If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
1 These authors have contributed equally to the work.
Wenqi Fan
Footnotes
1 These authors have contributed equally to the work.
Affiliations
National Clinical Research Center for Metabolic Diseases, Key Laboratory of Diabetes Immunology (Central South University), Ministry of Education, and Department of Metabolism and Endocrinology, The Second Xiangya Hospital of Central South University, Changsha, China
1 These authors have contributed equally to the work.
Haipeng Pang
Footnotes
1 These authors have contributed equally to the work.
Affiliations
National Clinical Research Center for Metabolic Diseases, Key Laboratory of Diabetes Immunology (Central South University), Ministry of Education, and Department of Metabolism and Endocrinology, The Second Xiangya Hospital of Central South University, Changsha, China
National Clinical Research Center for Metabolic Diseases, Key Laboratory of Diabetes Immunology (Central South University), Ministry of Education, and Department of Metabolism and Endocrinology, The Second Xiangya Hospital of Central South University, Changsha, China
National Clinical Research Center for Metabolic Diseases, Key Laboratory of Diabetes Immunology (Central South University), Ministry of Education, and Department of Metabolism and Endocrinology, The Second Xiangya Hospital of Central South University, Changsha, China
National Clinical Research Center for Metabolic Diseases, Key Laboratory of Diabetes Immunology (Central South University), Ministry of Education, and Department of Metabolism and Endocrinology, The Second Xiangya Hospital of Central South University, Changsha, China
National Clinical Research Center for Metabolic Diseases, Key Laboratory of Diabetes Immunology (Central South University), Ministry of Education, and Department of Metabolism and Endocrinology, The Second Xiangya Hospital of Central South University, Changsha, China
To characterize the exosomal miRNA profiles of latent autoimmune diabetes in adults (LADA) and evaluate the biomarker potential of selected miRNAs to distinguish LADA from type 2 diabetes (T2D).
Methods
Plasma-derived exosomal miRNA expression profiles were measured in patients with LADA (N = 5) and control subjects (N = 5). Five differentially expressed miRNAs were selected to validate their expression levels and assess their diagnostic potential by quantitative real-time PCR (qRT–PCR) in a larger cohort.
Results
Seventy-five differentially expressed plasma-derived exosomal miRNAs were identified in LADA patients compared to healthy subjects. The expression levels of three exosomal miRNAs (hsa-miR-146a-5p, hsa-miR-223-3p and hsa-miR-21-5p) were significantly different between the LADA group and the T2D group. The three miRNAs exhibited areas under the receiver operating characteristic curves of 0.978, 0.96 and 0.809, respectively.
Conclusions
This study uncovers the miRNA profiles of plasma-derived exosomes from LADA patients and identifies exosomal miRNAs as potential biomarkers to discriminate LADA from T2D for the first time. Our data demonstrate the function of exosomal miRNAs in the development of LADA and contribute to an in-depth understanding of the precise mechanisms underlying the pathogenesis of LADA.
Patients with latent autoimmune diabetes in adults, abbreviated LADA, exhibit clinical features of both type 1 diabetes (T1D) and type 2 diabetes (T2D) [
]. It is a slowly progressive autoimmune disease. Patients with LADA have serum immune markers of T1D but can be treated initially without insulin injection [
]. As its name implies, LADA is an adult-onset diabetes, and individuals with LADA have at least one organ-specific autoantibody, including autoantibodies against glutamic acid decarboxylase (GADA), insulin (IAA), insulinoma antigen-2 (IA-2A), islet-specific zinc transporter isoform 8 (ZnT8A), and tetraspanin 7, of which GADA is considered the most sensitive marker for LADA, as approximately 90% of patients with LADA are GADA positive [
]. However, given that the islet autoantibody assay is not universal in every hospital, some patients can progress to diabetes mellitus without islet autoantibody positivity [
]. In addition, because the early signs of LADA patients are similar to those of T2D patients, LADA patients are often misdiagnosed with T2D. It has been reported that 10–14% of West Africans with apparent T2D are misdiagnosed with LADA, a higher percentage than that in Europe (8–10%) [
]. Therefore, the identification and etiology detection of LADA are imperative to improve healthcare management.
