Validation of a Bead-Based Multiplex Assay for miRNA Quantification in Rat Liver Inflammation

Abstract

MicroRNAs (miRNAs) are key biomarkers for the diagnosis, prevention, and treatment of various diseases. Most miRNAs interact with one another to influence disease occurrence, making multiplex detection essential for miRNA analysis. In this study, we developed a quantitative analysis method for nine miRNAs using a bead-based flow cytometry system. The method was validated using calibration curves, working range, precision, accuracy, specificity, and carry-over assessments in accordance with bioanalysis guidelines. We analyzed the expression of eight inflammation-related miRNAs, including miR-21, miR-24, miR-29a, miR-34a, miR-122, miR-125a, miR-146a, and miR-155, using cel-miR-39 as an internal control. To demonstrate the applicability of the proposed method, we analyzed nine miRNAs in the livers of male and female rats administered CpG ODN 1826 and LPS and compared the results obtained using the RT–qPCR. The results revealed that the calibration standards for the nine miRNAs included nine points, covering a quantification range of 0.20 to 51.20 pg/μL, with a five-parameter logistic (5-PL) curve fit within ±20%. The quality control (QC) samples for all nine miRNAs met the accuracy and precision criteria, with a %RE within ±20% (25% at the LLOQ and ULOQ) and a %CV of ≤20% (25% at the LLOQ and ULOQ). The total error at each QC level was within the acceptance criteria of ≤30% (40% at the LLOQ and ULOQ). Specificity and carry-over evaluations indicated that specific analytes in the multiplex assay were successfully quantified without interference from other analytes, with no carry-over effects observed. Finally, a total of nine miRNAs were quantitatively analyzed in the livers of rats with inflammation induced by CpG ODN 1826 and LPS, compared with the control group. Of these, the expression of miR-21, miR-34a, miR-146a, and miR-155 tended to increase during liver inflammation in both male and female rats. These results were consistent with the RT–qPCR data. In conclusion, this analytical method, validated in accordance with FDA guidelines, enables accurate and quantitative multiplex detection of inflammation-related miRNA biomarkers in liver tissue. It demonstrates strong reliability in current tissue analyses and provides a foundation for developing multiplex miRNA assays capable of analyzing body fluids, with potential applications in future clinical studies.

Introduction

MicroRNAs (miRNAs) are small noncoding RNA molecules composed of approximately 22 nucleotides that play crucial roles in the regulation of gene expression. miRNAs bind to specific mRNAs, inhibiting their translation or promoting their degradation, thereby regulating various physiological processes and metabolic pathways within cells. This function is essential for various cellular processes, including cell proliferation, differentiation, survival, and apoptosis, making miRNAs critical for maintaining normal cellular functions.

Recent studies have increasingly highlighted the importance of miRNAs as biomarkers. miRNAs can be reliably detected in various biological samples, including blood, saliva, and tissues, making them useful indicators for early diagnosis and prognosis of diseases. , Notably, changes in miRNA expression patterns have been observed in various diseases, such as cancer, neurological disorders, and metabolic disorders. These alterations have significant implications for monitoring disease progression and therapeutic responses. Moreover, miRNAs interact with multiple target mRNAs to regulate complex gene expression networks. Understanding the overall biological context of diseases exclusively on the basis of changes in individual miRNAs can be challenging. Thus, analyzing the signatures of multiple miRNAs that act together is crucial. For this reason, the development of multiplex technologies for miRNA biomarker analysis is essential, as it will facilitate a deeper understanding of the interactions among various miRNAs and their associated biological significance.

miRNA detection methods include a variety of technologies and methodologies that are primarily used for quantifying and analyzing miRNA expression levels. Conventional miRNA detection methods include reverse transcription–qPCR (RT–qPCR), microarray, next-generation sequencing (NGS), Northern blot (NB), and in situ hybridization (ISH). RT–qPCR is considered the gold standard method; this technique is highly sensitive, but it requires the interconnection of multiple steps and cannot perform multiple detection. Microarrays can analyze hundreds of targets but have lower sensitivity. NGS enables multiplex analysis of hundreds to thousands of targets. However, its data analysis is complex, and the associated costs are relatively high.

The bead-based multiplex assay developed and validated in our study offers a quantitative analysis method that provides high-throughput screening, resulting in time and sample savings, as well as high specificity. Unlike the RT–qPCR method, this system allows direct quantitative detection of miRNAs from total RNA samples without cDNA synthesis or nucleic acid amplification step, thereby minimizing sample manipulation and potential degradation, and enhancing experimental convenience. The bead-based system developed in this study was established based on Luminex’s xMAP technology, which utilizes magnetic microspheres (5.6 μm in diameter) internally dyed with two fluorescent substances to create up to 100 distinguishable bead sets. Each microsphere surface is coated with capture probes specific to target analytes, and a chimeric probe is prepared for use in conjunction with these probes. The chimeric probe consists of RNA and DNA components, where the RNA segment is complementary to the target miRNA and the DNA segment binds to the bead capture probe. After mixing, the biotinylated chimeric probe reacts with streptavidin and R-phycoerythrin conjugate (SAPE) to generate a fluorescent signal, which can be quantitatively analyzed on the basis of the intensity of the signal. This bead-based xMAP system has been proposed as an alternative method for verifying miRNA expression. ,

