close

Вход

Забыли?

вход по аккаунту

?

j.lungcan.2017.10.002

код для вставкиСкачать
Lung Cancer 114 (2017) 6–11
Contents lists available at ScienceDirect
Lung Cancer
journal homepage: www.elsevier.com/locate/lungcan
Research paper
Identification of a serum microRNA expression signature for detection of
lung cancer, involving miR-23b, miR-221, miR-148b and miR-423-3p
MARK
Ying Zhua,1, Tao Lia,1, Gang Chenb, Guifang Yana, Xiaojing Zhanga, Ying Wanb, Qijing Lic,
⁎
⁎
Bo Zhua, , Wenlei Zhuoa,
a
b
c
Institute of Cancer, Xinqiao Hospital, Third Military Medical University, Chongqing 400037, China
Biomedical Analysis Center, Third Military Medical University, Chongqing 400038, China
Department of Immunology, Duke University Medical Center, Durha, USA
A R T I C L E I N F O
A B S T R A C T
Keywords:
Serum
MicroRNA
Lung cancer
Biomarker
Diagnosis
Objectives: Serum mircoRNAs (miRNAs), with their noticeable stability and unique expression pattern in patients
with various diseases, are powerful novel non-invasive biomarkers for cancer detection. The objective of this
study was to identify specific serum miRNAs as potential diagnostic markers for detection of lung cancer.
Materials and methods: The expression of serum miRNA from treatment-naive lung cancer patients (LC), benign
pulmonary disease patients (PD) and healthy controls (HC) were examined by PCR array. The study was divided
into two phases: the biomarker-screening phase and the biomarker-validation phase. Logistic regression and
receiver operating characteristics curve analyses were used to identify differentially expressed miRNA signatures
that could distinguish LC from PD and HC. In addition, target genes of miRNAs were predicted using bioinformatic assays.
Results: Ten miRNAs (let-7f, miR-126-3p, miR-148b, miR-151-5p, miR-199a-3p, miR-221, miR-23b, miR-26a,
miR-27b, and miR-423-3p) in LC were significantly increased compared to those in PD and HC in biomarkervalidation phase (P < 0.05). Bioinformatic analyses showed that predicted targets of these miRNAs might have
a correlation with formation and development of cancer. Furthermore, we have developed classifiers including 4
miRNAs (miR-23b, miR-221, miR-148b and miR-423-3p) that can be demonstrated as a signature for LC detection, yielding a receiver operating characteristic curve area of 0.885.
Conclusion: our findings define a distinct miRNA expression profile in LC cases. These 4-miRNA signatures (miR23b, miR-221, miR-148b and miR-423-3p) may be considered as novel, non-invasive biomarker for LC diagnosis.
1. Introduction
Lung cancer is the leading cause of cancer death worldwide.
Incidence of lung cancer has been increasing in Asian countries. In
China, it ranks the first in men and women, respectively [1]. At the time
of diagnosis, the majority of lung cancers are often in an advanced
stage. The curative effect especially long-term outcome is still very
poor. Since reports showed that surgical resection for early stage or
locally advanced lung cancer adjuvant with other treatment are the
keys to improve curative effect and survival rate in lung cancer patients
[2], early detection might be a critical step that influences the prognosis
of lung carcinoma. For instance, the program U.S. screening study
(NLST) had provided growing evidence for early detection of lung
cancer [3].
Recently, more attention has been focused on serum biomarkers for
early detection of malignancies. Evidence indicated the surprising discovery that serum contains a large amount of stable microRNAs
(miRNAs) derived from various tissues [4]. miRNAs are a class of highly
conserved, single-stranded, small (19–25 nucleotides) noncoding RNAs
that can negatively regulate gene expression at the post-transcriptional
or translational level via pairing with complementary messenger RNA
bases, leading to mRNA degradation or translational inhibition. They
play important roles in diverse cellular processes such as cell differentiation, proliferation and apoptosis. miRNAs are resistant to RNaseA
digestion and other harsh conditions, which potentially explain the
robustness of serum miRNAs [5].
Abbreviations: SQ, squamous carcinoma; AD, adenocarcinoma; S, small cell lung cancer; HC, health control; LC, lung cancer; PD, pulmonary disease; MiR, microRNA; ROC, receiver
operating characteristic; COPD, chronic obstructive pulmonary disease
⁎
Corresponding authors.
E-mail addresses: b.davis.zhu@gmail.com (B. Zhu), zhuowenlei@tmmu.edu.cn (W. Zhuo).
1
Contributed equally.
http://dx.doi.org/10.1016/j.lungcan.2017.10.002
Received 5 April 2017; Received in revised form 7 September 2017; Accepted 3 October 2017
0169-5002/ © 2017 Elsevier B.V. All rights reserved.
