Factors influencing degradation kinetics of mRNAs and half-lives of microRNAs, circRNAs, lncRNAs in blood in vitro using quantitative PCR

RNAs are rapidly degraded in samples and during collection, processing and testing. In this study, we used the same method to explore the half-lives of different RNAs and the influencing factors, and compared the degradation kinetics and characteristics of different RNAs in whole blood and experimental samples. Fresh anticoagulant blood samples were incubated at room temperature for different durations, RNAs were extracted, and genes, including internal references, were amplified by real-time quantitative PCR. A linear half-life model was established according to cycle threshold (Ct) values. The effects of experimental operations on RNA degradation before and after RNA extraction were explored. Quantitative analysis of mRNA degradation in samples and during experimental processes were explored using an orthogonal experimental design. The storage duration of blood samples at room temperature had the greatest influence on RNA degradation. The half-lives of messenger RNAs (mRNAs) was 16.4 h. The half-lives of circular RNAs (circRNAs), long non-coding RNAs (lncRNAs) and microRNAs (miRNAs) were 24.56 ± 5.2 h, 17.46 ± 3.0 h and 16.42 ± 4.2 h, respectively. RNA degradation occurred mainly in blood samples. The half-life of mRNAs was the shortest among the four kinds of RNAs. Quantitative experiments related to mRNAs should be completed within 2 h. The half-lives of circRNAs and lncRNAs were longer than those of the former two.


Materials and methods
Treatment of experimental specimens. Fresh anticoagulant whole blood was used as experimental samples. Next, genomic RNA was extracted using a high efficiency blood RNA extraction kit (Beijing TIAN-GEN Biotechnology, batch number DP190813) according to the instructions supplied with the kit. After extraction, RNA concentration and purity were calculated using a microspectrophotometer (Vastech Inc.Wilmington, USA) according to the absorbance at 260 and 280 nm (A260 and A280). Reverse transcription was carried out in a 20 µL system with primers and reverse transcriptase (The system contains primers, RNA templates, 2 × ES Reaction Mix, RT Enzyme Mix and gDNA Remover). A 1 µL sample of the reverse transcription product was used for real-time quantitative PCR, and each gene was amplified by PCR. Primer sequences are shown in Table 1. Real-time quantitative PCR reaction system is shown in Table 2.
Quantitative performance evaluation of detection system. Detection using a high efficiency blood total RNA extraction kit. Whole blood was diluted to a series of concentrations, RNA was extracted using a high efficiency total blood RNA extraction kit, and a linear relationship between RNA concentration and dilution ratio was observed. The experiment was repeated three times. Real-time fluorescence quantitative PCR. Extracted RNA was reverse-transcribed into cDNA and diluted to a series of concentrations. The GAPDH gene was amplified by real-time quantitative PCR, and a linear relationship between its Ct value and dilution ratio was observed. The experiment was repeated three times.
The influence of three factors on RNA degradation rate determined by an orthogonal method. Three factors and three levels of orthogonal experimental methods were employed. The three factors were the storage time of fresh whole blood at room temperature, the storage time of RNA at room temperature, and the storage time of cDNA after reverse transcription at − 20 °C. The Ct values of internal reference genes were determined by real-time quantitative PCR and the results were analysed by multivariate ANOVA and regression analysis. The optimal combination was obtained according to the results of orthogonal experiments. Whole blood was diluted 1:2, 1:4, 1:8, 1:16 and 1:32 with normal saline. RNA extraction and other related experiments were carried out under optimal combination conditions. GAPDH, β-actin and U6 genes were amplified by real-time quantitative PCR.

Comparison of the half-lives of mRNAs and ncRNAs.
When whole blood samples were incubated at room temperature, the RNA concentration gradually decreased with time. The half-life was calculated using a microspectrophotometer and real-time fluorescence quantitative PCR. Firstly, according to the equation N t = 2 (Ct0−Ctt) , the relative template concentration (N t ) was calculated from the Ct values (t represents the t th 12-h period of incubation after the specimen is placed; Ct 0 represents the Ct value at 0 h of specimen placement; Ctt represents the Ct value of the specimen after it has been placed for a period of time). K derived from the linear regression equation between the natural logarithm of the concentrations (N t ) and the incubation time, is the slope of the equation. The half-life(T 1/2 ) was obtained using the equation 36 Statistical analysis. All data were analysed using SPSS17.0 statistical software 38 . Multi-factor ANOVA was performed on the orthogonal test results. The storage time of fresh whole blood at room temperature, the storage time of RNA at room temperature, and the time of reverse transcription after cDNA storage at − 20 °C were considered as independent variables, and the concentration was taken as a dependent variable to investigate the effects of the three factors on RNA degradation at different levels. A p-value < 0.05 indicated statistically significant differences.

