4620 MOLECULAR MEDICINE REPORTS 13: 4620-4626, 2016

Circulating microRNAs as biomarkers for the early

diagnosis of childhood tuberculosis infection
MENGYAO ZHOU, GUANGYUAN YU, XIANTAO YANG, CHAOMIN ZHU, ZHENZHEN ZHANG and XUE ZHAN

The Children's Hospital of Chongqing Medical University,

Ministry of Education Key Laboratory of Child Development and Disorders, Key Laboratory of Pediatrics in Chongqing,
Chongqing International Science and Technology Cooperation Center for Child Development and Disorders,
Chongqing 400000, P.R. China

Received April 3, 2015; Accepted February 9, 2016

DOI: 10.3892/mmr.2016.5097

Abstract. MicroRNAs (miRNAs) are a class of highly miRNAs exhibited increased diagnostic value compared with
conserved, single‑stranded RNA molecules (length, 18‑25 nt) any single miRNA. To the best of our knowledge, the present
that regulate the expression of their target mRNAs. Previous study is the first to identify the expression profile of circulating
studies have demonstrated that miRNAs may be novel miRNAs in childhood TB and demonstrated that miRNAs may
biomarkers in the diagnosis of certain diseases. In order to be a novel, non‑invasive and effective biomarker for the early
evaluate the diagnostic value of miRNAs in childhood tubercu- diagnosis of childhood TB.
losis (TB), the circulating miRNA profile was determined using
microarray analysis. An miRNA‑gene network was constructed Introduction
to identify closely associated miRNAs and these miRNAs were
validated using reverse transcription‑quantitative polymerase Tuberculosis (TB) results in high rates of mortality and
chain reaction (RT‑qPCR). A receiver operational curve (ROC) morbidity globally. According to global TB reports in 2014,
was used to evaluate the diagnostic sensitivity and specificity an estimated 9.6 million individuals developed TB and
of confirmed miRNAs. The microarray data demonstrated that 1.5 million mortalities were caused by TB (1). Despite prog-
29 miRNAs were altered with 15 upregulated and 14 down- ress being made on TB, childhood TB remains an epidemic
regulated. The network showed indicated 14  miRNAs that in numerous developing countries. There were an estimated
are critical in childhood TB. RT‑qPCR validated that miR‑1, 53,000 TB cases among children (<15 years of age) and
miR‑155, miR‑31, miR‑146a, miR‑10a, miR‑125b and miR‑150 74,000 TB‑associated mortalities in 2012, which account for
were downregulated in while miR‑29 was upregulated in chil- the 6 and 8% of the global totals, respectively. It has been
dren with TB compared with uninfected children. The ROC reported that there was no marked decrease in the prevalence
curve data indicated the diagnostic value of single miRNA was of childhood TB from 1979 to 2000 in China, and patients
as follows: miR‑150>miR‑146a>miR‑125b>miR‑31>miR‑10a with bacteriologically‑negative TB remain a large proportion
>miR‑1>miR‑155>miR‑29. Notably, a combination of these of TB patients (2,3). There remains a long way to go for China
to achieve their 2015 targets of eliminating TB. From 2016, the
aim is to end the global TB epidemic by implementing the End
TB Strategy. Adopted by the World Health Assembly in May
Correspondence to: Dr Zhenzhen Zhang or Professor Xue Zhan, 2014, the strategy serves to reduce the number of TB mortali-
The Children's Hospital of Chongqing Medical University, ties by 90% by 2030 (compared with 2015 levels), decrease
Ministry of Education Key Laboratory of Child Development and the number of new cases by 80% and ensure that no family is
Disorders, Key Laboratory of Pediatrics in Chongqing, Chongqing financially burdened due to TB (1). To achieve this goal, novel
International Science and Technology Cooperation Center for and effective methods for the early development of childhood
Child Development and Disorders, 136 Zhongshan Second Street, TB are required.
Yuzhong, Chongqing 400000, P.R. China
MicroRNA (miRNA) is a class of highly conserved,
E‑mail: 39003614@qq.com
single‑stranded RNA molecules (length, 18‑25 nt) that regu-
E‑mail: zhanxue@cqmu.edu.cn
late expression levels of their target mRNAs (4,5). Emerging
Abbreviations: miRNA, microRNA; TB, tuberculosis; MTB, research has demonstrated that serum miRNA is stable to
Mycobacterium tuberculosis; ROC, receiver operational curve; AUC, repeated freezing and thawing as well as heat, acidic and
area under the curve; CI, confidence interval alkaline conditions and other extremes. It may have poten-
tial as a useful biomarker for disease diagnosis, the effects
Key words: childhood tuberculosis, microRNA, microarray, of therapeutics and prognosis (6‑10). Previous studies have
diagnosis, biomarker demonstrated that miRNA is associated with numerous
diseases, including cancer and heart, immune and infec-
tious diseases (11‑14). Altered levels of miRNAs following
 ZHOU et al: CIRCULATING microRNAs TO DIAGNOSE CHILDHOOD TUBERCULOSIS 4621

