Unmethylated Cytosine phosphate guanine DNA (CpG-DNA) exacerbates liver Kupffer cells (KCs) inammation through Toll-like receptor 9 (TLR9) in diabetic rats

Recently, gut microbiota for various pathogens has attracted attention. The present study investigated the role of gut microbiota unmethylated cytosine phosphate guanine DNA (CpG-DNA) on liver Kupffer cells (KCs) inammatory cytokine interleukin-1β (IL-1β) in diabetic rats. We induced diabetic rats models and sequenced the gut microbiota composition of fecal samples. We also applied CpG-DNA and TLR9 inhibitor on KCs to investigate the regulation of inammatory cytokine IL-1β and Toll-like receptor 9 (TLR9) signaling pathway. We found a signicant difference of gut microbiota between the control and the diabetic rats with increased Clostridium. Meanwhile, diabetes could upregulate TLR9 in KCs and increase IL-1β concentration. Furthermore, high concentration of unmethylated CpG-DNA could signicantly increase IL-1β secretion while it was suppressed by TLR9 inhibitor in KCs cultured in high glucose medium. Our study suggests that unmethylated CpG-DNA, which was highly expressed in diabetic rats, activated KCs through TLR9, and induced IL-1β secretion in vitro and in vivo which plays an important role in diabetic liver inammation. It may contribute to the progress of the diabetes.


Introduction
Diabetes mellitus (DM) is one of the most prevalent metabolic disorders which induced by both genetic and environment factors [1]. Accumulating evidences indicated that altered gut microbiota existed in patients with DM who had an increase of pathogens and a decrease of probiotics. It may contribute to impaired glucose tolerance and insulin signaling [2,3]. Pathogen and its metabolites such as lipopolysaccharide (LPS) and unmethylated cytosine phosphate guanine DNA (CpG-DNA) could reach the liver through the portal vein, arousing low-grade in ammation, while hyperglycemia drives intestinal barrier dysfunction which might contribute to the DM [4]. Unmethylated CpG-DNA is the speci c component of bacterial DNA containing CpG motifs called pathogen associated molecular pattern (PAMP) [5]. The length, number and lateral sequence of unmethylated CpG-DNA that differ from different microbiota and determine their immunocompetence [6]. Pathogens such as Clostridia and Escherichia coli express higher unmethylated CpG-DNA that had stronger immunocompetence compared to the probiotics [7]. Recent study identi ed unmethylated CpG-DNA as a pathogen-dependent antigen that activated macrophage and induced in ammatory factors secretion [8]. PAMP can be speci cally identi ed by pattern recognition receptor (PRR) including toll-like receptors (TLRs) in rats and human [5]. TLR9 is a major receptor of TLRs family that expresses in macrophage such as Kupffer cells (KCs) in liver and speci cally recognizes unmethylated CpG-DNA [9,10]. Activated KCs secreted various proin ammatory cytokines including interleukin-1β (IL-1β), a well-known in ammatory factor contributed to the low-grade in ammation and insulin resistance (IR) [11,12]. IL-1β over expression promoted phosphorylation and degradation of IRS, leading to the suppression of insulin signaling [13,14] which promoted IR as well as the development of DM. However, the role of gut microbiota CpG-DNA on the development of DM is still elusive. The present study investigated the role of CpG-DNA on liver KCs to explore the possible mechanism of CpG-DNA activating KCs by TLR9 that induce liver in ammation in the progress of DM.

Experimental animal models
Male SD rats (120-150g) (Shanghai laboratory animal center, Shanghai, China) were randomly divided into the control group (CT) and the diabetes mellitus group (DM). 24 h after fasting, the streptozotocin (65 mg/kg body weight, dissolved in 0.01 M citrate buffer, pH 4.5; Catalog: S8050, Solarbio, Beijing, China) were intraperitoneal injected to the rats of DM group while CT group were injected with equal volume of the vehicle. 72 h after diabetes conduction, the rats with blood glucose above 16.7 mmol/L were considered the successful diabetes models.
After six weeks, all rats were anesthetized by sodium pentobarbital (45 mg/kg body weight). Blood samples were obtained from hepatic portal vein to evaluate the level of IL-1β. The liver and the body weight were also recorded. The liver of half rats in each group were removed and xed in 4% paraformaldehyde solution overnight for hematoxylin-eosin (H&E) and immunohistochemical staining, while another half liver in each group were preserved in liquid nitrogen for Western blot and ELISA assay.

