Identification of the differential expression of serum microRNA in type 2 diabetes

Abstract

The identification of disease-specific alterations in miRNA expression and the ability to detect miRNAs in serum furnish the basis for identified potential research value. This study was aimed to characterize the expression of miRNAs in the serum samples from people with type 2 diabetes mellitus (T2DM) and healthy individuals in order to detect the differential expression of miRNAs in T2DM. In total, 582 participants were recruited. Microarray-based miRNA expression profiles were screened in pooled serum samples from two groups (T2DM and healthy control). The candidates’ miRNAs were validated by reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR). Five significantly different serum miRNAs were identified in T2DM patients (hsa-miR-320d, hsa-miR-4534, hsa-miR-3960, hsa-miR-451a, and hsa-miR-572) compared to those in the serum of healthy controls. This study provided evidence that serum miRNAs had differential expressions between healthy controls and T2DM patients. These five differential expression miRNAs might be of help for subsequent study in T2DM.

Graphical abstract

This research found five significant disease-special serum miRNAs (hsa-miR-320d, hsa-miR-4534, hsa-miR-3960, hsa-miR-451a, hsa-miR-572) in T2DM compared with healthy control.

Key words:

• T2DM
• microRNA
• serum
• microarray
• qRT-PCR

Diabetes mellitus (DM) has affected over 350 million people worldwide.¹⁾ Diabetes is a complex metabolic disorder that is the result of a combination of environmental and genetic factors that induce individuals fail to maintain blood glucose homeostasis.²⁾ T2DM, which accounts for almost 90% of all cases of DM, is characterized by peripheral insulin resistance and β-cell dysfunction. Pancreatic β-cells are required to produce much more insulin to satisfy the increased demand when insulin resistance occurs.³⁾ As one’s age increases, β-cells are damaged by decreased functions, which finally leads to T2DM.⁴⁾ Multiple chronic macrovascular and microvascular complications (e.g. diabetic nephropathy, cardiomyopathy, neuropathy, and retinopathy) caused by T2DM give rise to considerable global morbidity and mortality rates as a result of strokes, myocardial infractions, blindness, and foot ulcers.^5–8) Serious complications can devastate one’s quality of life and life expectancy.^6,9) Diabetes has become a major socioeconomic health problem, but there is still no efficient prevention measures or treatments. It is quite essential to explore a novel research direction for T2DM.

MiRNAs were initially identified in early 1990s. They are endogenous, single-stranded RNA molecules consisting of 21–23 nucleotides. It has been suggested that miRNAs regulate gene expression at the post-transcriptional level by binding to the three untranslated regions of the target mRNAs.^10–12) Mitchell P. S. and others found that miRNAs detected in serum/plasma were present in a remarkably stable form that was protected from endogenous RNase activity.^13–20) Recent studies have suggested that miRNA expressions were potentially disease specific and could be used to identify various types of diseases, including breast cancer, colorectal cancer, large β-cell lymphoma, tissue injury, and non-small cell lung cancer.^{14, 21–25)} These characteristics of circulating miRNAs have great promise as novel, non-invasive biomarkers for diseases.^{6, 20, 26, 27)}

In order to study the relationship between serum miRNAs and T2DM, the expression levels of miRNAs were detected in the serum of T2DM patients and those in the healthy control group. This research investigated 1887 serum microRNA expression profiles with independent validation in a large cohort of 582 participants. The main purpose of the study was to identify the differential expressions of serum miRNAs in T2DM patients.