Mycobacterium tuberculosis (MTB) infection have been Table I. Clinical characteristics of childhood TB.
reported, however, these studies focused on adults rather than
children (15‑18). Therefore, the present study aimed to identify Culture‑positive Culture‑negative
and validate the altered levels of circulating miRNAs in child- Characteristic TB TB
hood TB. Furthermore, the present study aimed to determine
the diagnostic value of single and combined miRNAs in child- Total 14 14
hood TB. Age (years, 8.4±5.7 6.7±4.9
mean ± SD)
Materials and methods Gender (male/female) 11/3 5/9
History of close 7 4
Ethics statement. The present study was reviewed and TB contact
approved by the ethics committee of Chongqing Medical TST positive 4 10
University (Chongqing, China). Written informed consent was Cough for >2 weeks 8 5
obtained from participants' parents at the time of enrollment Fever for >2 weeks 8 10
in the present study. Night sweats for 0 0
>2 weeks
Diagnostic process. Children with the following clinical Weight loss/failure 0 2
manifestations were enrolled as suspected TB cases: i) Cough, to thrive
fever and weight loss lasting >2  weeks; ii)  pneumonia not
Radiographic features 14 14
responding to antibiotics; iii) and other clinical findings,
of TB
including hydrothorax, tuberculin test (+), interferon‑γ release
Culture‑positive TB 14 0
assay (+), anti‑tuberculosis antibody (+), PCR (+) and chest
radiography suggestive of TB. Preliminary screening was HIV‑positive 0 0
conducted according to previously described guidelines (19). HBV‑positive 0 0
Cases that failed to meet the criteria, had pulmonary infec-
TB, tuberculosis; TST, tuberculin skin test; SD, standard deviation;
tion other than TB or had other preliminary diagnoses were
HIV, human immunodeficiency virus; HBV, hepatitis B virus.
excluded. The patients' sputum and stomach lavage fluid were
collected to isolate MTB using Lowenstein‑Jensen medium
and an acid‑fast bacillus test. Cases with successful isolation
of MTB were regarded as culture‑positive tuberculosis, while
cases with no successful isolation of MTB were regarded as at room temperature for 5 min. The supernatant was discarded
culture‑negative tuberculosis. Healthy children were selected and the sediment was white blood cells.
from a child care center at the Children's Hospital of Chongqing
Medical University (Chongqing, China) between February 2012 RNA extraction. Total RNA was isolated from peripheral
and July 2013. Written informed consent was obtained from white blood cells using TRIzol (Invitrogen; Thermo Fisher
the patient's family. These children did not recently suffer from Scientific, Inc., Waltham, MA, USA) according to the manu-
any infectious diseases and did not have any immunodeficiency facturer's protocols. Each sample was dissolved in 50  µl
diseases. RNAase‑free water. RNA quantity and purity was measured
using a NanoDrop spectrophotometer (NANODROP 1000;
Sample collection and handing. A total of 129 children were Thermo Fisher Scientific, Inc.). In solution, the high quantity
recruited from the Children's Hospital of Chongqing Medical RNA had an A260/A280 ratio of 1.8‑2.1. The total RNA
University (Chongqing, China). Following preliminary concentration was 100‑400 ng/µl.
screening, 28 patients and 24 healthy children were included
in the present study. Three active pulmonary TB cases Analysis of miRNA microarray data. miRNA microarray
(confirmed by MTB culture) and three healthy samples were assays were performed using the Agilent Human miRNA
used for microarray detection. The remaining 25 patients and microarray platform (Agilent Technologies, Inc., Santa Clara,
21 healthy children had samples taken for use in reverse tran- CA, USA). Labeling and hybridization were performed
scription‑quantitative polymerase chain reaction validation. according to the protocols in the Agilent miRNA microarray
There was no significant difference in age or gender between system. The microarray was scanned by Agilent miRNA
the two groups. The demographic and clinical characteristics Scanner and the microarray data was normalized using
of patients are summarized in Table I. Fresh blood samples R‑language programming (http://www.r‑project.org). The
were collected from the participants in 2.0 ml anticoagulant normalized data was analyzed according to log 2‑transformed
tubes. Within 4 h of collection, red blood cell lysis buffers expression level and the expression levels were tested by
produced in our laboratory were used to collect peripheral Student's t‑test and the Wilcoxon signed rank sum test. The
white blood cells for total RNA extraction. A total of 3 ml miRNAs that indicated significantly different expression in
1X red blood cell lysis buffer per ml of blood was added and the two tests and had at least a 2‑fold changes were filtered. An
mixed sufficiently. Centrifugation was then performed at miRNA may target different genes and a gene may be targeted
6,000 x g at room temperature for 8‑10 min and the superna- by different miRNAs. In order to understand the complicated
tant was discarded. Subsequently, 1 ml 1X red blood cell lysis associations between miRNAs and target genes, Targetscan
buffer per tube was added, mixed and centrifuged at 6,000 x g (http://genes.mit.edu/targetscan/index.html) and Pictar
4622 MOLECULAR MEDICINE REPORTS 13: 4620-4626, 2016

