Identification of miR-151a as a novel endogenous control for small extracellular vesicle cargo normalization in human cancer

doi:10.21203/rs.3.rs-2731367/v1

Download PDF

Research Article

Identification of miR-151a as a novel endogenous control for small extracellular vesicle cargo normalization in human cancer

https://doi.org/10.21203/rs.3.rs-2731367/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: Small extracellular vesicles (sEVs) in the blood of cancer patients contain higher amounts of tumor markers than those identified as free-circulating. miRNAs have significant biomedical relevance due to their high stability and feasible detection. However, there is no reliable endogenous control available to measure sEVs-miRNA content, impairing the acquisition of standardized consistent measurements in cancer liquid biopsy.

Results: In this study, we identified three miRNAs from a panel of nine potential normalizers that emerged from a comprehensive analysis comparing the sEV-miRNA profile of six lung and ovarian human cancer cell lines in the absence of or under different conditions of chemotherapy. Their relevance as normalizers was tested in 26 additional human cancer cell lines from nine different tumor types undergoing chemotherapy or radiotherapy treatment. The validation cohort was comprised of 172 prospective plasma and ascitic fluid samples from three different human tumor types. Variability and normalization properties were tested in comparison to miR-16, the most used control to normalize free-circulating miRNAs in plasma.

Conclusion: Our results indicate that miR-151a is consistently represented in small extracellular vesicles with minimal variability compared to miR-16, providing a novel normalizer to measure small extracellular vesicle miRNA content that will benefit liquid biopsy in cancer patients.

sEVs

miRNAs standardization

miRNA endogenous control in liquid biopsy

The use of micro-RNAs (miRNAs) from small extracellular vesicles (sEVs) as cancer biomarkers has attracted the interest of researchers around the world ever since Valadi et al. reported the presence of miRNAs in small extracellular vesicles in 2007 (1). MiRNAs are short non-coding RNAs of 18–25 nucleotides that regulate gene expression at the post-transcriptional level. MiRNAs bind RNAs such as messenger-RNA, or long non-coding-RNAs by sequence complementarity, inhibiting their translation and promoting their degradation (2, 3). Each mRNA, in turn, can be regulated by different miRNAs (4, 5). This makes them especially interesting since they are functional elements by themselves, able to produce changes in the expression of numerous genes (6). Specifically, their potential as biomarkers from sEVs compared to circulating miRNAs lies in their easy accessibility by non-invasive methods, greater stability, and high enrichment in the blood of cancer patients (7–9). Therefore, understanding their dynamic/expression levels, is important for a strong characterization of their use as biomarkers. SEVs are nanosized (30–150 nm) vesicles secreted by most cell types as a mechanism of cellular communication. It has been reported that their content is not arbitrary and cargo selection depends on the cell type, the physiological context and the stimuli that promote their secretion (10). For example, Zhang et al. postulated that rat cardiomyoblasts release sEVs enriched in cardioprotective miRNAs under hypoxic conditions (11), while Zhou et al. found that breast milk sEVs carry immune-related miRNAs that could be essential for infant immune system development (12). Although research focusing on the biomedical potential of miRNAs from sEV has increased exponentially over the last decade, there is still a lack of standardized measurements providing reliable comparisons between different contexts and pathologies. To date, there is no endogenous control of sEV origin on the market providing a reproducible normalization of these elements. Only the correct selection of housekeeping and pathology-specific markers will provide accurate liquid biopsy tests to be used for any basic, translational, or clinical study based on sEV-miRNAs normalization. Currently, the endogenous controls to normalize sEV-miRNAs are not specific to the sEV compartment. Moreover, few studies to date have identified the best normalizer for their specific experimental context, such as miR-221 in hepatocellular carcinoma (13). Nonetheless, its stability and that of other normalizers identified in other pathologies varies as the context changes or when including more tumor types (14). In fact, the most common endogenous controls used to normalize sEV-miRNAs in cancer as well as in other pathologies (15–17) are miR-16 and small nucleolar RNAs (snoRNAs) including U6 (or RNU6B) (18–25), often without validating their stability in sEVs. The use of miR-16 usually refers to studies performed in plasma (26, 27) and the use of RNU6B normally is not even referenced, probably assuming that its stability in the whole cell can be extended to sEVs (20, 21). In fact, RNAs derived from nuclear or cytosolic compartments of tissues are unstable in serum/plasma, which is why many studies have considered miR-16 and RNU6B as inappropriate endogenous controls for the quantification of circulating microRNAs, questioning their normalizing properties (13, 28–31). sEV biogenesis takes place in a specific compartment within the cell and, given that its content is a selected formation, it is necessary to find a specific endogenous miRNA for these vesicles.

For this study, we wanted to establish a housekeeping miRNA useful to analyze sEV-miRNA cargo. We combined Next-Generation Sequencing (small-RNAseq) and TaqMan-based qRT-PCR-specific amplifications to characterize the levels of sEV-miRNAs in 31 human cancer cell lines, the immortalized embryonic kidney cell line 293T and 172 plasma and ascites samples from cancer patients and healthy donors. miR-151a emerged as the best housekeeping miRNA in sEVs, as it presented the lowest variability rate regarding its levels in all conditions tested. Our data supports that miR-151a can be used as a reliable endogenous reference for the study of sEV-miRNAs by liquid biopsy.

Characterization of sEVs

Accurate sEVs-miRNAs analysis requires the successful isolation of sEVs. To confirm their correct isolation from culture media, plasma and ascitic fluid, vesicles were characterized using three methodologies: Transmission electron microscopy (TEM), nanoparticle tracking analysis (NTA) and western blot (WB). TEM revealed the presence of approximately 100 nm cup or spherical shape structures in cell-derived sEVs as well as in plasma and ascites fluid (Figures 1A and 1B), consistent with previously reported sEV characteristics (32-35). These results were further confirmed with NTA, including plasma-derived sEVs isolated from different NSCLC, ovarian and glioblastoma patients (Figures 1C and 1D). Once more, NTA revealed that the size of the major particles was around 100 nm (Figure 1D). WB was used as a third method for sEV characterization using antibodies targeting two common sEV markers located in sEV membranes: tetraspanin molecules TSG101 and CD81. Both markers were enriched in sEVs isolated from the cisplatin resistant H23R, A2780R and 41MR cell cultures compared to whole cell extracts, confirming the successful sEV isolation (Figure 1E).

Identification of a sEV reference miRNA signature by comprehensive bioinformatics analysis.

To find a suitable endogenous control specific to normalizing the miRNA content of tumor-derived sEVs, we performed small RNA-seq on sEV-derived total RNA from H23S/H23R, A2780S/A2780R and 41S/41R cells, comparing their miRNA profile to identify candidates that were stable within two different conditions, sensitivity, and resistance to the most frequently used chemotherapy compound, cisplatin (CDDP). In addition, we included two different tumor types (lung and ovarian) and two different cell lines from the same tissue origin (A2780 and 41M) to identify common candidates with stable levels in the sEV compartment under the conditions and sample origins tested.

