Expression of microRNAs in leukocytes and serum of asbestosis patients

Participants of the underlying study comprised 36 male individuals with an average age of 73 years affected either by asbestos-related pleural plaques (J92.0; ICD-10-GM version 2022) or by asbestosis (J61 and J92.0; ICD-10-GM version 2022) and 15 healthy individuals who served as controls. The patient collective comprised male individuals since only men with occupational asbestos-related changes in pleura and/or lung presented to our clinic, which can be explained by the larger proportion of men working in the construction industry. The asbestosis group included patients who showed asbestos-related changes in both, pleura and lungs. Diagnosis was based on physical examination, pulmonary function testing, hematological and biochemical analyses of blood samples and chest X-ray radiography. X-rays of pleura and lungs were evaluated according to the International Classification of Occupational and Environmental Respiratory Diseases (ICOERD) and to the International Classification of Radiographs of Pneumoconiosis of the International Labour Organization (ILO) [49]. The latter is shown in Table 1.

According to the ILO or ICOERD, patients suffering from pleural plaques or asbestosis were classified into three groups considering the degree of severity (mild, moderate or severe).

The control group included 1 female and 14 male participants with an average age of 37 years who were not exposed to asbestos in the past and underwent a preventive medical check-up in our clinic.

The relatively young age of participants in the healthy control group compared to the group of patients affected by asbestos-related pleural plaques or by asbestosis, is based on the circumstance that the majority of older people who presented to our clinic, showed comorbidities (e.g., hypertension or type II diabetes mellitus). Unhealthy people were excluded from the control group. Besides, asbestos-related pleural plaques or asbestosis rarely emerges in younger people due to the long latency period. Hence, participants affected by asbestos-related pleural plaques or asbestosis were older compared to healthy control participants.

For serum microRNA analysis, whole blood was drawn from all 36 patients of whom 26 patients were affected by pleural plaques and 10 patients affected by asbestosis as well as from 15 healthy controls who were not exposed to asbestos. In reference to ICOERD and ILO, 21 of these patients (18 pleural plaques and 3 asbestosis) were mildly diseased (ILO: pleural changes 1 and/or parenchymal changes 1/1 and 1/2), 9 patients (5 pleural plaques and 4 asbestosis) were moderately diseased (ILO: pleural changes 2 and/or parenchymal changes 2/1, 2/2 and 2/3) and 6 patients (3 pleural plaques and 3 asbestosis) were severely diseased (ILO: pleural changes 3 and/or parenchymal changes 3/2, 3/3 and 3/+).

Leukocytes were obtained from the whole blood of 15 control individuals and from 34 out of 36 patients of whom 24 patients were affected by pleural plaques and 10 patients affected by asbestosis. Twenty of these patients (17 pleural plaques and 3 asbestosis) were classified as mildly diseased (ILO: pleural changes 1 and/or parenchymal changes 1/1 and 1/2), 8 patients (4 pleural plaques and 4 asbestosis) as moderately diseased (ILO: pleural changes 2 and/or parenchymal changes 2/1, 2/2 and 2/3) and 6 patients (3 pleural plaques and 3 asbestosis) as severely diseased (ILO: pleural changes 3 and/or parenchymal changes 3/2, 3/3 and 3/+).

Blood samples, which were analyzed for serum microRNA but not for leukocytes microRNA belonged to two patients with pleural plaques, one with mild disease and the other with moderate disease.

Blood sampling was approved by the local ethics commission of the department of medicine at the Justus-Liebig-University of Giessen, Germany (103/05 and 75/06). Patients gave written consent for their blood analysis in the underlying study. In total 36 patients and 15 healthy individuals were included in the study.

Whole blood was drawn by venipuncture on the inside of the elbow.

Blood samples of all 36 patients who were affected by either pleural plaques or asbestosis and 15 control individuals were analyzed for serum microRNAs as well as blood of 34 patients and 15 controls for leukocytes microRNAs. For leukocytes microRNA analysis three milliliters (ml) of whole blood was collected in Sarstedt Monovettes containing ethylenediaminetetraacetic acid (EDTA, Sarstedt, Nürmbrecht, Germany).

For serum analysis, 7 ml of whole blood was collected in Sarstedt S-Monovette® serum tubes (Sarstedt) without EDTA.