Table II. FC of miRs between children with tuberculosis and

healthy controls.

ID logFC

miR‑142‑5p 1.165432929
miR‑29b 1.289097829
miR‑21* 1.67673294
miR‑542‑5p 1.812354699
miR‑32 1.86960397
miR‑142‑3p 1.887657013
miR‑95 2.410210413
miR‑144 2.099441438
miR‑17* 2.128855333
miR‑141 2.278684837
miR‑33a 2.372570115
miR‑136 2.571642478
miR‑324‑5p 2.845130237
miR‑193a‑3p 3.94172976
miR‑503 4.518278633
miR‑31 ‑5.138973159
miR‑564 ‑3.577222134
miR‑1 ‑2.229848349
miR‑181a‑2* ‑2.048150209
miR‑874 ‑1.99302363
miR‑1305 ‑1.894597861
miR‑10a ‑1.824343578
miR‑1275 ‑1.756667644
Figure 1. Hierarchical clustering of circulating miRs from childhood TB and
miR‑125b ‑1.675341028 control groups. Peripheral white blood cells were clustered according to the
miR‑342‑5p ‑1.637504768 expression profiles of 29 differentially expressed miRs between children with
TB and healthy controls. The analyzed samples and the miRs were presented
miR‑150 ‑1.370167992
in rows. The miRNA clustering tree is shown on the right, and the sample
miR‑155 ‑1.282465779 clustering tree is presented at the top. The color scale in the right top corner
miR‑342‑3p ‑1.131121391 indicates the relative expression level of the miRs, with red representing a
high expression level and green representing a low expression level. miR,
miR‑146a ‑1.00258041 microRNA; TB, tuberculosis; N, healthy children; P, TB patients.

FC, fold change; miR, microRNA.

