Long non-coding RNA CRNDE as potential biomarkers facilitate inflammation and apoptosis in alcoholic liver disease

Search strategy and data collection, preprocessing, and normalization

GEO database (http://www.ncbi.nlm.nih.gov/geo) was electronically searched to collect eligible expression profiling from inception to April, 2019, with keywords including “Alcoholic Liver Diseases” [MeSH] OR “ALD” [MeSH] OR “Alcoholic Hepatitis” [MeSH] OR “AH” [MeSH] OR “Chronic Alcoholic Hepatitis” [MeSH] AND “Homo sapiens” [MeSH]. Only original experimental studies that screened for different expression profiling between ALD and normal in humans were the first choice for inclusion. Exclusion criteria included: (1) repeated reports by the same institute, hospital; (2) a non-expression gene chip; and (3) a nonwhole-genome chip. Next, we download the probe sequence from GEO and re-annotate the probe using bowtie based on the GENCODE Release 19 annotation for lncRNA. All datasets were normalized individually on the base-2 logarithm by Robust Multi-Array Average (RMA) and Linear Models for Microarray (LIMMA) package and annotated by converting different probe IDs to gene IDs. Finally, the 2 datasets TXT files of all gene lists sorted by logFC were integrated using the RobustRankAggreg (RRA) R package [13].

Target prediction and functional analysis of lncRNA CRNDE

The presumed targets of integrated-signature lncRNA CRNDE, firstly, the miRNA targets of lncRNA was predicted using the miRcode database (version 11; http://www.mircode.org/) and mRNA targets of miRNAs using miRTarBase (version 7.0), miRDB (version 5.0, http://mirdb.org/) and TargetScan (version 7.2; http://www.targetscan.org/). Unique genes with target sites in 3ʹUTR were incorporated. To assess the prospective functions of the most significant lncRNA CRNDE, we discharged the Kyoto Encyclopedia of Genes and Genomes (KEGG) using the Database for Annotation, Visualization and Integrated Discovery (DAVID). The P value that narrated KEGG pathway enrich the target gene less than 0.05 was defined as the cutoff criterion.

Preparation of ALD mice model

Adult male SPF BalB/c mice weighing 18-24g were provided by Nanjing Qinglongshan Breeding Farm. Mice were adaptively fed for 2 weeks, freely drinking and eating. 16 mice were randomly divided into dose group I (2 g·Kg^-1 group), dose group II (4 g·Kg^-1 group), dose group III (6 g·Kg^-1 group) and dose group IV (8 g·Kg^-1 group), each group of 4. After 2 weeks of adaptive feeding, when the mice are in a stable state, Gujing Gongjiu with an alcohol concentration of 45° is selected for gavage experiment on the mice. According to the calculation formula, calculate the amount of alcohol ingested by mice: alcohol intake (ml) = alcohol intake (g) ÷ (alcohol content × 0.8), where alcohol intake (g) = mouse body weight (Kg) × Dose (g·Kg^-1), and the alcohol content is 0.45, that is, alcohol intake (ml) = mouse weight (Kg) × dose (g·Kg^-1) ÷ 0.36. The above-mentioned gavage is performed every 3 days, 5 times, and the duration of the gavage is 15 days. Observe and record the behavior, physical signs and death of mice after alcohol gavage. A dose group with a lower mortality rate was selected to prepare for the formal preparation of a mouse alcoholic liver disease model. After 15 days of intragastric administration, the mortality of mice in dose group III and dose group IV was 75% and 100%, respectively. The mortality of mice in dose group III and dose group IV was too high to meet the requirements of the model preparation process. Stability and repeatability requirements. The survival rates of mice in dose group I and dose group II are 100% and 75%, respectively, which can meet the requirements for the preparation of alcoholic liver disease models in mice. Therefore, the gavage doses 2 g·Kg-1 and 4 g·Kg^-1 corresponding to dose group I and dose group II are selected as the actual gavage dose prepared for this alcoholic liver disease model, where 2 g·Kg^-1 is The gavage dose of the low-dose group, 4 g·Kg^-1 was the gavage dose of the high-dose group.

Next, 48 adult male SPF BalB/c mice were randomly divided into a blank control group, a model control group, and an alcoholic liver disease model group. The alcoholic liver disease model group was divided into a low-dose group according to different alcohol doses (2 g·Kg^-1), high-dose group (4 g·Kg^-1 body weight), 12 animals per group. The mice in the blank control group were not treated, and the mice in the model control group were given saline. The intake of physiological saline is equivalent to the alcohol intake of the middle-dose group, that is, the intake of physiological saline in the model control group (V health, ml) = 11.11 × mouse body weight (Kg). The different dose groups were given the corresponding alcohol intake of liquor, that is, the alcohol intake of the low-dose group (V low, ml)=5.55×the weight of the mouse (Kg), and the alcohol intake of the high-dose group (V, ml) = 11.11 × mouse body weight (Kg).