Exosomes are 30–200-nm extracellular vesicles (EVs) secreted by most cell types. They have lipid bilayer membranes and can be isolated from several body fluids, such as plasma, urine, cerebrospinal fluid, epididymal fluid, and breast milk [
]. As communication messengers that can interact with other cells, exosomes contain a large variety of contents, such as proteins, DNA, RNA and lipids, which can induce biological information transfer from donor cells to recipient cells to mediate various physiological activities or pathological changes [
]. Given that the number and content of exosomes depend on the originating cells and are strictly regulated by physiological and pathological factors, exosomal RNAs are ideal biomarkers [
]. In addition, the surrounding membrane confers stability to exosomes. A previous study reported that islet-derived exosomal RNAs in a T1D model exhibited a significantly altered expression profile after proinflammatory cytokine treatment, and the differentially expressed RNAs were predicted to be related to beta-cell function, which indicated the potential of exosomal RNAs as biomarkers [
] in T1D patients and found that some differentially expressed RNAs might play a role in the onset of T1D. Some studies have also indicated the biomarker potential of exosomes in some diseases. For example, it has been suggested that circulating EV-enriched fractions have a different miRNA profile in colon cancer patients than in normal controls and that EV-derived miRNAs are promising biomarkers to identify colon cancer at an early stage [
]. Recently, the expression levels of plasma-derived exosomal miR-205-5p, miR-429, and miR-375-3p were validated to be different between healthy controls and esophageal squamous cell carcinoma (ESCC) patients, and the three miRNAs were further found to have the ability to serve as potential biomarkers for ESCC diagnosis [
At present, there are no reports about exosomes in LADA patients. Our study not only profiles the plasma-derived exosomal miRNAs in LADA patients and evaluates their potential as biomarkers to distinguish LADA from T2D patients for the first time but also provides a foundation for their clinical application.
2. Subjects, materials and methods
2.1 Study subjects
We recruited 5 LADA patients and 5 age-matched (P = 0.58) and sex-matched (P = 1.00) healthy subjects in the study from the Second Xiangya Hospital of Central South University (CSU) (Table 1). Diabetes was diagnosed according to the World Health Organization (WHO) 1999 criteria. LADA was diagnosed based on (1) onset age > 30 years, (2) the presence of at least one diabetes-associated autoantibody (GADA, IA-2A, or ZnT8A) [
] no ketosis or ketoacidosis. Patients with pregnancy; malignant tumors; other autoimmune diseases; or a recent cardiovascular event including myocardial infarction, ischemic heart disease, cardiac hypertrophy or heart failure in the past 3 months were excluded given that these cardiovascular diseases influence the cargoes and numbers of exosomes [
]. The T2D patients met the WHO 1999 diabetes diagnostic criteria, and their islet autoantibodies were negative. The healthy subjects had no autoimmune diseases, cancers, or family history of diabetes. Additionally, we used a new cohort to verify the expression levels of sequencing data, including 15 patients with LADA and 15 age-matched (P = 0.412) and sex-matched (P = 0.705) healthy subjects. Moreover, 15 patients with LADA and 15 age-matched (P = 0.091) and sex-matched (P = 0.713) T2D patients were enrolled to screen out potential biomarkers that could distinguish LADA patients from T2D patients (Table 2). The study was approved by the institutional ethics review board of the National Clinical Research Center for Metabolic Diseases, and all experiments complied with the ethical principles of the Declaration of Helsinki. All the participants provided written informed consent.
Table 1Characteristics of the patients with LADA and healthy subjects in the study.
Characteristic
LADA (n = 5)
Control (n = 5)
P value
Male sex, % (n)
60.00 (3)
60.00 (3)
1.00
Age, years
46.80 ± 8.70
49.40 ± 4.83
0.58
Duration, months
13.9 ± 10.27
NA
NA
BMI, kg/m2
23.64 ± 1.63
25.60 ± 3.70
0.32
HR, bpm
82.4 ± 10.55
81.00 ± 11.17
0.85
Systolic blood pressure, mmHg
123.80 ± 12.70
120.75 ± 12.47
0.73
Diastolic blood pressure, mmHg
78.40 ± 7.13
72.00 ± 6.06
0.20
FBS, mmol/L
6.48 ± 1.22
5.23 ± 0.39
0.06
HbA1c, %
7.40 ± 2.00
5.37 ± 0.23
0.04
HbA1c, mmol/mol
68.18 ± 23.34
36.28 ± 2.12
0.04
TC, mmol/L
4.2 ± 1.29
3.9 ± 0.56
0.68
HDL, mmol/L
1.18 ± 0.28
1.05 ± 0.21
0.42
LDL, mmol/L
2.63 ± 1.09
2.43 ± 0.47
0.72
TG, mmol/L
1.09 ± 0.66
1.57 ± 0.87
0.35
CREA, μmol/L
69.14 ± 8.75
73.28 ± 3.95
0.36
BUN, mmol/L
4.82 ± 1.34
5.62 ± 1.10
0.34
UA, umol/L
344.94 ± 99.95
307.78 ± 34.97
0.46
UACR, mg/g
8.03 ± 7.18
NA
NA
Nephropathy, % (n)
0.00 (0)
NA
NA
Neuropathy, % (n)
0.00 (0)
NA
NA
Retinopathy, % (n)
0.00 (0)
NA
NA
Dyslipidemia, % (n)
0.00 (0)
40.00 (2)
0.48
Hypertension, % (n)
0.00 (0)
0.00 (0)
1.00
Currently smoke, % (n)
20.00 (1)
NA
NA
Cigarettes per day, n
5.00
NA
NA
Duration of smoking, years
13.00
NA
NA
Note: The data are presented as the mean ± SD and % (n).