Materials and methods

Study design and group population

This project was approved by the First Affiliated Hospital of Wenzhou Medical University, and written informed consent was obtained from all volunteers. All serum samples used for the present study were obtained from the First Affiliated Hospital of Wenzhou Medical University between 2012 and 2014. This study was approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University. All of the T2DM patients conformed to the T2DM inclusion and exclusion standard. The T2DM inclusion standards were as follows: symptoms of diabetes and blood glucose ≥11.1 mmol/L (200 mg/dL); fasting blood glucose (FBG) ≥7 mmol/L (126 mg/dL); and glucose tolerance testing 2-h blood glucose ≥11.1 mmol/L (200 mg/dL). The patients included in the T2DM group had to comply with one of the above standards on the day prior to review testing. The T2DM exclusion standards were as follows: patients who had endocrine disease; patients who had abnormal endocrine system-related indicators; patients who had tumor or abnormal tumor markers. All of the people included in the healthy controls group had clinical indicators that were within the medical reference value range and did not have diagnosed diseases, family histories of hereditary diseases, T2DM, or other malignant diseases. A multi-stage, case-control study was designed to identify different expressions of serum miRNAs in order to distinguish T2DM patients from those in the healthy control group (Fig. 1). In the screening stage, a total of 96 individuals were categorized accordingly into two groups, T2DM (n = 40) and healthy controls (n = 56), without complications. Serum samples were screened with a microarray platform (Table 1). There was no significant difference in the distribution of age and gender in the two groups (healthy and T2DM). The two groups were mixed, respectively, for microarray. Subsequently, individual RT-qPCR was performed in the selection and validation stages to further screen the candidate miRNAs (Table 2). There was no significant difference in the distribution of age and gender between the selection and validation stages for the healthy control group and the T2DM group. During the selection stage, 263 samples were tested to selected candidate miRNAs, 74 from T2DM patients and 189 from the healthy control group. In the last stage, 223 more samples were added in order to validate the selected candidate miRNAs, 69 from T2DM patients and 154 from those in the healthy control group.

Fig. 1. Systematic overview of the workflow.

Table 1. Characteristics of study participants in the screening stages.

Characteristic	Control	T2DM	P value
Sex			>0.05
Male	32	27
Female	24	13
Age, years			>0.05
Mean	63	58.7
SD	9.49	13.5

Note: Sex of T2DM compared with control group analyzed by χ² test, and ages analyzed by t-test.

Abbreviations: T2DM, type 2 diabetes mellitus; SD, standard deviation.

Table 2. Characteristics of study participants in the selection and validation stages.

	Selection (n = 263)		Validation (n = 223)		P value
Variable	No	%	No	%
Healthy count	74		69
Age, years					>0.05
Mean	53.6		54.2
SD	10.3		10.7
Sex					>0.05
Male	36	49	30	43
Female	38	51	39	57
T2DM	189		154
Age, years					>0.05
Mean	60.8		61.1
SD	12		12.4
Sex					>0.05
Male	98	52	70	45
Female	91	48	84	55

Note: Sex of T2DM compared with control group analyzed by χ² test, and ages analyzed by t-test.

Abbreviations: T2DM, type 2 diabetes mellitus; SD, standard deviation.

MiRNA isolation

For the study of miRNAs, 10-mL serum tubes with clot activators were collected. After allowing the content of the serum tubes to clot at room temperature for 1 h, the serum was prepared by centrifugation at 1000 × g (2300 rpm) for 15 min at 4 °C. The serum supernatant was slowly removed, leaving 0.5 cm remaining to avoid disturbing the serum–clot interface, and stored at −80 °C. Serum was thawed on ice before use.

The MiRNA extracted from serum using a miRNeasy kit (Qiagen miRNeasy Mini Kit) according to the manufacturer’s recommendations. The extracted miRNAs were stored at −80 °C for microRNA microarrays and qRT-PCR.