Here, we developed and validated an analytical method that enables the simultaneous quantification of nine miRNAs (miR-21, miR-24, miR-29a, miR-34a, miR-122, miR-125a, miR-146a, miR-155, and cel-miR-39) in CpG ODN 1826 and LPS-induced rat liver. miR-39, derived from Caenorhabditis elegans, was used as a spike-in control normalizer, whereas the other eight miRNAs are associated with increased release of cytokines during inflammation. To achieve this objective, we established a bead-based multiplex assay for the quantification of miRNAs with high accuracy, precision, specificity, and no carry-over on the basis of the United States Food and Drug Administration (US FDA) guidelines. We subsequently applied this method to actual samples in which inflammatory responses were induced, successfully detecting miRNAs and demonstrating the feasibility and practicality of the approach. Finally, we confirmed the accuracy and reliability of our bioanalytical method by comparing the results obtained from RT–qPCR analysis of the samples.

Results

Determination of the Calibration Curve for the Bead-Based Multiplex Assay

In this study, a bead-based multiplex assay was developed for the quantification of nine miRNAs, and the overall analytical workflow is illustrated in Figure A.

1.

Bead-based multiplex assay workflow and representative calibration curve. (A) Schematic diagram illustrating the workflow of the bead-based multiplex assay for miRNA quantification. (B) The calibration curves were generated for each of the nine miRNAs, depicting the relationship between concentration (pg/μL) on the x-axis and the corresponding measured signal (MFI) on the y-axis. Each curve represents the mean values from six measurements, with error bars indicating the standard deviation (SD).

To confirm the suitability of the experimental procedure for the bead-based bioanalytical method, calibration standards were prepared by diluting nine synthetic miRNAs. These standards were diluted 1:2 to produce ten calibrator concentrations, all of which exhibited the same detection range of 0.10 to 51.20 pg/μL. The five-parameter logistic (5-PL) model was used for calibration curve fitting, and a total of nine calibration curves were evaluated to thoroughly assess calibration performance (Figure B). Each calibration curve included the lowest accurately measurable concentration standard, defined as the lower limit of quantification (LLOQ, 0.20 pg/μL), and the highest measurable miRNA concentration, referred to as the upper limit of quantification (ULOQ, 51.20 pg/μL). A calibrator of 0.10 pg/μL, which is below the LLOQ, was used as the anchor point, establishing a quantification range from 0.20 pg/μL to 51.20 pg/μL.

Calibration curves for nine miRNAs were measured six times on different days, with reproducibility analyzed on the basis of the accuracy (percent relative error, %RE) and precision (percent coefficient of variation, %CV) of the concentration of the back-calculated calibration standards. The results indicated that the %RE and %CV of the calibration curve for the nine biomarkers ranged from −4.17% to 10.83% and 0.95% to 10.45%, respectively (Table ). The results suggest that our assay met the acceptance criteria of %RE (within ±20%, 25% at the LLOQ and ULOQ) and %CV (≤20%, 25% at the LLOQ and ULOQ).

1. 6-Replicate Analysis Results of Calibration Curve of Nine miRNAs.

nominal concentration (pg/μL)	inter-assay %RE
	miR-21	miR-24	miR-29a	miR-34a	miR-122	miR-125a	miR-146a	miR-155	cel-miR-39
0.10	–8.33	0.00	0.00	0.00	–1.67	–6.67	0.00	–1.67	–6.67
0.20	10.83	5.83	2.50	–0.83	4.17	10.83	5.00	4.17	5.83
0.40	–2.92	–2.08	–0.42	–1.67	–0.42	–4.17	–2.08	–1.67	0.42
0.80	–0.63	–2.29	–0.83	0.21	–2.08	0.21	–1.25	0.42	–1.04
1.60	1.04	2.50	–0.42	2.50	0.10	–0.83	0.31	–0.63	–0.31
3.20	–0.89	0.73	–0.05	–0.26	0.10	1.35	0.16	–0.31	0.21
6.40	1.38	0.89	3.39	–0.03	2.73	1.25	2.29	1.30	1.20
12.80	0.08	–3.93	–2.29	–3.16	–0.89	–1.18	–2.25	–0.53	0.03
25.60	–0.64	5.40	1.00	6.21	–1.65	0.48	0.47	–0.07	–0.36
51.20	0.41	–2.33	0.11	–3.43	1.22	0.23	0.24	0.17	0.41