5 times for Real‑Time PCR reaction. The total volume of the

reaction mixture was 20 µl, including 10 µl of 2X All‑In‑One
(http//pictar.bio.nyu.edu) databases were used to predict the qPCR mix, 2  µl All‑In‑One qPCR Primer (2  µM), 2  µl
target genes. The miRNA‑gene network was built to observe Universal Adaptor PCR Primer (2 µM), 2 µl cDNA and 4 µl
closely associated miRNAs following TB infection using dH2O. The PCR was conducted with incubation at 95˚C for
Cytoscape (version 3.0.1; http://www.cytoscape.org/). 10 min, and 40 cycles of 95˚C for 10 sec, 55˚C for 20 sec and
72˚C for 10 sec using a CFX96 Real‑Time system (Bio‑Rad
Reverse transcription‑quantitative polymerase chain reaction Laboratories, Inc., Hercules, CA, USA). Each sample was run
(RT‑qPCR). Total RNA (500 ng) was isolated and miRNA was in triplicate. For the RT‑qPCR data, the relative expression
obtained by poly (A) tail method. miRNAs in the center of level of each miRNA was calculated according to the different
the network were selected for validation using a SYBR green cycle quantification (Cq) values between the target miRNAs
RT‑qPCR assay (All‑In‑One miRNA qRT‑PCR Detection kit, and miR‑U6 by using the 2‑ΔΔCq method (20).
GeneCopoeia, Inc., Rockville, MD, USA). The miRNA primers
were synthesized by GeneCopoeia, Inc. and the expression Statistical analysis. Statistical analysis was performed with
level was normalized to U6. The miRNA was polyadenylated SPSS software (version 21; IBM SPSS, Armonk, NY, USA).
by poly(A) polymerase and reverse transcribed to cDNA under The data are presented as the mean ± standard deviation. An
conditions of 37˚C for 60 min followed by 85˚C for 5 min. The unpaired t‑test was used for statistical analysis and P<0.05
RT reaction was conducted in a reaction volume of 10 µl, which was considered to indicate a statistically significant difference.
contained 2.5 U/µl PolyA polymerase (0.4 µl), 0.4 µl RTase Receiver operating characteristic (ROC) curve was generated to
mix, 2 µl 5X Reaction Buffer, 500 ng total RNA and variable evaluate the diagnostic value of miRNAs. The area under the
volume of RNase Free dH2O. Synthesized cDNA was diluted curve (AUC) and 95% confidence intervals (CI) were calculated
 ZHOU et al: CIRCULATING microRNAs TO DIAGNOSE CHILDHOOD TUBERCULOSIS 4623

Figure 2. miR‑gene network of childhood tuberculosis. The different target genes are shown in blue and the different miRs are shown by the triangles. The
associations between miRs and genes are presented using straight lines. All miRs with differential expression and their target genes were analyzed, the
remaining 15 miRs were distributed sporadically and, thus, excluded from the network. miR, microRNA.

to determine the specificity and sensitivity. To increase the miRNAs that are closely associated with childhood TB, an
diagnostic accuracy, a combination of circulating miRNAs was miRNA‑gene network was constructed. miRNAs and their
analyzed. The ROC curve and multivariate logistic regression in target genes have a complex association as miRNAs may target
the ROC curves were calculated with MedCalc 11.4.3.0 statis- genes directly or indirectly. The miRNAs and genes closest
tical software (MedCalc Software bvba, Ostend, Belgium). to the center are more likely to have altered expression levels.
The network demonstrated miR‑1, miR‑155, miR‑31, miR‑146a,
Results miR‑10a, miR‑125b, miR‑150, miR‑29b, miR‑141, miR‑17*,
miR‑32, miR‑33a, miR‑503 and miR‑144 were in the center of
Identification of circulating miRNAs from children with the network (Fig. 2).
TB. Following preliminary screening, a total of 29 miRNAs
were altered between the TB group and the control. Of these, Validation of the altered miRNAs in independent samples.
15 miRNAs were upregulated and 14 were downregulated when Fourteen candidate miRNAs were selected for further verifi-
compared with the healthy control group (Table II). Cluster cation in 25 children with TB and 21 healthy controls. When
analysis based on differentially expressed miRNAs indicated compared with healthy controls, miR‑1, miR‑155, miR‑31,
marked distinctions between the two groups (Fig. 1). miR‑146a, miR‑10a, miR‑125b and miR‑150 were downregu-
lated while miR‑29 was upregulated in the TB cases (Fig. 3).
Construction of the miRNA‑gene network. To investigate
whether the unique miRNA profile and their target genes are Circulating miRNA to diagnose childhood TB. ROC was
associated with the progression of TB infection, and to identify performed to evaluate the potential of microRNA as a
4624 MOLECULAR MEDICINE REPORTS 13: 4620-4626, 2016

Figure 3. Validation of miRs with different expression levels in children with tuberculosis and healthy controls. miRs in the center of the network were selected
for validation. The 2‑ΔΔCq method was used to normalize the relative gene expression data in the reverse transcription‑quantitative polymerase chain reaction
assay and U6 served as the reference gene. **P<0.01; *P<0.05. miR, microRNA.