For this purpose, we systematically organized the NGS data listing the miRNAs represented in all the cell lines, with at least one count per million reading in all the experimental groups tested and the lowest difference in absolute values between the resistant and sensitive phenotype for each of the cell lines (Mean Internal Variability (MIV), log2FC RvsS), obtaining a list of nine candidate miRNAs that potentially could be used as endogenous control in tumor-derived sEVs (Table 1). We also analyzed sEV levels of miR-16 and miR-451a as controls for comparison of variability. The former is typically used as a normalizer in studies of free circulating miRNAs; the second was chosen as an external miRNA to be normalized over miR16 in comparison with the novel identified miRNA-signature, as it is commonly overrepresented in the sEV compartment and it shows a high variability in different biological contexts (36) and also in the overall data obtained from our microRNome sequencing. In addition, we examined their variability and levels in 30 cancer types from The Cancer Genome Atlas (TCGA) to increase the strength of the candidates chosen (Table 1). Following the first filtering, we identified four miRNAs with the lowest MIV: miR-151a-3p, miR-22-5p, miR-502-3p and miR-221-3p, discarding the miRNAs with a variability equal or greater than 0.26 between S and R cells. Then, considering the degree of representativeness in terms of transcripts per kilobase million reads (TPM) and the variability among the 30 types of tumors analyzed in the TCGA database as subsequent filters, resulted in the selection of miR-151a-3p, miR-221-3p and miR-22-5p as those candidates with the greatest potential as endogenous sEV controls. It is important to note that the variability observed for miR-16, conventionally used as reference control, is the highest (0.477) in comparison with the list of candidates identified with less variability among all miRNAs analyzed by NGS (mean value: 0.275; 0.084; 0.329), especially when compared with miR-151a (0.084 vs 0.477), a situation that is maintained when analyzing lung and ovarian cell lines independently. As expected, this difference in variability was less marked when analyzing the variance in the sequencing results obtained from the 30 cancer types studied from the TCGA, as the latter derive from tumor analysis and not from liquid biopsy.

Table 1 List of the nine potential endogenous miRNAs identified by small RNAseq. In addition, miR-451a, the gold standard miR-16 normalizer and an external miRNA were used for normalizing purposes. R: Tumor cells resistant to platinum-based chemotherapy treatment. S: Tumor cells sensitive to platinum-based chemotherapy treatment. TPM: Transcripts per kilobase million. TCGA: Tumor Cancer Genome Atlas.
MATURE-ID	Accession Number	Length (pb)	Mean Internal Variability (\|log2FC RvsS\|)	Mean Internal Variability in ovarian cancer samples (\|log2FC RvsS\|)	Mean Internal Variability in lung cancer samples (\|log2FC RvsS\|)	TPM per sample	Log2(Mean Expression) Variance in 30 cancer types (TCGA)
hsa-miR-151a-3p	MI0000809	21	0.084	0.105	0.041	6759	0.32
hsa-miR-22-5p	MI0000078	22	0.169	0.066	0.376	39.33	0.58
hsa-miR-502-3p	MI0003186	22	0.193	0.263	0.052	39.5	0.61
hsa-miR-221-3p	MI0000298	23	0.257	0.356	0.054	851.67	1.67
hsa-miR-1183	MI0006276	27	0.276	0.35	0.129	112.5	NA
hsa-miR-27a-3p	MI0000085	21	0.278	0.296	0.242	747.67	6.67
hsa-let-7i-5p	MI0000434	22	0.297	0.441	0.011	9586.5	0.6
hsa-miR-411-5p	MI0003675	21	0.323	0.196	0.579	54.17	3.05
hsa-miR-196b-5p	MI0001150	22	0.329	0.218	0.55	87	3.67
hsa-miR-16-5p	MI0000070	22	0.477	0.347	0.737	103.34	0.58
hsa-miR-451a	MI0001729	22	3.014	3.37	2.304	8884.5	1.24

miR-151a displays the lowest coefficient of variation in secretome-derived sEV content, emerging as the best candidate for normalization in this compartment.

To validate constant levels with an alternative methodology, we performed a nonspecific retrotranscription of all miRNAs from each sample as this allows for a reduction in the bias associated with individual retrotranscriptions, followed by individual qRT-PCR for each candidate. The low variability observed by NGS for miR-151a-3p (Figure 2A), miR-22-5p (Figure 2B), miR-221-3p (Figure 2C) and miR-16 (Figure 2D) levels in the sEV compartment was confirmed in the sEVs from the secretome of the three paired cell lines H23S/R, A2780S/R and 41M/MR (Figure 2 A-F). The analysis was extended to 19 additional human cancer cell lines representing a variety of human tumor types, including OVCAR3 as an additional CDDP sensitive/resistant cell line, and the 293T non-tumor cell line. We also include the effect of two other different cancer treatments over the miRNA levels: chemotherapy with carboplatin (CBDCA), another platinum-derived compound, in the cell lines H23R, A2780R, and three different doses of radiotherapy in 41M ovarian cancer cells (2, 4 and 6 Gy) (Figure 2).

The lowest variability among the different samples analyzed was found in miR-151a, with a standard deviation (SD) of 1.73 cycles and a coefficient of variation (CV) of 0.061, compared to candidates miR-22-5p (SD: 2.06; CV: 0.072) and miR-221-3p (SD: 2.17; CV: 0.093) (Figure 2A-C). In addition, we assessed the homogeneity of miR-16 in the same experimental groups, observing higher variability in comparison with miR-151a and similar to that observed for miR-221 (SD: 2.44; CV: 0.10) (Figure 2C-D and Supplementary Table 2).

To compare variability within the different miRNAs tested, we performed a mean Ct normalization of each miRNA in every cell line. This comparison showed that in the case of miR-151, for each cell line analyzed, the values remained close to the overall mean obtained from all the cell lines analyzed (Figure 2E). The distance to the mean (DM) was statistically significant when compared to miR-16 and -221 (p<0.05); however, no significant differences were observed with miR-22 (p=0.125) (Figure 2F). The variability observed in the DM was not statistically different between miR-16 and miR-221 (p=0.765) (Figure 2F and statistics table from Figure 2). Interestingly, we also observed an overall decrease between two and seven cycles in the levels of all four reference miRNAs tested after treatment with different doses of radiotherapy compared to their respective untreated control cells (41M cells) (Figure 2A-D). Using the values of untreated wild-type 41M cells as a calibrator, we observed that the amplification of miR-151a showed the least variation in the number of cycles, and the most stable values compared to the rest of miRNAs analyzed (Figure 2G). Most importantly, the number of cycles observed in the miR-151a amplification varied 2.7 times less compared to that observed for miR-16 (2.6 + 0.055 versus 7.0 + 0.075; p<0.001); 2.5 times less compared to that observed for miR-221 (2.6 + 0.055 versus 6.5 + 0.054; p<0.001) and 1.99 times less compared to that observed for miR-22 (2.6 + 0.055 versus 5.2 + 0.63; p<0.05) (Figure 2G). The cycle variation in the case of miR-221 and -22 was also significantly lower than that observed in miR-16 itself for the three radiotherapy doses tested (5.21+0.63 and 6.47+0.05 versus 7.0+0.075 respectively; p<0.05), (Figure 2G and Supplementary Table 3). We also analyzed the levels of miR-151 in sEVs derived from a group of human cancer cell lines cultured in the absence of FBS and compared them with those obtained in culture with depleted serum. We did not observe differences in the levels of miR-151 between sEVs-depleted FBS and the absence of FBS, showing a positive correlation between both conditions in the same cell line (R²= 0.08718) (Supplementary Figure 2).

To further determine if the relative sEVs levels of any miRNA can vary depending on the endogenous control used to normalize, we measured the levels of miR-451a, which presented values ranging from Ct 18 to Ct 30 in the different samples analyzed by miRNA-seq. miR-451a was normalized in all the cell lines and conditions tested. The normalization of miR-451a levels relative to the candidate miR-151a and the miR-16 standard used showed a significant correlation (p<0.001), with an intermediate correlation rate of 0.685 (Figure 2H). MiR-451a was also normalized with miR-22 and miR-221 showing as expected a slightly better correlation rate with miR-151a normalization (R=0.798 and 0.741 respectively) (Supplementary Figure 3A).