nominal concentration (pg/μL)	inter-assay %CV
	miR-21	miR-24	miR-29a	miR-34a	miR-122	miR-125a	miR-146a	miR-155	cel-miR-39
0.10	8.21	0.00	0.00	0.00	4.15	8.75	0.00	4.15	5.53
0.20	9.21	4.65	5.12	2.06	3.61	10.45	4.26	3.61	6.28
0.40	3.79	2.98	3.34	2.08	2.93	5.39	1.92	3.83	2.45
0.80	3.54	3.47	2.21	2.29	2.87	2.00	2.26	1.51	3.14
1.60	1.78	2.31	2.68	2.73	1.00	3.37	1.17	0.97	1.67
3.20	2.56	1.58	2.54	1.07	2.84	1.98	1.15	1.82	2.34
6.40	1.90	2.92	5.39	2.43	2.66	2.75	2.10	2.24	1.99
12.80	3.41	2.62	3.60	2.28	2.52	4.16	2.26	2.49	3.23
25.60	4.78	5.12	6.63	4.42	4.36	6.73	2.24	2.39	6.66
51.20	2.01	2.31	3.09	2.88	2.06	2.88	0.95	1.06	2.76

The acceptance criteria for each calibrator are as follows:

^a%RE within ±20% (25% at the LLOQ and ULOQ), and

^b%CV ≤20% (25% at the LLOQ and ULOQ).

^cAnchor point is not included in the acceptance criteria.

Evaluation of Accuracy and Precision for Multiplex Assay Validation

Five quality controls (QCs) were prepared to evaluate the accuracy and precision of the multiplex assay. The assessment was based on interday variations and involved a total of six tests. The results showed that between-run accuracy (%RE) for five QCs (LLOQ, low quality control; LQC, middle quality control; MQC, high quality control; HQC, and ULOQ) for nine miRNAs ranged from −7.88% to 9.74%. Additionally, the between-run precision (%CV) for five QCs (LLOQ, LQC, MQC, HQC and ULOQ) for nine miRNAs ranged from 3.78% to 12.21%. Finally, the total error of all the QCs was determined to be within the range from 5.17% to 19.15% between runs. Thus, the results in Table indicate that the accuracy and precision of the analytical method for nine miRNAs met the acceptance criteria.

2. Inter-Assay Accuracy and Precision of the Multiplex Assay for Nine miRNAs.

%RE	QCs	miR-21	miR-24	miR-29a	miR-34a	miR-122	miR-125a	miR-146a	miR-155	cel-miR-39
between-run accuracy	LLOQ	1.82	–4.55	–4.55	–2.27	–0.45	–4.32	–2.95	–1.36	–2.50
	LQC	1.89	0.98	1.44	3.48	3.86	–1.21	2.27	2.80	1.59
	MQC	1.38	1.53	6.19	2.21	4.22	2.79	4.80	2.34	3.00
	HQC	–0.13	3.95	9.74	2.09	–0.71	0.37	9.56	–1.54	0.64
	ULOQ	–2.89	–7.88	–1.42	–7.75	0.56	0.49	–0.11	–0.10	–3.07

%CV	QCs	miR-21	miR-24	miR-29a	miR-34a	miR-122	miR-125a	miR-146a	miR-155	cel-miR-39
between-run precision	LLOQ	11.60	12.09	8.54	7.53	9.79	12.21	7.73	10.40	9.46
	LQC	7.32	10.25	10.28	8.55	9.03	9.33	8.89	9.18	8.96
	MQC	3.78	6.97	8.03	6.25	5.01	6.25	6.82	7.02	4.48
	HQC	5.73	4.57	7.20	5.17	5.59	6.42	4.78	9.37	7.15
	ULOQ	6.54	8.42	6.50	11.40	7.24	5.84	8.85	10.87	9.07

%total error	QCs	miR-21	miR-24	miR-29a	miR-34a	miR-122	miR-125a	miR-146a	miR-155	cel-miR-39
between-run	LLOQ	13.41	16.63	13.09	9.80	10.25	16.53	10.69	11.76	11.96
	LQC	9.21	11.23	11.72	12.03	12.89	10.54	11.16	11.98	10.55
	MQC	5.17	8.51	14.22	8.46	9.22	9.04	11.62	9.37	7.47
	HQC	5.86	8.53	16.93	7.25	6.30	6.79	14.34	10.92	7.80
	ULOQ	9.43	16.30	7.92	19.15	7.80	6.34	8.96	10.97	12.14

The acceptance criteria are as follows:

^a%RE within ±20% (25% at LLOQ and ULOQ),

^b%CV ≤20% (25% at LLOQ and ULOQ), and

^c%total error ≤30% (40% at LLOQ and ULOQ) at each QC level.

Specificity Assessment for Multiplex Assays

The specificity of a multiplex assay refers to its ability to accurately identify and quantify target analytes without interference from other substances, making the establishment of multiplex assays with high specificity important. To assess specificity, a calibration curve for the specific miRNA was prepared alongside QCs containing the remaining eight miRNAs. Using bead-based analysis, the QC of the specific miRNA was evaluated to determine compliance with the acceptance criteria for %RE and %CV. For the specificity assessment, LQC (n = 3) and HQC (n = 3) samples were utilized. The results indicated that the %RE for the LQC and HQC of the nine miRNAs ranged from −5.00% to 15.00% and from −17.82% to 10.11%, respectively, whereas the %CV for the LQC and HQC ranged from 1.47% to 7.15% and from 0.76% to 9.03%, respectively. Thus, our findings demonstrate that our assay reacts exclusively with the target miRNA and not with other miRNAs (Table ).

