C. elegans nucleolar RG repeats are sufficient for nucleolar accumulation but insufficient for sub-nucleolar compartmentalization

27 Intrinsically disordered arginine-glycine (RG) repeat domains are enriched in multilayered 28 biomolecular condensates such as the nucleolus. C. elegans nucleolar RG repeats are 29 dispensable for nucleolar accumulation and instead contribute to the organization of sub-30 nucleolar compartments. The sufficiency of RG repeats to facilitate sub-nucleolar 31 compartmentalization is unclear. In this study, we drive expression of full-length RG repeats in 32 the C. elegans germline to test their ability to localize to nucleoli and organize into nucleolar 33 sub-compartments in vivo. We find that repeats accumulate within germ cell nucleoli but do not 34 enrich in the correct sub-compartment. Our results suggest RG repeats may indirectly influence 35 nucleolar organization by creating an environment favorable for sub-nucleolar 36 compartmentalization of proteins primarily based on their function within the nucleolus. 37 38

39 The organization of functionally related proteins and nucleic acids into biomolecular 40 condensates (BMCs) has emerged as a central theme in cell biology. Dozens of unique BMCs 41 have been described, ranging in composition from one to hundreds of proteins (Rostam et al., 42 2023). Some, such as nucleoli and germ granules, are comprised of multiple, immiscible 43 compartments (Brangwynne et al., 2009, 2011; Phillips et al., 2012; Feric et al., 2016; Banani et 44 al., 2017; Wan et al., 2018; Marnik and Updike, 2019; Manage et al., 2020). The mechanisms 45 by which proteins are targeted to the correct BMC and organized into sub-compartments are 46 unclear, especially in the cells of living animals. 47 48 Many BMCs form through condensation of proteins with intrinsically disordered regions (IDRs). 49 IDRs lack a stable 3-dimensional structure and shuffle through thousands of conformations, 50 making it challenging to characterize their contribution to BMC dynamics (Holehouse and 51 Kragelund, 2023). Although IDRs are enriched in BMC proteins, IDRs are not necessarily 52 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 3 determinants of condensation and their presence is not always required for targeting of a protein 53 to the correct BMC (Girard et al., 1994; Snaar et al., 2000; Marnik et al., 2019; Martin and 54 Holehouse, 2020; Spaulding et al., 2022). 55 56 One type of IDR consists of arginine-glycine (RG) repeats. RG repeats are enriched in many 57 BMCs, including nucleoli, germ granules, and stress granules (Thandapani et al., 2013). RG 58 repeats bind to RNA and form both homotypic and heterotypic interactions with other IDRs 59 through electrostatic interactions, hydrogen bonding, and p-stacking (Hanakahi et al., 1999; 60 Takahama et al., 2011; Chong et al., 2018). The strength of these interactions increases with 61 multivalency. Many RG sequences include regularly spaced aromatic residues, such as tyrosine 62 and phenylalanine, which contribute directly to condensation (Lin et al., 2017; Wang et al., 63 2018). 64 65 The complex in vivo environment influences condensation and BMC dynamics (Kim et al., 66 2023). In this study, we take advantage of the visually accessible and well-characterized C. 67 elegans germline to study RG repeat function in a living animal (Figure 1A). C. elegans germ 68 cells reside in a shared cytoplasm, with each nucleus surrounded by a nuclear membrane 69 spotted with germ granules. Germ granules sit on top of nuclear pores, where they are sites of 70 small RNA production and post-transcriptional gene regulation (Phillips and Updike, 2022) 71 (Figure 1B). Inside each germ cell nucleus is a 2-3 micron-wide nucleolus. Nucleoli of mammals 72 and other complex eukaryotes are partitioned into 3-5 sub-compartments that are thought to 73 correspond to specific stages of ribosome biogenesis (Feric et al., 2016; Stenström et al., 2020; 74 Shan et al., 2023). Nucleoli of lower eukaryotes, such as yeast and C. elegans, appear to 75 contain only 2 compartments: (1) the fibrillar center (FC) where rRNA is transcribed and 76 chemically modified and (2) the granular component (GC) where ribosomal subunits are 77 assembled (Lafontaine et al., 2020; Spaulding et al., 2022; Tartakoff et al., 2022). The RG 78 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 4 repeat-containing proteins Fibrillarin (FIB-1) and GAR1 (GARR-1) are found in the FC, where 79 they function in rRNA methylation and pseudouridylation, respectively (Tollervey et al., 1993; 80 Bousquet-Antonelli et al., 1997). The RG repeat-containing protein Nucleolin (NUCL-1) is found 81 in the GC where it functions in several aspects of ribosomal processing (Bouvet et al., 1998; 82 Ginisty et al., 1998; Roger et al., 2003; Rickards et al., 2007). Approximately 50% of nucleoli in 83 the adult germline contain nucleolar vacuoles, which are void of nucleolar components (Xu et 84 al., 2023) (Figure 1B,C). 85 86 We recently cataloged all RG repeats in C. elegans and found that nucleolar and germ granule 87 repeats contain distinctive phenylalanine and tyrosine-rich consensus motifs, respectively. 88 Despite the presence of a nucleolar phenylalanine-rich motif, RG repeats are dispensable for 89 localization of proteins to the nucleolus. Instead, repeats organize proteins into nucleolar sub-90 compartments, and their removal leads to FC/GC mixing (Figure 1D) (Spaulding et al., 2022). 91 How RG repeats direct sub-nucleolar compartmentalization is unclear, although recent work 92 demonstrates they are not sufficient for compartmentalization in Xenopus oocytes (Lavering et 93 al., 2023). In this study we drive expression of full-length nucleolar and germ granule RG 94 repeats in the C. elegans germline and perform super-resolution imaging of living worms to test 95 (1) if repeats independently accumulate in nucleoli and (2) if nucleolar repeats contain an 96 intrinsic ability to organize into the correct sub-nucleolar domain (Figure 1D). 97 98

