PQN-59 antagonizes microRNA-mediated repression and functions in stress granule formation during C. elegans development

microRNAs (miRNAs) are potent regulators of gene expression that function in a variety of developmental and physiological processes by dampening the expression of their target genes at a post-transcriptional level. In many gene regulatory networks (GRNs), miRNAs function in a switch-like manner whereby their expression and activity elicit a transition from one stable pattern of gene expression to a distinct, equally stable pattern required to define a nascent cell fate. While the importance of miRNAs that function in this capacity are clear, we have less of an understanding of the cellular factors and mechanisms that ensure the robustness of this form of regulatory bistability. In a screen to identify suppressors of temporal patterning phenotypes that result from ineffective miRNA-mediated target repression during C. elegans development, we identified pqn-59, an ortholog of human UBAP2L, as a novel factor that antagonizes the activities of multiple heterochronic miRNAs. Specifically, we find that depletion of pqn-59 can restore normal development in animals with reduced miRNA activity. Importantly, inactivation of pqn-59 is not sufficient to bypass the requirement of these regulatory RNAs within the heterochronic GRN. The pqn-59 gene encodes an abundant, cytoplasmically localized and unstructured protein that harbors three essential “prion-like” domains. These domains exhibit LLPS properties in vitro and normally function to limit PQN-59 diffusion in the cytoplasm in vivo. Like human UBAP2L, PQN-59’s localization becomes highly dynamic during stress conditions where it re-distributes to cytoplasmic stress granules and is important for their formation. Proteomic analysis of PQN-59 complexes from embryonic extracts indicates that PQN-59 and human UBAP2L interact with orthologous cellular components involved in RNA metabolism and promoting protein translation and that PQN-59 additionally interacts with proteins involved in transcription and intracellular transport. Finally, we demonstrate that pqn-59 depletion results in the stabilization of several mature miRNAs (including those involved in temporal patterning) without altering steady-state pre-miRNAs levels indicating that PQN-59 may ensure the bistability of some GRNs that require miRNA functions by promoting miRNA turnover and, like UBAP2L, enhancing protein translation. AUTHOR SUMMARY Bistability plays a central role in many gene regulatory networks (GRNs) that control developmental processes where distinct and mutually exclusive cell fates are generated in a defined order. While genetic analysis has identified a number of gene types that promote these transitions, we know little regarding the mechanisms and players that ensure these decisions are robust. and in many cases, irreversible. We leveraged the robust genetics and phenotypes associated with temporal patterning mutants of C. elegans to identify genes whose depletion would restore normal regulation in animals that express miRNA alleles that do not sufficiently down-regulate their targets. These efforts identified pqn-59, the C. elegans ortholog of the human UBAP2L gene. Like UBAP2L, PQN-59 likely forms a hub for a number of RNA/RNA-binding protein mediated processes in cells including translational activation and in the formation of stress granules in adverse environmental conditions. Finally, we also demonstrate that pqn-59 depletion stabilizes mature miRNA levels further connecting this new family of RNA-binding proteins to translation and miRNA-mediated gene regulation.


114
animals continually express LIN-14 throughout larval development and reiterate L1 specific patterns 115 of cell differentiation for each of the somatic blast cells at subsequent molts. Importantly, while the 116 lin-4(ma161) allele generates a small regulatory RNA, its developmental phenotypes are 117 indistinguishable from those completely lacking the lin-4 gene [5,6]. As a consequence of these 118 temporal patterning defects, lin-4(ma161) animals lack vulval structures required for normal egg 119 laying and also fail to induce the expression of an adult-specific col-19::GFP transcriptional reporter 120 after the fourth larval molt ( Figure 1A and B). To identify suppressors, we exposed lin-4(ma161) 121 animals harboring the col-19::GFP reporter to individual clones of a genomic scale RNAi library and 122 identified dsRNAs that could restore normal col-19::GFP expression during adulthood. One of these 123 clones generated dsRNA against pqn-59, a highly conserved and uncharacterized C. elegans gene, 124 that robustly suppressed the reiterative heterochronic phenotypes of lin-4(ma161) mutants to a similar 125 level as other previously described suppressors (including lin-14, lin-28, and lin-42) (Table 1)..

