated the secretion of IFN-γ (P ​< ​0.0001) and IL-17 (P ​= ​0.0004) in the co-culture supernatants, compared to co-culture with BMDCs from the shN ​+ ​L group (Fig. 5). Additionally, BMDCs loaded with MOG_35-55 peptide and stimulated by LPS were co-cultured with spleen CD4+T cells from the EAE mice to assess their capacity to mediate antigen-specific T responses. The data demonstrated that Talin1 knockdown significantly inhibited the ability of BMDCs to stimulate CD4+T cell proliferation (P ​< ​0.05) and differentiation into Th1 (P ​< ​0.01) and Th17 (P ​< ​0.01) cells, whereas it increased the proportion of Treg cells (P ​< ​0.05) (Supplemental Fig. 5). Consequently, Talin1 knockdown significantly impaired the ability of BMDCs to promote allogeneic and MOG_35-55-specific CD4+ T cell responses, suggesting its potential application in EAE management.

Fig. 5.

Talin1 knockdown impairs the capacity of BMDCs to stimulate allogeneic CD4+T cells proliferation and differentiation. BMDCs from C57BL/6 mice transduced with control or Talin1-knockdown lentiviral vectors were treated with control medium (shN vs. shT) or LPS (1 ​μg/mL, shN ​+ ​L vs. shT ​+ ​L) for 12 ​h. BMDCs with different intervenes were further co-cultured with spleen CD4+T cells from BALB/c mice at a ratio of 1:5 for 72 ​h. a, CFSE staining with flow cytometry for CD4+T cells proliferation analysis. b-e, Flow cytometry analysis for CD4+T cells differentiation into Th1 (CD4+IFN-γ+), Th2 (CD4+IL-4+), Th17 (CD4+IL-17+), and Treg (CD4+CD25+Foxp3+) cells. f-i, Cytometric Bead Array detected levels of IFN-γ, IL-4, IL-17, and IL-10 in co-culture supernatants. Data are expressed as the mean ​± ​SD. NC, negative control; ns, no statistical difference; ∗P ​< ​0.05, ∗∗P ​< ​0.01, ∗∗∗P ​< ​0.001, ∗∗∗∗P ​< ​0.0001.

Regulatory BMDCs induced by Talin1 knockdown protect against EAE by modulating inflammatory responses

Talin1 knockdown BMDCs alleviate EAE pathology and neurologic deficits

We initially assessed the impact of EAE induction on Talin1 expression. The data showed that Talin1 protein levels in the lymph nodes (P ​= ​0.0055) and spinal cord (P ​= ​0.0278) of EAE mice were significantly higher than those in normal controls (Supplemental Fig.6l), suggesting that dysregulated Talin1 expression may be associated with the pathogenesis of EAE.

Considering that Talin1 knockdown significantly downregulated the expression of co-stimulatory molecules, MHCII and pro-inflammatory cytokines of BMDCs, and impaired their capacity to stimulate the responses of CD4+T cells, we investigated whether in vivo administration of Talin1 knockdown BMDCs could mitigate the severity of EAE in mice. Given that early signs of EAE were observed on days 10–12 post-immunization, different groups of BMDCs preloaded with MOG_35-55 peptide and stimulated by LPS were injected into the mice through tail vein on days 8, 12, and 16 (preclinical intervention) or days 12 and 16 (therapeutic intervention) after EAE induction, and the mice were sacrificed on day 20 post-immunization (Fig. 6a, Supplemental Fig. 6a). For preclinical treatment (Fig. 6b–g), Talin1 knockdown BMDCs significantly delayed the development of EAE, as evidenced by alleviated weight loss (from day 10–11 to day 20 after EAE induction), reduced mean scores (from day 11 to day 20 after EAE induction), lower maximal scores (P ​< ​0.0001) and cumulative scores (P ​< ​0.0001), as well as delayed disease onset (P ​< ​0.0001). For therapeutic intervention, Talin1 knockdown BMDCs effectively alleviated the progression of EAE, as demonstrated by reduced weight loss (from day 19 to day 20 after EAE induction), lower mean scores (from day 16 to day 20 after EAE induction) and maximum scores (P ​< ​0.0001), and improved cumulative scores (P ​= ​0.0002) despite similar onset times (days 10.38 ​± ​0.74 vs. 10.25 ​± ​1.04, Supplemental Fig. 6b–e). Overall, these findings indicated that regulatory BMDCs induced by Talin1 knockdown attenuated the development and progression of EAE in mice. Alternatively, to determine the impact of Talin1 knockdown BMDCs on the severity of inflammation and demyelination in the CNS of EAE mice, lumbar enlargement of spinal cord tissues was processed for HE and LFB staining. For both preclinical and therapeutic intervention, the inflammation scores and demyelination scores in the white matter of the spinal cord were significantly diminished in Talin1-knockdown BMDCs-treated mice compared with the control groups (Fig. 6h-m, Supplemental Fig. 6f–k). Taken together, our data suggest that Talin1 knockdown BMDCs attenuate the severity of neurological impairments in EAE mice by mitigating the pathological manifestations of inflammatory infiltration and demyelination in the CNS.

