30 Nov 2018

Hexanucleotide expansions in the first intron of the C9ORF72 gene are the most common genetic cause of amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). Scientists entertain several possible explanations for what makes these run-on repeats toxic. C9ORF72 function is kaput, RNA foci form, dipeptide repeats coalesce into cytoplasmic aggregates. But which of these is it? At the 5th RNA Metabolism in Neurological Disease Conference, held November 1–2 in San Diego, researchers realized it might be all of the above.

Scientists drew pivotal connections, not distinctions, between these mechanisms. They saw that a cell’s inability to express C9ORF72 due to the cumbersome expansions taxes its ability to execute autophagy and this, in turn, allows dipeptide proteins (DPRs) translated from the repeat sequence to pile up. They described a scenario in which repeat-laden RNA transcripts tip off the cellular stress response, bolstering the process of repeat-associated non-AUG (RAN) translation that produces the toxic DPRs. In dissecting mechanisms of RAN translation, scientists discovered unexpected consequences. Therapeutic ideas for halting these vicious cycles emerged at the conference, as well.

Nicolas Charlet–Berguerand of the University of Strasbourg, France, had previously reported that the protein translated from C9ORF72 interacts with the Rab GTP/GDP exchange factor SMCR8; both together promote flux in autophagy. A dip in C9ORF72 expression—as has been observed recently in brain tissue of expansion carriers (Aug 2018 news)—therefore disrupts this flow. By itself, this dampening of autophagy does not wreak havoc on neurons. Instead, it is made worse by a second known genetic modifier of ALS—Ataxin-2 polyglutamine repeats—which build up when C9ORF72 expression is low (Sellier et al., 2016). 

In San Diego, Charlet-Berguerand told fellow scientists that he also sees this harmful synergy with the very DPRs translated from the repeat-hobbled C9ORF72 gene. Using HEK293 cells transfected with different C9ORF72-hexanucleotide repeat expansion (HRE) constructs, he found that RAN translation of DPRs is a particularly inefficient process. However, when he deployed siRNA to silence C9ORF72 in those repeat-expressing cells, or when he inhibited autophagy, cellular DPR accumulation ramped up to become deleterious.

Charlet-Berguerand proposed a “double hit” mechanism, in which low levels of DPRs translated via RAN accumulate when autophagy wanes. In vivo evidence in support of this mechanism came from Qiang Zhu, a postdoc in Don Cleveland’s lab at the University of California, San Diego. Zhu and colleagues generated transgenic mice expressing a bacterial artificial chromosome carrying the human C9ORF72 gene strapped with 450 GGGGCC repeats. These repeats were expressed separately from the endogenous mouse gene, and did not produce a functional C9ORF72 protein, only products of the repeat sequence (RNA foci and DPRs). The researchers had previously reported that these animals had pathological and behavioral phenotypes, including gliosis, fewer neurons in the hippocampus, poor performance on memory tests, and weak motor skills (Apr 2016 news). 

Zhu then crossed these mice to C9ORF72-deficient mice, which already had enlarged spleens, an indicator of inflammation. In San Diego, he reported that all of the repeat-associated disease phenotypes were more extreme in the crosses. The scientists got similar results when they used adeno-associated virus to express HRE sequences in the brains of C9ORF72-deficient mice. To Zhu, this implies that reduced autophagy due to an absence of C9ORF72 expression rendered the DPR accumulation more toxic. Zhu and Charlet-Berguerand agreed that their data complement each other in support of this idea.

Could boosting autophagy counteract damage by C9ORF72-HRE? This is an active area of preclinical drug development, and scientists at the conference took different views. Charlet-Berguerand considers the pursuit worthwhile but cautioned that despite considerable effort to develop autophagy-enhancing drugs, their toxicity remains a problem. Zhu favors the more direct approach of taking down expression of DPRs with antisense oligonucleotides, an approach he has employed to degrade repeat-containing C9ORF72 transcripts while sparing expression of the normal copy of the gene. Yet other researchers called autophagy a viable target that could have benefit in any disease marked by pathological protein accumulation.

Nathaniel Safren, a postdoc in Sami Barmada’s lab at the University of Michigan in Ann Arbor, described a new autophagy-screening system. He tags LC3 proteins, which associate with autophagosomes throughout their maturation process, with fluorescent dyes that change color with time. This allows him to gauge autophagic flux dynamically. Safren screened 24,000 compounds for hits that enhance autophagy in rat primary cortical neurons. One is NVP-BEZ235, aka dactolisib, an old PI3K-mTOR inhibitor and investigational cancer treatment. In Safren’s hands, NVP-BEZ235 extended survival of rat cortical neurons overexpressing TDP-43.

Besides waning autophagy, there is another way in which DPRs could end up accumulating in neurons. RAN translation itself is somehow revved up in cells bearing C9ORF72 HREs in their genome. Despite being tucked within an intron, with nary an AUG initiation codon upstream, this expansion somehow manages to be transcribed and translated. While scientists are still hammering out the particulars of this non-canonical process, they do know that translation starts at non-AUG codons—a process usually considered far less efficient than the canonical AUG start.

