Background
Lung cancer is the leading cause of cancer-related death in industrialized countries [
1]. Systemic treatment of lung cancer patients includes chemotherapy, inhibitors of angiogenesis and inhibitors of EGFR signaling. However, since the effect of these drugs is only transient, the overall five-year survival rate is less than 15%. Non-small cell lung carcinoma (NSCLC) accounts for 80% of lung cancer and is further subdivided into two major types, squamous cell carcinoma and adenocarcinoma [
2]. Squamous cell carcinoma usually arises from the major bronchi, whereas adenocarcinoma arises from distant airway bronchioles and alveoli. These tumours show frequent alterations of genes involved in cell cycle control or apoptosis including
k-RAS,
EGFR,
c-Myc,
cyclin D1 (
CCND1),
TP53,
retinoblastoma (
Rb),
p16INK and
Bcl2[
3], but the relevant molecular mechanisms driving the aggressive biological behaviour of these tumours are largely unknown.
miRNAs are small regulatory RNA molecules at the post-transcriptional level and are implicated in a wide variety of biological processes including proliferation, differentiation and apoptosis [
4]. Notably, miRNAs form networks to regulate the expression of individual components of the cell cycle control machinery. Many of these miRNAs including the
let-7 family [
5],
miR-34[
6],
miR-15a/16[
7],
miR-221/222[
8,
9],
miR-17-92[
10],
miR-107 and
miR-185[
11] are frequently dysregulated in lung cancer and therefore constitute promising targets for specific anticancer intervention (reviewed by Negrini et al. [
12]).
Many miRNAs are implicated in cell cycle progression or apoptosis, but surprisingly little information is available if these miRNAs are able to interact with each other to co-ordinately regulate these cellular processes. In addition, it is poorly understood why miRNAs often share common targets despite the fact that they constitute a relatively small family of RNAs encoded by less than 1000 genes. In this study we have analysed two miRNAs,
miR-15a/16 and
miR-34, which are located at chromosomal regions 13q14 and 1p34, respectively. Although these miRNAs contain completely unrelated seed sequences, they are functionally related since they are both able to induce G
1-G
0 cell cycle arrest and apoptosis [
7,
13‐
15]. In addition, they share common targets including
CCND1,
CDK4,
CDK6,
E2F3 and
Bcl2. However, other targets also exist which are unique to
miR-15a/16 (c
yclin E1 (
CCNE1),
cyclin D2 (
CCND2) or
cyclin D3 (
CCND3)) or
miR-34a (
c-Myc,
n-Myc, and
c-Met) [
7,
16‐
18].
To investigate if these miRNAs are able to interact with each other for the regulation of cellular processes, they were overexpressed in NSCLC cell lines. Here we demonstrate that miR-15a/16 and miR-34 act synergistically to induce cell cycle arrest in a Rb-dependent manner. The synergistic effect can be explained by the fact that in concerted action, miRNAs are able to down-regulate more targets than each miRNA alone. Thus, it may be important to analyse miRNAs in a combinatorial mode as this may provide additional information on their role in specific cellular processes. Consistent with these findings, both miRNAs are frequently down-regulated in adenocarcinomas and squamous cell carcinomas of the lung. Our results suggest that targeting a combination of miRNAs involved in the same pathway may potentiate the therapeutic effect of each individual miRNA.
Discussion
Cell cycle progression critically depends on numerous regulatory processes which are often deregulated in cancer (reviewed by Evan et al. [
28]). miRNAs contribute to the complexity of cell cycle control by interfering with a variety of different components of the cell cycle machinery allowing the coordinated regulation of gene expression at the post-transcriptional level (reviewed by Bueno et al. [
29] and Carleton et al. [
30]).
miR-15a/16 and
miR-34a share overlapping functions. They both induce cell cycle arrest in G
1-G
0 and share common targets including
CCND1,
CDK4 and
CDK6. In addition, the ability of either one of these miRNAs to induce cell cycle arrest in G
1-G
0 largely depends on the expression of Rb (Figure
3 and ref. [
7]). Cyclin D in complexes with CDK4 or CDK6, and cyclin E in a complex with CDK2 regulate progression through the G
1-S boundary of the cell cycle. These complexes phosphorylate and thereby prevent Rb from binding to E2F, which on release, drives cells from G
1 to S phase (reviewed by Morgan et al. [
31]). From these results we may conclude that functionally relevant targets of either type of miRNA must be upstream of Rb. These include CCNE1 and CCND3, which are unique to miR-15a/16, c-Myc and c-Met, which are unique to
miR-34a and CCND1, CDK4 and CDK6, which are common to both miRNAs. With the exception of CDK4 and c-Myc, all these genes are confirmed targets of
miR-15a/16 and
miR-34a in NSCLC cells [
7,
14,
16,
26]. In contrast, experimentally validated targets downstream of Rb including E2F1, E2F2, E2F3, E2F7, WEE1, CHK1 and CARD10 [
25,
32,
33] seem to be less relevant for the regulation of cell cycle progression by
miR-15a/16 or
miR-34a, at least in NSCLC cells.