Gut microbiota analysis
Fresh fecal samples were collected at the time of blood sampling, snap frozen in liquid nitrogen, and stored at -80°C until analysis. Sequencing and data processing were conducted by Illumina MiSeq . Raw fastq les were demultiplexed, quality-ltered by Trimmomatic and merged by FLASH with the following criteria: (i) The reads were truncated at any site receiving an average quality score <20 over a 50 bp sliding window. (ii) Primers were exactly matched allowing 2 nucleotide mismatching, and reads containing ambiguous bases were removed. (iii) Sequences whose overlap longer than 10 bp were merged according to their overlap sequence. Operational taxonomic units (OTUs) were clustered with 97 % similarity cutoff using UPARSE version 7.1 http://drive5.com/uparse/) and chimeric sequences were identi ed and removed using UCHIME. The taxonomy of each 16S rRNA gene sequence was analyzed by RDP Classi er algorithm (http://rdp.cme.msu.edu/) against the Silva (SSU123) 16S rRNA database using con dence threshold of 70%.

Hematoxylin-eosin staining (H&E) and immunohistochemical staining
The livers (in 4% paraformaldehyde solution) were embedded in para n, then cut into 6-μm section. The sections were respectively stained with H&E to determine pathological changes. Images were observed by the microscope (BX-53, Olympus Corp, Tokyo, Japan).
The rest sections from each group were stained with immunohistochemical assay and incubated with the primary antibodies against TLR9 (TLR9, Catalog: ab134368, 1:200, Abcam, USA). After overnight incubation at 4 ℃, sections were washed with PBS (3 × 5 min) and incubated with biotinylated and a nity-puri ed IgG secondary antibodies for 1 h at room temperature. Images were observed by the microscope (BX-53, Olympus Corp, Tokyo, Japan) and analyzed by Image-Pro Plus 6.0 (Media Cybernetics, USA).
KCs were cultured with TLR9 agonist and inhibitor

Western blot analysis
The expression of proteins in liver and cultured cells were detected by Western blot assay. Equal amounts of protein were separated by 10 % SDS-PAGE. It was transferred from the gel to nitrocellulose membrane, then blocked by 5 % skimmed milk (Tris-buffered saline of 0.1 % Tween 20) for 1 h at room temperature.
ELISA analysis ELISA analysis of IL-1β levels in tissue homogenates, serum and cell culture supernatant of KCs were detected by Rat IL-1β ELISA Kit (Catalog: E-EL-R0012c, Elabscience, Wuhan, China). Sample collection, reagent preparation and assay procedure were strictly according to the kit instruction. The microplate reader (Thermo, USA) was used to assay the optical density (OD) values at 450 nm.

Statistical analysis
The data are presented as mean ± standard deviation (SD) and analyzed by SPSS 18.0 software (SPSS Inc., USA). The differences among both groups were performed with one-way analysis of variance (ANOVA). P values <0.05 were considered statistically signi cant.

Gut microbiota changes in diabetic rats
To validate the difference of gut microbiota communities, we investigated fetal samples from rats with or without DM. Results showed that bacteria diversity reduced in DM rats, compared to the normal rats (Fig.  1a). Abundance of different bacteria class (Fig. 1b), Sample distances heatmap (Fig. 1c) and hierarchical clustering tree (Fig. 1d) all presented signi cant difference between the control group and the diabetes group. The diabetic rats had more pathogens (such as Clostridiawere) and less probiotics (such as Bacteroidia and Gammaproteobacteria) than the normal rats (P<0.05, vs control rats).

Pathogen-dependent unmethylated CpG-DNA activates KCs through TLR9
We adopted H&E staining to con rm the in ammatory response in diabetic liver. We found in ammatory in ltration and cellular morphology changed in diabetic rats (Fig. 2a). Kupffer cells (KCs), the major macrophage in liver, playing an important role in in ammation. As a speci c receptor of unmethylated CpG-DNA, immuno uorescent staining (Fig. 2b) and western blot assay (Fig. 2c-d) revealed that TLR9 was signi cantly increased by unmethylated CpG-DNA in DM rats (P<0.05, vs control rats). Thus, increasing TLR9 might present the activation of KCs in diabetic rats.