Or is it? In San Diego, Laura Ranum, University of Florida, Gainesville, shared her new discovery that the C9ORF72 repeat-laden transcripts trigger the cell’s integrated stress response. Ranum first described RAN translation from CAG repeat expansions seven years ago (Zu et al., 2011). The ISR starts up when the master kinase PKR phosphorylates elongation initiation factor 2α (eIF2α). It then shuts down canonical protein translation in favor of non-canonical pathways such as RAN. Ranum speculated that the repeat-containing transcripts trigger the stress response because they form hairpin structures akin to those formed by viral RNA.

At the conference, Ranum discussed initial data from two distinct strategies her lab pursued to take down DPRs. The scientists removed DPRs with a new poly-GA antibody found by screening the plasma of super-agers. They also shut down stress-induced DPR production by way of a dominant-negative form of PKR. Ranum tested both in the C9ORF72-BAC mouse model her group developed (May 2016 news). These mice recapitulate aspects of ALS/FTD, including development of RNA foci and DPRs, motor neuron degeneration, motor deficits, and early death. In San Diego, Ranum reported that both strategies substantially ameliorated these phenotypes. Surprisingly, although the antibody only targeted the poly-GA, it somehow triggered the clearance of other DPRs as well.

Peter Todd of the University of Michigan in Ann Arbor said that his lab’s preliminary data jibe with Ranum’s. RNA transcribed from the C9ORF72 repeat expansions instigated the eIF2α stress response, which, in turn, promoted RAN translation of DPRs.

Stalling—A Virtue in Biology?
Instead of presenting that data, though, Todd unveiled an unexpected consequence of RAN translation. It stalls production of FMRP. Located on the X chromosome, the FMR1 gene is implicated in both the neurodevelopmental disorder Fragile X Syndrome as well as the late-onset neurodegenerative disease Fragile X-associated tremor/ataxia syndrome (FXTAS). People with FXTAS have a string of 55 to 200 CGG repeats located upstream of the open reading frame of the FMR1 gene, and RAN translation of this expansion occurs in both the sense and antisense directions (for review, see Glineburg et al., 2018). In San Diego, Todd reported a surprising finding. These RAN translation products are detectable even at the normal repeat size in people, and both the repeat and their translation are highly conserved in mammals. This suggests that RAN translation of this sequence is a normal phenomenon.

What purpose could it serve? Todd found that the CGG repeats and RAN translation snagged scanning ribosomes, preventing them from reaching and engaging the proper AUG start site for the FMR1 gene. This dampened expression of the downstream FMR1 protein, FMRP. Most interestingly, Todd found that certain forms of synaptic plasticity suppress this RAN-mediated stalling as a method of upregulating FMRP synthesis in neuronal dendrites, allowing for local regulation of translation of this synaptic protein.

In collaboration with Ionis Pharmaceuticals in Carlsbad, California, Todd and colleagues have developed antisense oligonucleotides to block the non-cognate codons where RAN translation initiates upstream of the FMR1 gene. As predicted, shutting down RAN with these ASOs led to a spike in FMRP, and also a blockade of translation of the toxic repeat protiens at expanded repeats. In turn, this treatment lengthened the survival time of neurons derived from patients with expanded CGG repeats, Todd showed.

More broadly, Todd proposed that RAN translation and nucleotide repeats could serve physiological roles in regulating gene expression throughout the genome. He noted that there more than two million tandem microsatellte repeat sequences in the human genome, most of which are unstudied. Future work is needed to understand the mechanisms by which these repeats might influence brain function and disease, Todd believes.

Translation Right Off Transcript’s Trash?
Getting back to C9ORF72-HRE, prior studies have ignored the elephant in the room: The repeat expansion resides within an intron, and introns are removed from mRNA before ribosomes get to work. How is the expansion translated? Most scientists favor the idea that the intron is aberrantly retained as the raw mRNA transcript is being spliced, leading to translation starting at non-cognate codons just upstream of the repeat sequence (Jan 2018 news on Green et al., 2017 and Tabet et al., 2018; Sznajder et al., 2018). However, Shuying Sun, Johns Hopkins University, Baltimore, has evidence for a different mechanism. She believes the intron is mostly spliced out, and RAN translation occurs using this excised bit as a transcript. Introns lack the 5' m7G caps that usually anchor ribosomes and their associated translational machinery. Even so, Sun recently reported, these excised C9ORF72 introns were able to yield DPRs, especially, once again, when the integrated stress response was switched on (Jan 2018 news on Cheng et al., 2018). 

In San Diego, Sun bolstered this idea with additional data. She created a stir when she showed striking visuals of RAN translation in progress but, like many scientists intimidated by scientific journals, declined to share an image prior to formal publication. In collaboration with Bin Wu, also at Johns Hopkins, (Wu et al., 2016), who developed a technique to visualize RNA translation, Sun generated genetic constructs in which exons and the repeat-containing intron of C9ORF72 were tagged with either blue or red fluorescent dyes, respectively, after transcription. At the same time, she labeled the translating peptides with a SunTag (no relation to her name), a form of GFP that emits an amplified signal. In living cells expressing all these components, Sun was able to visualize and distinguish between translation occurring from the intron alone (red), exons alone (blue), or from sequences containing both.