Results
In order to evaluate the quantitative performance of the detection system, linear observation was carried out. The results showed revealed a good linear relationship between the natural logarithm of the concentration value, the Ct value, and the natural logarithm of the dilution factor, with correlation coefficient R of 0.9972, 0.9931, 0.9925, 0.9987, 0.9997, 0.9993, 0.9892, 0.9759 and 0.9914, respectively, all greater than 0.95 (Fig. 1). After statistical calculation, p-values of the three groups were all less than 0.001, showing statistical differences. Therefore, the experimental results obtained from subsequent experiments based on the above experimental instruments were reliable.  Table 3. Direct comparison of the nine experimental results in Table 3 shows that the Ct value of Experiment 1 was the smallest. With increasing time, the Ct value of genes gradually increased (i.e., the more factors involved, the greater the impact on RNA degradation, and the faster RNA degradation). Multivariate ANOVA and linear regression analysis were conducted for the above orthogonal experiments, and the analysis results are shown in Tables 4 and 5.
In the table comparing inter-subject effects, we compared the p-values of fresh whole blood storage time at room temperature, RNA storage time at room temperature, and cDNA storage time at − 20 °C, and the p-values of the three influencing factors were all less than 0.05, and less than 0.01 for two of the influencing factors. Thus, these three factors had significant effects on RNA degradation, and the differences were statistically significant. However, there were some differences in the degree of influence. According to the regression analysis coefficient results, the storage time of RNA at room temperature had the least influence on RNA degradation, while the storage time of fresh whole blood at room temperature had the largest influence on RNA degradation. The effects   According to the results of orthogonal experiments, the optimal combination of factors in this experiment was A1B1C1 (i.e., when the storage time of fresh whole blood at room temperature, the storage time of RNA at room temperature, and the storage time of reverse transcription cDNA at − 20 °C were all 0 h, the Ct value of genes was lowest). Therefore, in the process of RNA extraction and reverse transcription from whole blood, the time should be shortened as soon as possible, otherwise the subsequent results will be affected to a certain extent. Furthermore, blood was diluted to a series of concentrations for RNA extraction, followed by the amplification of GAPDH, β-actin, and U6 genes, and Ct values were recorded, all under optimal conditions. Taking the natural logarithm of the relative template concentration as a linear relationship with the dilution ratio, the correlation coefficient R values were 0.95, 0.99 and 0.93, respectively, all greater than 0.9, showing a good linear relationship.
Half-life calculation. For half-life calculation based on microspectrophotometry, fresh anticoagulant whole blood was placed at room temperature, RNA was extracted every 12 h, and the concentration was measured by microspectrophotometry. The results are shown in Fig. S1A. Linear regression analysis of the natural logarithm of concentration and time yielded the concentration equation y = − 0.0481x + 4.172, R = 0.972 (Fig. S1B), and the calculated half-life was 14.4 h.
For half-life calculation based on the real-time fluorescence quantitative PCR method, the degradation data at room temperature are shown in Tables 6, 7 and 8. The RNA concentration in whole blood decreased with increasing incubation duration. The half-life was calculated by performing linear regression analysis on the natural logarithm of the initial concentration and the incubation duration. The regression analysis diagram is shown in Fig. 2. The correlation coefficients (R) of all linear fitting equations are greater than 0.9. The R and the equations were shown in Table 9. And p-values of each group were all less than 0.05, showing statistical differences. The half-lives of mRNA and miRNA in whole blood were calculated by the formula to be 16.4 h and 16.42 ± 4.2 h, respectively, indicating that the half-lives of mRNA were similar to those of miRNA. The half-lives of circRNA and lncRNA in whole blood were 24.56 ± 5.2 h and 17.46 ± 3.0 h, respectively. When the confidence

