Della Valle et al. · Stem Cell Reports · 2020

Transdifferentiation of Mouse Embryonic Fibroblasts into Dopaminergic Neurons Reactivates LINE-1 Repetitive Elements

Della Valle F, Thimma MP, Caiazzo M, Pulcrano S, Celii M, Adroub SA, Liu P, Alanis-Lobato G, Broccoli V, Orlando V

The study

What was asked, and what was found

Della Valle et al., Stem Cell Reports, 2020 asked whether LINE-1 retrotransposons, the repeat family that makes up between 15% and 20% of the mammalian genome, are merely reactivated as a side effect when a cell changes identity, or whether their activity is part of the change. The model chosen makes that question answerable. Mouse embryonic fibroblasts were converted directly into induced dopaminergic neurons over 14 days by three transcription factors, with no pluripotent stage in between, so the genome and the epigenome are not wiped and anything that moves has moved during the conversion itself.

LINE-1 transcript, protein and genomic copy number all rose during conversion, and whole-genome sequencing found roughly 800 intergenic and 111 intragenic insertions unique to the converted cells, with the intragenic ones falling in genes enriched for neuronal function. To test whether that mattered, the authors blocked LINE-1 in two independent ways. One was lamivudine, a reverse transcriptase inhibitor. The other was AUMsilence sdASOs, four of them, against four different regions of LINE-1 ORF1, taken up by gymnosis with no transfection reagent and run beside a scramble control.

Both treatments cut the yield of tyrosine hydroxylase positive neurons by 30-35% and disturbed the dopaminergic marker genes. The two blocks then separated: the oligonucleotides prevented the full-length LINE-1 transcript accumulating as well as preventing retrotransposition, while lamivudine prevented only the expansion and left expression intact. The two LINE-1 proteins were blotted in treated, untreated and scramble treated cells, in Figure 3F, though the paper states no direction for that panel in its text. Supplying a LINE-1 element the oligonucleotides cannot bind partly rescued the conversion, which is a direct specificity control. The insertions themselves landed in loci with more accessible chromatin and more nearby long non-coding transcription, and the authors conclude that this is a non-random process accompanying direct reprogramming.

Key findings

Four AUMsilence sdASOs against LINE-1, taken up by gymnosis with no transfection reagent, stopped the full-length LINE-1 transcript accumulating and stopped the element copying itself into the genome.(Results, Block of L1 Dynamics Affects the Efficiency of iDA Cell Transdifferentiation, Fig. 3D to 3F)
Silencing LINE-1 while the fibroblasts were converting cut the yield of tyrosine hydroxylase positive dopaminergic neurons by 30-35% and disturbed the neuronal marker genes, which is how the paper places LINE-1 activity inside the conversion programme rather than beside it.(Results, Block of L1 Dynamics Affects the Efficiency of iDA Cell Transdifferentiation)
The oligonucleotides and the reverse transcriptase inhibitor separated cleanly: the oligonucleotides removed the transcript as well as the copying, while the drug removed only the copying, which is the difference between silencing the RNA and blocking the enzyme.(Results, Block of L1 Dynamics Affects the Efficiency of iDA Cell Transdifferentiation)
Putting back a LINE-1 element the oligonucleotides cannot bind partly restored the conversion, which is a specificity control on the silencing rather than an argument from the phenotype alone.(Results, Block of L1 Dynamics Affects the Efficiency of iDA Cell Transdifferentiation, Fig. S2D)
The paper blots the two LINE-1 proteins in treated cells beside untreated and scramble treated cells, with TBP as the loading control. Panel 3F itself is the only evidence for what happened to the protein, because the paper states no direction for it anywhere in its text.(Fig. 3 caption, panel F)
Across the study, LINE-1 elements woke up early in the conversion, inserted themselves into loci enriched for neuronal genes, and those recipient loci showed more open chromatin and more nearby long non-coding transcription.(Abstract)

Product: AUMsilence sdASO
Targets: LINE-1 (L1) retrotransposon, ORF1 regions of the full-length transcript · mouse · retrotransposon RNA, full-length bicistronic transcriptLINE-1 (L1) retrotransposon, full-length RNA · mouse · RNA knockdown, as stated in the Results
Models: mouse embryonic fibroblasts carrying a tyrosine hydroxylase promoter driven green fluorescent protein reporter, converted directly to induced dopaminergic neurons over 14 days · primary cellsfibroblasts treated while they were converting, not before and not after · primary cells, treatment window
Delivery: gymnosis, meaning naked uptake from the culture medium with no transfection reagent, following the supplier's instructions
Readouts: multiplex TaqMan RT-qPCR for LINE-1 expression and copy number, northern blot, immunoblot for LINE-1 ORF1 and ORF2 protein, immunofluorescence for tyrosine hydroxylase, TUJ1 and VMAT2, neomycin resistance retrotransposition assay with crystal violet colony counting, whole-genome sequencing, RNA sequencing, ATAC sequencing, PCR and Sanger sequencing validation of insertion junctions
Measured: loss of tyrosine hydroxylase positive dopaminergic neurons when LINE-1 was blocked: 30%-35% (either the LINE-1 oligonucleotides or lamivudine, measured at day 14 of conversion)number of distinct oligonucleotides used against LINE-1: four (all four sequences are printed in the methods)biological replicates behind the transdifferentiation efficiency count: 6 fields from 3 independent experiments (tyrosine hydroxylase positive cell counts after treatment)replicates behind the marker gene expression panels: n = 6 (En1, FoxoA1, FoxoA2, Tuj1 and Th expression at day 14)length of the conversion protocol: 14 days (mouse embryonic fibroblast to induced dopaminergic neuron conversion)share of the mammalian genome made up of LINE-1 retrotransposons: between 15% and 20% (mammals, as stated in the paper's opening sentence)de novo LINE-1 insertions unique to the converted cells, by whole-genome sequencing: about 800 intergenic and 111 intragenic (induced dopaminergic neuron genome, absent from fibroblasts and from lamivudine treated cells)recipient genes of the intragenic insertions: 111 acceptor genes, enriched for neuronal functions (gene ontology analysis of the converted cell genome)
Figures: Figure 3 The silencing experiment: the treatment scheme, tyrosine hydroxylase staining and cell counts after the oligonucleotides or the drug, LINE-1 transcript and copy number, the LINE-1 protein immunoblot, and five dopaminergic marker genes.Figure 3C The count of transdifferentiated tyrosine hydroxylase positive cells after treatment, quantified from six fields across three independent experiments.Figure S2D Partial rescue of the conversion when an oligonucleotide-resistant LINE-1 element is supplied.
Citation: Della Valle et al. Transdifferentiation of Mouse Embryonic Fibroblasts into Dopaminergic Neurons Reactivates LINE-1 Repetitive Elements. Stem Cell Reports 2020;14:60-74. doi:10.1016/j.stemcr.2019.12.002

For research use only. Not for use in diagnostic or therapeutic procedures.