Detection of serum biochemical indicators in mice

According to the peak concentration time of ethanol absorption in the body and the peak concentration time of some biochemical indicators, 9 hours after the completion of the modeling, the mice were taken from the eyeballs and blood was collected, and the blood was collected in an EP tube. The blood sample in the EP tube was balanced and placed in a high-speed refrigerated centrifuge., 1200g, 4° C centrifugation for 10min, after centrifugation, take the supernatant (serum) and store it in the -80° C refrigerator. Using Hitachi 7600 automatic biochemical analyzer and Biyuntian Biotechnology Co., Ltd. detection kit to determine serum biochemical indicators, including aspartate aminotransferase (AST), alanine aminotransferase (ALT), total bilirubin (TBIL), and direct bilirubin (DBIL).

Tissue collection

After the mice was sacrificed, the mice’s chest and abdomen were opened, and the mice’s liver, brain, heart, lung, spleen, pancreas, and kidney were quickly taken out and placed in a numbered 4ml EP tube, placed on dry ice for quick freezing and then placed in the refrigerator at -80° C. Take part of the liver tissue and place it in a 10% paraformaldehyde solution for fixation for 24 hours. The remaining organs are stored in a refrigerator at -80° C for inspection. Take out the liver tissue fixed with 10% paraformaldehyde solution, put it face down into the embedding box, dehydrate it on a dehydrator and then use xylene for transparency. Finally, the transparent brain tissue is immersed in wax and wrapped. After hematoxylin-eosin (HE) staining of liver tissues of each group of mice, the Leica microscope was used to observe and collect images.

CBA method to detect mice serum cytokines

Take blood from mouse eyeballs and place them in 2.5ml EP tube at 1500rpm×15min. After centrifugation, the upper layer is serum, and the lower layer is blood clot. Take the upper layer of serum into a 1.5ml EP tube and place it in a refrigerator at -80° C. Starting from the highest concentration standard, add 300μl multiples of diluent one by one, and dilute the standard gradually. Take 300μL of the solution from the highest concentration standard tube and add it to the 1:2 tube, pipette to mix, and then take 300μl from the 1:2 tube to the 1:4 tube, and so on, until the 1:256 tube. FCM software was used to analyze and calculate the expression of IL-6, IL-10, MCP-1, TNF, IFN-γ, and IL-12p70 on each group of cells.

Determination of the apoptosis rate of mice hepatocytes

The hepatocytes in the logarithmic growth phase were processed according to the aforementioned grouping method, and washed with PBS. Then add an appropriate amount of trypsin to digest the cells, discard the trypsin, add the culture medium to resuspend the cells and count. Use flow cytometry to detect cell apoptosis rate.

Statistical analysis

Data are all expressed by (mean ± SEM); homogeneity of variance data, one-way analysis of variance (one-way ANOVA) is used for multiple comparisons between groups, Student’s-t test is used for pairwise comparisons between groups. All data were statistically analyzed by SPSS 17.0 (SPSS, Chicago, IL) software. P<0.05 indicates that the difference is statistically significant. The test level α=0.05, P<0.05 indicates that the difference is statistically significant.

Integrated analysis of lncRNA CRNDE in 2 datasets

We located and manually curated 359 published and publicly available ALD datasets. We identified 2 datasets (GSE28619 [14] and GSE143318 [15]) meet the criteria including 20 ALD samples and 12 normal. The study selection flow chart for this integrated analysis is shown in Figure 1A. Using p <0.05 and [logFC]>1 as cut-off criterion, a total of 39 lncRNAs were significantly deregulated in GSE143318 datasets, and 35 in GSE28619 datasets (Figure 1B, 1C). After merging the datasets, 3 differential lncRNAs were identified including 2 downregulated lncRNAs (BMS3-AS3 and LINC01088) and 1 upregulated lncRNA (lncRNA CRNDE) (Figure 2A). Therefore, we focus on lncRNA CRNDE.

