Involvement of miRNAs in cellular responses to radiation

Abstract

Purpose

Exposure of living cells to ionizing radiation has different consequences, depending on the dose and cell type. Changes in gene expression at the level of transcription and translation, including those regulated by microRNAs (miRNAs), play a role in the intrinsic radiosensitivity of different cells and define their fate, survival or death. The aim of our work was to examine how ionizing radiation may influence the expression of genes regulated by different miRNAs and miRNA biogenesis.

Materials and methods

The work was performed on cultured human melanoma Me45 cells, transiently transfected with plasmids containing Renilla luciferase reporter gene targeted by miRNAs Let-7, miR-21 or miR-24. The levels of reporter mRNAs and mRNAs coding for proteins participating in miRNA biogenesis were assayed at different time points in irradiated and non-irradiated cells using RT-qPCR, and reporter protein by luciferase activity assays. MiRNA-targeted motifs in mRNAs coding for proteins engaged in miRNA biogenesis were extracted from the miRTarBase database.

Results

Messenger RNA and protein levels of transfected luciferase genes fluctuated in time in patterns that depended on the type of miRNA regulation and changed upon irradiation of the cells. The average levels of reporter mRNAs were higher in irradiated cells, whereas the levels of proteins changed in either direction. Radiation also influenced the levels of miRNAs and the expression of genes engaged in their biogenesis suggesting that the changes in gene expression following ionizing radiation result mainly from these changes in expression of genes regulating miRNA biogenesis and the influence of miRNA on mRNA translation.

Conclusions

Currently, the responses of cells to ionizing radiation are mainly ascribed to changes in their redox conditions and increased intracellular levels of ROS, but the experiments described here suggest that a further important factor is modulation of translation through changes in biogenesis and levels of miRNAs.

Keywords:

• mRNA and miRNA expression
• miRNA biogenesis
• ionizing radiation
• reporter genes
• Me45 cells

Introduction

Ionizing radiation, like other types of radiation such as UV, induces changes of cellular redox status and damage to macromolecules and influences cell proliferation and metabolism. Radiation induces biological damage by two main mechanisms: targeted effects resulting from direct traversals of cells by ionizing radiation tracks, and non-targeted effects caused by a release of signals from directly hit cells. Targeted effects are mainly mediated by the induction of reactive oxygen species (ROS) and radicals by ionizing radiation tracks and ROS damage to cellular macromolecules, activating different signaling pathways, and reprogramming cell activities by changes in gene expression (Azzam et al. 2012). Studies of the response to different doses of ionizing and UV radiation have shown different types of DNA damage (Ward 1998; Widel et al. 1999; Branzei and Foiani 2008; Shibata and Jeggo 2019), inhibition of the cell cycle and proliferation (Beucher et al. 2009; Saenko et al. 2013; Liu et al. 2019), and increased apoptosis or senescence (Kumala et al. 2003; Jeggo and Löbrich 2006; Barazzuol et al. 2015; Poleszczuk et al. 2015). Killing cells by irradiation has become the main therapy for many types of cancers, and therefore understanding processes induced by radiation and the possibility of predicting cellular response has become very important. For this reason, many attempts to create models that would predict the biological effects of certain doses of radiation on certain types of cells are of great importance. These predictive models rely mainly on measurements of DNA damage and repair mechanisms, which may vary by cell type and type of radiation and determine the intrinsic radiosensitivity of different cells to ionizing radiation (McMahon and Prise 2021). Precise prediction of radiosensitivity for a populations of cells such as a cancer is complicated by non-targeted effects caused by communication between irradiated and non-irradiated cells, the phenomena termed bystander effects (Mothersill and Seymour 1997; 2001; Widel et al. 2015), and non-irradiated cells which receive signals from irradiated neighbors show similar but weaker responses to those in directly irradiated cells (Lyng et al. 2002; Maguire et al. 2005; Rzeszowska-Wolny et al. 2009). One of the interesting phenomena related to bystander effects is releasing exosomes and microvesicles by irradiated cells which transport many different signaling molecules among which microRNAs and other regulatory RNAs were detected (Jella et al. 2014; Zhang et al. 2015; Al-Abedi et al. 2021). Reprograming of cells in response to ionizing (and UV) radiation starts from changes in the transcriptome (Snyder and Morgan 2004; Rzeszowska-Wolny et al. 2009), and the global changes in transcript levels in directly-irradiated and non-irradiated bystander cells are similar (Rzeszowska-Wolny et al. 2009).