Therefore, of the three miRNAs analyzed so far and compared with the gold standard, miR-151a is the one with the best values as a potential normalizer, after having been tested in a significant number of cell lines from different tumor types and after having been exposed to experimentally varied conditions, many of which are known to have the potential to modify miRNA levels and expressions, such as different types of chemotherapy or radiotherapy.

Translational validation of miR-151a confirms its role as endogenous control for sEVs- miRNA normalization in plasma specimens across different tumor types

To assess whether miRNA normalization ability was maintained in the analysis of human plasma sEVs-miRNAs, we first tested its variability in an initial cohort of 30 samples from early- and advanced-stage NSCLC, ovarian and glioblastoma patients as well as in 5 healthy donor individuals. To characterize the miRNA amplification in an alternative source, our patient cohort included an additional biological fluid for sEV isolation as it was ascitic fluid paired to the plasma sample from the same ovarian cancer patient. Our results revealed that, as observed in cell lines, miR-151a is the candidate with the lowest variability and coefficient of variation within the different samples analyzed (SD: 1.16 cycles; CV: 0.062), followed by miR-22 (SD: 1.89 cycles; CV: 0.075), miR-221 (SD: 2.17; CV: 0.077) and finally miR-16 (SD: 2.44: CV: 0.076) (Figure 3A-D and Supplementary Table 4). These differences were statistically significant when comparing miR-151a with miR-16 (p=0.02; Levene test). The mean CT normalization of each miRNA in every independent sample tested revealed that miR-151a values remained the closest to the overall mean obtained from all the patients tested (Figure 3E), with statistically significant DM values different from all miRNAs tested of p=0.001 compared to miR-16, p<0.005 to miR-221 and p<0.05 to miR-22 (Figure 3F). We found a very weak correlation rate of R=0.513 (Spearman correlation) when the levels of miR-451a were normalized to miR-151a or miR-16, (Figure 3G and Supplementary Figure 3B). These results indicate that the most reliable sEVs-miRNA levels come from the normalization using miR-151a, coinciding with the results previously observed in secretome-derived sEV content.

The validation of the normalizing role of miR-151a in patient samples was then performed on a larger cohort of 172 patient samples, including 79 NSCLC patients (51 advanced and 28 early stage), 12 glioblastoma patients and 13 samples from healthy volunteers. To test miR-151a suitability as a normalizer in sEVs from an alternative fluid we also included 33 ascitic fluid samples paired with plasma samples from the same patients, reaching a total of 68 samples from ovarian cancer patients (Figure 4A and Supplementary Figure 4). In addition, we studied the miR-16 levels in those patients (Figure 4B), along with miR-451a levels (Figure 4C and Supplementary Tables 5-7)

Once again, we observed the lowest variation in cycle amplification when analyzing miR-151a compared to miR-16 within the different tumor types assessed and type of samples collected (SD: 1.66 cycles; CV: 0.056, versus SD: 1.91 cycles; CV: 0.080), finding nearly four fewer cycles of variation in miR-151a amplification (8.5 versus 12.3 cycles) (Figure 4A, B). Compared to miR-451a amplification, which lacks normalizing features, these differences are three times higher (8.5 versus 19.3 cycles) (Figure 4A and 4C). The distance from the mean of all miR-151a individual values obtained from each sample was statistically significant with respect to miR-16 (p<0.001) (Figure 4D). This global variability of 8.5 cycles for miR-151a drops to 3.2 and 3.3 cycles in the case of healthy individuals and patients with localized stages NSCLC respectively (Supplementary Figure 4 A-B), and to 5.9 cycles in the case of patients with advanced stages NSCLC (Supplementary Figure 4 C), which is to be expected when having a more aggressive disease associated with a greater number of genetic and epigenetic changes. In patients with glioblastoma, a slightly lower variability is identified (5.5 cycles) (Supplementary 4D) and finally the greatest variability is observed in patients with ovarian cancer with variations of 7.5 and 7.8 depending on the plasma or ascitic fluid origin (Supplementary Figure 4E-F).

This consistent window of low variability within amplification cycles observed in miR-151a was maintained when analyzing the individual SD and CV for each type of tumor in plasma samples. Of special interest is the fact that the coefficient of variation of the control samples is almost doubled in the case of miR-16 amplification (SD: 1.43 cycles; CV: 0.06 versus SD: 0.96 cycles; CV: 0.033) (Figure 4E).

We also compared the plasma and ascetic fluid from ovarian cancer samples, we observed that miR-151a values remained closer to the overall mean (DM) than those of the miR-16 in the ovarian cancer plasma samples (p=0.007) and, although this difference was not found when analyzing the ascitic fluid (p=0.851), the overall SD and CV remained lower when analyzing miR-151a in this fluid (Figure 4E-F). Lastly, we used GeNorm to identify the best miRNA to be used as normalizers. Despite that this and other tools would only identify normalization genes as stable (housekeeping) reference genes from a set of candidate reference genes tested in a given sample panel, running the algorithm GeNorm to calculate the expression stability value of the different reference genes. We found that miR-151a has the lowest standard deviation in our sample set (0.79 vs. 0.82 for miR-22-5p, 1.35 for miR-221-5p and 1.49 for miR-16), supporting our RNAseq and qRT-PCR data.

A moderate correlation rate, primarily associated with those samples with the highest values of miR-451a, was maintained in this extended cohort when its levels were normalized by miR-151a or miR-16 (R²=0.732), while substantial differences were not observed in those samples with intermediate or low miR-451a after normalization with miR-16 (Figure 4G). In fact, when referring the values of miR-451a compared to the control samples as calibrator, we observed that the distribution of the patient samples differs depending on whether they are normalized against miR-151a or -16. The calibration against miR-151a would provide clear differentiation between two groups of patients behaving differently from the control value, with values above the 75th or below the 25th percentiles (Figure 4H). This difference disappears when normalization is obtained against miR-16, with all the samples appearing with the same profile, and values in all cases below those of the reference control samples (Figure 4H) (p=0.0002878- Levene test). Therefore, the miR-151a normalizing properties observed in the secretome-derived sEVs are maintained in the translational approach with plasma-derived sEVs from cancer patients and controls, irrespective of the tumor types tested or the biological fluid analyzed.

In the study of biological processes, the relevance of the results is determined by the comparison with an established reference level. To be able to differentiate a pathological process from a normal one, defining if a biomarker is overexpressed or inhibited, it is essential to have a stable parameter that serves as a reference. In the cellular context, reference parameters have been widely studied and established, preferring structural and essential components for cellular maintenance such as the cytoskeleton proteins, ribosomal RNAs or the GAPDH gene and protein (37, 38). However, such references have not yet been established in recently explored areas of knowledge, as is the case of sEVs. In this regard, given that sEV content has been found to be selective with respect to cellular content (39), an incorrect choice of endogenous control may bias the study results and lead to erroneous conclusions. The endogenous controls most frequently used to normalize the sEV-miRNAs content are miR-16 and small nuclear RNA (snRNA) U6 or RNU6B in cancer (18, 20–24) and other pathologies (15–17), often without validating their homogeneity in these nanovesicles. The use of miR-16 is justified by its use in studies carried out in free plasma miRNAs (26, 27); with RNU6B there is no reference to the reason for its choice as endogenous control, probably assuming that its stability in the overall context of the cell can be extended to its use in sEVs (20, 21). However, some studies have demonstrated the inadequacy of these non-coding RNAs as endogenous controls in sEVs (13, 28, 29). This scenario reveals the risk of disparity in the results published to date, centered on levels of miRNAs from sEV. It also highlights the urgent need to identify a miRNA with stable levels in sEV that can serve as an endogenous control for the analysis of the levels of miRNAs of interest.

