Transposable elements and miRNA: Regulation of genomic stability and plasticity

ABSTRACT

Transposable elements, the class of mobile DNA sequences that change their copies or positions within the genome have an ever increasing role in shaping the genetic and evolutionary landscape. Approximately half of the mammalian genome is composed of repetitive elements, including LINE-1 (L1) elements. Because of their ability to “copy and paste” into other regions of the genome, their activation represent an opportunity as well as a threat, as L1-induced mutations results in genomic instability and plasticity. On one hand L1 retrotransposition and integration fosters genomic diversity and on the other, de-repressed L1 functions as a driver of diseases such as cancer. The regulation of L1 is an area of intense research and novel epigenetic mechanisms have recently been discovered to now include DNA methylation, histone modifications, and miR-induced L1 silencing. During development, reprogramming and in transformed cells, specific classes of repetitive elements are upregulated, presumably due to the loss of epigenetic regulation in this process, increasing the risk of L1-induced mutations. Here we discuss how miR regulation of L1 activation fits into the complex picture of L1 repression in somatic cells and touch on some of the possible implications.

Long-Interspaced elements-1 (LINE-1 or L1)

L1 belong to the most abundant class of autonomous transposable elements comprising as much as ∼20% of mammalian genomes.^6,30 Human L1 is ∼6 kilobase pairs (kb) in length and contains 2 open-reading frames, ORF1 and ORF2. ORF1 encodes a 40-kDa protein with RNA-binding and nucleotide acid chaperone activity.³⁸ ORF2 encodes a protein with endonuclease and reverse transcriptase activities. L1 mobilizes replicatively from one place in the genome to another by a “copy and paste” mechanism, and are thought by some to be a remnant of an ancient retrovirus.²⁹ Active and inactive L1s are implicated in the evolution of mammalian genomes including activity of Alu elements and the generation of pseudogenes and have been linked to cell-based diseases, including cancer.^6,13,44 In addition, somatic L1 insertions are biased toward regions of cancer specific DNA hypomethylation, suggesting that L1 insertions may provide a selective advantage during tumorigenesis.³¹ Mechanisms that function at different levels in gene expression hierarchies have been selected to control transposition-mediated mutagenesis and diminish the potential negative impact of newly-inserted elements. In germ cells specific small RNA subtypes (piRNAs) efficiently counteract L1 activity, by directing de novo cytosine methylation and RNA degradation of active transposable elements including L1s.³ However, in somatic cells where the piRNA class of small RNAs is not expressed, attenuation of element mobilization is preferentially achieved by DNA methylation of the L1 promoter.⁴⁶ Hypomethylated somatic cells including cancer cells and during reprogramming, are therefore characterized by enhanced vulnerability to L1-induced mutagenesis, as L1 promoter activity is restored and L1 transcription reactivated.^{19,25,36,44,52} Intriguingly, the microprocessor complex, an integral part of the microRNA processing and biosynthesis pathway can repress L1 activity by binding to an RNA loop in 5′ UTR of L1 mRNA transcript that resembles a typical primary microRNA structure, leading to cleavage and subsequently repression of L1. We recently reported a novel mechanism of L1 regulation in non-germ cells, establishing L1 as a bona fide target of the cellular microRNA miR-128, which directly binds and downregulates L1 at the protein and mRNA levels.¹⁹

microRNA (miRNA, miR)