126
Examination of adult lin-4(ma161) animals that had been exposed to pqn-59 dsRNAs exhibited 127 normal temporal seam cell developmental division patterns and were now able to generate alae 128 production on adult cuticles, indicative of normal seam cell temporal patterning ( Figure 1C).

129
Furthermore, in contrast to control RNAi animals, pqn-59 RNAi also suppressed the vulvaless 130 phenotypes of lin-4(ma161) animals, enabling these animals to lay eggs ( Figure 1D). Surprisingly, 131 depletion of pqn-59 activity in wild-type animals did not induce precocious deposition of adult-specific 7 132 alae at the L4 molt which distinguishes it from other previously characterized lin-4 suppressors) (Table   133 1). Furthermore, pqn-59 depletion in wild-type backgrounds only induced a mild, early expression of 134 the col-19::GFP reporter in hypodermal cells found in the head and tail regions (H0, H1, and T cells) 135 of approximately 17% of late L4-staged animals (L4.5 or later [7]).  (Table 1). Furthermore, adult-specific alae formation and Po animals were exposed to bacteria expressing indicated dsRNAs and adult F1 animals were scored for the indicated phenotype c Presence and quality of cuticular alae structures were assayed by Normarski DIC optics. Only one side of each animals was scored. d let-7(n2853) animals exhibit some col-19::GFP expression in lateral seam cells and were scored positive if there was any expression in any hypodermal cells. When animals were treated with either pqn-59, lin-42, lin-14, or lin-28

232
Western blots from homozygous pqn-59(tm2960) animals suggest that this mutation is a null allele of 233 pqn-59 that generates no detectible PQN-59 protein and protein products of the predicted size for 234 PQN-59(tm2960) are not seen in pqn-59(0); cshIs38 :GFP] ( Figure S1). Homozygous pqn-235 59(tm2650) animals that segregate from a balancer strain develop very slowly and exhibit a severe 236 reduction in brood size (Figures 3D and E). The reduction in fecundity includes a reduced capacity to 237 fertilize oocytes as well as a reduction in embryonic and early larval viability. Surprisingly, we did not 238 observe any appreciable alterations in post-embryonic cell lineage in pqn-59 mutants or precocious 239 expression of adult-specific transcriptional reporters (Table I) vivo [29]. As a consequence, proteins harboring these IDR and or "PrD-like" domains exhibit a number 267 of interesting biochemical properties including the ability to be co-precipitated by biotinylated-268 isoxazole (b-isox), which forms crystals in a temperature-dependent manner in aqueous solution and 269 co-precipitates diverse proteins harboring low complexity domains (LCDs) from cell lysates [30,31].

270
We tested if PQN-59 is precipitated by b-isox by using two approaches. First, we demonstrated that 271 b-isox precipitates PQN-59 from whole worm lysates ( Figure 4B). Because PQN-59 could be co-272 precipitated with other endogenous C. elegans proteins that are directly precipitated by the b-isox 273 compound, we purified a fragment of PQN-59 that includes the "prion-like" domains as a GFP fusion protein and demonstrated that these domains are also precipitated by b-isox compound ( Figure 4C).

275
The b-isox does not precipitate soluble GFP [31] or proteins from whole-worm extracts that cross 276 react with anti-PQN-59 antibodies or the minor E. coli proteins that co-purify with GFP-PQN-59 (PrD1-277 3) indicating the specificity of b-isox for proteins harboring LCD and PrDs.

279
To determine if the "prion-like" domains of PQN-59 also exhibit LLPS properties, purified fluorescence when a portion of the condensate is bleached (n>10). This rapid recovery was also seen 295 in condensates formed with 150mg/mL Ficoll (n = 20) (Movie S1) indicating that GFP-PQN-59 (PrD1-296 3) forms phase-separated liquid droplets in vitro ( Figure 4G). Interestingly, prolonged incubation of 297 these condensates (> 1 week) at 4°C leads to the formation of a hydrogel-like material that was 298 incapable of being re-solubilized at room temperature in aqueous buffers ( Figure 4H).