High-throughput transcriptome studies show that the response to radiation at the level of messenger RNAs (mRNAs) is not uniform, and directly after radiation their levels may be down- or up-regulated (Jaksik et al. 2019). Early studies, performed nearly 60 years ago, suggested that irradiation may reduce activities of some RNases (Hunt et al. 1962; Slobodian et al. 1962; Rokushika et al. 1972; Tokarskaya et al. 1975) and in this way could cause increased levels of some mRNAs. Analysis of the sequences of transcripts whose levels change after radiation reveals that up- and down-regulated genes differ by the presence of microRNA (miRNA)-targeted sequence motifs (Sassen et al. 2008); transcripts with a higher abundance of such sequences showing increased levels after irradiation (Jaksik et al. 2014).

MiRNAs are single-stranded, non-coding regulatory RNAs built usually from 21-23 nucleotides which associate with Argonaute (AGO) proteins forming RNA- induced silencing complexes (RISCs), which serve as post-transcriptional regulators of many genes. MiRNAs arise by a quite complicated biogenesis pathway divided into a few main phases starting from a transcription of a primary transcript (pri-miRNA) by RNA polymerase II or III (Lee et al. 2004; Borchert et al. 2006; Ha and Kim 2014) which is processed to a shorter form (pre-miRNA); both these steps are carried out in the nucleus (Figure 1). The conversion of pri-miRNAs into pre-miRNAs is carried out with the help of the Drosha-DGCR8 complex, and the pre-miRNAs are then transported to the cytoplasm by Exportin 5 (XPO5) where they are processed to mature, double-stranded ∼20 nucleotides long miRNA-miRNA* (Borchert et al. 2006; Bumgarner 2013; Ha and Kim 2014). One strand of this double-stranded miRNA molecule is loaded into one of the AGO proteins, forming a RISC. Digestion of pre-mi-RNA to its mature form and loading into AGO needs the enzyme Dicer, part of a protein complex containing two double-stranded RNA-binding proteins TRBP and PRKRA (Chendrimada et al. 2005; Borchert et al. 2006; Bumgarner 2013; Ha and Kim 2014) (reviewed in (Kurzynska-Kokorniak et al. 2015). AGO proteins may further interact with others to form complexes containing, among others, members of the trinucleotide repeat-containing 6 protein family (TNRC6) which seem to be the most important for miRNA-induced regulation of translation because they create a platform for further proteins and formation of differently composed RISCs (Borchert et al. 2006; Huntzinger and Izaurralde 2011; Bumgarner 2013; Ha and Kim 2014; Mathys et al. 2014).

Figure 1. Scheme of miRNA biogenesis.

Interaction of an mRNA with a RISC may result in mRNA degradation and/or inhibition of its translation. Further, miRNAs can serve as an address for RISC to join complementary sequences on its target mRNAs (Huntzinger and Izaurralde 2011). Each miRNA may target many genes and a specific gene may be targeted by many miRNAs. A mathematical model simulating the role of miRNAs in changes of the transcriptome after exposure of different types of cells to ionizing radiation suggests that the action of a few out of hundreds of miRNAs is responsible for the induction of half of the changes observed experimentally (Mura et al. 2019).

Here we ask the question if ionizing radiation may influence miRNA biogenesis and how targeting of particular mRNAs by miRNAs may influence their response to radiation. We compared the expression of miRNA-targeted or non-targeted reporter genes during 12 hours and the changes induced by exposure of cells to 4 Gy of ionizing radiation. The levels of mRNA from both miRNA-targeted and non-targeted reporter genes fluctuated in time, and irradiation changed the fluctuation dynamics and increased the average reporter mRNA levels. Irradiation also influenced the levels of miRNAs and of mRNAs coding for proteins involved in miRNA biogenesis.

