(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Structured and disordered regions of Ataxin-2 contribute differently to the specificity and efficiency of mRNP granule formation [1]

['Arnas Petrauskas', 'Trinity College Institute Of Neuroscience', 'School Of Genetics', 'Microbiology', 'Smurfit Institute Of Genetics', 'School Of Natural Sciences', 'Trinity College Dublin', 'Dublin', 'Daniel L. Fortunati', 'Arvind Reddy Kandi']

Date: 2024-05

The structured PAM2 domain of Atx2 is necessary for the correct mRNA and protein content of Atx2 granules

A recent work from our lab used Targets of RNA-Binding Proteins Identified by Editing (TRIBE) technology to identify mRNAs associated with Atx2 in the Drosophila adult brain [83]. In vivo, the ability of an Atx2-fusion with ADARcd (the catalytic domain of an RNA-editing enzyme, ADAR), to edit a group of 256 target mRNAs was found to be dependent on the presence of the Atx2-cIDR, previously shown to be necessary for the formation of neuronal mRNP granules in vivo. In contrast, Atx2-ADARcd mutants lacking the LSm domain, both edited Atx2 TRIBE target RNAs and formed mRNP granules in cultured Drosophila S2 cells more efficiently than the wild-type. Thus, Atx2-ADARcd editing of target mRNAs occurs in and is reflective of mRNP granule assembly. While demonstrating a role for LSm-domain interactions in antagonizing cIDR-mediated granule assembly, these observations did not address mechanisms by which Atx2 target mRNAs are selected, or whether and how Atx2 played any role in determining the composition of RNP granules. New experiments presented here address these outstanding questions.

Previous TRIBE analyses showed that LSm and LSm-AD regions have no major role in the recognition or selection of the Atx2-target mRNAs [83]. We therefore tested whether the third conserved region of Ataxin-2, a PAM2 motif known to associate with PABP (polyA binding protein), played any role in this process [85,86].

We used Gal80ts-controlled elav-Gal4 to express Atx2ΔPAM2-ADARcd (deleted for the PAM2 motif) in brains of adult Drosophila for 5 days and used RNA-Seq to identify edited RNAs in polyA selected brain RNA and compared it with Atx2-ADARcd using procedures described earlier (Fig 1A) [87]. ADAR-edits, which convert Adenosine to Inosine on RNAs, are identified as A to G changes in TRIBE analyses. Each sample was sequenced to obtain 20 million reads (Table A in S1 File). The edits were only considered from the regions of the transcriptome that contained at least 20 reads. Genes with edits identified at a threshold above 15% in two biological replicates were considered as high confidence true targets. We compared edit frequency and edited gene identity in the brains of flies expressing Atx2ΔPAM2-ADARcd with those in brains expressing Atx2-ADARcd.

PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 1. The PAM2 domain facilitates the selection of Atx2 RNA targets. (A). The flowchart illustrates the TRIBE analysis pipeline. Atx2ΔPAM2-ADARcd was expressed in adult Drosophila brains using the Gal4-responsive UAS-Atx2 transgene, under the control of elav-Gal4 and TubGal80ts. The temperature-sensitive TubGal80ts enables us to induce the expression of the UAS-Atx2 transgene selectively in the adult fly brains by shifting the temperature from 18°C to 30°C. Total brain RNA was isolated and RNA edits were identified and compared to Atx2-ADARcd, similar to Singh et al 2021[83]. (B) Domain map of Atx2-ADARcd constructs used for TRIBE analysis. (C) Comparisons of genes and edits identified by TRIBE between Atx2-ADARcd and Atx2ΔPAM2-ADARcd targets. (D) Most Atx2ΔPAM2 targets identified by TRIBE are unique and not edited in Atx2WT, suggesting that these new targets bound by Atx2ΔPAM2 are not native Ataxin-2 granule targets. (E) Comparisons of the editing efficiency ratio of common edits between Atx2WT vs Atx2ΔPAM2 show a much lower editing efficiency in Atx2ΔPAM2 compared to Atx2WT. This is despite comparable expression levels of the Atx2 fusion protein forms in brains (panel A of I Fig in S1 File). (F) PAM2 deletion results in loss of 3’UTR specificity seen in Atx2WT and LSm deletion TRIBE target mRNAs. Atx2WT and Atx2ΔLSm-ADARcd data are extracted from [83]. Controls and comparisons of replicates are summarised in I Fig in S1 File. https://doi.org/10.1371/journal.pgen.1011251.g001