3. Specificity Evaluation of Bead-Based Multiplex Assay for Nine miRNAs.

QCs %RE	miR-21	miR-24	miR-29a	miR-34a	miR-122	miR-125a	miR-146a	miR-155	cel-miR-39
LQC	10.00	2.22	11.11	13.33	–5.00	4.44	15.00	1.67	1.67
HQC	–3.76	0.78	4.08	–9.32	3.64	–17.82	10.11	–5.78	–3.73

QCs %CV	miR-21	miR-24	miR-29a	miR-34a	miR-122	miR-125a	miR-146a	miR-155	cel-miR-39
LQC	6.94	3.39	2.29	1.47	6.33	4.02	3.83	2.84	7.15
HQC	9.03	1.63	5.91	0.76	5.93	1.53	2.99	3.48	4.35

The acceptance criteria are as follows:

^a%RE within ±20%, and

^b%CV ≤20% at each QC levels.

Carry-Over Evaluation in a Bead-Based Flow Cytometry System

Carry-over is an assessment to determine whether changes in measured concentration occur because of residual analyte from a preceding sample in the instrument. To evaluate carry-over, a blank sample was analyzed immediately after the ULOQ of the calibration curve, and whether the mean fluorescence intensity (MFI) of the blank sample was lower than the LLOQ was assessed. The ULOQs for the nine miRNAs were high, ranging from 15,257 to 27,803, whereas the MFI of the blank samples was indeed lower than the LLOQ of the calibration curve. These findings indicate that there is no carry-over between analytes in the analytical method (Table ).

4. Carry-Over Assessment of Multiplex Assay.

item	QC	miR-21	miR-24	miR-29a	miR-34a	miR-122	miR-125a	miR-146a	miR-155	cel-miR-39
MFI	ULOQ	22274	26545	25351	27803	19896	24785	25616	15257	25317
	LLOQ	441	511	511	542	349	566	507	316	604
	blank	190	169	147	105	187	290	189	161	246

^aMFI signal of QCs at the calibration curves.

Multiplex Assay with Inflammatory Rat Liver

Next, we analyzed nine miRNAs using our validated method. Among the nine miRNAs, cel-miR-39 serves as the internal control, whereas the remaining eight miRNAs (miR-21, miR-24, miR-29a, miR-34a, miR-122, miR-125a, miR-146a, and miR-155) are known to regulate cytokine production associated with inflammation and cytokine release syndrome. , The samples were derived from a rat model of liver inflammation induced by the administration of CpG ODN 1826 and LPS.

The results revealed that the quantitative range of the calibration curve for all nine miRNAs was 0.20 pg/μL to 51.20 pg/μL, with a %RE within ±20% and a %CV of ≤20%. Additionally, the QCs (LQC, MQC and HQC) satisfied the criteria of %RE within ±20% and %CV ≤20% (results not shown). Although the expression levels of nine miRNAs were within the calibration curve range, miR-122 expression was significantly greater than that of the other miRNAs, above the ULOQ of 51.20 pg/μL in female rat samples. Consequently, the miR-122 expression in female rats was reanalyzed after being diluted 2.2 times. In contrast, miR-155 expression was below the LLOQ (0.20 pg/μL) in both the vehicle control and negative control groups; however, its expression was increased in the CpG ODN 1826 and LPS treatment groups, allowing for quantification within the calibration curve range.

The results of the multiplex analysis shown in Figure confirmed that the concentration of cel-miR-39, which was used as an internal control, was consistently 2.30 pg/μL across all groups of female and male rats. Among the eight miRNAs, the expression of miR-21, miR-34a, miR-146a, and miR-155 tended to increase in the livers of both male and female rats treated with CpG ODN 1826 and LPS. In particular, compared with those in the vehicle and negative control groups, the levels of miR-21 and miR-155 in the livers of both male and female rats with CpG ODN 1826 and LPS significantly increased. On the other hand, the expression of miR-34a and miR-146a was significantly increased in female rats treated with CpG ODN 1826 and LPS compared with the vehicle and negative control groups.

2.

Detection of target miRNAs using a bead-based multiplex assay. The expression levels of miRNAs were assessed in the livers of male and female rats in the vehicle control (VC, saline), negative control (NC, control CpG ODN 1826), and treatment (T, CpG ODN 1826 and LPS) groups (n = 4 rats per group). Data are shown as the mean ± SD. Statistical significance is indicated as follows: *p ≤ 0.05 (VC vs T), ^# p ≤ 0.05 (NC vs T), and ⁺ p ≤ 0.05 (VC vs NC).

