The regulatory effect of chitooligosaccharides on islet inflammation in T2D individuals after islet cell transplantation: the mechanism behind Candida albicans abundance and macrophage polarization

ABSTRACT

Islet cell transplantation (ICT) represents a promising therapeutic approach for addressing diabetes mellitus. However, the islet inflammation during transplantation significantly reduces the surgical outcome rate, which is related to the polarization of macrophages. Chitooligosaccharides (COS) was previously reported which could modulate the immune system, alleviate inflammation, regulate gut microecology, and repair the intestinal barrier. Therefore, we hypothesized COS could relieve pancreatic inflammation by regulating macrophage polarization and gut microbiota. First, 18S rDNA gene sequencing was performed on fecal samples from the ICT population, showing abnormally increased amount of Candida albicans, possibly causing pancreatic inflammation. Functional oligosaccharides responsible for regulating macrophage polarization and inhibiting the growth of Candida albicans were screened. Afterwards, human flora-associated T2D (HMA-T2D) mouse models of gut microbiota were established, and the ability of the selected oligosaccharides were validated in vivo to alleviate inflammation and regulate gut microbiota. The results indicated that ICT significantly decreased the alpha diversity of gut fungal, altered fungal community structures, and increased Candida albicans abundance. Moreover, Candida albicans promoted M1 macrophage polarization, leading to islet inflammation. COS inhibited Candida albicans growth, suppressed the MyD88-NF-κB pathway, activated STAT6, inhibited M1, and promoted M2 macrophage polarization. Furthermore, COS-treated HMA-T2D mice displayed lower M1 macrophage differentiation and higher M2 macrophage numbers. Additionally, COS also enhanced ZO-1 and Occludin mRNA expression, reduced Candida albicans abundance, and balanced gut microecology. This study illustrated that COS modulated macrophage polarization via the MyD88/NF-κB and STAT6 pathways, repaired the intestinal barrier, and reduced Candida albicans abundance to alleviate islet inflammation.

GRAPHICAL ABSTRACT

Introduction

The global prevalence of diabetes among adults soared to 529 million in 2021, with projections indicating a staggering increase to 1.31 billion by 2050.¹ Notably, type 2 diabetes (T2D) constitutes a significant proportion of these cases, accounting for estimates ranging between 90% and 95%.² The pathogenesis of T2D primarily implicates insulin resistance (IR) and pancreatic β-cell apoptosis.^2–4 Presently, the prevailing clinical interventions for T2D encompass oral hypoglycemic agents such as thiazolidinediones, biguanides, and sulfonylureas, alongside subcutaneous insulin injections and islet cell transplantation (ICT). In clinical practice, ICT is not only a cure for type 1 diabetes, but also can be used for severe secondary diabetes with serious complications. Although insulin significantly reduces or delays diabetes, serious or fragile T2D patients still present fluctuating blood glucose levels and even sudden death. ICT effectively overcomes the shortcomings of insulin therapy, presenting an ideal treatment for insulin-dependent diabetes. Due to the side effects and low patient compliance of drug therapy, ICT has gradually become a better option for radical T2D treatment. ICT represents a promising therapeutic approach for addressing diabetes mellitus and mitigating associated complications, contributing to stabilize blood glucose levels. However, islet inflammation during transplantation is believed to primarily contribute to early islet damage and loss after transplantation.⁵

Gut fungi regulate host immunity. Leonardi et al. found that when opportunistic human symbiotic Candida albicans colonizes in the gut of mice, mononuclear phagocytes (MNPs) expressing CX3CR1 in myeloid cells of the gut upregulates CD40 and CD86 co-stimulatory molecule expression, and further regulates macrophage polarization.⁶ Macrophages form a crucial part of the innate immune system and are involved in various immune functions in organisms. Subtypes include the classically activated M1 type and the alternatively activated M2 type.^7,8 In pancreatic islets, inflammation demonstrates a close correlation with macrophage polarization,^9,10 which is influenced by gut fungi and bacteria. The gut microbiota foster macrophage accumulation and polarization toward a pro-inflammatory M1 phenotype.¹¹ Extracellular vesicles (EVs) released by Lactobacillus murinus within the intestinal milieu induce macrophage conversion from the M1 to M2 phenotype. This conversion is facilitated by the activation of Toll-like receptor 2 (TLR2), which prompts IL-10 release, consequently contributing to the restoration of the intestinal barrier.¹² Thus, we assumed that alterations in the composition of gut microbiota and macrophage polarization-induced inflammation affect ICT efficacy and the duration of insulin independence.

Chitooligosaccharides (COS), carbohydrates formed by glucosamine linked by β-1,4 glycosidic bonds, serve as representative functional oligosaccharides and constitute a crucial role in regulating gut microecology.¹³ Additionally, they exhibit a range of activities including anti-inflammatory and immune regulatory functions.^14–16 Treatment with COS significantly diminishes blood glucose levels in db/db mice, reverses IR, and mitigates gut microecological imbalance in diabetic mice by fostering Akkermansia growth and impeding Helicobacter proliferation. Hence, the intervention with functional oligosaccharides, such as COS, may mitigate islet inflammation through the regulation of gut microecology and alleviating macrophage inflammation, potentially enhancing the success rate of ICT.

This study aims to elucidate the relationship between gut fungal abnormalities, macrophage polarization, and islet inflammation in T2D population after ICT. And it is proposed that functional oligosaccharides, mainly COS, could reduce islet inflammation. This study explores the association between islet inflammation, abnormal polarization of macrophages, and the disorder of gut fungi (targeting Candida albicans) after ICT. Moreover, the effect of COS on the pancreatic inflammation induced by Candida albicans-induced macrophage polarization was investigated. Additionally, we confirm the role of COS in regulating gut microecology and alleviating islet inflammation in vivo, providing a theoretical basis for the intervention of gut microbiome-targeted foods to alleviate inflammation after ICT.