Lung Cancer 114 (2017) 6–11
Y. Zhu et al.
Serum miRNAs might be suitable biomarkers for distinguishing
cancer patients from control subjects due to their characteristic expression profile in different disorders [6]. Several published studies
have been conducted to discover possible miRNAs as biomarkers to
differentiate lung cancer from healthy population [7]. However, ethnic
variation was not considered in these studies. Also, only cancer group
and healthy control group were enrolled in the studies, but it is very
important to differentiate malignant and benign pulmonary diseases
clinically. In the present study, we aimed to more comprehensively
profile circulating miRNAs as potential markers for lung cancers.
2.2. RNA extraction
2. Materials and methods
2.3. MiRNA expression profiling and miRNA quantitative PCR
2.1. Patient recruitment and sample collection
For quantitative analysis of the mature miRNA expression, E. coli
poly A polymerase (Epicentre) was utilized to produce polyadenylated
tails at the 3′-end of all RNA molecules. Reverse transcription (RT) was
performed using the qScript Flex cDNA synthesis kit (Quanta
Biosciences) based on the manufacturer’s instructions. A point worth
emphasizing is that 0.02 pmol synthetic C. elegans miRNA (cel-miR-39)
was introduced before the poly A plus step to teach sample to surveillance technical variations in Reverse Transcription.
Then, a SYBR Green–based real-time PCR method was used to
quantify the relative expression of mature miRNAs as previously described [9]. In the miRNA expression profiling array, a total of 355
mature miRNAs were evaluated. The qRT-PCR was carried out on ViiA7
PCR Thermocycle Instrument, as described in the reference [10].
Threshold cycle (Ct) for each miRNA was extracted using MXPro software (Stratagene) by setting threshold and automatically defined
baseline. The −ΔΔCt method was used to analyze the relative quantitative expression levels of miRNAs. miRNA expression was normalized
by geometric mean–based global normalization using the Realtime
StatMiner (Integromics) analysis software.
RNA was isolated from serum by using the miRNeasy Serum/Plasma
Kit (Qiagen) as described previously [8]. In brief, two hundred microliters of serum was mixed with denaturing buffer in the volumes described in the manufacturer's instructions. The homogenate was incubated at 25 °C for 5 min. Then, 0.02 pmol synthetic C. elegans miRNA
(cel-miR-39) was introduced into the mixture. Next, the manufacturer's
protocols were followed for RNA extraction. Finally, RNA was recovered in 14 μL of RNase-free water. The RNA concentration was
quantified by NanoDrop ND-1000 (Nanodrop, USA).
Approval for the present study was obtained from the ethics committee and participants have given written informed consent. The studies were approved by the Xinqiao hospital of Third Military Medical
University. A multiphase, case–control study was designed to identify
serum miRNAs as biomarkers for lung cancer detection. This study included a total of two phases: a biomarker-screening phase and a biomarker-validation phase. The screening cohort comprised 10 lung
cancer patients (LC) and 10 healthy controls (HC), while the validation
cohort contained 40 LC, 40 benign pulmonary disease patients (PD) and
40 HC. LC and PD patients were enrolled from Oncology Department
and Respirology Department of Xinqiao hospital, and HC cases were
enlisted from the health examination center of Xinqiao hospital from
February 2013 to April 2015.
For the lung cancer group, the diagnosis was based on radiological
examination and histologically confirmed by biopsy and pathological
analysis and staged according to the tumor-node-metastasis (TNM)
staging system of the International Union Against Cancer. The benign
pulmonary disease group was diagnosed by clinical criteria and verified
by imageological examination, etiological examination or pulmonary
function test. For the HC group, blood samples were collected from
individuals who were proven to be cancer-free or benign lung diseasefree after chest CT scan. PD group and HC group were matched to the
lung cancer patients according to age, gender and race. Clinical information was collected for each participant at the time of blood collection and summarized in Table 1. Cell-free serum was isolated from
all blood samples using a two-step centrifugation protocol (2000 rpm
for 10 min, 12,000 rpm for 3 min) at room temperature to prevent
contamination by cellular nucleic acids. Then, the supernatant sera
were stored at −80 °C until analyzed.
2.4. MiRNA target gene prediction and bioinformatics analysis
Target genes of miRNAs were predicted using the miRWalk database
(http://www.ma.uniheidelberg.de/apps/zmf/mirwalk/).
Candidate
mRNAs were selected if they were identified as miRNA targets in at
least 5 of 8 databases. To determine the functions of the predicted
target genes, DAVID (Database for Annotation, Visualization and
Integrated Discovery) was used. This database allowed us to assign
predicted target genes to functional groups based on molecular function, biological process, and specific pathways. Gene annotation and
KEGG Pathway of the predicted target genes were analyzed. Pvalue < 0.05 was considered statistically significant. However, in order
to reduce the screening range, a P-value of < 0.01 was chosen as a
cutoff for the target prediction and their functional annotation.