Given that chemotherapy treatment represents external damage to the cell, inducing aggressive changes in gene/miRNA expression and the methylation profile (40–42), we considered this scenario ideal, first of all, to identify candidate miRNAs that maintained stable levels in these extreme situations, sensitivity and resistance to cisplatin, the most frequently used chemotherapy compound, especially if these cell lines share the same genetic and epigenetic background. For this reason and after corroborating the correct isolation of 100 nm vesicles from the cell media using three alternative methodologies (43), we conducted a comparison within the miRNA profiles obtained from two experimental conditions tested in three selected human cancer cell lines: resistant (treated with increased doses of cisplatin) and sensitive (untreated), searching for a strong endogenous sEV control signature. This selection also included additional variability in the study when considering different types of tumors and different cell lines for the NGS analysis.

The selection of miR-151a-3p, miR-22 and miR-221 candidates was prioritized based on their mean absolute internal variability being close to zero (|log2FC RvsS|), considering those with the lowest variability among the different scenarios. In addition, it is essential to discard candidates with a very low average TPM, because a “bona fide” endogenous control must be at high levels to ensure its detection. An additional filter included the miRNA expression in 30 tumor types collected in the TCGA database, strengthening their possible sEV content stability. The bioinformatics analysis revealed that the MIV of miR-151a-3p was lower than any other miRNA analyzed, specifically 5.7 times less variable than miR-16. This variation decreases to 1.8 times when cell content is analyzed (0.32 vs. 0.58), suggesting that the differences in miRNA levels are much more evident when sEV content in culture medium is analyzed compared to cell content. These results would explain why to date miR-16 has been routinely used as endogenous control in this source, although its use for normalizing the content in liquid biopsy, and specifically in sEVs, may not be the most appropriate (13, 28, 29).

An initial individual quantitative evaluation of the selected candidates including miR-16 in a total of 32 cell lines subjected or not to different external factors, confirmed that the candidate with the lowest coefficient of variation was miR-151a-3p, stability that is maintained when CT values are normalized in terms of each cell line-overall mean and considering the distance from the mean. The stability of miR-151a was maintained within the different tumor types, extending its use in the general context of sEVs in cancer. The strength of this second approach lies both in the use of a large number of cell lines and in their exposure to different antitumor agents, which involve external damages that have been shown to strongly modify the cellular transcriptome (40, 44–46). Of particular interest are our results regarding ionizing radiation damage. Radiotherapy is one of the primary modalities of cancer treatment and seriously damages cell genomic integrity by producing intermediate ions and free radicals that cause DNA double-strand breaks, triggering a dramatic number of cellular changes, especially those regarding DNA damage response (DDR). One mechanism by which radiotherapy affects the DDR is by altering the expression of translational regulators including miRNAs, which have complex regulatory roles in response to radiation damage, stemming from their involvement in the DNA damage response (47–50). These changes in the expression of certain miRNAs caused by radiotherapy treatment have also been described in the sEV content in different tumor types such as brain tumors, NSCLC, or cervical cancer (51–53). We therefore considered radiotherapy as an ideal additional extreme scenario for the validation of our miRNA candidates. Although we observed a generalized earlier amplification from 2 to 7 cycles in all miRNAs tested after radiotherapy, miR-151a-3p appeared to be the candidate with less cycle number variation, followed by miR-22, miR-221 and finally miR-16. This variation in the number of cycles is of great importance since one numeric difference in ∆Ct affects by 3.33 times the relative levels of any molecule, meaning that the observed variations referred to the same endogenous value could have an important influence on the generation of robust values for any biomarker being tested.

The suitability of miR-151a-3p as an endogenous reference for any sEV-miRNAs was then tested by normalizing miR-451a levels using miR-151a-3p and − 16 as formal endogenous controls. We selected miR-451a as it has been reported to be selectively enriched in sEVs (36) and in our data obtained from the miRNome sequencing. Partial correlation between both normalizations was expected since, as observed in the bioinformatics analysis, the variability of miR-16 is lower than that of any tested miRNA without normalizing potential, which ratifies the normalizing role of miR-151a. Nonetheless, the increase from 2 to 7 cycles of variation experienced by miR-16 explains that the differences in absolute levels of miR-451a within the cell lines tested were more striking after standardization with miR-151a, providing more accurate results after normalization.

The analysis of miR-151a stability in sEV-derived cell lines was especially relevant when we analyzed its use in plasma samples. We interrogated miR-151a variability in early and in advanced NSCLC, ovarian cancer, and glioblastoma samples as well as in control samples with no pathologies, which are known to release less sEVs (54–56). In fact, the statistical significance of miR-151a in terms of DM in the first patient cohort increased from p < 0.05 to p < 0.001 in comparison to miR-16. In addition, miR-151a values were significantly different with respect to the rest of potential miRNA candidates tested in human samples. miR-151a is the candidate with the greatest stability between human samples including healthy volunteers and patients with different tumor types, different cancer stages and even between different sources such as plasma and ascitic fluid, regardless of collective analysis or within the same condition, where its variability is further reduced. As previously observed in the in vitro approach, we identified a correlation where both miR-16 and miR-151a are used as normalizers for miR-451a. However, this correlation is very low in the case of sEVs of plasma origin, and the differences in absolute levels observed within the different cell lines analyzed were even more noticeable following standardization with miRNA-151 in the plasma samples, which would allow for the establishment of more sensitive differential levels between samples.

Once miR-151a was determined to have the best characteristics as a potential normalizer in cancer, we further confirmed its wide-ranging use as endogenous sEV control through an analysis of its stability in an extended cohort of samples. The statistical significance in terms of the DM was maintained as the sample size of the cohort increased. In fact, the homogeneity of miR-151a was maintained in the sEVs extracted from healthy volunteers, displaying the values with the least dispersion in number of cycles. If we consider that the control values are those that will be most used subsequently as calibration values, the fact that the least variability is observed in this experimental group is highly relevant, reinforcing the endogenous role of miR-151a.

It is important to consider that our study focuses on circulating exosomal content in blood, and that this compartment contains relatively low levels of miRNA content, as has been described in different pathological situations such as breast cancer, hepatocellular carcinoma, or hepatitis (13, 57, 58). This situation causes most miRNAs to amplify around cycle 30, including those that, to date, have been included as gold standard normalizers. Therefore, we can consider that miRNA-151 is stable and relatively abundant in the sEVs.

It is interesting to note that normalization against miR-151a allows us to see differences in the origin of the sample, since the samples presented a different variability depending on whether their origin is plasmatic or from ascitic fluid. The fact that certain differences are observed in the values according to their origin in the case of miR-151a, could be beneficial in its clinical use by normalizing the values according to the origin of the sample, given that miR-16 does not discriminate between samples from plasma or ascitic fluid. This margin of variability in the values of an endogenous control depending on the origin of a sample was something to be expected in this study, as it has already been described on numerous occasions for different endogenous references of tissue origin used to normalize gene expression (59, 60), or miRNAs whose values vary depending on certain external conditions and origins, even between tissues in fresh or in paraffin samples (61, 62).

It could be considered as a limitation in this study the use of miR-151a as endogenous control in a fluid different from plasma, however it was also expected that these variations will be manifested between different biological fluids or origins. Even so, although an increase in SD is still observed when comparing the values obtained in plasma and ascitic fluid for the measurement of miR-151a, this is still less than that observed with miR-16.

