Tolerogenic dendritic cell-mediated regulatory T cell differentiation by Chinese herbal formulation attenuates colitis progression

Graphical abstract

Highlights

• •CDD-2103, a standardized Chinese herbal formulation modified from si-jun-zi decoction was developed by our team and showed promising efficacy and safety profile in mouse models of colitis.
• •Using multiple strategies including co-transfer of naïve CD4⁺ CD45RB^hi and regulatory CD4 + Foxp3^GFP T cells as well as depletion Foxp3^DTR T cells in T-transferred Rag 1^-/- mice, we confirmed the suppressive effect of CDD-2103 was associated with an increase of colonic T_reg cell population.
• •Further single-cell transcriptomic analysis revealed the comprehensive picture of the immunological regulatory mechanism of CDD-2103 on T cell-mediated Rag 1^-/- mice. An increase of tolerogenic dendritic cells via regulating electron transport chain that related to oxidative phosphorylation was responsible for the increase T_reg differentiation in T-cell mediated colitis.
• •Our study also identified berberine and palmatine derived from CDD-2103 which particularly retain in mesenteric lymph node of colitis mice may contribute to the inhibitory effect of CDD-2103 on colitis.

Abstract

Introduction

Ulcerative colitis (UC) is a chronic inflammatory disease characterized by loss of immune tolerance to luminal antigens and progressive intestinal tissue injury. Thus, the re-establishment of immune tolerance is crucial for suppressing aberrant immune responses and UC progression.

Objectives

This study aimed to investigate the mechanisms underlying the action of CDD-2103 and its bioactive compounds in mediating immune regulation in mouse models of colitis.

Methods

Two experimental colitis models, chronic 2,4,6-trinitrobenzene sulfonic acid (TNBS)- and T-cell transfer-induced Rag1^-/- mice, were used to determine the effects of CDD-2103 on colitis progression. Single-cell transcriptome analysis was used to profile the immune landscape and its interactions after CDD-2103 treatment. Liquid chromatography-mass spectrometry (LC-MS) was used to analyze the major components interacting with lymphoid cells. A primary cell co-culture system was used to confirm the effects of bioactive component.

Results

CDD-2103 dose-dependently suppresses the progression of colitis induced by chemicals or T cell transplantation in Rag1^-/- mice. The effect of CDD-2103 is primarily attributable to an increase in the de novo generation of regulatory T cells (Tregs) in the lamina propria (LP). Single-cell transcriptomic analysis revealed that CDD-2103 treatment increased the number of tolerogenic dendritic cells (DCs). Mechanistically, CDD-2103 promoted tolerogenic DCs accumulation and function by upregulating several genes in the electron transport chain related to oxidative phosphorylation, leading to increased differentiation of Tregs. Further LC-MS analysis identified several compounds in CDD-2103, particularly those distributed within the mesenteric lymph nodes of mice. Subsequent studies revealed that palmatine and berberine promoted tolerogenic bone marrow-derived dendritic cells (BMDC)-mediated Treg differentiation.

Conclusion

Overall, our study demonstrated that the clinically beneficial effect of CDD-2103 in the treatment of UC is based on the induction of immune tolerance. In addition, this study supports berberine and palmatine as potential chemical entities in CDD-2103 that modulate immune tolerance.

Introduction

UC is a highly heterogeneous inflammatory intestinal disease that primarily affects the colon. Disease exacerbation may eventually lead to gut ulcerations and severe intestinal bleeding. The prevalence of UC in Eastern Asia has increased over the years, with approximately 1.5 million individuals in China living with the disease by 2025 [1]. The current treatment-to-target strategy for UC involves the induction and maintenance of clinical and endoscopic remission. However, most patients with UC constantly experience relapse and remission throughout the course of the disease, suggesting a modest clinical benefit achieved with current therapies. The inability to achieve sustained remission is associated with a high risk of relapse and the need for corticosteroid intervention or even colectomy [2]. Thus, there is an urgent need for a novel therapeutic approach to effectively maintain remission in patients with UC.

UC is primarily characterized by tissue injury caused by aberrant immune responses secondary to intestinal barrier breakdown, gut dysbiosis, and genetic mutations. Its pathogenesis is primarily attributed to dysregulated mucosal homeostasis as a result of the aberrant infiltration of gut-homing effector T cells and reduced immune tolerance, including a reduction in immunosuppressive clusters of differentiation 4⁺ forkhead box P3⁺ (CD4⁺Foxp3⁺) regulator T cells (Tregs). Recent single-cell analyses of colonic mucosal tissues from patients with UC have confirmed the expansion of T lymphocytes expressing CD8 and interleukin-17 (IL-17), along with disease activity [3]. Although only constituting–5–10 % of peripheral T cells are present, Tregs play a critical role in the maintenance of mucosal tolerance [4]. Impaired function and a reduced number of Tregs lead to disease exacerbation. Adoptive transfer of ex vivo-cultured polyclonal Treg cells for 12 weeks significantly improved the clinical response of patients with refractory UC. This transfer is accompanied by an increase in the intestinal Foxp3⁺ Treg population in the mucosal layer [5]. Clinical trials on visilizumab, the monoclonal antibody targeting on T-cell receptor (TCR) [6] and vedolizumab, an anti-α4β7 monoclonal antibody that blocking the T cell homing [7] have shown promising clinical outcomes on UC patients. However, a later study showed a minimal effect of visilizumab in patients who were refractory to corticosteroids, and patients experienced severe vascular and cardiac-related side effects [8]. Although the modulatory role of T lymphocytes in UC progression is compelling, a safe and effective Treg-targeted therapy for UC management is yet to be developed.

Accumulating evidence indicates the efficacy of Chinese herbal formulations in reducing the clinical symptoms of UC and improving the quality of life of patients [9]. As an alternative to the current mainstream treatment, the classical Chinese herbal formulation si-jun-zi decoction has long been used by the Chinese and Japanese communities for the treatment of gastrointestinal ulcerations [10]. CDD-2103, composed of nine herbs that were modified from the si-jun-zi decoction, was developed by our team and showed promising efficacy and safety profiles in mouse models of colitis. Palmatine and berberine are naturally occurring isoquinoline alkaloids with various pharmacological properties, and are the major compounds found in CDD-2103. Numerous studies have demonstrated that palmatine has anti-oxidant [11] and anti-inflammatory properties [12], [13]. A growing body of research has indicated that berberine exhibits a wide range of pharmacological activities, including anti-oxidant [14], anti-inflammatory [15], and modulation of intestinal microecology [16]. Although the formulation was developed based on previous clinical experience and knowledge of traditional Chinese medicine (TCM), a randomized trial with a placebo control is currently being carried out in Hong Kong to investigate the safety and efficacy of CDD-2103 (ChiCTR2100043200). Opening the label of this trial in the near future is expected to support the cost-effective use of CDD-2103 in Chinese patients with UC.

In this study, we systematically investigated the mechanism underlying the effects of CDD-2103 in a mouse model of colitis. We observed a significant reduction in colitis severity following CDD-2103 treatment of chronic 2,4,6-trinitrobenzene sulfonic acid TNBS-induced colitis and T-cell transfer-induced colitis in Rag1^-/- mice. Single-cell transcriptome analysis was used to profile the immune landscape of mice with colitis after CDD-2103 treatment. The results showed that the effect of CDD-2103 was dependent on tolerogenic dendritic cells (DCs) that favored the de novo generation of Tregs from CD4⁺ T cells. Additionally, palmatine and berberine may be responsible for the effects of CDD-2103. Our findings indicate that CDD-2103 and its chemical compounds are promising therapeutic agents for the management of UC by promoting immune tolerance.