Materials and methods

Materials

The purity of COS was higher than 90%, including COS2 (MW = 341.16 Da), COS3 (MW = 502.22 Da), COS4 (MW = 663.29 Da), COS5 (MW = 824.36 Da), and COS6 (MW = 985.43 Da), which were prepared in our laboratory.¹⁷ COS mixture displayed a polymerization degree of 2–9, a deacetylation degree of >95%, a purity higher than 95%, and an average molecular weight of 2087 Da. Fructooligosaccharides (FOS) with a polymerization degree of 2–14, a purity of 90%, and an average molecular weight of 2556 Da was procured from Shanghai Maclin Biochemical Technology Co. Ltd. (China, Shanghai). Turanose (TOS) with a purity of 98% was acquired from Shanghai Yuanye Biotechnology Co. Ltd. (China, Shanghai). Cellobiose (CEOS) with a purity higher than or equal to 99% was purchased from Shanghai Maclin Biochemical Technology Co., Ltd. (China, Shanghai). The detection results and related information of the COS and other functional oligosaccharides can be found in the supplementary information (Appendix Figures S1-S8, Appendix Table S1). The streptozocin (STZ) was obtained from Sigma-Aldrich (USA, St. Louis).

Subject recruitment and clinical sample collection

All patients were recruited from the Hepatobiliary and Pancreatic Department of Shanghai Chang-Zheng Hospital (China, Shanghai) and provided signed informed consent forms. This study adhered to the principles outlined in the Declaration of Helsinki and met the criteria established by the Ethics Committee (Ethics Approval number: ECUST-2022-104; Shanghai, China). The volunteers who contributed fecal samples comprised 16 patients with T2D (CON group) and 33 patients with T2D who underwent ICT (ICT group). The selection, inclusion, and exclusion criteria for patients are provided in the supplementary information. During the fecal sample collection process, volunteers completed health status forms. Feces were collected using sterile collection tubes.

The inclusion criteria for patients with T2D were as follows: individuals aged between 18 and 65 with insulin-dependent diabetes. The inclusion criteria for patients with T2D undergoing ICT were as follows: individuals aged between 18 and 65 within three months post-ICT, receiving immunosuppressive medication post-surgery, and exhibiting dependence on insulin therapy with poor blood glucose control (HbA1c > 8.0% or time in range (TIR)<70%). The exclusion criteria for the project participants included the use of antibiotics, probiotics, prebiotics, or other microbiological agents, as well as the utilization of microbiota-related medications or the post-surgical adherence to a specific dietary regimen (such as yogurt consumption).

Analysis of the gut fungal diversity

DNA extraction from fecal samples was performed using the TGuide S96 magnetic column soil/fecal genomic DNA extraction kit (Tiangen, DP712, China). Subsequently, DNA was amplified using 18S full-length primers, including the (5’−3’) AACCTGGTTGATCCTGCCAGT forward primer and (5’−3’) GATCCTTCTGCAGGTTCACCTAC reverse primer at 95°C for 4 min. Then, 98°C 15 s, 55°C 30 s, 72°C 2 min cycle 35 times, and finally 72°C 7 min ¹⁸. The concentration and integrity of the amplified products were assessed, and libraries were constructed for the qualified samples. Following the validation of the library test, the on-machine libraries were merged with the PacBio Binding kit prior to processing, while the subsequent reaction products were purified using AMpure PB Beads. Sequencing was conducted on a PacBio Sequel II sequencer. CCS sequences were obtained through the utilization of smart-link analysis software. Barcode identification, length filtering, and chimerism removal were employed to acquire effective CCS sequences, which were subsequently aggregated and denoised. The statistical results of sample sequencing data processing were shown in Table S2. The library underwent sequencing utilizing them. The Illumina sequencing platform (Biomarker Biotechnology Co. Ltd., China, Beijing) was used for library sequencing. BMK Cloud was utilized for composition analysis, alpha and beta diversity assessments, and microbial-related analyses of the fungal microorganisms in the CON and ICT group samples.¹⁸

Cell culture and treatments

RAW264.7 cells and 293T cells were procured from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China) and cultured in a complete medium supplemented with 1% penicillin and streptomycin, respectively, in a 5% CO₂ cell incubator at 37°C.^19,20 The RAW264.7 cells were seeded into 6-well plates, and following a 48-h intervention with various functional oligosaccharides, such as COS, TOS, CEOS, and FOS, a suspension of 1 × 10⁶ CFU/mL Candida albicans was added, followed by cell-line infection for 12 h. Additionally, the supernatant was collected for ELISA to detect the expression of the inflammatory factors and other related proteins.

Candida albicans culture and treatments

The Candida albicans (deposit number: ATCC90028) was cultured in a Sabouraud dextrose broth medium at 30°C for 24–48 h. Subsequently, equal volumes of the fungal suspension were added to culture media containing 10 mg/mL of COS, TOS, CEOS, and FOS, respectively, thoroughly mixed, and dispensed into 96-well plates, using the same concentration of glucose as a control. The fungal density was measured every 2 h at 600 nm and 30°C.²¹ Following inactivation at 121°C for 15 min, the fungal cells were harvested and utilized to construct the M1 inflammatory cell model.

Animal experiment

The ARRIVE guidelines 2.0 were followed for the experimental protocols, and all animal treatments and experiments were strictly conducted in compliance with the National Research Council’s Guide for the Care and Use of Laboratory Animals. The procedures were approved by the East China University of Science and Technology Bioethics Committee (Ethics Approval number: ECUST-2022-104; Shanghai, China). Male mice were selected for the STZ-induced diabetes model due to their higher sensitivity to islet cytotoxin compared to female mice.²²

SPF grade C57BL/6J male mice (Qualification NO.: 20180004074931) weighing 16–20 g and aged 4–5 weeks were procured from Shanghai Jesjie Laboratory Animal Co. Ltd. The mice were accommodated in a new environment at 21 ± 2°C with a 12-h light–dark cycle for 1 week.