Materials and methods

Cell culture and irradiation

Me45 cells (melanoma, Oncology Center, Gliwice; Kramer-Marek et al. 2006) were grown in DMEM/F12, (PAN Biotech, Aidenbach, Germany) supplemented with 10% fetal bovine serum (Eurx, Gdansk, Poland) and penicillin-streptomycin (Sigma-Aldrich, St. Louis, USA), at 37 °C in a humidified atmosphere with 5% CO₂. On the day of experiments the medium was changed for fresh medium and the cells were irradiated at room temperature with 4 Gy of X-ray photons generated by a Clinac 600 GMV (Varian, Palo Alto, CA, USA) at 1 Gy/min. Irradiated and control cells were incubated in standard conditions and harvested by trypsinization at 1, 2, 4, 6, 8, 10, 12 and 24 hours after irradiation.

Plasmids and transfection

The day before transfection cells were seeded in 96-well plates at 20 000 cells per well. They were transfected with psiCHECK-2 plasmids (Promega, Madison, USA) containing two luciferase genes, a reference Firefly gene and a reporter Renilla gene containing eight tandem repeats of the target sequences for miRNAs Let-7, miR-21 and miR-24 in their 3’UTRs. The Let-7 target sequence containing the motif TCGAGACTATACAAGGATCTACCTCAG with average complementarity of 71.75% to various mature let-7 family members was a kind gift from Martin Simard (Université Laval, Québec). The miR-21 target TCAACATCAGTCTGATAAGCTAAA was 100% complementary to the mature miR-21 sequence; the two last AAs form a spacer to limit complementarity for nonspecific binding. The miR-24 target sequence contained the motif ATACGACTGGTGAACTGAGCCG with 68% complementarity to the mature miR-24. Synthesis, sequencing, insertion, and verification of miR-24 and miR-21 motifs were performed by BLIRT (Gdansk, Poland). The unmodified plasmid was used as an unregulated control. Transfection was performed with LipofectamineTM 2000 (ThermoFisher Scientific/Invitrogen, Waltham, USA) according to the supplier’s protocol.

RNA assay

Cells from three wells were detached with trypsin (Sigma-Aldrich, St. Louis, USA), suspended in Fenozole (A&A Biotechnology, Gdynia, Poland) and extracted with a Total RNA Mini kit (A&A Biotechnology, Gdynia, Poland). Reverse transcription was done with an iScriptTM cDNA Synthesis kit (Bio-Rad, USA) and qPCR on a CFX96 Touch Real Time PCR System (BioRad, USA) with the RT PCR Mix SYBR® A kit (A&A Biotechnology, Gdynia, Poland) (Table 1).

Table 1. Real-Time PCR primer sequences.

Gene name	Forward primer	Reverse primer
Renilla Luciferase	ACAAGTACCTCACCGCTTGG	GACACTCTCAGCATGGACGA
Firefly Luciferase	GCTAAGAGCACCCTGATCG	CCTCTGGGGTAATCAGAATGG
GAPDH	TTTGGCTACAGCAACAGGGTG	TTCCTCTTGTGCTCTTGCTGG
ADAR	TCCAAAAGGCAATCCGGGAAG	AATTCATTCCGTGGCCCATCC
Ago1/EIF2C1	TTGAGACATGGGGTTTGGCTG	GATGCTTGCTTCAAGGCCAAC
Ago2/EIF2C2	ACAGCCAGCATCGAACATGAG	TTGAAATCTGGGACGGAAGGC
Ago3/EIF2C3	AGCCAACTGCTTTTTGTGCG	GTGCTGCCAAGCAACTTGAG
Ago4/EIF2C4	ACGCTCGACCTCGAACAATAC	AAATAGGCCTCCCGGCTTTTC
ATM	GGTGTTGCTTTTGAACAGGGC	ACCTTCAACACCCGTAATGCC
DGCR8	AAATGTGTTCCTGCCGGAGTC	AACGCTGGACTCTTGCTTCAG
DICER1	TTGATCCCCCTGTGAATTGGC	GCCAGCATGCAGTCTTTTGTC
PRKRA	TTGCTGAACTGTCCACCAGC	TGCAAAGCATTGTGAGCTGC
TARBP2	AATTGGCAAAGCGGAATGCG	AGCCCACACCAATGGAGAAG
TNRC6A	TTTGGCCCTCAACAGGTAGC	TTGCTGCTGCGCTAACAATC
XPO5	TGAGGCTGCAGATGAAAACCC	TGTGGTCAGCACCCTTCTTTG

Luciferase activity assays

Luciferases were assayed by the Dual-Luciferase Reporter Assay System (Promega, USA, cat. num.: E1960) according to the producer’s protocol using replicate samples from three wells. Luminescence was measured with an Infinite F200 Pro microplate reader (Tecan, Männdorf, Switzerland) in all-white flat-bottom 96 well plates (Falcon/Corning, USA).