All blood samples were processed within 30 minutes (min) after collection. They were centrifuged for 10 min at 1900×g and 4 °C. Subsequently, separated plasma or serum was transferred into the respective RNase-free micro tubes (Sarstedt). To ensure cell-free plasma or serum samples, samples were centrifuged for another 10 min at 16,000×g and 4 °C and supernatants transferred into fresh RNase-free micro tubes (Sarstedt) before being stored at −80 °C pending further processing. From plasma separated EDTA blood was processed immediately for RNA isolation from leukocytes. Plasma samples were not analyzed for the underlying study.

Prior to RNA isolation, erythrocytes in the cellular portion of blood obtained after centrifugation as described above, were lyzed using maximally 1.5 ml of blood cells and 10 ml of erythrocytes lysis buffer (Buffer EL, Qiagen, Hilden, Germany). The mixture was incubated in RNase-free reaction tubes for 15 min on ice and then centrifuged for 10 min at 400×g and 4 °C. Subsequently, supernatants were removed, 5 ml of erythrocyte lysis buffer added to the pellet, resuspended and the mixture centrifuged again for 10 min at 400×g and 4 °C. Supernatants were removed again, the cell pellet resuspended in 1 ml erythrocyte lysis buffer, samples bisected and 500 microliters (µl) each transferred into two RNase-free reaction tubes (Sarstedt). Another 500 µl of erythrocyte lysis buffer was added to each tube and cells centrifuged for 10 min at 400×g and 4 °C. Supernatants were removed, leukocytes homogenized in 700 µl Qiazol (Qiagen) and incubated for 5 min at room temperature. Then 140 µl of chloroform (Carl Roth, Karlsruhe, Germany) was added, the samples were vigorously shaken or vortexed for 15 seconds (s) and then incubated for 3 min at room temperature. The mixtures were centrifuged for 15 min at 12000×g and 4 °C. Clear upper phase was transferred into a fresh RNase-free reaction tube (Sarstedt), 1.5 volumes of 100% ethanol added and RNA isolated using the miRNeasy Mini Kit (Qiagen) according to the manufacturer’s protocol. Isolated RNA was eluted in 30 µl RNase-free water and stored at −80 °C pending further processing.

RNA was isolated from 100 µl of blood serum. Serum was mixed with 500 µl Qiazol (Qiagen) and incubated for 5 min at room temperature. After homogenization of serum samples, 3.5 µl Ce-miR-39_1 Spike-In Control (Qiagen) diluted to 1.6 × 10⁸ in RNase-free water containing 10 ng/µl carrier RNA from bacteriophage MS2 (Roche) was added to serum lysates to monitor microRNA recovery and reverse transcription efficiency in the following steps. Subsequently, 100 µl of chloroform was added and the samples were vigorously shaken or vortexed for 15 s and incubated for 2–3 min at room temperature. Afterwards, samples were centrifuged for 15 min at 12000×g and 4 °C. Clear upper phase was transferred into a fresh RNase-free reaction tube (Sarstedt) and 1.5 volumes of 100% ethanol added before RNA was isolated using the RNeasy Mini Elute columns and the miRNeasy Serum/Plasma Kit (Qiagen) according to manufacturer’s protocol.

cDNA was synthesized by reverse transcription of 250 ng of purified RNA from leukocytes or 1.5 µl of purified RNA from serum containing Ce-miR-39_1 Spike-In Control. RNA samples were reverse transcribed into cDNA for 24 hours at 37 °C using the miScript II RT Kit (Qiagen) with HiSpec buffer (Qiagen) and a cycler (Mastercycler gradient, Eppendorf, Hamburg, Germany). Volumes of components used for the reverse transcription reaction mix corresponded to manufacturer’s instructions. Two-hundred µl of RNase-free water were added to cDNA samples and microRNA expression was analyzed using real-time reverse transcription (RT)-polymerase chain reaction (PCR).

The miScript miRNA PCR Array Human Fibrosis (MIHS-117Z, Qiagen), which comprised 84 microRNAs expressed in fibrosis and additional normalization controls, reverse transcription controls as well as positive PCR controls, was used to pre-screen for microRNAs expressed in leukocyte or serum samples of two patients affected by moderate or severe pleural plaques, respectively. The miScript miRNA PCR Array Human Fibrosis was placed in a C1000 Touch™ Thermal Cycler with a CFX96 Real-time PCR Detection System (Bio-Rad, Munich, Germany) and cycling conducted according to manufacturer’s protocol as described below.

