Pancreatology 14 (2014) 159e166
Contents lists available at ScienceDirect
Pancreatology journal homepage: www.elsevier.com/locate/pan
Original article
Identification of serum microRNAs as diagnostic and prognostic biomarkers for acute pancreatitis Pi Liu a,1, Liang Xia a,1, Wei-long Zhang b, Hua-jing Ke a, Tao Su a, Li-bing Deng b, You-xiang Chen a, Nong-hua Lv a, * a b
Department of Gastroenterology, The First Affiliated Hospital of Nanchang University, Nanchang 330006, China Nanchang University, Nanchang 330031, China
a r t i c l e i n f o
a b s t r a c t
Available online 28 March 2014
Background/objectives: To identify serum microRNA (miRNA) as diagnostic and prognostic biomarkers for acute pancreatitis (AP). Materials and methods: Sera microRNA expression was profiled from 12 AP patients with varying disease severity and three healthy controls. Differentially expressed miRNAs were validated in a larger cohort of patients and controls. The diagnostic and prognostic potentials of differentially expressed miRNAs were evaluated using receiver operating characteristic (ROC) curve analysis and compared to that of classic prognostic markers for AP. Results: miRNA microarray analyses identified 205 differentially expressed miRNAs between sera from AP patients and that from controls. Nine miRNAs were differentially expressed between severe and mild AP patients. Further validation confirmed the down-regulation of miR-92b, miR-10a, and miR-7 in AP patients, and ROC analysis revealed that these miRNAs can differentiate AP from health cases. Furthermore, the serum miR-551b-5p level was significantly higher in patients with disease complications or a low plasma calcium level. ROC analysis showed that the serum miR-551b-5p level can distinguish between severe and mild AP. Conclusion: The expressions of miR-92b, miR-10a, and miR-7 in AP might be used for the early diagnosis of AP and miR-551b-5p may be used for predicting AP severity. Copyright Ó 2014, IAP and EPC. Published by Elsevier India, a division of Reed Elsevier India Pvt. Ltd. All rights reserved.
Keywords: microRNA Acute pancreatitis Early diagnosis Prognosis Biomarker Prediction
1. Introduction Acute pancreatitis (AP) is a reversible inflammatory process of the pancreas with a wide range of clinical variations. Mild AP (MAP) is often self-limited and has a very low mortality rate ( 2, and (2) p value < 0.01. Hierarchical clustering of differentially expressed miRNAs was analyzed as a heat map, using MEV (version 4.6). Furthermore, RT-qPCR data were analyzed using the comparative CT method. DCT for each target miRNA was defined as CTtarget miRNA e CT miR-16. Two-sided, two-tailed Student t-tests were performed to determine the statistical significance (p < 0.05) of differences between groups. ROC curve analysis was performed using GraphPad Prism version 5.
2.4. miRNA microarray 3. Results Total serum miRNAs including six SAP, six MAP, and three control cases were used for miRNA profiling. The miRCURYÔHy3Ô/ Hy5Ô Power labeling kit (Exiqon, Vedbaek, Denmark) was used for miRNA labeling according to the manufacturer’s instructions. The labeled miRNA samples were hybridized onto the miRCURYÔ LNA Array (v. 16.0, Exqo). The complete procedures of hybridization, incubation, and washing were conducted according to the manufacturer’s instructions. The resulting signals were scanned using an Axon GenePix 4000B microarray scanner. 2.5. miRNA quantification by RT-qPCR Total serum miRNAs (15 ng) isolated from the validation cohort (18 controls and 62 patients) were reverse transcribed using the One Step PrimeScript miRNA cDNA Synthesis Kit (Takara Biotechnology, D350A) with a universal adaptor primer. The RT reaction was performed at 37 C for 60 min, followed by 85 C for 5 s. The synthesized cDNA was then amplified by quantitative PCR, using SYBR Premix EX Taq II (Takara Biotechnology, DRR081A), Uni-miRQ PCR Primer (Takara Biotechnology), and primers specific to each individual miRNA (Table 1). The reactions were conducted using PRISM 7500 System (Applied Biosystems, USA). Reactions with no template were used as negative controls, and has-miR-16 was used as internal control [17]. The reaction conditions were 95 C for 30 s followed by 40 cycles of 95 C for 15 s and 60 C for 60 s (ramping was set to 1.6 C/sec). 2.6. Statistical analysis The miRNA array results were extracted and analyzed using GenePix Pro 6.0 software (Axon). The miRNA expression signals were normalized using median normalization. miRNAs with signal values higher than 50 in all groups were included in further analyses. The signals from multiple probes for each miRNA were averaged to represent the expression of the miRNA. Two-tailed ManneWhitney unpaired test was used to determine the