Materials and methods

Animals

Wild-type C57BL/ 6 J mice (6–8 weeks old) were purchased from the Laboratory Animal Services Center of the Chinese University of Hong Kong, China. B6.129S7-Rag1^tm1Mom/J (Rag 1^-/-), B6.129(Cg)-Foxp3<^{tm3(DTR/GFP)Ayr}>/J (Foxp3), and B6.SJL-Ptprc^a Pepc^b/BoyJ (CD45.1) mice were purchased from Jackson Laboratory (Maine, USA). Mice were bred and maintained at the Hong Kong Baptist University Animal Facility under specific pathogen-free (SPF) conditions. All experiments were approved by the Committee on the Use of Human and Animal Subjects in Teaching & Research at Hong Kong Baptist University (REC/19–20/0296) and were performed in accordance with the Animals (Control of Experiments) Ordinance of the Department of Health, Hong Kong SAR, China.

Preparation of CDD-2103

CDD-2103 in granular form (batch no. KA00301) is a standardized formulation composed of Codonopsis pinosula (Franch.) Nannf. (Dang Shen), Poria cocos (Schw.) Wolf (Fu Ling), Atractylodis macrocephalae Koidz. (Fu Chao Bai Zhu), Glycyrrhizae uralensis Fisch. (Zhi Gan Cao), Paederia scandens (Lour.) Merr. (Ji Shi Teng), Bletilla striata (Thunb.) Rchb.f. (Bai Ji), Curcuma longa L. (Jiang Huang), Coptis chinensis Franch. (Huang Lian), and Cornus officinalis Sieb. et Zucc. (Shan Zhu Yu). The formulation has been patented (Chinese patent number: 202110927451.X). An Agilent 6470 Triple Quad Liquid Chromatography-Mass Spectrometry (LC-MS) system equipped with an electrospray ionization (ESI) ion source was used for fingerprinting in both negative and positive ion modes. Berberine, curcumin, loganin, and morroniside were used for quality control of CDD-2103.

Chemicals and reagents

Anti-mouse antibodies (Abs) were purchased from Biolegend (San Diego, CA) including anti-mouse CD4 (100407), CD3 (100214), CD3ε (100301), CD45RB (103305), and CD45 (157605). Brefeldin A (420601) and recombinant mouse transforming growth factor-β (TGF-β; 763102) were from Biolegend. APC-conjugated anti-mouse Foxp3 antibody was purchased from eBioscience (San Diego, CA, USA). Adalimumab (T9901) was purchased from TargetMol (Boston, MA, USA). Sulfasalazine (S129986) was purchased from Aladdin (Shanghai, China). Diphtheria Toxin (149A) was purchased from List Laboratories (CA, U.S.). 5-Aminosalicylic acid (5-ASA) (89–57-6), dithiothreitol (DTT; D0632), oligomycin (O4876), carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP; C2920), phorbol 12-myristate 13-acetate (PMA; P8139), ionomycin (I9657), paraformaldehyde (PFA; 30525–89-4), and antimycin A (A8674) were purchased from Sigma-Aldrich (U.S.). Ethylenediaminetetraacetic acid (EDTA; 17892) was purchased from Thermo fisher Scientific (Waltham, MA). The anti-PE MicroBeads (130–048-801) were purchased from Miltenyi Biotec (Germany). A Luminescent ATP Detection Assay Kit (ab113849) was purchased from Abcam (Cambridge, UK). Diphtheria toxin (DT, CAY19657) was purchased from Cayman Chemical Company (Michigan, US).

Bone marrow-derived dendritic cell (BMDC) culture

BMDCs were generated from the bone marrow of C57BL/6J mice [17]. Briefly, the mice were euthanized, and the femurs and tibiae were isolated to prepare bone marrow cells. Red blood cells were removed using a lymphoma density gradient medium. Bone marrow cells were cultured for 6 d in a complete Roswell Park Memorial Institute (RPMI) 1640 medium containing 10 % fetal bovine serum (FBS), 10 ng/mL recombinant mouse granulocyte–macrophage colony-stimulating factor (GM-CSF) (RP-8620, Invitrogen) and 10 ng/mL recombinant mouse IL 4 (404-ML/CF, RD) in 37 °C, 5 % CO₂ incubator. On day 7, loosely adherent cells were harvested as immature BMDCs and used in subsequent experiments. Immature BMDCs (5 × 10⁴) were seeded into U-shaped 96-well plates and stimulated with a series of CDD-2103 concentrations in the presence or absence of 10 nM rotenone (HY-B1756, MCE) for 48 h. The cells were collected for phenotypic analysis using flow cytometry. Supernatants were harvested for cytokine detection. In the DC-T co-culture experiments, purified unfractionated CD4⁺ T cells were co-cultured with BMDCs pre-treated with CDD-2103 at a ratio of 1:3 or 1:10 (BMDCs: CD4⁺ T cells) for 3 d. Referring to the DC-T indirect contact co-culture experiment, a 0.4-µm tissue culture plate insert (TCS-001012, BIOFIL) was used to separate the BMDCs and unfractionated CD4⁺ T cells.

Seahorse assay

The bioenergetic function of BMDCs was measured using an Agilent Seahorse XFe96 Analyzer (Seahorse Biosciences, USA) [18]. BMDCs were cultured in the six-well plate in the presence or absence of CDD-2103 for 48 h in 37 °C, 5 % CO₂ humidified incubator. The XF extracellular flux sensory cartridge was hydrated one day prior to the assay. BMDCs were harvested after 48 h and resuspended in XF RPMI assay medium (pH 7.4) supplemented with 100 mM sodium pyruvate, 1.0 M glucose, and 200 mM glutamine. Next, 35,000 cells were inoculated into the cell culture plate of the Seahorse analyzer and incubated at 37 °C without CO₂ for 1 h. Finally, the plate was loaded into the XFe96 Analyzer and Oligomycin (0.5 μM), FCCP (4 μM), and the mixture of rotenone and antimycin A (1 μM) were injected sequentially according to the instrument setting procedure, and the data analysis was performed.

Isolation of myeloid cells from the lamina propria (LP) and mesenteric lymph nodes

Single cell suspensions from the spleen and mesenteric lymph nodes (MLN) were prepared by filtration through a 70-µm cell strainer (BD Labware, San Jose, CA, USA). Spleen cells were lysed using 1X red blood cell (RBC) lysis buffer (00–4333-57; Invitrogen). CD4⁺ T cells were purified from single-cell suspensions of the spleen and MLN using MACS cell separation technique. Single-cell suspensions of the spleen and MLN were labelled with anti-CD4-PE antibody and anti-PE conjugated with microbeads before being applied to a Miltenyi LS column (130–122-729, Miltenyi Biotec). CD4⁺ cells were collected as the flow-through cells. LP-derived myeloid cells were isolated as previously described [19]. Briefly, the colon was rinsed with phosphate-buffered saline (PBS) and cut into approximately 0.3-cm pieces. The intestinal epithelial cells were removed by incubation in PBS supplemented with 5 mM EDTA and 1 mM DTT. Then, colon tissues were washed with PBS and further digested using RPMI 1640 containing 10 % FBS, 600 U collagenase type 4 (17104019, Gibco) and 0.15 mg/mL RNAse I (11284932001, Roche) for 45 min at 37 °C. The LP cells were centrifuged using a 40/80 Percoll density gradient (17089101, Cytiva). The fatty tissue floating on the liquid surface was removed, and the LP cells at the interface were carefully collected in a new tube for further analysis.

T Cell activation and proliferation assay

CD4⁺ T cells were enriched from the spleen of C57BL/6J mice and labelled with CellTrace™ CFSE (C34554, Invitrogen). Cells (1x10⁵) were seeded in a u-shaped 96-well plate and stimulated with a series concentration of CDD-2103 (31.25, 62.5, 125, 250, and 500 µg/mL) in the presence of mouse CD3/CD28 Dynabeads (11452D, Gibco) for 72 h. The cells were harvested and analyzed using a BD FACSCelesta cell analyzer. In the T-cell activation assay, the supernatant was collected, and the levels of cytokines were examined using the LEGENDplex Mouse Inflammation Panel kit (740150, Biolegend) according to the manufacturer’s instructions.