Gene expression microarrays and identification of miRNA motifs in mRNA

To examine up- or down-regulation of genes after irradiation we compared the levels of their transcripts using Affymetrix microarrays, as described in our earlier studies (Rzeszowska-Wolny et al. 2009; Herok et al. 2010). The results are available in the ArrayExpress database under accession number E-MEXP-2623 (Athar et al. 2019). All data are MIAME compliant. Microarray data quality was assessed using the simpleaffy Bioconductor package (Wilson and Miller 2005). Raw HG-U133A microarray data from two experiments were processed using Brainarray EntrezGene specific custom CDF (v22) (Dai et al. 2005) in R using the RMA algorithm implemented in the affy-Bioconductor library (Gautier et al. 2004). Changes of expression were determined 1 and 12 h after irradiation. MiRNA-targeted motifs in mRNAs up- or down-regulated after irradiation and in mRNAs coding for proteins engaged in miRNA biogenesis (DROSHA, DGCR8, XPO5, DICER1, PRKRA, TRBP2, TNRC6A,B and C, AGO1, AGO2, AGO3, AGO4 and ADAR) were extracted from the miRTarBase database as previously described (Jaksik et al. 2014).

Statistical analyses

Results are expressed as means ± SD from at least three replicates. The two-sided Student’s t-test was used for the identification of statistically significant differences (p < 0.05) between irradiated and non-irradiated cells.

Results

Exposure of cells to ionizing radiation causes changes in the levels of specific groups of mRNAs

We showed previously that transcriptome changes occur after exposure of Me45, HCT116 or K562 cells to 4 Gy of ionizing radiation (Rzeszowska-Wolny et al. 2009; Herok et al. 2010; Mura et al. 2019), causing an increased level of some transcripts and a decrease of others. These two groups differ both in nucleotide composition (GC content) and in the frequency of miRNA-targeted transcripts (Figure 2), suggesting that changes of mRNA levels induced by irradiation are not completely random and that some transcripts are specifically down-regulated, protected, or degraded. Irradiation of Me45 cells resulted in enrichment of up-regulated transcripts in sequences targeted by miRNAs (Figure 2B), and in the next experiments, we concentrated on possible roles of miRNA in this response to radiation.

Figure 2. Messenger RNAs up- or down-regulated by ionizing radiation differ in nucleotide composition. (A) Relationship between fold change of radiation-induced effect on transcript level and G and C nucleotide content in the transcript’s 3’UTR; each point represents one transcript. (B) Average numbers of miRNA-targeted sequences in transcript groups show lower, higher, or unchanged levels 1 h after irradiation.

The levels of reporter gene transcripts show fluctuations which are influenced by miRNA targeting and change after irradiation

Searching for a possible role of miRNA in the response of transcript levels to radiation, we performed a series of experiments on the expression of reporter genes transfected into Me45 cells. We used Renilla luciferase genes into which additional sequences targeted by three different miRNAs were inserted in the 3’UTRs. The levels of reporter mRNAs and proteins were assayed at different time points, and a comparison of the miRNA-targeted and non-targeted mRNAs allowed us to estimate how miRNA targeting influenced their expression and changes after irradiation (Figure 3). During 12 h we observed fluctuations of mRNA levels in the case of both miRNA non-targeted and targeted Renilla transcripts (Figure 3A). The levels of proteins also fluctuated during that time (Figure 3D). These fluctuations were not regular, and reproducible results were obtained only when the experiments were performed on the same passage of cultured cells. The results in Figure 3A suggest that interaction of transcripts with some miRNAs may result in their protection against degradation because the average levels of miRNA-targeted transcripts were higher than those of untargeted transcripts. This did not occur in the case of protein levels; targeting by miRNA-21 and Let-7 caused a significant decrease of about two orders of magnitude and some fluctuations were seen even with low levels of protein production.