99 Driving RG repeat expression in the C. elegans germline 100 To visualize RG repeats exclusively within the C. elegans germline, we modified a system in 101 which endogenous GLH-1 (Vasa) drives expression of a super-folder GFP (sGFP) (Goudeau et 102 al., 2021). GLH-1 is one of the most highly expressed genes in the germline and its use as a 103 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 5 driver promotes abundant and uniform expression of RG repeats (Campbell and Updike, 2015). 104 In this system GLH-1 is separated from sGFP by a self-cleaving T2A peptide (GLH-105 1::T2A::sGFP) and GLH-1 expression is left intact. CRISPR/Cas9 genome editing was used to 106 place full-length RG repeats followed by the SV40 nuclear localization signal (NLS) directly after 107 sGFP in this strain. Upon T2A self-cleavage in the resulting strain (GLH-108 1::T2A::sGFP::RG::NLS), the sGFP-tagged RG repeat with NLS is released from GLH-1, 109 resulting in fluorescently-tagged RG repeats in the nuclei of germ cells. As a control for possible 110 GFP nucleolar enrichment and compartmentalization, sGFP with only the SV40 NLS was also 111 included in each experiment (GLH-1::T2A::sGFP::NLS) (abbreviated GFP::NLS) (Figure 1E). 112 Due to incomplete T2A cleavage from germ granule localized GLH-1, we expect to observe 113 some residual GFP signal in germ granules within all strains using the GLH-1 driver system. 114 115 We inserted full-length RG repeat domains from three nucleolar proteins, NUCL-1, FIB-1, and 116 GARR-1, into the expression system. The NUCL-1 repeat is the longest nucleolar RG repeat in 117 C. elegans at 176 amino acids in length. The FIB-1 repeat is the third longest nucleolar repeat 118 at 106 amino acids in length. Both repeats contain a phenylalanine-rich nucleolar consensus 119 motif (“FRGGDRGGFR”) (Spaulding et al., 2022). The GARR-1 protein has two RG repeat 120 domains, one at the N- and one at the C-terminus. The N-terminal 48 amino acid-long domain is 121 phenylalanine-rich but does not contain the nucleolar consensus motif. The C-terminal 70 amino 122 acid-long domain does contain the motif. We included the N-terminal RG repeat in this study to 123 test if a nucleolar repeat without the consensus motif can accumulate and compartmentalize 124 within nucleoli (Supplemental Figure 1A). 125 126 RG repeats from C. elegans germ granule proteins contain a tyrosine-rich consensus motif 127 (“RGGRGGYRGGD”) (Spaulding et al., 2022). To test if an RG repeat with a germ granule motif 128 will accumulate in nucleoli, we inserted the 53 amino acid-long PGL-1 RG repeat into the 129 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 6 expression system (Supplemental Figure 1A). To mark nucleoli, wrmScarlet was placed on the 130 C-terminus of endogenous GARR-1 (GARR-1::wSc) in each GFP::RG::NLS strain and the 131 GFP::NLS strain. Finally, in a strain containing wrmScarlet-tagged GARR-1, we also placed 132 sGFP at the C terminus of endogenous NUCL-1 (NUCL-1::GFP) to provide examples of nucleoli 133 with correctly partitioned FC and GC sub-compartments. Western blots confirmed the expected 134 size of each RG repeat insert. GFP::GARR-1 RG::NLS and GFP::PGL-1 RG::NLS repeats are 135 expressed at the same level as GFP::NLS, but GFP::NUCL-1 RG::NLS and GFP::FIB-1 136 RG::NLS are expressed at lower levels (Supplemental Figure 1B,C). Evolutionarily conserved 137 strict regulation of the synthesis and degradation of proteins with large IDRs has been 138 demonstrated in organisms from yeast to humans (Gsponer et al., 2008). As the same 139 endogenous driver is used for expression, lower levels of NUCL-1 and FIB-1 RGs suggest there 140 may be cellular mechanisms in place to limit over-expression of the longest IDRs. 141 142 RG repeats accumulate in germ cell nucleoli 143 Because RG repeats are dispensable for nucleolar accumulation of proteins in C. elegans germ 144 cells, we asked if they are sufficient for accumulation. To visualize localization of RG repeats, 145 we performed confocal imaging of live, adult worms. Endogenously-labeled GARR-1 and NUCL-146 1 proteins accumulate in germ cell nucleoli throughout the entire germline (Supplemental Figure 147 2A). The expected background GFP signal is visible in germ granules in all strains using the 148 GLH-1 driver system, including the control GFP::NLS strain. This signal is a useful marker that 149 defines the nuclear boundary (Supplemental Figure 2B, white arrow). GFP::NLS is observed 150 throughout the nuclei of germ cells and a slight accumulation is visible in some nucleoli 151 (Supplemental Figure 2B). In contrast, GFP-labeled RG repeats from NUCL-1, FIB-1, GARR-1, 152 and PGL-1 are enriched in all germ cell nucleoli (Supplemental Figure 2C-F). 153 154 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 7 To measure nucleolar enrichment we performed super-resolution confocal imaging of meiotic, 155 pachytene-stage germ cells and calculated the ratio of mean nucleolar fluorescence intensity to 156 mean nucleoplasmic fluorescence intensity for at least 50 cells per strain. Endogenous GARR-1 157 and NUCL-1 demonstrated high nucleolar enrichment, as expected (Figure 2A). In contrast, 158 GFP::NLS demonstrates a nucleolar enrichment of close to zero, which is significantly lower 159 than endogenous GARR-1 in the same germ cells (Figure 2B). This result confirms that the GFP 160 tag alone does not contribute to nucleolar enrichment in the GFP::RG repeat strains. The 161 NUCL-1, FIB-1, GARR-1, and PGL-1 RG repeats all enrich in nucleoli, but with decreased 162 efficiency compared to endogenous GARR-1 in the same cells (Figure 2C-F). 163 164 Directly comparing the nucleolar:nucleoplasmic fluorescence intensity ratios of all strains 165 confirms that each RG repeat accumulates in nucleoli less efficiently than GFP-tagged NUCL-1 166 or wrmScarlet-tagged GARR-1 (Figure 2G), but more efficiently than GFP::NLS (Figure 2H). 167 The NUCL-1 RG repeat is the longest repeat and displays the greatest nucleolar enrichment, 168 most likely due to its high multivalence and capacity to engage in condensation-promoting 169 molecular interactions. The germ granule PGL-1 RG repeat shows the second-highest nucleolar 170 enrichment, despite being shorter than the FIB-1 RG repeat. This result may point to the 171 stronger contribution of tyrosine residues to condensation compared to phenylalanine (Bremer 172 et al., 2022). The GARR-1 repeat is phenylalanine-rich but does not contain the nucleolar 173 consensus motif. Regardless, the GARR-1 repeat enriches in nucleoli, suggesting that non-motif 174 phenylalanines in the GARR-1 repeat contribute to condensation behavior in a manner similarly 175 to the consensus motif. In summary, nucleolar repeats both with (NUCL-1 and FIB-1) and 176 without (GARR-1) the consensus motif and ranging in size from 176 to 48 amino acids 177 accumulate in germ cell nucleoli. In addition, the germ granule PGL-1 RG repeat with a tyrosine-178 rich consensus motif accumulates in germ cell nucleoli. 