Table 1
Demographic and clinical characteristics of the patients and control individuals in the
screening cohort and validation cohort.
Characteristics
LC (n = 50)
PD (n = 40)
HC (n = 50)
Gender (F/M)
Age (year)
TNM stage
11/39
61.3 ± 10.1
I 2 (4%)
II 6 (12%)
III 12 (24%)
IV 30 (60%)
50 (100%)
SCLC 8 (16%)
SQ 13 (26%)
AD 23 (46%)
Other type 6 (12%)
8/32
57.2 ± 15.6
20/30
44.0 ± 13.8
40 (100%)
Tuberculosis 15 (37.5%)
COPD 10 (25%)
Pneumonia 15 (37.5%)
50 (100%)
Ethnic(Han)
Pathologic type
2.5. Statistical analysis
The patients’ demographics were reported as mean ± standard
deviation or frequencies and percentages for continuous and categorical
variables, respectively. Hierarchical clustering analysis was performed
using Cluster 3.0. The Mann-Whitney test or one-way analysis was used
to compare the differences in plasma miRNA expression between
groups. The P-value < 0.05 was considered as statistically significant
differences. Logistic regression was used to develop a combined miRNA
panel to predict the probability of lung cancer. Receiver-operating
characteristic (ROC) curves and the area under the ROC curve (AUC)
were used to evaluate the diagnostic values of miRNA biomarkers for
the diagnosis of lung cancer. The optimal diagnostic point of the signature was assessed at cut-off values with the largest Youden’s index
(sensitivity + specificity − 1). Statistical analysis was performed using
SPSS version 18.0 software (SPSS Inc., Chicago, IL).
LC: Lung cancer; PD: Pulmonary disease control; HC:Health control; SQ: Squamous
Cancer; SCLC: Small Cell Lung Cancer; AD: adenocarcinoma; COPD: Chronic obstructive
pulmonary disease.
7
Lung Cancer 114 (2017) 6–11
Y. Zhu et al.
3. Results
above 6 SCLC, 9 SQ and 19 AD pathological subgroups (Table 3 and
Fig. 1c). Hence, we chose these 10 miRNAs for further investigation.
3.1. Characteristic of the subjects
3.4. Diagnostic value of the miRNA classifier in lung cancer
A total of 50 LC patients, 40 PD cases, and 50 HC donors were enrolled in the present study. As shown in Table 1, there were no significant differences between the three groups in the distribution of most
clinical parameters, such as gender (P = 0.056) and ethnicity
(P > 0.05). In the screening cohort, there are 2 SCLC and 8 NSCLC
(including 4 squamous carcinoma and 4 adenocarcinoma) in the LC
group, and 10 HC donors were enrolled as controls. In the validation
cohort, there are 6 SCLC and 34 NSCLC (including 9 squamous carcinoma,19 adenocarcinoma and 6 other pathology type) in the LC group,
and 40 PD cases (including 15 tuberculosis, 10 COPD and 15 pneumonia) as well as 40 HC donors were enlisted as controls.
To evaluate the sensitivity and specificity of these miRNAs in diagnosing lung cancer, we used Logistic regression and ROC analyses to
determine the best combination of miRNAs to predict lung cancer.
ROC curves were constructed to compare the relative expression of
the 10 miRNAs for lung cancer patients and controls that including PD
and HC, which have yielded the following AUCs: let-7f, 0.778(95% CI,
0.683–0.872); miR-126-3p, 0.792(95% CI, 0.696–0.888); miR-148b,
0.825(95%
CI, 0.742–0.908);
miR-151-5p,
0.784(95%
CI,
0.686–0.882); miR-199a-3p, 0.790(95% CI, 0.693–0.887); miR-221,
0.784(95% CI, 0.679–0.889); miR-23b, 0.725(95% CI, 0.624–0.825);
miR-26a, 0.792(95% CI, 0.703–0.881); miR-27b, 0.792(95% CI,
0.702–0.882); miR-423-3p,0.767 (95% CI, 0.676–0.868) (Fig. 2a). The
ROC area under the curve (AUC) analysis identified the statistically
significant 10 miRNAs for differentiation between the lung cancer and
the control groups (P < 0.05) and the expression of any single miRNA
may be used to predict lung cancer to a certain degree. Of the 10
miRNAs investigated, miR-148b displayed the highest AUC of 0.825 for
diagnosis of lung cancer.