Comparison between Bead-Based Multiplex Assays and RT–qPCR

We compared the results obtained from our validated analytical method for eight inflammation-associated miRNAs with those from the well-established RT–qPCR method. For this comparison, we performed RT–qPCR after cDNA synthesis of the same samples that were utilized in the multiplex assay. The results in Figure indicate that the expression patterns of the eight miRNAs were nearly identical across the two methods. Among the eight miRNAs, miR-21, miR-34a, miR-146a, and miR-155 exhibited increased expression levels in both male and female rats treated with CpG ODN 1826 and LPS compared with those in both the vehicle and negative control groups. In contrast, the expression of the remaining miRNAs, namely, miR-24, miR-29a, miR-122, and miR-125a, did not significantly change across all the CpG ODN 1826 and LPS treatment groups compared with that in both the vehicle and negative control groups. This finding was consistently confirmed using both analytical methods.

3.

Comparative analysis of miRNA expression using two different technique. The expression levels of eight miRNAs obtained from the bead-based multiplex assay (left y-axis, bar graph) and RT–qPCR (right y-axis, line graph) were presented. Data are shown as the mean ± SD, Vehicle control (VC, saline), negative control (NC, control CpG ODN 1826), and treatment (T, CpG ODN 1826 and LPS) groups (n = 4 rats per group).

Discussion

In this study, we developed and validated a multiplex analytical method to quantify miRNAs using a bead-based flow cytometry system. Our multiplex assay for miRNAs demonstrated accuracy, precision, and specificity, allowing for the simultaneous quantitative analysis of nine miRNAs without the influence of carry-over effects. Utilizing this validated method, we analyzed the expression of nine miRNAs in the livers of male and female rats administered with CpG ODN 1826 and LPS. We detected increases in the expression of four miRNAs, namely, miR-21, miR-34a, miR-146a, and miR-155, in both male and female rats with inflammation induced by CpG ODN 1826 and LPS. These findings were consistent with those obtained using the well-established RT–qPCR method, supporting the validity of our results. Thus, our validated multiplex assay is suitable for the quantitative analysis of miRNA biomarkers in rat liver tissues.

The bead-based xMAP system has been widely used for multiplex analyses of proteins, DNA, and miRNAs, and has also been applied to various biological studies, including those on inflammation, liver injury, and cancer biomarkers. − However, most miRNA assays based on this platform have remained at a research-use level without systematic validation of analytical robustness. In this study, we optimized the xMAP-based miRNA assay by adjusting hybridization time and probe concentrations, and validated the method according to FDA bioanalytical guidelines. Through these optimizations, we successfully established a reliable and reproducible multiplex assay for the simultaneous quantification of eight inflammation-related miRNAs, demonstrating an analytically optimized implementation of xMAP technology rather than a simple application of existing protocols.

miRNAs serve as valuable biomarkers for disease diagnosis and prognostic assessment in variety of conditions, including cancer, and the field of miRNA research continues to receive increasing interest. , Specific miRNA expression patterns are observed across different diseases, indicating that miRNAs are involved in complex networks that regulate the expression of specific genes. For example, in gastric cancer, combinations of miRNAs such as miR-224/452 and miR-181c/-340 have been shown to contribute to the regulation of genes such as KRAS and MECP2. Yaqub et al., reported that the expression of miR-4539, miR-372–3p, miR-566, and miR-7106–5p wew significantly altered in both cognition and dementia, affecting intracellular signaling and synaptic function. Additionally, TAmiRNA GmbH, a biotechnology company, provides diagnostic, predictive, and prognostic insights by analyzing miRNA expression patterns related to specific diseases such as liver disease, osteoporosis, and cardiovascular diseases. These findings suggest that analyzing multiple miRNAs simultaneously is essential for better understanding the mechanisms of specific diseases and for the development of diagnostic and therapeutic strategies. Therefore, multiple analysis methods should be considered for miRNA biomarker analysis rather than single analysis methods.

Our validated analytical method demonstrates high accuracy, precision, and specificity, allowing for the simultaneous screening of nine types of miRNA profiles. Compared with RT–qPCR, this method provides several practical advantages, including direct quantitative detection of miRNAs from total RNA without cDNA synthesis or nucleic acid amplificationthereby minimizing sample handling and degradationreduced analysis time (approximately 6 h for nine targets), minimal sample requirement (2.5 μL RNA), lower cost through multiplexing, a convenient single-step hybridization–detection process, and customizable assay design for up to 100 miRNA targets. These features suggest that the developed method can serve as a practical and efficient alternative to RT–qPCR for high-throughput miRNA screening in future clinical and toxicological studies.

