The PUF RNA-binding protein, FBF-2, maintains stem cells without binding to RNA

Abstract

Like all canonical PUF proteins, Caenorhabditis elegans FBF-2 binds to specific RNAs via tripartite recognition motifs. Here, we report that an FBF-2 mutant protein that cannot bind to RNA is nonetheless biologically active and maintains stem cells. This unexpected result challenges the conventional wisdom that RBPs must bind to RNAs to achieve biological activity. Also unexpectedly, FBF-2 interactions with partner proteins can compensate for the loss of RNA binding. FBF-2 only loses biological activity when its RNA-binding and partner interactions are both defective. These findings highlight the complementary contributions of RNA-binding and protein partner interactions to the activity of an RNA-binding protein.

Keywords: C. elegans; PUF RNA-binding protein; PUF partnerships; germline stem cells; sperm/oocyte cell fate decision.

FIGURE 1.

FBF-2 in vivo function requires both RNA- and protein-binding residues. (A) FBF binds to RNA and protein partners via distinct interfaces (magenta for RNA, blue for partners) to control various activities: RNA repression, blunted end; RNA activation, arrowhead; RNA binding, square end. (ORF) Open reading frame; (FBE) FBF-binding element. (B,C) Crystal structure of FBF-2 (B, surface; C, ribbon) binding to RNA (PDB: 3K5Y) (Wang et al. 2009). RNA-binding residues, magenta; Y479 partner interface, blue. TRMs in each PUF repeat mediate RNA binding; TRM7 is highlighted in C. (D) fbf-2 mRNA and FBF-2 protein features. Untranslated regions (gray boxes), coding regions (white boxes), introns (peaked lines), PUF repeats (black ovals). Sites for insertion of 3xFLAG and relevant mutations are indicated. (E) FBF-2 TRM mutant effects on germline fates, scored in DAPI-stained extruded gonads. GSCs, percentage of germlines with stem cells maintained to the distal end; s/o switch, percentage of germlines with a successful sperm-to-oocyte switch. n, number of gonads scored. TRM7^mut = (S453A, H454A, E457A), TRM6^mut= (N415A, Y416A, Q419A). (F–I) Representative z-projection images of extruded adult gonads, stained for DNA (DAPI, cyan), sperm (αSP56, green), and a cell cycle marker, phosphohistone H3 (αPH3, yellow). αPH3 marks cells in both mitotic and meiotic G2/M phase. GSC maintenance is inferred from mitotic divisions in the distal gonad (αPH3-positive staining, yellow arrows); spermatogenic meiotic divisions occur more proximally (αPH3-positive staining, green arrows), where SP56-positive staining indicates sperm differentiation. Dotted line marks gonad boundary; asterisk marks distal end. Twenty micrometers of scale bar in F applies to F–I.

FIGURE 2.

Y479A TRM7^mut behaves like a null when assayed in the presence of wild-type FBF-1. (A) Representative z-projection images of extruded adult gonads, stained for FLAG:FBF-2 (αFLAG, green) and DNA (DAPI, gray). Dotted line marks gonad boundary; asterisk marks distal end. Twenty micrometers of scale bar in the top left image applies to all images. Yellow line marks approximate length of the progenitor zone (PZ). (B) FBF-2 mutant defects in the presence of wild-type FBF-1. PZ, progenitor zone size; fertile, animals capable of producing self-progeny. All animals that were not fertile were feminized (only oocytes, Fog).

FIGURE 3.

TRM7^mut eliminates FBF-2 binding to target RNAs in vivo. (A) TRM mutations do not change FBF-2 protein abundance or immunoprecipitation efficiency. Representative western blot for RIP-qPCR experiment. Left, input lysates (1%); right, FLAG IP (1%). FLAG-tagged FBF-2 variants and GFP are immunoprecipitated; negative control (GAPDH) is not immunoprecipitated. (B) TRM mutations abolish FBF-2 mRNA binding. Heatmap depicts results from quantitative PCR of FBF target mRNAs and control mRNAs after αFLAG IP, using 3xFLAG::FBF-2 for variants and 3xFLAG::GFP for the control. Mean mRNA abundance in input (left) and IPs (right) was calculated with the comparative C_T method $\left(2^{-\mathrm{\Delta \Delta }{C}_T}\right)$ (Schmittgen and Livak 2008), using rps-25 for normalization and making all comparisons to the wild-type sample. $2^{-\mathrm{\Delta \Delta }{C}_T}$= 1, no change in mRNA level compared to wild-type; $2^{-\mathrm{\Delta \Delta }{C}_T}$< 1, less mRNA than wild-type; $2^{-\mathrm{\Delta \Delta }{C}_T}$> 1 more mRNA than wild-type (gray level scale indicated to right of heatmap). Because no specific signal was seen for negative controls eft-3 and tbb-2 in the wild-type IP sample, $2^{-\mathrm{\Delta \Delta }{C}_T}$ was not calculated (boxes containing black X). (C) Effect of FBF-2 mutations on binding to gld-1 RNA. Example bar graph of $2^{-\mathrm{\Delta \Delta }{C}_T}$ values for one of the target RNAs tested, gld-1. No significant differences in gld-1 levels were seen in the input. The gld-1 RNA abundance is significantly different in immunoprecipitated samples. (**) P = 0.0017, (****) P < 0.0001.

FIGURE 4.

Models for effects of RNA binding and protein partner on FBF-2 function. (A–D) Models to illustrate how RNA binding and protein partner interactions modulate FBF-2 function. Target RNAs (straight lines) with a cap (circle) at 5′ end, open reading frame (ORF), and 3′ untranslated region (3′ UTR) containing an FBF-binding element (FBE). (GSCs) Germline stem cells; (s/o switch) sperm/oocyte fate switch. (A) Wild-type FBF-2 binds to an FBE in its target RNA and also interacts with protein partner complexes. These two binding interactions are sufficient for both GSC self-renewal (left) and the s/o switch (right). Both GSC-promoting (yellow) and s/o switch- promoting (green) partner complexes interact via the R7/R8 loop interface and Y479 (blue). (B) The FBF-2 mutant protein, Y479A, does not bind to R7/R8 loop-dependent partner complexes, and its binding to certain RNAs is affected. Y479A cannot execute the s/o cell fate switch but remains capable of GSC self-renewal. See Carrick et al. (2024) for details. (C) The FBF-2 mutant protein, TRM7^mut, does not bind an FBE but does bind R7/R8 loop-dependent protein partner complexes. TRM7^mut remains capable of GSC self-renewal (left) but not for the s/o switch (right). See text for further explanation. (D) TRM7^mut Y479A double mutant no longer binds RNA or protein partners and loses its ability to accomplish either germline function.

Brian H. Carrick