2.2 Clinical features, biochemical measurements and autoantibody assays
Body height and weight were measured by clinical physicians, and body mass index (BMI) was calculated by dividing weight by the square of height. Heart rate (HR) was recorded in beats per minute (bpm), and blood pressure was recorded in mmHg under resting states. The levels of blood glucose, C-peptide (CP), hemoglobin A1c (HbA1c), four lipids (triglycerides TGs, total cholesterol TC, high-density lipoprotein HDL, low-density lipoprotein LDL), creatinine (CREA), blood urea nitrogen (BUN), uric acid (UA) and the urinary albumin/creatinine ratio (UACR) were measured at the Department of Clinical Laboratory in the Second Xiangya Hospital. Fundus examination and sensory testing were performed in the Second Xiangya Hospital. Hypertension was identified as systolic blood pressure ≥ 140 or diastolic blood pressure ≥ 90 mmHg, or current taking medication for hypertension. Dyslipidemia was identified as serum TGs ≥ 1.7 mmol/L, HDL < 1.04 mmol/L for men and < 1.30 mmol/L for women, or currently use of antihyperlipidemic medication. GADA, IA-2A and ZnT8A were detected by radioligand assays in a core laboratory (Diabetes Center, CSU) as previously described [
]. The sensitivity and specificity were 82% and 96.7% for GADA, 76% and 100% for IA-2A, and 76% and 100% for ZnT8A, respectively, according to the Islet Autoantibody Standardization Program (IASP) 2020. Based on the 99th percentile observed in 405 healthy participants, the cutoff values were 3.3 U/ml for IA-2A positivity and 18 U/ml (WHO unit) for GADA positivity. The threshold antibody index for ZnT8A was 0.011.
2.3 Isolation and characterization of exosomes
Human peripheral blood samples from LADA subjects, T2D subjects and healthy subjects who had been fasting for at least 8 h were collected in EDTA tubes between 8:00 and 9:00 am and separated at 3000 × g for 15 min at 4 °C to acquire plasma within 30 min. The plasma samples were then stored at −80 °C until use. We used size exclusion chromatography to isolate exosomes from the plasma samples and used transmission electron microscopy (TEM), nanoparticle tracking analysis (NTA), and western blotting (WB) to verify exosomes. The detailed method has been described in our previous study [
]. Briefly, plasma samples were diluted with PBS and purified with Exosupur® columns (Echobiotech, China). The samples were further eluted according to the manufacturer’s instructions. The eluate fractions were then concentrated by using 100-kDa-molecular-weight-cutoff Amicon® Ultra spin filters (Merck, Germany). Moreover, we performed a bicinchoninic acid (BCA) assay for exosomal protein quantification by a commercial BCA assay kit (Thermo ScientificTM Pierce™ BCA Protein Assay Kit, catalog no. 23225, USA). In addition, we used a nanoflow cytometer (N30E Nanoflow Analyzer, NanoFCM Inc., Xiamen, China) to measure the size distribution and particle concentration of exosomes at EchoBiotech Co. Ltd., Beijing, P. R. China. Briefly, the side scatter intensity (SSI) was measured by loading standard polystyrene nanoparticles (250 nm) into the nanoflow cytometer. Then, the isolated sEV sample diluted with PBS (according to the BCA protein assay results, the exosomes were diluted to 1–10 ng/μl) was loaded into the nanoflow to measure the SSI. Finally, the concentration of EVs was calculated according to the ratio of the SSI to the particle concentration in the standard polystyrene nanoparticles. For size measurement, standard silica nanoparticles with mixed sizes (68 nm, 91 nm, 113 nm, 155 nm) were loaded into the nanoflow cytometer to generate a standard curve, followed by loading of the exosome sample. The size distribution was calculated on the basis of the standard cure.
2.4 RNA extraction and miRNA sequencing
The exosomal RNA was prepared using an miRNeasy Serum/Plasma Advanced Kit (Qiagen, Cat. No. 217204). The extraction procedure was conducted according to the kit instructions. The RNA Nano 6000 Assay Kit of an Agilent Bioanalyzer 2100 System (Agilent Technologies, CA, USA) was used to determine the concentration and purity of RNA. Sequencing was performed on an Illumina HiSeq platform after library construction and evaluation.