After pre-screening for microRNAs expressed, real-time RT-PCR analyses of all patient samples were performed with the miScript SYBR Green PCR Kit (Qiagen) and miScript Primer Assays (Qiagen) in a Lightcycler System (Roche Diagnostics, Mannheim, Germany). MicroRNAs analyzed in leukocytes and serum were hsa-miR-32-5p (Accession No. MIMAT0000090), hsa-miR-143-3p (Accession No. MIMAT0000435), hsa-miR-145-5p (Accession No. MIMAT0000437), hsa-miR-146b-5p (Accession No. MIMAT0002809), hsa-miR-204-5p (Accession No. MIMAT0000265), and hsa-miR-451a (Accession No. MIMAT0001631). Volumes of components used for the real-time RT-PCR reaction mix corresponded to manufacturer’s instructions.

The cycling procedure started with an initial activation step at 95 °C for 15 min, followed by 40 cycles of denaturation at 94 °C for 15 s, annealing at 55 °C for 30 s and extension at 70 °C for 30 s. A melting curve was performed to check the purity of RT-PCR products. Negative controls received water instead of cDNA. For evaluation of microRNA expression in leukocytes or serum, delta cycle threshold (CT) values were analyzed using the pseudogene hsa-RNU6-6P as reference (Accession No. NG_034215) or the Spike-In Control Ce-miR-39_1 (Accession No. MIMAT0000010), respectively.

Table 2 summarizes all primers used for real-time RT-PCR.

For the evaluation of individual results of patients suffering from pleural plaques or asbestosis, delta CT values were compared to the 90% confidence interval of delta CT values of the healthy control group who were not exposed to asbestos. Values of diseased patients below or above the range of the 90% confidence interval of the healthy control group were considered up- or down-regulated.

Statistical analyses and the generation of graphs were carried out using the statistics program SPSS (version 28.0; SPSS Institute Inc., Chicago, USA). The Kolmogorov–Smirnov test was applied to determine the normal distribution of delta CT values of the leukocyte analysis. Since real-time RT-PCR delta CT values of miR-143-3p, miR-145-5p, miR-146b-5p and miR-204-5p were normally distributed, one-way analysis of variance (one-way ANOVA) with Bonferroni correction was performed. Delta CT values of miR-32-5p and miR-451a did not meet the normal distribution. Therefore, the Kruskal–Wallis test was used.

Delta CT values of miR-145-5p and miR-451a in serum samples were normally distributed and evaluated by using one-way ANOVA with Bonferroni correction. For the comparative analysis of delta CT values of miR-145-5p and miR-451a between leukocytes and serum, which were not normally distributed, the Kruskal–Wallis test was applied.

To evaluate the effect of statistically significant differences, effect sizes [partial eta-squared (η²_p) and Cohen’s f] according to one-way ANOVA were determined. η²_p values < 0.06, 0.06–0.14 or > 0.14 and effect sizes according to Cohen’s f of 0.10, 0.25 or 0.4 indicate small, medium or large effects, respectively. Small, medium or large effect size for the evaluation of Kruskal–Wallis test (r) are indicated by 0.1 ≤ r < 0.3, 0.3 ≤ r < 0.5 or r > 0.5, respectively. Moreover, values of difference and the 95% confidence interval were calculated for differences between groups.

A receiver operating characteristic (ROC) curve was performed and the area under the curve (AUC) calculated to evaluate the ability to discriminate between patients suffering from pleural plaques and healthy control participants using miR-146b-5p expression.

The coefficient of determination (R²) was calculated to evaluate regression regarding miR-145-5p or miR-451a expression between leukocytes (dependent variable) and serum (independent variable) using linear regression analysis [scatter (XY) plots]. An R² value of 0 indicates no correlation and an R² of 1 absolute correlation.

The microRNAs miR-32-5p, miR-143-3p, miR-145-5p, miR-146b-5p, miR-204-5p and miR-451a, known to be involved in fibrosis, were detected in leukocytes of patients suffering from pleural plaques or asbestosis as well as in the healthy control group.