Fig. 6.

Talin1-knockdown BMDCs alleviate the development, neurological deficits, and pathology of EAE mice. C57BL/6 mice induced into the EAE model were randomized and treated with saline (EAE, n ​= ​10) or BMDCs transduced with control lentiviral vectors (EAE ​+ ​shN, n ​= ​10) or BMDCs transduced with Talin1-knockdown lentiviral vectors (EAE ​+ ​shT, n ​= ​10) by tail vein injection. BMDCs for intervention in EAE models were loaded with MOG_35-55 peptide (20μg/ml) and stimulated by LPS (1ug/ml) for 24 ​h. a, Flowchart of BMDCs preclinical intervention in EAE mice. b-c, The changes of body weight and neurological scores of each group. d, Log-rank (Mantel-Cox) test comparing the onset time of each group. e-g, Comparison of time to onset, maximal scores and cumulative scores among different groups. On day 20 post-immunization, mice in each group were sacrificed and their lumbar spinal cord tissue were collected for pathological analysis. h-j, HE staining and analysis of inflammation infiltration in the white matter of spinal cord (n ​= ​6; h, Scale bar ​= ​500 ​μm; i, Scale bar ​= ​50 ​μm). k-m, LFB staining and analysis of demyelination in the white matter of spinal cord (n ​= ​5–6; k, Scale bar ​= ​500 ​μm; l, Scale bar ​= ​50 ​μm). Data are expressed as the mean ​± ​SD. ※ indicates P ​< ​0.05, compared to EAE group, and # indicated P ​< ​0.05, compared to EAE ​+ ​shN group. ns, no statistical difference; ∗P ​< ​0.05; ∗∗∗∗P ​< ​0.0001.

Talin1-knockdown BMDCs treatment inhibits proinflammatory (Th1 and Th17 ​cells) and enhances anti-inflammatory (Treg cells) responses in EAE mice

Th1 and Th17 ​cells play a pivotal role in the pathogenesis of EAE, driving a variety of pro-inflammatory processes in the peripheral and CNS, which prompted us to further explore whether Talin1 knockdown in BMDCs could alleviate EAE by modulating the above cardinal cellular responses. For this purpose, splenocytes were restimulated with MOG_35-55 peptide, and the proportions of Th or Treg cells were detected by flow cytometry. Meanwhile, the infiltration of IFN-γ-, IL-17- and IL-10-positive cells in the lumbar enlargement of spinal cord tissue was assessed by IHC staining. As shown in Fig. 7a-h, Talin1 knockdown BMDCs dramatically decreased the differentiation into Th1 and Th17 ​cells and increased the proportion of Treg cells, as well as reduced the concentrations of IFN-γ and IL-17 and enhanced the level of IL-10 in the supernatants. Consistently, Talin1 knockdown BMDCs notably reduced the IHC scores of IFN-γ-positive and IL-17-positive cells and increased the IHC score of IL-10-positive cells in the spinal cord (Fig. 7 i-k). To identify the primary cellular sources of IL-10 in the spinal cord (T cells vs. DCs), we performed dual immunofluorescence staining for CD3⁺IL-10⁺ and CD11c⁺IL-10⁺ cells. The data demonstrated that the percentage of IL-10-producing T cells (CD3⁺IL-10⁺) in the spinal cord of the EAE ​+ ​shT group was significantly higher than that in the EAE group, while the percentage of IL-10-producing DCs (CD11c⁺IL-10⁺) in the spinal cord of the EAE ​+ ​shT group was comparable to that in the EAE group (Supplemental Fig. 7). These findings indicated that most of the increased IL-10 in the spinal cord of EAE mice treated with Talin1 knockdown BMDCs were predominantly derived from T cells, with Treg cells representing the most likely subset. Overall, our data suggest that Talin1 knockdown in BMDCs alleviates the neuropathological deficits of EAE by fine-tuning the homeostasis of Th1/Th17/Treg cells and associated cytokines.