2.5 miRNA analysis
We used Bowtie tools software to align the clean reads with the Silva database, GtRNAdb database, Rfam database and Repbase database sequence alignment. Then, we filtered ribosomal RNA, transfer RNA, small nuclear RNA, small nucleolar RNA and other ncRNA and repeats. The known miRNA and novel miRNA were predicted by miRbase and Human Genome (GRCh38) using the remaining reads. The mapping results were analyzed to acquire the read count for each miRNA, and the transcripts per million (TPM) value was calculated.
2.6 Quantitative real-time PCR (qRT–PCR) analysis
To verify the sequencing data, we selected five miRNAs (hsa-miR-122-5p, hsa-miR-146a-5p, hsa-miR-16-5p, hsa-miR-21-5p, and hsa-miR-223-3p) to perform qRT–PCR analysis between LADA patients and heathy subjects based on the two criteria: (1) the high expression level in exosomes in the present study and significant differential expression between different groups and (2) a previous report of association of plasma levels with LADA [
]. Moreover, the five miRNAs were also analyzed for their potential as biomarkers to distinguish LADA from T2D patients. A PrimeScript™ RT Reagent Kit (Perfect Real Time) (TAKARA, RR037A) was used to reverse-transcribe the RNA to cDNA. A TaqMan® probe was used to detect the target gene expression via real-time qPCR. The sequences of the primers and probes are shown in Supplementary Table 1. A t test was used to compare the expression levels of miRNAs between the two groups. A P value < 0.05 was considered to indicate statistical significance.
2.7 Gene ontology (GO) and kyoto encyclopedia of genes and genomes (KEGG) pathway enrichment analysis
The potential target genes of miRNAs were predicted using the multiMiR R package (version 2.3.0), and the target genes predicted by at least three databases (software) or verified by at least one experiment were selected as the prediction results. Moreover, GO enrichment analysis of the target genes of differentially expressed miRNAs was performed with the topGO R package. KEGG pathway enrichment was performed with the Python program KOBAS [
2.8 Construction of the protein–protein interaction (PPI) network
To better detect the cross-actions among the target genes of differentially expressed miRNAs, three miRNAs (hsa-miR-146a-5p, hsa-miR-21-5p, and hsa-miR-223-3p) were selected to construct a PPI network. The STRING database (v11.5) was used for the analysis, and Cytoscape software (v3.9.1) was used for visualization. The minimum required interaction score was 0.4.
2.9 Statistical analysis
The data are expressed as the mean ± SEM, median (25th–75th percentile) or % (n). SPSS software version 20.0, GraphPad Prism version 7.0 and the R package were used to perform the statistical analyses. GPower software version 3.1 [
] was used to calculate the sample size, and the minimal significance (alpha value) and the study power (1 - beta) were set at 0.05 and 0.80, respectively. According to the results of miRNA sequencing, 1.5 fold change of expression level was assumed. Based on these inputs, the sample size of 15 for each group is sufficient for our study. Continuous variables were compared by one-way ANOVA. Categorical variables were compared using Fisher’s exact test or a χ2 test when appropriate. Nonparametric tests were performed using the Mann–Whitney U test. An unpaired T test was used to compare the expression of miRNAs between patients with islet autoantibody positivity and negativity. Receiver operating characteristic (ROC) curve analysis was used to check the diagnostic accuracy of validated differentially expressed miRNAs, and the Pearson correlation coefficient was used to evaluate the correlation between the differentially expressed miRNAs and clinical characteristics. Adjustment for the difference in BMI was performed with analysis of covariance. A P value<0.05 was considered to indicate statistical significance.
3. Results
3.1 Characterization of exosomes
Exosomes isolated from the plasma of a healthy control subject (age: 51 years old; sex: female) were analyzed by NTA, TEM and WB. TEM showed the characteristic cup-shaped morphology of exosomes (Fig. 1a). NTA revealed a normal size arrangement of exosomes (30–200 nm), and the mean and SD of exosome size was 102.3 ± 49.5 nm (Fig. 1b). The main peak of particle size was 80.3 nm, and the percentage of the main peak was 96.4%. Additionally, the concentration of exosomes was 3.9*1012 particles/ml in this sample. WB confirmed that the plasma-derived exosomes expressed the characteristic exosomal markers Tsg101, HSP70 and CD63, whereas they did not contain calnexin, the negative exosomal marker, on their surface (Fig. 1c). These results confirmed the presence and good preparation of the plasma-derived exosomes. In addition, to identify the potential difference in exosomes from different groups, TEM, BCA assays and nanoflow cytometry, a high-sensitivity detection technique for exosomes [
Measuring particle concentration of multimodal synthetic reference materials and extracellular vesicles with orthogonal techniques: Who is up to the challenge?.