The finding that both miRNAs share highly related functions is further illustrated by the fact that both miRNAs are co-regulated in all adenocarcinoma samples. In the majority of NSCLC cases, both miRNAs are significantly down-regulated indicating that they play an important role as tumour suppressor. Tumours can escape the concerted action of
miR-15a/16 and
miR-34a by down-regulation of both miRNAs or, alternatively, by down-regulation of Rb. Mechanisms which may lead to dysregulation of
miR-15a/16 or
miR-34a in cancer include deletion of the respective miRNA loci [
7,
34], defects in miRNA processing [
21], altered promoter methylation [
35], or altered expression of transcriptional regulators [
36‐
38]. Defects in miRNA processing may account for only a subgroup of NSCLC, since the majority of tumours either expressed normal or high levels of miR-21. p53 is a potent transactivator of
miR-34a[
39,
40], and is implicated in the processing of
miR-16[
40]. However, no correlation was observed between the mutation status of p53 and the expression level of miR-34a [
6] or
miR-15a/16[
41]. In addition, the possibility that both miRNAs are able to mutually regulate their expression can be excluded (Figure
1C). Thus, it rather seems that several independent mechanisms may account for the dysregulation of
miR-15a/16 and
miR-34a in NSCLC.
Why is there a relatively high degree of redundancy between miRNAs? To address this question we co-transfected cells with
miR-15a/16 and
miR-34a and demonstrated that both miRNAs act synergistically to induce cell cycle arrest in G
1-G
0. In contrast, the concerted action of these miRNAs on common mRNA targets was additive rather than synergistic. Thus, there seems to be little interference in binding of these miRNAs to the same target molecule and each miRNA contributes to the mRNA stability in an independent manner. The synergistic effect can rather be explained by the fact that in addition to their targets common to both miRNAs they are also able to bind to targets unique to either type of miRNA. Thus, in a combinatorial mode, both miRNAs are able to down-regulate more targets than each miRNA alone. This is based on the finding that knocking down
CCNE1, a target unique to
miR-15a/16, by RNA interference, abrogated the synergistic effect exerted by the combination of both miRNAs (Figure
7). These effects were not cell-line-specific, since comparable results were obtained with A549 and H1299 cells. This model is in agreement with our results that
miR-34a and
miR-15a/16 acted synergistically under both saturating and non-saturating conditions. In contrast, if the synergistic effect of these miRNAs were due to a more efficient repression of individual targets, we would expect such an effect to occur only under non-saturating conditions. miRNAs exert fine-tuning regulatory functions, in most cases leading only to a modest repression of target mRNAs and proteins [
24]. Our results suggest that miRNAs can potentiate their impact on the regulation of cellular processes by acting in a combinatorial mode.
Surprisingly, we were unable to detect any synergistic effect on apoptosis. Although both miRNAs are able to target
Bcl2, only
miR-34a was able to induce apoptosis. This may be due to quantitative differences in their ability to down-regulate
Bcl2. Alternatively, other anti-apoptotic genes besides
Bcl2, which are targeted by
miR-34a, but not
miR-15a/16, may have to be down-regulated in order apoptosis can occur. It is noteworthy, however, that the observed effects may depend on the cell system as
miR-15a/16 was able to induce apoptosis in CLL [
15].
There are only few examples of miRNAs in the literature that act in a synergistic manner. Ivanosvska and Cleary were the first to investigate the concerted action of miR-16 and miR-34a on cell cycle arrest. However, based on their results it was not clear if both miRNAs acted in an additive or synergistic manner [
42].
miR-84 and
let-7 promote terminal differentiation of the hypodermis and cessation of molting in
C. elegans in a synergistic manner [
43]. However,
miR-84 and
let-7 share identical seed sequences, suggesting that they regulate the same set of target genes. In addition, pairs of a cytomegalovirus derived miRNA and a host cell derived miRNA acted on the same gene (
MICB) through site proximity in a synergistic manner [
44]. Thus different mechanisms may exist that may lead to a synergistic action of miRNAs.
Therapeutic strategies for the treatment of human cancer based on modulation of miRNA activity in cancer tissues have gained much attention in the past few years [
12,
45‐
49]. In a recent publication, a new formulation is described that allows the reintroduction of miRNAs, depleted in cancer cells, in order to reactivate cellular pathways that drive a therapeutic response [
50]. The authors demonstrated that formulated
miR-34a blocked tumour growth in a mouse model of NSCLC. Our results suggest that administering formulated
miR-34a in combination with formulated
miR-15a/16 may lead to a significant increase in the therapeutic impact. This strategy may be particularly effective for the treatment of NSCLC, since both types of miRNAs are normally down-regulated in this class of tumours.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
NB performed all experiments, participated in the conception and design of the study and helped to draft the manuscript. EV was responsible for the conception and design of the study and wrote the manuscript. Both authors read and approved the final manuscript.