Fig. 7.

Talin1-knockdown BMDCs regulate Th1/Th17/Treg responses in the spleen, and IFN-γ-, IL-17-, and IL-10-positive cells infiltration in the spinal cord of EAE mice. EAE mice in each group were sacrificed on day 20 post-immunization. Their splenocytes were collected and re-stimulated with MOG_35-55 peptide. a-d, Flow cytometry analysis of proportions of splenic Th1 (CD4+IFN-γ+), Th2 (CD4+IL-4+), Th17 (CD4+IL-17+), and Treg (CD4+CD25+Foxp3+) cells (n ​= ​5/group). e-h, Levels of IFN-γ, IL-4, IL-17, and IL-10 in the cell supernatants (n ​= ​5/group). i-k, Their lumbar enlargement sections of spinal cord were collected for the immunohistochemistry assessments of IFN-γ-, IL-17-, and IL-10-positive cells (Scale bar ​= ​50 ​μm, n ​= ​5/group). Data are expressed as the mean ​± ​SD. ns, no statistical difference; ∗P ​< ​0.05, ∗∗P ​< ​0.01, ∗∗∗P ​< ​0.001, ∗∗∗∗P ​< ​0.0001.

Discussion

Talin1, encoded by TLN1, is a well-established focal adhesion cytoskeletal protein that participates in cell migration and adhesion through binding and activating integrins [14]. Studies have demonstrated that Talin1 is essential for the migration and infiltration of a wide range of immune cells to sites of injury or inflammation [15,16,18,[32], [33], [34]]. Beyond its role in cell motility, Talin1 has been implicated in several immunological processes. For instance, cytokines produced by macrophages and the proliferation, differentiation, and cycle of acute myeloid leukemia (AML) cells are dependent on Talin1 [33,35]. Besides, it has been reported that mice with a DC-specific Talin1 deficiency exhibit impaired support for CD47-deficient cell-associated antigen-induced T cell proliferation and differentiation [36]. These findings collectively suggest Talin1's involvement in the regulation of functional properties of immune cells. Furthermore, a recent study revealed Talin1's involvement in the formation of pre-assembled TLR4 complexes at steady state, resulting in a faster and stronger response to TLR ligand binding [18]. Given that TLR4-mediated downstream pathways control the activation or maturation of DCs, we sought to identify the impacts of Talin1 knockdown on various phenotypic and functional hallmarks of both inactivated and activated DCs. Meanwhile, given that hyperactivated DCs are engaged in the pathogenesis of MS by stimulating the activation, proliferation, and differentiation of CD4+T cells, while TolDCs exert a potentially protective role in MS/EAE by restoring the equilibrium of pro- and anti-inflammatory profiles, we aimed to investigate whether Talin1 knockdown DCs are protective against EAE and reveal the underlying regulatory mechanisms involved.

Firstly, our data revealed that Talin1 knockdown did not alter the cellular morphology and ultrastructure, surface MHCII and co-stimulatory molecules expression, cytokines secretion, as well as biochemical features of steady-state (inactivated) BMDCs. In contrast, Talin1 knockdown resulted in LPS-stimulated (activated) BMDCs exhibiting sparse/short protrusions, reduced organelles, lower levels of MHCII, co-stimulatory molecules, and pro-inflammatory cytokines, and a markedly divergent biochemical composition, with phenotypic signatures similar to those of TolDCs. These findings indicate that Talin1 knockdown suppresses the phenotypic activation programs of BMDCs without affecting their steady-state phenotypic characteristics, suggesting that down-regulation of Talin1 may represent a viable strategy for TolDCs induction without compromising their intrinsic immunological properties.