Treg differentiation assay

Purified CD4⁺ T cells (1x10⁵ p) were plated in a 5-µg/mL anti-CD3 antibody-pre-coated U-shaped 96-well plate. Next, the cells were treated with a series concentration of CDD-2103 (15.6, 31.25, 62.5, and 125 µg/mL) in the presence of recombinant TGF-β (0.5 ng/mL) in 37 °C, 5 % CO₂ incubator. After culturing for 72 h, the cells were collected and permeabilized using Foxp3/Transcription Factor Fixation/Permeabilization solution. The percentage of CD4⁺Foxp3⁺ cells was determined using a BD FACSCelesta cell analyzer.

Flow cytometry analysis

The cells were then incubated with the appropriate diluted antibodies at 4 °C for 30 min. The antibody dilution information is provided in Supplementary Table 1. For intracellular cytokines staining, cells were stimulated with 50 ng/mL PMA, 1 µg/mL ionomycin, that activate T cell non-specifically and 5 µg/mL brefeldin A which enhanced detection of intracellular cytokines, at 37 °C, 5 % CO₂ incubator for 4 h and processed by Foxp3/Transcription Factor Fixation/Permeabilization Kit (00–5523-00, Invitrogen). Finally, the cells were analyzed using BD Celesta (BD Biosciences, Mountain View, CA, USA), and data analysis was performed using FlowJo software (Tree Star Inc., Ashland, OR, USA).

Single-cell clustering, annotation, and relationship analyses

For the scRNA-seq of CD45⁺ cells, colon tissues from the vehicle and CDD-2103-treated Rag1^-/- mice (12 weeks old) were pooled from six biological replicates and processed as discussed mentioned. FACS-sorted CD45⁺ cells were encapsulated into droplets and pooled libraries were prepared using the 10 × Genomics platform. Libraries were sequenced using DNBSEQ PE100. The output file was aligned with mm10 (Cell Ranger mouse reference genome mm10-2020) and filtered. Following quality control using Seurat in R, the filtered cells (8955 cells in the vehicle group and 9734 cells in the CDD-2103 group) were collected. The thresholds were set as follows: 1) gene number in each cell ranging 200–6000, 2) percentage of mitochondrial DNA-derived or ribosome-associated genes < 20 %), and 3) hemoglobin-related genes < 5 %. The gene expression was processed by DESeq2-related normalization and uniform manifold approximation and projection (UMAP) clustering using the “Dimplot” function. Fourteen cell clusters (Cluster 0–13) were identified and determined using PanglaoDB according to the representative cell markers in each group. The top five highly-weighted markers of each cluster were shown in the heatmap (p < 0.001) using the “FindAllMarkers” function. The relationship between clusters was determined using “lm” linear regression function and visualized using the “ggplot” package in R software.

Evaluations of colitis severity by disease activity index (DAI)

Body weight, stool consistency, and fecal occult blood were monitored to evaluate the severity of colitis. The DAI was calculated based on weight loss, stool consistency, and bloody stool using a standard protocol [20] (Table 1).

Table 1.

DAI.

Weight loss (%)	Stool consistency	Bloody stool[21]	Score
0	Normal	None	0
1–5	Soft and shaped	weak positivity (±)	1
5–10	Loose stools	Positive bleeding (+)	2
10–15	Liquid stools	Strong positivity (++)	3
>15	diarrhea	Gross bleeding	4

Hematoxylin and eosin (H&E) analysis

For histopathological analysis, the 5-μm PFA-embedded colon tissues were sectioned, and stained with H&E (G1120, Solarbio). Histopathological scores were assessed using a previously described [22]. Histology was scored on a four-point scale, including the degree of epithelial injury and inflammatory infiltrates, as shown in Table 2.

Table 2.

Histopathological scoring scheme.

Inflammatory cell infiltrate Extent	Epithelial changes	Score
None	None	0
Basal 1/3 crypt damage or minimal goblet cell loss	Mucosal infiltration	1
Basal 2/3 crypt damage or goblet cell loss	Mucosal and submucosal infiltrate of inflammatory cells	2
Ulceration with only surface epithelium intact	Mucosal and submucosal infiltration with crypt abscesses	3
Epithelium loss	Transmural inflammatory cells	4

Animal experiments

• (i)TNBS-induced colitis C57BL/6 mice (8–9 weeks old) were used to construct the TNBS-induced colitis model [23]. Briefly, 1 mg of TNBS (P2297, Sigma-Aldrich) in 50 % ethanol was administered intrarectally into the mouse rectum through a mouse intragastric needle No.10. On day 7, DAI was tested and mice were randomly grouped into five groups that receiving either vehicle, 100 mg/kg 5-ASA, 1.94 g/kg CDD-2103, 3.88 g/kg CDD-2103, or 7.76 g/kg CDD-2103. 5-ASA, the first-line medication for treating mild-to-moderate UC in clinical settings, was used as a positive control in this model. The human equivalent dose of 5-ASA for managing UC is 1.5 g/d, which is converted to a mouse equivalent dose of approximately 308 mg/kg. Previous studies reported that a dose of 5-ASA as low as 50 mg/kg showed significant efficacy in ameliorating TNBS-induced colitis [24], [25]. Therefore, we selected a 5-ASA dose of 100 mg/kg. The dosage of CDD-2103 was determined based on our ongoing randomized placebo-controlled clinical trial in Hong Kong (ChiCTR2100043200). In this trial, the crude dose of CDD-2103 was 90 g/d. Because 3.02 g of crude drug yields 1 g of CDD-2103 powdered substances, the human-to-mouse dosage conversion equates to 3.88 g/kg/d of the powdered substances of CDD-2103. Our study therefore utilized doses of 1.94, 3.88, and 7.76 g/kg as low, middle, and high doses, respectively. On days 8 and 15, 2 mg TNBS in 50 % ethanol was repeatedly administered to the mice. Daily oral administration was initiated on day 8 and continued until day 19. The animals were monitored every 4 d for weight loss, diarrhea, and bloody stools. On day 20, the mice were sacrificed. Blood and colon samples were collected, and the length of the colon was recorded.
• (ii)T-cell transfer model of colitis: Single cell suspensions were prepared from the spleen and MLN of C57BL/6J mice and CD4⁺CD45RB^hi T cells were sorted using BD FACSAria™ III Cell Sorter. Naive CD4⁺CD45RB^hi T cells (5 × 10⁵) were intraperitoneally injected to Rag1^-/- mice [26]. After T-cell adoptive transfer, mice were grouped into five groups that receiving either vehicle, 25 mg/kg infliximab, 1.94 g/kg CDD-2103, 3.88 g/kg CDD-2103, or 7.76 g/kg CDD-2103. Infliximab, approved for moderate-to-severe UC, is a monoclonal antibody that neutralizes TNF-α and has been shown to reverse disease severity in the T-cell transfer model of colitis [27] was used as positive control. Naive mice did not undergo T cell transfer. Oral administration of CDD-2103 was initiated once daily on day 1. Mice in the positive control group were intraperitoneally administered infliximab (25 mg/kg) twice weekly. The animals were monitored weekly for weight loss, diarrhea, and bloody stools.
• (iii)Treg depleted Rag 1^-/- mouse model: 5 × 10⁵ CD4⁺CD45RB^hi T cells purified from the spleen and MLN of B6.129(Cg)-Foxp3^{<tm3(DTR/GFP)Ayr>}/J donor mice were intraperitoneally injected into Rag 1^-/- mice to construct a colitis model. During the experiment, 10 µg/kg DT was intraperitoneally injected into Rag 1^-/- mice every 7 d from week 1, onward. B6.129(Cg)-Foxp3^{<tm3(DTR/GFP)Ayr>}/J mice, which are Foxp3^DTR knock-in mice. These mice contained a human DTR and GFP inserted downstream of the internal stop codon of the X-linked Foxp3 gene. Foxp3⁺ Tregs in these mice highly express DTR-GFP. Administration of DT to these mice resulted in the depletion of Treg cells in lymphoid tissues within 2 d post-injection.
• (iv)Teff and Treg co-transfer Rag 1^-/- mouse model: Naïve CD4⁺ CD45RB^hi T cells and CD4⁺ Foxp3^GFP Treg cells were prepared and sorted from B6 CD45.1 mice and Foxp3^DTR-GFP mice (CD45.2), respectively. Naïve CD4⁺ CD45RB^hi T cells (5 × 10⁵) and CD4⁺ Foxp3^GFP Tregs (2 × 10⁴) were mixed and intraperitoneally injected into Rag 1^-/- mice to induce colitis.

