Sage Journals: Discover world-class research

Abstract

BACKGROUND:

Increasing studies have identified a series of circulating mircoRNAs (miRNAs) as biomarkers for disease detection due to their stability in the blood. The aim of the present study was to identify specific plasma miRNAs as potential biomarkers for nasopharyngeal carcinoma (NPC) detection.

MATERIALS AND METHODS:

Relative public microarray data were obtained and analyzed for screening of the plasma differentially expressed miRNAs (DEMs) between NPC patients and controls. This study contained two phases: a screening phase and a validation one. Logistic regression and receiver operating characteristics curve (ROC) analyses were used to identify DEM signatures. Moreover, targeted genes of the selected DEMs were predicted and their functions were annotated by using bioinformatic analysis.

RESULTS:

Both the screening and the validation phases showed that three miRNAs (miR-548q, miR-630 and miR-940) in the plasma of NPC patients were up-regulated compared to those of controls. They can be used as biomarkers for discriminating NPC patients from non-cancerous controls. Moreover, we found a classifier including only two miRNAs (miR-548q and miR-940) that can be used as a diagnostic signature for NPC, achieving an area under curve (AUC) of 0.972, a sensitivity of 0.94, and a specificity of 0.925.

CONCLUSIONS:

The present study demonstrated that three miRNAs (miR-548q, miR-630 and miR940) might be novel and useful biomarkers for NPC detection. A two-miRNA signature (miR-548q and miR940) may be considered as a better biomarker for NPC detection with relatively high sensitivity and specificity. Future studies with large sample sizes are needed for further validation.

Keywords

Nasopharyngeal carcinoma diagnosis microRNA accuracy plasma

1. Introduction

Nasopharyngeal carcinoma (NPC) is a head and neck cancer rare throughout most areas of the world except for Southeast Asia, which is often accompanied by early lymph node metastasis and high rate of distant metastasis when diagnosed [1]. Evidence showed that virus infection such as Epstein-Barr virus (EBV) and human papillomavirus (HPV) may be one of the important etiological factors for NPC [2]. Thus, the popular clinical present screening method for NPC has been detection of the circulating EBV-associated index such as EBV-DNA and EBV-VCA-IgA [3]. However, not all NPC cases have a correlation with virus infection. Additionally, EBV infection has also been indicated to have a relation with other diseases such as lymphoma and gastric cancer [4], and HPV infection has been suggested to increase breast [5] and cervical cancer [6] risk, which lowered the specificity of virus infection index as a biomarker for NPC diagnosis. In addition, no convincible evidence suggests EBV-associated index as a useful biomarker for NPC screening [7]. Therefore, to find new biomarkers for detection of NPC is required.

Recently, much attention has been focused on the role of microRNA (miRNA) as biomarkers for disease diagnosis. miRNAs are a family of short noncoding RNA molecules that can bind to mRNAs and suppress their expression, which can be secreted into the circulation and exhibit aberrant expression under different physiological and pathological conditions and exist stably [8]. Hence, circulating miRNA expression profiles has been regarded as a potential biomarker for diagnosis of disorders, particularly carcinomas. For example, evidence indicates that a six-miRNA signature in plasma might act as a non-invasive biomarker in diagnosis of pancreatic cancer [9]. Combination of circulating miR-21 and let-7a as a biomarker has advantages over Computed Tomography-Guided Core-Needle Biopsy in lung cancer diagnosis [10]. Plasma miR-125a-3p may be a promising diagnosis biomarker for early-stage colon cancer [11]. Thus, circulating miRNAs might also act as a diagnostic biomarker for NPC.

In the present study, we aimed to screen the dysregulated plasma miRNAs in NPC cases and further assess their diagnostic values. First, microarray-based datasets were obtained from Gene Expression Omnibus (GEO) database. Then, differentially expressed miRNAs (DEMs) were screened out. The obviously and markedly up-regulated or/and down-regulated miRNAs were selected for further validation. Afterwards, the diagnostic values of the selected DEMs were also assessed.