In terms of functional profiles, as classical antigen-presenting cells, activated or mature DCs promote T cell activation, proliferation, and preferential polarization of CD4+ T cells into Th1 and Th17 lineages through the high expression of MHCII and co-stimulatory molecules, as well as the secretion of a large number of inflammatory factors. These processes are crucial for the body to resist and eliminate pathogenic factors under physiological conditions, whereas over-activation of the above systems trigger the aberrant activation and proliferation of autoreactive T cells, which in turn leads to a variety of autoimmune diseases [2,3]. On the other hand, TolDCs exhibit phenotypic and functional signatures similar to immature or semi-mature DCs, including hypo-expression of co-stimulatory and pro-inflammatory molecules and hyper-expression of anti-inflammatory factors, which promote the development of Treg cells and induces tolerance to maintain the immune homeostasis [37]. In this study, inactivated BMDCs exhibited limited ability to stimulate CD4+T cell proliferation and differentiation irrespective of Talin1 expression, whereas Talin1 knockdown in activated BMDCs significantly inhibited CD4+T cell proliferation and differentiation into Th1 and Th17 ​cells. These findings suggest that Talin1 knockdown effectively disrupts the functional activation programs of BMDCs, and the functional characteristics of Talin1-deficient BMDCs are similar to those previously reported for TolDCs [38,39]. Accordingly, the above findings reveal that Talin1 knockdown could effectively inhibit the activation of BMDCs, thereby providing a potential approach for the generation of TolDCs.

Next, we sought to investigate the potential molecular mechanisms by which Talin1 knockdown disrupts the phenotypic and functional activation programs of BMDCs (also known as the induction of TolDCs). RNA-sequencing and enrichment analysis revealed that Talin1 knockdown dysregulated DCs activation-induced upregulation of multiple immune/inflammation-related biological processes and signaling pathways, especially the TLR/NF-κB pathway. The subsequent validation study unveiled that Talin1 knockdown substantially mitigated the activation of key molecules of the TLR4/MyD88/NF-κB pathway, as well as TLR4/IRAK4 interaction after LPS stimulation for 1 ​h. Consistently, a recent study by Lim et al. uncovered that Talin1 knockout significantly inhibited TLR4/MyD88 and TLR4/IRAK4 interactions in DCs or Langerhans cells intervened by LPS for 30min [18]. Mechanistically, the interaction between Talin1 and MyD88 has been reported to serve as a scaffold for the TLR4 complex under steady-state conditions and during the first 10 ​min of LPS stimulation. Afterward, upon release of Talin1 from the preassembled TLR4/MyD88 complex, the intermediate structural domain of MyD88 enables IRAK4 recruitment 30 ​min after LPS stimulation. Building upon the above findings, a spatiotemporal sequence can be inferred regarding the binding and release of Talin1 from the TLR4/MyD88 complex and the interaction of IRAK4 with this complex, with MyD88 undergoing a series of conformational adjustments to accommodate IRAK4 binding after the release of Talin1. In our study, Talin1 knockdown was confirmed to result in either unaffected TLR4/MyD88 interactions or significantly suppressed TLR4/IRAK4 interactions in BMDCs stimulated with LPS for 1 ​h. The slight inconsistency regarding TLR4/MyD88 interactions may stem from differences in the method and extent of Talin1 interference (knockdown vs. knockout), in addition to the following reasons: 1) the residual amount of Talin1 present after Talin1 knockdown may be sufficient for the formation of a steady-state TLR4/MyD88 complex or amplified certain compensatory mechanisms, which led to uninhibited TLR4/MyD88 interactions, given the persistency of some TLR4/MyD88 complexes even with Talin1 knockout in Lim et al.'s study (suggesting there may exist other compensatory mechanisms in the formation and maintenance of TLR4/MyD88 complex); 2) the small number of Talin1 available may be sufficient to form an abundant TLR4/MyD88 complex over a relatively long period of time, and the discrepancies between these two studies may be due to the difference in the length of LPS stimulation (1 ​h vs. 30min). Besides, although TLR4/MyD88 interactions in BMDCs were unaffected, TLR4/IRAK4 interactions were inhibited by Talin1 knockdown, which may be attributed to the fact that relatively lower levels of Talin1 delayed and/or diminished the formation of MyD88 structural domains that can be specifically bound by IRAK4. Accordingly, Talin1 controls activation of the TLR4/MyD88/NF-κB pathway in BMDCs by influencing key cascade reactions or events associated with this pathway.