179 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 8 180 RG repeats do not display large-scale sub-nucleolar compartmentalization 181 Because RG repeats are required for enrichment of nucleolar proteins into the FC and GC, we 182 asked if they are sufficient for this enrichment. To measure sub-nucleolar compartmentalization 183 of RG repeats, we again used the super-resolution images of pachytene germ cells. The 184 coefficient of variation (CV) is a measure of fluorescence variation and was calculated by 185 dividing the standard deviation by the mean intensity of GFP fluorescence in individual nucleoli. 186 In each nucleolus, the CV was also measured for endogenous GARR-1 as a positive control for 187 compartmentalization. A high CV indicates heterogeneous distribution and more 188 compartmentalization, while a low CV indicates homogeneous distribution and less 189 compartmentalization (Spaulding et al., 2022). 190 191 In the strain with endogenously-tagged NUCL-1 and GARR-1, NUCL-1 enriches in the GC and 192 GARR-1 enriches in the FC, as expected. NUCL-1 displays a slightly higher CV compared to 193 GARR-1, most likely because the larger GC contains more areas of enrichment and depletion 194 throughout nucleoli (Figure 3A). As expected, GFP::NLS shows no compartmentalization within 195 nucleoli and has a significantly lower CV than endogenous GARR-1 (Figure 3B). NUCL-1, FIB-196 1, and GARR-1 RG repeats do not display large-scale compartmentalization and have 197 significantly lower CVs than endogenous GARR-1 (Figure 3C-E). Endogenous NUCL-1 protein 198 is restricted to the GC, but the NUCL-1 RG repeat localizes throughout the entire nucleolus and 199 its expression overlaps with endogenous GARR-1 in the FC (Figure 3C, white arrows). 200 Endogenous FIB-1 and GARR-1 proteins are restricted to the FC, but both FIB-1 and GARR-1 201 RG repeat expression spills over into areas of endogenous GARR-1 depletion that correspond 202 to the GC (Figure 3D,E white arrows). In some nucleoli, RG repeat expression is lightest in 203 spots that also lack endogenous GARR-1 expression, areas which most likely correspond to 204 nucleolar vacuoles (Figure 3D-F, pink arrows) (Spaulding et al., 2022; Xu et al., 2023). As a 205 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 9 native component of germ granules, the PGL-1 RG repeat was not expected to enrich in a 206 specific nucleolar sub-compartment and shows significantly decreased CV compared to 207 endogenous GARR-1 (Figure 3F). Although RG repeats have significantly lower CVs compared 208 to endogenous GARR-1 or NUCL-1 (Figure 3G), they do have significantly higher CVs than 209 GFP::NLS, suggesting some micro-organization within nucleoli (Figure 3H). 210 211 As a second way to measure RG repeat organization, we measured the Mander’s 212 Colocalization Coefficient (MCC) for wSc and GFP using the same images used for CV 213 analysis. MCC measures the fraction of one fluorophore that colocalizes with another. The 214 Mander’s Overlap Coefficient 1 (M1) reports on the fraction of wSc fluorescence in areas with 215 GFP fluorescence. The larger M1, the stronger the evidence for colocalization. Confirming the 216 CV analysis, the overlap of GARR-1::wSc with GFP::NLS, NUCL-1RG, FIB-1RG, GARR-1RG, 217 and PGL-1RG is significanctly higher than with endogenous NUCL-1 (Supplemental Figure 3). 218 219 As a third way to measure RG repeat compartmentalization, we created profile plots of mean 220 wrmScarlet and GFP fluorescence across at least 50 pachytene nucleoli from each strain. The 221 large standard deviation of endogenous NUCL-1 and GARR-1 fluorescence demonstrates the 222 heterogeneous distribution of NUCL-1 in the GC and GARR-1 in the FC (Supplemental Figure 223 4A). In contrast, GFP::NLS and GFP-tagged RG repeats display smaller standard deviations 224 across nucleoli, indicating more homogeneous distribution and a lack of precise 225 compartmentalization (Supplemental Figure 4B-F). In summary, although nucleolar RG repeats 226 are required for sub-nucleolar organization, they are insufficient to recognize and enrich in the 227 appropriate nucleolar sub-compartment. 228 229

230 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 10 RG repeat-containing nucleolar proteins such as Nucleolin, Fibrillarin, and GAR1 are precisely 231 partitioned into immiscible sub-nucleolar compartments (Figure 4A). In C. elegans, deletion of 232 endogenous nucleolar RG repeats leads to a loss of sub-nucleolar compartmentalization (Figure 233 4B) (Spaulding et al., 2022). How do RG repeats direct sub-nucleolar organization? Repeats 234 may form interactions with proteins and/or RNA in one sub-compartment, thereby enriching the 235 protein in that compartment and excluding it from others. If this were the case, we would expect 236 full-length repeats to be capable of independently forming those interactions and enriching in 237 the correct sub-nucleolar compartment. To the contrary, when ectopically expressed in Xenopus 238 oocytes, RG repeats from Nucleolin and Fibrillarin do not enrich in a specific nucleolar sub-239 domain (Lavering et al., 2023). Our study confirms these results in C. elegans. One potential 240

of this study comes from the lower expression of NUCL-1 and FIB-1 RG repeats. 241 Although it is challenging to precisely control expression levels in a living animal, studying RG 242 repeats in their natural context is crucial for determining physiologically relevant functions. As 243 demonstrated through visualization of endogenous GARR-1, concentration thresholds for sub-244 nucleolar phase separation have been met and nucleoli are correctly compartmentalized. Thus, 245 regardless of over-expression levels, RG repeats should be free to enrich in pre-existing sub-246 compartments if they are capable. Instead, our study provides additional evidence from a living 247 animal that full-length RG repeats do not independently enrich in sub-nucleolar compartments 248 even when endogenous proteins are present and correctly partitioned (Figure 4C). 249 250 Our findings support a model in which RG repeats indirectly influence nucleolar organization by 251 creating an environment conducive to compartmentalization that is primarily driven by other 252 functional protein domains. For example, the methyltransferase domain of human FBL binds to 253 nascent pre-rRNA and enriches FBL in the correct sub-nucleolar compartment. The FBL RG 254 repeat controls larger-scale self-association (Yao et al., 2019). Recent work by King et al, also 255 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 11 points to the role of functional RNA and DNA binding domains tethered to acidic domains (D/E 256 tracts) as determinative factors in organization of nucleolar proteins (King et al., 2024). 