299
14 300 Under normal growth conditions, PQN-59::GFP exhibits reduced mobility in the cytoplasm and 301 this feature requires the "prion-like" domains 302 We next sought to examine the in vivo properties of PQN-59::GFP by comparing its diffusibility 303 to that of a soluble, monomeric GFP. To accomplish this, we performed FRAP experiments using a 304 strain harboring a single copy PQN-59::GFP (cshIs38) or a transgene driving soluble GFP driven from 305 the glh-1 promoter. We focused on measuring fluorescence recovery in developing oocytes where 306 both proteins are localized in large and accessible cytoplasmic compartment ( Figure 4I). As would be 307 expected for a small, soluble protein, photobleached regions rapidly recover fluorescence signal in 308 strains harboring soluble, monomeric GFP ( Figure 4I). In contrast, photobleached cytoplasmic 309 regions in oocytes expressing PQN-59::GFP recover fluorescence exceptionally slowly suggesting 310 that PQN-59::GFP normally exhibits a limited diffusibility in vivo. We also performed FRAP on PQN-311 59::GFP in hypodermal cells and found that recovery was also slower than that observed for similar 312 experiments with soluble GFP ( Figure S2). We then sought to determine if the "prion-like" domains 313 contributed to this feature. We integrated a PQN-59::GFP transgene that lacked the three PrDs,  , and transport (nuclear pore complex, motor proteins, microtubule binding) ( Figure 6A). In 386 addition to core complexes involved in translation, PQN-59 interacts with ALG-1, a core miRISC 387 component whose mutation elicits phenotypes that are suppressed by pqn-59(RNAi) (Figure 2), and 388 GTBP-1 which colocalizes with PQN-59 in stress granules ( Figure 5).

389
The structural relationships between PQN-59 and human UBAP2 and UBAP2L (and Lingerer) 390 ( Figure S3) as well as their relationship to stress granule formation suggests that these proteins may  Figure S4).

408
In order to both visualize PQN-59 interacting proteins into functional groups, we took further reduced. To test this hypothesis, we measured the levels of multiple miRNAs in animals that were 427 exposed to control or pqn-59 dsRNAs and extracted total RNA from late L4-staged animals. We then 428 examined both miRNA processing and steady state levels using quantitative northern blots. As shown 429 in Figure 7A, RNAi-mediated depletion of pqn-59 increases the relative abundance of multiple, fully-    Figure 7B). In contrast to what happens in wild-type animals, animals 520 lacking full activity of heterochronic miRNAs (as exemplified for lin-4(ma161) mutants) are unable to 521 dampen the expression of their target mRNA below critical threshold required for the bistable 522 transition in temporal cell fate ( Figure 7B). As a consequence, lin-4(ma161) animals exhibit 523 heterochronic phenotypes that are indistinguishable from those in animals that completely lack lin-4 524 (as exemplified for lin-4(e912) mutants) [6]. As demonstrated in Figure 7A, pqn-59 depletion results 525 in the stabilization of many mature miRNAs. We hypothesize that the potentially generalized 526 stabilization of mature miRNAs elicited by depleting pqn-59 expression may enable the levels of 527 critical miRNAs to increase to a level where they can now effectively dampen lin-14 expression 528 ( Figure 7C model 1). In addition to this potential mechanism, PQN-59, like UBAP2L, may promote 529 general protein translation. In this context, depletion of pqn-59 reduces the normal expression levels 530 of miRNA target genes (like lin-14) to a level that is closer to the threshold that defines the bistable 531 switch between cell fate specification ( Figure 7C model 2). Therefore, pqn-59 may function to 532 normally assure that this bistable switch that defines the L1 to L2 transition in C. elegans larval 533 development is not inappropriately crossed unless a sufficient miRNA is expressed. Both of these 534 models are consistent with pqn-59 functioning outside of the normal heterochronic GRN and 24 535 furthermore explain the observation that pqn-59(RNAi) is an efficacious suppressor of multiple miRNA 536 loss-of-function phenotypes but that pqn-59 depletion is incapable of bypassing the activities of these 537 genes.