Studies focused on miR-151a are limited and controversial. Its involvement in prostate cancer has been reported as an miRNA tumor suppressor capable of inhibiting proliferation and migration in cell lines (63), although it was also found overexpressed in esophageal carcinoma compared to normal controls (64). Variable or stable levels of miRNA 151 have also been identified in the plasma or serum of patients with different types of cancer such as head and neck cancer, hepatocellular carcinoma, or lung adenocarcinoma (65–67) but in all these cases the study has been performed on free circulating miRNAs and not on extracellular vesicles. In fact, the presence of miR-151a in sEVs has been described only in a study conducted in glioblastoma, in which high levels were associated with a greater response to temozolomide (68). Nonetheless, our findings are supported not only by the stability that we found in the sEV cargo in 32 human cancer cell lines and more than 170 human cancer samples, but also in the stability of this miRNA in 30 tumor types of the TCGA, and the fact that two studies focused on the search for normalizing miRNAs in different contexts, serum of patients with colorectal adenocarcinoma and dried cytological smears of thyroid lesions, identified miR-151a in their initial screenings as one of the candidates with greater stability among the samples analyzed (69, 70).

In this study we have identified miR-151a as the best endogenous control described to date for the normalization of miRNA levels in the sEV content under normal and tumor conditions. Its use will provide reliable liquid biopsy tests to be used in any basic, translational, or clinical studies based on sEV-miRNAs normalization.

Cell culture

A total of 31 human cell lines representing the most frequent human cancer types were used including lung (H23, H1299, H727, H1975 and A549), colon (HT-29 and LoVo), pancreas (IMIM-PC2 and PANC-1), breast (MCF7), liver (Hep EPG2), bladder (SW 780), head and neck (OSC-20, CAL27 and CAL-33), ovarian cancers (A2780 and OVCAR-3) and glioblastoma (T98G, LN-229, U-251 and U-87), as well as 293T cells. These cell lines were purchased from the ATCC (Manassas, VA, USA) or ECACC (Sigma-Aldrich, Spain) and authenticated by the Genomics Department of the Alberto Sols Biomedical Research Institute (iiBm) CSIC-UAM, through microsatellite analysis of specific loci within the Human Genome using the GenePrint 10 kit (Promega, USA) (Supplementary Table 1). Cisplatin (CDDP) resistant variants H23R, OVCAR-3R and A2780R were established previously by our group (41, 71-74). Carboplatin (CBDCA) resistant subtypes H23 and A2780 were established by exposing cells to increasing doses of carboplatin as described in the literature (74), obtaining a final CBDCA concentration of 2.5 and 1.25 µg/mL, respectively (Supplementary Figure 1). The CDDP-sensitive and resistant ovarian cancer cell lines 41M and 41MR were kindly provided by Dr. Kelland (UK). FBS was depleted of bovine sEVs by ultracentrifugation at 100,000 × g for 16h at 4ºC. All cell lines were maintained in RPMI or DMEM supplemented with 10% sEV-depleted FBS, according to manufacturer specifications. For each experiment, the original sensitive and resistant subtypes were cultured for 72h in sEV-depleted FBS media before sEV isolation. Additionally, a panel of these cell lines was cultured in the absence of FBS to compare with sEV-depleted FBS. The IC50 (µg/mL) values comparing resistant versus sensitive phenotypes are also included in Supplementary Figure 1. The CDDP-sensitive 41M cell line was irradiated at doses of 2, 4 and 6 Gy using a Mark I-30 Cesium-137 irradiator (J.L. Shepherd and Associates, CA, USA). Cells were cultured for 72 h following irradiation, when the media was collected for sEV isolation. Individual assays were performed in duplicate. All cell lines under different experimental conditions were used to validate the normalizing properties of the miRNA candidates identified after the miRNome screening performed in the paired cisplatin-sensitive and resistant cell lines H23, A2780 and 41M along with the analysis of miR-16 levels.

Sample collection

A total of 172 plasma samples were collected prospectively before receiving any antitumoral treatment from 51 advanced NSCLC patients (stages IIIA to IV), 28 patients at localized NSCLC stages (stages IA to IIIA), 12 glioblastoma patients (GB), 35 plasma and ascites fluid paired samples from ovarian cancer patients and plasma from 13 healthy donors, following the criteria of the medical oncology division of La Paz University Hospital. All samples were processed following the standard operating procedures with the appropriate approval of the Human Research Ethics Committees, including informed consent within the context of research (HULP: PI-3508 and -5063). All samples were processed within the first 30 minutes of collection using Vacutainer EDTA blood collection tubes in the case of blood and hemolyzed samples were discarded for the study as recommended (75).

SEV characterization and RNA extraction

Unbroken cells and debris were removed from the culture media by centrifugation at 7480 rpm for 5 min; for both the ascitic fluid and the blood plasma samples, two centrifugations at 5130 rpm and 14000 rpm for 10 min and 20 min respectively were performed. Samples were stored at –80C° until use. Cell line-derived sEVs used for small RNA-seq and for qRT-PCR validation, were isolated by miRCURY Exosome Cell Isolation Kit starting from 1 mL of sEV depleted FBS cell culture media (Qiagen, Germany). Circulating sEVs were obtained with the exoRNeasy Serum/Plasma Midi Kit (Qiagen, Germany). All samples followed a 0.22 µm filtration step that was included in the protocol directions. Transmission Electron Microscopy (TEM), Nanoparticle Tracking Analysis (NTA) and Western Blot (WB) were used for sEV characterization.

Transmission Electron Microscopy.

The analysis of the size and morphology of the sEVs by Transmission Electron Microscopy was conducted in the TEM Department of the Autonomous Univesity of Madrid School of Medicine. SEVs from cisplatin-resistant cells H23R, A2780R and 41MR were fixed in 100 µl of 0.1 M paraformaldehyde (PFA) of 2% phosphate buffer (pH 7.4) and samples were then transferred onto Formvar/carbon-coated copper grids. The grids were subsequently fixed with 1% w/v glutaraldehyde in PBS and washed several times in distilled water. After the washes, the grids were contrasted with 2% uranyl oxalate and soaked in a mixture of 0.4% uranyl acetate and 1.8% methylcellulose. The samples were examined in a JEOL JEM-1010 transmission electron microscope (JEOL, USA), using a working voltage of 100 kV and a 120,000 magnification. Analysis of the samples was conducted using the DigitalMicrograph program for the visualization and processing of images.

Nanoparticle tracking analysis

Nanoparticle tracking analysis (NTA) was performed using a NanoSight LM10 microscope equipped with NTA software v3.0 (Malvern, UK) in the cisplatin-resistant sEV derived cells H23, A2780 and 41M, and in a representation of plasma sEVs isolated from NSCLC, ovarian and glioblastoma patients. Background extraction was applied and the automatic setting for minimum expected particle size, minimum track length and blur settings were employed. Three 30 s recordings at 30 frames per second were taken for each sample, which were diluted at 1:400-1:100 depending on the sample concentration in sterile filtered PBS.

Western Blot

20 µg of cisplatin-resistant H23R, A2780R and 41MR whole cell extracts and their sEV-rich pellets obtained from the cell media were resuspended in 1X RIPA buffer, separated on a polyacrylamide gel, and transferred to a polyvinylidene difluoride (PVDF) membrane. The membrane was blocked with bovine serum albumin and incubated first with TSG101 antibody (ab30871, Abcam, 1: 1000) and CD81 antibody (sc23962, Santa Cruz, 1:1000) and then with species-specific horseradish peroxidase-labelled secondary antibodies (Thermo Scientific, MA, USA). Proteins were detected by enhanced chemiluminescence (Thermo Scientific, MA, USA).

