Elaborate security TRAINing to fight against expression of genomic junk

Abstract

A new study shows that the expression of two classes of repetitive elements in the mouse genome is controlled through two complementary mechanisms: DNA methylation and p53-mediated transcription suppression.¹ When both lines of defense fail, expression of the repeats yields large quantities of double-stranded RNA, triggering interferon response that leads to caspase-dependent cell death. These notable findings highlight two fundamental trends: tight coupling of defense and cell death mechanisms that appears to be universal in cellular life and the exploitation of the expression of “junk” DNA as a signal triggering “altruistic” cell suicide.

Double-stranded (ds) RNA is a major distress signal in eukaryotic cells.² Transcripts of regular eukaryotic genes do not normally form extensive duplexes. The presence of long stretches of dsRNA in a eukaryotic cell is a telltale sign of virus infection that is sensed by multiple defense (innate immunity) systems, including RNA interference that appears to be ancestral in eukaryotes^3,4 and the interferon (IFN) pathways that are vertebrate-specific.^5,6 A new study from Andrei Gudkov’s group reveals novel, unanticipated aspects of the dsRNA response in mammals and in a striking manner links this form of antivirus defense with tumorigenesis and with mammalian genome evolution. Leonova et al. describe a previously unsuspected phenomenon they denote TRAIN, for transcription of repeats activates interferon.¹ This discovery brings together, in a fine-tuned circuit, two distinct mechanisms of mammalian gene expression control, namely transcription suppression by p53 and by DNA methylation, with the IFN-dependent pathway of programmed cell death.

Approximately 60% of mammalian genomes consists of various classes of repetitive elements, most of which propagate via reverse transcription followed by ectopic integration.^7,8 Typically, the expression of repeats is tightly controlled, in particular by DNA methylation at CpG sites that suppresses transcription.⁹ Leonova et al.¹ have shown that this form of control of repeat transcription is complemented by a distinct mechanism that depends on p53, the key tumor suppressor known as a universal “genome guardian.”^10,11 In p53-knockout cells, but not in wild-type cells, a DNA-demethylating agent causes a major increase in the expression of several classes of repeats, in particular short interspersed nuclear elements, SINEs (represented by B1 and B2 elements in the mouse genome) and tandem satellite repeats (GSAT). The transcripts of both the SINEs and the GSATs form dsRNA owing to extensive secondary structure and/or symmetrical transcription. Leonova et al. directly demonstrate that the resulting level of dsRNA is sufficient to trigger IFN response that induces caspase-dependent cell death.¹ In tumor cells, demethylation also induces repeat transcription, and moreover, in many tumors repeats are transcribed spontaneously, but the interferon response does not follow, apparently because this mechanism of apoptosis induction is impaired in tumors.

These discoveries are remarkable from several perspectives. Perhaps the least surprising, but nevertheless notable, is the expansion of the functions of p53 as a “genome guardian” to include prevention of repeat expression. Leonova et al. have detected p53-binding sites in different classes of repeats and have further shown that these sites mutate less often than the rest of the repeat sequence. There is also evidence that p53 can bind these sites, albeit apparently weakly, compatible with direct suppression of transcription.¹ Most strikingly, the p53-binding sites were detected both in the SINE elements from mouse (B1 and B2) and those from human (Alu) that do not share a common origin.^12,13 This finding implies parallel evolution of the p53-binding sites in the primate and rodent lineages that conceivably was driven by strong positive selection for curtailment of repeat expression in the face of repeated bursts of repetitive elements propagation.^14,15

On a more fundamental level, the results of Gudkov’s group shed light on the long-standing problem of “junk DNA.” The discovery of pervasive transcription predictably reinvigorated the idea that complex eukaryotic genomes, in particular those of mammals, contain no or little junk, and that the lack of evolutionary conservation notwithstanding, (nearly) all transcripts perform unique, albeit mostly enigmatic functions.^16,17 I believe that the tight, double control of repeat expression reported by Leonova et al.¹ is excellent evidence against these speculations. Moreover, these findings, along with previously reported data on piRNAs that inhibit transposon expression in mammalian germ line,^18,19 imply that a substantial fraction of the genome is not simply junk but, literally, toxic waste that requires secure containment. These findings are fully compatible with the non-adaptationist view of the evolution of the genome complexity^20,21 but add an important new twist: the poisonous genomic garbage can be employed as a signal of distress triggering “altruistic” cell suicide.

Finally, the discovery of the coupling between the defense against repeat expression with the suicide decision caused by the failure of defense systems appears to reflect a fundamental feature of cellular life. This discovery complements numerous other findings that implicate p53 in apoptosis induction in response to genotoxic stress.^22,23 Furthermore, analogous coupling between immunity and cell suicide is likely to exist in bacteria and archaea.²⁴ All cells that evolve in an incessant arms race with parasitic selfish elements²⁵ face these life or death decisions.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.