The blood and colon were harvested, and the length of the colon was recorded. Flow cytometry analysis was conducted to examine the changes in immune cells in the colon and lymphoid tissues. Colon samples were promptly fixed in 4 % PFA and embedded in paraffin.

Ethics statement

All experiments were approved by the Committee on the Use of Human and Animal Subjects in Teaching & Research at Hong Kong Baptist University (REC/19–20/0296) and were performed in accordance with the Animals (Control of Experiments) Ordinance of the Department of Health, Hong Kong SAR, China.

LC-MS analysis

Cells were isolated from MLN of CDD-2103-treated Rag1^-/- mice. The cells were harvested in 10 mL absolute methanol and frozen at −20 °C overnight. Next, samples were centrifuged at 15,000 g for 15 min at 4 °C and the supernatant was collected to evaporate the residual methanol. Finally, the sediments were thoroughly re-dissolved in 60 μL absolute methanol.

Compounds in the cells were determined using ultra-high-performance liquid chromatography (UPLC) with a Triple Quadrupole Mass Spectrometer (Agilent, CA, USA). The Mass Spectrometer was operated in both positive and negative ESI modes with multiple reaction monitoring (MRM) mode. MS data were acquired using a Dual Agilent Jet Stream ESI source operating at capillary voltages of 3500 V (positive) and 4000 V (negative). The nozzle voltage of the positive polarity was 500 V, and that of the negative polarity was 1000 V. The nitrogen drying gas temperature was maintained at 300 °C and the flow rate was set at 7 L/min. The nebulizer pressure was set to 45 psi. The sheath gas temperature was 350 °C, and the flow rate was 10 L/min.

The chromatographic separation was performed using a 100 mm × 2.1 mm Acquity UPLC BEH C18 1.7-μm column with temperature maintained at 40 °C. The injection volume of the standard was 1 μL with 3 μL of each sample. Mobile phase A was 0.1 % formic acid in H₂O and mobile phase B was 0.1 % formic acid in acetonitrile. The gradient was as follows: 5–10 % B 0–2 min, 10–15 % B 2–8 min, 15–17 % B 8–10 min, 17–20 % B 10–16 min, 20–50 % B 16–21 min, 50–54 % B 21–24 min, 54–100 % B 24–25 min, and 100 % B 25–28 min, with flow maintained at 0.4 mL/min. The MassShunter software (version 10.00) was used for data acquisition and processing. Each compound was identified based on its retention time and the corresponding mass spectra of the reference standard compounds. Table 3 summarizes the specific MRM transitions and MS/MS parameters of the compounds identified in the MLN of colitis mice treated with CDD-2103.

Table 3.

Nine compounds in MLN of CDD-2103-treated colitis mice identified using LC-MS.

Compound name	Formula	m/z	MRM transitions	RT (min)	Fragmentor (V)	Ion type
Glycyrrhizic acid	C42H62O16	823.4	823.4 → 453.4	20.826	170	[M−H]+
Palmatine	C21H22NO4+	352.2	352.2 → 336.1, 308.0	17.503	170	[M]+
Magnoflorine	C20H24NO4+	342.2	342.2 → 297.1, 265.0	5.419	130	[M]+
Columbamine	C20H20NO4+	338.1	338.1 → 323.1, 294.1	13.288	170	[M]+
Jatrorrhizine	C20H20NO4+	338.1	338.1 → 294.1, 265.0	13.716	170	[M]+
Berberine	C20H18NO4+	336.1	336.1 → 320.1, 292.1	17.229	170	[M]+
Epiberberine	C20H18NO4+	336.1	336.1 → 320.1, 292.1	13.397	170	[M]+
Coptisine	C19H14NO4+	320.1	320.1 → 292.0, 262.0	12.995	170	[M]+
Atractylenolide III	C15H20O3	249.1	249.1 → 231.1, 163.0	22.213	50	[M−H]+

Statistical analysis

Statistical analyses were conducted using GraphPad Prism 8 software and all experimental values were expressed as mean ± standard error of the mean (SEM). The significance level was determined using a two-tailed Student’s t-test. Statistical significance was set at P < 0.05.

Results

CDD-2103 suppresses TNBS-induced colitis accompanied increase in DCs and CD4⁺ T cells in the LP

To investigate the efficacy of CDD-2103 in chronic colitis, we established a TNBS-induced chronic colitis mouse model in which C57BL/6 mice were stimulated with 1 mg TNBS on day 1 and induced with 2 mg TNBS on days 8 and 15. The oral administration of CDD-2103 was initiated on day 8 and continued until day 19 (Fig. 1a). Although mice treated with three doses of CDD-2103 showed minimal changes in body weight compared to the vehicle control group (Fig. 1b), the DAI, as determined by the sum score of body weight change, fecal occult blood, and stool consistency, was significantly lower in the high-dose CDD-2103-treated mice than in the vehicle group on days 11 and 18. This was particularly evident from the reduction in stool consistency (Fig. 1c). At the end of the experiment, the mice were sacrificed, and colon length was measured from the cecum to the end of the distal colon. The colon length of the middle and high dose CDD-2103 groups was significantly longer than that of the vehicle group, suggesting an anti-colitis effect of CDD-2103 (Fig. 1d). Histological evaluation consistently showed that TNBS-treated mice exhibited significant crypt distortion and goblet cell loss, suggesting chronic tissue injury. However, mice treated with middle and high doses of CDD-2103 showed alleviated epithelial damage and preserved colonic architecture. A high dose of CDD-2103 was more effective than 5-ASA in improving the disrupted histology (Fig. 1e). Flow cytometry analysis of LP-derived myeloid cells revealed that treatment with CDD-2103 led to a significant increase in CD11C⁺ DCs and CD4⁺ T cells, whereas other cell types, including CD19⁺ B cells and F4/80⁺ macrophages, remained unchanged (Fig. 1f). Taken together, these results indicate that CDD-2103 maintains disease remission by reducing disease severity in mice with chronic colitis, and that this effect may be associated with an increase in DCs and CD4⁺ T cells in the LP.

Fig. 1.

CDD-2103 suppresses TNBS-induced colitis accompanied by the increase of DCs and CD4⁺ T cells in lamina propria. (A) The schematic diagram of TNBS-induced chronic colitis mice model. (B) The body weight change; (C) disease activity index; and (D) colon length of the chronic TNBS-induced C57BL/6 mice following CDD-2103 treatment. (E) The representative images of H&E and Alcian blue staining of colon from chronic TNBS-induced C57BL/6 mice. The histological score was shown on the right. (F) Quantification of lamina propria-derived CD45⁺ cells (as percentage of live cells), CD3⁺ CD4⁺ T cells, CD11c⁺ DCs, CD11b⁺ F4/80⁺ macrophages, CD19⁺ B cells (as percentage of CD45⁺ cells) in chronic TNBS-induced C57BL/6 mice (n = 6–8 per group). Data were expressed as mean ± SEM. ∗ p < 0.05, ∗∗p < 0.01 versus vehicle group and n.s., no significant differences.