Table 1 miRNA-specific primers. 0
0
ID
miRNA
Sequence (from 5 to 3 )
Number of nucleotides
01 02 03 04 05 06 07 08 09 10 11 12
hsa-miR-92b hsa-miR-10a ebv-miR-BART6-3p hsa-miR-132* hsa-miR-146b-5p hsa-miR-30c hsa-miR-7 hsa-miR-890 hsa-miR-551b-5p hsa-miR-514b-3p hsa-miR-214 hsa-miR-16
AGGGACGGGACGCGGTG GTACCCTGTAGATCCGAATTTGTG CGGGGATCGGACTAGCCTTAGA ACCGTGGCTTTCGATTGTTAC GGCTGAGAACTGAATTCCATAGGC CGCTGTAAACATCCTACACTCTCAG CGGTGGAAGACTAGTGATTTTGTTG CGTACTTGGAAAGGCATCAGTTG GAAATCAAGCGTGGGTGAGACC CCATTGACACCTCTGTGAGTGGA TGCCTGTCTACACTTGCTGTGC TAGCAGCACGTAAATATTGGCG
17 24 22 21 24 25 25 23 22 23 22 22
3.1. Serum miRNA expression profiling in AP To identify differentially expressed miRNAs in blood from AP patients, we performed miRNA microarray analyses on serum miRNA samples from 3 healthy control and 12 AP patients (6 SAP and 6 MAP). All blood samples were collected within the first 24 h after hospital admission. MiRNAs with expression fold change greater than 2 with a p value less than 0.01 were considered as differentially expressed. Using these criteria, we identified 205 differentially expressed miRNAs in AP. Among these miRNAs, 62 (30%) were up-regulated and 143 (70%) were down-regulated (Fig. 2A and B). Between SAP and MAP cases, nine miRNAs were differentially expressed, with four up-regulated and five downregulated in SAP compared to MAP patients (Fig. 2C). These results revealed unique serum miRNA expression profiles in AP, in particular in SAP, therefore suggesting that serum miRNAs may serve as biomarkers for AP diagnosis and prognosis. 3.2. Selection of serum miRNA candidates for AP diagnosis To identify serum miRNAs with diagnostic potential, we further screened miRNA candidates using the following four criteria. First, a candidate gene must be differentially expressed (fold change greater than 2 and p value less than 0.01) between AP patients and controls or between SAP and MAP cases. Second, the trend of altered expression of a given candidate gene between AP patients and controls must be the same as that between the SAP and MAP groups. Third, miRNAs with greater fold changes or lesser p values between the different groups are preferred. Last, miRNAs expressed with high abundance in serum are preferred. Although the first criterion is required, the other three criteria are optional. Based on these criteria, we identified 11 candidate miRNAs (Table 2), including seven down-regulated miRNAs and four up-regulated miRNAs, in the blood from AP patients compared to that from the controls. Among the four up-regulated miRNAs, miR-551b-5p and miR-890 also showed significantly higher expression in the blood from SAP patients than in that from MAP patients (Table 2). 3.3. Validation of candidate miRNAs in an independent cohort of AP patients and controls Next, we sought to validate the differential expression of serum miRNA candidates by RT-qPCR in an independent cohort that included 62 AP patients (37 MAP cases and 25 SAP cases) and 18 controls. Although the MAP and SAP groups were of comparable age and gender composition, significant differences were observed in their serum calcium concentrations, CRP levels, and CT and APACHE II scores (Table 3). We reasoned that the internal control miRNA for RT-qPCR should be consistently expressed regardless the origin of the miRNA sample and it should be present in high abundance in serum. After analyzing the 15 miRNA profiles from our initial
162
P. Liu et al. / Pancreatology 14 (2014) 159e166
Fig. 2. Hierarchical clustering analysis of differentially expressed serum miRNAs in AP. A, Two hundred five miRNAs were differentially expressed between acute pancreatitis patients (AP, n ¼ 12) and controls (N, n ¼ 3). B, Thirty-one most differentially expressed serum miRNAs between AP patients and healthy controls. C, Nine differentially expressed serum miRNAs between MAP and SAP cases (n ¼ 6 for each). Red indicates high level expression, and blue indicates low level expression.
