SEARCH

SEARCH BY CITATION

Keywords:

  • mandibular prognathism;
  • microRNA;
  • serum;
  • marker

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Subjects and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. Author contributions
  10. References

Objectives

The aim of the study was to determine whether the expression levels of specific circulating serum microRNAs (miRNAs) are associated with mandibular prognathism (MP).

Methods

Sixty subjects in the early permanent dentition stage and 23 in the mixed dentition stage with MP were identified. Sixty-eight normal control subjects in the early permanent dentition stage and 24 in the mixed dentition stage were recruited for comparison. According to the microarray-based expression profiling, four serum miRNAs (let-7i-3p, miR-595, miR-16-2-3p, and miR-367-5p) were validated.

Results

In the MP groups, let-7i-3p was significantly over-expressed in subjects in the early permanent (< 0.0005) and mixed (< 0.001) dentitions, and miR-595 was significantly under-expressed (< 0.004) in subjects in the early permanent (< 0.004) and mixed (< 0.0005) dentitions, compared with normal control groups. Multiple logistic regression analysis and receiver operating characteristic curve analysis revealed that let-7i-3p and miR-595 were able to significantly discriminate MP subjects from normal controls.

Conclusion

Let-7i-3p and miR-595 could be potential, non-invasive biomarkers for the accurate early detection and diagnosis of MP, which may result in improved clinical management.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Subjects and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. Author contributions
  10. References

Mandibular prognathism (MP) is a skeletal type of occlusal variation caused by excessive mandibular growth (Proffit and Fields, 2000), producing a concave facial profile and a skeletal Class III malocclusion. The prevalence of MP varies among different ethnic groups, but it is more common in Asian populations and Mongolian regions. The incidence ranges between 4% and 14% among Chinese people (Allwright and Burndred, 1964; Ishii et al, 1987). MP is a functional inconvenience, and the person's appearance may cause severe psychological problems. And treatment of patients with this skeletal pattern remains one of the most challenging problems confronting practicing orthodontists because of the great diversity in anatomic craniofacial structures and the unpredictable growth of the mandible (Dietrich, 1970; Guyer et al, 1986; Williams and Andersen, 1986; Chang et al, 2006). During mandibular growth, early orthopedic and orthodontic treatment responses can vary greatly, and some patients may still need subsequent surgery to correct their prognathic facial profiles. Therefore, novel non-invasive prognostic biomarkers that accurately represent the biological characteristics of MP could improve the early clinical management of affected persons.

Consisting of only 22–25 nucleotides, microRNAs (miRNAs) regulate a variety of biological processes including developmental timing, signal transduction, cell growth, and death (Hwang and Mendell, 2006). Recent reports demonstrate the roles of serum miRNA expressions as potential diagnostic or prognostic markers for lung (Chen et al, 2008; Hu et al, 2010), prostate (Mitchell et al, 2008), colorectal (Huang et al, 2010b), ovarian (Resnick et al, 2009), breast (Heneghan et al, 2010), and, in particular, oral (Wong et al, 2008; Lin et al, 2010; Liu et al, 2010) cancers. Nevertheless, there is little information available on specific miRNA expression patterns and their roles in MP. It is hypothesized that the expression levels of specific serum miRNAs may be useful to monitor the progression of MP, and therefore, this pilot study explores the feasibility of using serum miRNAs as non-invasive diagnostic and potential prognostic biomarkers in young persons.

Subjects and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Subjects and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. Author contributions
  10. References

Sample recruitments

Based on the clinicopathologic findings from radiographs, facial profiles, and molar occlusions, serum samples were collected from 60 subjects (33 boys and 27 girls) in the early permanent dentition stage aged from 10 to 14 years and 23 subjects (10 boys and 13 girls) in the mixed dentition stage aged from 8 to 10 years. Selection criteria included the following: (i) a prognathic mandible, (ii) a normal-sized maxillary arch, and (iii) no congenital deformities. Normal control subjects were recruited and serum samples obtained from 68 age-matched healthy subjects (36 boys and 32 girls) in the early permanent dentition stage aged from 10 to 14 years and 24 subjects (12 girls and 12 girls) in the mixed dentition stage aged from 8 to 10 years with normal molar occlusions. All subjects were recruited from the Department of Orthodontics, School of Stomatology, Nanjing Medical University, during July 2011 to August 2012. Informed written consents were obtained from all subjects and parents before participation in the study, and the study protocol was approved by the Institutional Review Board of Nanjing Medical University (2011-03-16). Serum was stored at −80°C until miRNA extraction.