Figure 4. ROC curve analysis of single and combined miRs. AUC, area under curve; ROC, receiver operational curve; miR, microRNA.
 ZHOU et al: CIRCULATING microRNAs TO DIAGNOSE CHILDHOOD TUBERCULOSIS 4625

biomarker to diagnose childhood TB. The data demonstrated miR‑342‑5p, miR‑193a‑3p, miR‑10 and miR‑33a. This may
the diagnostic value of single miRNA was as follows: miR‑1 be a result of different experimental protocols, experimental
50>miR‑146a>miR‑125b>miR‑31>miR‑10a>miR‑1>miR‑155 conditions and samples.
>miR‑29. The data demonstrated miR‑1, miR‑10a, miR‑125b The miRNA‑gene network was constructed to observe
miR‑146a, miR‑150, miR‑155 and miR‑31 exhibited a reli- key miRNAs. The further validation data demonstrated eight
able discrimination yielding AUC values of 0.926 (95% CI, of the miRNAs in the center of the network had significantly
0.809‑0.982), 0.950 (95% CI, 0.843‑0.993), 0.962 (95% CI, different expression levels. Of these miRNAs, miR‑1, miR‑155,
0.860‑0.996), 0.964 (95% CI, 0.862‑0.997), 0.989 (95% CI, miR‑31, miR‑146a, miR‑10a, miR‑125b, miR‑150, miR‑141,
0.902‑1.0), 0.918 (95% CI, 0.799‑0.978) and 0.952 (95% CI, miR‑144, miR‑17*, miR‑32 and miR‑503 were downregulated
0.844‑0.994) respectively. Using an optical cutoff (maximum while miR‑29 was upregulated in the TB group. The diag-
Youden index; sensitivity +, specificity ‑1), the sensitivity of nostic values of these miRNAs were analyzed using ROC
the seven miRNAs was 76, 76, 100, 100, 100, 96 and 95.8% curves. The data demonstrated each of the miRNAs, excluding
respectively and the specificity was 100, 100, 81, 81, 90.5, 85.7 miR‑29b, had a reliable diagnostic value and they all exhibited
and 90.5% respectively. The miR‑29b had a lower discrimina- moderate specificity and sensitivity. Furthermore, the ROC
tion yielding an AUC value of 0.697 (95% CI, 0.544‑9.824), curve data indicated the diagnostic value of single miRNA
with optical cutoff, the sensitivity and specificity were 56 and was as follows: miR‑150>miR‑146a>miR‑125b>miR‑31>
90.5% respectively. Logistic regression analysis with the ROC miR‑10a>miR‑1>miR‑155>miR‑29.
curves was used to identify combined miRNAs, which may aid However, a combination of the eight miRNAs demonstrated
in the diagnosis of childhood TB, the result indicated marked increased diagnostic value with an AUC of 0.996 (95% CI,
separation between TB and healthy groups with an AUC value 0.914‑1.0), the sensitivity and specificity were 95.8 and 100%,
of 0.996 (95% CI, 0.914‑1.0), the sensitivity was 95.8% and the respectively. It indicated combined identification of miR‑1,
specificity was 100% (Fig. 4). miR‑155, miR‑31, miR‑146a, miR‑10a, miR‑125b, miR‑150
and miR‑29 may be a novel early diagnostic biomarker.
Discussion In conclusion, the present study is the first, to the best of
our knowledge, to analyze the circulating miRNA profile in
Childhood TB cases are currently diagnosed by history of childhood TB. Results from the current study suggest multiple
close contact with a patient with TB, clinical manifestations, miRNAs may be suitable to serve as potential biomarkers for
chest radiography, tuberculin test, interferon‑γ release assay, the diagnosis of childhood TB.
anti‑tuberculosis antibody or PCR. However, children typically
lack clinical symptoms, with >50% children asymptomatic at Acknowledgements
early stages of the disease (21,22). The gold standard to diag-
nose TB is microbiological isolation of MTB; however, the The authors would like to thank all the patients and healthy