In contrast to Atx-2 forms lacking LSm or LSm-AD domains [83], Atx2ΔPAM2-ADARcd edited significantly fewer RNA targets than wild-type Atx2-ADARcd (108 genes and 165 edits vs 256 genes and 317 edits, Fig 1B and 1C, and Table B in S1 File). This is despite comparable protein and RNA expression levels (panels A and B of I Fig in S1 File). More striking, the cohort of mRNAs edited by the ΔPAM2 mutant form differed extensively from the largely overlapping cohorts edited by either wild-type forms of Atx2 (panels C and D of I Fig in S1 File). Of the 108 genes edited by Atx2ΔPAM2-ADARcd, 36 were also targets of Atx2-ADARcd, the remaining 72 were unique. (Fig 1C and 1D, and Table B in S1 File). Fifty edit sites were common between the Atx2ΔPAM2 and Atx2WT targets. Those sites were edited with much lower efficiency in Atx2ΔPAM2 as compared to Atx2WT (Fig 1E).

The location of the edits made by Atx2ΔPAM2-ADARcd also differed dramatically as to where they occurred relative to the coding sequences of the target mRNAs (Fig 1F). While edits made by wild-type and ΔLSm forms of Atx2-ADARcd were greatly enriched in the 3’UTR of the mRNAs, Atx2ΔPAM2 targets were edited indiscriminately all along the mRNA length (Fig 1F).

Taken together, these data identify the PAM2 motif as necessary for Atx2 engagement with its correct mRNA targets. The PAM2 motif interacts with PABP, which binds polyA tracts at the 3’ end of mRNAs [88]. The distribution of targets across the expression level spectrum of all sequenced mRNAs further supports the role of the PAM2 domain in specifically associating with a subset of mRNAs (panel C of I Fig in S1 File). Therefore, the data point to a role for the structured PAM2:PABP interaction in guiding the association of Atx2 with mRNAs and for the subsequent inclusion of these mRNAs in Atx2-containing granules.

If Atx2-ADARcd edits of target mRNAs occur predominantly in the mRNP granules [83], then the ability of Atx2ΔPAM2-ADARcd to edit some target mRNAs would suggest that the PAM2 motif is not essential for mRNP granule formation per se. To examine this, we expressed wild-type and ΔPAM2 mutant forms of GFP-tagged Atx2 under the control of a native genomic promoter in Drosophila S2R+ cells. Atx2 overexpression in S2 cells induced the formation of mRNP granules closely related to SGs, containing endogenous Atx2 and various SG proteins as previously reported (Fig 2A) [61,83]. Similar expression of Atx2ΔPAM2-GFP also induced granule formation. However, these granules were compositionally distinct from those induced by Atx2-GFP. While they contained some SG markers present on Atx2-granules, e.g., Me31B and Rox8 (Drosophila homologs of DDX6 and TIA1), they did not contain others such as PABP, Caprin and dFMRP that are present in both wild-type Atx2-GFP and Atx2ΔLSm-GFP granules (Fig 2B, and III Fig in S1 File).

PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. The presence of the PAM2 domain affects the protein composition of Atx2-GFP granules in S2 cells. (A) Over-expression of Atx2-GFP in unstressed Drosophila S2 cells induces the formation of Atx2GFP granules to which various SG markers, namely, Caprin, dFMRP, PABP, Me31B and Rox8, co-localize. (B) Deletion of the PAM2 affects the Atx2-GFP granule composition. Over-expression of Atx2ΔPAM2-GFP in S2 cells still induces the formation of granules, but some SG markers fail to co-localize in these, notably dFMR, Caprin and PABP. In contrast, co-localisation with dFMR, Caprin and PABP is not affected by the deletion of the LSm domain in analogous experiments (III Fig in S1 File). (C) Atx2-GFP granule formation in S2 cells relies primarily on the cIDR. Deletion of the cIDR in Atx2WT, Atx2ΔPAM2 and Atx2ΔLSm, removes their ability to form granules upon overexpression. See panels A-B of II Fig in S1 File, for quantification. The scale bar in (A) applies to (B) and (C). Scale bar = 5 μm. https://doi.org/10.1371/journal.pgen.1011251.g002

RNP-granules induced by expression of wild-type, LSm and PAM2 deficient forms of Atx2GFP required the presence of the c-terminal IDR (Fig 2C). Thus, while largely dispensable for efficient mRNP assembly, the PAM2 domain plays a significant role in determining both mRNA and protein components of mRNP granules. One possibility is that the PAM2 motif directly recruits PABP and associated mRNAs to granules and indirectly recruits other proteins through their interactions with either PABP or mRNAs brought to RNP granules through Atx2PAM2:PABP interactions.

[END]
---
[1] Url: https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1011251

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.