Though a single miRNA could be used to distinguish lung cancer
from healthy individuals, a combination of several miRNAs might
provide a stronger differentiation power than individual miRNA. The
miRNA classifiers for defining lung cancer were demonstrated by logistic regression analysis. We found that a linear combination of the
expression levels of all the 10 miRNA produced the best model to
predict lung cancer. The AUC for the signature comprising the 10
miRNA was 0.892 (95% CI, 0.830–0.954), representing a significant
improvement in comparison to each single miRNA marker (Fig. 2b).
However, a combination of 10 miRNAs seemed to be complicated.
Then, we calculated the AUC of combination of other miRNAs to get a
more simplified combination of miRNAs for lung cancer prediction.
After the permutation and combination, we found that the AUC for the
signature comprising 4 miRNA (miR-23b, miR-221, miR-148b and miR423-3p) achieved 0.885 (95% CI, 0.824–0.947), which is the highest in
any other 4 combinations (Fig. 2c). It showed that the 4 miRNA
achieved a similar performance compared with all the 10 miR signature, indicating that the 4 miRNA signature might be an easier and
suitable model and could be a potential biomarker for detection of lung
cancer. Further investigations enrolling more samples are necessary to
get other miRNA signatures with better diagnostic values.
3.2. Expression profiling of microRNAs in lung cancer serum
To identify a diagnostic serum miRNA signature for lung cancer, a
real-time PCR- based high-throughput PCR array was used to compare
the serum miRNA expression profiles between 10 lung cancer patients
and 10 health controls. A miRNA was defined to be differentially expressed if the fold change in comparison to the control was > 30 and Pvalue < 0.05. As a result, a distinct miRNA signature including a set of
21 miRNAs that were differentially expressed in lung cancer patients
relative to healthy controls was identified, and all of them were upregulated (Table 2 and Fig. 1a).
3.3. Validation of miRNAs by quantitative PCR
To validate the screening results and confirm the pattern of miRNA
expression between the patient and the control samples, the expression
levels of the 21 candidate miRNAs were tested in an independent validation cohort on samples from patients with lung cancer (n = 40),
benign pulmonary diseases (n = 40), and healthy controls (n = 40). A
miRNA was defined to be differentially expressed if the fold-change in
comparison to the control was more than 4 that means the −DDct was
more than 2. Based on this definition, among the 21 candidate miRNAs,
10 serum miRNAs (let-7f, miR-126-3p, miR-148b, miR-151-5p, miR199a-3p, miR-221, miR-23b, miR-26a, miR-27b, and miR-423-3p) were
significantly up-regulated in the lung cancer group compared to the
benign pulmonary disease and healthy control. In addition, no significant difference in any of the 10 candidate miRNAs was observed
between the benign pulmonary disease and healthy control group
(Table 2). Hierarchical clustering, based on the selected 10 differentially expressed miRNAs, showed that the 10 serum miRNA signature
was able to distinguish lung cancer from control subjects (Fig. 1b). The
cancer cases included 9 SCLC and 34 NSCLC (9 SQ, 19 AD and 6 cases of
other pathological cancer types). Notably, no significant difference in
any of the 10 candidate miRNAs expression was found between the
3.5. Prediction of miRNA target genes associated with lung cancer
With the use of miRWalk (http://www.umm.uni-heidelberg.de/
apps/zmf/mirwalk/) that combines the output of multiple prediction
algorithms, a total of 3108 predicted targets of the 10 miRNAs classifier
Table 2
Top 10 differentially expressed miRNAs (lung cancer vs HC and PD) in the biomarker-validation phase and their respective fold-changes from qPCR expression profiling.
miRNA
DCt-LC Mean ± SD
DCt-HC Mean ± SD
DCt-PD Mean ± SD
FC (LC-HC)
FC (LC-PD)
FC (PD-HC)
P1 (LC-HC)
P2 (LC-PD)
P3 (PD-HC)
let-7f
miR-126-3p
miR-148b
miR-151-5p
miR-199a-3p
miR-221
miR-23b
miR-26a
miR-27b
miR-423-3p
0.26
0.06
1.65
2.50
2.99
4.87
4.05
1.66
2.76
3.13
3.76
2.84
4.74
4.92
6.05
7.41
6.07
4.82
5.26
5.70
2.75
3.20
4.79
5.60
6.29
7.89
6.31
4.82
5.37
5.37
11.31
6.89
8.53
5.33
8.36
5.80
4.06
8.90
5.66
5.95
5.64
8.84
8.82
8.59
9.83
8.10
4.79
8.90
6.10
4.72
2.00
0.78
0.97
0.62
0.85
0.72
0.85
1.00
0.93
1.26
7.00E-08
2.95E-06
3.46E-08
3.08E-05
1.96E-06
7.70E-06
9.76E-04
6.43E-07
2.22E-06
3.23E-06
2.12E-04
1.10E-06
3.40E-07
1.43E-06
1.19E-06
1.06E-06
1.20E-03
3.49E-06
6.72E-06
2.32E-04
5.65E-02
4.24E-01
9.11E-01
1.46E-01
6.29E-01
2.37E-01
6.51E-01
1.00E+00
8.11E-01
4.97E-01
±
±
±
±
±
±
±
±
±
±
3.33
3.19
2.84
3.06
3.44
3.09
3.41
3.21
2.56
2.7
±
±
±
±
±
±
±
±
±
±
2.13
1.72
1.58
1.80
2.02
1.47
1.88
2.19
1.72
1.88
±
±
±
±
±
±
±
±
±
±
2.61
2.19
2.23
2.37
2.30
1.97
2.89
2.67
2.22
2.54
LC: lung cancer. PD: Pulmonary disease. HC: Health control. P1: Comparison between LC and HC. P2: Comparison between LC and PD. P3: comparison between HC and PD. FC: Fold
change.