First, we established human flora associated type 2 diabetes (HMA-T2D) mice models (Figure 1). Subsequently, 48 mice were randomly allocated to two groups: the CON group (n = 8) and the HMA-T2D group (n = 40) (Figure 1). The mice in the CON group were fed a standard diet and provided with sterile water. Mice in the HMA-T2D group were fed a high-fat diet for 8 weeks and received intraperitoneal injections of STZ at a dose of 110 mg/kg, dissolved in 0.1 M citrate buffer at pH 4.5.²³ Random blood glucose levels of the mice were measured on the 3rd and 7th days post-STZ injection. The diabetes model was deemed successful if the glucose level exceeded 16.7 mmol/L.²⁴ The mice were then subjected to HMA. Pseudo-sterile mice were prepared by adding broad-spectrum antibiotics in their drinking water for 3 weeks, which was changed twice weekly.¹⁵ Following this period, fresh mouse feces were collected in sterile centrifuge tubes to prepare a diluent, which was inoculated onto Colombia agar medium using the plate coating method and cultured for 18–24 h. Subsequently, bacterial growth was observed to determine the bacterial concentration in the feces. A 90% reduction in the viable bacteria in the mouse feces indicated the successful establishment of the antibiotic-treated mouse model. Next, the HMA mouse model was established. The fecal bacterial suspension prepared from the T2D patients was administered to the mice via gavage at a volume of 100 μL per administration (OD = 1). This gavage procedure was repeated once every 2 weeks, with continuous gavage for 3 days each time, totaling 3 gavages.²⁵

Figure 1.

Development of the HMA-T2D mouse model and the experimental intervention group design.

The HMA-T2D mice were randomly divided into five groups, consisting of eight mice each: (a) Model group: Received 0.1 mL of solvent every other day. (b) RA group: Gavaged with rapamycin (RAPA) (5 mg/kg every other day). (c) RC group: Gavaged with RAPA (5 mg/kg every other day) and COS (350 mg/kg/day). (d) CA group: Gavaged with Candida albicans (2 × 10⁸ CFU/0.2 mL every other day). (e) CC group: Gavaged with Candida albicans (2 × 10⁸ CFU/0.2 mL every other day) and COS (350 mg/kg/day).^24,26

After the intervention experiment, blood was collected from the intraorbital venous plexus following 12 h fasting period. Blood was collected using 1.5 mL centrifuge tubes without anticoagulants, which was allowed to stand at room temperature for 2 h, and centrifuged at 7,000 g for 15 min to separate the serum. Following blood collection, the mice were euthanized via spinal dislocation to minimize pain and promptly dissected, and the livers, kidneys, and pancreas were collected and weighed. A portion of the pancreas was fixed with a 4% paraformaldehyde fixation solution, while other tissues were rapidly frozen in liquid nitrogen and stored at −80°C in a refrigerator. The sample collection, processing, and determination were conducted randomly to avoid the influence of confounding factors on the experiment.

Immunofluorescence

Immunofluorescence (IF) staining was conducted on pancreatic sections from the mice in the various treatment groups. The paraffin sections were dewaxed and subjected to antigen retrieval. BSA was added and incubated for 30 min to block nonspecific-binding sites. Primary antibodies against iNOs (GB11119, Servicebio, Wuhan, China) and CD206 (GB113497, Servicebio, Wuhan, China) were added and incubated overnight at 4°C. Following decolorization and washing, the corresponding secondary antibodies were applied and incubated at room temperature for 50 min. The sections were decolorized again, after which a DAPI dye solution (Servicebio, Wuhan, China) was added and incubated at room temperature for 10 min. Images of the sections were captured using confocal fluorescence microscopy (Ti-U, NIKON, Japan).²⁷

Flow cytometry

The RAW264.7 cells were washed with PBS, and the cell concentration was adjusted to 1 × 10⁶/mL. The live and dead cells were labeled with Fixable Viability Dye eFluor 780 (65-0865-14, Thermo, USA) and stained with CD80-FITC (11-0801-81, Thermo, USA) for 30 min. CD206-APC²⁸ (17-2061-80, Thermo, USA) was utilized for an additional 30 min staining, after which the cells were washed twice with PBS, and the degree of cell polarization was analyzed using a Cytoflex S Flow Cytometry system (Beckman, USA).

RNA extraction and quantitative real-time PCR

Total RNA was extracted using a TransZol Up Plus RNA Kit (Transgen, ER501, China). Subsequently, the concentration and quality of RNA were assessed utilizing), while a multifunctional microplate reader was employed to assess the RNA concentration and quality. Then, the RNA was reverse-transcribed into cDNA using TransScript All-in-One First-Strand cDNA Synthesis Super Mix for qPCR (Transgen, AT341, China). The mRNA expression levels of the genes associated with macrophage polarization were determined using qPCR. The primer sequences can be found in Appendix Table S2. Primers referenced Primer Bank (https://pga.mgh.harvard.edu/primerbank.).

Enzyme-linked immunosorbent assay (ELISA) assay

The cell supernatant and tissue homogenate were collected. The expression levels of TNF-α, IL-10, IL-1β, TGF-β, and other proteins were assayed using an ELISA kit (MlBio, Shanghai, China) at an optical density of 450 nm according to the protocols of the manufacturer.

Molecular docking simulation

The molecular structure of COS5 was downloaded from PubChem, while the 3D structures of the MyD88 (PDB code: 7I6W) and STAT6 (PDB code: 4Y5U) were obtained from the RCSB Protein Data Bank (www.rcsb.org.). The receptor protein was dehydrated, and the ligands were removed using PyMol, followed by docking analysis conducted with AutoDock Vina 1.1.2. PLIP was employed to analyze the docking results, which were visualized using PyMol^30.