RNA extraction and Small RNAseq (miRNAseq)

miRNA content from circulating sEVs was obtained from plasma and ascites fluid samples with an exoRNeasy Serum/Plasma Midi Kit (Qiagen, Germany). Cisplatin-sensitive/resistant paired H23S/H23R, A2780S/A2780R and 41S/41R cells (Supplementary Figure 1) were used for the Small-RNAseq. RNA was extracted using a miRCURY RNA Isolation Kit -Cell and Plant (Exiqon, Denmark) according to manufacturer instructions. RNA was quantified with a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, USA) and with a Qubit 4 fluorometer (Invitrogen, Thermo Fisher Scientific, USA), then analyzed with Arraystar (Arraystar Inc., MD, USA) as follows: Total RNA of each sample was used to prepare the miRNA sequencing library; 3' and 5' adapter ligation, cDNA synthesis, PCR amplification and size selection between 130 and 150 bp (corresponding to 15‒35 nucleotides (nt) of miRNAs) were performed. The DNA was sequenced with 51 cycles using the Illumina HiSeq 2000 sequencer (Illumina Inc., CA, USA). The trimmed reads (length >= 15 nt and adapter removal) were aligned to human pre-miRNA in miRBase 21 (76) using NovoAlign software (77) and miRNA read counts were normalized as tag counts per million (TPM) alignments (76). Calculating the miRNA levels, reads with counts of less than 2 were discarded. Fold Change (log2FC) and p-values were computed for comparisons between samples (resistant versus sensitive phenotype). Those with the lowest Log2FC value with at least one count per million reads under all experimental conditions were then selected, and the hg19 version was used as a reference annotation. GEO repository number: GSE204944 (restricted access until article acceptance).

qRT-PCR

The nonspecific retrotranscription of all miRNAs from each sample was conducted using the TaqMan TM Advanced miRNA cDNA Synthesis Kit (Thermo Fisher Scientific, USA), according to manufacturer instructions. Quantitative analysis of each specific miRNA was performed using the next TaqMan Advanced miRNA assay for each candidate: hsa-miR-151a-3p: 477919_mir; hsa-miR-22-5p: 477987_mir; hsa-miR-221-3p: 477981_mir; hsa-miR-16-5p: 477860_mir; and hsa-miR-451a: 478107_mir). TaqMan Universal PCR Master Mix (Thermo Fisher Scientific, USA) was used for qPCR amplification. All samples were analyzed in triplicate with a HT7900 Real-Time PCR System thermocycler (Applied Biosystems, USA) according to these settings: 10 min at 95ºC and 40 cycles of 15 s at 95ºC followed by 1 min at 60ºC. The analysis of the results was performed with RQ Manager software (Thermo Fisher Scientific, USA) and the relative quantification of miR-451a levels was based on the comparative method 2-ΔCt, calculating the value of ΔCt by subtracting endogenous Ct value to miR-451a Ct value. In the case of the radiotherapy treatment, the values from 41M mock cells were used as calibrator.

Statistical Analysis

miRNA levels were represented by the mean values ± S.D. from at least three replicates. The coefficient of variation (CV) was calculated as the ratio of the standard deviation to the mean (CV=SD/mean). Variance analysis was performed using Levene test. Statistical analysis of the effect of radiotherapy in cycle number variation was performed using Student’s t test (paired; two-tailed) with a significance of p < 0.001 in all cases.

For a specific miRNA, triplicate Ct values obtained by qRT-PCR from each sample were normalized against the mean value of the same miRNA analyzed in the total samples evaluated. These normalized values were used to calculate the distance from the mean (DM) in terms of absolute values, performing the Wilcoxon signed-rank test to analyze them. Correlation analysis between normalized levels of miR-451a using miR-151a or miR-16 as endogenous controls were carried out with a non-parametric Spearman test.

All statistical analyzes were performed with SAS 9.3 (SAS Institute, Cary, NC, USA) and RStudio (version 1.1.423). Values of p<0.05 were considered as statistically significant, being * p<0.05, ** p<0.01 and *** p<0.001.

Ethics approval and consent to participate: All samples were processed following the standard operating procedures with the appropriate approval of the Human Research Ethics Committees, including informed consent within the context of research (HULP: PI-3508 and -5063).

Consent for publication: Not applicable

Availability of data and materials: The datasets generated and/or analyzed during the current study are available in the GEO repository, number GSE204944.

Competing interests: Authors declare that they have no competing interests.

Funding: Instituto de Salud Carlos III and the European Regional Development Fund/European Social Fund FIS [ERDF/ESF], Una Manera de Hacer Europa. Under Grants: PI18/00050; PI21/00145; DTS20/00029. HHRR from ISCIII: CP19/00063; CM19/00100; FI19/00061; CD22/00040. This work was also supported by Caixa-Impulse Validate program under CI20-00182. This work was also supported by Fundación Mutua Madrilena AP180852022.

Authorship: We declare that all the authors of this study have directly participated in the planning, execution, or analysis of the study, and all the authors have read and approved the final version submitted, adhering to the guidelines of the ICMJE.

Acknowledgments: The authors thank Morote translation services for its assistance with the language editing and HULP-IdiAPZ Biobank for sample processing and the cell culture facility at IdiPAZ. We also thank the financial support.