2. Material and methods

2.1 Data source

To obtain circulating DEMs between NPC compared to healthy controls, we downloaded the public gene expression profile (GSE43329) from the GEO database (http://www.ncbi.nlm.nih.gov/geo/).

This dataset was deposited by Zheng et al. in 2013, containing plasma samples from 31 NPC patients and 19 healthy controls. The experiment type was a Non-coding RNA profiling based on Agilent-028035 hsa_miRNA14_Virus platform.

2.2 Screening of DEMs

The data of the dataset was downloaded and analyzed using dChip software (v.2011.01) (http://www.hsph.harvard.edu/). T test was used to screen out circulating DEMs between NPC and Healthy controls, with a threshold of $p$ -value $<$ 0.05 and fold-change $\geqslant$ 1.2. The DEMs with |fold-change| of $\geqslant$ 3 were selected for further analysis.

2.3 MiRNA target gene prediction and functional annotation

For prediction of the relationships between miRNAs and their target mRNAs, several computational microRNA-target prediction tools have been developed. Nevertheless, each tool has its own strengths and limitations. To minimize the shortcomings of individual databases, mirDIP database including integration of different tools has been established. Thus, target genes of the screened miRNAs were predicted though the mirDIP database [12].

Candidate mRNAs were selected if they were identified as miRNA targets in this database. To annotate the functions of the predicted target genes, Gene Ontology (GO) and KEGG pathway enrichment analysis were conducted through Kobas tool [13]. $p$ -value $<$ 0.05 was considered statistically significant.

Table 1
Demographic and clinical characteristics of the patients and controls in the screening and validation cohort

Characteristics	NPC ( $n=$ 55)	Controls ( $n=$ 45)
Gender (F/M)	22/33	21/24
Age (year)	48 (26–71)	46 (24–67)
Histological types	Squamous carcinoma
TNM stage
I	12
II	8
III	7
IV	7
Unknown	21
Ethnic (Han)	55 (100%)	45 (100%)

Figure 1.

The dysregulated genes for the microarray-based miRNA expression profile GSE43329. The horizontal axis stands for the name of sample; the right vertical stands for the clustering of mRNAs. Red stands for up-regulated mRNAs, while green stands for down-regulated ones.

2.4 Patient recruitment and sample collection

The NPC patients enrolled were recruited from Chongqing Cancer Hospital and Southwest Hospital during the period 2014 to 2018. They were histopathologically confirmed as NPC. Peripheral blood samples were drawn before any radiation and chemotherapy treatment. Patients of vocal cord polyp without any cancers from Southwest Hospital were enrolled as controls. The characters of the participants were listed in Table 1. Approval for the present study was obtained from the ethics committee and participants have provided informed consent.

This study included two phases: a screening phase and a validation phase. The screening cohort comprised 5 NPC patients and 5 controls, while the validation cohort contained 50 NPC patients, and 40 controls.

Plasma was isolated from all blood samples. Peripheral blood (3 ml) was drawn into EDTA tubes, and these tubes were centrifuged at 2,000 rpm for 10 min. Then, the plasma were transferred to 1.5 ml-tubes and centrifuged at 12,000 rpm for 3 min to remove any remaining cellular debris. Afterwards, the supernatant plasma were transferred to fresh tubes and stored at $-$ 80 ${}^{\circ}$ C.

2.5 RNA extraction

RNA was isolated from 200 $\mu$ L of plasma using the mirVANA PARIS miRNA Isolation Kit (Ambion, Austin, TX, USA) according to the manufacturer’s protocol. 5 $\mu$ L of 5 nM Synthetic cel-miR-39 was used for normalization and extraction efficiency control. The RNA concentration was quantified using a NanoDrop ND-1000 spectrophotometer (Nanodrop, Waltham, MA, USA). The total RNA samples were stored at $-$ 80 ${}^{\circ}$ C for future use.