Statistical analysis

All data were presented as mean ± standard deviation. A t-test was employed to ascertain statistically significant differences between the two groups. For comparisons involving more than two groups, One-way ANOVA and the Tukey multiple-comparison test were utilized, with p-values below 0.05 considered statistically significant.

Results

ICT increased the Candida albicans in the fecal samples

To investigate the impact of gut fungi before and after ICT, 16 fecal samples from the CON group and 33 fecal samples from the ICT group were analyzed using 18S rDNA gene sequencing (Figure 2). As illustrated in Figure 2(a), the community richness and diversity of gut fungal communities before and after ICT were assessed via alpha diversity analysis. Compared to the CON group, the Chao1 and ACE indexes in the ICT group decreased significantly (p < 0.05). Conversely, the Shannon index displayed an upward trend, while the Simpson index significant substantially (p < 0.05). These findings suggested that ICT reduced both the richness and diversity of the fungal community.

Figure 2.

(Continued).

The microbial community structure served as a crucial indicator in microbial flora analysis. Compared with the CON group, the microbial community structure of the ICT group showed significant changes (Figure 2(b)). PLS-DA was used for analysis, showing similarities within each group and significant differences between the groups (p < 0.05).

After ICT, the gut fungi were analyzed at the phylum, genus, and species levels to elucidate the composition of the gut fungi. At the phylum level, compared with the CON group, Basidiomycota (CON:ICT, 37%:14%) exhibited a significant decrease in the ICT group (p < 0.05), while Ascomycota (4%:30%) and Ciliophora (2%:0.1%) increased substantially (p < 0.05) (Figures 2(c,d)). At the genus level, Russula (24%:6%) and Quercus (16%:4%) decreased significantly in the ICT group compared to the CON group (p < 0.05), while Candida (0%:17%) exhibited a significant increase (p < 0.05) (Figures 2(e,f)). Compared to the CON group, Quercus lobata and Russula exalbicans in the ICT group experienced significant decreases (p < 0.05), while Candida albicans showed a significant increase (p < 0.05) (Figures 2(g,h)).

LEfSe was employed to analyze the distinct microorganisms present in the two groups, using an LDA score >2 as the standard (Figure 2(i)). At the family level, Leucosporidiaceae and Polygonaceae clustering was evident in the CON group, while Debaryomycetaceae and Cystofilobasidiaceae clustered in the ICT group. At the genus level, Russula and Leucosporidium increased significantly in the CON group, while Amphisiella, Cystofilobasidium, and Candida displayed significant increases in the ICT group (p < 0.05).

The species-level analysis (Figure 2(h)) showed a significant increase in the abundance of Hannaella luteola, Leucosporidium yakuticum, and Rheum spiciforme in the CON group, while the Candida albicans, Amphisiella magnigranulosa, Candida dubliniensis, Malassezia globosa, and Candida glabrata in the ICT group. Notably, Candida albicans demonstrated the most pronounced variance, it was selected for subsequent experimental investigation. The mRNA expression of Candida albicans in the human fecal samples was validated using qPCR to determine its relative abundance after ICT (Figure 2(i)). The mRNA levels of the Candida albicans were significantly higher in the ICT group than in the CON group (p < 0.05), while its intestinal abundance increased significantly (Figure 2(j)). The above results indicated that ICT decreased the fungal abundance and diversity and changed the fungal community structure. Candida albicans had the most significant accumulation in the ICT group after transplantation.

Figure 2.

The microbial composition and differences between the ICT and CON (CON group: n = 16; ICT group: n = 33). (a) The alpha diversity analysis evaluated the diversity and richness of microbial communities in CON and ICT samples. (b) The beta diversity analysis evaluated the difference of microbial structure in CON and ICT samples. (p<0.05). (c) Comparison of microbial distribution histogram and microbial distribution histogram among samples. (e) and (g), comparison of microbial distribution histogram at genus, species level. (d), (f) and (h) the significant analysis of gut fungi. (i) The LEfSe was employed to analyze the diverse microorganisms present in the CON and ICT groups. The LDA value greater than 2 and a p-value less than 0.05 served as the threshold criteria for the screening process. (j) The relative expression levels of Candida albicans mRNA in CON and ICT groups.

Inactivated Candida albicans leaded to macrophage polarization and inflammation

The impact of Candida albicans on macrophage polarization was assessed via flow cytometry, which involved detecting the CD80 and CD206 marker proteins on various macrophage subtypes (Figure 3). The number of M1 macrophages (CD80⁺CD206⁻) induced by Candida albicans was significantly higher in the MOD group than in the CON group (p < 0.001). The expressions of IL-1β and TNF-α were significantly increased while that of IL-10 and TGF-β were significantly decreased. The results revealed that Candida albicans polarized the RAW264.7 cells into M1 macrophages and leading to an inflammatory response.

Figure 3.

The typing and expression levels of the inflammatory-related factors in the RAW264.7 cells induced by Candida albicans (n = 5). (a) The flow cytometry analysis of the RAW264.7 cell M1 (CD80+CD206-)/M2 (CD80-CD206+) macrophages induced by Candida albicans. (b) The mRNA and protein expression levels of IL-1β, TNF-α, IL-10, and TGF-β. *p < 0.05, **p < 0.01, and ***p < 0.001.