CDD-2103 orchestrates regulatory T cell (Treg) differentiation

Considering that UC is highly dependent on T cell homeostasis, we further investigated the role of CDD-2103 in T cells by adopting the T cell transfer Rag1 mouse colitis model, which recapitulates human diseases. To establish this model, CD4⁺CD45RB^hi cells generated from wild-type donor mice were intraperitoneally transferred to Rag1^-/- recipient mice, which lacked B and T lymphocytes (Fig. 2a, 0.5 % naïve Foxp3⁺ cells in total transferred cells). Administration of high doses of CDD-2103 to Rag1^-/- mice significantly improved body weight loss starting from week 3 post-T cell transfer, and all doses of CDD-2103 reduced disease activity severity at week 5 (Fig. 2b). Mice treated with CDD-2103 exhibited longer colons than the vehicle group, especially at the middle and high doses of CDD-2103 (Fig. 2d). Histological analysis showed that both doses of CDD-2103 improved T cell-induced histological injury, and the high dose showed greater efficacy in promoting histological restoration compared to the mice treated with infliximab (Fig. 2c). Flow cytometry analysis revealed a significant increase in the percentage of Foxp3⁺-expressing CD4⁺ T cells in the LP of Rag1^-/- mice and no significant change in the mesenteric lymph nodes compared to vehicle-treated mice by the end of the experiment (Fig. 2e). Next, we investigated whether CDD-2103 induces Treg proliferation or the differentiation of naïve T cells. Thus, we established another T cell co-transfer model in which naïve CD4⁺ CD45RB^hi from B6 CD45.1 mouse and regulatory CD4⁺ Foxp3^GFP T cells from Foxp3^DTR-GFP mice at a ratio of 20:1 was intraperitoneally co-transferred to Rag1^-/- recipient mice, and post-transferred mice were orally administered CDD-2103 daily (Fig. 2f). The vehicle-treated mice developed gradual body weight loss and disease activity starting from week 6 following T cell transfer, suggesting a delayed disease onset compared to their counterparts that were transferred with only naïve CD4⁺ CD45RB^hi cells. However, mice treated with daily CDD-2103 showed a significant reversal in body weight loss and disease severity (Fig. 2g). CDD-2103 consistently ameliorated colitis in terms of colon length (Fig. 2i) and colon tissue histology (Fig. 2h). Notably, CDD-2103 exhibited a minimal effect on the population of CD4⁺ CD45.2⁺ cells while significantly enhancing the frequency of CD45.2^- CD4⁺ Foxp3⁺ T cells in the LP (Fig. 2j), suggesting that CDD-2103 did not promote Treg proliferation but promoted Treg differentiation. Collectively, these data suggest that CDD-2103 ameliorates experimental colitis by promoting Treg differentiation.

Fig. 2.

CDD-2103 orchestrates in vivo Treg differentiation but not proliferation. (A) The schematic diagram of CD4⁺ CD45RB^hi T-transferred Rag1^-/- mice model. (B) The body weight change and disease activity index; (C) the representative images of H&E staining of colon and the histological score were shown on the right; and (D) colon length of CD4⁺ CD45RB^hi T-transferred Rag1^-/- mice following CDD-2103 treatment. (E) Quantification of mesenteric lymph node and lamina propria-derived CD4⁺ Foxp3⁺ T cells (as percentage of CD45⁺ cells) in CD4⁺ CD45RB^hi T-transferred Rag1^-/- mice (n = 7–8 per group). (F) The schematic diagram of CD4⁺ CD45RB^hi and CD4⁺Foxp3⁺ T-co-transferred Rag1^-/- mice model. (G) The body weight change and disease activity index; (H) the representative images of H&E staining of colon and the histological score was shown on the right; and (I) colon length of CD4⁺ CD45RB^hi and CD4⁺ Foxp3⁺ T-co-transferred Rag1^-/- mice following CDD-2103 treatment. (J) Quantification of lamina propria-derived CD4⁺ T cells (as percentage of CD45.2⁺ cells) and CD4⁺Foxp3⁺ T cells in CD4⁺ CD45RB^hi and CD4⁺Foxp3⁺ T-co-transferred Rag1^-/- mice (n = 6–7 per group). Data were expressed as mean ± SEM. ∗ p < 0.05, ∗∗p < 0.01 versus vehicle group and n.s., no significant differences.

Depletion of Tregs abolishes the effect of CDD-2103 on colitis

The Treg population seemingly primarily supports the effect of CDD-2103 on colitis suppression. We thus examined the efficacy of CDD-2103 following Treg depletion by injecting DT to Rag1^-/- recipient mice transferred with naïve CD4⁺ CD45RB^hi T cells from Foxp3^DTR-GFP donor mice (Fig. 3a). Weekly DT injections significantly abolished the effects of CDD-2103 (Fig. 3b). T-cell-depleted mice treated with CDD-2103 exhibited a significantly shorter colon length and distorted colonic crypt architecture as compared to T non-depleted counterparts (Fig. 3c and d). These data confirmed that the inhibitory effect of CDD-2103 on experimental colitis was Treg-dependent. To determine whether CDD-2103 directly supports the role of T cells and regulates T cell differentiation, we incubated splenocyte-derived CD4⁺ T cells in vitro with serial doses of CDD-2103. At doses of 250 and 500 µg/ml, CDD-2103 significantly reduced T cell viability (Fig. 3e). However, non-toxic doses of CDD-2103 showed minimal effects on CFSE⁺ CD4⁺ T cell proliferation and activation as observed from T-cells produced IFN-γ and IL-6 (Fig. 3f and g). Furthermore, treatment with CDD-2103 resulted in minimal changes in Foxp3 expression compared to vehicle-treated cells (Fig. 3h). Thus, the T cell-dependent anti-colitis effect of CDD-2103 was not attributed to the direct action of CDD-2103 on effector T cells.

Fig. 3.

Depletion of Treg abolish the effect CDD-2103 on colitis, but CDD-2103 showed no direct effect on T cell function. (A) The schematic diagram of Tregs depletion in CD4⁺ CD45RB^hi T-transferred Rag1^-/- mice model. (B) The body weight change and disease activity index; and (C) colon length of CD4⁺ CD45RB^hi T-transferred Rag1^-/- mice following CDD-2103 treatment with or without intraperitoneal injection of diphtheria toxin (10 μg/kg). (D) The representative images of H&E staining of colon and the histological score is shown on the right. (E) CD4⁺ T cells were treated with serial dose of CDD-2103 in vitro. After 72 h, MTT assay was applied to test the cell viability. (F and G) CFSE -labelled CD4⁺ T cells were treated with serial dose of CDD-2103 in the presence of mouse CD3/CD28 dynabeads. After 72 h, the proliferation of T cell was analyzed by flow cytometry (F) and the concentration of IFN-γ and IL-6 in the supernatant were detected by LEGENDplex mouse Inflammation Panel kit (G). (H) CD4⁺ T cells were stimulated with anti-CD3 in the presence of TGFβ. After treatment with or without CDD-2103 for 72 h, Foxp3⁺ Treg were analyzed by flow cytometry. Typical FACS plots (gating on CD4⁺ cells) and a summary of the percentage of Foxp3⁺ Treg was shown in the right. Data are expressed as mean ± SEM. *p < 0.05, **p < 0.01, CDD-2103 versus vehicle; ^#p < 0.05, CDD-2103 versus CDD-2103 + DT and n.s., no significant differences.

scRNA analysis reveals the involvement of DCs in CDD-2103-mediated Treg differentiation

