Identification of novel nesprin-1 binding partners and cytoplasmic matrin-3 in processing bodies

Abstract

Nesprins are highly conserved spectrin repeat-containing scaffold proteins predominantly known to function at the nuclear envelope (NE). However, nesprin isoforms are emerging with localizations and scaffolding functions at sites away from the NE, suggesting their functions are more diverse than originally thought. In this study, we combined nesprin-1 coimmunoprecipitations with mass spectrometry to identify novel nesprin-1 binding partners for isoforms that localize to subcellular compartments beyond the NE. We show that one of these interactors, matrin-3 (matr3), localizes to mRNA processing bodies (PBs), where we have previously shown a nesprin-1 isoform to localize. Furthermore, we show that Matr3 is part of PB mRNP complexes, is a regulator of miRNA-mediated gene silencing, and possibly shuttles to stress granules in stressed cells. More importantly, we identify a new C-terminally truncated Matr3 isoform that is likely to be involved in these functions and PB localization. This study highlights several novel nesprin-1 binding partners and a new function and localization for Matr3 in cytoplasmic RNA granules.

FIGURE 1:

Identification of novel nesprin-1 binding partners. (A) Schematic illustrating the location of pAbN4 and pAbN5 relative to the nesprin-1 giant, nesprin-1β, and smaller KASH-less nesprin-1 isoforms. (B) SDS–PAGE and silver-stained gel of pAbN4 and pAbN5 co-IPs. Table summarizes interacting partners identified by mass spectrometry that were present in both nesprin-1 immune complexes but not the IgG control. (C) Validation of mass spectrometry binding partners by co-IPs and Western blotting (N = 4). (D) GST pull downs using SRs located in the central rod region of nesprin-1 at bait (N = 3).

FIGURE 2:

Matr3 is required for miRISC function. (A) Western blotting demonstrating efficient knockdown of matrin3, DDx5, and p54 using their respective siRNAs (N = 4). (B) Cells depleted of Matr3 have attenuated miRISC function of a Let-7a reporter assay. p54 knockdown cells served as a positive control (N = 4). (C) Endogenous Let-7a does not suppress activity of the reporter under any knockdown condition when the Let-7a binding sites are mutated (N = 4). (C, D) Cells depleted of Matr3 have attenuated miRISC function of CXCR4 reporter assay (N = 3). *, p < 0.05; **, p < 0.01; ***, p < 0.001; one-way ANOVA, Dunnett’s post hoc test.

FIGURE 3:

Matr3 localizes to PBs. (A) Schematic representation illustrating the structure of matr3 and the location of M3N and M3C antibodies. (B) M3C detects the nuclear matrix in U2OS cells. M3N detects the nuclear matrix and cytoplasmic foci that colocalize with PB markers Hedls (N = 100–300 over four experimental repeats).

FIGURE 4:

Unique matr3 is present within PBs. (A) Co-IPs with M3N and M3C show that only M3N immune complexes contain PB proteins. PSF served as a positive control (N = 4). (B) M3N detects two Western bands in U2OS whole-cell lysates. Subcellular fractions show the larger ∼95 kDa band is present in the nuclear fraction, and the smaller ∼50 kDa band is present in the cytosol. Lamin B1 and α-tubulin were used to measure purity of nuclear and cytoplasmic fractions respectively (N = 4). (C) PB proteins coimmunoprecipitate with the smaller matr3 variant, while PSF co-IPs with the larger form (N = 4). (D) si-M3N knocks down both M3N bands detected by Western blotting, while M3C only knocks down the larger molecular weight band (N = 5). (E) si-M3N knocks down cytoplasmic foci and nuclear matr3 staining. si-M3C does not affect PB but does eliminate nuclear staining. Cytoplasmic foci knockdown was determined by cells having <4 foci (N = 400 over four experimental repeats). Nuclear matr3 levels were measured by nuclear matr3 fluorescence integrated density (N = 100 over four experiments). *, p < 0.05; ****, p < 0.0001; one-way ANOVA, Dunnett’s post hoc test.

FIGURE 5:

Matr3 ZnF1 and RBD1 are required for PB localization. (A) Flag-tagged C-terminal truncation constructs were created and transfected into U2OS cells. The table indicates which constructs had cytoplasmic and nuclear localizations (N = 30–50 over three experimental repeats). Flag-469 did have weak nuclear staining in a small population of transfected cells (13.33% of cells). (B) Flag-469 foci formed colocalize with PB marker p54 and induce SG formation, as indicated by PABP-1 staining. After heat shock (42°C, 45 min), Flag-469 moves from PBs to SGs. Colocalization of Flag-469 with p54 PBs and PABP-1 SGs before and after SGs were quantified using Pearson’s colocalization coefficient (N = 15 over three experimental repeats). (C) GST-469 pulls down Ago2 and p54 from U2OS cell extract (N = 3). (D) Flag-469 has reduced interactions with Ago2 and p54 after heat shock. *, p < 0.05; **, p < 0.01; ****, p < 0.0001; Student’s t test.

FIGURE 6:

After heat shock, endogenous matr3 does not colocalize with SGs; endogenous matr3 foci have reduced colocalization with Hedls PBs; and Matr3 foci have enhanced colocalization with PABP-1 after heat shock, even though the degree of colocalization remains weak. Furthermore, nuclear matr3 levels were reduced after heat shock. Colocalizations quantified using Pearson’s colocalization coefficient (N = 15 over three experimental repeats). ****, p < 0.0001; Student’s t test.