miRNA microarray

Serum used for clustering contained 36 samples of several mixtures obtained from subjects in the early permanent dentition. Eighteen MP samples (combined as three experimental chips) and 18 normal occlusion samples (combined as three control chips) of serum were recruited for miRNA microarray analysis as a fast, efficient technology to study, firstly, biological information. Each chip contained six serum samples of either the MP group or the normal group. The consistent relationship between each experimental and control sample was evaluated (Table 1). The labeling and hybridization work were completed by KangChen Bio-tech Inc. (Shanghai, China). To confirm the microarray data and to determine the clinical significance of the differential expression of miRNAs in the subjects with MP, specific miRNA expression levels were then validated among the remaining samples.

Table 1. Correlation coefficients (r) of normal samples (normal1, normal2, and normal3) and mandibular prognathism (MP) samples (MP1, MP2, and MP3)
 normal1normal2normal3MP1MP2MP3
normal11.000.970.930.930.950.93
normal20.971.000.940.930.950.91
normal30.930.941.000.950.950.90
MP10.930.930.951.000.910.96
MP20.950.950.950.911.000.91
MP30.930.910.900.960.911.00

miRNA isolation

Three hundred microlitre serum was obtained from each of the remaining 65 MP samples (42 in the early permanent dentition and 23 in the mixed dentition) and 74 normal samples (50 in the early permanent dentition and 24 in the mixed dentition). miRNA isolation was performed using a NucleoSpin miRNA Plasma Kit (Macherey-Nagel GmbH & Co. KG, Düren, Germany) according to the manufacturer's instructions, and final miRNAs were dissolved in 50 μl RNase-free water. The quantity and quality of samples were evaluated with a ND-1000 NanoDrop spectrophotometer at 260 and 280 nm (Thermo Fisher Scientific Inc., Waltham, MA, USA).

Quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR)

(i) Probes: miRNA-specific (including internal control miRNA) probes were ordered as part of a TaqMan miRNA assay kit (Applied Biosystems, Foster City, CA, USA). (ii) RT-PCR: Total RNA was retrotranscribed using a TaqMan microRNA Reverse Transcription Kit (Applied Biosystems). In brief, each 15-μl RT reaction system consists of 7 μl RT master mix, 3 μl RT probe, and 5 μl RNA sample. The 7 μl RT master mix contained 0.15 μl dNTPs, 1 μl reverse transcriptase, 1.5 μl reverse transcription buffer, 0.19 μl RNase inhibitor, and 4.16 μl nuclease-free water. Reactions were incubated for 30 min at 16°C, 30 min at 42°C, and 5 min at 85°C and then held at 4°C. (iii) qPCR: qPCR was performed in duplicate using a 7900 Fast Real-Time PCR Machine (Applied Biosystems). In brief, each 20-μl PCR reaction system consisted of 1 μl PCR probe, 1.33 μl product from the RT reaction, 10 μl TaqMan Universal PCR Master Mix, and 7.67 μl nuclease-free water. The thermal cycle started with 2 min at 50°C, 10 min at 95°C, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. Cel-miR-238 was selected as the internal control miRNA, because of the absence of homologous sequences in humans (Mitchell et al, 2008; Xiea et al, 2010), to normalize the cDNA levels of different samples. A blank control was run in parallel to determine the absence of contamination within each experiment.

Target gene analysis

To further investigate the possible functions of let-7i-3p (a minor form of let-7i) and miR-595, the list of genes predicted to be targeted by the candidate miRNAs was obtained using TargetScan (www.targetscan.org). The predicted target genes were analyzed for different signaling pathways or functions by the NCBI DAVID server (http://david.abcc.ncifcrf.gov/tools.jsp) (Fu et al, 2011).