257 258 C. elegans NUCL-1, FIB-1, and GARR-1 RG repeats are insufficient for sub-nucleolar 259 compartmentalization, perhaps because they do not bind functional partners. When 260 endogenous RG repeats from NUCL-1 or GARR-1 are deleted (NUCL-1DRG or GARR-1DRG), 261 large-scale sub-nucleolar compartmentalization is lost. Surprisingly, NUCL-1DRG worms are 262 healthy and fertile and GARR-1DRG worms show only mild fertility defects (Spaulding et al., 263 2022). This resilience may be restricted to organisms with a 2-compartment nucleolus and lost 264 in animals with more complex nucleolar organization. In C. elegans, NUCL-1DRG and GARR-265 1DRG may still bind functional partners and take part in ribosome biogenesis despite the 266 collapse of large-scale partitioning (Figure 4B). In summary, our data support the idea that sub-267 nucleolar organization is primarily driven by the functional interactions of structured protein 268 domains, while RG repeats contribute to large-scale assembly of compartments by influencing 269 the biophysical properties of nucleoli through additional specific or nonspecific interactions 270 (Protter et al.) (Figure 4C). 271 272 Our previous work demonstrated that repeats are dispensable for nucleolar accumulation of 273 endogenous proteins (Spaulding et al., 2022). If RG repeats are not required for nucleolar 274 accumulation and are insufficient for sub-nucleolar compartmentalization, what is the 275 significance of RG nucleolar and germ granule consensus motifs? In this study, nucleolar 276 repeats with and without the phenylalanine-rich motif accumulated within nucleoli. The germ 277 granule PGL-1 RG repeat contains a tyrosine-rich motif and also accumulates within nucleoli. 278 Molecular interactions formed by RG repeats will vary in strength depending upon the identity 279 and number of aromatic residues involved (Lin et al., 2017). Thus, distinctive consensus motifs 280 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 12 may impart nucleolar and germ granule RG repeats with individual characteristics that help tune 281 the biophysical conditions of each BMC to its function. 282 283 Defining the functions of IDRs inside a living animal is a crucial step in determining how 284 mutations in these domains alter BMC dynamics and drive human disease (Tsang et al., 2020; 285 Borcherds et al., 2021; Mensah et al., 2023). Our in vivo work supports a modulatory role of RG 286 repeat domains in condensate organization and points to functional interactions as the primary 287 drivers of nucleolar sub-compartmentalization (Choi et al., 2020; Savojardo et al., 2020; Feng et 288 al., 2021). Extensive RG repeats like those found in NUCL-1, FIB-1, GARR-1, and their human 289 homologs may act as flexible scaffolds, creating a nucleolar environment condusive to large 290 scale compartmentalization. Disease-linked mutations in RG repeats likely disrupt the finely 291 tuned biophysical properites of BMCs and lead to widespread functional consequences (Sheikh 292 et al., 2022). 293 294

295 296 Strain generation and maintenance: 297 C. elegans strains were maintained using standard protocols (Brenner 1974). Strains created for 298 this study include DUP277 glh-1(sam168[glh-1::T2A::sGFP2(1-10)::M3::NUCL-1RGG+NLS]) I; 299 DUP281 glh-1(sam171[glh-1::T2A::sGFP2(1-10)::M3::FIB-1RGG+NLS]) I; DUP282 glh-300 1(sam172[glh-1::T2A::sGFP2(1-10)::M3::PGL-1RGG+NLS]) I; DUP284 glh-1(sam174[glh-301 1::T2A::sGFP2(1-10)::M3::GARR-1NtermRGG+NLS]) I; DUP293 garr-1(sam179[garr-302 1::wrmScarlet(1-11)]) IV; DUP294 glh-1(sam168[glh-1::T2A::sGFP2(1-10)::M3::NUCL-303 1RGG+NLS]) I; garr-1(sam179[garr-1::wrmScarlet(1-11)]) IV; DUP295 glh-1(sam171[glh-304 1::T2A::sGFP2(1-10)::M3::FIB-1RGG+NLS]) I; garr-1(sam179[garr-1::wrmScarlet(1-11)]) IV; 305 DUP296 glh-1(sam172[glh-1::T2A::sGFP2(1-10)::M3::PGL-1RGG+NLS]) I; garr-1(sam179[garr-306 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 13 1::wrmScarlet(1-11)]) IV; DUP297 glh-1(sam174[glh-1::T2A::sGFP2(1-10)::M3::GARR-307 1NtermRGG+NLS]) I; garr-1(sam179[garr-1::wrmScarlet(1-11)]) IV; DUP305 glh-1 (sam182[glh-308 1::T2A::sGFP2(1-11)::NLS]) I; DUP307 glh-1 (sam182[glh-1::T2A::sGFP2(1-11)::NLS])1; garr-309 1(sam179[garr-1::wrmScarlet(1-11)]) IV; DUP311 garr-1(sam179[garr-1::wrmScarlet(1-11)]) 310 IV;nucl-1(sam186[nucl-1::sGFP(1-11)])IV. Sequence files for CRISPR-generated alleles are 311 stored on figshare (see Data Availability Statement). All strains generated for this study and 312 their associated sequence files are available upon request. 313 314 CRISPR strain construction 315 CRISPR/Cas9 genome editing was used to place a split superfolder-GFP11 tag (M3), a flexible 316 linker sequence, a full-length RG sequence, and an SV40 nuclear localization signal on the C 317 terminus of sGFP1(1-10) in the DUP223 background (GLH-1::T2A::sGFP2(1-10)). Creation of 318 the DUP223 glh-1(sam129[glh-1::T2A::sGFP2(1-10)]) I allele was previously described 319 (Goudeau et al., 2021). The 176 amino acid NUCL-1 RG repeat was inserted to create DUP277, 320 the 107 amino acid FIB-1 RG repeat was inserted to create DUP281, the 48 amino acid long N-321 terminal GARR-1 RG repeat was inserted to create DUP284, and the 53 amino acid PGL-1 RG 322 repeat was inserted to create DUP282. DUP305 was created by inserting a split superfolder-323 GFP11 tag, flexible linker, and SV40 nuclear localization signal on the C terminus of sGFP(1-324 10) in the DUP223 background. DUP293 was created by injecting GARR-1::wrmScarlet 325 CRISPR constructs into the N2 laboratory strain. DUP294, DUP295, DUP296, DUP297, and 326 DUP307 were created by crossing DUP277, DUP281, DUP282, DUP284, and DUP305 into 327 DUP293, respectively. DUP311 was created by inserting sGFP onto the C-terminus of NUCL-1 328 in DUP293. CRISPR techniques for efficient genome editing in C. elegans were followed as 329 described (Ghanta et al., 2021). All CRISPR reagents (Cas9 (Cat# 1081058), trRNA (Cat# 330 1072532), crRNAs (2nmol), and dsDNA repair templates (HDR donor blocks)) were ordered 331 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 14 from Integrated DNA Technologies, Inc (San Diego, CA). Sequences for the guide RNA and 332 repair templates are stored on figshare 333 (https://figshare.com/articles/dataset/CRISPR_reagents_for_Spaulding_Updike_2024/26662102334 ?file=48495868). All edits generated for this study were sequence verified, and sequence files 335 are stored on figshare 336 (https://figshare.com/articles/figure/Sequence_files_for_Spaulding_Updike_2024/26789935). All 337 strains generated for this study are available upon request. 