Delta CT values of miR-146b-5p in leukocytes were significantly down-regulated in patients affected by pleural plaques, but not in patients affected by asbestosis when compared to controls (Fig. 1D). The effect size calculation for miR-146b-5p down-regulation resulted in η²_p = 0.150 and f = 0.42, indicating a large effect. The value of difference was 0.725 and the 95% confidence interval was 0.070–1.381 when comparing healthy controls to patients suffering from pleural plaques. No significant differences in microRNA expression were detected comparing patients affected by pleural plaques to patients suffering from asbestosis. Expression of miR-32-5p, miR-143-3p, miR-145-5p, miR-204-5p and miR-451a showed slight differences between patients affected by pleural plaques or asbestosis as well as in controls. However, differences were not statistically significant between all three groups (Fig. 1A–C and E–F).

Figure 2 displays in how many patients each analyzed microRNA was either up- or down-regulated or else in the range of the 90% confidence interval of controls in leukocytes. All six microRNAs analyzed were detectable. In most patients affected by the disease, these microRNAs were down-regulated. Several patients showed microRNA expression in the range of healthy non-exposed control individuals (Fig. 2).

Table 3 shows the individual variability of microRNA expression, summarized as frequency of up- or down-regulated microRNAs or unchanged expression compared to controls in leukocytes of each patient. In contrast to the majority of patients in whom microRNAs were down-regulated, in six patients most microRNAs analyzed were up-regulated (Table 3).

To compare microRNA expression in respect to mild, moderate or severe disease, patients suffering from pleural plaques or asbestosis were pooled in the respective group. MicroRNA miR-146b-5p was significantly down-regulated in mildly diseased patients compared to the control group (Fig. 3D). Results of the effect size calculation for miR-146b-5p down-regulation indicated a large effect with η²_p=0.178 and f=0.465. The value of difference was 0.848 and the 95% confidence interval 0.097–1.599 when comparing healthy controls to mildly diseased patients. No statistically significant differences were detectable when comparing miR-146b-5p expression of mildly diseased patients to that of moderately or severely diseased patients (Fig. 3D). Moreover, expression of microRNAs miR-32-5p, miR-143-3p, miR-145-5p, miR-204-5p and miR-451a did not differ in regard to the extent of pleural plaques or severity of asbestosis (Fig. 3A–C and E–F).

To evaluate discrimination ability between patients suffering from pleural plaques and healthy controls using miR-146b-5p, ROC curve analysis was performed and the area under the ROC curve (AUC) was determined. ROC curve and area under the curve (AUC) allow discrimination between diseased and healthy patients. Generally, a ROC curve located closer to the upper left-hand corner of the image indicates greater discrimination ability between diseased and healthy individuals. The area under the curve (AUC) summarizes the overall accuracy of the test and ranges from 0 to 1 (two-dimensionally). A value of 0 indicates low accuracy and a value of 1 indicates high accuracy, the latter correlates to a high ability to differentiate between diseased and healthy individuals. An AUC of 0.5 suggests no discrimination. An AUC of 0.7–0.8 is considered acceptable, 0.8–0.9 is considered excellent and above 0.9 is considered outstanding [50].

In the underlying study an AUC of 0.757 for miR-146b-5p expression in leukocytes indicated an acceptable ability to discriminate between patients suffering from pleural plaques and healthy controls (Fig. 4).

MicroRNAs miR-32-5p, miR-143-3p, miR-146b-5p and miR-204-5p were neither detectable in serum samples of patients affected by pleural plaques or asbestosis nor in the control group, but microRNA miR-451a was detectable in all serum samples analyzed. In contrast, miR-145-5p was expressed only in 10 of 26 patients exhibiting pleural plaques and in 5 of 10 asbestosis patients as well as in 10 of 15 study participants in the control group. However, microRNAs were not significantly (up- or down)-regulated in serum samples of either group, independent from the degree of severity of the disease (Fig. 5).

Still, considering the number of patients, in serum of most patients, the expression of miR-451a was down-regulated or in the range of controls as shown in Fig. 6. Nevertheless, expression levels were similar to those of the control group, resulting in no significant differences. In serum of most patients, miR-145-5p was not detected or in the range of healthy non-exposed controls (Fig. 6).

Individual microRNA regulation profiles are presented in Table 4. It shows in how many patients each analyzed microRNA was up- or down-regulated or in the range of the 90% confidence interval of the controls in serum. In most patients, microRNAs analyzed were down-regulated or not detectable.

