The Zinc-Finger Protein SOP1 Is Required for a Subset of the Nuclear Exosome Functions in Arabidopsis

Abstract

Correct gene expression requires tight RNA quality control both at transcriptional and post-transcriptional levels. Using a splicing-defective allele of PASTICCINO2 (PAS2), a gene essential for plant development, we isolated suppressor mutations modifying pas2-1 mRNA profiles and restoring wild-type growth. Three suppressor of pas2 (sop) mutations modified the degradation of mis-spliced pas2-1 mRNA species, allowing the synthesis of a functional protein. Cloning of the suppressor mutations identified the core subunit of the exosome SOP2/RRP4, the exosome nucleoplasmic cofactor SOP3/HEN2 and a novel zinc-finger protein SOP1 that colocalizes with HEN2 in nucleoplasmic foci. The three SOP proteins counteract post-transcriptional (trans)gene silencing (PTGS), which suggests that they all act in RNA quality control. In addition, sop1 mutants accumulate some, but not all of the misprocessed mRNAs and other types of RNAs that are observed in exosome mutants. Taken together, our data show that SOP1 is a new component of nuclear RNA surveillance that is required for the degradation of a specific subset of nuclear exosome targets.

Fig 1. sop mutations suppress the pas2-1 growth defect via restoration of Acyl-CoA dehydratase activity.

(A) Picture of 12-day-old seedlings of the indicated genotype grown in petri dishes in long day conditions. Only pas2-1 and pas2-1^Y presented misformed cotelydons (inset, white arrow head). Bar = 5mm. (B) Analysis of Acyl-CoA composition from the genotypes presented in 1A. Synthesis of very-long-chain-fatty-acids (VLCFA) is impaired in pas2-1 mutants which accumulate hydroxylated synthesis intermediates. The phenotype is restored by the sop mutations. Error bars represent standard deviation (n = 3). (C) RT-PCR analysis of PAS2 mRNA splicing from the genotypes presented in 1A. (D) RT-PCR analysis of pas2-1 mRNA isoforms accumulating in the indicated genotypes. The sop1-5 mutant also accumulates the PAS2-1^LONG mRNA isoform. (E) Phenotype of 7-day-old seedlings of the indicated genotype showing that sop1-5 specifically suppresses pas2-1 but not any other VLCFA-deficient mutant.

Fig 2. sop1 accumulates different levels of pas2-1 splice isoforms, some of which encode a functional PAS2 protein.

(A) Phenotype of 7-day-old pas2-1 seedlings expressing wild-type PAS2 or the various PAS2-1 isoforms under the control of the endogenous PAS2 promoter. Only the expression of PAS2^WT and PAS2-1^MIDa isoforms complement the growth defect of pas2-1. (B) RT-PCR analysis of PAS2 mRNA isoforms in the seedlings presented in 2A. (C) Quantification of RNA-seq reads matching the different pas2-1 mRNA isoforms produced in 12-day-old seedlings of the indicated genotypes. The values are normalized by the number of reads per million reads. Details regarding the pas2-1 mRNA isoforms and the corresponding specific consensus sequences can be found in S3 Fig.

Fig 3. sop1 do not suppress other splice mutants and is independent of the NMD pathway.

(A) Phenotype of 7-day-old seedlings of ton2-12 alone or in combination with sop1-5. SOP1 loss of function does not suppress the phenotype caused by the mis-spliced ton2-12 allele. (B) RT-PCR analysis of ton2-12 mRNA isoforms in the indicated genotypes. The sop1-5 mutant did not induce the accumulation of mispliced ton2-12 mRNA. (C) Phenotype of 7-day-old seedlings of the indicated genotype showing specific suppression of pas2-1 by sop1-5 but not by non-sense mediated decay upf1-5 and upf3-1 mutants. (D) RT-PCR analysis of PAS2 mRNA isoforms in the indicated genotypes showing that the PAS2-1^LONG isoform is not targeted for NMD.

Fig 4. SOP proteins encode nuclear proteins involved in RNA quality control.

(A) Diagram depicting protein identity and structure of SOP1, SOP2 and SOP3 with their color-coded annotated domains. Mutations characterized in this study are also displayed. (B) Confocal laser scanning imaging of root cells from plants stably co-expressing SOP1, SOP2 and SOP3 proteins in fusion with either GFP or RFP. (bar = 5μm). (C) Quantification of nuclear foci containing either HEN2 or SOP1 alone or both (n = 508 foci from 37 root epidermal cells).

Fig 5. SOP1 partly overlaps with SOP2 and SOP3 in RNA processing and degradation.

(A) Effect of sop mutations on post-transcriptional gene silencing (PTGS) of the 35S::GUS reporter line Hc1. Percentage represent the number of silenced plants (n = 96). (B) Northern blot showing the accumulation of 5.8S rRNA precurors in sop1-5, sop2-1 or sop3-1 single mutants compared to wild type (Col) and mtr4. The radiolabelled oligoprobe is complementary to the region directly 3’ of mature 5.8S rRNA. A methylene blue stain of the membrane is shown as loading control. (C) RT-qPCR analysis of transcript accumulation in sop1-5, sop2-1 or sop3-1 single mutants compared to wild type (Col 0) and hen2-4. Expression relative to wild type is presented on a log-scale, error bars represent SD (n = 2).

Fig 6. sop1 accumulates a subset of exosome targets.

(A) Venn Diagrams showing number of genes differentially expressed in the indicated genotypes analyzed by RNA seq. Upregulated genes (in red) or downregulated genes (in green) in pas2-1^Y and pas2-1^Ysop1-1 compared to wild type plants. (B) Functional classification based on the GO term « molecular process » of the 114 and 201 genes specifically up- or down-regulated in sop1-1 as compared to wild type plants. (C) RT-PCR of selected sop1-dependent genes in the indicated genotypes. (D) RT-qPCR analysis of transcript accumulation in sop1-5, sop2-1 or sop3-1 single mutants compared to wild type (Col 0) and hen2-4. Expression relative to wild type is presented on a log-scale, error bars represent SD (n = 2).