screening and conducting a literature review, we selected has-miR16 as the internal control. RT-qPCR for all 11 candidate miRNAs confirmed the downregulation of three miRNAs, miR-92b, miR-10a, and miR-7, in AP (Fig. 3). A low level of serum calcium is known to correlate with poor prognosis in AP patients [18], and high CRP (>150 mg/L) is a marker of severe AP [19,20]. In addition, a plasma PCT level greater than 0.5 ng/mL has been used to determine infection status in AP and predict disease severity [21,22]. Therefore, we investigated whether the serum miR-551b-5p level correlates with any of these AP severity biomarkers. As demonstrated in Fig. 4, miR-551b-5p expression was significantly higher in patients who had low plasma calcium levels or presented with complications (Fig. 4).
When patients were stratified by PCT or CRP levels, miR-551b-5p expression was not different between different groups (Fig. 4). Taken together, these results confirmed the down-regulation of miR-92b, miR-10a, and miR-7 in the blood of AP patients and demonstrated that elevated miR-551b-5p expression correlated with some severity markers, suggesting that these miRNAs may serve as biomarkers for AP detection and severity prediction. 3.4. Diagnostic potential of differentially expressed miRNAs in AP Given that the expression levels of miR-92b, miR-10a, and miR-7 were significantly down-regulated in blood samples from AP patients collected within 24 h of hospital admission, we reasoned that these three miRNAs might serve as biomarkers for the early
Table 2 List of candidate miRNAs. ID
145897 17561 10306 29490 13485 145634 42923 147809 11014 42781 28047
Name
hsa-miR-92b ebv-miR-BART6-3p hsa-miR-146b-5p hsa-miR-7 hsa-miR-10a hsa-miR-132* hsa-miR-30c hsa-miR-514b-3p hsa-miR-214 hsa-miR-551b-5p hsa-miR-890
AP vs Normal
SAP vs MAP
Fold change
p Value
Mean of normal
Mean of AP
Fold change
p Value
Mean of MAP
Mean of SAP
0.25 0.35 0.38 0.43 0.32 0.36 0.43 8.81 13.54 5.43 3.93
1.16E-05 1.62E-05 2.01E-05 5.35E-05 9.04E-05 1.06E-04 1.86E-03 2.36E-03 5.33E-03 9.61E-03 5.00E-02
1.87 4.28 7.99 17.90 8.03 1.60 4.29 0.03 0.04 0.06 0.22
0.47 1.49 3.04 7.69 2.59 0.58 1.85 0.23 0.50 0.34 0.88
2.00 2.38
3.30E-03 6.44E-03
0.23 0.52
0.46 1.24
P. Liu et al. / Pancreatology 14 (2014) 159e166
4. Discussion
Table 3 Clinical characteristics of patients with MAP and SAP.