8
Lung Cancer 114 (2017) 6–11
Y. Zhu et al.
Fig. 1. (a) Hierarchical clustering analyses of 21
differentially expressed miRNAs (10 lung cancer vs
10 health control) in biomarker-screening phase. The
10 lung cancer group includes 4 SQ, 4 AD, and 2 S
cases. (b) Hierarchical clustering analyses of 10 differentially expressed miRNAs (40 LC vs control including 40 HC and 40 PD) in biomarker-validation
phase. Clustering indicates significant distinction
between LC and control group. (c) Hierarchical
clustering analyses of 10 differentially expressed
miRNAs in 34 lung cancer including 6 SCLC, 9 SQ
and 19 AD in biomarker-validation phase. Clustering
indicates no significant distinction between SCLC, SQ
and AD group.
SQ: Squamous carcinoma; AD: Adenocarcinoma; S:
Small cell lung cancer; HC: Health control; LC: lung
cancer. PD: Pulmonary disease.
Table 3
Ten differentially expressed miRNAs in 34 lung cancer including 6 SCLC, 9 SQ and 19 AD in biomarker-validation phase.
miRNA
DCt-AD mean ± sd
DCt-SQ mean ± sd
DCt-S mean ± sd
P1 (SQ-AD)
P2 (SQ-S)
P3 (AD-S)
let-7f
miR-126-3p
miR-148b
miR-151-5p
miR-199a-3p
miR-221
miR-23b
miR-26a
miR-27b
miR-423-3p
0.76
0.62
2.04
3.17
3.79
5.41
4.42
2.23
3.39
3.62
0.14 ± 3.8
−0.02 ± 3.33
1.37 ± 3
2.29 ± 3.24
2.86 ± 3.96
4.51 ± 2.69
4.12 ± 3.92
1.83 ± 3.77
2.66 ± 2.69
2.81 ± 2.93
−1.00 ± 5.31
−1.4 ± 5.15
1.31 ± 4.69
1.53 ± 5.09
1.62 ± 5.03
3.85 ± 4.11
3.33 ± 6.27
0.09 ± 4.97
1.48 ± 4.2
2.58 ± 3.89
0.965
0.938
0.913
0.849
0.891
0.806
0.995
0.986
0.843
0.849
0.955
0.913
1.000
0.982
0.939
0.979
0.989
0.848
0.899
0.999
0.817
0.740
0.975
0.829
0.690
0.774
0.963
0.684
0.648
0.897
±
±
±
±
±
±
±
±
±
±
2.34
2.42
2.5
2.35
2.64
2.84
2.14
2.17
1.77
2.37
SQ: Squamous carcinoma. AD: Adenocarcinoma. S: Small cell lung cancer. P1: Comparison between SQ and AD. P2: Comparison between SQ and S. P3: Comparison between AD and S.
Fig. 2. ROC curve analysis. (a): The 10 individual
miRNAs achieved AUC of 0.725–0.825;(b): The
classifier of all 10 miRNA yielded an AUC of 0.892
(95% CI: 0.830–0.954) with 77.5% sensitivity,
90.0% specificity and Youden's index 0.675. (c):The
classifier of 4 miRNA(including miR-23b, miR221,miR-148b and miR-423-3p) yielded an AUC of
0.885 (95% CI: 0.824–0.947) with 80.0% sensitivity,
82.5% specificity and Youden's index 0.625.
were identified. Then, we used DAVID to analyze the possible functions
of the targeted genes. The results of Gene Ontology (GO) classification
indicate that most of the common target genes were involved in transcription-related items, signal transduction, and cell movement. In addition, pathway enrichment analysis (Table 4) showed that the most
significant dysfunctional pathways were pathways in cancer (82 genes).
Other significant pathways included focal adhesion (53 genes), adherens junction (25 genes), etc. The results indicated that the 10 miRNA
classifiers may participate in the occurrence and development of lung
cancer through regulation of the series of pathways.