Table 2
The top ten DEMs screened from GSE43329

Probe set	NPC mean	Control mean	Fold change	$p$
hsa-miR-135a	492.54	141.61	3.48	$<$ 0.05
hsa-miR-122	1879.51	543.48	3.46	$<$ 0.05
hsa-miR-548q	740.97	216.51	3.42	$<$ 0.05
hsa-miR-630	533.87	164.1	3.25	$<$ 0.05
hsa-miR-940	998.43	317.66	3.14	$<$ 0.05
hsa-miR-572	1212.11	409.17	2.96	$<$ 0.05
hsa-miR-19a	559.71	200.67	2.79	$<$ 0.05
hsa-miR-1290	382.1	140.62	2.72	$<$ 0.05
hsa-miR-483-5p	977.55	358.76	2.72	$<$ 0.05
hsa-miR-185	281.58	110.3	2.55	$<$ 0.05

2.6 miRNA quantitative PCR

Reverse transcription (RT) was performed using a TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA) based on the manufacturer’s instructions. The human TaqMan MicroRNA Assay Kit (Applied Biosystems, Foster City, CA, USA) was used to quantify the amount of miRNAs. Quantification of miRNA was performed by qPCR using an Applied Biosystems 7900HT thermocycler (Applied Biosystems): incubation at 95 ${}^{\circ}$ C for 10 min, followed by 40 cycles of 95 ${}^{\circ}$ C for 15 s, and 60 ${}^{\circ}$ C for 1 min. qPCR data was analyzed by using the RQ Manager software (Applied Biosystems). miR-16 served as the internal reference. All reactions were conducted in triplicate. The 2 ${}^{-\Delta\Delta\text{Ct}}$ method was used to analyze the relative expression levels of miRNAs.

Figure 2.

Go analysis (a) and KEGG enrichment analysis (b).

2.7 Statistical analysis

The differences expression levels of miRNAs in plasma between cancer patients and healthy controls were analyzed by the Mann-Whitney U-test. Logistic regression was used to develop a combined miRNA panel to predict the probability of NPC.

The diagnostic accuracy of miRNA was measured by the receiver-operating characteristic (ROC) curves and the area under the ROC curve (AUC). The optimal diagnostic point of the signature was assessed at cut-off values with the largest Youden’s index (sensitivity $+$ specificity $-$ 1). A $p$ -value of less than 0.05 was considered as statistically significant differences. Statistical analysis was performed using SPSS version 18.0 software (SPSS Inc., Chicago, IL, USA) and Medcalc Version 11.1.1.0 software (Maria Kerke, Belgium).

3. Results

3.1 Characteristic of the subjects

A total of 55 NPC patients, and 45 non-cancerous controls were enrolled in the present study. As shown in Table 1, in the screening cohort, 5 NPC patients and 5 controls were enrolled. In the validation cohort, the remaining of them (50 NPC cases and 40 controls) was involved.

3.2 Identification of plasma DEMs between NPC patients and controls

On the basis of the public microarray dataset GSE43329, the T test in dChip Software was used to analyze the gene expression profiles and screen the DEMs between NPC group and control group with the criteria described. As a result, a total of 31 DEMs met the criteria and were screened out (Fig. 1). Interestingly, these DEMs happened to be up-regulated in the plasma of NPC patients compared with those in the controls.

The top ten DEMs were listed in Table 2. There were five miRNAs, namely, miR-135, miR-122, miR-548q, miR-630 and miR-940, met the criteria and were selected for further analysis.

3.3 Prediction of miRNA target genes associated with NPC

Using the mirDIP database that combines the output of multiple prediction algorithms, we identified a total of 3150 predicted targets of the 5 miRNAs classifier.

To determine the functions of these target genes, we submitted these gene IDs to Kobas tool for analysis. As shown in Fig. 2, the top ten significant GO items have an association with metabolic processes, such as organic substance metabolic process, cellular metabolic process, and primary metabolic process. The pathway enrichment analysis showed that the most significant dysfunctional pathways were Pathways in cancer, MAPK signaling pathway, and Proteoglycans in cancer. The results suggested that the 5-miRNA classifiers may participate in the metabolic process of cells, and might be involved in the development of NPC through regulation of the series of pathways.

Figure 3.

The expression levels of the 5 miRNAs in the plasma of 5 NPC patients and 5 controls, as determined by qRT-PCR in the screening phase. (a) miR-548q, miR-630 and miR-940; (b) miR-122 and miR-135a.