The comparison of microRNA expression between leukocytes and serum showed significant differences. Not all microRNAs evidenced in leukocytes were detectable in serum samples. Only miR-145-5p and miR-451a were expressed in leukocytes and serum. Therefore, a comparison of expression differences between leukocytes and serum was performed for these two microRNAs exclusively (Fig. 7).

MicroRNA miR-145-5p was significantly down-regulated in serum compared to leukocytes of patients suffering from pleural plaques or asbestosis (Fig. 7A), independent from the severity of the disease (Fig. 7C) as well as in controls. Effect size (r) calculation according to the Kruskal–Wallis test for miR-145-5p down-regulation when comparing serum to leukocytes of patients suffering from pleural plaques, asbestosis or controls resulted in values of r = 0.38, 0.50 or 0.51, respectively, and indicated a medium effect. Considering disease severity, effect size for miR-145-5p down-regulation between serum and leukocytes in mildly diseased patients was r = 0.43 and indicated a medium effect as well. Neither pleural plaques nor asbestosis or degree of severity did affect the expression of miR-451a (Fig. 7B and D).

Regression analysis revealed an R² value of 0.004 for miR-145-5p and an R² of 0.043 for miR-451a, indicating no correlation between leukocytes and serum (Fig. 8).

The Pearson correlation (r) supported the result of the R² analysis and revealed values of r = 0.061 and r = 0.208 for miR-145-5p and miR-451a, respectively, indicating no correlation as well.

Partial correlation in leukocytes and serum showed a value of −0.02 for miR-145-5p and of 0.159 for miR-451a when controlling for pleural plaques or asbestosis. When controlling for disease severity, partial correlation revealed a value of −0.009 for miR-145-5p and 0.166 for miR-451a. Thus, regarding miR-145-5p and miR-451a expression, results of the partial correlation analysis confirmed no correlation between leukocytes and serum when controlling for pleural plaques, asbestosis or disease severity, accordingly.

In the present study, microRNA miR-146b-5p was significantly down-regulated in leukocytes of patients exhibiting pleural plaques. The majority of patients in this study showed a radiological mild extent of pleural plaques. However, in patients affected by asbestosis, no significant differences were detected in microRNA expression.

Pleural plaques and asbestosis are progressive diseases [53, 55]. Progression of fibrosis can change the regulation of microRNAs [56]. Interestingly, with regard to severity, miR-146b-5p was down-regulated only in patients impaired by mild pleural plaques or mild asbestosis. However, most asbestosis patients showed moderate or severe changes of pleura and lung (7 out of 10). Considering moderately or severely affected patients, miR-146b-5p was neither up- nor down-regulated in leukocytes.

Weber et al. (2017) detected no significant differences in the regulation of miR-146b-5p when comparing plasma samples of mesothelioma patients with plasma samples of asbestos-exposed control participants [57]. This result is in accordance with our finding in leukocytes, which showed no significant differences in miR-146b-5p regulation when comparing patients affected by moderately and severely asbestos-related pleural plaques or asbestosis as well as non-exposed healthy controls. Accordingly, down-regulation of miR-146b-5p in leukocytes could be interpreted as an indicator for mild asbestos-related diseases.

Interestingly, microRNA miR-146b-5p was down-regulated in patients suffering from pleural plaques, but not from asbestosis. In patients suffering from asbestosis the lung parenchyma is fibrotic, mostly accompanied by pleural plaques [58]. Lung parenchyma differs from tissue of the pleura. Whereas lung parenchyma consists of epithelial cells [59], pleura tissue comprises mesothelial cells [60]. Presumably, different cell types might display different microRNA expression patterns, which could explain that miR-146b-5p was significantly down-regulated only in patients suffering from pleural plaques, but not in asbestosis patients. Although pleural plaques were present in asbestosis patients in this study as well, the lung parenchyma is proportionally larger than the pleural tissue. Thus, effects of pleural plaques on microRNA regulation in leukocytes might have been less pronounced in asbestosis patients.

Leukocytes are of mesenchymal origin, like mesothelial cells of the pleura. Presumably, the down-regulation of miR-146b-5p only in leukocytes of patients suffering from asbestos-related pleural plaques could suggest a correlation between microRNA regulation in leukocytes and pleural plaques manifestation.