Valadi H, Ekstrom K, Bossios A, Sjostrand M, Lee JJ, Lotvall JO. Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells. Nat Cell Biol. 2007;9(6):654-9.
Visone R, Croce CM. MiRNAs and cancer. Am J Pathol. 2009;174(4):1131-8.
Malumbres M. miRNAs and cancer: an epigenetics view. Mol Aspects Med. 2013;34(4):863-74.
Kanwal R, Gupta S. Epigenetics and cancer. J Appl Physiol (1985). 2010;109(2):598-605.
Bartel DP. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell. 2004;116(2):281-97.
Mallick R, Patnaik SK, Yendamuri S. MicroRNAs and lung cancer: Biology and applications in diagnosis and prognosis. J Carcinog. 2010;9.
Zhao L, Yu J, Wang J, Li H, Che J, Cao B. Isolation and Identification of miRNAs in exosomes derived from serum of colon cancer patients. J Cancer. 2017;8(7):1145-52.
Ge Q, Zhou Y, Lu J, Bai Y, Xie X, Lu Z. miRNA in plasma exosome is stable under different storage conditions. Molecules. 2014;19(2):1568-75.
Szczepanski MJ, Szajnik M, Welsh A, Whiteside TL, Boyiadzis M. Blast-derived microvesicles in sera from patients with acute myeloid leukemia suppress natural killer cell function via membrane-associated transforming growth factor-beta1. Haematologica. 2011;96(9):1302-9.
Minciacchi VR, Freeman MR, Di Vizio D. Extracellular vesicles in cancer: exosomes, microvesicles and the emerging role of large oncosomes. Semin Cell Dev Biol. 2015;40:41-51.
Zhang J, Ma J, Long K, Qiu W, Wang Y, Hu Z, et al. Overexpression of Exosomal Cardioprotective miRNAs Mitigates Hypoxia-Induced H9c2 Cells Apoptosis. Int J Mol Sci. 2017;18(4).
Zhou Q, Li M, Wang X, Li Q, Wang T, Zhu Q, et al. Immune-related microRNAs are abundant in breast milk exosomes. Int J Biol Sci. 2012;8(1):118-23.
Li Y, Zhang L, Liu F, Xiang G, Jiang D, Pu X. Identification of endogenous controls for analyzing serum exosomal miRNA in patients with hepatitis B or hepatocellular carcinoma. Dis Markers. 2015;2015:893594.
Occhipinti G, Giulietti M, Principato G, Piva F. The choice of endogenous controls in exosomal microRNA assessments from biofluids. Tumour Biol. 2016;37(9):11657-65.
Byun JS, Hong SH, Choi JK, Jung JK, Lee HJ. Diagnostic profiling of salivary exosomal microRNAs in oral lichen planus patients. Oral Dis. 2015;21(8):987-93.
Lange T, Stracke S, Rettig R, Lendeckel U, Kuhn J, Schluter R, et al. Identification of miR-16 as an endogenous reference gene for the normalization of urinary exosomal miRNA expression data from CKD patients. PLoS One. 2017;12(8):e0183435.
Lv LL, Cao YH, Ni HF, Xu M, Liu D, Liu H, et al. MicroRNA-29c in urinary exosome/microvesicle as a biomarker of renal fibrosis. Am J Physiol Renal Physiol. 2013;305(8):F1220-7.
Tanaka Y, Kamohara H, Kinoshita K, Kurashige J, Ishimoto T, Iwatsuki M, et al. Clinical impact of serum exosomal microRNA-21 as a clinical biomarker in human esophageal squamous cell carcinoma. Cancer. 2013;119(6):1159-67.
Tokuhisa M, Ichikawa Y, Kosaka N, Ochiya T, Yashiro M, Hirakawa K, et al. Exosomal miRNAs from Peritoneum Lavage Fluid as Potential Prognostic Biomarkers of Peritoneal Metastasis in Gastric Cancer. PLoS One. 2015;10(7):e0130472.
Li XJ, Ren ZJ, Tang JH, Yu Q. Exosomal MicroRNA MiR-1246 Promotes Cell Proliferation, Invasion and Drug Resistance by Targeting CCNG2 in Breast Cancer. Cell Physiol Biochem. 2017;44(5):1741-8.
Que R, Ding G, Chen J, Cao L. Analysis of serum exosomal microRNAs and clinicopathologic features of patients with pancreatic adenocarcinoma. World J Surg Oncol. 2013;11:219.
Warnecke-Eberz U, Chon SH, Holscher AH, Drebber U, Bollschweiler E. Exosomal onco-miRs from serum of patients with adenocarcinoma of the esophagus: comparison of miRNA profiles of exosomes and matching tumor. Tumour Biol. 2015;36(6):4643-53.
Samsonov R, Shtam T, Burdakov V, Glotov A, Tsyrlina E, Berstein L, et al. Lectin-induced agglutination method of urinary exosomes isolation followed by mi-RNA analysis: Application for prostate cancer diagnostic. Prostate. 2016;76(1):68-79.
Wang J, Zhou Y, Lu J, Sun Y, Xiao H, Liu M, et al. Combined detection of serum exosomal miR-21 and HOTAIR as diagnostic and prognostic biomarkers for laryngeal squamous cell carcinoma. Med Oncol. 2014;31(9):148.
Matsumura T, Sugimachi K, Iinuma H, Takahashi Y, Kurashige J, Sawada G, et al. Exosomal microRNA in serum is a novel biomarker of recurrence in human colorectal cancer. Br J Cancer. 2015;113(2):275-81.
Kanaoka R, Iinuma H, Dejima H, Sakai T, Uehara H, Matsutani N, et al. Usefulness of Plasma Exosomal MicroRNA-451a as a Noninvasive Biomarker for Early Prediction of Recurrence and Prognosis of Non-Small Cell Lung Cancer. Oncology. 2018;94(5):311-23.
Xie Z, Chen G, Zhang X, Li D, Huang J, Yang C, et al. Salivary microRNAs as promising biomarkers for detection of esophageal cancer. PLoS One. 2013;8(4):e57502.
Gouin K, Peck K, Antes T, Johnson JL, Li C, Vaturi SD, et al. A comprehensive method for identification of suitable reference genes in extracellular vesicles. J Extracell Vesicles. 2017;6(1):1347019.
Ni Q, Stevic I, Pan C, Muller V, Oliveira-Ferrer L, Pantel K, et al. Different signatures of miR-16, miR-30b and miR-93 in exosomes from breast cancer and DCIS patients. Sci Rep. 2018;8(1):12974.
Xiang M, Zeng Y, Yang R, Xu H, Chen Z, Zhong J, et al. U6 is not a suitable endogenous control for the quantification of circulating microRNAs. Biochemical and biophysical research communications. 2014;454(1):210-4.
Benz F, Roderburg C, Vargas Cardenas D, Vucur M, Gautheron J, Koch A, et al. U6 is unsuitable for normalization of serum miRNA levels in patients with sepsis or liver fibrosis. Exp Mol Med. 2013;45(9):e42.
Silverman JM, Reiner NE. Exosomes and other microvesicles in infection biology: organelles with unanticipated phenotypes. Cell Microbiol. 2011;13(1):1-9.
Ye SL, Li WD, Li WX, Xiao L, Ran F, Chen MM, et al. The regulatory role of exosomes in venous thromboembolism. Front Cell Dev Biol. 2022;10:956880.
Mostafazadeh M, Samadi N, Kahroba H, Baradaran B, Haiaty S, Nouri M. Potential roles and prognostic significance of exosomes in cancer drug resistance. Cell Biosci. 2021;11(1):1.
Howes LG, Krum H, Louis WJ. Increase in plasma cholesterol on subjects' standing correlates with the increase in plasma norepinephrine. Clin Chem. 1988;34(5):988.
Guduric-Fuchs J, O'Connor A, Camp B, O'Neill CL, Medina RJ, Simpson DA. Selective extracellular vesicle-mediated export of an overlapping set of microRNAs from multiple cell types. BMC Genomics. 2012;13:357.
Barber RD, Harmer DW, Coleman RA, Clark BJ. GAPDH as a housekeeping gene: analysis of GAPDH mRNA expression in a panel of 72 human tissues. Physiol Genomics. 2005;21(3):389-95.
Eisenberg E, Levanon EY. Human housekeeping genes, revisited. Trends Genet. 2013;29(10):569-74.
Villarroya-Beltri C, Baixauli F, Gutierrez-Vazquez C, Sanchez-Madrid F, Mittelbrunn M. Sorting it out: regulation of exosome loading. Semin Cancer Biol. 2014;28:3-13.
Cortes-Sempere M, de Miguel MP, Pernia O, Rodriguez C, de Castro Carpeno J, Nistal M, et al. IGFBP-3 methylation-derived deficiency mediates the resistance to cisplatin through the activation of the IGFIR/Akt pathway in non-small cell lung cancer. Oncogene. 2013;32(10):1274-83.
Soto JA, Rodriguez-Antolin C, Vera O, Pernia O, Esteban-Rodriguez I, Dolores Diestro M, et al. Transcriptional epigenetic regulation of Fkbp1/Pax9 genes is associated with impaired sensitivity to platinum treatment in ovarian cancer. Clinical epigenetics. 2021;13(1):167.