Next, in order to predict the possible potential functions of lncRNA CRNDE, we first predicted the target genes of lncRNA CRNDE, and used these target genes to intersect the difference genes of GSE28619. A total of 114 genes were found (Figure 2B). In addition, we use GSEA to analyze the function of 114 genes, and GSEA GO analysis revealed lncRNA CRNDE is significantly involved in the extrinsic apoptotic signaling pathway via death domain receptors, intrinsic apoptotic signaling pathway in response to DNA damage, regeneration, and regulation of adherens junction organization (Figure 2C). GSEA KEGG pathways analysis revealed lncRNA CRNDE is significantly involved in the MAPK signaling pathway, apoptosis, wnt signaling pathway and hematopoietic cell lineage (Figure 2D). In summary, lncRNA CRNDE may play a important role in the apoptosis pathway.

Preparation of ALD mice model

In order to verify the function of lncRNA in vitro, we have established an ALD model. As shown in Table 1, the weight of mice in the control group and the model group showed an upward trend. The weight of the mice in the low-dose group and the high-dose group was lower than that of the blank control and the model control group, and the weight of the mice in the low-dose and high-dose groups changed steadily, especially the high-dose group. The survival rate of mice in the blank control group and the model control group was 100%. The survival rates of the mice in the low-dose group and the high-dose group were 83.33% and 58.33%, respectively. The survival rate of the mice in the high-dose group was obvious lower in the lower-dose group (Table 2). Next, we detect the serum biochemical indicators in mice. Compared with the blank control group, the serum TP, GLOB, ALT and AST of the mice in the high-dose group were all increased (P<0.05). The serum ALB of the mice in the high-dose group was lower than that of the blank control group (P<0.05). Compared with the control group, mouse serum TBIL, DBIL and GGT did not change significantly (Table 3).

Next, in order to verify whether alcohol can affect the liver structure, we use HE staining. The results of the blank group and the model showed that the structure of the liver lobules was clearly distinguishable, the liver cords were arranged radially, the liver cells were present, the nuclei were large and round, regular in size, blue stained, cytoplasm powder stained, no steatosis or inflammatory cell infiltration (Figure 3A, 3B). The results of the low-dose group showed that the structure of the liver lobules was discernible, the hepatic cords were arranged irregularly, the erythrocytes in the sinusoids were stagnated, and a small amount of inflammatory cell infiltration around the blood vessels (Figure 3C). Liver cells are swollen, with nuclei of varying sizes, a small amount of nuclei disappeared, cytoplasm is loose, lightly stained and homogeneous, and vacuoles appear in the cytoplasm of some liver cells with varying sizes (Figure 3D). Moreover, the results of the high-dose group showed that the structure of the liver lobules is still indistinguishable, the hepatic cords are arranged disorderly, the red blood cells in the sinusoids are stagnated, and a large number of inflammatory cells infiltrate around the blood vessels (Figure 3E). The cytoplasm of hepatocytes is loose and lightly stained, the nucleus staining varies in depth and size, and some of the nuclei of hepatocytes disappear; the cytoplasm is lightly stained, and vacuoles appear in the cytoplasm of most hepatocytes with varying sizes (Figure 3F). In conclusion, the mice ALD model was validated by serum biochemical test and HE staining. The results showed that the animal model of chronic alcohol infusion in mice was successfully established, for the next experimental study of the function of lncRNA CRNDE and other animal models to lay a solid foundation.

Increased lncRNA CRNDE expression in ALD mice

To investigate the correlation between lncRNA CRNDE and ALD, we performed qRT-PCR to detect relative lncRNA CRNDE expression ALD mice, and found ALD tissues in mice had significantly higher lncRNA CRNDE levels than normal tissues (Figure 3G). Due to inflammatory mediators play a vital role in the occurrence and development of alcoholic liver disease, therefore, we use the CBA method to detect changes in the expression of mice serum cytokines (Figure 4A). The results showed that compared with the control group, the expression of IL-6 in the low-dose group was significantly increased (Figure 4B), while other inflammatory mediators such as IL-10, MCP-1, IFN-γ, TNF, IL-12p70 did not change significantly (Figure 4C–4G).

Increased lncRNA CRNDE facilitates apoptosis in ALD

In order to explore the possible mechanism of lncRNA CRNDE in ALD, we used flow cytometry to detect the apoptosis rate of mice hepatocytes. Data revealed that compared with the model control group, the apoptosis rate of mice hepatocytes was significantly increased. Moreover, with the increase of alcohol concentration, the higher the apoptotic rate of mice liver cells (Figure 5A, 5B). Next, mice receiving saline were injected with the same amount of empty virus (Sh-RNA group) and low-dose alcoholic were injected with the same amount of siRNA-CRNDE, and then perform flow cytometry detection on the model control group and low-dose group. We found that compared with the model control group, the mouse liver cell apoptosis rate in the low-dose group was significantly reduced (Figure 5C, 5D). In conclusion, lncRNA CRNDE may affect ALD through apoptotic pathway.