], were performed in 5 LADA patients, 4 T2D patients and 5 control subjects. As shown in Supplementary Fig. 1a, the exosomes from the three groups of subjects all showed cup-shaped morphology, which indicates that there was no difference in the shape of exosomes between groups. Similarly, there were no statistically significant differences in the size, quantity and protein concentration of exosomes between groups (control vs. LADA and LADA vs. T2D) (Supplementary Fig. 1b-d).
Fig. 1Identification of plasma-derived exosomes. The plasma-derived exosomes were analyzed by transmission electron microscopy (TEM) images (a). TEM showed the characteristic cup-shaped morphology of exosomes. The size distribution of exosomes was analyzed by nanoparticle tracking analysis (NTA) (b). NTA revealed a normal size arrangement of exosomes (30–200 nm), and the mean and SD of exosome size was 102.3 ± 49.5 nm. The main peak of particle size was 80.3 nm, and the percentage of the main peak was 96.4%. Additionally, the concentration of exosomes was 3.9*1012 particles/ml in this sample. The plasma-derived exosomes expressed Tsg101, HSP70 and CD63 but did not contain calnexin (c).
The procedures for miRNA sequencing and functional analysis are shown in Supplementary Fig. 2. A total of 1091 miRNAs, including 1060 known miRNAs and 31 novel miRNAs, were identified by examining 10 plasma-derived exosome samples from 5 LADA patients and 5 matched healthy controls. To screen potential differentially expressed miRNAs, the selection criteria were as follows: (1) |log2(FC)|≥0.58 and (2) P value ≤ 0.05. As shown in the hierarchical clustering analysis (Supplementary Fig. 3a) and volcano diagram (Supplementary Fig. 3b), 75 miRNAs were differentially expressed between the two groups, of which 51 were upregulated and 24 were downregulated in LADA patients (Supplementary Table 2). The results showed that miRNAs in plasma-derived exosomes were differentially expressed between LADA and control subjects.
3.3 Verification and functional analysis of differentially expressed exosomal miRNAs
Five miRNAs, including 2 downregulated differentially expressed miRNAs (hsa-miR-122-5p and hsa-miR-16-5p) and 3 upregulated differentially expressed miRNAs (hsa-miR-146a-5p, hsa-miR-21-5p, and hsa-miR-223-3p), were selected to verify the sequencing data in an independent batch of samples. Hsa-miR-146a-5p (P = 9.2e-05), hsa-miR-21-5p (P = 0.0026) and hsa-miR-223-3p (P = 0.034) were significantly upregulated in LADA patients compared to control subjects, which was in accordance with the sequence data (Fig. 2a-e). Therefore, our qRT–PCR results verified the reliability of the sequence data. We further assessed the diagnostic efficiency of the three miRNAs. The areas under the ROC curves (AUCs) between LADA and control subjects were 0.884, 0.836 and 0.769 for hsa-miR-146a-5p, hsa-miR-21-5p and hsa-miR-223-3p, respectively, as shown in Fig. 3c. Combining two miRNAs yielded AUCs between 0.796 and 0.942, and the three-miRNA panel showed an AUC of 0.956 (Fig. 3d).
Fig. 2The results of quantitative real-time PCR (qRT–PCR) analysis of five differentially expressed miRNAs between LADA patients and control subjects (LADA N = 15; control N = 15) (a-e) and between LADA and T2D patients (LADA N = 15; T2D N = 15) (f-j).
Fig. 3Area under the receiver operating characteristic (ROC) curves between LADA patients and T2D patients (a, b) and LADA patients and control subjects (c, d).
GO enrichment analysis and KEGG pathway analysis were conducted to explore the function of differentially expressed miRNAs between LADA patients and healthy subjects. As shown in Supplementary Fig. 4, the significantly enriched GO terms of the differentially expressed miRNAs were classified into three categories: biological process (BP), cellular component (CC), and molecular function (MF). Cellular component organization or biogenesis, nucleoplasm and cell adhesion molecule exhibited the highest enrichment factors for the BP (Supplementary Fig. 4a), CC (Supplementary Fig. 4b) and MF categories (Supplementary Fig. 4c), respectively. In KEGG analysis, the ErbB signaling pathway showed the greatest enrichment factor, followed by small cell lung cancer and the TNF signaling pathway (Supplementary Fig. 4d).