Discussion
In this study, a fast and reliable RNA half-life measurement method was established, based on quantitative real-time fluorescence PCR and supplemented by microspectrophotometry. Quantitative real-time fluorescence PCR can quickly and accurately determine the copy number of molecules in samples. qPCR was employed to detect changes in RNA content after incubating samples for different time periods, and the half-lives of mRNA and ncRNAs were compared 39 . Because incomplete or degraded RNA may interfere with the accuracy of RNA half-life detection, the length of the amplified fragment is important to determine the RNA half-life 40 . Regarding genes, we selected β-actin, GAPDH and other housekeeper genes as internal reference sequences. Since their expression levels in cells or copy number in the genome is constant, and is less affected by environmental factors, their quantitative results reflect the number of cells or genomes contained in the sample 41 .
In order to avoid errors caused by manual operation and differences in reagents and instruments, we conducted quantitative performance evaluation of the methods involved in the experimental process before the Table 7. Changes in lncRNAs concentration in whole blood at room temperature. The 95% confidence interval for T 1/2 was calculated to be (14.830, 20.090).    www.nature.com/scientificreports/ experiment. And the correlation coefficient R values were greater than 0.95, with good reliability (Fig. 1), and these good linear relationships laid a foundation for subsequent experiments.
In the analysis of key influencing factors, we employed an orthogonal experiment to investigate multiple factors accurately. The approach uses a set of normalised orthogonal tables to conduct tests, and the experimental results are subjected to statistical analysis to draw scientific conclusions. This design method is used for studying multi-factors and multi-levels rapidly and conveniently 42 . The orthogonal experiment results showed that the greater the number of factors involved in the experiment, the greater the impact on RNA degradation, and the faster RNA degradation. Multivariate ANOVA in the orthogonal experiment showed that the p-values of the three influencing factors were all less than 0.05, and p-values for two of the influencing factors were less than 0.01 (Table 4), indicating that the three factors had a significant impact on RNA degradation, and the differences were statistically significant. By comparing the standardisation coefficient, it could be seen that the storage duration of RNA at room temperature had the least effect on RNA degradation, while the storage duration of fresh whole blood at room temperature had the largest effect on RNA degradation. Therefore, we studied the storage duration as a factor, and investigated the half-life of mRNA, miRNA, lncRNA and circRNA in whole blood at room temperature.
When processing actual samples, especially during mass inspection, experimentalists often ignore the effects of intermediate processes from specimen collection to specimen treatment on RNA concentration 3 . Therefore, we chose room temperature as the most likely incubation temperature for blood samples after collection. In view of the fact that the target gene fragment could not be amplified after whole blood samples were incubated for a week in preliminary experiments, we tested observation times of 0 h, 12 h, 24 h, 36 h, 48 h, and 60 h. The half-life was calculated by two methods (microspectrophotometry and qPCR). First, RNA concentration was measured by microspectrophotometry, we found that RNA degradation was fastest in the first 24 h (the slope of the dotted line was the greatest), and the slope of the dotted line became more gentle after 24 h. This shows that RNA is particularly prone to degradation which is consistent with literature reports 43,44 . The linear equation was calculated and the half-life of whole blood RNA at room temperature was 14.4 h. Second, using real-time quantitative PCR, the relative concentration was calculated according to the formula after taking the median Ct value, linear regression analysis was performed on the natural logarithm of the concentration vs. time, and the correlation coefficients were all greater than 0.9 (Table 9). Therefore, we can say that at room temperature, the degradation of RNA in whole blood after a defined period of time conforms to the first-order kinetic law. The experimental results showed that the half-life of mRNA was 16.4 h, and the half-life of miRNA was 16.42 ± 4.2 h. Thus, the half-life of mRNA was shorter than that of miRNA, but only slightly. The half-lives of circRNA and lncRNA were longer than those of the other two. Therefore, circRNAs are relatively stable, which is consistent with literature reports 45,46 . However, we found that there seems to be no relationship between the length of the amplified fragments and stability because even though the amplified fragments are similar in length, their halflives are still very different (Tables 6,7,8).
In this experiment, the half-life results obtained by real-time fluorescence quantitative PCR and microspectrophotometry were essentially similar, and mRNAs and ncRNAs were measured simultaneously on the same platform. Therefore, we concluded that the storage method for samples directly affected the quality of RNA in the stored samples. RNA in blood is easily degraded during improper storage, and RNA is rapidly degraded by RNases present in blood or in the external environment. Storage at room temperature has a greater impact on RNA degradation 47 .
The mRNAs and ncRNAs studied in the present work were from blood samples. Most previous studies have focused on tissue samples, but blood samples have advantages including being clearer, non-invasive, convenient for sampling, and amenable to continuous detection 1 . With the development of biotechnology, detection methods will become more and more sensitive, hence it is particularly important to reduce variation during pre-treatment of specimens 48 . Due to the poor stability and short half-life of RNA, it is of great significance to select appropriate and professional methods to preserve blood samples to ensure the accuracy and reliability of test results. Especially in clinical trials, the acquisition of high-quality experimental samples is very important, hence more attention should be paid to the impacts of specimen analysis and processing 49 .

Data availability
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.