Data analysis

Pearson's correlation coefficient (r) was used to analyze the relationships between all chips of the microarray. Expression folds of miRNAs were calculated using the arithmetic formula inline image and tested statistically with Student's paired t-test. Each miRNA's expression level between male and female subjects was also tested with Student's paired t-test. For differentially expressed serum miRNAs validated by qRT-PCR, receiver operating characteristic (ROC) curves were generated. Area under curve (AUC) values and 95% confidence intervals (CI) were calculated to determine the specificity and sensitivity of their diagnostic efficiency to MP. To increase the diagnostic accuracy of combined changes in serum miRNA levels, multiple logistic regression analysis was carried out according to previously described methods (Redell et al, 2010). Data analysis was performed using the Statistical Package for the Social Sciences (version 17.0; SPSS Inc., Chicago, IL, USA). The probability level for statistical significance was set at α = 0.05.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Subjects and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. Author contributions
  10. References

Identification of differentially expressed miRNAs from miRNA microarray analysis

A custom-made mammalian miRNA microarray was used to evaluate the expression of human serum miRNAs. Hierarchical clustering was used to group samples from the MP group and normal controls (Figure 1). Twenty-eight miRNAs expressed up-regulated and 49 down-regulated with at least a twofold change between the MP and normal groups. Of these, three up-regulated and eight down-regulated miRNAs were identified as being statistically significant (Table 2). After considering the normalized reads of these miRNAs in serum, four miRNAs (let-7i-3p, miR-595, miR-16-2-3p, and miR-367-5p) with relatively more reads were selected to investigate their expression levels in the MP subjects and age-matched normal controls in the early permanent dentition stage. TaqMan miRNA assay kits with miRNA-specific (including internal control miRNA) probes are shown in Table 3.

Table 2. Expression levels and differentially normalized reads of miRNAs by microarray analysis with at least a twofold change between the mandibular prognathism (MP) group and normal group
 miRNAFold changeP-valueReads (normalized)
MP vs normalMP vs normalMPnormal
  1. a

    Chosen miRNAs identified by qRT-PCR.

Up-regulatedmiR-367-5pa2.3560.047207.5074.00
let-7i-3pa2.5650.003286.0097.33
miR-31422.5170.046143.8349.50
Down-regulatedmiR-16-2-3pa0.4960.002225.50396.67
miR-369-3p0.4510.00673.50142.00
miR-595a0.2090.02250.17212.50
miR-502-3p0.2010.01030.17131.33
miR-24-2-5p0.3610.04054.83129.33
miR-18a-3p0.1970.01816.0072.00
miR-500a-3p0.2580.04911.6739.67
miR-450a0.2960.01013.6741.17
Table 3. Probes of detected miRNAs (including internal control miRNA)
miRNA probeTaqMan assay IDmiRNA accession in miRBaseMature miRNA sequence
let-7i-3p002172MIMAT0004585CUGCGCAAGCUACUGCCUUGCU
miR-595001987MIMAT0003263GAAGUGUGCCGUGGUGUGUCU
miR-16-2-3p002171MIMAT0004518CCAAUAUUACUGUGCUGCUUUA
miR-367-5p002121MIMAT0004686ACUGUUGCUAAUAUGCAACUCU
cel-miR-238 (internal control miRNA)000248MI0000313UUUGUACUCCGAUGCCAUUCAGA
image

Figure 1. Part of the hierarchical clustering of differentially expressed miRNAs of the microarray. Three mandibular prognathism (MP) samples (MP1, MP2, and MP3) and three normal samples (normal1, normal2, and normal3) of serum were used for the miRNA microarray. Rows: miRNAs. Columns: experimental and control samples. Green, black, and red indicate the up-regulation, unchanged expression, and down-regulation of expression of the miRNAs, respectively

Download figure to PowerPoint

Differential expression of miRNAs in MP subjects

For the subjects in the early permanent dentition stage, the expression level of let-7i-3p in the MP group was up-regulated 2.346 times higher (< 0.0005), while the level for miR-595 was down-regulated 1.499 times lower (< 0.004) than in the normal group. However, no statistically significant differences were observed between the two groups for miR-16-2-3p (= 0.56) and miR-367-5p (= 0.13) expression levels (Figure 2). No statistically significant differences were found between boys and girls for each miRNA's expression level, either (all < 0.05).

image

Figure 2. Scatter dot plots with means comparing statistically significant relative expression levels between the MP group (n = 42) and normal group (n = 50) in subjects in the early permanent dentition stage calculated using inline image for four selected miRNAs (a), (b), (c), and (d). Lines indicate the mean value with error bar. *< 0.005, ns, not significant. MP, mandibular prognathism