Age (years) Sex (male) Ca (mmol/L) PCT (ng/ml) CRP (mg/L) CT A B C D E APACHE II
MAP (n ¼ 37)
SAP (n ¼ 25)
p Valuea
49.32 23 1.96 3.70 99.34
54.20 18 1.8 18.81 220.4
0.29 0.6447b 0.02 0.10 0.00
Contents lists available at ScienceDirect
Pancreatology journal homepage: www.elsevier.com/locate/pan
Original article
Identification of serum microRNAs as diagnostic and prognostic biomarkers for acute pancreatitis Pi Liu a,1, Liang Xia a,1, Wei-long Zhang b, Hua-jing Ke a, Tao Su a, Li-bing Deng b, You-xiang Chen a, Nong-hua Lv a, * a b
Department of Gastroenterology, The First Affiliated Hospital of Nanchang University, Nanchang 330006, China Nanchang University, Nanchang 330031, China
a r t i c l e i n f o
a b s t r a c t
Available online 28 March 2014
Background/objectives: To identify serum microRNA (miRNA) as diagnostic and prognostic biomarkers for acute pancreatitis (AP). Materials and methods: Sera microRNA expression was profiled from 12 AP patients with varying disease severity and three healthy controls. Differentially expressed miRNAs were validated in a larger cohort of patients and controls. The diagnostic and prognostic potentials of differentially expressed miRNAs were evaluated using receiver operating characteristic (ROC) curve analysis and compared to that of classic prognostic markers for AP. Results: miRNA microarray analyses identified 205 differentially expressed miRNAs between sera from AP patients and that from controls. Nine miRNAs were differentially expressed between severe and mild AP patients. Further validation confirmed the down-regulation of miR-92b, miR-10a, and miR-7 in AP patients, and ROC analysis revealed that these miRNAs can differentiate AP from health cases. Furthermore, the serum miR-551b-5p level was significantly higher in patients with disease complications or a low plasma calcium level. ROC analysis showed that the serum miR-551b-5p level can distinguish between severe and mild AP. Conclusion: The expressions of miR-92b, miR-10a, and miR-7 in AP might be used for the early diagnosis of AP and miR-551b-5p may be used for predicting AP severity. Copyright Ó 2014, IAP and EPC. Published by Elsevier India, a division of Reed Elsevier India Pvt. Ltd. All rights reserved.
Keywords: microRNA Acute pancreatitis Early diagnosis Prognosis Biomarker Prediction
1. Introduction Acute pancreatitis (AP) is a reversible inflammatory process of the pancreas with a wide range of clinical variations. Mild AP (MAP) is often self-limited and has a very low mortality rate ( 2, and (2) p value < 0.01. Hierarchical clustering of differentially expressed miRNAs was analyzed as a heat map, using MEV (version 4.6). Furthermore, RT-qPCR data were analyzed using the comparative CT method. DCT for each target miRNA was defined as CTtarget miRNA e CT miR-16. Two-sided, two-tailed Student t-tests were performed to determine the statistical significance (p < 0.05) of differences between groups. ROC curve analysis was performed using GraphPad Prism version 5.
2.4. miRNA microarray 3. Results Total serum miRNAs including six SAP, six MAP, and three control cases were used for miRNA profiling. The miRCURYÔHy3Ô/ Hy5Ô Power labeling kit (Exiqon, Vedbaek, Denmark) was used for miRNA labeling according to the manufacturer’s instructions. The labeled miRNA samples were hybridized onto the miRCURYÔ LNA Array (v. 16.0, Exqo). The complete procedures of hybridization, incubation, and washing were conducted according to the manufacturer’s instructions. The resulting signals were scanned using an Axon GenePix 4000B microarray scanner. 2.5. miRNA quantification by RT-qPCR Total serum miRNAs (15 ng) isolated from the validation cohort (18 controls and 62 patients) were reverse transcribed using the One Step PrimeScript miRNA cDNA Synthesis Kit (Takara Biotechnology, D350A) with a universal adaptor primer. The RT reaction was performed at 37 C for 60 min, followed by 85 C for 5 s. The synthesized cDNA was then amplified by quantitative PCR, using SYBR Premix EX Taq II (Takara Biotechnology, DRR081A), Uni-miRQ PCR Primer (Takara Biotechnology), and primers specific to each individual miRNA (Table 1). The reactions were conducted using PRISM 7500 System (Applied Biosystems, USA). Reactions with no template were used as negative controls, and has-miR-16 was used as internal control [17]. The reaction conditions were 95 C for 30 s followed by 40 cycles of 95 C for 15 s and 60 C for 60 s (ramping was set to 1.6 C/sec). 2.6. Statistical analysis The miRNA array results were extracted and analyzed using GenePix Pro 6.0 software (Axon). The miRNA expression signals were normalized using median normalization. miRNAs with signal values higher than 50 in all groups were included in further analyses. The signals from multiple probes for each miRNA were averaged to represent the expression of the miRNA. Two-tailed ManneWhitney unpaired test was used to determine the