A validated analytical method was used to detect nine miRNAs in rat livers. As shown in Figure , compared with the vehicle and negative control groups, the hepatic expression levels of miR-21, miR-34a, miR-146a, and miR-155 were significantly increased in rats administered with CpG ODN 1826 and LPS. These miRNAs are known to play critical roles in the progression and pathological changes of inflammatory liver diseases, , and their expression has been reported to increase in both serum and liver or hepatocytes across various liver diseases. , miR-122 is widely recognized as a representative biomarker for liver diseases, with elevated levels typically observed in serum or plasma. However, its expression pattern in liver tissue varies depending on the type and severity of liver injury. Bala et al., reported that CpG ODN 1826 and LPS-treated mice showed decreased hepatic miR-122 levels despite an increase in circulating miR-122, suggesting hepatocyte release into circulation. In contrast, in our rat model under identical treatment conditions, we observed elevated plasma miR-122 levels (results not shown) but no significant change in hepatic expression. This discrepancy may reflect species-specific differences in miRNA regulation between mice and rats. Indeed, meta-analysis studies have shown that miRNA expression patterns can differ substantially across species even under comparable pathological conditions.

To confirm the accuracy and reliability of the validated analytical method, we compared the results with those obtained using the RT–qPCR method (Figure ). The results revealed that the expression trends of the eight miRNAs in each group was consistent between the methods. Additionally, among the eight miRNAs analyzed, miR-21 and miR-155 showed markedly elevated expression compared with the other miRNAs, with consistent trends across both analytical methods. In the case of miR-155, our method indicated that its expression was below the LLOQ in the vehicle and negative control samples, making it unquantifiable, and the RT–qPCR method also yielded a Ct value of approximately 30. In contrast, both analytical methods revealed high miR-122 expression. Our method measured it at or above the ULOQ, and the RT–qPCR method yielded a Ct value of approximately 19, indicating its expression was significantly greater than that of the other seven miRNAs. Therefore, comparison of our multiplex analysis method with the RT–qPCR approach confirmed the reliability and accuracy of our method and further demonstrated its effectiveness.

This study has two main limitations. First, the relatively narrow quantification range (0.20–51.20 pg/μL) and the sigmoidal calibration curves may limit the accuracy for highly abundant miRNAs, such as miR-122. Second, the use of synthetic miRNA standards prepared in buffer does not fully account for potential matrix effects in complex biological samples. To address these limitations, samples exceeding the ULOQ were appropriately diluted and reanalyzed, and exogenous spike-in cel-miR-39 was added to evaluate RNA recovery and potential matrix effects. The consistent cel-miR-39 signal across all samples suggested stable RNA recovery and limited matrix interference, providing indirect support that the use of synthetic standards was reasonably reliable, although it does not fully account for all potential matrix effects. In future studies, calibration strategies that are matrix-matched or tissue-appropriate will be considered to more accurately account for potential matrix effects, such as signal suppression or enhancement, and to improve the quantitative accuracy of highly abundant endogenous miRNAs in tissue samples.

Conclusions

Our results demonstrate the quantification of nine miRNAs (miR-21, miR-24, miR-29a, miR-34a, miR-122, miR-125a, miR-146a, miR-155, and cel-miR-39) in rat liver tissues administered with CpG ODN 1826 and LPS using a bead-based multiplex assay. This method meets the bioanalytical method validation defined by the FDA guidelines, ensuring high reliability. Importantly, this assay requires a small sample volume of only 2.5 μL and does not require nucleic acid amplification step, allowing for efficient analysis of the nine miRNAs within approximately 6 h. The validated assay enabled detection of specific miRNA changes associated with liver inflammation, and these results were consistent with RT–qPCR data, further supporting the accuracy and reliability of the method.

miRNAs can be measured noninvasively in body fluids such as blood and urine, and they play a crucial role in disease diagnosis and prognosis prediction. Body fluid-derived miRNAs often exist in complexes with proteins or extracellular vesicles, which can reduce hybridization efficiency and detection accuracy. Moreover, their concentrations in body fluids are typically low, necessitating the development of more sensitive and specialized detection assays. In future studies, we will focus on improving the detection limit and optimizing sample extraction and preprocessing steps, while addressing matrix-related effects and structural considerations, to achieve more reliable and sensitive miRNA quantification.

In conclusion, our assay provides a simple and quantitative platform for multiplex miRNA detection in rat liver tissue, demonstrating reproducibility and reliability, and establishing a foundation for further investigations in biological fluids.

Methods

Animals Models and Treatments

Sprague–Dawley (SD) rats were obtained from Orient Bio Inc. (Seongnam-si, Republic of Korea), and maintained under controlled temperature (21–23 °C) and humidity (45–55%) on a 12 h light–dark cycle during the study. Animal testing was carried out in accordance with the guidelines of the American Association for Accreditation of Laboratory Animal Care (AAALAC, accredited since 1998) and following the protocol approved by the Institutional Animal Care and Use Committee (IACUC) of the Korea Institute of Toxicology (protocol code, IAC-23-01-0522-0294; date of approval, 07 November 2023).