Download figure to PowerPoint

As the best time for early treatment for MP is before the growth peak, it is critical to explore whether the differential expressions of these two miRNAs exist in subjects in the late mixed dentition stage. We further examined the expression levels of let-7i-3p and miR-595 in 23 MP patients and 24 controls aged 8–10 in mixed dentition. The expression level of let-7i-3p in the MP group was up-regulated 4.337 times (< 0.001) and miR-595 was down-regulated 2.283 times (< 0.0005) compared with those in normal group (Figure 3), which were consistent with the results in subjects in the early permanent dentition stage.

image

Figure 3. Scatter plots of relative expression levels of (a) let-7i-3p and (b) miR-595 between the MP group (n = 23) and normal group (n = 24) in subjects in the mixed dentition stage calculated using inline image . Lines indicate the mean value with error bar. *< 0.005. MP, mandibular prognathism

Download figure to PowerPoint

Evaluation of the diagnostic potential of miRNAs for MP

To investigate the characteristics of these miRNAs as potential diagnostic biomarkers for MP, multiple logistic regression analysis and ROC curve analysis were performed on the two miRNAs exhibiting significant differential expression. In subjects in the early permanent dentition stage, ROC curves of combined let-7i-3p and miR-595 reflected strong separation between the MP and normal subjects, with AUC values of 0.896 (95% CI = 0.817–0.976) and 0.923 (95% CI = 0.860–0.986), respectively (Figure 4). At an optimal cutoff point set at 0.65, let-7i-3p and miR-595 yielded sensitivity values of 83.3% and 85.7% and specificity values of 92.9% and 95.2%, respectively.

image

Figure 4. Receiver operating characteristic (ROC) curves of combined let-7i-3p and miR-595 to distinguish the MP patients from the normal controls in subjects in the early permanent dentition stage. MP, mandibular prognathism.

Download figure to PowerPoint

In subjects in the mixed dentition stage, strong separations of expressions of let-7i-3p and miR-595 between the two groups were also found in the ROC curve analysis, with AUC values of 0.960 (95% CI = 0.891–1.000) and 0.847 (95% CI = 0.730–0.964), respectively (Figure 5). At an optimal cutoff point set at 0.60, let-7i-3p and miR-595 yielded sensitivity values of 87.0% and 73.9% and specificity values of 87.0% and 82.6%, respectively.

image

Figure 5. ROC curves of combined let-7i-3p and miR-595 to distinguish the MP group from the normal group in subjects in the mixed dentition stage. MP, mandibular prognathism.

Download figure to PowerPoint

Target prediction

Using the TargetScan algorithm, 1072 target genes for let-7i and 183 for miR-595 were obtained. Gene ontology analysis showed that four genes directly targeted by miR-595 are involved in ossification and bone development, including BCL2, SATB2, CDH11, and SLC26A2. Other genes targeted by let-7i are associated with the regulation of tissue growth (TAF9B, ADRB1, ADRB3, NTN1, and TP53) and apoptosis (E2F2, FASLG, MDM4, STEAP3, AEN, ESPL1, and NTN1).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Subjects and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. Author contributions
  10. References

As Lawrie et al (2008) introduced miRNA as a new class of biomarker for cancers, serum miRNAs have gained increasing attention because of their consistent reproducible levels across individuals of the same species (Chen et al, 2008; Resnick et al, 2009). There is no significant gender difference in the expression levels of circulating miRNAs (Wang et al, 2010; Li et al, 2011), which circulate in a stable, cell-free form in the blood (both plasma and serum). Serum miRNAs exhibit high stability against external impacts such as RNase and DNase enzymatic degradation, multiple freeze–thaw cycles, and intense pH conditions (Gilad et al, 2008; Mitchell et al, 2008). Furthermore, isolated miRNAs are not affected by different storage temperatures, as investigated for storage at −20 and −80°C (Mraz et al, 2009). Easy and non-invasive access is an important criterion for using serum miRNAs as potential biomarkers.

Although most of the published studies using circulating miRNAs as biomarkers have focused on cancers, miRNAs also regulate various osteogenesis events in vertebrates (Li et al, 2008; Inose et al, 2009; Kapinas et al, 2009; Huang et al, 2010a). Osteoblast differentiation is a key step in proper skeletal development and the acquisition of bone mass, and the present study is the first report of a comprehensive interrogation of a skeletal developmental condition using serum miRNAs.