338 339 Nucleolar imaging and analysis 340 L4 worms were plated at 20oC the day prior to imaging. On the day of imaging live, young adult 341 worms were mounted on agarose pads in egg buffer (25mM HEPES (Fisher, cat#BP310-1), 342 120mM NaCl (Sigma, cat#S9888, 2mM MgCl2 (Sigma, cat#M9272, 50mM KCl (Fisher, 343 cat#S77375-1, and 10mM levamisole (Thermofisher, cat# AC187870100) between the slide and 344 a No.1.5 coverslip (Fisherbrand). Images were acquired using a point scanning confocal unit 345 (LSM 980, Carl Zeiss Microscopy, Germany) on a Zeiss Axio Examiner Z1 upright microscope 346 stand (ref: 409000-9752-000, Carl Zeiss Microscopy, Germany) equipped with a Plan-347 Apochromat 63X/1.4 Oil (ref:420782-9900-799, Carl Zeiss Microscopy, Germany) objective. 348 sGFP and wrmScarlet fluorescence were excited with the 488nm Diode (0.5% laser power) and 349 the 561nm DPSS laser (0.2% power), respectively. Fluorescence was collected with Airyscan2 350 with the following detection wavelengths: sGFP from 499 to 557nm and wrmScarlet from 573-351 627nm. Images of the adult germline were acquired using standard confocal mode. Images of 352 pachytene nucleoli were sequentially acquired in Super Resolution mode (SR) at zoom 10, with 353 a line average of 1, a resolution of 292x292 pixels, 0.043 x 0.043 µm pixel size, a pixel time of 354 0.69µs, in 16-bit, and in bidirectional mode. Z-stack images were collected with a step size of 355 0.170µm with the Motorized Scanning Stage 130x85 PIEZO (Carl Zeiss Microscopy) mounted 356 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 15 on Z-piezo stage insert WSB500 (Carl Zeiss Microscopy). Microscope was controlled using Zen 357 Blue Software (Zen Pro 3.1), Airy scan images were processed using the “auto” mode and 358 saved in CZI format. Z stacks of at least 10 pachytene germ cell nucleoli were taken from each 359 of 10 worms per genotype during at least 2 separate experiments. 360 361 ImageJ/Fiji was used to quantify the coefficient of variation (CV) within individual nucleoli. A 362 single plane of the Z stack was chosen for each nucleolus that contained the maximum 363 nucleolar area. A circle was placed within individual nucleoli that covered its maximum area 364 without including background space and the following macro code was used to calculate the 365 coefficient of variation (standard-deviation divided by the mean fluorescence): 366 367 getRawStatistics( N, mean, min, max, std ); 368 print( std / mean ); 369 370 The CV was measured from at least 10 nucleoli from each of 10 worms per genotype. Data was 371 analyzed including all individual nucleoli or the mean of all nucleoli measured per worm. 372 373 ImageJ/Fiji was used to quantify the nucleolar:nucleoplasmic fluorescence intensity ratio of at 374 least 50 nucleoli from at least 5 worms per strain using the same images as used for CV 375 analysis. A single plane of the Z stack was chosen for each nucleolus that contained the 376 maximum nucleolar area. A circle of the same size was used to measure the mean fluorescence 377 intensity of 3 spots within the nucleolus, the nucleoplasm, and the surrounding cytoplasm. This 378 was performed for both GFP and wrmScarlet. The average of the 3 cytoplasmic intensity spots 379 was subtracted from the average of the 3 nucleolar and nucleoplasmic intensity spots. The 380 background-subtracted nucleolar intensity was divided by the background-subtracted 381 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 16 nucleoplasmic intensity to determine the ratio. Data was analyzed including all individual 382 nucleoli or the mean of all nucleoli measured per worm. 383 384 Fiji was also used to create fluorescence intensity profile plots of pachytene germ cells using the 385 same images as used for CV and nucleolar enrichment analysis. Dual-channel images were 386 split into 2 images and the frames were synchronized. Brightness was auto-scaled for both GFP 387 and wrmScarlet channels. A single plane of the Z stack was chosen for each nucleolus that 388 contained the maximum cell area. A 6-micron straight line was placed across the center of a 389 single cell in the GFP channel. The “plot profile” feature was used to create a plot of gray value 390 vs distance in microns. The data was saved in the list format and imported into Prism. This was 391 repeated for the wrmScarlet channel. Profile plots were created for at least 50 cells from at least 392 5 worms per strain. 393 394 Fiji was used to measure the Mander’s Colocalization Coefficient using the same images used 395 for CV and profile plot measurements. Dual-channel images were split into 2 images and the 396 frames were synchronized. A single plane of the Z stack was chosen for each image that 397 contained the maximum number of nucleoli. The JACoP plugin was used to manually threshold 398 the wSc and GFP channels and measure M1 and M2. 399 400 Worm Crosses 401 Males were generated by plating 10 L4 hermaphrodites each on 10 plates and incubating at 402 30oC for 6 hours. Plates were then shifted to 20oC and males were picked 3 days later. 403 DUP277, DUP281, DUP282, DUP284, and DUP305 males were crossed into DUP293 404 hermaphrodites. 4 F1 worms were picked from plates with 50% males (indicating successful 405 mating). 9 F2 worms with both GFP and wrmScarlet expression were picked from each F1 clone 406 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 17 using a fluorescent dissecting microscope. Homozygosity of F2 worms was determined by 407 visually screening their progeny. 408 409 Western Blotting 410 100-150ul of worms of each strain were washed off plates with dH2O and flash frozen in liquid 411 nitrogen. 100ul of solubilization buffer (300mM NaCl, 50mM Tris-HCl [pH8.0], 10mM MgCl2, 412 1mM EGTA, ½ tablet Complete protease inhibitor, 1% Triton-X, 1mM PMSF) was added to each 413 frozen worm pellet and worms were homogenized on ice for 1-2 minutes using a Fisherbrand 414 cordless mixer with disposable pestle (Kimble, item#6CJ2ZNZ). Homogenate was left on ice for 415 1 hour and vortexed every 10 minutes, followed by centrifugation at 12K for 5 minutes at 4oC. 416 The aqueous layer was transferred to a new tube and mixed with 1X Laemmli buffer (10% beta-417 mercaptoethanol (Fisher, cat#BP176-100), 4% SDS (Bio-Rad, cat#161-0203), 20% glycerol 418 (Invitrogen, cat#15514-011), .004% bromophenol blue (Sigma, cat#B5525), 0.125M Tris-Cl 419 pH6.8 (Fisher, cat#BP153-500)). Samples were then boiled for 10 minutes and spun at 12K for 420 5 minutes at room temperature. 