Table 1 miRNA-specific primers. 0
0
ID
miRNA
Sequence (from 5 to 3 )
Number of nucleotides
01 02 03 04 05 06 07 08 09 10 11 12
hsa-miR-92b hsa-miR-10a ebv-miR-BART6-3p hsa-miR-132* hsa-miR-146b-5p hsa-miR-30c hsa-miR-7 hsa-miR-890 hsa-miR-551b-5p hsa-miR-514b-3p hsa-miR-214 hsa-miR-16
AGGGACGGGACGCGGTG GTACCCTGTAGATCCGAATTTGTG CGGGGATCGGACTAGCCTTAGA ACCGTGGCTTTCGATTGTTAC GGCTGAGAACTGAATTCCATAGGC CGCTGTAAACATCCTACACTCTCAG CGGTGGAAGACTAGTGATTTTGTTG CGTACTTGGAAAGGCATCAGTTG GAAATCAAGCGTGGGTGAGACC CCATTGACACCTCTGTGAGTGGA TGCCTGTCTACACTTGCTGTGC TAGCAGCACGTAAATATTGGCG
17 24 22 21 24 25 25 23 22 23 22 22
3.1. Serum miRNA expression profiling in AP To identify differentially expressed miRNAs in blood from AP patients, we performed miRNA microarray analyses on serum miRNA samples from 3 healthy control and 12 AP patients (6 SAP and 6 MAP). All blood samples were collected within the first 24 h after hospital admission. MiRNAs with expression fold change greater than 2 with a p value less than 0.01 were considered as differentially expressed. Using these criteria, we identified 205 differentially expressed miRNAs in AP. Among these miRNAs, 62 (30%) were up-regulated and 143 (70%) were down-regulated (Fig. 2A and B). Between SAP and MAP cases, nine miRNAs were differentially expressed, with four up-regulated and five downregulated in SAP compared to MAP patients (Fig. 2C). These results revealed unique serum miRNA expression profiles in AP, in particular in SAP, therefore suggesting that serum miRNAs may serve as biomarkers for AP diagnosis and prognosis. 3.2. Selection of serum miRNA candidates for AP diagnosis To identify serum miRNAs with diagnostic potential, we further screened miRNA candidates using the following four criteria. First, a candidate gene must be differentially expressed (fold change greater than 2 and p value less than 0.01) between AP patients and controls or between SAP and MAP cases. Second, the trend of altered expression of a given candidate gene between AP patients and controls must be the same as that between the SAP and MAP groups. Third, miRNAs with greater fold changes or lesser p values between the different groups are preferred. Last, miRNAs expressed with high abundance in serum are preferred. Although the first criterion is required, the other three criteria are optional. Based on these criteria, we identified 11 candidate miRNAs (Table 2), including seven down-regulated miRNAs and four up-regulated miRNAs, in the blood from AP patients compared to that from the controls. Among the four up-regulated miRNAs, miR-551b-5p and miR-890 also showed significantly higher expression in the blood from SAP patients than in that from MAP patients (Table 2). 3.3. Validation of candidate miRNAs in an independent cohort of AP patients and controls Next, we sought to validate the differential expression of serum miRNA candidates by RT-qPCR in an independent cohort that included 62 AP patients (37 MAP cases and 25 SAP cases) and 18 controls. Although the MAP and SAP groups were of comparable age and gender composition, significant differences were observed in their serum calcium concentrations, CRP levels, and CT and APACHE II scores (Table 3). We reasoned that the internal control miRNA for RT-qPCR should be consistently expressed regardless the origin of the miRNA sample and it should be present in high abundance in serum. After analyzing the 15 miRNA profiles from our initial
162
P. Liu et al. / Pancreatology 14 (2014) 159e166
Fig. 2. Hierarchical clustering analysis of differentially expressed serum miRNAs in AP. A, Two hundred five miRNAs were differentially expressed between acute pancreatitis patients (AP, n ¼ 12) and controls (N, n ¼ 3). B, Thirty-one most differentially expressed serum miRNAs between AP patients and healthy controls. C, Nine differentially expressed serum miRNAs between MAP and SAP cases (n ¼ 6 for each). Red indicates high level expression, and blue indicates low level expression.