After a 1 week acclimation period, male (n = 12) and female (n = 12) SD rats aged 7 to 8 weeks were randomly assigned to three groups, including the vehicle control (VC), negative control (NC), and treatment (T) groups, with 4 rats per group. The rats in the treatment group received a daily intraperitoneal injection of 2.5 mg/kg CpG ODN 1826 (TLR9 ligand: 5′- TCCATGACGTTCCTGACGTT-3′; Bioneer, Daejeon, Republic of Korea) for three consecutive days. On day 4, the treatment group was administered 0.5 mg/kg LPS from Escherichia coli O55/B55 (Sigma-Aldrich, St. Louis, MO, USA) and subsequently sacrificed. The VC group received saline injections, whereas the NC group was injected with 2.5 mg/kg control CpG ODN 1826 (5′- TCCATGAGCTTCCTGAGCTT-3′; Bioneer) for 3 days and received saline on day 4.

Total RNA Extraction

Frozen liver tissues were cut into 50 mg pieces on dry ice and transferred to a Precellys lysing kit (Bertin Technologies, Montigny-le-Bretonneux, France). The tissues were homogenized using Precellys 24 Touch (Bertin Technologies). Total RNA from liver tissues was extracted using TRIzol reagent (Invitrogen, ThermoFisher Scientific, Waltham, MA, USA) in accordance with the manufacturer’s instructions. The spike-in cel-miR-39 was added at a concentration of 119.2 pg/μL in 3.5 μL during the chloroform extraction step.

RNA purity was assessed by measuring the 260/280 and 260/230 absorbance ratios using a microplate reader (SpectraMax M3; Molecular Devices, Sunnyvale, CA, USA). The RNA concentration was quantified using a QuantiFluor RNA system (Promega, Madison, WI, USA).

Bead-Based Multiplex Assay Procedure

Nine MagPlex-TAG microspheres were obtained from Luminex (Austin, TX, USA). All miRNAs and biotin-labeled DNA/RNA chimeric probes were synthesized by Bioneer, and their sequences are listed in Table . The bead-based multiplex assay was developed by adapting the miRNA assay provided by the manufacturer, with modifications and optimizations to fit our specific experimental conditions. The analysis process is summarized as follows.

5. Sequences of Nine miRNAs and Chimeric Probes .

miRNA	mature miRNA sequence (5́ - 3́)	chimeric probes sequence (5́ - 3́)
rno-miR-21–5p	UAGCUUAUCAGACUGAUGUUGA	Biotin-UCAACAUCAGUCUGAUAAGCUAaatcaacacacaataacattcata
rno-miR-24–3p	UGGCUCAGUUCAGCAGGAACAG	Biotin-CUGUUCCUGCUGAACUGAGCCAcataatcaatttcaactttctact
rno-miR-29a–3p	UAGCACCAUCUGAAAUCGGUUA	Biotin-UAACCGAUUUCAGAUGGUGCUActttctcatactttcaactaattt
rno-miR-34a–5p	UGGCAGUGUCUUAGCUGGUUGU	Biotin-ACAACCAGCUAAGACACUGCCActaaacatacaaatacacatttca
rno-miR-122–5p	UGGAGUGUGACAAUGGUGUUUG	Biotin-CAAACACCAUUGUCACACUCCAatactttacaaacaaataacacac
rno-miR-125a–5p	UCCCUGAGACCCUUUAACCUGUGA	Biotin-UCACAGGUUAAAGGGUCUCAGGGAatctcaattacaataacacacaaa
rno-miR-146a–5p	UGAGAACUGAAUUCCAUGGGUU	Biotin-AACCCAUGGAAUUCAGUUCUCAtacttctttactacaatttacaac
rno-miR-155–5p	UUAAUGCUAAUUGUGAUAGGGGU	Biotin-ACCCCUAUCACAAUUAGCAUUAAcaaacaaacattcaaatatcaatc
cel-miR-39–5p	AGCUGAUUUCGUCUUGGUAAUA	Biotin-UAUUACCAAGACGAAAUCAGCUttaacaacttatacaaacacaaac

^aChimeric probe sequences: RNA (uppercase), DNA (lowercase).

For analysis, total RNA was prepared at a concentration of 2.2 μg/μL. A reaction mixture for each sample was prepared using 2.5 μL of total RNA or a blank (hybridization buffer), 16.25 μL of hybridization buffer (10 mM Tris, 200 mM sodium acetate, 5 mM EDTA, 0.05% Tween 20, pH 7.7), and 1.25 μL of 20 nM mixed chimeric probes in a 96-well PCR plate. The 96-well plate was then placed in a thermal cycler set to the StepOnePlus Real-Time PCR (Applied Biosystems, Thermo Fisher Scientific). The samples were incubated at 90 °C for 30 min, followed by a gradual decrease to 80 °C at a rate of 5 °C every 6 min until the temperature reached 50 °C. Afterward, 4 μL of mixed beads (1000 beads/region) was added to the samples and allowed to conjugate with the chimeric probes for 30 min at 37 °C. This was followed by a 1:500 dilution of RNase inhibitor (Promega) for 30 min at 30 °C to eliminate any unmatched chimeric probes. The samples were then transferred to black/clear flat bottom plates (Costar, Thermo Fisher Scientific) and placed on a magnetic washer (Bio-Rad Laboratories, Inc., Hillsborough, CA, USA) to remove the supernatant. Subsequently, 75 μL of a 1:100 diluted SAPE solution (Invitrogen) was added and incubated for 30 min at RT. The plate was subsequently washed twice with hybridization buffer, after which 80 μL of phosphate-buffered saline (pH7.4; Gibco, Thermo Fisher Scientific) was added to the plate. The MFI was measured using the Bio-Plex 200 (Bio-Rad Laboratories, Inc.).