3.4 Potential biomarkers for distinguishing LADA from T2D
To screen out the potential biomarkers that could distinguish LADA from T2D patients, we performed qRT–PCR in a cohort including 15 LADA patients and 15 T2D patients. Hsa-miR-146a-5p (P = 1.3e-05), hsa-miR-21-5p (P = 0.0014) and hsa-miR-223-3p (P = 0.0051) were significantly upregulated in LADA patients compared to T2D patients (Fig. 2f-j). Moreover, given the differences in BMI between LADA patients and T2D patients, the BMI was adjusted. The results showed that hsa-miR-146a-5p (P < 0.001) and hsa-miR-21-5p (P < 0.001) were still significantly upregulated in LADA patients compared to T2D patients after BMI adjustment, which indicated that the differential expression of the two miRNAs between LADA and T2D could not be ascribed to BMI. Notably, the expression of hsa-miR-122-5p showed no difference between LADA and T2D patients when we excluded the outlier. Therefore, we did not examine it further in the following study. There was no difference in these selected exosomal miRNAs between T2D and control subjects (Supplementary Fig. 5). In addition, the AUCs between LADA and T2D patients were 0.978, 0.96 and 0.809 for hsa-miR-146a-5p, hsa-miR-21-5p and hsa-miR-223-3p, respectively, as shown in Fig. 3a. The AUCs of combining the two miRNAs hsa-mir-21-5p and hsa-miR-223-3p, hsa-mir-146a-5p and hsa-miR-223-3p, and hsa-mir-21-5p and hsa-miR-146a-5p were 0.964, 0.978, and 1, respectively. The three-miRNA panel (hsa-miR-146a-5p, hsa-miR-21-5p and hsa-miR-223-3p) showed an AUC of 1 (Fig. 3b). We further conducted Pearson correlation coefficient analysis to assess the associations between the differentially expressed miRNAs and clinical features. As shown in Supplementary Fig. 6 and Supplementary Fig. 7, hsa-miR-146a-5p and hsa-miR-223-3p were both positively correlated with postprandial blood glucose (R = 0.917 and R = 0.898, respectively) and negatively correlated with triglycerides (R = -0.418 and R = -0.352, respectively). The expression of hsa-miR-146a-5p was significantly upregulated in patients with GADA, IA-2A and ZnT8A positivity (P < 0.05). The expression of hsa-miR-223-3p was only significantly upregulated in patients with GADA and IA-2A positivity (P < 0.05).
To further detect the potential biological function of hsa-miR-146a-5p, hsa-miR-21-5p and hsa-miR-223-3p in plasma-derived exosomes, we performed GO enrichment analysis and KEGG pathway analysis. As shown in Supplementary Fig. 8, the terms cellular response to growth factor stimulus, nucleoplasm and cell adhesion molecule binding exhibited the highest enrichment scores for BP, CC and MF, respectively. In addition, we generated a KEGG pathway enrichment scatter diagram to show the reliability and significance of different pathways (Supplementary Fig. 8d). The colorectal cancer pathway showed the greatest enrichment factor, followed by pancreatic cancer and the p53 signaling pathway. Moreover, to better understand the cross-action between the target genes of hsa-miR-146a-5p, hsa-miR-21-5p and hsa-miR-223-3p, a PPI network was constructed according to the susceptibility genes of autoimmune diabetes mellitus [
] that included 24 nodes and 67 edges (Supplementary Fig. 9). For example, PTPN20A was related to INSRR via the PTPN20A-PTPN2-INSR-INSRR pathway.
4. Discussion
At present, LADA is defined as a slowly progressive autoimmune disease with the presence of diabetes-associated autoantibodies that distinguish LADA from T2D. However, due to the initial noninsulin dependence and the fact that not every LADA patient has islet autoantibodies measured at the onset of diabetes mellitus, LADA patients are often misdiagnosed with T2D and treated with oral hypoglycemic drugs, which results in accelerated failure of beta-cell function. As a significant form of diabetes, LADA should be identified, understood and managed appropriately to delay or even reverse the progression of the disease. Therefore, identifying a novel and specific biomarker for LADA patients is imperative.
Large amounts of evidence in recent years have suggested a crucial role of exosomes in other types of diabetes. Regarding T1D, researchers found that exosomes released by islet-derived mesenchymal stem cells may function as autoimmune triggers in nonobese diabetic mice [
Extracellular vesicles in immune system regulation and type 1 diabetes: cell-to-cell communication mediators, disease biomarkers, and promising therapeutic tools.
]. For example, lymphocyte-derived exosomal miRNAs can be transferred in active form to beta cells to promote autoimmune attack and recipient beta-cell apoptosis [
]. Insulin resistance is an important characteristic of T2D. Plasma-derived exosomal miR-20b-5p is significantly elevated in T2D patients and can impair insulin-stimulated glycogen accumulation in human skeletal muscle cells [
]. Moreover, placental exosomes from gestational diabetes mellitus pregnancies reduce insulin-stimulated migration and glucose uptake in primary skeletal muscle cells obtained from patients with normal insulin sensitivity [
]. Another study has found that natural killer cell-derived exosomal miR-1249-3p can decrease cellular insulin sensitivity and relieve inflammation in a T2D mouse model [
]. Taken together, these data suggest a crucial role of exosomes and exosomal miRNAs in the onset and development of diabetes. However, to the best of our knowledge, there have been no reports on the function of exosomes in LADA. In this study, we detected the exosomal miRNA profiles of LADA patients and identified plasma-derived exosomal miRNA biomarkers to discriminate LADA patients from T2D patients in a minimally invasive manner for the first time.