screening and conducting a literature review, we selected has-miR16 as the internal control. RT-qPCR for all 11 candidate miRNAs confirmed the downregulation of three miRNAs, miR-92b, miR-10a, and miR-7, in AP (Fig. 3). A low level of serum calcium is known to correlate with poor prognosis in AP patients [18], and high CRP (>150 mg/L) is a marker of severe AP [19,20]. In addition, a plasma PCT level greater than 0.5 ng/mL has been used to determine infection status in AP and predict disease severity [21,22]. Therefore, we investigated whether the serum miR-551b-5p level correlates with any of these AP severity biomarkers. As demonstrated in Fig. 4, miR-551b-5p expression was significantly higher in patients who had low plasma calcium levels or presented with complications (Fig. 4).
When patients were stratified by PCT or CRP levels, miR-551b-5p expression was not different between different groups (Fig. 4). Taken together, these results confirmed the down-regulation of miR-92b, miR-10a, and miR-7 in the blood of AP patients and demonstrated that elevated miR-551b-5p expression correlated with some severity markers, suggesting that these miRNAs may serve as biomarkers for AP detection and severity prediction. 3.4. Diagnostic potential of differentially expressed miRNAs in AP Given that the expression levels of miR-92b, miR-10a, and miR-7 were significantly down-regulated in blood samples from AP patients collected within 24 h of hospital admission, we reasoned that these three miRNAs might serve as biomarkers for the early
Table 2 List of candidate miRNAs. ID
145897 17561 10306 29490 13485 145634 42923 147809 11014 42781 28047
Name
hsa-miR-92b ebv-miR-BART6-3p hsa-miR-146b-5p hsa-miR-7 hsa-miR-10a hsa-miR-132* hsa-miR-30c hsa-miR-514b-3p hsa-miR-214 hsa-miR-551b-5p hsa-miR-890
AP vs Normal
SAP vs MAP
Fold change
p Value
Mean of normal
Mean of AP
Fold change
p Value
Mean of MAP
Mean of SAP
0.25 0.35 0.38 0.43 0.32 0.36 0.43 8.81 13.54 5.43 3.93
1.16E-05 1.62E-05 2.01E-05 5.35E-05 9.04E-05 1.06E-04 1.86E-03 2.36E-03 5.33E-03 9.61E-03 5.00E-02
1.87 4.28 7.99 17.90 8.03 1.60 4.29 0.03 0.04 0.06 0.22
0.47 1.49 3.04 7.69 2.59 0.58 1.85 0.23 0.50 0.34 0.88
2.00 2.38
3.30E-03 6.44E-03
0.23 0.52
0.46 1.24
P. Liu et al. / Pancreatology 14 (2014) 159e166
4. Discussion
Table 3 Clinical characteristics of patients with MAP and SAP.
Age (years) Sex (male) Ca (mmol/L) PCT (ng/ml) CRP (mg/L) CT A B C D E APACHE II
MAP (n ¼ 37)
SAP (n ¼ 25)
p Valuea
49.32 23 1.96 3.70 99.34
54.20 18 1.8 18.81 220.4
0.29 0.6447b 0.02 0.10 0.00
