An epigenetic gene silencing pathway selectively acting on transgenic DNA in the green alga Chlamydomonas

Juliane Neupert,Sorel T Fitz-Gibbon,Michael Schroda,Yinghong Lu,Ralph Bock,Sean D Gallaher,Rouhollah Barahimipour,Sabeeha S Merchant,Daniela Strenkert,Na’Ama Segal

doi:10.1038/s41467-020-19983-4

Abstract

Silencing of exogenous DNA can make transgene expression very inefficient. Genetic screens in the model alga Chlamydomonas have demonstrated that transgene silencing can be overcome by mutations in unknown gene(s), thus producing algal strains that stably express foreign genes to high levels. Here, we show that the silencing mechanism specifically acts on transgenic DNA. Once a permissive chromatin structure has assembled, transgene expression can persist even in the absence of mutations disrupting the silencing pathway. We have identified the gene conferring the silencing and show it to encode a sirtuin-type histone deacetylase. Loss of gene function does not appreciably affect endogenous gene expression. Our data suggest that transgenic DNA is recognized and then quickly inactivated by the assembly of a repressive chromatin structure composed of deacetylated histones. We propose that this mechanism may have evolved to provide protection from potentially harmful types of environmental DNA.

Highlights

Introduction of theSir2-type histone deacetylase (SRTA) wild-type allele restores poor transgene expression capacity
Transgenic DNA is associated with transcriptionally active chromatin in UVM4 and UVM11
Nucleosome occupancy was determined with antibodies against unmodified histone H3, active chromatin was probed with antibodies against acetylated histones H3 (K9/K14) and H4 (K5/K8/K12/K16), and repressive chromatin formation was assessed with antibodies against monomethylated lysine at position 9 of histone 3 (H3K9me[1]; refs. 20,21)

Summary

Introduction

Introduction of theSRTA wild-type allele restores poor transgene expression capacity. (hereinafter referred to as srta-1) and strain UVM4 (hereinafter referred to as srta-2) and to confirm our assumption that srta-1 and srta-2 represent loss-of-function alleles, complementation analyses were undertaken To this end, the wild-type cDNA sequence of SRTA and that of the srta-2 mutant allele were cloned in an expression cassette containing the native SRTA regulatory sequences (Fig. 4a). The wild-type cDNA sequence of SRTA and that of the srta-2 mutant allele were cloned in an expression cassette containing the native SRTA regulatory sequences (Fig. 4a) Both expression strains were transformed with these constructs and, with a FLAG-tagged version of SRTA (Fig. 4a; Table 1). The UVM4 and UVM11 strains were transformed with the mutated allele srta-2 (pJR85; Fig. 4a) As expected, this did not result in complementation of the expression phenotype of the strains, as revealed by supertransformation with the YFP cassette (Supplementary Table 7)

Methods

Results

Conclusion