COS inhibited Candida albicans proliferation

The Candida albicans growth curves were examined to determine the ability of functional oligosaccharides to directly inhibit its proliferation. As shown in Figure 4(a), the Candida albicans in the glucose group (GLU) exhibited a lag phase from 0 to 2 h, followed by an exponential phase from 2 to 20 h, entering a stationary phase after 20 h. COS significantly decreased the Candida albicans concentration, which inhibited its growth and prevented it from reaching the stationary phase. Furthermore, to assess the ability of COS to prolong the lag phase or reduce the concentration reaching the stationary phase, both the lag- and exponential-phase Candida albicans was reintroduced into the COS-containing medium for re-culturing. COS reduced the fungal concentration during the stationary phase, indicating that it inhibited Candida albicans growth (Figures 4(b-e), p < 0.05).

Figure 4.

The effect of functional oligosaccharides on Candida albicans growth (n = 5). (a) The effect of functional oligosaccharides on the growth of Candida albicans. (b-c) the effect of COS on the growth during the lag and exponential phases inoculated with Candida albicans. (d-e) the fungal concentration of the Candida albicans inoculated during lag and exponential phases when proliferating to the stationary phase.

COS ameliorated Candida albicans-induced polarization in RAW264.7 cells

The expression levels of by CD80 and CD206 were measured to evaluate macrophages polarization (Figure 5(a)). Compared to the MOD group, the M1-type macrophages decreased significantly in the COS, FOS, TOS, and CEOS groups (p < 0.05), while the M2-type macrophages (CD80-CD206+) were significantly higher in the COS and TOS groups (p<0.05). COS, TOS, and CEOS significantly decreased the M1 macrophages and increased the M2 macrophages (p < 0.05), as indicated by the M1-to-M2 macrophage ratio (CD80/CD206).

Figure 5.

The RAW264.7 polarization induced by Candida albicans after functional oligosaccharide treatment (n = 5). (a) Flow cytometry was used to detect the marker proteins CD80 and CD206 of the different types of macrophages, the expression level of CD80, the expression level of CD206, and the ratio of CD80-to-CD206 expression. (b) The mRNA and protein expression levels of IL-1β, TNF-α, IL-10, and TGF-β. (c) Flow cytometry was used to detect the marker proteins CD80 and CD206 of the different types of macrophages, the expression level of CD80, the expression level of CD206, and the ratio of CD80-to-CD206 expression. (d) The mRNA and protein expression levels of IL-1β, TNF-α, IL-10, and TGF-β. Chitooligosaccharides (COS), Turanose (TOS), Cellobiose (CEOS), Fructooligosaccharides (FOS).

Figure 5.

(Continued).

Only COS both inhibited Candida albicans and regulated macrophage polarization. Therefore, subsequent research focused on COS intervention. COS significantly decreased the expression levels of the IL-1β and TNF-α pro-inflammatory factors (p < 0.05, Figure 5(b)) while elevating that of the IL-10 and TGF-β anti-inflammatory factors (p < 0.05) induced by Candida albicans, consequently alleviating the inflammatory response.

Subsequently, the influence of specific degree of polymerization COS on macrophage polarization was investigated, and the COS with the most potent anti-inflammatory effect was identified. As illustrated in Figure 5(c-d), the CD80/CD206 ratio of COS2–5 showed an overall downward trend, so COS5 was used for subsequent mechanism research. Furthermore, COS5 attenuated the levels of inflammatory factors IL-1β and TNF-α induced by Candida albicans (p < 0.05), promoted the expression of anti-inflammatory factors IL-10 and TGF-β (p < 0.05), and mitigated inflammation.

Oral COS reduced the Candida albicans-induced inflammation and macrophage polarization of pancreatic islets in T2D mice

To detect the in vivo effects of COS, an HMA-T2D mouse model was constructed, and the body weight and blood glucose levels were evaluated daily (Figures 6(a,b)). In comparison to the CON group, as depicted in Figure 6(b), the random blood glucose level of mice in other groups was significantly elevated (p < 0.05). There was no discernible difference in the blood glucose level of mice in the RA and CA groups when compared to the MOD group. However, the blood glucose levels of the mice in the RC and CC groups of the COS intervention groups exhibited a decline after week 13, and reaching a significantly lower level than that of the RA and CA groups at week 15. After 5 weeks of intervention, as depicted in Figure 6(c,d), the area under the curve (AUC) in the MOD, RA, and CA groups was significantly elevated compared to the CON group (p < 0.05). Conversely, the AUC in the CC group was significantly diminished compared to the CA group (p < 0.05). These findings suggested that COS might enhance glucose tolerance. Observing the tissue weight of mice in each group, as depicted in Figure 6(e), compared with the CON group, the weight of pancreas from mice in MOD, RA and CA groups decreased significantly (p < 0.05), and the RC and CC groups intervened by COS could reverse this change (p < 0.05), indicating that COS could reverse the reduction of pancreas weight caused by T2D. Thus, COS were able to regulate blood glucose balance and alleviate the loss of the weight of pancreatic.

Figure 6.

(Continued).

Figure 6.

COS improved the islet inflammation and macrophage polarization induced by Candida albicans (n = 8). (a) The daily monitoring of the body weight. (b) The daily monitoring of the blood glucose. (c-d) the glucose tolerance test and AUC after COS intervention. (e) The ratio of the pancreas or kidneys to the body weight. (f) The levels of the inflammation-related factors IL-1β, TNF-α, IL-10, and TGF-β in the serum of the mice. (g) The mouse pancreatic H&E staining. The black arrows signify cell vacuoles, while the green arrows denote nuclear shrinkage. (h) The if staining of the iNOS and CD206 in the mouse pancreas (observed at 200× magnification). (i) The quantitative area percentage of the iNOS in the different treatment groups. (j) The quantitative area percentage of CD206.