First, we found 75 differentially expressed miRNAs in the plasma-derived exosomes between the LADA and healthy subjects, including 51 upregulated and 24 downregulated miRNAs. The differential expression of exosomal miRNAs between LADA patients and healthy control subjects indicated that these miRNAs may be involved in diabetic progression. Moreover, qRT–PCR showed that hsa-miR-146a-5p, hsa-miR-21-5p and hsa-miR-223-3p were significantly upregulated in LADA patients compared with healthy controls, which further validated the accuracy and reliability of the sequence data. Second, to identify the potential functions of the 75 differentially expressed miRNAs, we performed GO enrichment and KEGG pathway analyses. Some identified biological pathways have been found to be involved in diabetes. For example, the ErbB signaling pathway, which was the most enriched, is related to diabetes mellitus. The expression level of ErbB2 has been found to be negatively related to hemoglobin A1c and diabetes mellitus [
]. In addition, tumor necrosis factor alpha, a proinflammatory cytokine, was found to induce beta-cell death by receptor interacting protein kinase 1 (RIPK1) and receptor interacting protein kinase 3 (RIPK3) in INS-1 beta cells and mouse islet cells [
]. These results implied that these differentially expressed miRNAs are involved in diabetes mellitus, although more experiments are warranted.
To validate the changes and further identify biomarkers that could distinguish LADA from T2D at the level of plasma-derived miRNA, five miRNAs (hsa-miR-122-5p, hsa-miR-146a-5p, hsa-miR-16-5p, hsa-miR-21-5p, and hsa-miR-223-3p) were selected for analysis by qRT–PCR. The results showed that the expression levels of hsa-miR-146a-5p, hsa-miR-21-5p and hsa-miR-223-3p were not only significantly different but also exhibited a high AUC between the LADA group and the T2D group. But hsa-miR-146a-5p and hsa-miR-21-5p were still significant different after the BMI adjustment between the two groups. The disappearance of differential expression for hsa-miR-223-3p may be explained by that adipose tissues were also an important source for circulating exosomes [
]. The AUC of combing three miRNAs hsa-miR-146a-5p, hsa-miR-21-5p and hsa-miR-223-3p was 1 and the AUC of combing two miRNAs hsa-miR-146a-5p and hsa-miR-21-5p was still 1. Additionally, the expression level of hsa-miR-146a-5p was significantly elevated in patients with islet autoantibody positivity and the expression level of hsa-miR-223-3p was only significantly elevated in patients with GADA and IA-2A positivity. Interestingly, in previous studies, miR-146a-5p was found to be a plasma and serum biomarker for T2D, and miR-223-3p was found to be a potential tissue biomarker for T2D [
]. In addition, previous research has found that cellular miRNAs can be selectively sorted to exosomes by certain mechanisms, including RNA-binding proteins and membranous proteins involved in exosome biogenesis [
], indicating that plasma-derived exosomal miRNAs and plasma miRNAs are two completely different kinds of biomarkers. Moreover, plasma-derived exosomal miRNAs were found to show higher diagnostic value than miRNAs directly derived from plasma [
]. In addition, there are some other advantages to using plasma-derived exosomes as biomarkers, including (1) the relationship between exosomes and/or exosomal miRNAs and disease, (2) the stability over time of plasma-derived exosomal miRNAs [
]. Therefore, the plasma-derived exosomal hsa-miR-146a-5p, hsa-miR-223-3p and hsa-21-5p identified in our study may be autoimmune indicators and promising biomarkers to discriminate LADA from T2D patients.
To further determine the biological roles of the identified miRNAs, additional GO enrichment and KEGG pathway analyses were performed. The cellular response to growth factor stimulus that exhibited the highest enrichment score was found to be related to beta-cell function. For example, a defect in vascular endothelial growth factor (VEGF)-A in beta cells results in abnormal insulin secretion [
]. Beta-cell loss or dysfunction plays a role in almost all types of diabetes. Emerging evidence suggests that the transforming growth factor beta (TGF-beta) signaling pathway participates in the development, proliferation, apoptosis, dedifferentiation, and function of islet beta cells [
]. In addition, Smad3 is considered an important transcription factor of TGF-beta signaling and has been shown to promote diabetes by suppressing beta-cell proliferation. Diabetic mice transplanted with Smad3 knockout islets show lower blood glucose and HbA1c levels and better protection from diabetic kidney injury than mice transplanted with Smad3 wild-type islets [
Smad3 deficiency improves islet-based therapy for diabetes and diabetic kidney injury by promoting β cell proliferation via the E2F3-dependent mechanism.