load of MTB is often low in tested fluids, including sputum, children who participated in the present study at the Children's
stomach lavage fluid and bronchoalveolar lavage fluid. In addi- Hospital of Chongqing. The present study was supported
tion, the culture of clinical specimens requires long incubation by NSFC (grant no. 81071406) and the Medical Scientific
times (4‑8 weeks), by which point it is too late to influence Research of Chongqing (grant no. 20142046).
clinical strategies of treatment and the success rate of the
treatment is low, <30‑40% (17,23,24). Identification of a novel References
biomarker to allow early diagnosis of childhood tuberculosis
is required.  1. World Health Organization: Global Tuberculosis Report 2015.
20th edition. World Health Organization, 2015.
There are numerous studies that describe the miRNA profile  2. Li L: The national epidemiological sampling survey analysis
following TB infection and miR‑155, miR‑155, miR‑200C, of childhood tuberculosis from 1979 to 2000. Chinese Medical
miR‑193a‑3p, miR‑595, miR‑432, miR‑9, miR‑582‑5p, miR‑144 Journal 80: 1678, 2004 (In Chinese).
 3. World Health Organisation: Global tuberculosis report 2013,
and miR‑29b have been validated in other studies (15,18,25‑28). 2013.
The microarray data from the present study is largely incon-  4. Krishnarao A (ed): MicroRNAs ‑ From Basic Science to Disease
sistent with previous studies (17,27,29,30). For example, data Biology. Cambridge University Press, Cambridge, UK, 2010.
 5. He L and Hannon GJ: MicroRNAs: Small RNAs with a big role
from the present study indicated that miR‑155 was downregu- in gene regulation. Nat Rev Genet 5: 522‑531, 2004.
lated in the TB group, however, Wu et al (30) demonstrated  6. Chim SS, Shing TK, Hung EC, Leung TY, Lau TK, Chiu RW and
that miR‑155 was upregulated in the purified protein deriva- Lo YM: Detection and characterization of placental microRNAs
in maternal plasma. Clin Chem 54: 482‑490, 2008.
tive challenged peripheral blood mononuclear cells of patients  7. Alevizos I and Illei GG: MicroRNAs as biomarkers in rheumatic
with active TB. Consistent with previous studies, miR‑141, diseases. Nat Rev Rheumatol 6: 391‑398, 2010.
miR‑32 and miR‑29b were overexpressed in the TB group  8. Gilad  S, Meiri  E, Yogev  Y, Benjamin  S, Lebanony  D,
Yerushalmi N, Benjamin H, Kushnir M, Cholakh H, Melamed N,
of the present study. The expression level of miR‑144 has et al: Serum microRNAs are promising novel biomarkers. PloS
varied in previous studies, Wang et al (27) suggested miR‑144 One 3: e3148, 2008.
is upregulated in TB patients, while Wu et al (30) observed  9. Chen X, Hu Z, Wang W, Ba Y, Ma L, Zhang C, Wang C, Ren Z,
Zhao Y, Wu S, et al: Identification of ten serum microRNAs from
downregulation. In the present study, no significant difference a genome‑wide serum microRNA expression profile as novel
was observed in the miRNA between TB children and healthy noninvasive biomarkers for nonsmall cell lung cancer diagnosis.
children. Furthermore, microarray data from the present Int J Cancer 130: 1620‑1628, 2012.
10. Kosaka N, Iguchi H and Ochiya T: Circulating microRNA in
study has suggested a difference in expression in a number of body fluid: A new potential biomarker for cancer diagnosis and
miRNAs that have not yet been reported, including miR‑31, prognosis. Cancer Sci 101: 2087‑2092, 2010.
4626 MOLECULAR MEDICINE REPORTS 13: 4620-4626, 2016