Pernia O, Belda-Iniesta C, Pulido V, Cortes-Sempere M, Rodriguez C, Vera O, et al. Methylation status of IGFBP-3 as a useful clinical tool for deciding on a concomitant radiotherapy. Epigenetics. 2014;9(11):1446-53.
Poupardin R, Wolf M, Strunk D. Adherence to minimal experimental requirements for defining extracellular vesicles and their functions. Adv Drug Deliv Rev. 2021;176:113872.
Koussounadis A, Langdon SP, Harrison DJ, Smith VA. Chemotherapy-induced dynamic gene expression changes in vivo are prognostic in ovarian cancer. Br J Cancer. 2014;110(12):2975-84.
Bravata V, Cava C, Minafra L, Cammarata FP, Russo G, Gilardi MC, et al. Radiation-Induced Gene Expression Changes in High and Low Grade Breast Cancer Cell Types. Int J Mol Sci. 2018;19(4).
Iyer G, Wang AR, Brennan SR, Bourgeois S, Armstrong E, Shah P, et al. Identification of stable housekeeping genes in response to ionizing radiation in cancer research. Sci Rep. 2017;7:43763.
Li L, Story M, Legerski RJ. Cellular responses to ionizing radiation damage. Int J Radiat Oncol Biol Phys. 2001;49(4):1157-62.
Metheetrairut C, Slack FJ. MicroRNAs in the ionizing radiation response and in radiotherapy. Curr Opin Genet Dev. 2013;23(1):12-9.
Czochor JR, Glazer PM. microRNAs in cancer cell response to ionizing radiation. Antioxid Redox Signal. 2014;21(2):293-312.
May JM, Bylicky M, Chopra S, Coleman CN, Aryankalayil MJ. Long and short non-coding RNA and radiation response: a review. Transl Res. 2021;233:162-79.
Zhu Z, Li Q, Xu M, Qi Z. Effect of Whole-Brain and Intensity-Modulated Radiotherapy on Serum Levels of miR-21 and Prognosis for Lung Cancer Metastatic to the Brain. Med Sci Monit. 2020;26:e924640.
Lv J, An J, Zhang YD, Li ZX, Zhao GL, Gao J, et al. A three serum miRNA panel as diagnostic biomarkers of radiotherapy-related metastasis in non-small cell lung cancer. Oncol Lett. 2020;20(5):236.
Cho O, Kim DW, Cheong JY. Plasma Exosomal miRNA Levels after Radiotherapy Are Associated with Early Progression and Metastasis of Cervical Cancer: A Pilot Study. J Clin Med. 2021;10(10).
Elsharkawi F, Elsabah M, Shabayek M, Khaled H. Urine and Serum Exosomes as Novel Biomarkers in Detection of Bladder Cancer. Asian Pac J Cancer Prev. 2019;20(7):2219-24.
Huang T, Deng CX. Current Progresses of Exosomes as Cancer Diagnostic and Prognostic Biomarkers. Int J Biol Sci. 2019;15(1):1-11.
Kosaka N, Kogure A, Yamamoto T, Urabe F, Usuba W, Prieto-Vila M, et al. Exploiting the message from cancer: the diagnostic value of extracellular vesicles for clinical applications. Exp Mol Med. 2019;51(3):1-9.
McDermott AM, Kerin MJ, Miller N. Identification and validation of miRNAs as endogenous controls for RQ-PCR in blood specimens for breast cancer studies. PLoS One. 2013;8(12):e83718.
Zhao J, Zhu XC, Wu XS, Wang L, Zhu CC, Yang K, et al. Identification of miR-4644 as a suitable endogenous normalizer for circulating miRNA quantification in hepatocellular carcinoma. J Cancer. 2020;11(23):7032-44.
Ayakannu T, Taylor AH, Willets JM, Brown L, Lambert DG, McDonald J, et al. Validation of endogenous control reference genes for normalizing gene expression studies in endometrial carcinoma. Mol Hum Reprod. 2015;21(9):723-35.
Racz GA, Nagy N, Tovari J, Apati A, Vertessy BG. Identification of new reference genes with stable expression patterns for gene expression studies using human cancer and normal cell lines. Sci Rep. 2021;11(1):19459.
Pamedytyte D, Leipute E, Zilaitiene B, Sarauskas V, Dauksiene D, Dauksa A, et al. Different stability of miRNAs and endogenous control genes in archival specimens of papillary thyroid carcinoma. Mol Med. 2020;26(1):100.
Inada K, Okoshi Y, Cho-Isoda Y, Ishiguro S, Suzuki H, Oki A, et al. Endogenous reference RNAs for microRNA quantitation in formalin-fixed, paraffin-embedded lymph node tissue. Sci Rep. 2018;8(1):5918.
Zhang Y, Hao T, Zhang H, Wei P, Li X. [Over-expression of miR-151a-3p inhibits proliferation and migration of PC-3 prostate cancer cells]. Xi Bao Yu Fen Zi Mian Yi Xue Za Zhi. 2018;34(3):247-52.
Zhang K, Wu X, Wang J, Lopez J, Zhou W, Yang L, et al. Circulating miRNA profile in esophageal adenocarcinoma. Am J Cancer Res. 2016;6(11):2713-21.
Pastorino R, Sassano M, Danilo Tiziano F, Giraldi L, Amore R, Arzani D, et al. Plasma miR-151-3p as a Candidate Diagnostic Biomarker for Head and Neck Cancer: A Cross-sectional Study within the INHANCE Consortium. Cancer Epidemiol Biomarkers Prev. 2022;31(12):2237-43.
Okajima W, Komatsu S, Ichikawa D, Miyamae M, Kawaguchi T, Hirajima S, et al. Circulating microRNA profiles in plasma: identification of miR-224 as a novel diagnostic biomarker in hepatocellular carcinoma independent of hepatic function. Oncotarget. 2016;7(33):53820-36.
Lv S, Xue J, Wu C, Wang L, Wu J, Xu S, et al. Identification of A Panel of Serum microRNAs as Biomarkers for Early Detection of Lung Adenocarcinoma. J Cancer. 2017;8(1):48-56.
Zeng A, Wei Z, Yan W, Yin J, Huang X, Zhou X, et al. Exosomal transfer of miR-151a enhances chemosensitivity to temozolomide in drug-resistant glioblastoma. Cancer Lett. 2018;436:10-21.
Zheng G, Wang H, Zhang X, Yang Y, Wang L, Du L, et al. Identification and validation of reference genes for qPCR detection of serum microRNAs in colorectal adenocarcinoma patients. PLoS One. 2013;8(12):e83025.
Titov SE, Demenkov PS, Ivanov MK, Malakhina ES, Poloz TL, Tsivlikova EV, et al. Selection and validation of miRNAs as normalizers for profiling expression of microRNAs isolated from thyroid fine needle aspiration smears. Oncol Rep. 2016;36(5):2501-10.
Vera O, Rodriguez-Antolin C, de Castro J, Karreth FA, Sellers TA, Ibanez de Caceres I. An epigenomic approach to identifying differential overlapping and cis-acting lncRNAs in cisplatin-resistant cancer cells. Epigenetics. 2018;13(3):251-63.
Vera-Puente O, Rodriguez-Antolin C, Salgado-Figueroa A, Michalska P, Pernia O, Reid BM, et al. MAFG is a potential therapeutic target to restore chemosensitivity in cisplatin-resistant cancer cells by increasing reactive oxygen species. Transl Res. 2018;200:1-17.
Pernia O, Sastre-Perona A, Rodriguez-Antolin C, Garcia-Guede A, Palomares-Bralo M, Rosas R, et al. A Novel Role for the Tumor Suppressor Gene ITF2 in Tumorigenesis and Chemotherapy Response. Cancers. 2020;12(4).
Vera O, Jimenez J, Pernia O, Rodriguez-Antolin C, Rodriguez C, Sanchez Cabo F, et al. DNA Methylation of miR-7 is a Mechanism Involved in Platinum Response through MAFG Overexpression in Cancer Cells. Theranostics. 2017;7(17):4118-34.
McDonald JS, Milosevic D, Reddi HV, Grebe SK, Algeciras-Schimnich A. Analysis of circulating microRNA: preanalytical and analytical challenges. Clin Chem. 2011;57(6):833-40.
Kozomara A, Griffiths-Jones S. miRBase: annotating high confidence microRNAs using deep sequencing data. Nucleic Acids Res. 2014;42(Database issue):D68-73.
Novocraft technologies. http://www.novocraft.com/ [Available from: http://www.novocraft.com/.

No competing interests reported.

BurdieletalSupplementaryMaterialsClinEpi.doc

Download PDF

Version 1

posted

You are reading this latest preprint version

Identification of miR-151a as a novel endogenous control for small extracellular vesicle cargo normalization in human cancer

Status:

Version 1

Abstract

Figures

Introduction

Results

Discussion

Methods

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1