]. Moreover, TGF-beta/Smad3 signaling can promote apoptosis and loss of beta-cell mass while leading to beta-cell dysfunction and glucose intolerance [
]. Smad3 influences not only beta-cell proliferation but also insulin synthesis and secretion. TGF-beta/Smad3 signaling has been found to be an important regulator of insulin gene transcription [
]. In summary, these data suggest that the TGF-beta signaling pathway may contribute to the development of LADA and provide a novel target point for the treatment of LADA.
The pancreatic cancer pathway was the main enriched KEGG pathway in the current study. Diabetes mellitus, or even fasting blood glucose that does not meet the diagnostic criterion of diabetes mellitus, is associated with an increased risk of pancreatic cancer [
]. Moreover, CD4+ and CD8+ effector T cells (Teffs) have been found to play a role in autoimmune diabetes mellitus. p53 is involved in the simplified DNA damage response to regulate the apoptosis of Teffs. To better distinguish the connection and identify hub genes among the target genes of hsa-miR-146a-5p, hsa-miR-21-5p and hsa-miR-223-3p, a PPI network was constructed. Twenty-four genes were screened out among approximately sixty susceptible genes of autoimmune diabetes mellitus [
]. One example was the human protein tyrosine phosphatase nonreceptor 22 (PTPN22) gene, which has been found to be significantly associated with the onset of autoimmune diabetes in different populations [
]. The results further indicated that the three miRNAs were related to autoimmune reactions and had potential as autoimmune biomarkers from the genetic background perspective. In brief, the three miRNAs hsa-miR-146a-5p, hsa-miR-21-5p, and hsa-miR-223-3p are associated with the occurrence and development of diabetes mellitus.
Current research detecting the role of exosomal miRNAs in diabetes, especially with regard to LADA, is still in its infancy. Notably, there were some limitations in this study. First, it is certainly a small number of respondents. Second, we evaluated only five differentially expressed miRNAs by qRT–PCR. There are still other miRNAs that require validation in the future. Third, as mentioned, the potential functions of the exosomal miRNAs were only predicted by bioinformatics analysis; more in vivo and in vitro experiments need to be conducted to enable more definitive conclusions.
Taken together, our findings reveal the profiles of plasma-derived exosomal miRNAs in LADA patients and new promising biomarkers for distinguishing LADA from T2D for the first time. These findings will help clinicians discriminate LADA patients from T2D patients early by detecting the expression levels of miRNAs from plasma-derived exosomes. In addition, they provide new targets for molecular drugs with which to treat LADA patients early in order to protect beta cell function and delay or even avoid complications given the emerging roles of engineered exosomes for drug delivery. Moreover, the aberrant characteristics of exosomal miRNAs from LADA plasma and the bioinformatics analysis results may help improve the understanding of the underlying molecular mechanisms of exosomal miRNAs involved in the onset and development of LADA.
5. Submission declaration and verification
The manuscript has not been published previously and is not under consideration for publication elsewhere. Its publication is approved by all authors and tacitly or explicitly by the responsible authorities where the work was carried out. If accepted, it will not be published elsewhere in the same form, in English or in any other language, including electronically, without the written consent of the copyright holder.
Funding
This study was supported by the National Natural Science Foundation of China (82070812, 81820108007, 81800745) and the Science and Technology Innovation Program of Hunan Province (2020RC4044).
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 material
The following are the Supplementary data to this article:
Measuring particle concentration of multimodal synthetic reference materials and extracellular vesicles with orthogonal techniques: Who is up to the challenge?.
Extracellular vesicles in immune system regulation and type 1 diabetes: cell-to-cell communication mediators, disease biomarkers, and promising therapeutic tools.
Smad3 deficiency improves islet-based therapy for diabetes and diabetic kidney injury by promoting β cell proliferation via the E2F3-dependent mechanism.
00045-1&id=gr1.jpg){kind=link}
00045-1&id=gr2.jpg){kind=link}
00045-1&id=gr3.jpg){kind=link}
00045-1&id=fx1.jpg){kind=link}
00045-1&id=fx2.jpg){kind=link}
00045-1&id=fx3.jpg){kind=link}
00045-1&id=fx4.jpg){kind=link}
00045-1&id=fx5.jpg){kind=link}
00045-1&id=fx6.jpg){kind=link}
00045-1&id=fx7.jpg){kind=link}
00045-1&id=fx8.jpg){kind=link}
00045-1&id=fx9.jpg){kind=link}