Figure 4.

The expression levels of the 3 miRNAs in the plasma of 50 NPC patients and 40 controls, as determined by qRT-PCR in the validation phase. (a) miR-548q; (b) miR-630; (c) miR-940.

3.4 Screening of miRNAs by quantitative PCR

To screen plasma DEMs for NPC, a real-time PCR was used to compare the 5 candidate plasma miRNAs between 5 NPC patients and 5 controls.

The results showed that 3 of them, miR-548q, miR-630 and miR-940 were up-regulated in the plasma of NPC patients relative to those of controls, respectively (Fig. 3a). The expression levels of the remaining two miRNAs, miR-135a and miR-122 were not dysregulated compared to the controls, respectively (Fig. 3b). Thus, miR-548q, miR-630 and miR-940 were enrolled in the validation phase.

3.5 Validation of miRNAs by quantitative PCR

In this phase, three miRNAs, namely, miR-548q, miR-630 and miR-940 were involved for validation. A total of 50 NPC patients and 40 non-cancerous controls were included.

The results showed that three miRNAs were significantly up-regulated in the plasma of NPC patients compared with those of the controls, respectively (Fig. 4).

Table 3
Diagnostic accuracy of the selected plasma miRNAs in NPC patients

miRNAs	Sensitivity	Specificity	AUC	Standard error	95% CI of AUC
miR-548q	0.840	0.725	0.855	0.0391	0.765–0.920
miR-630	0.820	0.825	0.847	0.0424	0.755–0.914
miR-940	0.920	0.700	0.879	0.0363	0.793–0.938
miR-548q $+$ miR-630	0.740	0.925	0.905	0.0298	0.824–0.956
miR-548q $+$ miR-940	0.940	0.925	0.972	0.0138	0.913–0.995
miR-630 $+$ miR-940	0.980	0.775	0.949	0.0206	0.881–0.984
miR-548q $+$ miR-630 $+$ miR-940	0.980	0.875	0.982	0.00959	0.929–0.998

Figure 5.

Candidate miRNAs as potential biomarkers for NPC diagnosis. (a) The 3 individual miRNA achieved an AUC of 0.847–0.879. (b) The classifier of 2 miRNAs (miR-548q $+$ miR-940) yielded an AUC of 0.972 (95% CI: 0.913–0.995), with a sensitivity of 0.940 and a specificity of 0.925. (c) The classifier of 3 miRNAs (miR-548q $+$ miR-630 $+$ miR940) yielded an AUC of 0.982 (95% CI: 0.929–0.998), with a sensitivity of 0.980 and a specificity of 0.875.

3.6 Diagnostic value of the miRNA classifier in NPC

To determine the diagnostic values of these miRNAs in NPC, we used Logistic regression and ROC analyses to determine the best combination of miRNAs to predict NPC.

ROC curves were constructed to compare the relative expression of the 3 miRNAs for NPC patients and controls. As a result, the following AUCs were generated as shown in Table 3: miR-548q, 0.855 (95% CI, 0.765–0.920); miR-630, 0.847 (95% CI, 0.755–0.914); and miR-940, 0.879 (95% CI, 0.793–0.938) (Fig. 5). Of the 3 miRNAs, miR-940 displayed the highest AUC of 0.879 for diagnosis of NPC.

Though a single miRNA could be used as a biomarker for distinguishing NPC patients from healthy controls, a combination of several miRNAs might provide a better differentiation power than individual miRNA. The miRNA signatures for defining NPC were generated by logistic regression analysis. Consequently, compared with that of any single miRNA, higher AUCs were observed when the combined miRNAs were used. As shown in Fig. 5, the AUCs for miR-548q $+$ miR-630, miR-548q $+$ miR-940, and miR-630 $+$ miR-940 were 0.905 (95% CI, 0.824–0.956), 0.972 (95% CI, 0.913–0.995), and 0.949 (95% CI, 0.881–0.984), respectively. Combination of the three miRNAs achieved the highest AUC as 0.982 (95% CI, 0.929–0.998).