Multiple biologically and pharmacologically induced TolDCs have been widely used in the treatment of autoimmune disease or related animal models such as EAE [[40], [41], [42]]. In the present study, Talin1 knockdown BMDCs exhibited phenotypic and functional profiles similar to those of TolDCs. Therefore, we investigated whether such cells could protect against EAE and the mechanisms involved. As expected, preclinical and therapeutic intervention with Talin1 knockdown BMDCs could efficiently mitigate the development and progression of EAE in mice by alleviating inflammatory infiltration and demyelination in the CNS. This may be attributed to the depression of peripheral Th1 and Th17 ​cell differentiation and enhancement of Treg cell differentiation mediated by Talin1 knockdown BMDCs, which in turn significantly reduced pro-inflammatory IFN-γ+ and IL-17+ cells and increased anti-inflammatory IL-10+ cells infiltrating in the spinal cord. These findings further substantiate that Talin1 knockdown BMDCs share similar functional profiles with TolDCs.

Our study presents several limitations that should be acknowledged. Importantly, the temporal dynamics of the influence of Talin1 on the TLR4 complex and subsequent signaling events remain to be fully characterized. Besides, the precise molecular mechanisms by which Talin1 knockdown BMDCs hindered Th1 and Th17 ​cells differentiation in EAE mice have not been systematically clarified. Nonetheless, our findings establish Talin1 as a precise immunomodulatory target for DCs activation and provide a DC-centered neurotherapeutic approach for counteracting immune dysregulation in MS.

In conclusion, Talin1 mediates the activation landscapes of BMDCs through the TLR4/MyD88/NF-κB pathway. Talin1-knockdown BMDCs ameliorate EAE neuropathology by modulating Th1/Th17/Treg homeostasis, which is expected to yield novel insights for the targeted treatment of MS and other related autoimmune disorders.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval

All experimental protocols were conducted in accordance with the guidelines of the Animal Care and Use Committee of KangTai Medical Laboratory (Hebei, China).

Consent for publication

Not applicable.

Author Contributions

Jia Liu contributed with: conception and design, experimental conduction, analysis and interpretation of data, statistical analysis, and drafting and revision of the manuscript. Xiaorui Guan and Yuanbo Cao contributed with: experimental conduction, analysis of data, and revision of the manuscript. Zhen Jia and Bin Li contributed with analysis of data and revision of the manuscript. Kazuo Sugimoto contributed with: study supervisor.

Funding

This study was supported by the Beijing Natural Science Foundation (Grant No. 7222114), the National Natural Science Foundation of China (Grant No. 82205034), and the Basic scientific research operation cost project of Beijing University of Chinese Medicine (2023-JYB-JBQN-017).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article.

Supplemental Fig. 1.

Confirmation of Talin1 knockdown by lentiviral vectors. A, Sequence of control (shN) and Talin1 knockdown (shT1-3) lentiviral vectors. B, Effect of different Talin1 knockdown lentiviral vectors on Talin1 expression in BMDCs detected by western blotting, indicating Talin1 protein was significantly knocked-down by shT2. Data are expressed as the mean ​± ​SD. ns, no statistical difference; ∗∗P ​< ​0.01, ∗∗∗∗P ​< ​0.0001.

Supplemental Fig. 2.

Cluster plot of the first three principal components for the Reman profiles of BMDCs. BMDCs transduced with control (shN) or Talin1-knockdown (shT) lentiviral vectors were treated with control medium (shN vs. shT) or LPS (1 ​μg/mL, shN ​+ ​L vs. shT ​+ ​L) for 12h. A, Cluster plot of the first three principal components for the analysis of BMDCs from groups of shN and shT. The corresponding first three loadings are shown on the right. B, Cluster plot of the first three principal components for the analysis of BMDCs from groups of shN ​+ ​L and shT ​+ ​L. And the corresponding first three loadings are shown on the right. The purple, green, red, and blue points in the plots represent samples from shN, shT, shN ​+ ​L, and shT ​+ ​L groups, respectively.

Supplemental Fig. 3.