To investigate the impact of COS on islet inflammation, the levels of IL-1β, TNF-α, IL-10, and TGF-β were quantified to investigate the impact of COS on islet inflammation (Figure 6(f)). The IL-1β and TNF-α levels were significantly higher in the MOD and CA groups than in the CON group (p < 0.05). The RA group was significantly increased displayed a considerably elevated TNF-α level (p < 0.05), while a negligible IL-1β increase was evident (p > 0.05). After COS intervention, the inflammatory factors in the RC and CC groups were significantly lower than in the RA and CA groups (p < 0.05). These results indicated that COS alleviated the inflammatory response by inhibiting inflammatory factor activity and promoting anti-inflammatory factor production.

To investigate the effect of COS on the pathological status of the mouse pancreas, the changes in its tissue structure were examined by HE staining. In Figure 6(g), it was observed that the pancreas of mice in the CON group had a complete shape with a clear structure, and the islet cells arranged neatly and tightly. In contrast, the MOD, RA, and CA groups showed obvious pathological damage to the pancreas. There were prominent cell vacuoles (black arrow), nuclear pyknosis (green arrow), and interstitial elevation. Compared with the MOD group, the pancreatic injury was more severe in the RA and CA groups, indicating that both RAPA and Candida albicans exacerbated pancreatic tissue injury. After COS intervention, the pathological pancreatic damage was reversed. Islet cells were observed to arrange in an orderly manner, with reduced interstitial and cellular vacuoles, suggesting that COS could alleviate islet damage and improve its physiological function. Thus, our results indicated that RAPA and Candida albicans are main factors which increased the inflammatory response in T2D mice, while COS significantly reduced the inflammatory response and balanced the gut microecology.

To explore the relationship between islet inflammation and macrophage polarization, immunofluorescence staining for iNOs and CD206 was performed to characterize the polarization of macrophages to M1 and M2, respectively. The results are shown in Figure 6(h-j). The expression of CD206 was decreased (p < 0.05), indicating that the M1 polarization of macrophages was increased and the M2 polarization was decreased. After COS intervention, the expression of iNOs in RC and CC groups was decreased (p < 0.05), and the expression of CD206 was increased (p < 0.05), indicating that COS intervention could reduce the number of M1 macrophages and increase the number of M2 macrophages in the pancreas, thereby alleviating inflammatory response.

COS-induced pancreatic islet macrophage differentiation into anti-inflammatory types via the MyD88- NF-κB and STAT pathways

COS5 was docked with MyD88 and STAT6 molecules (Figure 7(a)). The results revealed that COS5 formed hydrogen bonds with THR272 and ARG288 in the TIR domain of MyD88, and established salt bridge (SB) interactions with HIS248, which was in according with the research of Clabbers.²⁹ The interface score of COS5 with MyD88 was −4.3 kcal/mol, indicating that COS5 bound to the TIR domain of MyD88 to inhibit MyD88 expression. Concurrently, COS5 formed a hydrogen bond with amino acids SER564, ASP565, SER566, GLU567, GLN590, and ASP596 in the active pocket of STAT6, and a salt bridge with amino acids LYS544 and ARG562. The docking score of COS5 and STAT6 was −5.1 kcal/mol, which suggested that COS5 tightly bound to the active pocket in STAT6, thereby activating STAT6 activity and promoting M2 polarization of cells.

Figure 7.

The effect of COS on the polarization-related macrophage pathways in the pancreas and the expression of intestinal barrier-related mRNA (n = 8). (a) The results of the molecular docking between COS5 and MyD88. (b-c) the expression of the macrophage polarization-related mRNA and protein in the cells. (d-e) the expression of the macrophage polarization-related mRNA and protein in the mouse pancreas.

To further elucidate the potential mechanism, the related mRNA and protein expression in the MyD88-NF-κB and STAT pathways was assessed in RAW264.7 cells and pancreatic tissues (Figure 7(b,c)). In the MOD group induced by Candida albicans, the mRNA expression levels of MyD88 and NF-κB were significantly increased (p < 0.05), while STAT6 were significantly decreased (p < 0.05). Following COS5 intervention, the expression levels of MyD88 and NF-κB were significantly reduced (p < 0.05), whereas STAT6 expression levels were significantly increased (p < 0.05). These results suggested that COS5 reduced M1-type polarization by inhibiting the MyD88/NF-κB pathway and promoted M2-type polarization by activating the STAT6 pathway in vitro.

The in vivo MyD88 and NF-κB expression levels in the RA and CA groups were significantly higher (p < 0.05) than in the CON group, while that of STAT6 displayed a considerable decrease (p < 0.05) (Figures 7(d,e)). COS significantly reduced the MyD88 and NF-κB mRNA expression levels (p < 0.05) while substantially elevating the STAT6 content (p < 0.05). Notably, the Mod group exhibited expression patterns similar to those of the CON group, further indicating that both RAPA and Candida albicans contributed to M1-type polarization. While, COS intervention reversed these expression patterns and alleviated inflammation.

Figure 7.

(Continued).

COS reduced Candida albicans, improved fungal community diversity, and restored gut microbiota balance

To investigate the effect of COS on the gut flora of humanized diabetic mice, 18S rDNA sequencing was used to identify the gut fungi across the different treatment groups. Alpha diversity analysis showed a significant decrease (p < 0.05) in the Shannon index of the RA group treated with RAPA, with no evident differences in the ACE index, indicating that RAPA reduced the intestinal fungal diversity in the mice (Figure 8(a)). COS intervention significantly increased the fungal diversity (p < 0.05).

Figure 8.

Analysis of the gut fungal community in the HMA-T2D mice. (a) The alpha diversity analysis of the gut fungal community in the HMA-T2D mice. (b-c) analysis of the gut fungal community abundance differences between the RA and RC groups of HMA-T2D mice (LEfSe analysis). (d) The bar charts of the gut fungal distribution comparison among the different treatment groups of HMA-T2D mice. (e) The significance analysis of the fungi in the RA and RC groups of HMA-T2D mice. (f) The effect of COS on the mRNA expression in the colon midgut barrier-related genes in the HT2D mice (n = 5).