Our study revealed no statistically significant differences in expression of miR-32-5p, miR-143-3p, miR-145-5p, miR-204-5p and miR-451a in leukocytes of patients exhibiting mild, moderate or severe extent of pleural plaques or asbestosis. Nevertheless, looking at patients individually, these microRNAs were predominantly down-regulated or in the range of controls who were not exposed to asbestos.

However, in a small group of patients (6 of 34), most microRNAs were up-regulated although no explanation could be found for that finding in relation to anamnesis, disease severity, hemogram and biochemical parameters. The same parameters did not differ noticeably between patients with up- or down-regulated microRNA expression.

The regulation of certain microRNAs can differ between younger and older individuals [61]. Although age itself can influence the microRNA regulation, to the best of our knowledge, the regulation of miR-146b-5p has not been shown to be affected by increased age. Given that we did not detect significant differences in the regulation of miR-32-5p, miR-143-3p, miR-145-5p, miR-204-5p and miR-451a, we conclude that the age difference between diseased patients and healthy control participants did not influence the regulation of microRNAs analyzed in the underlying study.

In serum samples of patients affected by pleural plaques or asbestosis and control participants, miR-32-5p, miR-143-3p, miR-146b-5p and miR-204-5p were either not traceable or their expression was at a low, not evaluable detection level (CT value > 35). Therefore, they do not seem suitable for evaluating the progression of asbestos-related pleural plaques or asbestosis in serum samples.

Although all microRNAs analyzed were shown to be involved in fibrosis, in this study only miR-145-5p and miR-451a were detectable in serum samples of patients affected by asbestos-related pleural plaques or asbestosis as well as in control individuals who were not exposed to asbestos. This indicates that these microRNAs are physiologically expressed in serum and that microRNA expression in serum of patients suffering from pleural plaques or asbestosis did not differ from controls.

Damaged tissue can release microRNAs into serum [28] and there is generally a positive correlation between the quantity of microRNAs that are released and damage severity [62]. Thus, since the majority of patients in this study were only mildly affected by asbestos-related pleural plaques, less microRNAs might have been released into serum. This could explain the similar expression of miR-145-5p and miR-451a between diseased patients and healthy control individuals. Unexpectedly, in serum of patients suffering from moderate or severe asbestosis, we detected no significant differences in microRNA expression compared to non-exposed healthy controls, contradicting the positive correlation between the quantity of microRNAs that are released and damage severity. Unlike in leukocytes this indicates that disease severity did not influence microRNA expression in serum of patients suffering from asbestos-related pleural plaques or asbestosis.

Our study revealed differentially expressed microRNAs in leukocytes and serum. MicroRNA miR-146b-5p was significantly down-regulated in leukocytes but not detected in serum. Interestingly, miR-145-5p was significantly differently regulated in leukocytes and serum. MicroRNA miR-145-5p expression was higher in leukocytes than in serum.

Pascut et al. (2019) suggested that expression of microRNAs in serum could be affected by the disease and not by leukocytes [24], since tissue can release microRNAs into serum. Thus, leukocytes or serum could be analyzed independently to obtain information about the progression of the disease. Considering a correlation between higher microRNA levels in serum and increased severity of tissue damage, serum might support detection of moderate or severe asbestos-related pleural plaques or asbestosis. However, even more severely affected patients than the individuals who participated in this study were not available. Presumably, microRNA analysis in serum would be more suitable to evaluate disease progression in mesothelioma or asbestos-related lung cancer. Thus, since we could not detect significant differences in microRNA expression in serum of mildly, moderately or severely diseased patients, leukocytes seem more suitable for the microRNA analysis to evaluate progression of asbestos-related pleural plaques or asbestosis and a possible cancer risk.

MicroRNA miR-451a was equally regulated when comparing leukocytes and serum of patients affected by pleural plaques or asbestosis, and this was independent of the level of severity.

One might consider the rather small sample size of nine moderately and six severely diseased patients as well as the relatively large age gap between patients and control participants as limitations in this study. However, since asbestos-related diseases show long latency, patients usually are of advanced age. The older individuals who presented to our clinic showed comorbidities, so the control group consisted of comparably young and healthy individuals. However, the respective results that showed no significant differences when comparing older diseased participants with younger healthy control individuals suggest that age did not affect microRNA regulation in the underlying study.