The recent discovery and characterization of miRs function in cells has revolutionized our understanding of gene control.^4,10 miRs exemplify the emerging view that non-coding RNAs (ncRNAs) may rival proteins in regulatory importance. Over half of the human transcriptome is predicted to be under miR regulation, and emerging technologies coupling biochemical isolation and next generation sequencing have been instrumental in biochemically identifying gene networks controlled by specific miRs.⁴ Mounting evidence points to critical roles of the miR pathway to the post-transcriptional regulation of every major gene cascade. Given their far-reaching role, it is not surprising that disruption of miR function contributes to many human diseases, including cancer, heart disease and neurological dysfunctions.^28,45,47 For this reason miRs pose an attractive pharmaceutical target, and miR mimics or inhibitors (anti-miRs) are being developed by a growing industry hoping to harness the power of RNA-guided gene regulation to combat disease and infection.^2,24,48,49 miRs function as short RNA guides (∼21nt) that are loaded onto specific Argonaute (Ago) proteins, which, along with other effector proteins constitute the miR-inducing silencing complex (miRISC). miRs guide the miRISC complexes to target mRNAs by imperfect basepairing between the miR and an mRNA target site sequence, resulting in reduced protein expression through a variety of mechanisms that involve mRNA destabilization and translational repression.^4,27 The major determinant of the miR target recognition is the “seed," nucleotides 2-8 at the 5′ end of the miR which perfectly complement the mRNA target site.^4,17 The short length of the major recognition sequence leads to a large number of potential mRNA targets for every miR, complicating the computation prediction of targets. To address this challenge, elegant high-throughput methodologies have been developed that biochemically isolate the sites of interaction between miR and target and identify them by next generation sequencing (CLIP-seq, PAR-CLIP, iCLIP).^8,11,18,58 Results from studies employing miR knockout mutants and benefiting from the high-throughput biochemical isolation studies have indeed confirmed predictions that whole gene networks can pivot around a single miR for their regulation. Interestingly, the miR-mediated gene regulation seems to have an unimpressive effect on individual target mRNAs; the vast majority of targets seem to be regulated between 1.5 and 2-fold; only a handful of genes are modulated above the 2-fold threshold.^15,20,24 This could be attributed to the overall redundancy of miR-mediated regulation, since most mRNA transcripts can be regulated by multiple miRs and thus the effect of single miR is dampened. An attractive theory that encompasses these observations, proposes that single miRs often may have relatively minute effects on target mRNAs but exert their role on entire gene networks and thus small expression changes in multiple nodes of the pathway can have an additive, even synergistic, effect on the pathway outcome.²¹ Expanding on this concept, a potential role for miRs would be to act as guardians of the transcriptome against aberrantly overexpressed transcripts, by downregulating their expression. In this role, miRs monitor the expression levels of targeted transcripts and act as first line responders by moderating their expression, long before any other regulatory mechanism can be activated. Thus, an overexpressed transcript that may be detrimental to the cellular physiology and homeostasis, such as oncogenes and transposon elements, can be moderated as long as they bear miR target sites, which, given the promiscuity of the seed sequence, is quite probable for any given sequence. Furthermore, one can envision that miR-mediated regulation of aberrantly overexpressed transcripts, including transposon elements, can be enhanced by modulating the miR levels, as a response to external stimuli, such as cytokines, nuclear receptor signaling^7,55 or internal processes and stimuli, such as interferon signaling,^16,37,39 defending cellular homeostasis.

miRs seem to have a role in limiting L1 activity, as abolishing miRs by using Dicer-, DGCR8- and Ago-deficient cells has led to significantly increased L1 activity.²³ These proteins all affect miR-dependent mechanisms in cells.

miR-Induced L1 repression and it's role in genomic stability and plasticity

We recently reported a novel function of the miR pathway in somatic cells, namely guarding the genomic integrity of the cell, in a manner reminiscent of the piRNAs function in germ cells.¹⁹ In somatic cells, the miRNA machinery has evolved an L1 regulatory function to compensate for the lack of piRNAs, which are not expressed in non-germ cells. In these cells, the miRNA induced silencing complex (miRISC) is guided by miR-128 to bind directly to a target site residing in the ORF2 RNA of L1. This interaction results in destabilization of the L1 transcript and, subsequently, translational repression of the L1 protein as evidenced in cancer cell lines, cancer stem cells and induced pluripotent stem cells (iPSCs), thereby reducing risk of L1-mediated mutagenesis (Fig. 1).

Figure 1.

miR-128 regulates L1 activity in cancer cells and induced pluripotent stem cells (iPSCs). miR-128 regulates multiple cellular regulators, some of which represses tumorigenesis by inhibiting stemness and epithelial to mesenchymal transition (EMT) and induces differentiation and apoptosis. We have demonstrated that miR-128 also directly targets L1 RNA, thereby repressing L1-induced mutagenesis in cancer cells, tumor-initiating cells and in induced pluripotent stem cells (iPSCs). In addition, it is likely that miR-128 attenuate the expression of cellular proteins, which feeds into the regulatory circuit of L1, thus inhibiting L1-activity by a multi-functional mechanism.