What did she see? Mostly translating peptides whirring off either red or blue transcripts. Out of 380 cells Sun and Wu visualized, only 22 contained very few co-localized introns and exons in the cytoplasm. Importantly, Sun reported that the excised intron was exported only when it contained the repeat expansion.

Researchers complimented the beauty of Sun’s data. They also wondered how relevant these artificially tagged RNAs are to physiological transcripts in vivo. Susan Ackerman of the University of California, San Diego, asked whether nuclear export of excised introns was a broad phenomenon, or only occurred in specific cell types. Roy Parker of the University of Colorado, Denver, asked whether the secondary structure of the repeat-containing RNAs might somehow stabilize the excised intron. Does it form the characteristic, lasso-like lariat structure, as most excised introns do? Ranum asked whether other instances of RAN translation from excised introns have been reported. Sun is working on these aspects. Meanwhile, she said, her findings provide evidence that RAN translation from excised introns can happen, at least under certain conditions.—Jessica Shugart

Comments

No Available Comments

Make a Comment

To make a comment you must login or register.

References

News Citations

1. In C9ORF72 Expansion Carriers, Protein Levels Drop in Cerebellum 17 Aug 2018
2. Paper Alert: Could Antisense Therapy One Day Squelch Toxic Repeats in ALS or FTD? 29 Apr 2016
3. New C9ORF72 Mice Develop Symptoms Resembling ALS/FTD 3 May 2016
4. Stressed-Out Cells Translate C9ORF72 Repeats, Unleash Toxic Peptides 23 Jan 2018

Paper Citations

1. Sellier C, Campanari ML, Julie Corbier C, Gaucherot A, Kolb-Cheynel I, Oulad-Abdelghani M, Ruffenach F, Page A, Ciura S, Kabashi E, Charlet-Berguerand N. Loss of C9ORF72 impairs autophagy and synergizes with polyQ Ataxin-2 to induce motor neuron dysfunction and cell death. EMBO J. 2016 Jun 15;35(12):1276-97. Epub 2016 Apr 21 PubMed.
2. Zu T, Gibbens B, Doty NS, Gomes-Pereira M, Huguet A, Stone MD, Margolis J, Peterson M, Markowski TW, Ingram MA, Nan Z, Forster C, Low WC, Schoser B, Somia NV, Clark HB, Schmechel S, Bitterman PB, Gourdon G, Swanson MS, Moseley M, Ranum LP. Non-ATG-initiated translation directed by microsatellite expansions. Proc Natl Acad Sci U S A. 2011 Jan 4;108(1):260-5. Epub 2010 Dec 20 PubMed.
3. Glineburg MR, Todd PK, Charlet-Berguerand N, Sellier C. Repeat-associated non-AUG (RAN) translation and other molecular mechanisms in Fragile X Tremor Ataxia Syndrome. Brain Res. 2018 Aug 15;1693(Pt A):43-54. Epub 2018 Feb 14 PubMed.
4. Green KM, Glineburg MR, Kearse MG, Flores BN, Linsalata AE, Fedak SJ, Goldstrohm AC, Barmada SJ, Todd PK. RAN translation at C9orf72-associated repeat expansions is selectively enhanced by the integrated stress response. Nat Commun. 2017 Dec 8;8(1):2005. PubMed.
5. Tabet R, Schaeffer L, Freyermuth F, Jambeau M, Workman M, Lee CZ, Lin CC, Jiang J, Jansen-West K, Abou-Hamdan H, Désaubry L, Gendron T, Petrucelli L, Martin F, Lagier-Tourenne C. CUG initiation and frameshifting enable production of dipeptide repeat proteins from ALS/FTD C9ORF72 transcripts. Nat Commun. 2018 Jan 11;9(1):152. PubMed.
6. Sznajder ŁJ, Thomas JD, Carrell EM, Reid T, McFarland KN, Cleary JD, Oliveira R, Nutter CA, Bhatt K, Sobczak K, Ashizawa T, Thornton CA, Ranum LP, Swanson MS. Intron retention induced by microsatellite expansions as a disease biomarker. Proc Natl Acad Sci U S A. 2018 Apr 17;115(16):4234-4239. Epub 2018 Apr 2 PubMed.
7. Cheng W, Wang S, Mestre AA, Fu C, Makarem A, Xian F, Hayes LR, Lopez-Gonzalez R, Drenner K, Jiang J, Cleveland DW, Sun S. C9ORF72 GGGGCC repeat-associated non-AUG translation is upregulated by stress through eIF2α phosphorylation. Nat Commun. 2018 Jan 4;9(1):51. PubMed.
8. Wu B, Eliscovich C, Yoon YJ, Singer RH. Translation dynamics of single mRNAs in live cells and neurons. Science. 2016 Jun 17;352(6292):1430-5. Epub 2016 May 5 PubMed.

Further Reading

Papers

1. Cleary JD, Pattamatta A, Ranum LP. Repeat-associated non-ATG (RAN) translation. J Biol Chem. 2018 Oct 19;293(42):16127-16141. Epub 2018 Sep 13 PubMed.