Furthermore, LEfSe was used to analyze intestinal fungi in RA group and RC group, and the fungus with significant differences were screened. As shown in Figure 8b and c, the relative abundance of Candida albicans, Unclassified mattesia, Hanseniaspora uvarum and Kazachstania telluris in RA group were significantly increased (p < 0.05), while the relative abundance of mucor racemosus, Debaryomyces hansenii, Yarrowia lipolytica, Trichosporon japonicum, Naganishia albida, Rhodotorula mucilaginosa, and Candida glabrata in RC group significantly increased (p < 0.05). These suggested that RAPA caused an increase in the relative abundance of Candida albicans in the gut.

The gut fungi of the mice in the different treatment groups were analyzed at the species level to determine the relative abundance of Candida albicans. As shown in Figures 8(d,e), COS significantly reduced the relative abundance of Candida albicans compared with the CA group, indicating that the administration of RAPA led to the abnormal increase of Candida albicans and the decrease of microbial diversity, and COS reduced the abnormal increase of Candida albicans and increased microbial diversity.

To further explore how COS protected the pancreas from intestinal flora damage, the expression of intestinal barrier-related genes in the colon was assessed. As shown in Figure 8(f), the expression levels of ZO-1 and Occludin mRNA significantly decreased in the RA and RC groups (p < 0.05). This indicated that the intestinal barrier was compromised, leading to intestinal fungal leakage and translocation, consequently causing islet inflammation. COS safeguarded the pancreas by upregulating the expression of ZO-1 and Occludin (p < 0.05). Previous studies have also demonstrated that COS enhances the integrity of the intestinal barrier via calcium-sensing receptors.³⁰

Discussion

Research has shown that organ and cell transplantation reduce gut microbial diversity and alters community structures. Even after two decades, recipients of liver and kidney transplants still exhibit gut microbial dysbiosis, characterized by reduced microbial diversity and increased levels of harmful microbial species such as Escherichia coli.³¹ Furthermore, the gut microbial diversity of recipients undergoing allogeneic blood cell transplantation is intricately linked to mortality, with higher microbial abundance correlating with lower mortality rates.³² Consistent with the reports, we found that after ICT, the Alpha and Beta diversity of gut fungi decrease in microbial richness and diversity, accompanied by changes in fungal community structure.

Immunosuppressants are essential drugs after transplantation surgery, which play an important role in regulating the immune process of the body after organ and cell transplantation. For example, rapamycin promotes the induction of regulatory T cells (Tregs) by B7-H1 dependent and IL-1β dependent dendritic cells by inhibiting the mTOR mediated signal transducer and transcriptional activator 3 and inhibits the activation and proliferation of autoreactive T cells in the body, thus regulating the immunity of the body.²⁵ However, immunosuppressants significantly reduce the gut microbial abundance and facilitate structural changes after transplantation. Tacrolimus, a commonly prescribed immunosuppressant in clinical practice, is associated with a significant increase in the relative abundance of Bacteroides and Lactobacillus and a considerable decline in that of Clostridium and Ruminococcus in the human gut.^33,34

Immunosuppressants also affect macrophage polarization. Fetuin (FetA), acts as an immunomodulatory agent, promotes the expression of immunosuppressive molecule PD1 and polarization of M2 macrophages, thereby inhibiting excessive inflammation and alleviating HO and related bone loss in ectopic ossification (HO) lesions.³⁵ Therefore, further researches are needed to investigate the dual effects of GA immunosuppressants on transplant patients and provide a basis for clinical medication.

By further analyzing the gut fungi of pancreatic islet cell transplantation population using immunosuppressants, we found that the relative abundance of Hannarella luteola, Leucosporidium Yakuticum, and Rheum spiciforme significantly decreased, and the relative abundance of Candida albicans, Amphisiella magnigranulosa, Candida dubliniensis, Malassezia globusa, as well as Candida glabrata significantly increased after transplantation. It is worth noting that the most significant difference observed in Candida albicans. It was speculated that the increase in abnormal Candida albicans was attributed to the use of immunosuppressants after ICT, causing pancreatic inflammation on two levels. First, studies have shown that intestinal barrier disruption caused microbial component leakage into the circulatory system and distant organs of the host. Metabolic microbial products exacerbate tissue inflammation and metabolic dysregulation. EVs transport various substances, including RNA, DNA, proteins, and lipids, from one cell type to adjacent or distant cells. Similarly, Gao et al. find that, after the intestinal barrier is disrupted by obesity, EV containing microbial DNA from the microbiome exerted a role in mediating pancreatic inflammation and β-cell dysfunction.³⁶ Therefore, we suggested that some microbial composition of Candida albicans enters the circulation through intestinal leakage, triggering inflammation and immune responses. Additionally, Candida albicans induced macrophages to polarize toward the M1 inflammatory phenotype in in vitro experiments, which may be due to the cell wall components of Candida albicans were recognized by pattern recognition receptors (Pattern recognition receptor, PRRs), thereby causing macrophage polarization and promoting the expression and production of inflammatory cytokines TNF-α, IL-6, and IL-1, as well as chemokines CXCL-1 and CXCL-2, and recruitment of leukocytes.³⁷ The abnormal gut microbiome and higher intestinal permeability allowed the entry of metabolites such as LPS, bile acids and their derivatives, and tryptophan and its metabolites into the bloodstream, which participated in systemic circulation and prompted pancreatic macrophage polarization toward the inflammatory M1 phenotype. These hypotheses and related metabolites will be further confirmed in our subsequent studies.