To gain a deeper understanding on how CDD-2103 promotes Treg differentiation, we performed single-cell RNA sequencing on Fluorescence-activated cell sorting (FACS)-sorted CD45⁺ cells from the LP of Rag1^-/- mice receiving either vehicle or CDD-2103 (Fig. 4a). Unsupervised hierarchical clustering was performed on 8955 and 9734 cells from vehicle- and CDD-2103-treated mice, respectively, and UMAP identified 13 clusters in an unbiased manner (Fig. 4b). Analysis of the percentile distribution of the clusters demonstrated that clusters 1 and 8 were particularly enriched in CDD-2103-treated mice, whereas cluster 10 was enriched in vehicle-treated mice (Fig. 4c). Based on differential marker gene expression, we identified FcRg-expressing DCs (cluster 1) and tnfrsf4 and Il2ra-expressing Tregs (cluster 8), whereas cluster 10 expressed high levels of cytotoxic genes, including gzma, gzmb, and klrd1 (Fig. 4d). In addition, a significant positive correlation was observed between the genes in clusters 1 and 8 in the CDD-2103-treated mice (Fig. 4e). Further analysis of the correlated gene expression data revealed that several Treg transcripts significantly correlated with DCs. These transcripts are associated with various Treg processes, including Treg cell antigen processing (H2-ab1, H2-aa, and Cd74), protein folding (HSPs, Pdrg1, and Ahsa1), activation (Cd3d and Slamf7), and adhesion (Itga4, Itgb2, and Tln2) (Fig. 4f). Importantly, gene ontology analysis of the biological processes revealed that cluster 1 was enriched with genes involved in anti-inflammatory IL-10 production, tolerance induction, and defense responses to bacteria (Fig. 4g). This indicates the typical features of tolerogenic DCs.

Fig. 4.

Scrna analysis reveals the involvement of dendritic cells in cdd-2103-mediated treg differentiation. (A) The schematic diagram of CD4⁺ CD45RB^hi T-transferred Rag1^-/- mice model. (B) UMAP plot of CD45⁺ cell showing annotated clusters in vehicle-treated and CDD-2103-treated mice. (C) Staggered bar plot of annotated cell clusters between vehicle and CDD-2103- treated mice. (D) Heat map showing scaled expression of DEGs among 13 clusters of annotated cells. (E) Correlation between the expressing genes in cluster 1 and 8 from CDD-2103-treated mice. (F) Correlations of curated genes with Treg. (G) Enriched Gene Ontology (GO) analysis on biological process in cluster 1.

CDD-2103-induced DCs exhibit an immuno-tolerogenic phenotype

We further characterized DCs population from the mesenteric lymph nodes and LP of Rag1^-/- mice following CDD-2103 treatment. CD11c⁺ MHCII⁺ DCs were significantly enriched in the MLN and LP of mice treated with CDD-2103 (Fig. 5a and 5c). CD11c⁺ MHC-II ⁺ DCs in CDD-2103-treated colitis mice expressed reduced levels of the costimulatory molecule CD86 (Fig. 5b and d). One characteristic of tolerogenic DCs is their poor immunogenicity and low expression of costimulatory molecules [28]. Based on these observations and the scRNA-seq data, we sought to determine whether CDD-2103 increased DCs tolerogenic activity. We induced DCs in vitro from the bone marrow hematopoietic progenitors of C57BL/6 mice using GM-CSF and IL-4 containing medium and then with or without CDD-2103 intervention for 2 d (Fig. 5e). Non-toxic dose CDD-2103-treated BMDCs presented reduced expression of the costimulatory molecule CD86 (Fig. 5f). We examined whether CDD-2103 enhanced the cognate interaction between DCs and unfractionated CD4⁺ T cells, which further promoted the differentiation of unfractionated CD4⁺ T cells into Tregs in vitro. GM-CSF- and IL-4-cultured BMDCs were co-cultured with CD4⁺ T cells sorted from splenocytes at ratios of 1:3 and 1:10 for 3 d. It was observed that CDD-2103-pre-treated DCs significantly enhanced the differentiation of unfractionated T cells into Tregs (Fig. 5g, 11.37 % of Foxp3⁺ cells in unfractionated cells), and this phenomenon was not significant in the transwell co-culture system of DCs and unfractionated CD4⁺ T cells (Fig. 5h). These results indicate that CDD-2103 strengthens the interaction between BMDCs and CD4⁺ T cells, triggering antigen-specific Treg differentiation in a contact-dependent manner.

Fig. 5.

CDD-2103-induced DCs exert an immuno-tolerogenic phenotype. (A, B) Quantification of mesenteric lymph node-derived CD11c⁺ MHCII⁺ DCs gating on CD45⁺ cells (A) and CD86 expression gating on CD11c⁺ MHCII⁺ DCs (B) in CD4⁺ CD45RB^hi T-transferred Rag1^-/- mice (n = 6–7 per group). (C, D) Quantification of lamina propria-derived CD11c⁺ MHCII⁺ DCs gating on CD45⁺ cells (C) and CD86 expression gating on CD11c⁺ MHCII⁺ DCs (D) in CD4⁺ CD45RB^hi T-transferred Rag1^-/- mice (n = 5–7 per group). (E) The schematic diagram of the generation of BMDC. After treatment with serial dose of CDD-2103 for 48 h and MTT assay was applied to test the viability of BMDC. (F) The expression of CD86 of BMDC with or without CDD-2103 was analyzed by flow cytometry. (G, H) BMDCs pretreated with CDD-2103 or vehicle were co-cultured with unfractionated CD4⁺ T cells at a ratio of 1:3 or 1:10 (BMDCs: CD4⁺ T cells) for 3 days without (G) or with 0.4 µm tissue culture plate insert (H). Cells were collected and the percentage of CD4⁺Foxp3⁺ Tregs were detected by using flow cytometry. Data were expressed as mean ± SEM. ^∗p < 0.05, ^∗∗p < 0.01 versus indicated group and n.s., no significant differences.

CDD-2103-induced oxidative phosphorylation promotes DCs activation and function

Next, we characterized the signaling pathways involved in the DC-mediated Treg differentiation. Differential gene expression analysis between DCs (cluster 1) from CDD-2103- and vehicle-treated chronic Rag1^-/- colitis mice showed that there were 538 genes up-regulated, whereas 781 genes were downregulated following CDD-2103 treatment (Fig. 6a). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of CDD-2103-regulated differential genes in DCs revealed the downregulation of several inflammation-related pathways, including the MAPK, TNF, and NF-kappa B (NF-κB) signaling pathways, whereas oxidative phosphorylation and antigen processing and presentation were found to be upregulated (Fig. 6b). Gene set enrichment analysis revealed that oxidative phosphorylation was one of the most upregulated pathways after CDD-2103 treatment (Fig. 6c). Within the oxidative phosphorylation (OXPHOS), the major transcripts related to the complex I-V in electron transport chain were up-regulated in DCs from CDD-2103-treated mice (Fig. 6d). Tolerogenic DCs show a metabolic profile switch to OXPHOS, which further maintains tolerogenic cytokines and increases Treg induction [29]. Therefore, we systematically assessed whether CDD-2103 induces tolerogenic properties in DCs by enhancing their mitochondrial metabolism. We analyzed the oxygen consumption rate (OCR) of CDD-2103-treated BMDCs at 48 h and the seahorse profile showed that the DCs respiration rate was significantly enhanced following CDD-2103 treatment (Fig. 6e). The CDD-2103-treated BMDCs exhibited increased intracellular ATP production (Fig. 6f) and ROS generation (Fig. 6g). Mitochondrial ROS levels significantly increased, as determined by MitoSOX staining (Fig. 6h). These data indicated that CDD-2103 enhanced mitochondrial respiration in DCs. To further examine whether the CDD-2103-induced OXPHOS activation in DCs supports its function, we incubated DCs with rotenone, a small molecule that interferes with complex I in the OXPHOS electron transport chain. It was observed in the CD86 of CDD-2103-treated DCs was significantly inhibited in the presence of rotenone (Fig. 6i). The co-culture of CDD-2103 and DCs with pre-treated rotenone showed a reduced capacity to promote Treg differentiation in unfractionated CD4⁺ T cells (Fig. 6j). Collectively, our data demonstrate that CDD-2103-induced OXPHOS in DCs leads to DC-mediated Treg differentiation.

Fig. 6.