MicroRNA microarray

The miRNA samples were spiked by the MicroRNA Spike-In Kit (Agilent Technologies) to help distinguish biological data from processing issues of the labeling and hybridization. 100 ng of miRNA per sample was dephosphorylated by alkaline calf intestine phosphatase incubating the reaction at 37 °C. T4 RNA ligase was used to label the dephosphorylated RNA incorporated with cyanine 3-cytidine biphosphate (miRNA Complete Labeling and Hyb Kit; Agilent Technologies). The labeled miRNA samples were hybridized to human miRNA microarrays (Human miRNA 8 × 60 K V18.0; Agilent Technologies) at 55 °C for at least 20 h. After washing the microarray slides (Gene Expression Wash Buffers; Agilent Technologies), they were dried. Scanning was performed with the Agilent G2565CA Microarray Scanner, and images were extracted using the Feature Extraction 10.7.3.1 Software (Agilent Technologies). All of the steps described were performed according to the manufacturer’s instructions.

qRT-PCR assay

For the testing of candidate miRNAs acquired on microarrays, qRT-PCR was performed using SYBR-Green assays (Applied Biosystems) in a final reaction volume of 20 μl (All-in-One miRNA Q-PCR Detection Kit, GeneCopoeia). The reaction mix was transferred into 96-well plates and run on the qRT-PCR instrument (Applied Biosystems). The amplification curves were analyzed using version 2.1 of the StepOne Software (Applied Biosystems) for both the determination of cycle threshold (Ct) value and for melting curves analysis.

Statistical analyses

For quality control, statistical analysis, miRNA annotation, and visualization purposes, we used Feature Extraction 10.7.3.1 Software and GeneSpring GX11.5.1 (both Agilent Technologies). Cluster analysis was performed using the unweighted pair group method with arithmetic means based on Euclidean distance with hierarchical clustering applied to the normalized data. Absolute fold changes were calculated to identify different expressions between the T2DM and control groups. A T-test was performed between the T2DM and healthy control groups.

Version 2.1 of the StepOne Software (Applied Biosystems) was utilized for instrument control, data acquisition, and raw data analysis. In order to calculate the relative expression levels of the target miRNAs, we performed ΔCt normalization, which was calculated using the following mathematical formula: ΔCt = Ct_sample − Ct_endogenous. Unpaired T-tests (with significance levels set at p < 0.05) were calculated between the T2DM and healthy control groups.

Statistical analysis was performed using SPSS 17.0, and presented with GraphPad Prism 5.0 software. Results were considered statistically significant when p < 0.05.

Results

Differential microRNA expression based on microarray data

In order to gain an expression profile of serum miRNAs that is specific to T2DM, the microarray was used to identify the differentially expressed miRNAs in those in the healthy control group (n = 56) and those in the T2DM group (n = 40) in the initial screening stage (Supplemental Fig. 1). We applied a stringent filtering approach that compared T2DM with healthy control (absolute fold change > 2.0). Among 1887 miRNAs analyzed, we identified 54 up-regulated and 4 down-regulated miRNAs (Supplemental Table 1). Through the results of the microarray and the relative reports^{26, 28)} observably differentially expressed microRNAs were selected and identified as candidates for further qRT-PCR. They were the following: hsa-miR-4454, hsa-miR-4497, hsa-miR-4530, hsa-miR-4534, hsa-miR-4713–3p, hsa-miR-4716-3p, hsa-miR-762, hsa-miR-1183, hsa-miR-1305, hsa-miR-134, hsa-miR-3198, hsa-miR-320c, hsa-miR-320d, hsa-miR-3676-5p, hsa-miR-4443, hsa-miR-572, hsa-miR-575, hsa-let-7b-3p, hsa-miR-21-5p, hsa-miR-223-3p, hsa-miR-3960, hsa-miR-4281, hsa-miR-4459, hsa-miR-4514, hsa-miR-5581-5p, and hsa-miR-451a.

qRT-PCR validation

In order to characterize the differential expression of serum miRNAs between those in the T2DM and control groups, the reference gene miR-1228 and the 26 differently expressed miRNAs were detected. The differential expression of the target miRNAs between the T2DM and control groups was summarized (Supplemental Table 2). The expression levels of nine miRNAs (hsa-miR-1183, hsa-miR-320c, hsa-miR-320d, hsa-miR-4530, hsa-miR-4534, hsa-miR-3960, hsa-miR-451a, hsa-miR-4443, and hsa-miR-572) were different in T2DM group compared to the healthy control group.