Notably, though the AUC for the combination of the three miRNAs was the highest, the AUC for miR-548q $+$ miR-940 (0.972) was close to that of the three-miRNA combination (0.982). In addition, the specificity of miR-548q $+$ miR-940 (0.925) was higher than that of the three-miRNA combination (0.875). The data showed that the combination of miR-548q and miR-940 had a better performance than the entire three-miRNA signature, suggesting that this two-miRNA signature might be an easy and appropriate model for NPC detection.

4. Discussion

Circulating miRNAs have been indicated to be biomarkers for cancer diagnosis. To find potential miRNAs for NPC detection, we screened possible candidate DEMs from the public microarray-based data, and then validate their expressions in the plasma of NPC patients. Interestingly, three miRNAs, which might be up-regulated in the plasma of NPC patients, have been suggested to be the candidate DEMs. Each of them can be used to distinguish NPC patients from healthy controls. Interestingly, we found that the combination of 2 miRNAs, miR-548q and miR-940, were shown to have the highest sensitivity and specificity, and relatively high AUC for predicting NPC through logistic regression and ROC analyses.

By analyzing the data of the public microarray, we found that a total of 31 DEMs were up-regulated in NPC patients relative to the healthy controls, indicating that these up-regulated miRNAs might play a critical role in the genesis and progression of NPC. Of these DEMs, five top DEMs that have a fold-change of more than 3.0 were selected to be up-regulated in the plasma of NPC patients. GO analysis showed that the selected miRNAs might have a relationship with metabolic process of cells, implying that metabolic disturbance might be one of the key mechanisms involved in NPC genesis. Evidence suggested that EBV-LMP1 may trigger multiple pathways and induce metabolic reprogramming that might support tumorigenicity and malignancy [14]. Moreover, since alteration of energy metabolism is an essential hallmark of cancer, disturbance of glucose metabolism has been indicated to contribute to NPC pathogenesis [15]. This may help clarify the roles of metabolic disturbance and multiple signaling pathways in the development of NPC.

In the screening phase, only three out of the five DEMs were tested to be up-regulated in the plasma of NPC patients compared to those of controls. Hence, they were tested in the validation phase. A single miRNA can regulate multiple target genes, and conversely, a single gene could be regulated by a number of miRNAs. Thus, a network can be constructed by the miRNAs and the target genes, which might intensively and accurately regulate various biological processes. miR-548q has been rarely studied and only a few relative literature can be searched. A report showed that over-expression of the non-classical human leukocyte antigen G (HLA-G) can contribute to renal cell cancer progression, which can be refrained by miR-548q through enhancement of NK cell-mediated HLA-G-dependent cytotoxicity [16]. Previous report had shown that miR-548q might act as a biomarker for NPC detection [17], in accordance with the present study. Hence, miR-548q can act as either an oncogene or a tumor suppressor under different conditions. Likewise, miR-630 also plays dual roles in tumor progression in various human cancers. For example, low miR-630 expression can confer cisplatin resistance in lung cancer [10]. Conversely, in ovarian carcinoma, over-expression of miR-630 was detected, and its inhibition resulted in increased chemosensitivity of the cancer cells [18]. For miR-940, a number of studies have investigated its roles in cancers. Reports showed that low expression of miR-940 in the serum may be a biomarker for breast cancer [19]. However, miR-940 was up-regulated in cervical cancer, which led to cancer progression [20]. Taken together, these miRNAs may act as either oncogenes or tumor suppressors under different biological conditions. It is worth noting that only up-regulated miRNAs were identified in the screening phase. A possible reason for clarifying this discrepancy might be that activation of oncogenes may play a predominant role in the genesis and progression of NPC. In addition, the sample difference and the sensitivity of the microarray for detecting the gene expression levels might also influence the results.