Method Validation Parameters

The assay was validated according to the FDA guidelines for bioanalytical methods.

Calibration Curves and QCs

Synthetic miRNA oligonucleotides (Bioneer) were used as calibration standards and QC samples. Ten consecutive concentrations with 2-fold serial dilutions of each miRNA were ranged from 0.10 pg/μL to 51.20 pg/μL.

A 5-PL model was used to construct a sigmoid curve in Bio-Plex Manager software v6.2 (Bio-Rad). To determine the optimal standard, Bio-Plex Manager software v6.2 (Bio-Rad) automatically converted MFIs for the corresponding miRNA concentrations. The acceptance criteria state that nonzero calibrators, excluding the anchor, should have a %RE within ±20% and a %CV of ≤20% of the nominal concentrations in each run. For the LLOQ and ULOQ, the calibrators should have a %RE within ±25% and a %CV of ≤25% of the nominal concentrations.

QCs were prepared at five concentrations: 0.20 pg/μL (LLOQ), 0.60 pg/μL (LQC), 12.80 pg/μL (MQC), 40.00 pg/μL (HQC), and 51.20 pg/μL (ULOQ), corresponding to the ranges of the calibration curve.

Accuracy and Precision

The accuracy and precision were evaluated by analyzing the QCs and calculating the %RE and %CV from the observed concentration of each QC. Between-run accuracy and precision were assessed by performing multiple analyses (n ≥ 3) of each QC across 6 different runs.

For each QC level, except the LLOQ and ULOQ, replicate measurements were evaluated to demonstrate an accuracy (%RE) within ±20%, a precision (%CV) of ≤20%, and a total error of ≤30%. The total error is the absolute value of (%RE + %CV). In the case of the LLOQ and ULOQ, the accuracy (%RE) of the replicate measurements was established to be within ±25%, with a precision (%CV) of ≤25% and a total error of ≤40%.

Specificity

LQC and HQC containing a mixture of nine miRNAs were prepared, and the signal values of the QC samples (n = 3) corresponding to each single-plex assay were extrapolated to their respective calibration curves. The specificity was verified by confirming whether the QCs met the acceptance criteria of %RE (±20%) and %CV (≤20%).

Carry-Over

To evaluate carry-over, the high concentration sample of the calibration curve was analyzed, followed by the analysis of a blank sample. The MFI signal of each blank sample for the nine miRNAs was lower than the respective LLOQ MFI signal within the calibration curve for each miRNA.

cDNA Synthesis and qPCR

To analyze the miRNA, isolated total RNA was subjected to poly­(A) tailing by polyadenylation of 0.5 μg of RNA with 5 U of poly­(A) polymerase (Ambion, Thermo Fisher Scientific) at 37 °C for 1 h. Poly­(A)-tailed miRNAs were reverse-transcribed using M-MLV reverse transcriptase (Invitrogen, Thermo Fisher Scientific) following the manufacturer’s instructions. For the initial step, 1 μg of poly­(A)-tailed miRNAs was mixed with 1 μL of a 10 mM RT linker (CTGTGAATGCTGCGACTACGA-18 dTs) in a total volume of 9.5 μL, and the mixture was heated at 65 °C for 5 min to disrupt secondary structures. The reverse transcription reaction was then initiated by adding 5 μL of 10 mM dNTPs, 4 μL of 5× RT buffer, 0.5 μL of RNase inhibitor, and 1 μL of M-MLV reverse transcriptase. The samples were incubated at 42 °C for 90 min, followed by enzyme inactivation at 94 °C for 2 min. The resulting cDNA served as a template for real-time quantitative PCR using SYBR Green Universal Master Mix (Applied Biosystems, Thermo Fisher Scientific) on a StepOnePlus Real-Time PCR System (Applied Biosystems, Thermo Fisher Scientific). The relative expression levels of the miRNAs were analyzed using the 2^–ΔΔCt method, with all miRNA expression levels normalized to that of 5S rRNA. All primers, except for cel-miR-39, were designed using rat-derived sequences and are shown in Table S1.

Statistical Analysis

The results are expressed as the mean ± standard deviation (SD). GraphPad Prism version 8.0 (GraphPad Software, San Diego, CA, USA) was used for all the statistical analyses. Normality was assessed using the Shapiro–Wilk test. The homogeneity of variances was evaluated using Brown–Forsythe test. For normally distributed data with equal variances, one-way ANOVA was performed, followed by Tukey’s HSD post hoc test to identify group differences. Statistical significance was defined as a p value of less than 0.05.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c09640.

• Primer sequences of nine miRNAs (Table S1) (PDF)

The authors declare no competing financial interest.