TLR9 regulated KCs IL-1β secretion in virto and in vivo
KCs activated initiate immune response for secreting a series of proin ammatory cytokines including IL-1β, a well-known proin ammatory cytokine. The western blot assay con rmed that an increased level of IL-1β in liver of DM rats, compared to the normal rats (P<0.05, Fig. 2c, 3a). ELISA analysis consistently showed the same results in both serum (P<0.05, Fig. 3b) and liver tissues (Fig.3c). In addition, ELISA assay showed that after 12 hours incubation with unmethylated CpG-DNA (CU-CPT17e), KCs signi cantly increased IL-1β expression. However, IL-1β expression was higher than the diabetic group without agonist treatment (P<0.05). In contrast, E6446 dihydrochloride, an inhibitor of TLR9, suppressed IL-1β expression in the diabetic KCs signi cantly (P<0.05, Fig.3d). These results indicated that activated KCs by unmethylated CpG-DNA increased IL-1β expression and inhibitor of TLR9 could reduce IL-1β secretion.

Discussion
Diabetes mellitus, characterized by glucose and insulin metabolism disturbances, is a rapidly increasing worldwide prevalence [15,16]. Recently, studies showed that gut microbiome was involved in various pathogens via generating bioactive metabolites [17]. Gut microbiota composition has a strong in uence on weight gain and metabolic dysfunction [18]. Consistent with previous study [3] the present sequencing results revealed that the composition of gut microbiota in DM rats was different from the control rats, with signi cantly decreasing probiotics such as Bacteroidia and increasing pathogen such as Clostridia. Accordingly, several treatments based on supplement of probiotics also exhibited positive effects on DM patients [19,20].
Pathogen and its metabolites might contribute to in ammation and diabetes development [4]. The studies have favored the roles of LPS in diabetes, and the results showed that LPS could initiate lowgrade in ammation and impair glucose tolerance [21,22]. However, as another major metabolite of pathogen, unmethylated CpG-DNA had the same capability as LPS [4]. Since unmethylated CpG-DNA is a speci c component of bacteria DNA concentrating in pathogen, the concentration of unmethylated CpG-DNA would increase with more pathogens and less probiotics in diabetic rats. KCs, functioned as the dominant liver intravascular phagocytes, response to the pathogens and their metabolites [23]. In addition, as the speci c receptor of unmethylated CpG-DNA, TLR9 was expressed in Kupffer cells providing a solid foundation for activation of KCs by unmethylated CpG-DNA, and induction of interleukin-1β in mice [11].
The present study con rmed that liver in ammation occurred and altered gut microbiota composition in diabetic rats. Interestingly, we found that diabetes promoted expression of TLR9 in KCs. As we know, KCs were the major source of IL-1β in liver [12] and played an important role in in ammation and DM development [13,14]. Increasing IL-1β could down regulate PI3K/Akt insulin signal pathway and lead to the progress of DM. We found that IL-1β was not only present at higher concentration in portal blood and liver tissue in DM rats, but in the cultured KCs treated by high concentration unmethylated CpG-DNA. Furthermore, TLR9 inhibitor treatment could suppress the CpG-DNA induced IL-1β secretion. The present data suggested that upregulation of TLR9 could activate KCs to promote in ammatory response and 1β secretion which may play a role in DM development.
In conclusion, the present results demonstrated that gut microbiota unmethylated CpG-DNA could activate TLR9 of KCs and increase IL-1β secretion in the diabetic liver and cultured KCs, leading to liver in ammation. It may contribute to the DM development. The current study for the rst time provided a potential target for the therapy of DM with improving gut microbiota.

Declarations
Funding: The study is supported by the Natural Science Foundation of Zhejiang Province China (LY18H170002) and the National Natural Science Foundation of China (81472149).

Con ict of Interest:
The authors declare that there are no con icts of interest.
Availability of data and material: The datasets used or analyzed during the current study are available from the corresponding author on reasonable request.
Code availability: Not applicable.

Authors contributions:
All the authors contributed to conception and study design. Cai contributed to acquisition of the data, Li and Cai contributed to analysis of the data. All the authors contributed to drafting and critical revision of the manuscript of important intellectual content. All the authors have given nal approval of the version to be published and agreed to be accountable for all aspects of the work.

Ethics approval
The animal experiments were approved by the Committee of Animal Experiment Center of Zhejiang University (Hangzhou, China) which complies with the guidelines of the National Institutes of Health guide for the care and use of Laboratory animals (NIH Publications No. 8023, revised 1978).