To date, there has been no reliable serum miRNA internal reference generally accepted [21]. Which factor was the best endogenous control for circulating miRNAs detection has remained controversial. Previous studies explored the stabilities of some miRNAs as internal references. For example, Zalewski et al. [22] indicated that miR-93-5p and miR-425-5p were robust in plasma of vulvar intraepithelial neoplasia lesion and vulvar squamous cell carcinoma patients. However, these two miRNAs (miR-93-5p and miR-425-5p) seem to be unstable because they might act as the biomarkers monitoring the therapy of head and neck cancer [23]. U6 has been widely used as internal reference when circulating miRNAs were detected [24, 25]. Nevertheless, Xiang et al. [26] and Benz et al. [27] proposed that U6 is not suitable to be the endogenous control for the quantification of circulating microRNAs because the level of U6 in the blood was unstable. Instead, miR-16 has been indicated by a number of studies to be the most stably expressed reference miRNA for detection of circulating miRNAs in cancers [28, 29, 30]. In addition, in the previous report regarding circulating miRNAs in NPC patients, miR-16 has been also selected as a relatively reliable reference [31]. Therefore, on the basis of the above reasons, miR-16 was selected as the internal reference in our study.

A previous study by Zheng et al. in 2014 [17] used the same dataset for analysis. However, though miR-548q, miR-630 and miR-940 have been detected to be dysregulated in NPC, the diagnostic values of miR-630 and miR-940 had not been evaluated in their study. The roles of miR-630 and miR-940 have rarely been investigated in NPC. To our knowledge, we for the first time evaluated the diagnostic values of each miRNA as well as the combinations of them by using a cohort with a relatively large sample size.

Notably, the present study only focused on detection of NPC. If the plasma of the patients after treatment can be obtained, it would be interesting to assess the diagnostic value of these miRNAs for the early detection of relapse upon treatment. Hence, future studies are needed to explore the significance of the miRNA expression variations in the plasma of NPC patients before and after treatment.

Several limitations might be involved in the present study. First, the sample sizes are rather small, leading to any selection bias. Second, the confounding factors such as smoking and drinking were not assessed in this study on account of the limited sample size. Third, since there were not standard accepted internal references for circulating miRNA detection, the results might be interpreted with care. Nevertheless, the validated three miRNAs may be of great significance for NPC detection.

In conclusion, three miRNAs (miR-548q, miR-630, and miR-940) could be used as biomarkers for discriminating NPC from non-cancer individuals. A combination of two miRNAs (miR-548q and miR-940) had high discriminatory power and simple operability for NPC diagnosis. Future studies containing larger sample sizes are needed to confirm the conclusions.

Footnotes

Acknowledgments

This study was supported by the Fund of Chongqing Health and Family Planning Commission (20141020).

Conflict of interest

None declared.

References

Wang

Zhang

Chen

Cai

Zeng

Xiong

Wang

and Xiang

, Vimentin is a crucial target for anti-metastasis therapy of nasopharyngeal carcinoma, Mol Cell Biochem 438 (2018), 47–57.

Ngan

H.L.

Wang

K.W.

and Lui

V.W.Y.

, Genomic landscapes of EBV-associated nasopharyngeal carcinoma vs. HPV-associated head and neck cancer, Cancers (Basel) 10 (2018).

Kim

K.Y.

Q.T.

Yom

S.S.

R.H.W.

Chan

K.C.A.

Bratman

S.V.

Welch

J.J.

Divi

R.L.

Petryshyn

R.A.

and Conley

B.A.

, Clinical utility of epstein-barr virus DNA testing in the treatment of nasopharyngeal carcinoma patients, Int J Radiat Oncol Biol Phys 98 (2017), 996–1001.

Tan

G.W.

Visser

Tan

L.P.

van den Berg

and Diepstra

, The microenvironment in epstein-barr virus-associated malignancies, Pathogens 7 (2018).

Kouloura

Nicolaidou

Misitzis

Panotopoulou

Smyrniotis

Corso

Veroneci

and Arkadopoulos

, HPV infection and breast cancer. Results of a microarray approach, Breast 40 (2018), 165–169.

McGregor

Saulo

Brotherton

Liu

Phillips

Skinner

S.R.

Luey

Oliver

Stewart

Tabrizi

S.N.

Garland

and Kaldor

J.M.

, Decline in prevalence of human papillomavirus infection following vaccination among Australian Indigenous women, a population at higher risk of cervical cancer: The VIP-I study, Vaccine 36 (2018), 4311–4316.