30ug of protein as determined by Bradford protein 421 quantification assay was loaded onto a mini-PROTEAN TGX stain-free gel (Bio-Rad, 422 cat#4568084) and run at 200V for 25 minutes in SDS running buffer (.2501M Tris base (Fisher, 423 cat#BP15201), 1.924M glycine (Fisher, cat#G48-500), .0347M SDS (Bio-Rad, cat#161-0203)). 424 The gel was exposed to UV light for 5 minutes for total protein quantity detection and then the 425 contents of the gel were transferred using a trans-blot turbo transfer pack (Bio-Rad, 426 cat#1704156) on the Bio-Rad Trans-Blot Turbo Transfer System. The PVDF membrane was 427 imaged to determine total protein levels and then blocked in 5% nonfat milk in TBST (20mM 428 Tris-HCl [pH7.4], 150mM NaCl, 0.1% Tween) at room temperature for 1 hour. The membrane 429 was incubated overnight at 4oC with rabbit polyclonal anti-GFP, 1:2000 (Invitrogen, cat#A-6455) 430 in 5% milk and then washed 6X 10 minutes in TBST at room temperature. The membrane was 431 then incubated in goat anti-rabbit IgG-HRP, 1:10,000 (Bio-Rad, cat#170-6515) in 5% milk for 1 432 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 18 hour at room temperature and washed 6X 10 minutes in TBST. Finally, the membrane was 433 developed using Clarity Western ECL Substrate (Bio-Rad, cat#1705060S) and imaged on a 434 Syngene G:Box gel and blot imaging system. For images of uncropped blot see Source Data. 435 436 Statistics and Reproducibility: All imaging experiments investigating nucleolar accumulation 437 and sub-nucleolar organization (Figures 2,3; Supplemental Figures 2,3) were performed on at 438 least two independent occasions and similar results were always obtained. Imaging and image 439 analysis was not done blinded to genotype because it was performed sequentially as each 440 strain was created. When imaging DUP311 worms we would always observe some worms with 441 less precise FC/GC organization (approximately 25% of total imaged worms, observed at each 442 imaging session). This phenomenon may indicate that the sGFP and wrmScarlet tags on 443 NUCL-1 and GARR-1, respectively, are interfering with the stability of nucleolar substructure. 444 Worms with less precise FC/GC organization were still included in all analyses. Western blots 445 were performed on three independent occasions and similar results were always obtained 446 (Supplemental Figure 1). For all pairwise comparisons (Figures 2a-f,3a-f) unpaired, 2-tailed t 447 tests with Welch’s correction was performed. For comparisons of three or more groups (Figures 448 2g-i,3g-i and Supplemental Figure 1c and 3) one-way ANOVA tests with multiple comparisons 449 were performed. Statistical analysis was done using Prism software. 450 451 DATA AVAILABILITY: 452 For CRISPR/Cas9 editing experiments, sequences for the guide RNA and repair templates are 453 stored on figshare 454 (https://figshare.com/articles/dataset/CRISPR_reagents_for_Spaulding_Updike_2024/26662102455 ?file=48495868). All edits generated for this study were sequence verified and sequence files 456 are stored on figshare 457 (https://figshare.com/articles/figure/Sequence_files_for_Spaulding_Updike_2024/26789935 458 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 19 ). All strains generated for this study are available upon request. Source data for all experiments 459 are provided with the paper. 460 461

462 We would like to acknowledge support from Chris Smith in the MDI Biological Laboratory 463 (MDIBL) Sequencing Facility, Dr. Frederic Bonnet in the MDIBL Light Microscopy Facility, and 464 plate pouring services provided by the MDIBL Animal Resources Core. Schematics created with 465 BioRender.com. 466 467 FUNDING: 468 NIH NRSA F32GM143851 (E.L.S.) and NIH R35GM152109 (D.L.U.). Research reported in this 469 publication was supported by the NIGMS under the following grants: Maine INBRE NIH 470 P20GM103423 and the MDIBL COBRE NIH P20GM104318. 471 472 AUTHOR CONTRIBUTIONS: 473 E.L.S. generated worm strains, performed imaging, data analysis, and experimental design, and 474 wrote the manuscript. D.L.U. provided conceptualization and experimental design and edited 475 the manuscript. 476 477 COMPETING INTERESTS: 478 The authors declare no competing interests. 479 480 481 482 483 484 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 20

485 Banani, SF, Lee, HO, Hyman, AA, and Rosen, MK (2017). Biomolecular condensates: Organizers 486 of cellular biochemistry. Nat Rev Mol Cell Biol. 487 Borcherds, W, Bremer, A, Borgia, MB, and Mittag, T (2021). How do intrinsically disordered 488 protein regions encode a driving force for liquid-liquid phase separation? HHS Public Access. 489 Curr Opin Struct Biol 67, 41–50. 490 Bousquet-Antonelli, C, Henry, Y, Gélugne, JP, Caizergues-Ferrer, M, and Kiss, T (1997). A small 491 nucleolar RNP protein is required for pseudouridylation of eukaryotic ribosomal RNAs. EMBO J 492 16, 4770–4776. 493 Bouvet, P, Diaz, JJ, Kindbeiter, K, Madjar, JJ, and Amalric, F (1998). Nucleolin interacts with 494 several ribosomal proteins through its RGG domain. J Biol Chem 273, 19025–19029. 495 Brangwynne, CP, Eckmann, CR, Courson, DS, Rybarska, A, Hoege, C, Gharakhani, J, Jülicher, F, 496 and Hyman, AA (2009). Germline P granules are liquid droplets that localize by controlled 497 dissolution/condensation. Science (1979). 498 Brangwynne, CP, Mitchison, TJ, and Hyman, AA (2011). Active liquid-like behavior of nucleoli 499 determines their size and shape in Xenopus laevis oocytes. Proc Natl Acad Sci U S A 108, 4334. 500 Bremer, A, Farag, M, Borcherds, WM, Peran, I, Martin, EW, Pappu, R V., and Mittag, T (2022). 501 Deciphering how naturally occurring sequence features impact the phase behaviors of 502 disordered prion-like domains. Nat Chem 14, 196. 503 Campbell, AC, and Updike, DL (2015). CSR-1 and P granules suppress sperm-specific 504 transcription in the C. elegans germline. Development 142, 1745. 505 Choi, JM, Holehouse, AS, and Pappu, R V. (2020). Physical Principles Underlying the Complex 506 Biology of Intracellular Phase Transitions. Annu Rev Biophys 49, 107–133. 507 Chong, PA, Vernon, RM, and Forman-Kay, JD (2018). RGG/RG Motif Regions in RNA Binding and 508 Phase Separation. J Mol Biol 430, 4650–4665. 509 Feng, Z, Jia, B, and Zhang, M (2021). Liquid-Liquid Phase Separation in Biology: Specific 510 Stoichiometric Molecular Interactions vs Promiscuous Interactions Mediated by Disordered 511 Sequences. Biochemistry 60, 2397–2406. 