The assays were further applied on 223 samples for nine candidate miRNAs by qRT-PCR. Finally, five significantly changed serum miRNAs were identified (hsa-miR-320d, hsa-miR-4534, hsa-miR-3960, hsa-miR-451a, and hsa-miR-572) in T2DM patients compared to those from the healthy control group (Fig. 2).

Fig. 2. Relative expression levels of nine candidate miRNAs.

Notes: (A–I) Different expressions of miR-1183, miR-320c, miR-320d, miR-4530, miR-4534, and miR-3960, miR-451a, miR-4443, miR-572 were confirmed by RT-PCR in groups. Data are mean ± SEM. * indicated p < 0.05.

Discussion

Diabetes is a complex metabolic disorder disease. This complex disease is based on the consequence of interaction between several factors, including genetics, lifestyle, age, and ethnic and environmental factors. T2DM is characterized by insulin resistance. Enhanced insulin resistance or reduced compensatory insulin secretion bring about decreased glucose tolerance and gradually progress to diabetes. Then, blood glucose levels began to rise. Hie et al. investigated the effects of insulin-dependent DM on osteoblastogenesis and osteoclastogenesis used diabetic rats two weeks after the streptozotocin injection. They found that the level of the receptor activator of NF-κB (RANK) protein increased and the runt-related transcription factor 2 (Runx2) protein decreased in diabetic rats, and the change could be reversed by insulin treatment.²⁹⁾ Subsequently, Hu et al. found miR-3960/miR-2861 promotes osteoblast differentiation by targeting histone deacetylase 5 and increases Runx2 protein production.³⁰⁾ In this study, miR-3960 was found to have a differential expression in the serum of T2DM patients. Based on these studies, it can be seen that miR-3960 might play an important role in diabetes.

This study demonstrated four other differential expressions of serum miRNAs in T2DM patients (miR-451a, miR-320d, miR-4534, and miR-572). Murata K. et al. compared the expression of miR-451 in neutrophils from patients with rheumatoid arthritis (RA) and healthy controls to identify the role of miR-451 in autoimmune arthritis. The research found that miR-451 suppresses neutrophil chemotaxis via p38 MAPK and was a potential target in the treatment of RA.³¹⁾ Map kinase kinase 3 (MKK3)–p38 MAPK signaling was a pathogenic factor in the progression of diabetic nephropathy in db/db mice.³²⁾ Mir-451 might have potential research value in the study of diabetic nephropathy. Zhang et al. found that miR-320d was under-expressed in their research, which might be involved in the regulation of stem cell differentiation,³³⁾ but miR-320d was never reported in T2DM. Through deep sequencing, Jima et al. found that miR-4534 has differential expression in normal and malignant human β-cells, but the function of miR-4534 was not detected in this study.²⁸⁾ MiR-572 was reported in a colorectal microRNAome study, but the function of miR-572 was not clear in this research.³⁴⁾

In conclusion, we identified serum miRNAs that differentiate T2DM patients from healthy individuals with a high degree of accuracy in a large number of participants. Our study demonstrates that serum miRNAs have differential expressions in T2DM patients, and that serum miRNAs dysregulation might be associated with the pathogenesis of diabetes.

Author contributions

Conceived and designed the experiments: Lianjin Zhong and Jianxin Lu.

Analyzed the data: Linchao Ding, Ruihao Wu, and Tao Zhang. Contributed reagents/materials/analysis tools: Li Jing

Performed the experiments and wrote the manuscript: Linchao Ding and Dongdong Ai.

Funding

This work was supported by the Program for Zhejiang Leading Team of S&T Innovation [grant number 2010R50048].

Supplemental material

The supplemental material for this paper is available at http://dx.doi.org/10.1080/09168451.2015.1107460.

Disclosure statement

No potential conflict of interest was reported by the authors.