Yang

Zhou

and Chen

X.Y.

, Screening for nasopharyngeal cancer, Cochrane Database Syst Rev (2015), CD008423.

Wang

Peng

Wang

Qin

and Xue

, Circulating microRNAs as potential cancer biomarkers: The advantage and disadvantage, Clin Epigenetics 10 (2018), 59.

Zhou

Wang

Huang

Zhu

and Miao

, Plasma miRNAs in diagnosis and prognosis of pancreatic cancer: A miRNA expression analysis, Gene (2018).

10.

Chen

M.J.

D.W.

Wang

G.C.

Wang

Y.C.

Chen

C.Y.

and Lee

, MicroRNA-630 may confer favorable cisplatin-based chemotherapy and clinical outcomes in non-small cell lung cancer by targeting Bcl-2, Oncotarget 9 (2018), 13758–13767.

11.

Wang

Yan

Zhao

Zhan

Wang

Zhang

and Huang

, Circulating exosomal miR-125a-3p as a novel biomarker for early-stage colon cancer, Sci Rep 7 (2017), 4150.

12.

Tokar

Pastrello

Rossos

A.E.M.

Abovsky

Hauschild

A.C.

Tsay

and Jurisica

, mirDIP 4.1-integrative database of human microRNA target predictions, Nucleic Acids Res 46 (2018), D360–D370.

13.

Xie

Mao

Huang

Ding

Dong

Kong

Gao

C.Y.

and Wei

, KOBAS 2.0: A web server for annotation and identification of enriched pathways and diseases, Nucleic Acids Res 39 (2011), W316–22.

14.

Luo

Hong

Cheng

Zhao

Shi

Liu

Fan

Zhou

Bode

A.M.

and Cao

, DNMT1 mediates metabolic reprogramming induced by Epstein-Barr virus latent membrane protein 1 and reversed by grifolin in nasopharyngeal carcinoma, Cell Death Dis 9 (2018), 619.

15.

Zhang

Jia

Tsang

C.M.

and Tsao

S.W.

, EBV infection and glucose metabolism in nasopharyngeal carcinoma, Adv Exp Med Biol 1018 (2017), 75–90.

16.

Jasinski-Bergner

Reches

Stoehr

Massa

Gonschorek

Huettelmaier

Braun

Wach

Wullich

Spath

Wang

Marincola

F.M.

Mandelboim

Hartmann

and Seliger

, Identification of novel microRNAs regulating HLA-G expression and investigating their clinical relevance in renal cell carcinoma, Oncotarget 7 (2016), 26866–78.

17.

Zheng

X.H.

Cui

Ruan

H.L.

Xue

W.Q.

Zhang

S.D.

Y.Z.

Zhou

X.X.

and Jia

W.H.

, Plasma microRNA profiling in nasopharyngeal carcinoma patients reveals miR-548q and miR-483-5p as potential biomarkers, Chin J Cancer 33 (2014), 330–8.

18.

Eoh

K.J.

Lee

S.H.

Kim

H.J.

Lee

J.Y.

Kim

S.W.

Kim

Y.T.

and Nam

E.J.

, MicroRNA-630 inhibitor sensitizes chemoresistant ovarian cancer to chemotherapy by enhancing apoptosis, Biochem Biophys Res Commun 497 (2018), 513–520.

19.

Liu

Guan

and Meng

, Clinical potential of miR-940 as a diagnostic and prognostic biomarker in breast cancer patients, Cancer Biomark 22 (2018), 487–493.

20.

Wang

C.F.

Zhang

Cai

Y.J.

Zhang

Y.Y.

and Zhao

, miR-940 upregulation contributes to human cervical cancer progression through p27 and PTEN inhibition, Int J Oncol (2017).

21.

Q.S.

Weiland

R.Q.

Gao

X.H.

Poisson

L.M.

and Zhou

, Identification of mouse serum miRNA endogenous references by global gene expression profiles, PLoS One 7 (2012), e31278.

22.