Table 4
The enriched KEGG pathway of predicted genes.
Term
Count
P Value
hsa05200:Pathways in cancer
hsa04510:Focal adhesion
hsa05014:Amyotrophic lateral sclerosis (ALS)
hsa04520:Adherens junction
hsa05222:Small cell lung cancer
hsa05215:Prostate cancer
hsa04144:Endocytosis
hsa05212:Pancreatic cancer
hsa04210:Apoptosis
hsa04350:TGF-beta signaling pathway
hsa05210:Colorectal cancer
hsa04512:ECM-receptor interaction
hsa05220:Chronic myeloid leukemia
hsa04910:Insulin signaling pathway
hsa04010:MAPK signaling pathway
82
53
20
25
26
27
46
22
25
25
24
24
22
34
59
1.77E-05
1.73E-04
3.25E-04
6.30E-04
1.04E-03
1.13E-03
1.68E-03
3.36E-03
3.99E-03
3.99E-03
5.27E-03
5.27E-03
5.68E-03
6.70E-03
7.14E-03
4. Discussion
In this study, we tried to find serum miRNAs as biomarkers that can
be used to distinguish lung cancer from healthy controls. A total of 21
miRNAs were retrieved in the screen process. Nevertheless, only 10
miRNAs (let-7f, miR-126-3p, miR-151-5p, miR-199a-3p, miR-221, miR23b, miR-26a, miR-27b, miR-423-3p and miR-148b) were identified to
be associated with lung cancer in the validation process. Interestingly,
we found that the combination of 4 miRNAs, miR-23b, miR-221, miR9
Lung Cancer 114 (2017) 6–11
Y. Zhu et al.
Several limitations might be included in the present study. Firstly, in
both the screening and validation cohort, the sample sizes are rather
small. The subjects involved in this study were mainly persons lived in
southwest area of China. Secondly, the confounding factors such as
smoking and drinking were not considered in this study. Thirdly, the
clinical factors such as TMN stages, age and sex were not evaluated.
Fourthly, the proportion of the early stage lung cancer cases was small
in this cohort, the results might be interpreted with care for early detection. However, the selected miRNA signature may be of great significance for lung cancer detection. To our knowledge, the present research for the first time identified miRNAs as biomarkers for lung
cancer diagnosis.
In summary, 10 miRNAs (let-7f, miR-126-3p, miR-148b, miR-1515p, miR-199a-3p, miR-221, miR-23b, miR-26a, miR-27b, and miR-4233p) could be used as biomarkers for discriminating lung cancer from
non-cancer individuals. A combination of 4 miRNAs (miR-23b, miR221, miR-148b and miR-423-3p) had high discriminatory power for
lung cancer cases and healthy controls. Further studies are needed to
confirm the conclusions.
148b and miR-423-3p, were proven to have the highest sensitivity and
specificity for predicting lung cancer via logistic regression and ROC
analyses.
Recent evidence has shown that dysregulated miRNAs might act as
prognostic factors for lung carcinoma [11]. MiRNAs, such as miR-137
[12] and miR-148b [13] might play critical roles in the development of
lung cancer. In the present study, after both the screen and validation
processes, 10 important miRNAs have been proven to be a biomarker
discriminating lung cancer from benign pulmonary disease patients and
healthy individuals. The 10 miRNAs were significantly over-expressed
in the serum of lung cancer patients relative to those of benign pulmonary disease patients and healthy controls, indicating that these
miRNAs might play a critical role in lung cancer genesis and progression. The upregulated miRNAs might regulate multiple target genes and
result in activation of oncogenes and suppression of cell apoptosis.
Generally speaking, miRNAs could induce mRNA cleavage and inhibition of translation by binding to the seed site of the 3′-UTR of target
gene mRNA [14]. The target gene could be easily predicted by bioinformation programs. In theory, a single miRNA could regulate multiple target genes, and conversely a single gene could be targeted by
several miRNAs. The results of the programs showed that the target
genes of each miRNA vary among different softwares. The findings of
the three programs indicated that the miRNA might interact with each
other in determining the balance between oncogenes and tumor suppression factors. Notably, real-time PCR analysis revealed that 10 elevated miRNAs were validated in lung cancer serum. It is worthy of note
that only upregulated miRNAs rather than downregulated ones were
identified in the present study. A probable reason for this deviancy
might be that activation of oncogenes may play a major role than inactivation of tumor suppression factors in the genesis of lung carcinoma.
When we combined the 10 miRNAs classifier for diagnosing lung
cancer, the diagnostic accuracy was improved substantially as shown by
ROC analysis. However, this combination is rather complex due to the
large number of miRNAs used for test. To simplify the steps, we found
that combination of 4 miRNAs could achieve the similar diagnostic
efficiency, with high specificity and sensitivity. Thus, these 4 miRNAs
seemed to associate with the genesis and development of lung cancer.