The potential relevance of miR-128 in tumorigenesis has previously been reported. miR-128 is localized on human chromosome 2q21.3 and 3p22.3, both which generate mature miR-128.^5,9,33 The possibility that miR-128 function as a tumor suppressor miR is strongly supported by the fact that 96% of lung cancers and 87% of breast cancers are characterized by a 3p loss of heterozygosity (LOH).^34,40,51,53 In addition miR-128 has been determined to be increased in metastatic tumors including ovarian cancer⁵⁴ and reduced in chemo-resistant breast cancers.⁵⁷ miR-128 mRNA targets which are likely responsible for the tumor suppressor activity include (Bmi-1, Nanog, HIF-1, VEGF, TGFBR1, EGFR), which function by inhibiting stemness, epithelial to mesenchymal transition (EMT), and inducing differentiation and apoptosis,^{26,32,35,40,43,56} all required functions for the success of a cancer cells. Furthermore, p53 and miR-128 has been demonstrated to take part in an amplification loop, in which p53 induces miR-128 expression levels and miR-128 targets SIRT1 resulting in p53 induction.¹ However, as is often the case, the opposite role for a gene in tumorigenesis has been described as well, including colon and pancreatic cancers, in which miR-128 expression has been found to be increased.⁵⁰ We anticipate that loss of miR-128s role as a L1-induced mutagenesis restriction mediator may be yet another important factor involved in allowing cancer cells to develop and progress. This may also be true in cancer initiating cells (cancer stem cells). However, it is also possible that miR-128s regulation of retroviral elements in general affect tumor-based therapies, as recently described (Fig. 2).^12,41

Figure 2.

miR-128 regulates genomic instability and genomic plasticity. L1-induced mutagenesis can result in cancer-associated genetic changes, as well as induce beneficial genomic changes leading to genomic diversity of the human brain. The finding that miR-128 represses L1 activity indicates a role for miR-128 in the regulation of genomic stability as well as genomic plasticity and enhanced diversity in the brain.

The importance of L1 de-repression in induced pluripotent stem cells (iPSCs) and cells derived from iPSCs are still unclear. However, the potential for increased risk of genomic instability induced by L1-mutagenesis warrants in depth analysis in a kinetic manner. This is especially true for cells, which are in consideration of being used in cellular therapies.

In the human brain, L1 retrotransposition has surprisingly been observed to be de-repressed in specific neuronal progenitors paving the way for neuronal plasticity with potential great importance for brain functions including memory.¹⁴ As miR-128 is highly enriched in the brain^22,42 it will be interesting to examine a role for miR-128-induced regulation of L1 in the human brain (Fig. 2).

Finally, as miRs function by repressing hundreds of mRNA targets, which often work in concert to fine-tune specific cellular responses, we anticipate that miR-128 in addition to directly targeting L1 RNA, regulates cellular mRNA targets which indirectly affect L1 activity in the cell. Our recently published data support that this is indeed the case. When comparing the effect of miR-128 on L1 retrotransposition and integration by functional assays such as colony formation assays, to our analysis of miR-128s effect by directly binding to L1 RNA (using L1 ORF2 mutants), we observed that attenuating direct binding to L1 RNA only partly rescued L1 activity. We have initiated the identification of additional miR-128 targets, by performing different RNA-seq analysis (using miR-128 Ago immune-purification and miR-128 mESC DGCR8 KO approaches). These experiments will help dissect the miR-128 network, which affect L1-induced mutagenesis involved in genomic stability and plasticity, in somatic cells. We anticipate that miR-128 regulates key proteins involved in areas of L1 activity, such at proteins crucial for RNA binding and stability, nuclear transport and potentially regulators of other cellular L1 restriction factors (Fig. 1). As retroviral elements including L1 can be viewed as a cellular insult or a cellular-encoded pathogen, it is likely that regulation of restriction factors such as miR-128 is strictly and potently coordinated following L1 de-repression. Additional studies are underway to characterize the regulatory circuits of miR-128 during retroviral element de-repression.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Funding

This work was supported by University of California Cancer Research Coordinating Committee 55205 (I.M.P.), American Cancer Society – Institutional Research Grant 98-279-08 (I.M.P.), University of California Irvine Institute for Memory Impairments and Neurological Disorders grant (I.M.P.)