512 Feric, M, Vaidya, N, Harmon, TS, Kriwacki, RW, Pappu, R V, Correspondence, CPB, Mitrea, DM, 513 Zhu, L, Richardson, TM, and Brangwynne, CP (2016). Coexisting Liquid Phases Underlie 514 Nucleolar Subcompartments Article Coexisting Liquid Phases Underlie Nucleolar 515 Subcompartments. Cell 165, 1686–1697. 516 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 21 Ghanta, KS, Ishidate, T, and Mello, CC (2021). Microinjection for precision genome editing in 517 Caenorhabditis elegans. STAR Protoc 2, 100748. 518 Ginisty, H, Amalric, F, and Bouvet, P (1998). Nucleolin functions in the first step of ribosomal 519 RNA processing. EMBO J 17, 1476. 520 Girard, JP, Bagni, C, Caizergues-Ferrer, M, Amalric, F, and Lapeyre, B (1994). Identification of a 521 segment of the small nucleolar ribonucleoprotein-associated protein GAR1 that is sufficient for 522 nucleolar accumulation. Journal of Biological Chemistry 269, 18499–18506. 523 Goudeau, J, Sharp, CS, Paw, J, Savy, L, Leonetti, MD, York, AG, Updike, DL, Kenyon, C, and 524 Ingaramo, M (2021). Split-wrmScarlet and split-sfGFP: tools for faster, easier fluorescent 525 labeling of endogenous proteins in Caenorhabditis elegans. Genetics 217. 526 Gsponer, J, Futschik, ME, Teichmann, SA, and Madan Babu, M (2008). Tight regulation of 527 unstructured proteins: from transcript synthesis to protein degradation Europe PMC Funders 528 Group. Science (1979) 322, 1365–1368. 529 Hanakahi, LA, Sun, H, and Maizels, N (1999). High Affinity Interactions of Nucleolin with G-G-530 paired rDNA. Journal of Biological Chemistry 274, 15908–15912. 531 Holehouse, AS, and Kragelund, BB (2023). The molecular basis for cellular function of 532 intrinsically disordered protein regions. Nature Reviews Molecular Cell Biology 2023, 1–25. 533 Kim, T, Yoo, J, Do, S, Hwang, DS, Park, Y, and Shin, Y (2023). RNA-mediated demixing transition 534 of low-density condensates. Nat Commun 14, 2425. 535 King, MR, Ruff, KM, Lin, AZ, Pant, A, Farag, M, Lalmansingh, JM, Wu, T, Fossat, MJ, Ouyang, W, 536 Lew, MD, et al. (2024). Macromolecular condensation organizes nucleolar sub-phases to set up 537 a pH gradient. Cell 187, 1889-1906.e24. 538 Lafontaine, DLJ, Riback, JA, Bascetin, R, and Brangwynne, CP (2020). The nucleolus as a 539 multiphase liquid condensate. Nat Rev Mol Cell Biol. 540 Lavering, ED, Gandhamaneni, M, and Weeks, DL (2023). Intrinsically disordered regions are not 541 sufficient to direct the compartmental localization of nucleolar proteins in the nucleus. PLoS 542 Biol 21. 543 Lin, Y, Currie, SL, and Rosen, MK (2017). Intrinsically disordered sequences enable modulation 544 of protein phase separation through distributed tyrosine motifs. J Biol Chem 46, 19110–19120. 545 Manage, KI, Rogers, AK, Wallis, DC, Uebel, CJ, Anderson, DC, Nguyen, DAH, Arca, K, Brown, KC, 546 Rodrigues, RJC, de Albuquerque, BFM, et al. (2020). A tudor domain protein, SIMR-1, promotes 547 sirna production at pirna-targeted mrnas in C. Elegans. Elife 9. 548 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 22 Marnik, EA, Fuqua, JH, Sharp, CS, Rochester, JD, Xu, EL, Holbrook, SE, and Updike, DL (2019). 549 Germline maintenance through the multifaceted activities of GLH/Vasa in caenorhabditis 550 elegans P Granules. Genetics. 551 Marnik, EA, and Updike, DL (2019). Membraneless organelles: P granules in Caenorhabditis 552 elegans. Traffic. 553 Martin, EW, and Holehouse, AS (2020). Intrinsically disordered protein regions and phase 554 separation: sequence determinants of assembly or lack thereof. Emerg Top Life Sci 4, 307–329. 555 Mensah, MA, Niskanen, H, Magalhaes, AP, Basu, S, Kircher, M, Sczakiel, HL, Reiter, AMV, Elsner, 556 J, Meinecke, P, Biskup, S, et al. (2023). Aberrant phase separation and nucleolar dysfunction in 557 rare genetic diseases. Nature 614, 564–571. 558 Phillips, CM, Montgomery, TA, Breen, PC, and Ruvkun, G (2012). MUT-16 promotes formation 559 of perinuclear Mutator foci required for RNA silencing in the C. elegans germline. Genes Dev 26, 560 1433. 561 Phillips, CM, and Updike, DL (2022). Germ granules and gene regulation in the Caenorhabditis 562 elegans germline. Genetics 220. 563 Protter, DSW, Rao, BS, Treeck, B Van, Lin, Y, Mizoue, L, Rosen, MK, and Parker, R Intrinsically 564 Disordered Regions Can Contribute Promiscuous Interactions to RNP Granule Assembly RNP 565 Granule Assembly Interactions Promiscuous Disordered Interactions Competitor Proteins 566 Synergy. 567 Rickards, B, Flint, SJ, Cole, MD, and LeRoy, G (2007). Nucleolin Is Required for RNA Polymerase I 568 Transcription In Vivo. Mol Cell Biol 27, 937. 569 Roger, B, Moisand, A, Amalric, F, and Bouvet, P (2003). Nucleolin provides a link between RNA 570 polymerase I transcription and pre-ribosome assembly. Chromosoma 111, 399–407. 571 Rostam, N, Ghosh, S, Fung, C, Chow, W, Hadarovich, A, Landerer, C, Ghosh, R, Moon, H, 572 Hersemann, L, Mitrea, DM, et al. (2023). CD-CODE: crowdsourcing condensate database and 573 encyclopedia. Nature Methods 2023 20:5 20, 673–676. 574 Savojardo, C, Martelli, PL, and Casadio, R (2020). Protein–Protein Interaction Methods and 575 Protein Phase Separation. Https://DoiOrg/101146/Annurev-Biodatasci-011720-104428 3, 89–576 112. 577 Shan, L, Xu, G, Yao, R-W, Luan, P-F, Huang, Y, Zhang, P-H, Pan, Y-H, Zhang, L, Gao, X, Li, Y, et al. 578 (2023). Nucleolar URB1 ensures 3ʹ ETS rRNA removal to prevent exosome surveillance. 526 | 579 Nature | 615. 580 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 23 Sheikh, TI, Harripaul, R, Vasli, N, Ghadami, M, Santangelo, SL, Ayub, M, Sasanfar, R, and Vincent, 581 JB (2022). Heterozygous De Novo Truncating Mutation of Nucleolin in an ASD Individual 582 Disrupts Its Nucleolar Localization. Genes (Basel) 13. 583 Snaar, S, Wiesmeijer, K, Jochemsen, AG, Tanke, HJ, and Dirks, RW (2000). Mutational Analysis of 584 Fibrillarin and Its Mobility in Living Human Cells. J Cell Biol 151, 653. 585 Spaulding, EL, Feidler, AM, Cook, LA, and Updike, DL (2022). RG/RGG repeats in the C. elegans 586 homologs of Nucleolin and GAR1 contribute to sub-nucleolar phase separation. Nat Commun 587 13. 588 Stenström, L, Mahdessian, D, Gnann, C, Cesnik, AJ, Ouyang, W, Leonetti, MD, Uhlén, M, Cuylen-589 Haering, S, Thul, PJ, and Lundberg, E (2020). Mapping the nucleolar proteome reveals a 590 spatiotemporal organization related to intrinsic protein disorder. Mol Syst Biol 16. 591 Takahama, K, Kino, K, Arai, S, Kurokawa, R, and Oyoshi, T (2011). Identification of Ewing’s 592 sarcoma protein as a G-quadruplex DNA- and RNA-binding protein. FEBS Journal 278, 988–998. 593 Tartakoff, A, DiMario, P, Hurt, E, McStay, B, Panse, VG, and Tollervey, D (2022). The dual nature 594 of the nucleolus. Genes Dev 36, 765–769. 