Figure 3. Time course fluctuations of levels of reporter mRNAs untargeted or targeted by different miRNAs and the influence of irradiation. (A,B) mRNA and (C,D) protein levels. (A,B) Transcript levels in non-irradiated and irradiated cells; (E) fold changes of mRNA levels at different times after irradiation; (C,D) protein levels in non-irradiated and irradiated cells; (F) fold changes of protein levels at different times after irradiation. Bars and whiskers show means and standard deviations from 3 experiments. Asterisks denote statistical significance of differences between irradiated and control samples with a p-value <.05.

In our experiments, half of the transfected cells were exposed to ionizing radiation and we followed the changes of miRNA-targeted reporter mRNAs and proteins isolated from irradiated cells at different times after irradiation (Figure 3B,D). Non-targeted and miRNA-targeted transcripts showed time course fluctuations that changed after irradiation, and the irradiation-induced fold change of mRNA levels differed for mRNAs targeted by different miRNAs (Figure 3D).

Messenger RNA levels fluctuated in both irradiated and non-irradiated cells, but the maximal and minimal values appeared at different time points. Most reporter transcripts had higher levels in irradiated than in non-irradiated cells, but lower levels were seen at some moments, (e.g. reporter mRNAs regulated by miRNA-21 at 6 and 8 h, and by miRNA-24 at 8 h, Figure 4B), and the influence of radiation was different for transcripts targeted by different miRNAs. However, the average transcript level from all time points was higher in irradiated cells. Protein levels early after irradiation showed some increase (best seen for Let-7 regulated genes), although, unlike average mRNA levels, they were lower during the 12 h after irradiation.

Figure 4. The levels of miRNAs that target the reporter genes shown in Figure 3 in control and irradiated cells, assayed by microarrays 12 h after irradiation.

Micro RNA levels and expression of genes participating in miRNA biogenesis change after irradiation

A possible explanation for the differences between effects of different miRNAs could be radiation-induced changes in levels of miRNAs that target reporter genes. Our microarray data on miRNA levels in irradiated Me45 and other types of cells (K562 and HCT116) performed 12 h after (Rzeszowska-Wolny et al. 2009; Herok et al. 2010) suggested that the levels of most miRNAs did not change significantly, and among those whose level changed by at least 10% nearly all decreased (Table 2).

Table 2. Changes of miRNA levels after irradiation.

miRNA	Type of change	Cell type
		K562	Me45	HCT116
Number of miRNAs with >10% change of expression	Decrease	126	183	10
	Increase	1	1	5

In Me45 cells transfected with reporter genes, miRNA-21, miRNA-24, and miRNAs from the Let7 group which was expressed in these cells also showed a ∼30% decrease in level after irradiation (Figure 4). Micro RNAs are relatively small molecules less prone to a direct influence of radiation (Singh and Pollard 2017), but the different steps in their biogenesis could be sensitive to changed cellular conditions induced by irradiation. To understand better the influence of irradiation on miRNA biogenesis, we examined the expression of genes coding for proteins participating in the biogenesis and activity of miRNAs (Figure 1). The levels of mRNAs coding for DGCR8, DROSHA, Exportin 5, DICER1, TRBP, PRKRA and TNRC6 proteins, the main participants in miRNA biogenesis, were assayed by RT-qPCR in control and irradiated cells. Twelve hours after irradiation the levels of nearly all these transcripts had increased in irradiated cells, suggesting that miRNA biogenesis could be enhanced by the exposure of cells to radiation (Figure 5).

Figure 5. Irradiation-induced changes in levels of transcripts coding for proteins participating in miRNA biogenesis and loading into RISC, 12 h after irradiation of cells (a.u., arbitrary units). Bars and whiskers show means and standard deviations from 3 experiments. Asterisks denote statistical significance of differences between irradiated and control samples with a p-value <.05.

The increases of the levels of mRNAs coding for proteins participating in miRNA biogenesis were not large but were reproducible. The exceptions were DICER1 and PRKRA enzymes, which participate in the final processing of miRNAs and their loading to RISC complexes and showed significantly higher variability after irradiation with an increase or decrease in different experiments, suggesting that regulation of their expression in response to radiation differs from that of other genes coding for steps in miRNA biogenesis. Among these genes, transcripts of ADAR showed a high and stable increase in transcript level after irradiation, suggesting that this enzyme may also play some role in responses to radiation.