Zalewski

Misiek

Kowalik

Bakula-Zalewska

Kopczynski

Zielinska

Bidzinski

Radziszewski

Gozdz

and Kowalewska

, Normalizers for microRNA quantification in plasma of patients with vulvar intraepithelial neoplasia lesions and vulvar carcinoma, Tumour Biol 39 (2017), 1010428317717140.

23.

Summerer

Niyazi

Unger

Pitea

Zangen

Hess

Atkinson

M.J.

Belka

Moertl

and Zitzelsberger

, Changes in circulating microRNAs after radiochemotherapy in head and neck cancer patients, Radiat Oncol 8 (2013), 296.

24.

Jones

Nourse

J.P.

Keane

Bhatnagar

and Gandhi

M.K.

, Plasma microRNA are disease response biomarkers in classical Hodgkin lymphoma, Clin Cancer Res 20 (2014), 253–64.

25.

Komatsu

Ichikawa

Takeshita

Tsujiura

Morimura

Nagata

Kosuga

Iitaka

Konishi

Shiozaki

Fujiwara

Okamoto

and Otsuji

, Circulating microRNAs in plasma of patients with oesophageal squamous cell carcinoma, Br J Cancer 105 (2011), 104–11.

26.

Xiang

Zeng

Yang

Chen

Zhong

Xie

and Zeng

, U6 is not a suitable endogenous control for the quantification of circulating microRNAs, Biochem Biophys Res Commun 454 (2014), 210–4.

27.

Benz

Roderburg

Vargas Cardenas

Vucur

Gautheron

Koch

Zimmermann

Janssen

Nieuwenhuijsen

Luedde

Frey

Tacke

Trautwein

and Luedde

, U6 is unsuitable for normalization of serum miRNA levels in patients with sepsis or liver fibrosis, Exp Mol Med 45 (2013), e42.

28.

Song

Bai

Han

Zhang

Meng

Han

and Zhang

, Identification of suitable reference genes for qPCR analysis of serum microRNA in gastric cancer patients, Dig Dis Sci 57 (2012), 897–904.

29.

Wang

Liu

Jiang

Zheng

Wang

Zhang

Liu

Pan

Yang

and Wang

, Identification and validation of reference genes for the detection of serum microRNAs by reverse transcription-quantitative polymerase chain reaction in patients with bladder cancer, Mol Med Rep 12 (2015), 615–22.

30.

Chang

K.H.

Mestdagh

Vandesompele

Kerin

M.J.

and Miller

, MicroRNA expression profiling to identify and validate reference genes for relative quantification in colorectal cancer, BMC Cancer 10 (2010), 173.

31.

Liu

Cui

R.X.

Sun

Guo

Mao

Y.P.

Tang

L.L.

Jiang

Liu

Cheng

Y.K.

Q.M.

Cho

W.C.

Liu

L.Z.

and Ma

, A four-miRNA signature identified from genome-wide serum miRNA profiling predicts survival in patients with nasopharyngeal carcinoma, Int J Cancer 134 (2014), 1359–68.

Plasma microRNA expression signature involving miR-548q,miR-630 and miR-940 as biomarkers for nasopharyngeal carcinoma detection

Abstract

BACKGROUND:

MATERIALS AND METHODS:

RESULTS:

CONCLUSIONS:

Keywords

1. Introduction

2. Material and methods

2.1 Data source

2.2 Screening of DEMs

2.3 MiRNA target gene prediction and functional annotation

Table 1 Demographic and clinical characteristics of the patients and controls in the screening and validation cohort

2.5 RNA extraction

Table 2 The top ten DEMs screened from GSE43329

3. Results

3.1 Characteristic of the subjects

3.2 Identification of plasma DEMs between NPC patients and controls

3.3 Prediction of miRNA target genes associated with NPC

3.5 Validation of miRNAs by quantitative PCR

Table 3 Diagnostic accuracy of the selected plasma miRNAs in NPC patients

4. Discussion

Footnotes

Acknowledgments

Conflict of interest

References

Table 1
Demographic and clinical characteristics of the patients and controls in the screening and validation cohort

Table 2
The top ten DEMs screened from GSE43329

Table 3
Diagnostic accuracy of the selected plasma miRNAs in NPC patients