11. O'Connell  RM, Kahn  D, Gibson  WS, Round  JL, Scholz  RL, 21. Anderson ST, Kaforou M, Brent AJ, Wright VJ, Banwell CM,
Chaudhuri AA, Kahn ME, Rao DS and Baltimore D: Chagaluka G, Crampin AC, Dock rell HM, French N,
MicroRNA‑155 promotes autoimmune inf lammation by Hamilton MS, et al: Diagnosis of childhood tuberculosis and
enhancing inflammatory T cell development. Immunity 33: host RNA expression in Africa. N Engl J Med 370: 1712‑1723,
607‑619, 2010. 2014.
12. Ura S, Honda M, Yamashita T, Ueda T, Takatori H, Nishino R, 22. Eamranond  P and Jaramillo  E: Tuberculosis in children:
Sunakozaka H, Sakai Y, Horimoto K and Kaneko S: Differential Reassessing the need for improved diagnosis in global control
microRNA expression between hepatitis B and hepatitis C leading strategies. Int J Tuberc Lung Dis 5: 594‑603, 2001.
disease progression to hepatocellular carcinoma. Hepatology 49: 23. Zar HJ, Connell TG and Nicol M: Diagnosis of pulmonary
1098‑1112, 2009. tuberculosis in children: New advances. Expert Rev Anti
13. Gidlöf O and Erlinge D: MicroRNAs in the failing heart‑novel Infect Ther 8: 277‑288, 2010.
therapeutic targets? Scand Cardiovasc J 48: 328‑334, 2014. 24. Wallis  RS, Pai  M, Menzies  D, Doherty TM, Walzl G,
14. Ji J, Shi J, Budhu A, Yu Z, Forgues M, Roessler S, Ambs S, Perkins MD and Zumla A: Biomarkers and diagnostics for
Chen Y, Meltzer PS, Croce CM, et al: MicroRNA expression, tuberculosis: Progress, needs and translation into practice.
survival and response to interferon in liver cancer. N Engl J Lancet 375: 1920‑1937, 2010.
Med 361: 1437‑1447, 2009. 25. Maertzdorf  J, Weiner  J III, Mollenkopf  HJ; TBornotTB
15. Fu Y, Yi Z, Wu X, Li J and Xu F: Circulating microRNAs in Network, Bauer  T, Prasse  A, Müller‑Quernheim  J and
patients with active pulmonary tuberculosis. J Clin Microbiol 49: Kaufmann SH: Common patter ns and disease‑related
4246‑4251, 2011. signatures in tuberculosis and sarcoidosis. Proc Natl Acad Sci
16. Spinelli SV, Diaz A, D'Attilio L, Marchesini MM, Bogue C, USA 109: 7853‑7858, 2012.
Bay ML and Bottasso OA: Altered microRNA expression levels 26. Qi  Y, Cui  L, Ge  Y, Shi Z, Zhao K, Guo X, Yang D, Yu H,
in mononuclear cells of patients with pulmonary and pleural Cui L, Shan Y, et al: Altered serum microRNAs as biomarkers
tuberculosis and their relation with components of the immune for the early diagnosis of pulmonary tuberculosis infection.
response. Mol Immunol 53: 265‑269, 2013. BMC Infect Dis 12: 384‑393, 2012.
17. Zhang X, Guo J, Fan S, Li Y, Wei L, Yang X, Jiang T, Chen Z, 27. Wang C, Yang S, Sun G, Tang X, Lu S, Neyrolles O and Gao Q:
Wang C, Liu J, et al: Screening and identification of six serum Comparative miRNA expression profiles in individuals with
microRNAs as novel potential combination biomarkers for latent and active tuberculosis. PloS One 6: e25832, 2011.
pulmonary tuberculosis diagnosis. PloS One 8: e81076, 2013. 28. Yi  Z, Fu  Y, Ji  R, Li R and Guan Z: Altered microRNA
18. Abd‑El‑Fattah AA, Sadik NA, Shaker OG and Aboulftouh ML: signatures in sputum of patients with active pulmonary tuber-
Differential microRNAs expression in serum of patients with culosis. PloS One 7: e43184, 2012.
lung cancer, pulmonary tuberculosis and pneumonia. Cell 29. Liu  Y, Wang  X, Jiang  J, Cao Z, Yang B and Cheng X:
Biochem Biophys 67: 875‑884, 2013. Modulation of T cell cytokine production by miR‑144* with
19. Subspecialty Group of Respiratory Diseases; Society of elevated expression in patients with pulmonary tuberculosis.
Pediatrics, Chinese Medical Association; Editorial Board, Mol Immunol 48: 1084‑1090, 2011.
Chinese Journal of Pediatrics: Diagnostic standards and thera- 30. Wu  J, Lu  C, Diao  N, Zhang S, Wang S, Wang F, Gao Y,
peutic recommendations for pulmonary tuberculosis in children. Chen J, Shao L, Lu J, et al: Analysis of microRNA expression
Zhonghua Er Ke Za Zhi 44: 249‑251, 2006 (In Chinese). profiling identifies miR‑155 and miR‑155* as potential diag-
20. Schmittgen TD and Livak KJ: Analyzing real‑time PCR data by nostic markers for active tuberculosis: A preliminary study.
the comparative C(T) method. Nat Protoc 3: 1101‑1108, 2008. Hum Immunol 73: 31‑37, 2012.