Genes participating in miRNA biogenesis are regulated by miRNAs whose levels decrease in cells exposed to ionizing radiation

All the transcripts mentioned above coding for proteins participating in miRNA biogenesis are themselves controlled by miRNAs and contain sequences targeted by many different miRNAs; some such as TNRC6 and AGO transcripts contain sequence motives targeted by over a hundred of different miRNAs (Table 3).

Table 3. Numbers of different miRNA that may target transcripts implicated in miRNA biogenesis pathways*.

Transcript	Number of different miRNAs
DROSHA	24
DGCR8	19
XPO5	26
DICER1	75
PRKRA	2
TRBP2	25
TNRC6 A, B and C	185/200/69
AGO1	113
AGO2	300
AGO3	181
AGO4	55
ADAR	57

* from the miRTarBase database

Some miRNAs target more than one transcript coding for proteins from the biogenesis pathway, and in Table 4 we list those miRNAs which target more than four transcripts encoding proteins involved in miRNA maturation and function that are expressed in Me45 cells. Their high number and the fact that some of them regulate more than one of these transcripts suggests specific regulation of gene expression at the posttranscriptional level.

Table 4. MiRNAs targeting at least 4 types of transcripts encoding proteins in the miRNA biogenesis pathways*.

miRNA	Targeted genes
miR-let-7	AGO1-4, DGCR8, DICER1, TNRC6A, TNRC6B, XPO5
miR-148a/b-3p	AGO1-3, DICER1, DROSHA, TNRC6A, TNRC6B
miR-548-5p	ADAR, AGO2-4, DICER1, DROSHA, TNRC6A, TNRC6B
miR-103a-3p	AGO1-3, DICER1, TARBP2, TNRC6B
miR-107	AGO1-3, DICER1, TARBP2, TNRC6B
miR-130a/b-3p	ADAR, AGO1, AGO3, DICER1, TNRC6A, TNRC6B
miR-195-5p	AGO1-2, AGO4, DICER1, TARBP2, TNRC6B
miR-20a/b-5p	ADAR, AGO1, AGO3-4, TNRC6A, TNRC6B
miR-484	ADAR, AGO1, AGO4, TARBP2, TNRC6A, TNRC6B
miR-519-3p	ADAR, AGO1, AGO3, DICER1, TNRC6A, TNRC6B
miR-106a/b-5p	ADAR, AGO1, AGO3, TNRC6A, TNRC6B
miR-152-3p	AGO2-3, DICER1, TNRC6A, TNRC6B
miR-15a/b-5p	AGO2, AGO4, DICER1, TARBP2, TNRC6B
miR-16-5p	AGO2, AGO4, DICER1, TARBP2, TNRC6B
miR-17-5p	ADAR, AGO1, AGO3, TNRC6A, TNRC6B
miR-19a/b-3p	AGO1, AGO3, DICER1, TNRC6A, TNRC6B
miR-30-5p	AGO1-2, DROSHA, TNRC6A, TNRC6C
miR-497-5p	AGO2, AGO4, DICER1, TARBP2, TNRC6B
miR-652-3p	AGO1-3, DGCR8, TNRC6A
miR-8485	AGO2-3, DROSHA, TNRC6B, TNRC6C
miR-93-5p	ADAR, AGO1, AGO3, TNRC6A, TNRC6B

* from the miRTarBase database

This irradiation-induced increase of the levels of transcripts coding for proteins participating in miRNA biogenesis could depend on changes of miRNAs that target these transcripts. Figure 6 presents the changes of some of these miRNA levels in cells 12 h after exposure of cells to ionizing radiation.

Figure 6. The levels of miRNAs that target transcripts coding for different proteins participating in miRNA biogenesis decrease 12 h after irradiation.

The levels of most miRNAs targeting transcripts coding for proteins important in miRNA biogenesis either do not change or decrease after irradiation, with Let-7i and Let-7a showing the biggest decrease. However, the level of Let-7b increased slightly, suggesting that different miRNAs of the same group may play different roles, even targeting the same transcripts.