However, the precise mechanisms were not fully understood. MiR-23b
was shown to have a close association with cancer metastasis and
chemoresistance [15]. Moreover, it might directly regulate the functions of tumor suppressors Nischarin and Proline oxidase and contribute
to cancer growth [16,17]. MiR-148b might be involved in immune
homeostasis and immune regulation during cancer progression [18].
MiR-423-3p could downregulate the expression of a tumor suppressor
p21Cip1/Waf1 and promote cell cycle progression at G1/S transition
[19]. In the present study, the combination of the 4 miRNAs for diagnosing lung cancer provided the highest AUC of the ROC test, suggesting that they might interact with each other in the genesis of lung
cancer. The possible underlying mechanisms are not clear and future
studies on this issue are required.
Measuring the circulating miRNAs is a non-invasive method for
detection of lung cancer, which might reduce the false positive of CT
scan screening. The tumor-associated endogenous plasma miRNAs were
stable in a form that is resistant 4 °C–37 °C incubation and RNAase
activity possibly because they were packaged inside exosomes that are
secreted from cancer cells [20,21]. Therefore, circulating miRNAs
might be a potential biomarker for lung cancer diagnosis. Notably,
malignant cancer cells can selectively release specific miRNAs into
blood stream, causing levels of particular miRNAs to go up in blood
derivatives such as serum and go down in the tumor cells from which
they originate [22,23]. In addition, circulating miRNAs might represent
the by-products of dying/dead cancer cells and are mainly circulating
outside of exosomes in plasma [24]. Hence, the origins and sources of
miRNAs are not clear and the problems should be resolved in future
investigations.
Conflicts of interests
The authors declare that there are no conflicts of interest.
Acknowledgement
The present study was supported by the National Science
Foundation of China (No.81620108023 and No.31270845).
References
[1] Q.Y. Hong, G.M. Wu, G.S. Qian, C.P. Hu, J.Y. Zhou, L.A. Chen, W.M. Li, S.Y. Li,
K. Wang, Q. Wang, X.J. Zhang, J. Li, X. Gong, C.X. Bai, Prevention and management
of lung cancer in China, Cancer 121 (Suppl. 17) (2015) 3080–3088.
[2] H.Y. Deng, Y.C. Wang, P.Z. Ni, G. Li, X.Y. Yang, Y.D. Lin, L.X. Liu, Radiotherapy,
lobectomy or sublobar resection? A meta-analysis of the choices for treating stage I
non-small-cell lung cancer, Eur. J. Cardiothorac. Surg. 51 (2017) 203–210.
[3] E.F. Patz Jr., E. Greco, C. Gatsonis, P. Pinsky, B.S. Kramer, D.R. Aberle, Lung cancer
incidence and mortality in National lung screening trial participants who underwent low-dose CT prevalence screening: a retrospective cohort analysis of a randomised, multicentre, diagnostic screening trial, Lancet Oncol. 17 (2016) 590–599.
[4] B. Ortiz-Quintero, Cell-free microRNAs in blood and other body fluids, as cancer
biomarkers, Cell Prolif. 49 (2016) 281–303.
[5] T. Kawaguchi, S. Komatsu, D. Ichikawa, M. Tsujiura, H. Takeshita, S. Hirajima,
M. Miyamae, W. Okajima, T. Ohashi, T. Imamura, J. Kiuchi, H. Konishi, A. Shiozaki,
K. Okamoto, E. Otsuji, Circulating microRNAs: a next-generation clinical biomarker
for digestive system cancers, Int. J. Mol. Sci. 17 (2016).
[6] X. Yu, Z. Li, Serum microRNAs as potential noninvasive biomarkers for glioma,
Tumour Biol. 37 (2016) 1407–1410.
[7] C. Zhao, F. Lu, H. Chen, F. Zhao, Z. Zhu, X. Zhao, Clinical significance of circulating
miRNA detection in lung cancer, Med. Oncol. 33 (2016) 41.
[8] W. Sohn, J. Kim, S.H. Kang, S.R. Yang, J.Y. Cho, H.C. Cho, S.G. Shim, Y.H. Paik,
Serum exosomal microRNAs as novel biomarkers for hepatocellular carcinoma, Exp.
Mol. Med. 47 (2015) e184.
[9] S. Jiang, C. Li, V. Olive, E. Lykken, F. Feng, J. Sevilla, Y. Wan, L. He, Q.J. Li,
Molecular dissection of the miR-17-92 cluster's critical dual roles in promoting Th1
responses and preventing inducible treg differentiation, Blood 118 (2011)
5487–5497.