CDD-2103-induced oxidative phosphorylation promotes DCs activation and function. (A) Volcano plot of differentially expressed genes between DCs (cluster 1) from CDD-2103 and vehicle-treated CD4⁺ CD45RB^hi T-transferred Rag1^-/- colitis mice. (B) Enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis in DCs cluster. (C) Gene set enrichment analysis on oxidative phosphorylation in DCs from CDD-2103-treated mice. (D) Heatmaps showing the expression changes of major transcripts related to the complex I-V in electron transport chain in DCs cluster between vehicle and CDD-2103 treated mice. (E) BMDCs were treated with or without CDD-2103 for 48 h, and oxygen consumption rate (OCR) of BMDCs was measured by a Seahorse XF analyzer over time. (F) Intracellular ATP production in CDD-2103-treated and control BMDCs were detected by Luminescent ATP Detection Assay. (G, H) Bar charts presented the percentage change in mean fluorescence intensity (MFI) of (G) intracellular ROS levels determined by DCFH-DA fluorescence probe and (H) mitoROS levels with mitoSOX probe in BMDCs treated with serial doses of CDD-2103 compared to unstimulated control. (I) BMDCs were treated with or without serial dose of CDD-2103 for 48 h in the presence or absence of rotenone (10 nM), and the expression of CD86 was analyzed by flow cytometry. (J) BMDCs pre-treated with CDD-2103 in the presence of absence of roterone were co-cultured with CD4⁺ T cells at a ratio of 1:3 or 1:10 (BMDCs: CD4⁺ T cells) for 3 days. The percentage of CD4⁺Foxp3⁺ Tregs were detected by using flow cytometry. Data were expressed as mean ± SEM. ^∗p < 0.05, ^∗∗p < 0.01 versus indicated group and n.s., no significant differences.

Palmatine and berberine from CDD-2103 that are distributed in mesenteric lymph nodes promote DC-mediated Treg differentiation

CDD-2103 is a standardized formulation composed of nine herbs with clinically determined efficacy in UC patients in remission. LC-MS analysis of its major ingredients revealed that CDD-2103 is primarily composed of berberine and magnoflorin from C. chinensis Franch., curcumin from C. longa L., and morroniside and loganin from C. officinalis Sieb. et Zucc (Fig. 7a). Therefore, these compounds serve as quality control markers for CDD-2103. Because of the significant changes in the MLN-derived CD11C⁺ DCs population in chronic Rag1^-/- colitis mice, we sought to determine the distribution of the components of CDD-2103 in the MLN of these mice (Fig. 7b). CDD-2103 was administered to mice and MLN were isolated 1 h post-administration. Nine compounds, confirmed by the corresponding reference standards, were present in the MLN of mice with colitis (Table 3), suggesting that these compounds could directly target MLN-derived immune cells in vivo. We further examined whether these compounds were responsible for DC-mediated Treg differentiation induced by CDD-2103. Surprisingly, berberine and palmatine suppressed CD86 expression in BMDCs at non-toxic doses (Fig. 7c and d). Co-culture of berberine- or palmatine-pre-treated BMDCs with unfractionated CD4⁺ T cells enhanced the Treg population, suggesting that these compounds may be the responsible components in CDD-2103 mediating DC-specific Treg differentiation (Fig. 7e). Further studies are required to investigate how palmatine and berberine interact with DCs in MLN.

Fig. 7.

The major compounds derived from CDD-2103 are highly distributed in mesenteric lymph nodes of mice with colitis. (A) LC-MS chromatogram (BPC) showing the major compounds contained in CDD-2103. (B) A schematic diagram illustrating the distribution investigation of CDD-2103 components in chronic Rag1^-/- colitis mice using LC/MS analysis. (C) LC chromatogram of berberine (left) and palmatine (right) in MLN of colitis mice compared with its standard compound. (D) The expression of CD86 of BMDC with or without berberine and palmatine was analyzed by flow cytometry. (E) BMDCs pretreated with berberine or palmatine or vehicle were co-cultured with unfractionated CD4⁺ T cells at a ratio of 1:3 (BMDCs: CD4⁺ T cells) for 3 days. Cells were collected and the percentage of CD4⁺Foxp3⁺ Tregs were detected by using flow cytometry. (F) The schematic illustration of the therapeutic effect of CDD-2103 on colitis through CDD-2103-induced oxidative phosphorylation in DCs and tolerogenic DCs-mediated Treg cell differentiation.

Discussion

The immunosuppressive characteristics of Tregs have long been recognized in the control of inflammation and auto-immune disease. An early study suggested an increased rate of apoptosis in LP-residing Treg cells in UC patients compared to non-inflamed control tissues (62 % ± 5.1 % vs. 3.5 ± 2.2 %, p < 0.01) [30], indicating an enhanced susceptibility to Treg loss during active disease. Furthermore, patients who responded to anti-TNFα therapy exhibited a decrease in Treg apoptosis (25.5 ± 0.4 % vs. 9.8 ± 1.9 %) after treatment, whereas patients who did not respond to the therapy showed unchanged Treg apoptosis (2.9 ± 0.4 % vs 2.8 ± 0.2), indicating a correlation between treatment response and Treg cell survival. Interestingly, the current anti-TNFα monoclonal antibodies, infliximab and adalimumab used for severe UC patients, have been reported to induce apoptosis in tumor necrosis factor receptor 2 (TNFR2)-expressing CD4⁺ T cells. In contrast, paradoxical reports suggest that functional impairment, instead of numerical changes in Treg cells, plays a more crucial role in UC pathogenesis. Several studies and recent single-cell transcriptomic analyses [31], [32] have suggested a significant increase in Treg cell clusters in patients with UC. The increased Treg population may represent an immune response aimed at counteracting ongoing chronic inflammation. Another study on IBD patients’ response to vedolizumab revealed an increase in α4β7-expressing β1⁺PI16⁺ Treg population at a certain dose of vedolizumab treatment, suggesting that these cells are primarily resistant to vedolizumab [33]. Enrichment of IL-23R in Tregs impairs their stability and suppressive functions [34]. These findings highlight the importance of impaired Treg suppressive function in UC exacerbation.

The role of tolerogenic DCs, which exhibit an anti-inflammatory phenotype and maintain immune tolerance by promoting the polarization of T cells into immunosuppressive Tregs, is highly appreciated in autoimmune disease [35]. A phase I clinical trial involving human autologous tolerogenic DCs has been conducted (NCT02903537) in patients with multiple sclerosis (MS). Tolerogenic DCs represent another subset of DCs characterized by low or absent maturity, as indicated by reduced amounts of costimulatory molecules and decreased secretion of pro-inflammatory cytokines. This leads to the absence of the co-stimulatory signals necessary for effector T cell differentiation. DCs can be programmed to develope either a tolerogenic or immunogenic phenotype, depending on environmental cues. Several studies have suggested that tolerogenic DCs mediate T cell phenotype switching through various mechanisms: 1) in a high-lactate environment established by autologous tolerogenic DCs, LDH kinetics are exhausted, leading to reduced glycolysis and differentiation of effector T cells [36]; and 2) limited antigen presentation triggers tolerogenic Treg differentiation while inducing anergy in autoreactive T cells [37]. 3) Tolerogenic DCs expressing indoleamine 2,3-dioxygenase support the development of antigen-specific Treg [38]. In summary, tolerogenic DCs promote a tolerant microenvironment by inducing anergy, reducing antigen-reactive T populations, and fostering Treg expansion or differentiation.