Discussion

In these and earlier experiments on cellular responses to ionizing radiation, we observed changes of the transcriptome and differences between up- and down-regulated groups of transcripts such as differences in GC nucleotide content or in the presence of sequence motives recognized by miRNAs (Jaksik et al. 2014) in Me45 cells. A short time after irradiation, transcripts that were up-regulated had a lower content of guanidine and cytidine nucleotides and differed in their content of sequences targeted by miRNAs. In experiments with reporter gene expression, we observed fluctuations of mRNA levels. At some time points, mRNA levels in irradiated cells were lower than in non-irradiated cells, but the average levels calculated for all time points during 12 h were always higher in irradiated cells, in accordance with observations of others (Tsai et al. 2005). One reason for the elevated mRNA levels in irradiated cells could be the inactivation of RNases which was reported for both purified RNase and RNase isolated from irradiated rats (Hunt et al. 1962; Slobodian et al. 1962; Rokushika et al. 1972; Tokarskaya et al. 1975). However, the global inactivation of some RNAses cannot explain the specific changes in groups differing in nucleotide composition. The differences in response to radiation between GC-rich and AT-rich groups of genes could result from different influences of radiation on RNA interactions with RNA binding proteins. GC- and AT-rich RNAs are differently expressed and distributed in cytoplasmic structures such as P-bodies (Kudla et al. 2006; Courel et al. 2019) and GC-rich transcripts are also less frequently targeted by miRNAs (Courel et al. 2019). The different distribution of miRNA targets between up- and down-regulated transcript groups suggests a role for miRNAs in responses to radiation. One explanation for the observed increase of the levels of miRNA-targeted transcripts could be a decrease in miRNA levels in irradiated cells, because targeting by miRNAs can lead to mRNA degradation (Bartel 2009). Our microarray data for three cell lines showed that levels of most miRNAs did not change significantly after irradiation, and those which did change mainly show a decrease 12 h after irradiation (Table 2). There are discrepancies in the results of other comparable studies; increases of some miRNAs have been reported and particular miRNAs which decreased after irradiation in some studies were found to increase in others (Cha et al. 2009; Shin et al. 2010; Wagner-Ecker et al. 2010) reviewed in (Czochor and Glazer 2014), and even in the same experiments different members of the Let-7 group differed in response to radiation dose (Simone et al. 2009). Such results could be explained by the influence of radiation on miRNA biogenesis. In Tables 3 and 4, we show that miRNA biogenesis depends on proteins whose expression is also regulated by different miRNAs and thus, a decrease in the some miRNA levels may increase miRNA biogenesis, which in turn would increase the miRNA level and inhibit biogenesis again. In a stable environment such feedbacks would cause permanent fluctuations and further fluctuations of other miRNA regulated gene expression.

Random temporal fluctuations in RNA and protein expression have been shown in many cell types and interpreted as intrinsic noise (shown as a ratio of variance to square of the mean) arising from stochastic events including promoter binding, mRNA decay and translation, and protein degradation (Raser and O'Shea 2005), reviewed in (Ebert and Sharp 2012). We also observed irregular fluctuations of mRNA levels which were not reproducible in experiments performed on different passages of cells. (Schmiedel et al. 2015) showed that miRNA may differentially influence this noise, decreasing it in lowly-expressed transcripts and increasing it in those highly expressed. In our experiments, both targeted and non-targeted Renilla and Firefly reporter genes had relatively high expression in transfected cells, and fluctuations of miRNA-targeted and non-targeted transcripts differed depending on the type of miRNA target. We also observed differences in response to radiation between miRNA-regulated and non-regulated transcripts; on average irradiation caused an increase of about 60% in miRNA-regulated Renilla luciferase transcripts whereas non-targeted Firefly luciferase mRNA levels showed about a 10% decrease. However, Firefly mRNA levels were characterized by a large dispersity ranging from 90% decrease to 40% increase at different time points. Our observation of a higher dispersity in mRNA levels than in protein levels suggests that interactions between mRNA and other macromolecules which lead to its synthesis and degradation are more random than those leading to production and degradation of proteins, and regulation by miRNA targeting seems to decrease random interactions between mRNA and other molecules but randomness in translation is less miRNA-dependent.