[10] L. Ma, X.Q. Zhang, D.X. Zhou, Y. Cui, L.L. Deng, T. Yang, Y. Shao, M. Ding,
Feasibility of urinary microRNA profiling detection in intrahepatic cholestasis of
pregnancy and its potential as a non-invasive biomarker, Sci. Rep. 6 (2016) 31535.
[11] Q. Liu, Z. Yu, S. Yuan, W. Xie, C. Li, Z. Hu, Y. Xiang, N. Wu, L. Wu, L. Bai, Y. Li,
Circulating exosomal microRNAs as prognostic biomarkers for non-small-cell lung
cancer, Oncotarget 8 (2017) 13048–13058.
[12] B. Zhang, Z. Ma, X. Li, C. Zhang, Y. Shao, Z. Liu, Y. Li, Y. Jin, Tanshinones suppress
non-small cell lung cancer through up-regulating miR-137, Acta Biochim. Biophys.
Sin. (Shanghai) 48 (2016) 768–770.
[13] R. Wang, F. Ye, Q. Zhen, T. Song, G. Tan, W. Chu, Y. Zhang, B. Lv, X. Zhao, J. Liu,
MicroRNA-148b is a potential prognostic biomarker and predictor of response to
radiotherapy in non-small-cell lung cancer, J. Physiol. Biochem. 72 (2016)
337–343.
[14] K. Felekkis, E. Touvana, C. Stefanou, C. Deltas, MicroRNAs: a newly described class
of encoded molecules that play a role in health and disease, Hippokratia 14 (2010)
236–240.
[15] Y.T. Park, J.Y. Jeong, M.J. Lee, K.I. Kim, T.H. Kim, Y.D. Kwon, C. Lee, O.J. Kim,
H.J. An, MicroRNAs overexpressed in ovarian ALDH1-positive cells are associated
10
Lung Cancer 114 (2017) 6–11
Y. Zhu et al.
microRNA: a diagnostic marker for lung cancer, Clin. Lung Cancer 10 (2009) 42–46.
[21] D. Zheng, S. Haddadin, Y. Wang, L.Q. Gu, M.C. Perry, C.E. Freter, M.X. Wang,
Plasma microRNAs as novel biomarkers for early detection of lung cancer, Int. J.
Clin. Exp. Pathol. 4 (2011) 575–586.
[22] K. Cuk, M. Zucknick, J. Heil, D. Madhavan, S. Schott, A. Turchinovich, D. Arlt,
M. Rath, C. Sohn, A. Benner, H. Junkermann, A. Schneeweiss, B. Burwinkel,
Circulating microRNAs in plasma as early detection markers for breast cancer, Int.
J. Cancer 132 (2013) 1602–1612.
[23] L. Pigati, S.C. Yaddanapudi, R. Iyengar, D.J. Kim, S.A. Hearn, D. Danforth,
M.L. Hastings, D.M. Duelli, Selective release of microRNA species from normal and
malignant mammary epithelial cells, PLoS One 5 (2010) e13515.
[24] A. Turchinovich, L. Weiz, A. Langheinz, B. Burwinkel, Characterization of extracellular circulating microRNA, Nucleic Acids Res. 39 (2011) 7223–7233.
with chemoresistance, J. Ovarian Res. 6 (2013) 18.
[16] L. Jin, O. Wessely, E.G. Marcusson, C. Ivan, G.A. Calin, S.K. Alahari, Prooncogenic
factors miR-23b and miR-27b are regulated by Her2/Neu, EGF, and TNF-alpha in
breast cancer, Cancer Res. 73 (2013) 2884–2896.
[17] W. Liu, O. Zabirnyk, H. Wang, Y.H. Shiao, M.L. Nickerson, S. Khalil, L.M. Anderson,
A.O. Perantoni, J.M. Phang, miR-23b targets proline oxidase, a novel tumor suppressor protein in renal cancer, Oncogene 29 (2010) 4914–4924.
[18] H. Chang, X. Zhou, Z.N. Wang, Y.X. Song, F. Zhao, P. Gao, Y. Chiang, H.M. Xu,
Increased expression of miR-148b in ovarian carcinoma and its clinical significance,
Mol. Med. Rep. 5 (2012) 1277–1280.
[19] J. Lin, S. Huang, S. Wu, J. Ding, Y. Zhao, L. Liang, Q. Tian, R. Zha, R. Zhan, X. He,
MicroRNA-423 promotes cell growth and regulates G(1)/S transition by targeting
p21Cip1/Waf1 in hepatocellular carcinoma, Carcinogenesis 32 (2011) 1641–1647.
[20] G. Rabinowits, C. Gercel-Taylor, J.M. Day, D.D. Taylor, G.H. Kloecker, Exosomal
11
Документ
Категория
Без категории
Просмотров
2
Размер файла
349 Кб
Теги
2017, lungcan, 002
1/--страниц
Пожаловаться на содержимое документа