RNA sequencing with enrichment analysis of BMDCs. BMDCs (n ​= ​3/group) transduced with control or Talin1-knockdown lentiviral vectors were treated with LPS (1 ​μg/mL, LPS vs. shTalin1-LPS) for 12h.The bubble plots of downregulated, upregulated and total differential genes of top 20 Rectome enrichment terms (A-C) and top 20 Wiki pathway terms (D-F) between groups of LPS and shTalin1-LPS.

Supplemental Fig. 4.

The ratio of key molecules of TLR4/MyD88/NF-κB pathway in BMDCs. BMDCs transduced with control (shN) or Talin1-knockdown (shT) lentiviral vectors were treated with LPS (1 ​μg/mL) for indicated times. Ratios of p-p65/p65 (A), p-IκB/IκB (B), and p-IRAK4/IRAK4 (C) in BMDCs with different treatment. Data are expressed as the mean ​± ​SD. ∗P ​< ​0.05.

Supplemental Fig. 5.

Knockdown of the Talin1 in BMDCs inhibits antigen-specific CD4+ T cell responses. BMDCs transduced with control or Talin1-knockdown lentiviral vectors were loaded with MOG_35-55 peptide and stimulated by LPS for 24h (shN ​+ ​L Vs. shT ​+ ​L). Then, BMDCs were co-cultured with spleen CD4⁺ T cells from EAE mice at a ratio of 1:5 for 72h. a, CFSE staining with flow cytometry examined CD4+T cells proliferation. b-d, Flow cytometry for the assessment of CD4⁺ T cell differentiation into Th1 (CD4+IFN-γ+), Th17 (CD4+IL-17A+), and Treg (CD4+CD25+Foxp3+) cells. Data are expressed as the mean ​± ​SD. NC, negative control; ∗P ​< ​0.05.

Supplemental Fig. 6.

Talin1-knockdown BMDCs suppresses the progression of EAE in mice. C57BL/6 mice induced into the EAE model were randomized and treated with saline (EAE, n ​= ​8) or BMDCs transduced with Talin1-knockdown lentiviral vectors (EAE ​+ ​shT, n ​= ​8) by tail vein injection. BMDCs for intervention in EAE models were loaded with MOG_35-55 peptide (20μg/ml) and stimulated by LPS (1ug/ml) for 24h. a, Flowchart of BMDCs therapeutic intervention in EAE mice. b-c, The changes of body weight and neurological scores of each group (EAE, blue symbol; EAE ​+ ​shT, green symbol). d-e, Comparison of maximal scores and cumulative scores between different groups. On day 20 post-immunization, mice in each group were sacrificed and their lumbar spinal cord tissue were collected for pathological analysis. f-g, j, HE staining and analysis of inflammation infiltration in the white matter of spinal cord (n ​= ​5–6; f, Scale bar ​= ​500 ​μm; g, Scale bar ​= ​50 ​μm). h-i, k, LFB staining and analysis of demyelination in the white matter of spinal cord (n ​= ​5–6; h, Scale bar ​= ​500 ​μm; i, Scale bar ​= ​50 ​μm). l, Western blot analysis of Talin1 expression in the cervical lymph nodes and spinal cord of EAE and normal control mice. Data are expressed as the mean ​± ​SD. # indicates P ​< ​0.05, compared to EAE group. Con, normal controls; ∗P ​< ​0.05, ∗∗P ​< ​0.01, ∗∗∗P ​< ​0.001, ∗∗∗∗P ​< ​0.0001.

Supplemental Fig. 7.

Talin1-knockdown BMDCs increased IL-10 derived from T cells in the spinal cord of EAE mice. On day 20 after induction, spinal cord tissue sections from EAE mice treated with saline (EAE) and Talin1-knockdown BMDCs (EAE ​+ ​shT) were stained with anti-CD11c (green) or anti-CD3 (green) antibodies and intracellular anti-IL-10 (red). Representative images from each group were shown. a, The percentage of IL-10-producing T cells (CD3⁺IL-10⁺) in the spinal cord of the EAE ​+ ​shT group was significantly higher than that in the EAE group (n ​= ​3; Scale bar ​= ​50 ​μm). b, The percentage of IL-10-producing DCs (CD11c⁺IL-10⁺) in the spinal cord of the EAE ​+ ​shT group was comparable to that in the EAE group (n ​= ​3; Scale bar ​= ​50 ​μm). Data are expressed as the mean ​± ​SD. ns, no statistical difference; ∗P ​< ​0.05.