Evidence that miRNAs may influence the response to radiation at the level of gene expression was presented earlier (Czochor and Glazer 2014) reviewed in (Podralska et al. 2020) and was explained by their targeting of transcripts encoding proteins connected to cellular processes such as proliferation, apoptosis, glycolysis, DNA repair, and the cell cycle (reviewed in (Podralska et al. 2020)) but there is not much information on the beginning of these processes. We suggest that modulation of changes in gene expression after irradiation is mainly caused by changes in the levels of miRNAs resulting from the effects of radiation on their biogenesis. Regulation of translation by miRNAs is complicated and based on specific RNA-protein and RNA-RNA interactions which depend on levels of different miRNAs (Mura et al. 2019). On the other hand, the levels of miRNA depend on those of proteins participating in their biogenesis and there is mutual regulation between miRNA levels and levels of proteins involved in miRNA biogenesis (model in Figure 7).

Figure 7. Influences of ionizing radiation on cellular processes and miRNA biogenesis.

Ionizing radiation changes the transcriptome and translation profiles of cells (Kumaraswamy et al. 2008). MiRNA profiles also change in irradiated cells (Simone et al. 2009), possibly as a result of changes in their biogenesis but also of their export from stressed cells in exosomes to signal to other cells in a population (Wang et al. 2010). Responses of cells to ionizing radiation are mainly ascribed to changes in their redox conditions and increasing intracellular levels of ROS. The experiments described here suggest that a further important factor in the responses may be targeting translation through changes in biogenesis and levels of miRNAs.

Regulation of the levels of proteins participating in miRNA biogenesis by miRNA must result in permanent fluctuations not only miRNAs but also other cellular miRNA regulated proteins, but in stable conditions, such fluctuations should not exceed some threshold values which would define the physiological state of the cell. Change of cellular conditions, as caused by an irradiation-induced change of redox status, may help to overcome threshold values for the levels of some components and induce signals leading to a new state of the cell. It seems that changes in the translation process inducing changes in miRNA levels could be one of the main elements of cellular response to radiation and possibly also to other stressing factors.

Acknowledgments

Ronald Hancock is acknowledged for critically reading and editing the manuscript. S.C. acknowledges support by the Young Author project BKM 02/040/BKM21/1014.

Disclosure statement

The authors declare no conflict of interest.

Additional information

Funding

This work was supported by the NCN under Grant 2016/23/D/ST7/03665; NCN under Grant 2019/35/N/ST6/04281; The Silesian University of Technology under Grant number 02/040/BK_21/1010; and Medical University of Silesia under Grant number KNW-2-O06/N/9/N.

Notes on contributors

Joanna Rzeszowska-Wolny

Joanna Rzeszowska-Wolny PhD, physicist specialized in molecular biology and genetics, 1997-2011 professor and head of the Radiobiology Department at the Institute of Oncology in Gliwice (Poland), presently professor at the Department of Systems Biology and Engineering in the Silesian University of Technology, interested in systems approaches to understanding the processes and mechanisms regulating gene expression.

Dorota Hudy

Dorota Hudy MSc, graduate of the Faculty of Automatic Control, Electronics and Computer Science at the Silesian University of Technology, PhD student at this university and research assistant at the Department of Medical and Molecular Biology in the Medical University of Silesia, scientific interests regulation of gene expression at the translation level, and role of miRNAs.

Krzysztof Biernacki

Krzysztof Biernacki MSc, graduate of the Faculty of Automatic Control, Electronics and Computer Science at the Silesian University of Technology in Gliwice, PhD student at the Biotechnology Center of the Silesian University of Technology, research assistant at the Department of Medical and Molecular Biology in the Medical University of Silesia, scientific interests regulation of gene expression and bioinformatic data analysis.

Sylwia Ciesielska

Sylwia Ciesielska MSc, a graduate of the Faculty of Automatic Control, Electronics and Computer Science at the Silesian University of Technology in Gliwice, PhD student and research assistant at the Department of Systems Biology and Engineering at this university, scientific interest reactive oxygen species and their role in cellular systems.

Roman Jaksik

Roman Jaksik, PhD in biocybernetics and biomedical engineering, assistant professor at the Silesian University of Technology in the Department of Systems Biology and Engineering, interested in the development of novel methods for the analysis of high throughput data and studies of intracellular processes related to the development and progression of cancer.