The clinical diagnosis of true skeletal MP, which determines the timing of orthodontic intervention and the treatment decision boundary between orthodontic camouflage and orthognathic surgery, is controversial. Serum protein biomarkers for the diagnosis of malocclusion have been used in other studies (Ghafari et al, 1995). But, as serum miRNAs have more stable characteristics, it is preferable to explore the association between the diagnosis of MP and serum miRNAs. If significant differences in the expression levels of serum miRNAs exist between MP patients and normal controls, then serum miRNAs could be used as potential, non-invasive biomarkers for the accurate early diagnosis and management of MP.

The results of the present pilot cohort study demonstrated that two miRNAs were differentially expressed after validation of TaqMan-based qRT-PCR. Following multiple logistic regression analysis and ROC analysis, let-7i-3p and miR-595 in two young age groups were shown to significantly increase the effectiveness of MP diagnosis with high sensitivity and specificity. In addition, specific serum miRNAs may also indicate the potential severity of MP and possibly the later dominant growth direction of the mandible. From the relative expression levels of serum let-7i-3p and miR-595, it may also be possible to determine those orthodontic patients at high risk of relapse. Further associations of the severity of MP with the relative expression levels of let-7i-3p and miR-595 need to be explored.

The popular hypothesis is that the origin of circulating miRNAs is from tumor cell death and lysis. The present study findings suggest that the increase in let-7i and the decrease in miR-595 in serum might have been caused by changes in leakage or secretion of miRNAs from hyperplastic mandibular skeletal cells. Differential expression of serum miRNAs somewhat indicates the function of these miRNAs in the development of disease (Zhao et al, 2010). The results from target gene prediction found that the two candidate miRNAs were involved in contributing to the overgrowth of tumor cells and possibly also in regulating the overgrowth of mandibular bone cells that causes MP. Researchers would not expect that only one or even a few target proteins of these two miRNAs play a key role in the development of MP. Instead, there should be a complex molecular network involved in MP, with let-7i-3P and miR-595 being potentially therapeutic targets. Through several pathways in the molecular mechanism, it potentially allows new modes of therapy using target gene avenues.

Four genes directly targeted by miR-595 are involved in ossification and bone development. For example, over-expression of BCL2 inhibits osteoblast differentiation and induces osteocyte apoptosis (Moriishi et al, 2011). Other research found that miR-34s inhibit osteoblast proliferation and differentiation by targeting SATB2 (Wei et al, 2012). As MP is caused fundamentally by the over-development of mandibular bone and dysfunction of ossification, it is assumed that miR-595 may inhibit osteoblast proliferation by targeting these related genes. And, therefore, the under-expression of miR-595 may result in over-differentiation and proliferation of osteoblasts in mandibular bone. Genes targeted by let-7i are found associated with the regulation of tissue growth and apoptosis. TP53, a popular gene, is associated with the repair of cell damage or with triggering apoptotic death if the damage is beyond repair (Jolliffe and Derry, 2012), while the dysfunction of programmed cell death can cause the eventual overgrowth of mandibular bone. The above reasoning may explain why the two miRNAs are expressed differentially from the gene level and indicate related genes to be explored by further research.

In summary, the present study strongly suggests that let-7i-3p and miR-595 are potential biomarkers for the accurate early diagnosis and management of MP patients. These two miRNAs may also provide new avenues for a sensitive and reliable clinical prognostic test for MP. Further studies should explore the mechanism of potential target genes, to create possible modes for gene therapy, which may provide a more convenient and effective method for future clinical therapy.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Subjects and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. Author contributions
  10. References

This study was supported by the Natural Science Foundation of Jiangsu Province (BK2010529); National Natural Science Foundation of China (81000457); and a project funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (2011-137).

Author contributions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Subjects and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. Author contributions
  10. References

J Ma and L Wang designed the experiment and performed the data analysis. H Zhang, Y Pan, and J Ni performed the child population recruitment sample. Y Gu, H Wang, and C Zhao performed the sample preparation. Y Zhang and Y Ni performed the validation of specific serum miRNAs. Y Gu and RJ Smales wrote and edited the manuscript.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Subjects and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of interest
  9. Author contributions
  10. References