595 Thandapani, P, O’Connor, TR, Bailey, TL, and Richard, S (2013). Defining the RGG/RG Motif. Mol 596 Cell. 597 Tollervey, D, Lehtonen, H, Jansen, R, Kern, H, and Hurt, EC (1993). Temperature-sensitive 598 mutations demonstrate roles for yeast fibrillarin in pre-rRNA processing, pre-rRNA methylation, 599 and ribosome assembly. Cell 72, 443–457. 600 Tsang, B, Pritišanac, I, Scherer, SW, Moses, AM, and Forman-Kay, JD (2020). Phase Separation as 601 a Missing Mechanism for Interpretation of Disease Mutations. Cell 183, 1742–1756. 602 Wan, G, Fields, BD, Spracklin, G, Shukla, A, Phillips, CM, and Kennedy, S (2018). Spatiotemporal 603 regulation of liquid-like condensates in epigenetic inheritance. Nature 557, 679. 604 Wang, J, Choi, J-M, Holehouse, AS, Pappu, R V, Alberti, S, Hyman, AA, Lee, HO, Zhang, X, Jahnel, 605 M, Maharana, S, et al. (2018). A Molecular Grammar Governing the Driving Forces for Phase 606 Separation of Prion-like RNA Binding Proteins Article A Molecular Grammar Governing the 607 Driving Forces for Phase Separation of Prion-like RNA Binding Proteins. Cell 174, 688–699. 608 Xu, D, Chen, X, Kuang, Y, Hong, M, Xu, T, Wang, K, Huang, X, Fu, C, Ruan, K, Zhu, C, et al. (2023). 609 rRNA intermediates coordinate the formation of nucleolar vacuoles in C. elegans. Cell Rep 42, 610 112915. 611 Yao, W, Xu, G, Wang, Y, Xing, Y-H, Yang, L, and Correspondence, L-LC (2019). Nascent Pre-rRNA 612 Sorting via Phase Separation Drives the Assembly of Dense Fibrillar Components in the Human 613 Nucleolus. Mol Cell 76, 767–783. 614 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint 24 615 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint a The adult hermaphrodite germline contains a progression of mitotic to meiotic germ cells that mature into oocytes ready for fertilization. b Schematic of 2 C. elegans pachytene germ cells. c Live, super-resolution (AiryScan) image of a germ cell nucleolus labeled with NUCL-1::GFP and GARR-1::wrmScarlet. White arrow points to a vacuole. Image is 1 plane of a 23-slice Z stack, cropped to focus on 1 nucleolus. d Deleting RG repeats causes FC/GC mixing (left-most nucleolus). Are RG repeats sufficient for nucleolar accumulation (center nucleolus)? Are RG repeats sufficient for sub-nucleolar compartmentalization (right-most nucleolus)? e Driver system for germline expression of RG repeats and GFP control. Figure 1: Studying RG repeat function in the C. elegans germline a b C. elegans germ cell nucleiC. elegans germline GFP::NLS control strain c e NUCL-1 (GC) GARR-1 (FC) Merge Super-resolution germ cell nucleolus d 2µm GFP::RG::NLS strains (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint Figure 2: Nucleolar enrichment of RG repeatspachytene- SR GARR-1::wSc pachytene- SR cytoplasm Merge GARR-1::wSc Merge pachytene- SR GARR-1::wSc Merge pachytene- SR pachytene- SR GARR-1::wSc Merge c d e f GARR-1::wSc GFP::NLS Mergeb g a GARR-1::wSc Merge pachytene- SR NUCL-1::GFP GFP::PGL-1 RG::NLS GFP::GARR-1 RG::NLS GFP::NUCL-1 RG::NLS GFP::FIB-1 RG::NLS GARR-1::wSc NUCL-1::GFP GFP::NLS GFP::NUCL-1 RG::NLS GFP::FIB-1 RG::NLS GFP::GARR-1 RG::NLS GFP::PGL-1 RG::NLS Indiv idual nucleoli Indiv idual worms h i Indiv idual nucleoli Indiv idual nucleoli Individual worms (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint a-f Single plane, super-resolution confocal images of 3-4 pachytene nucleoli and nucleolar enrichment quantification. a n=53 nucleoli from 7 worms (same image used in Fig 1c). b-c n=50 nucleoli from 7 worms. d n=55 nucleoli from 11 worms. e-f n=50 nucleoli from 5 worms. a-f Points on the left graph are individual nucleoli with means ± SD, points on the right graph are individual worms with means ± SD. g Comparison of nucleolar enrichment across all strains. Red asterisks compare against NUCL-1::GFP and green asterisks against GFP::NLS. h-i Comparison of RG repeat nucleolar enrichment against GFP::NLS, showing individual nucleoli (h) and individual worms (i). Data in g-i are the same data presented in a-f. a-f Unpaired t test with Welch’s correction. g-i One-way ANOVA with multiple comparisons. *p<.05, **p<.01, ****p<.0001. (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint Merge Merge Merge Merge Merge Figure 3: RG repeats do not display large-scale sub-nucleolar compartmentalization pachytene- SR pachytene- SRpachytene- SR pachytene- SRpachytene- SR c d e f b g GARR-1::wSc Mergea pachytene- SR GARR-1::wSc GARR-1::wSc GARR-1::wSc GARR-1::wSc GARR-1::wSc NUCL-1::GFP GFP::NLS GFP::NUCL-1 RG::NLS GFP::FIB-1 RG::NLS GFP::GARR-1 RG::NLS GFP::PGL-1 RG::NLS h i Indiv idual nucleoli Indiv idual worms Indiv idual nucleoli Indiv idual nucleoli Indiv idual worms (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint a-f Single plane, super-resolution confocal images of 4-5 pachytene nucleoli and CV quantification. The white- dashed circle in (a) is an example of the area measured for CV. a n= 114 nucleoli from 11 worms. b n=153 nucleoli from 10 worms. c n=170 nucleoli from 10 worms. d n=159 nucleoli from 10 worms. e n=153 nucleoli from 11 worms. f n=175 nucleoli from 10 worms. a-f Points on the left graph are individual nucleoli with means ± SD, points on the right graph are individual worms with means ± SD. g Comparison of nucleolar enrichment across all strains. Red asterisks are against NUCL-1::GFP and green asterisks are against GFP::NLS. h-i Comparison of RG repeat strains against GFP::NLS, showing individual nucleoli (h) and individual worms (i). Data in g-i are the same data presented in a-f. a-f Unpaired t test with Welch’s correction. g-i One-way ANOVA with multiple comparisons. *p<.05, **p<.01, ***p<.001, ****p<.0001. (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint a In a WT nucleolus, binding of nucleolar proteins to functional partners may drive enrichment in a sub- compartment, while RG repeats allow for large-scale assembly of compartments. b When endogenous RG repeats are not present, nucleolar proteins do not compartmentalize but may still bind functional partners. c RG repeats accumulate within nucleoli but do not enrich in a sub-nucleolar compartment. Figure 4: Model of RG repeat function in the C. elegans nucleolus a b cWild-type nucleolus Lacking endogenous RG repeats Overexpression of RG repeats in wild-type nucleolus NUCL-1 RG repeat FIB-1 RG repeat GARR-1 RG repeat PGL-1 RG repeat FC/GC mixing FC GC NUCL-1 FIB-1 GARR-1 FC GC (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted December 20, 2024. ; https://doi.org/10.1101/2024.12.19.629445doi: bioRxiv preprint