The pancreas and intestines are physiologically interconnected. The high abundance of gut microbiota and metabolites can be exchanged through the pancreatic duct. Previous studies indicated the presence of microorganisms in the pancreas and compared to the pancreas of patients with pancreatitis and pancreatic ductal adenocarcinoma, normal pancreas contains a large amount of Brevibacteria and Chlamydiales.³⁸ After the disruption of the intestinal microbiota, Muribaculaceae accumulated selectively in the intestine and migrated to the pancreas, causing local inflammation and β-cell damage.³⁹ The fungi displayed a similar effect in the pancreatic tumor model. The fungi migrated from the intestinal lumen to the pancreas, resulting in an approximate 3000-fold higher fungal level in pancreatic tumors than in normal pancreatic tissue, both in pancreatic ductal adenocarcinoma populations and mouse models. Malassezia showed marked enrichment. The removal of fungal communities prevented tumor growth, while Malassezia reproduction accelerated tumor occurrence. The underlying mechanism involved the interaction between the fungal wall and mannose-binding lectin (MBL), activating the complement cascade reaction and promoting cancer progression.⁴⁰ Thus, these evidences strongly support our speculation that intestinal fungal dysbiosis caused by Candida albicans translocation leads to pancreatic inflammation. However, the limitations of current microbial sequencing methods impede the absolute quantification of trace fungi in the pancreas. We will also pay attention to the development of sequencing technology and developing quantitative methods for trace fungal detection.

Since multiple studies have shown that the gut microbiome regulates homo-immunity and rejection, targeting microbial components is a potential therapeutic strategy to improve graft acceptance.⁴¹ Next, we screened functional oligosaccharides regulating the gut microbiota by inhibiting Candida albicans proliferation.

In previous studies, COS inhibits the formation and proliferation of Candida albicans biofilm. Moreover, when combined with antifungal drugs such as fluconazole and miconazole, COS exhibits a synergistic effect on Candida albicans biofilm. Compared to monotherapy with antifungal agents, the combination of COS with drugs effectively reduces Candida albicans resistance and inhibits fungal growth.⁴² In line with previous studies, we have found that COS inhibited the proliferation of Candida albicans, which was further validated by animal experiments (Figures 8(b-d)). These results showed that COS reduced the abnormal increase of Candida albicans and enhanced microbial diversity. COS, acting as a prebiotic, is selectively fermented by microorganisms in the gut. Previous studies have indicated that oligosaccharides can augment the population of beneficial bacteria, such as Lactobacillus and Bifidobacterium, thereby promoting gut microbial compositional equilibrium and diversity.⁴³

Moreover, COS exhibited significant effect in augmenting M2 macrophage levels while decreasing M1 macrophages. In addition to these oligosaccharides, compounds such as mannooligosaccharides (MRO)⁴⁴ and konjac glucomannan⁴⁵ have been found to reprogram macrophages in vitro, resulting in cell polarization toward the M2 type and displaying anti-inflammatory effects. These compounds share a common structure characterized by β-1,2 and β-1,4-glucoside bonds. However, other studies have demonstrated that (1→3)-β-d-glucan oligosaccharides (GOS), linked by β-1,3-glucoside bonds, induce macrophages toward the M1 type.⁴⁶ The β-glucans predominantly found in the cell wall of Candida albicans consist of β-1,3 and β-1,6 glucans. Therefore, COS is ideal for controlling cell polarization and regulating intestinal microbiota.

The molecular mechanism is investigated by examining the TLR4/NF-κB and JAK/STATs pathways involved in macrophage polarization.⁶ TLR4 activation triggers NF-κB) through the MyD88-dependent pathway, promoting inflammatory factor expression.¹⁰ Drugs such as berberine,⁴⁷ quercetin,⁴⁸ and Meis indigo⁴⁹ inhibit M1 polarization by targeting the MyD88/NF-κB pathway. Additionally, the STAT6 pathway is closely associated with M2 polarization, while IL-4 enhances the number of M2-type macrophages by upregulating STAT6 expression.^50,51 Therefore, COS intervention in the MyD88/NF-κB and STAT6 pathways may represent the main reason for regulating macrophage polarization and alleviating pancreatic inflammation.

Abbreviations

CEOS

Cellobiose

COS

Chitooligosaccharides

FOS

Fructooligosaccharides

H&E

Hematoxylin-eosin staining

IAPP

Islet amyloid polypeptide

ICT

Islet cell transplantation

IF

Immunofluorescence

IL-1β

Interleukin-1β

IL-10

Interleukin-10

IR

insulin resistance

STZ

Streptozocin

TGF-β

Transforming growth factor β

TLR2

Toll-like receptor 2

TNF-α

Tumor necrosis factor α

TOS

Turanose

T2D

Type 2 Diabetes

STAT6

Signal transducer and activator of transcription 6

Highlights

• High gut Candida albicans caused abnormal macrophage polarization and islet inflammation after T2D ICT.

• COS significantly inhibited Candida albicans growth during the stationary phase.

• COS shifted macrophage polarization from M1 to M2 and downregulated IL-1β.

• COS reduced islet inflammation in the HMA-T2D model by modulating macrophage polarization via the MyD88/NF-κB-STAT6 pathway and intestinal fungi.

Funding Statement

This work was supported by the Shanghai Sailing Program (23YF1409800), China Postdoctoral Science Foundation (2023M731087), Shanghai Post-doctoral Excellence Program (2022153) and Shanghai Frontiers Science Center of Optogenetic Techniques for Cell Metabolism (Shanghai Municipal Education Commission) and Shanghai Collaborative Innovation Center for Biomanufacturing Technology.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Author contributions

Y.Z. investigation, methodology, formal analysis, writing-original draft, writing-review and editing. X.J. conceptualization, data curation, funding acquisition, investigation, methodology, writing-review & editing. K.C. validation, investigation. H.Y. methodology. M.Z. conceptualization, methodology, formal analysis, supervision, writing-review and editing. L.Z. conceptualization, funding acquisition, project administration, resources, supervision, writing-review and editing.

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/19490976.2024.2442051