DCs reprogram across a spectrum of phenotypes to meet the specific metabolic demands for their downstream functions, including antigen presentation and T-cell polarization. Single-cell mapping of the metabolic status of in vitro culture DCs revealed that DCs with tolerogenic and immunosuppressive properties favored metabolic hyperactivity, characterized by increased OXPHOS, glycolysis, and fatty acid oxidation. Additionally, GM-CSF/IL-4-induced immature DCs exhibit increased mitochondrial dependency [39]. This study demonstrated that metabolic transitions and energy sources significantly influence the plasticity and functional properties of DCs. A recent study reported the role of vitamin D in patients with inflammatory bowel disease [40]. Treatment with active vitamin D3, a major immunomodulator of DCs, reduced the co-stimulatory factors CD80 and CD86, whereas upregulating genes related to OXPHOS and the tricarboxylic acid cycle (TCA). Vitamin D3-induced DCs exhibit a reduced ability to induce autoreactive T cells and promote Treg generation [41]. Tolerogenic DCs express high levels of genes in the OXPHOS pathway, particularly those responsible for complexes I-V [42]. Additionally, tolerogenic DCs exhibit distinct redox activities, with increased ROS levels and mitochondrial activity. Consistent with our current findings, CDD-2103 significantly increased ROS production and mitochondrial respiration in BMDCs in a dose-dependent manner. However, supplementation with rotenone, which blocks complex I of the electron transport chain, reversed all phenotypes observed in CDD-2103-intervened BMDCs.

Transfer colitis is a mouse model that mimics intestinal inflammation observed in IBD [43]. In the present study, we observed that the administration of CDD-2103 reversed disease severity in chemically induced colitis and T cell-transferred Rag 1^-/- mouse models. The co-transfer of naïve CD4⁺ CD45RB^hi and regulatory CD4⁺ Foxp3^GFP T cells as well as depletion Foxp3^DTR T cells in T-transferred Rag 1^-/- mice confirmed the suppressive effect of CDD-2103 on colitis activity, which was associated with an increase in the Treg cell population in the colon. A previous study reported a significant decrease in the number of immune cells, specifically Tregs, in the colonic tissue of patients with UC [30]. It can be hypothesized that an increase in Tregs induced by CDD-2103 leads to restoration of the immune system. In the DCs and T cells in vitro co-culture system, CDD-2103-treated DCs increased the Treg cell population through oxidative phosphorylation. Further studies are necessary to investigate not only the mechanism of action of CDD-2103 on DCs but also its downstream signaling pathway and potential modification of Tregs. Numerous studies have revealed that Traditional Chinese Medicine has the potential to treat UC [9]. In this study, we identified a Chinese herbal formula, CDD-2103, that enhances Tregs, suggesting the utility of natural products for the treatment of UC via Tregs. Compounds derived from Chinese herbal medicines or natural plants have been shown to modulate Treg cell activity and contribute to disease progression in animal models of IBD [44], further reinforcing the potential of the vast resource of natural products for the treatment of UC. The primary components of CDD-2103, palmatine and berberine, were found to be distributed in the MLN of mice with colitis, fostering tolerant DC-induced Treg differentiation. Palmatine and berberine are isoquinoline alkaloids that belong to the protoberberine class. Previous studies have reported that the inhibitory effects of palmatine and berberine on colitis are primarily attributed to their regulation of inflammatory signaling and the gut microbiome [45], [12], [46]. Berberine downregulates the expression of co-stimulatory molecules in DCs and suppresses Th17 T cells [47], [48]. However, the immunoregulatory effects of palmatine have rarely been reported. Further investigations into how palmatine regulates the roles and functions of DCs are warranted.

Previous clinical studies have highlighted the efficacy of Chinese herbal preparations in alleviating UC symptoms. For instance, Shen et al. reported significant improvements in patients with moderately active UC treated with the Qing-Chang-Hua-Shi Formula, which consists of 11 herbs, showing greater clinical response (31.48 % vs. 12.50 %) and remission rates (92.59 % vs. 72.92 %) than those receiving a placebo [49]. Our meta-analysis, which included 2311 articles on clinical studies of Chinese herbal medicine, revealed that the most commonly used herbs for UC are those considered to clear internal heat and dampness, invigorate the spleen, and hinder blood stasis [9]. For example, in CDD-2103, herbs, such as G. uralensis Fisch. and A. macrocephalae Koidz., which are known to nourish the spleen, have demonstrated anti-inflammatory activities. Water extracts of G. uralensis Fisch. significantly reduced inflammation in mice with colitis by inhibiting TLR4/MyD88/NF-kB and enhancing mucosal barrier integrity in vivo [50]. Its major components, glycyrrhizin, liquiritin, and glycyrrhetinic acid, have been shown to suppress various cytokines and the expressions of NF-kB, STAT-3, and STAT-6 [51]. Furthermore, atractylone from A. macrocephalae Koidz. has been reported to attenuate colitis progression by reducing the TNF-α and ROS level [52]. Other notable herbs include C. pinosula (Franch.) Nannf. and B. striata (Thunb.) Rchb.f. is known for its ability to address blood stasis and is rich in polysaccharides that induce immunomodulatory effects. Polysaccharides from C. Pinosula (branches) Nannf. enhances secretory immunoglobulin A and Lactobacilus populations in immunosuppressed mice [53], whereas those from B. striata (Thunb.) Rchb.f. exhibits immunostimulatory activity, including increased macrophage phagocytosis and serum immunoglobulin secretion [54]. Herbs known for their heat-clearing properties, such as C. longa L. and C. chinensis Franch., have also been studied. Curcuminoids in C. longa L. exhibit potent anti-oxidant effects, inhibit neutrophil migration, and protect colonic mucosal cells against apoptosis [55]. C. chinensis Franch., which is rich in alkaloids, has been reported to positively alters intestinal microbiota composition and metabolism, aiding in the restoration of gut microbiota homeostasis in colitis [56]. In summary, the complex chemistry of CDD-2103 encompasses a range of pharmacological activities, including anti-inflammatory, immunomodulatory, and anti-foxidant effects, in addition to the regulation of gut dysbiosis. These findings suggest that the major chemical constituents of these herbs play an indispensable role in the efficacy of CDD-2103 in delaying the progression of colitis.

In most patients, the course of UC is characterized by a cyclical pattern of relapse and remission, highlighting the limited efficacy of existing treatment [2]. Our ongoing randomized placebo-controlled clinical trial aimed to investigate the efficacy of CDD-2103 in maintaining remission in patients with UC (ChiCTR2100043200). To determine the efficacy of remission maintenance in a mouse model, we selected a mouse model of chronic colitis. This model involves repeated challenges with a low dose of TNBS after a certain recovery period, which accurately mimics the relapse and remission of UC [23]. Similar to the assessment of UC exacerbation in patients, our study evaluated the severity of diarrhea and fecal occult blood using the DAI, as well as through histopathological assessment. Although CDD-2103 intervention reduced the stool consistency score and preserved the colon architecture following multiple TNBS challenges, suggesting that CDD-2013 intervention may sustain disease remission, the preclinical colitis mouse model may not fully replicate the clinical context of patients in UC remission, as well as inform the efficacy of CDD-2103 in remission maintenance. Therefore, unveiling the label of the randomized controlled trial is crucial to confirm the effectiveness of CDD-2103 in sustaining remission. In addition, to better characterize the effect of CDD-2103 on immune tolerance, further studies should include a broader concentration range to determine the dose–response relationship of CDD-2103 on immune regulation.

In conclusion, our study indicates that CDD-2103, a Chinese herbal formulation, can effectively suppress the progression of colitis, primarily by enhancing the tolerogenic immune microenvironment. The cellular and molecular evidence of CDD-2103 will facilitate the future development of CDD-2103 as a new drug candidate for the management of UC.

Author statement

Zhaoxiang Bian and Hor-Yue Tan conceived and designed the study; Chunhua Huang and Cheng Lyu did the experiments; Heung-Lam Mok, Yiqi Xu and Die Hu facilitated animal experiments; Cheng Zhang performed computational analysis; Ka-Wing Cheng and Lin Zhu performed quality control of CDD-2103; Lin Zhu, Chengyuan Lin and Xin Chen provided essential insight on the study; Hor-Yue Tan and Xin Chen interpreted the results and wrote the manuscript. All authors approved the final manuscript.

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: