key: cord-0853718-jk2iatel
authors: Roden, Christine A.; Dai, Yifan; Seim, Ian; Lee, Myungwoon; Sealfon, Rachel; McLaughlin, Grace A.; Boerneke, Mark A.; Iserman, Christiane; Wey, Samuel A.; Ekena, Joanne L.; Troyanskaya, Olga G.; Weeks, Kevin M.; You, Lingchong; Chilkoti, Ashutosh; Gladfelter, Amy S.
title: Double-stranded RNA drives SARS-CoV-2 nucleocapsid protein to undergo phase separation at specific temperatures
date: 2021-06-15
journal: bioRxiv
DOI: 10.1101/2021.06.14.448452
sha: 01ee193cdf4303946046ad5396949a0a247e0240
doc_id: 853718
cord_uid: jk2iatel

Betacoronavirus SARS-CoV-2 infections caused the global Covid-19 pandemic. The nucleocapsid protein (N-protein) is required for multiple steps in the betacoronavirus replication cycle. SARS-CoV-2-N-protein is known to undergo liquid-liquid phase separation (LLPS) with specific RNAs at particular temperatures to form condensates. We show that N-protein recognizes at least two separate and distinct RNA motifs, both of which require double-stranded RNA (dsRNA) for LLPS. These motifs are separately recognized by N-protein’s two RNA binding domains (RBDs). Addition of dsRNA accelerates and modifies N-protein LLPS in vitro and in cells and controls the temperature condensates form. The abundance of dsRNA tunes N-protein-mediated translational repression and may confer a switch from translation to genome packaging. Thus, N-protein’s two RBDs interact with separate dsRNA motifs, and these interactions impart distinct droplet properties that can support multiple viral functions. These experiments demonstrate a paradigm of how RNA structure can control the properties of biomolecular condensates.

loop to restore the structure (Restore pairing). The Restore pairing mutant resembled wildtype levels of LLPS in all sequence contexts. Thus, we concluded that unpairing of the principal stemloops generally reduces LLPS and the specific primary sequence of the stem-loops does not play a significant role in LLPS.

To assess if addition of dsRNA was sensitive to the relative stoichiometry of RNA and protein, we tested wildtype and +dsRNAa in the context of 1-1000 in a small phase diagram. +dsRNAa was chosen as it is the same exact length as wildtype but drove more LLPS at 3.6µM N-protein and 15nM RNA. We observed that relative to wildtype (Fig. S1C) , +dsRNAa ( Fig. S1D and E) consistently drove more LLPS at 3µM N-protein ( Fig. S1C-D) indicating this enhancement is reproducible in multiple regimes. However, differences were observed at 1µM N-protein with only some conditions driving more LLPS, indicating a shifted phase boundary for the mutant. (Fig.  S1C-D) . We further confirmed that N-protein recruitment to droplets was higher for +dsRNAa by measuring the absorbance of the diffuse phase at 280nm (A280) following the phase separation assay (Fig. S1E ). In all three tested RNA concentrations at 3µM N-protein mutant RNA addition resulted in significantly higher A280 signal indicative of higher levels of droplet protein recruitment (Fig. S1E ).

We next determined which RNA binding domain of N-protein mediates the dsRNA-based LLPS enhancement. N-protein has two distinct RNA-binding domains; RBD1 is structured (Kang et al., 2020) and RBD2 is a lysine-rich IDR (Zinzula et al., 2021) . The single point mutant Y109A in RBD1 blocked LLPS with 5′end RNA (1-1000) and resulted in a 2000fold reduction in affinity for RNA (Kang et al., 2020) . Y109A mutant N-protein was incubated with the panel of mutant RNAs in the context of 1-1000. Only those mutations which resulted in more dsRNA could induce LLPS (Fig. 1G) . Notably, the droplets that form with these more structured RNAs and Y109A are smaller and flocculated (different morphology) suggesting key aspects of the material properties of droplets are lost with the loss of RBD1 activity. Thus, +dsRNA can drive LLPS independent of a functional RBD1 N-protein suggesting +dsRNA requires RBD2.

We sought to test whether the LLPS-promoting mutations in the RNA sequences were specific to N-protein or generalizable to any RNA-driven phase separating system. To this end, we tested all mutations in the 1-1000 context with recombinant Whi3 protein. Whi3 has previously been shown to undergo sequence-specific RNA-dependent LLPS (Langdon et al., 2018; Zhang et al., 2015) . We observed no obvious difference between any of the mutant RNAs and the wildtype 1-1000nt sequence with condensing Whi3 protein (Fig. S1F ). This indicates that the mutations are specifically acting through alteration of N-protein/RNA LLPS and not a general, non-specific RNA:protein interaction or trans RNA:RNA interaction. Taken together, addition of dsRNA enhances LLPS of N-protein specifically, and this enhancement is independent of RBD1 and RNA primary sequence.

Next, we sought to confirm our observations regarding RNA sequence/structure-mediated N-protein LLPS in cells to see if the sequences behave similarly in the more complex and crowded cellular environment. To this end, we first needed to control for the reported translational repressive effects (Tidu et al., 2021; Yuan et al., 2020) of non-structural protein 1 (NSP1) which was encoded in our 5′end 1-1000 fragment. Thus, we designed a mutation in the start codon of NSP1 (Start Mutant) which would preserve the structure of SL5 but block NSP1 translation (Fig.  S2A) . We then confirmed that the Start Mutant yielded similar levels of LLPS as wild-type (Fig.  S2B) . It was unnecessary to also mutate the NSP1 start codon of our structure mutant of the same length (+dsRNAa) as this mutation also resulted in premature stop codons in NSP1 protein (Fig. S2C) . Thus, we cloned wildtype 1-1000, Start Mutant, and +dsRNAa into a mammalian expression vector and co-transfected these plasmids with a plasmid driving N-protein: GFP in HEK293T cells (Fig. S2D) .

To determine if dsRNA addition altered LLPS in cells, we imaged co-transfected cells. We observed that at early timepoints (24 hours) +dsRNAa resulted in a significant increase in the number of puncta (4-5 per cell) per square micron in cells compared to Wildtype or Start Mutant control (2-3 per cell) (Fig.S2G ). However, this difference was reduced at 48 hours. Further, there was no significant difference in the mean fluorescence of N: GFP between the compared cells ( Fig. S2G ) so differences could not be explained by N-protein expression levels . Collectively, these results suggest that dsRNA addition accelerates N-protein LLPS in cells.

The apparent acceleration in droplet formation time, prompted us to examine differences in timing with the in vitro system for mutants in all 3 sequence contexts (shorter incubation time (2 hours) than shown in Figure 1 D-F (18 hours)). Consistent with the structure mutants accelerating N-protein LLPS in cells, the mutants which result in more LLPS at 18 hours had vastly more pronounced differences at 2 hours indicating that these structure mutants also accelerate droplet formation cell free (Fig. S2H) . Similar results as for 1-1000 (Fig. S2H) were obtained for both the 1-500 and 500-1000 context ( Fig. S2I and J) . Collectively, these data suggest that in cells and cell free addition of dsRNA accelerates N-protein phase separation with target RNAs.

The data thus far show that a primary driver of LLPS is dsRNA/RBD2 interactions but there are several lines of evidence that suggest additional interactions are mediated by RBD1. First, Y109A mutant N-protein does not undergo LLPS with wildtype 5′end sequence. Second, the Y109A+dsRNA droplets have altered morphology suggesting some interaction between N and RNA has been lost in the absence of RBD1 activity. Thus, we sought to identify what RNA sequence features are favored by RBD1.

Given the transcriptional regulatory sequence (TRS/SL3) is the reported binding site of RBD1 in MHV (Grossoehme et al., 2009) , and the 1-500 fragment which contains the TRS was better able to drive LLPS than the 500-1000nt fragment (Fig. 1C) , we reasoned the TRS may be the preferred binding site of RBD1. Thus, we sought to characterize the importance of the TRS in N-protein LLPS (Fig. 2) .

To test if the presence of the TRS was required for LLPS, in the context of 1-1000nt, we deleted the entire TRS stem-loop (TRS-del) or added an additional TRS motifs to the 3′ end (Add TRS-3′). These mutations effectively increased or decreased the putative RBD1 binding site by one ( Fig. 2A) . TRS-del almost completely blocked LLPS ( Fig. 2A) and reduced N-protein recruitment to droplets (more N-protein in the diffuse phase) (Fig. 2B) . These results are analogous to how in the absence of RBD1 activity (Y109A mutation), 5′end RNA 1-1000 could no longer drive N-protein LLPS . Fig. 1G) . Additionally, crosslinking experiments reveal reduced binding of Y109A N-protein at regions adjacent to the TRS-Loop ( Fig.  S3A and B . Conversely, the addition of a TRS-loop (Add TRS-3′) resulted in slightly larger droplets than wildtype and enhanced N-protein recruitment to droplets (Fig. 2B) . We conclude from these studies that the presence of TRS-loop facilitates N-protein LLPS with 5′end RNA.

We next wondered if N-protein could also bind sequences which were similar to the TRSloop. This is because N-protein can drive LLPS with other genomic RNA sequences (Carlson et al., 2020; Iserman et al., 2020; Jack et al., 2020; Lu et al., 2021) We hypothesized that the most favored RBD1 binding site was YYAAAY (Y= C or U) which is similar to the TRS-loop sequence.

In accordance with this hypothesis, crosslinking of N-protein is reduced in the region adjacent to the TRS-Loop for the RBD1 Y109A mutant N-protein ( Fig. S3A and B . The TRS-loop sequence, CUAAAC, occurs 16 times across the genome (Fig. 2C) , and a chemically similar sequence YYAAAY (Y= C or U), a TRS-loop-like sequence, occurs 114 times across the genome (Fig. 2D ). Binding to this sequence is suggested to occur in MHV N-protein experiments (Grossoehme et al., 2009 ). Thus, we tested TRS-del, or mutated the individual As in the TRS-Loop sequence, CUAAAC, (A68U, A69U, A70U named so for the corresponding nucleotides in MHV). We also mutated the Cs and Us in the TRS-Loop sequence based on their occurrence in the SARS-CoV-2 genome. A68U, like TRS-del, completely blocked LLPS, and A69U and A70U resulted in a decrease relative to wildtype (Fig. 2F) . Mutation of sequences to the rare (low frequency in the genome) or common Y (high frequency) sequence had negligible effects on Nprotein LLPS (Fig. 2F) . We conclude that while the A identity is critical for N-protein LLPS, the identity of the flanking Y sequence (C or U) is not. These results suggest that RBD1 may interact with TRS-loop-like sequences (YYAAAY) across the genome. Thus, individual mutations of specific, conserved nucleotides in the TRS-Loop sequence also resulted in reduced LLPS.

We next asked if the reduction of LLPS which occurred following primary sequence mutation was due to RNA structure-dependent alterations. This is because when we were designing A mutations, we noticed that the A68U/A69U mutations were predicted to favor a different structure than wildtype. (Fig. 2G) . This predicted alternative structural arrangement pairs the TRS stem-loop (SL3, green) with SL5 (magenta). Thus, the A to U mutants' impact on LLPS may occur via a sequence change, structure change or both. We reasoned we could make an allele that disentangles these confounding features by mutating nucleotides in SL5. This mutant would restore the wildtype structure while preserving the single nucleotide changes in the TRSloop (Rescue mutants, Fig. 2G ). Rescue wildtype RNA resulted in a mild enhancement of Nprotein LLPS in comparison to the wildtype. In contrast, Rescue A68U almost completely restored the LLPS and Rescue A69U completely restored the LLPS. We conclude that the identity of A68 and A70 are important for N-protein RBD-1 binding. In contrast to MHV N-protein preferences, A68 and A69U further suppress LLPS by causing a structural rearrangement. Thus, the sequence and structure of the TRS stem-loop is required for N-protein binding and LLPS.

N-protein can phase separate in the absence of RNA at high temperatures but, RNA addition lowers LLPS temperature. Thus N-protein displays lower critical solution temperature (LCST) behavior with respect to RNA . Notably, the addition of RNA, lowers the LCST to physiological temperatures. It is unclear if and how RNA sequence and structure dictates the temperature at which LLPS occurs. Given our observations from Figures  1 and 2, where we observed N-protein LLPS is controlled by RNA sequence/structure we wondered if LCST temperature was also controlled by RNA sequence/structure. To our knowledge, nothing is known about the role of RNA sequence in influencing LCST behavior in LLPS systems.

To test if LCST temperature is encoded by RNA sequence we first tested if N-protein condensation temperature changed as a result of the co-condensing RNA. Thus, we tested 3 different RNA sequences. To this end, we used a temperature-dependent ultraviolet-visible spectroscopy assay to map the saturation temperature, read out as turbidity with turbidity being a proxy for LLPS. We examined the following conditions: N-protein alone, N-protein + an RNA which does not drive LLPS (Frameshifting region RNA (FS) , N-protein + 5′end RNA (1-1000nt) which drove LLPS, or N-protein + Nucleocapsid RNA (drives LLPS but is a longer sequence then 5′end). We used these different sequences to examine if all RNA sequences influence N-protein LCST to the same degree.

We observed that N-protein+FS (which does not drive LLPS) and N-protein alone underwent phase separation at the same high temperature (Fig. 3A) of ~46°C. In contrast, the two LLPS-promoting RNAs both lowered temperature with Nucleocapsid RNA conferring a lower temperature then 5′end. The turbidity curves differ in shape depending on the specific RNAs such that LLPS-promoting RNAs display a more gradual turbidity increase. This may be due to heterotypic RNA-protein interactions. Thus, while different RNAs promoted different LCST behavior, this could be due to sequence and/or length-dependent effects.

We wanted to identify the RNA and protein features which were responsible for conferring N-protein LCST independent of RNA length (Nucleocapsid RNA is longer then 5′end). To disentangle the effects of RNA length and RNA sequence we tested the 1-1000nt mutant RNAs (similar or identical lengths only very slightly different sequences and structure) with Nprotein (Fig. 3B) . Notably, only those mutations which resulted in more secondary structure (+dsRNAa-d) significantly lowered the LCST (Fig. 3B and C) . Loss of the putative RBD1 binding site (TRS) had no significant impact on temperature. Additionally, raising the temperature resulted in mildly enhanced N-protein binding to 5′UTR RNA in RBD1 deficient (Y109A mutant) protein via EMSA ( Fig. S4A and B) suggesting N-protein's RNA binding activity at higher temperatures is independent of RBD1.

To confirm that RBD1 activity did not alter LLPS temperature, we additionally tested the Add TRS 3′ mutation which creates an additional RBD1 binding site. Add TRS 3′ did not lower the LCST ( Fig. S4C and D) via either a microscopy assay or diffuse phase measurement. These results suggested temperature sensitivity was conferred by RBD2/dsRNA interactions rather than generally enhanced binding to RNA (through RBD1 for example).

We confirmed the temperature-dependent turbidity results reflected the formation of droplets by examining assemblies under the microscope. +dsRNAa (one of the add more structure mutants with the same length as 1-1000 wildtype Fig. 3B ) lowered N-protein LLPS temperature to 25°C (Fig. 3D) . We confirmed that the diffuse phase measurement was perfectly anti-correlated with the imaging of the droplets at all temperatures (Fig. 3E) . This further suggests that dsRNA-dependent interactions can tune the LCST.

Given that the more structured mutants do not require RBD1 activity to undergo LLPS ( Fig. 1G ) but all other tested RNAs do, we reasoned that the LCST behavior of N-protein could be specified by RBD2. To assess this possibility, we purified N-protein with a deletion in RBD2 (red amino acids) (Fig. 3F ) which was predicted to preserve the conserved dimerization interface, (Fig. S4E) . We observed that N-protein RBD-2-Del's LCST behavior was significantly altered with both wildtype and mutant RNA (Fig. 3G ) and could undergo LLPS at all tested temperatures even without additional RNA. Further, similar levels of protein in the diffuse phase (A280 signal) were detected following the LLPS assay (Fig. S4F) . Reducing the N-protein and RNA concentration showed some degree of RNA dependence for the RBD-2-Del mutant ( Fig. S4G and H) . These data support that RBD2 interactions encode the temperature threshold for LLPS in this system.

We were surprised, however, to see that RBD2 deletion leads to increased LLPS. Based on the literature of SARS-CoV-1 N-protein RBD2 crystal structure (Chen et al., 2007) , we hypothesized that RBD2-del region may stabilize the formation of higher order oligomers of Nprotein. To address if RBD2-del mutation was destabilizing the formation of N-protein dimers (the reported oligomerization state of N-protein in the absence of nucleic acid (Zeng et al., 2020; Zhao et al., 2021a) we performed mass photometry. We observed that, consistent with previous studies, wildtype N-protein forms a dimer (Fig. 3H) whereas RBD-2 del is mostly a monomer (Fig.  3I) . We conclude from this that the RBD2-del mutation destabilizes the N-protein dimer which may lead to reduced temperature and less RNA dependence of LLPS. dsRNA addition can similarly perform the dimer destabilization (splitting) but only for wild-type N-protein (Fig. 3J) . Collectively, these data suggest that while RBD1 is required for 5′end to phase separate, RBD2 is required for LCST behavior at physiologically-relevant temperature and salt, and RBD2 encodes LCST behavior through preferentially binding dsRNA.

Given the +dsRNA structure mutant promoted puncta formation in cells (Fig. S2 ) and the SARS-CoV-2 genome is enriched in dsRNA, even in protein coding sequences (Huston et al., 2021; Lan et al., 2021; Sun et al., 2021) , we next asked if N-protein binding and LLPS could regulate target RNA translation. We reasoned the increase in LLPS due to dsRNA addition may be antagonistic to translation as some condensates can repress translation (Kim et al., 2019; Tsang et al., 2019) . Thus, an understanding of how N-protein regulates translation would be informative for the viral life cycle.

To address N-protein mediated protein translational regulation encoded by dsRNA, we sought to replicate our dsRNA/ssRNA addition experiments (Figure 1 ) in the context of the 5′UTR (Fig. 4A ) by altering stem-loop 4 (SL4). We observed that only the +dsRNAb (which results in 22nt additional sequence and 44nt of additional dsRNA) drove significant additional LLPS relative to wildtype (Fig. 4A) . +dsRNAd which adds 10 paired nucleotides to the base of SL4 (20 additional nucleotides total) also resulted in minor enhancement. All other non +dsRNA mutants had negligible effects. Thus, length dependent addition of dsRNA to the 5′UTR should be sufficient to enhance LLPS independent of the coding sequence when appended in cis.

To ask if 5′UTR or +dsRNA UTR could differentially regulate translation in droplets we fused either the wildtype 5′UTR or a more structured mutant (+dsRNAb) to nano luciferase (Fig.  4B) . To determine if 5′UTR structure affects LLPS for the fusions, we mixed 3.2µM N-protein with 25, 15, or 5 nM RNA. At the highest tested RNA concentration, 25nM, only the more structured mutant UTR resulted in LLPS (Fig. 4C) . Similarly, only 25nM RNA condition had a statistically significant difference in A280 absorbance in the diffuse phase (Fig. 4D) . Thus, consistent with results above, addition of dsRNA facilitates LLPS of nano luciferase fusion RNA. We next asked how LLPS conditions (3.2µM N-protein 25nM RNA) impact translation? To this end, we performed an in vitro translation assay +/-3.2µM N-protein. We observed that addition of 3.2µM N-protein almost completely blocked the translation of RNAs (Fig. 4E ) Collectively, these results suggest that N-protein droplet conditions block translation.

We next asked if translation inhibition depended on LLPS or N-protein binding in the diffuse phase? To this end, we repeated our in vitro translation assay this time with 0.3µM of Nprotein ( Fig. 4F over 10-fold less N-protein than in Fig. 4E ). In these conditions, translation of the wild-type UTR was moderately but significantly repressed by 0.3µM N-protein addition, and the translation of the +dsRNA UTR mutant was almost completely repressed translation (9.1X further reduction in translation compared to wildtype). This is consistent with phase behavior and Nprotein affinity differences for these two RNAs ( Fig. 4C and D) . Collectively, these data suggest that droplet promoting conditions completely block translation, diffuse state-promoting conditions partially block translation. Translational block is dependent on relative N-protein affinity for the translating RNA (Fig. 4G) .

We hypothesize that N-protein binding to the genome (particularly the 5′UTR) may act to halt protein translation and promote packaging in later stages on infection (Fig. 4H) . In support of this idea, N-protein RNA is low at early stages of infection and gradually increases (via generation of sub-genomic N-protein RNA at late stages of infection (Kim et al., 2020) . Thus, as time increases translatable genome should go down, thereby promoting a switch from translation to packaging in late-stage infection.

Given the key central role of N-protein in genome packaging, we next asked how what we have learned thus far about different types of N-protein/RNA interactions may impact packaging. Specifically, we postulated that the 5′end of the genome should have different patterning and/or affinity for N-protein binding sites than the genome center. We reason that this is because the non-structural genes in orf1ab need to be efficiently translated and LLPS of N-protein clearly can repress translation (Fig. 4H) . Further, sequence property differences between the ends and the center of the genome are supported by sequence analysis . Given the surfeit of in cell and cell free genome structural data for SARS-CoV-2 and related coronaviruses that is now available (Huston et al., 2021; Lan et al., 2021; Sun et al., 2021) , as well as the published genome sequence data (NC_045512.2) we next asked how the two "dsRNA stickers" we identified as important for N-protein LLPS (RBD1 binding structured YYAAAY and RBD2 binding dsRNA) (Fig. 5A) are distributed throughout the genome?

To determine the patterning of our two identified N-protein "stickers", we obtained published in cell SARS-CoV-2 DMS-MaP based RNA structure models to obtain locations of dsRNA. We also collected the density of YYAAAY from NC_045512.2 reference SARS-CoV-2 genome. Given our observation that structure of YYAAAY is important for binding (Figure 2 ) we restricted the YYAAAY motif to only those which overlap with structured RNA (80 structured YYAAAY motifs) (Fig. 5B) . We observed that both features were relatively uniformly distributed across the genome with enrichment of structured YYAAAY motifs (above black dashed line) at the 5′ and 3′ ends of the genome, perhaps to promote genome circularization (Seim et al., 2021) .

While the enrichment of LLPS-driving sequences at the 5′ and 3′ ends could suggest that LLPS may promote genome circularization Seim et al., 2021; Ziv et al., 2020) the final packaged genome is clearly not a single droplet. Based on high-resolution cryo-EM tomography, the genome of SARS CoV-2 is arranged inside virions in a so-called "birdsnest" arrangement with "eggs" made of RNP complexes that are ~14-20nm (Klein et al., 2020; Yao et al., 2020) . We previously observed that RNA derived from the center of the SARS-CoV-2 genome including RNA encoding the Frameshifting-region (FS) promoted N-protein solubilization at the microscopic level . We reasoned that the solubilizing effect of FS RNA may be conferred by the formation of diffraction limited clusters that may be distinct from LLPS or are arrested from coarsening into macroscopic droplets. If indeed small RNP-scale particles form in this cell free system this would indicate that N-protein binding to RNA, as dictated by RNA sequence, was sufficient to condense RNA independent of cellular machinery.

To address if N-protein mediated condensation is sufficient to compact RNA to RNPsize assemblies' cell free, we first asked what size particles form from FS RNA (1000nt in length)? We examined FS RNA as this RNA does not drive LLPS at 4µM N-protein 10-15nM RNA (Fig. 3A) . To this end, we measured the particles formed from 10nM FS RNA and 4µM Nprotein by dynamic light scattering. We chose 400X protein to RNA as this would be reminiscent of late-stage infection and packaging (Kim et al., 2020) . We observed that following 20-minute incubation time at room temperature FS RNA forms homogenously sized clusters 43.7-57.9 nM in diameter (Fig. 5C) suggesting RNA cluster generation can occur cell free.

To directly visualize cluster formation a second way, we used TEM. Indeed, after 20-min of incubation a relatively monodispersed population of symmetric, circular assemblies form that are centered on 42.9 nm diameter (Fig. 5D) . To assess if these formations were specific to FS RNA, we also examined N-protein 1-1000 RNA in conditions that do not support phase separation (room temperature). These formed similarly shaped and sized particles as the FS RNA (Fig. 5D) . The assemblies formed with both RNAs are more than double the size of the reported RNP (~14-20nM) diameter (Klein et al., 2020; Yao et al., 2020) .

We wondered what caused the >2-fold size discrepancy between these RNP assemblies and the RNPs seen in virions? It is established that some droplets age into gel-like or glass-like states that can be associated with compaction (Jawerth et al., 2020), we therefore asked how the clusters change with time. Indeed, at the 20-hour time point smaller, more similar sized clusters for both RNAs were formed (Fig. 5D) indicating the clusters are shrinking by ~15% over time, independent of RNA sequence (Fig. 5E ). Some larger, rarer clusters were detected at 20hours for both RNAs ( Fig. S5A and B) . Thus, N-protein and 1kB gRNA form monodispersed clusters cell free that compact over time. Both RNA and protein are required to form clusters (Fig. S5C) . The similar size distribution of 5′end and FS fragments may result from the similar length (1kb) and overall affinity for RBD1 (the temperature insensitive RNA binding domain). Therefore, condensation differences between 5′end and FS require temperature-sensitive RBD2 interactions.

We postulated that FS interactions with N-protein may be heavily dependent on RBD1 rather than RBD2. In support of this hypothesis, the FS RNA contains 7 YYAAAY motifs 5 of which are structured and FS does not engage with RBD2 in a way that alters LLPS temperature (Fig. 3A) . To confirm FS N-protein interactions are strongly RBD1 dependent, we performed RNP-map on FS with wildtype and Y109A mutant (RBD1 deficient) N-protein ( Fig. S5D and E) . We observed that the majority of the N-protein crosslinking peaks in FS were absent following incubation with Y109A mutation. Some Y109A-independent crosslinking was detected (purple boxes) and this tended to be adjacent to structured RNA. Thus, FS/N-protein interactions are primarily driven by RBD1 ( Fig. S5D and E) whereas 5′end N-protein interactions are driven by both RBD1 and 2 ( Fig. S3A and B) . RBD1 binding site patterning conferred by structured YYAAAY motifs may be required for RNP-sized cluster generation.

In this paper, we elucidate the RNA sequence and structure preferences of SARS-CoV-2 N-protein to understand how these features lead to condensate properties relevant to viral processes in cells. We show that 1) RBD1 prefers TRS-like sequences in an RNA structure dependent manner. 2) RBD2 prefers dsRNA in a sequence-independent manner. 3) RBD2 dsRNA interactions encode N-protein LCST behavior (Fig. 6A) . 4) RNA sequence/structure features specify N-protein interactions to regulate puncta formation rate, translation, and RNA cluster size cell free (RNP size) (Fig. 6B) . To our knowledge, this is the first example of a role of dsRNA in encoding "stickers" and specific droplet properties.

Our data (Fig. 6B) suggests a mechanism by which N-protein can perform multiple distinct functions over time in the same cytoplasm depending on N-protein concentration. Following viral entry, N-protein concentration is low. The low protein concentration allows for N-protein to dissociate from the condensed genome and for the initiation of translation. As infection progresses, N-protein's (and other structural proteins) accumulation is driven by production from sub-genomic transcripts (Finkel et al., 2021; Kim et al., 2020) . The accumulation of N-protein initiates a switch from translation to packaging, shutting down non-structural protein production while sparing the sub-genomic RNAs (which lack the most structured stem-loops 5 and on) (Kim et al., 2020; Sun et al., 2021) . The structure of our RBD1 motif, the TRS, on the sub-genomic RNA is also predicted to regulate translation in a structure dependent manner with unstructured TRS present on the highly translated N-protein sub-genomic RNA (Finkel et al., 2021; Sun et al., 2021) .The enrichment of high affinity N-protein binding sites at genome ends may allow for LLPSmediated circularization to promote single genome packaging Seim et al., 2021) . Finally, within double membrane vesicles, RNPs form as additional N-protein accumulates over time, with high concentration driving N-protein recruitment to low affinity sites in the genome center. The condensed genome ultimately matures into virions. We have identified unique dsRNA encoded "stickers" for N-protein conferred by the two RNA binding domains. The patterning and quality of the two N-protein dsRNA stickers can confer N-protein's multiple functions through concentration dependent binding and LLPS. Thus, biochemical complexity needed for viral replication can be achieved with minimal components.

Addition of dsRNA, independent of sequence ( Fig. 1D-F, Fig. S1A ), resulted in more LLPS at all tested temperatures (Fig. 3B) and conditions (Fig. S1C) . Reduction or addition of short ssRNA (comparable lengths to dsRNA mutants) sequences resulted in negligible enhancement of LLPS ( Fig. 1D-F, Fig. S1B ). Unpairing dsRNA generally reduced LLPS ( Fig. 1D-F) . There is likely an absolute length preference for RBD2 binding (Fig. S1A ) which is consistent with 5′UTR stem-loop length altering experiments leading to viral plaque reduction (Raman and Brian, 2005; Yang et al., 2011) . The lack of observed primary sequence specificity to N-protein RBD2 dsRNA binding (Fig1. D-F) may explain why previous stem-loop swap experiments, switching stem-loops from one betacoronavirus for another, generated functional virus (Guan et al., 2011; Kang et al., 2006a Kang et al., , 2006b . Although dsRNA length is important for RBD2 engagement, the specific sequence of the stem-loops is not critical (Fig. 1D-F) suggesting that N-protein may be able to engage with the entirety of the highly structured genome of SARS-CoV-2 (Huston et al., 2021; Lan et al., 2021; Sun et al., 2021) . The lack of dsRNA sequence specificity is also suggested by the nature of the RBD2 motif, a disordered lysine rich IDR (Fig. 3F, Fig. S4C ), which is unlikely to have primary RNA sequence specificity. Therefore, these data suggest that the sequence of the stem-loop does not matter for viral production and only minor differences in length are tolerated.

Importantly, our results suggest excessive differences in the length of the stem-loops appears to alter temperature encoding behavior with ~20-24nt of dsRNA (present in SL5, SL12, and SL13 (Fig. 1A ) encoding LLPS at 37°C and additional dsRNA (10nt+ -80nt+) lowering the temperature to as low as 25°C ( Fig. 3B and C) . The absolute length of the stem-loop must be under a degree of selective pressure. Importantly, the most structured stem-loops (5, 12 and 13) are highly conserved (Fig. 1B (Yang and Leibowitz, 2015) ) suggesting that stem-loop length mediated regulation is a universal feature for proper viral production, but subtle differences exist in the stem-loop length between individual viruses. This matches with experiments in MHV virus where altering the pairing of stem-loop 1 reduced the efficiency of viral production (Li et al., 2008) . This suggests that stem-loop length may be co-evolving with N-protein RBD2/dimerization sequence, protein amount or both.

RBD1 preferentially crosslinked adjacent to TRS and TRS-loop-like sequences (YYAAAY motifs) in the first 1000nt of the SARS-CoV-2 genome . Improperly structured TRS loop sequences resulting from point mutation or deletion of the TRS stem-loop completely abrogated N-protein LLPS with 5′end 1-1000nt RNA. This is consistent with RBD1 mutant N-protein no longer undergoing LLPS with 5′end RNA and previous findings in MHV suggesting TRS is specifically recognized by N-protein RBD1 not RBD2 (Grossoehme et al., 2009) . Unique to our findings of the TRS motif is the discovery that the secondary structure, in addition to the primary sequence, is important for SARS-CoV-2 N-protein LLPS and presumably binding. Our results suggest weak RBD1-mediated binding can occur at regions that are not the TRS stem-loop but have similar sequence and are structured (i.e., structured YYAAAY motifs).

We postulate that the native, more structured stem-loops of the genome ends (i.e. SL5, 13 and 14 in the 5′end) are the most efficient binding sites for RBD2 (as evidenced by RBD1 independent crosslinking adjacent to these stem-loops ( Fig. S3A and B ) and this binding promotes LLPS at human body temperature (37°C). We predict that the binding to RBD2 in combination with physiological temperature (37°C) allows for the dissolution of the dimerization domain adjacent to RBD2 (Fig. 4J) . Temperature is likely to facilitate the "unfolding" of the dimerization domain (the location of the fluorescent amino acids) as purified RBD2dimerization domain undergoes a structural change at ~50°C by differential scanning fluorometry (Zinzula et al., 2021) . Of note, the temperature of dimerization unfolding determined by Zinzula et al for purified RBD2 dimerization domain alone is very close to our observed full-length N-protein only turbidity temperature (46°C Fig. 3A ) suggesting temperature-dependent unfolding of this domain is critical to LCST behavior. The exposure of the hydrophobic core of the dimerization domain to the solution following temperature and RNA engagement may facilitate LLPS as hydrophobic regions tend to be insoluble.

The temperature at which dimerization occurs can be lowered via addition of dsRNA, potentially due to increasing the overall affinity of wildtype N-protein's two RBD2 domains or by offering additional sites of interaction (at a greater distance apart) on the same stem-loop. In support of the latter possibility, the two RNA binding domains are arranged diagonal to each other, and dsRNA binding may force the dimer apart (Fig. 3J) . Cryo-EM data of purified RBD-2/dimerization domain with ssRNA seems to agree with this hypothesis (Zinzula et al., 2021) with 7 base pairs of ssRNA spanning between two RBD2 motifs and a marked separation in the dimer region.

Our results also may explain why not all labs reporting N-protein LLPS have observed LCST behavior in N-protein RNA interactions as these results show that LCST behavior is specifically encoded in N-protein dsRNA interaction. N-protein dsRNA interaction is unlikely to be observed in reconstitution experiments conducted with less physiological, unstructured RNA (poly U for example). RBD2 also seems to regulate the dimerization domain of N-protein (Zhao et al., 2021a) . RBD2 dimerization is highly dependent on the salt concentration with only physiological salt concentrations (150 +/-~30) allowing for LCST behavior Lu et al., 2021; Zhao et al., 2021a) . Lower salt results in an increase in N-protein dimerization domain interactions (Jack et al., 2020) which might increase the required total solution concentration of N-protein for phase separation, thus also increasing the temperature boundary of the LCST behavior. Others have not observed LCST behavior using nearly identical RNA sequences and N-protein preparation methods, but they were using much lower salt (80) (Lu et al., 2021) . We conclude that because physiological levels of salt are more likely to be present in cells, LCST behavior of N-protein is relevant.

As RBD2 may recognize RNA through a complex interaction involving charge, disorder, and transient protein structure. It is highly likely that post-translational modifications (PTMs) play a huge role not only in RBD-2 binding RNA but also dimerization and LCST behavior. This may begin to explain why those labs which purify N-protein from mammalian cells did not observe LCST while those which purify N-protein from bacterial sources did Jack et al., 2020; Zhao et al., 2021a) . As packaged N-protein is specifically free of post-translational modifications (Fung and Liu, 2018; Wu et al., 2009) , we hold that the LCST behavior is likely still relevant for packaging with other N-protein compartments such as those regulating viral RNA transcription and translation being far more likely candidates to be regulated by PTMs (particularly those droplets that form outside double membrane vesicles associated with packaging). Future directions will explore how PTMs tune LLPS temperature to potentially sustain viral replication and viral RNA N-protein interactions during late-stage infection/fever temperatures.

Additionally, our results suggest the primary sequence of RBD2 dimerization region is critical for RNA binding and LCST behavior we would postulate that any mutation that arises and is selected for in these regions (in patient isolates or across species) would be particularly informative. Therefore, future directions will carefully characterize the sequence variation in this region.

Distinct N-protein "dsRNA stickers" are distributed throughout the genome (Fig. 5B) . This prompted us to hypothesize dsRNA sticker patterning could be relevant for packaging. Although N-protein clearly has tendencies to form macroscopic condensates in vitro and in cells, the packaged genome is instead packed into regularly-spaced RNPs which may be arrested in coarsening. Reconstituted N-protein mixed with 1000nt RNA fragments in physiological salt and pH was able to form clusters that were roughly 1.75X-2.5X the diameter of the RNP, the unit of packaging of the virion. This size difference suggests that either 1) the RNA content of the RNP is ~500-1000nt (to give a 14-20nm RNP diameter) or 2) further, compaction occurs in the cells. We suggest the former possibility is more likely as there is a number range of RNPs (30-35 by cryo-EM suggesting each RNP must contain less then 1000nt (~30kb genome/ ~35 RNPs) and there is likely a flexible linker region composed of RNA depleted in N-protein (less electron dense) between each RNP to facilitate compaction. These data suggest that the information needed to condense the RNA genome is contained within the genomic RNA sequence.

Future directions will involve modeling of the SARS-CoV-2 dsRNA and structured TRSloop-like sequences patterning across the genome to examine if indeed sequence element patterning is sufficient for RNP patterning (Choi et al., 2019; Seim et al., 2021) . Of note, the length of viral RNA fragments tested in this work (0.5 and 1kB) is highly relevant for this consideration as each RNP/egg is likely to contain <1kB. We and others have observed that longer RNAs (Jack et al., 2020) including RNA purified from infected cells containing SARS-CoV-2 genome results in a "string of pearls" type droplets rather than rounded droplets further suggesting that the formation of RNPs/eggs is recapitulated cell free. Thus, the fragments tested here are short enough to encode single RNP/egg like features but long enough to have sequence and structural complexity to allow for observable regional differences in LLPS.

Notably, increasing RNA order, rather than disorder, through additional RNA structure drives N-protein LLPS. These results contrast with those observed by Mayr lab where increased RNA disordered single stranded regions promoted intramolecular association and LLPS (Ma et al., 2021) . This discrepancy is likely due to differences in the proteins, specifically the preference of both of N-protein's RNA binding domains for the highly-structured RNA genome of SARS-CoV-2 (dsRNA stickers).

Our work suggests that reconstitution experiments of phase separating proteins with similar dsRNA preferences must be carried out with physiological RNA targets to capture biological behavior. DsRNA protein interactions are not captured with poly A or U. In short, RNA sequence and structure profoundly influence LLPS behavior. Finally, this work shows the complexity of the RNA-protein code in determining the kinetics, emergent properties and temperatures at which biomolecular condensates can form. We predict that this is the tip of the iceberg in terms of unraveling the information provided by RNA sequence to specify the form and function of condensates.

RNP-MaP experiments. We thank Benjamin Stormo for providing purified Whi3 protein. A.S.G., C.I. and C. A. R. were supported by NIH R01GM081506 and an HHMI faculty Scholar Award, C.A.R. was supported by NIH T32 CA 9156-43, F32GM136164 and L'OREAL USA for Women in Science Fellowship. This work was supported by a FastGrant. I. S. and A. S. G. were supported by Air Competing interests: K.M.W. is an advisor to and holds equity in Ribometrix, to which mutational profiling (MaP) technologies have been licensed. A. S. G. serves on the scientific advisory board to Dewpoint Therapeutics. CI is by now employed at Dewpoint Therapeutics. All other authors declare that they have no competing interests.

Limitations: This study addresses mechanisms of phase separation of components of the SARS-CoV-2 virus. However, because the work involved reconstitution experiments from purified components and expression of viral proteins in mammalian cells rather than in an actual infection it is still unclear what step(s) in the viral replication cycle may utilize the mechanisms described. The work is also is missing some factors such as M-protein and lipid membranes that may change the physical chemistry of the phase separation in an infection. Unpairing principal site adjacent stem-loops (grey -dsRNA) on the 5′ side reduces LLPS. Restoration of wildtype RNA structure (blue Restore pairing) but with a different sequence restores LLPS to wildtype levels. (G) Only those mutations who lead to an addition of dsRNA (+dsRNAa-d), retain the ability to induce phase separation following Y109A mutation and destruction of N-protein RBD1. For all images, scale bar indicates 10µm all experiments show representative images from at least 3 replicates and 2 independent batches of RNA. Transition temperature comparison (repeat of the experiment shown in (A) of wildtype 5′end or 11 mutants in the context of 1-1000nt. Bar length indicates the temperature in °Celsius at which the turbidity of the solution reaches ~0.1. Only those mutants which alter the dsRNA content (teal +dsRNA), lower the temperature at which OD reaches ~0.1 indicative of increased solution turbidity. (C) Temperature dependent turbidity tests for N-protein plus wildtype 5′end RNA as well as the 4 more structured mutants (+dsRNA) which lower the transition temperature. (D) Validation of the turbidity assay using droplet imaging ( Fig. 3B and C) . 3.4µM Wildtype N-protein was mixed with either 25nM of wildtype 5′end 1-1000 RNA, +dsRNAa (RBD1 independent Fig. 1G ) or water only added control (H20) and incubated at the indicated temperature 37, 30, or 25°C for a period of 20 hours prior to imaging. Consistent with previous results, +dsRNAa increases droplet size relative to wildtype at 37°C (Fig. 1F ) & induces LLPS at lower temperatures. (E) A280 measurement of remaining N-protein in the diffuse phase for Fig. 3D . At all temperatures +dsRNAa lowers A280 measurements relative to wildtype. Error bars mark standard deviation for the three replicates and * indicate significance students T test (*** p<0.001, ** p<0.01, *p<0.05, ns not significant) with brackets showing comparison for the indicated statistical test. (F) Protein sequence conservation of N-protein RBD2 and structure model of the RBD2 dimerization domain for SARS-CoV-2 (red sequences/ red ribbon) indicate the location of the deletion in the primary sequence tested in (G). (G) RBD2/Dimerization domain is required for proper N-protein LCST behavior at indicated temperature range. 3.4µM of N-protein RBD-del (green) was mixed with 25nM of either wildtype 1-1000, +dsRNAa, or water only control and incubated at the indicated temperatures for 16 hours. Droplet formation was observed in all conditions although RNA dependence was more evident at lower protein concentrations (Fig. S4 G-H 

A U A U U A A A GGUUUUU CCU C ||.|||| ||| CCGAAAA GGA C A C A C C A A C C U C C GAUCU U ||||| U CUAGA G U U UC GUUC U |||| A CAAG A U CA U U A A A A U C U C U CUGUGUAG UG CGCU GGC G |||.|||| || |.|| ||| C UC GACGCAUC AC GUGA CCG A U G U C U U U U A G-C U G-C A C-G A A-U A U.G C U A U U A C-G U G-C A U-A A C-G U A-U A G-C A G A A U U U U A C U C A A G U C U C U C A G A UAA G C G A UU UCGUUGACA AGA ACGAG CUC UCC UCUU CAUACCU GG U ||||||||| ||| ||.|| ||| ||. |||| ||.|||. || C AGCAACUGU UCU UGUUC GAG AGG AGAA GUGUGGG CC G GG C U U-A U G-C G-C A G A A A " " # # # $ $ " " " # " # ! ! " " ! ! !# $ $ " # # ! # # # ! ! # # ! ! # # ! " " " ! $ # " ! " ! " " $ " # $ # " ! " $ " " ! " ! " # # # ! $ # # ! " " " # # # # " ! " $ " $ " $ $ ! " $ " ! # ! " ! $ $ ! " $ ! # " $ ! " " # $ " $ ! #! " ! # ! $ ! # $ " # " # # " " # # " # # ! " # # " " # ! " $ " ! $ " " $ # ! # $ $ # ! # ! $ # $ " # # ! " ! $ " ! " # " ! " " ! " $ ! # $ $ ! " $ ! " " # ! $ $ " " " ! $ " ! ! $ " $" " $ ! # $ ! !$ # " ! # " ! # $ ! # ! # " ! " # $ $" " " !$ " ! ! $ $ $ " $ " $ # ! ! $ # # # $ $ " # # $ # " $ $ # $ # $ !! " " $ " ! ! ! " $ $ "" " ! # # ! $ # $ # # # # ! # ! # ! $ " ! ! # # ! " !# $ " " " $ ! ! " $ " " " " # ! # $ $ " " ! $ ! $ # ! $ " $ ! " ! $ " # ! $ " $ $ ! " " " $ $ # $ # ! " ! ! $ " $ $ #$ $ # $ $ " ! " " # " ! # $ # $ $ ! # ! $ " ! # # ! # " ! " " # # # $ # " $ $ ! # ! " " $ " $ $ ! " " #$ " # $ # # $ " " $ # ## # # $ $ ! $ " " " " $ ! ! " ! # # ! " " $ # # ! # $ ! ! ! " # " $ " $ " " ! # " ! # # # ! $ " " ! $ $ # " $ ! " ! $ # # ! " $ ! # ! ! " ! # " $ $ " ! # " $ " " # " $ $" " $ # $ ! " $ $ " # $ ! # $ # # ! " ! $ # # $ $ ! # " " ! # $ " # ! $ $ " ! $ " # $ " $ $ " $ # $ #! # ! " " $ $ " $ " ! ! " " $ " ! !! " ! # " $ " $ $ $ ! $ # # # " # ! ! # $ " $ $ ! " " # ! ! $ ! # # $ $ " " ! " " !" " ! $" # # $ # # ! $ $ " # # " # # # $ $ # $ ! " $ $ " $ $ ! ! # " # $ " " # ! $ $ ! $ ! ! $ # " ! " # # # $ " ! # " " " $ # ! " " # $ $ ! $ # ! $ # $ ! " " $ $ ! #! " $ # " ! ! " " # " $ # # $ # " " " " ! # # $ # # # # ! " $ $ # # ! # ! " # # # ! # " # $ ! # $ " $ $ " $ " " # ! ! ! $ " $ # # ! " ! # " $ ! $ " $ # $ ! " " # # ! $ $ # $ $ $ $ ! # " # ! # ! " ! $ ! " # " $ " ! $ # " # # ! # # ! " " ! " $ " $ $ ! ! ! " $ #" $ $ ! " # ! ! ! " ! " " $ # $ " $! # " " # # # $ # ! ! " " ! " # $ ! # !$ " $ ! " $ $ " # # # $ ! " " ! # " $ ! # ! " " " $ " ! ! $ # # ! # # ! " $ $ # ! " " " #" " $# ! # ! " # # $ # $ $ $ $ " $ " # " #! " $ ! " $ ! ! $ " $ ## ! # " $ # $ ! #" $ # # # " " $ ! " " $ $ " # ! # ! $ $ # # ! $ " " ! " . .& *& )& '& %& (& ,& -& +& .&& ..& .*& .)& .'& .%& .(& .,& .-& .+& *&& *.& **& *)& *'& *%& *(& *,& *-& *+& )&& ).& )*& ))& )'& )%& )(& ),& )-& )+& '&& '.& '*& ')& ''& '%& '(& ',& '-& '+& %&& %.& %*& %)& %'& %%& %(& %,& %-& %+& (&& (.& (*& ()& ('& (%& ((& (,& (-& (+& ,&& ,.& ,*& ,)& ,'& ,%& ,(& ,,& ,-& ,+& -&& -.& -*& -)& -'& -%& -(& -,& --& -+& +&& +.& +*& +)& +'& +%& +(& +,& +-& ++& +++ &7& .7& "/012345026.- 5 end RNA Principal 1 SL6 GCUGGUG U A GAUCUA A AAGGA GCCA AGUU CGGCGCC AAGUC U ||||| |||| ||.| || |||| ||||| UUCCU CGGU UCGA GCAGCGG UUCAG U A AGUCA A U U G A A G A U U U U C A A G A A A A C U G G A A C A C U A A AC GAA U ACAUAGC AGUG GUGUU CCGU CUCA G ||||||| |||| .||.. |||| |||| UGUAUCG UCAC UACGG GGCA GAGU C C A GGA AUUC G U U CUGUGUGGCUG CACUCGGC G |||.|||. || |||| .|| C GACGCACUCAC GUGAUUCG A U U A U A A U U A A U A A C U A A U U A C U G AC UAA G UCGUUGA CAGG ACGAG CUC UCUAUCUU-80nt- ||||||| |||| ||.|| ||| ||.||||| AGCAACU GUCC UGUUC GAG AGGUAGAA UUG C C SARS-CoV-2 SL5 SL4 SL3 SL2 SL1 Principal 2 1-500 500-1000 5 UTR GCUGGUG U A GAUCUA A AAGGA GCCA AGUU CGGCGCC AAGUC U ||||| |||| ||.| || |||| ||||| UUCCU CGGU UCGA GCAGCGG UUCAG U A AGUCA A U U G A A G A U U U U C A A G A A A A C U G G A A C A C U A A AC GAA U ACAUAGC AGUG GUGUU CCGU CUCA G ||||||| |||| .||.. |||| |||| UGUAUCG UCAC UACGG GGCA GAGU C C A GGA AUUC G U U CUGUGUGGCUG CACUCGGC G |||.|||. || |||| .|| C GACGCACUCAC GUGAUUCG A U U A U A A U U A A U A A C U A A U U A C U G AC UAA G UCGUUGA CAGG ACGAG CUC UCUAUCUU-A AGUCA A U U G A A G A U U U U C A A G A A A A C U G G A A C A C U A A AC GAA U cgauugcaucACAUAGC AGUG GUGUU CCGU CUCA G ||||||||||||||||| |||| .||.. |||| |||| gcuaacguagUGUAUCG UCAC UACGG GGCA GAGU C C A GGA AUUC G U U cgauugcaucCUGUGUGGCUG CACUCGGC G |||||||||||||.|||. || |||| .|| C gcuaacguagGACGCACUCAC GUGAUUCG A U U A U A A U U A A U A A C U A A U U A C U G AC UAA G agauugcaucUCGUUGA CAGG ACGAG CUC UCUAUCUU ||||||||||||||||| |||| ||.|| ||| ||.||||| ucuaacguagAGCAACU GUCC UGUUC GAG AGGUAGAA UUG C C E U U CUGUGUGGCUG CACUCGGC G |||.|||. || |||| .|| C GACGCACUCAC GUGAUUCG A U U A u U a A c A c U a U u A a A c U g A a u g u u c c A a C g U a A u A u U a U c A g C c U u G AC UAA G UCGUUGA CAGG ACGAG CUC UCUAUCUU ||||||| |||| ||.|| ||| ||.||||| AGCAACU GUCC UGUUC GAG AGGUAGAA UUG C C GCUGGUG U A GAUCUA A AAGGA GCCA AGUU CGGCGCC AAGUC U ||||| |||| ||.| || |||| ||||| UUCCU CGGU UCGA GCAGCGG UUCAG U A AGUCA A U U G A A u G a A c U c U c U a U u C a A c A g a u g G u A u A c A c A a C g U a G u G u A a A c C g A c C u U A A AC GAA U ACAUAGC AGUG GUGUU CCGU CUCA G ||||||| |||| .||.. |||| |||| UGUAUCG UCAC UACGG GGCA GAGU C C A GGA AUUC G F Wildtype U U gguguguuCUG CcagCGuC G || | . | C GACGCACUCAC GUGAUUCG A U U A U A A U U A A U A A C U A A U U A C U G AC UAA G Uauggug gcuu gauca agu cagcgagc | . . . AGCAACU GUCC UGUUC GAG AGGUAGAA UUG C C GCUGGUG U A GAUCUA A ggaag cugg cugc acGguag AAGUC U ||..| |||| ||.| || |||| ||||| ccuuc gacc gaug ugAcauc UUCAG U A AGUCA A U U G A A G A U U U U C A A G A A A A C U G G A A C A C U A A AC GAA U cacgcua cugc ugugg gaug CUCA G ||||||| |||| .||.. |||| |||| gugcgau gacg gcauu cuac GAGU C C A GGA AUUC G U U gguguguuCUG CcagCGuC G |||.|||. || |||| .|| C ccauacagCAC GgucUUaG A U U A U A A U U A A U A A C U A A U U A C U G AC UAA G UauggugA AGUCA A U U G A A G A U U U U C A A G A A A A C U G G A A C A C U A A AC GAA U cacgcua cugc ugugg gaug CUCA G |||| UGUAUCG UCAC UACGG GGCA GAGU C C A GGA AUUC G SL4 SL5 SARS-CoV-1 SL1-5 G A U U U U AAGUGA AUAGC |||||| ||||| U UUCACU UAUCG C C G C C C U C G C GUUCU U ||||| U CAAGA G C U UU UUG U ||| U AGC A A AA C U U A A A U A A A A U U U C G UU AUUGU UU GCCC G UGUU AGCGUAU GUU CAC GUCU GGUGGG GGCA A .||| | |||| |||.||| ||. ||| |.|| |.||.| |||| A UGGG C ACAA UCGUAUA CAG GUG CGGA CUACUC GUCU U C A U GUC UU UUCG A U U U G-C A G-C A C-G U A U C-G C A-U U U.G A A A A C U C-G U A-U G C-G A U.G A G-C U A A A C U U U A G-C U C-G U.G UUA CG U UU UC-G U UU UUUUCAG GAGCGU UG CUC GUACG CACAUCA GUC ||||||| |||||. || ||| .|||| ||||||| |.| U AAAAGUC CUCGCG AC GAG UAUGC GUGUAGU CGG C UUG C A UG U-A G-C G-C C-G G-C C-G G-C C A G A UGG MERS-Cov SL1-5 B 25nM

UC GUUC U |||| A CAAG A CA UC GUUC U |||| A CAAG A CA UC GUUC U |||| u CAAG A CA UC GUUC U |||| A CAAG u CA UC GUUC U |||| A CAAG A Cu UC GUUC c |||| A CAAG A CA Uu GUUC U |||| A CAAG A uA C 1-1000G AAA A U GGUU GGUUU UACCU C |||| ||||| ||||| C C CCAA CCAAA AUGGA G A A CA C U U A C-G C G-C U G.U U C U U U-A C C-G G A-U A C-G U U C G-C U U-A C C C U G.U U G-C G U-A U G-C A U.G G G-C A U-A U C-G C U U AACUUUAAAAAT A AU GUUCUCUAAACGA UAAUUA ||.|| | ||||| |||||| A CAGGACAGUUGCU AUUAAU GUC CA U ACGAG ||||| A UGCUC C A U A G G G U U A A A A U GGUUU UACCU C ||||| ||||| CCAAA AUGGA C A CA C A C C A A C U U U C C GAUCU U ||||| U CUAGA G U U UC GUUC U |||| A CAAG A U CA U U A A A A U U U CUGUGUGGCUG CACUCGGC G |||.|||. || |||| .|| C GACGCACUCAC GUGAUUCG A U U A U A A U U A A U A A C U A A U U A C U G AC UAA G UCGUUGA CAGG ACGAG CUC UCUAUCUU-80nt- ||||||| |||| ||.|| ||| ||.||||| AGCAACU GUCC UGUUC GAG AGGUAGAA UUG C C G AAA A U GGUU GGUUU UACCU C |||| ||||| ||||| C C CCAA CCAAA AUGGA G A A CA C U U A C-G C G-C U G.U U C U U U-A C C-G G A-U A C-G U U C G-C U U-A C C C U G.U U G-C G U-A U G-C A U.G G G-C A U-A U C-G C U U AACUUUAAAAAT A AU GUUCUCUuAACGA UAAUUA ||.|| |.||||| |||||| A CAGGACAGUUGCU AUUAAU GUC CA U ACGAG ||||| A UGCUC C A U A G AAA A U GGUU GGUUU UACCU C |||| ||||| ||||| C C CCAA CCAAA AUGGA G A A CA C U U A C-G C G-C U G.U U C U U U-A C C-G G A-U A C-G U U C G-C U U-A C C C U G.U U G-C G U-A U G-C A U.G G G-C A U-A U C-G C U U AACUUUAAAAAT A AU GUUCUCUuAACGA UAAUUA ||.|| ||||| |||||| A CAGGACccUUGCU AUUAAU GUC CA U ACGAG ||||| A UGCUC C A U A G AAA A U GGUU GGUUU UACCU C |||| ||||| ||||| C C CCAA CCAAA AUGGA G A A CA C U U A C-G C G-C U G.U U C U U U-A C C-G G A-U A C-G U U C G-C U U-A C C C U G.U U G-C G U-A U G-C A U.G G G-C A U-A U C-G C U U AACUUUAAAAAT A AU GUUCUCUAuACGA UAAUUA ||.|| | |||| |||||| A CAGGACAGUUGCU AUUAAU GUC CA U ACGAG ||||| A UGCUC C A U A G AAA A U GGUU GGUUU UACCU C |||| ||||| ||||| C C CCAA CCAAA AUGGA G A A CA C U U A C-G C G-C U G.U U C U U U-A C C-G G A-U A C-G U U C G-C U U-A C C C U G.U U G-C G U-A U G-C A U.G G G-C A U-A U C-G C U U AACUUUAAAAAT A AU GUUCUCUAuACGA UAAUUA ||.|| |||| |||||| A CAGGACccUUGCU AUUAAU GUC CA U ACGAG ||||| A UGCUC C A U A G AAA A U GGUU GGUUU UACCU C |||| ||||| ||||| C C CCAA CCAAA AUGGA G A A CA C U U A C-G C G-C U G.U U C U U U-A C C-G G A-U A C-G U U C G-C U U-A C C C U G.U U G-C G U-A U G-C A U.G G G-C A U-A U C-G C U U AACUUUAAAAAT A AU GUUCUCUAAACGA UAAUUA ||.|| ||||| |||||| A CAGGACccUUGCU AUUAAU GUC CA U ACGAG ||||| A UGCUC C A U A G G G U U A A A A U GGUUU UACCU C ||||| ||||| CCAAA AUGGA C A CA C A C C A A C U U U C C GAUCU U ||||| U CUAGA G U U UC GUUC U |||| A CAAG A U CA U U A A A A U U U CUGUGUGGCUG CACUCGGC G |||.|||. || |||| .|| C GACGCACUCAC GUGAUUCG A U U A U A A U U A A U A A C U A A U U A C U G AC UAA G UCGUUcc CAGG ACGAG CUC UCUAUCUU ||||| |||| ||.|| ||| ||.||||| AGCAACU GUCC UGUUC GAG AGGUAGAA UUG C C G G G U U A A A A U GGUUU UACCU C ||||| ||||| CCAAA AUGGA C A CA C A C C A A C U U U C C GAUCU U ||||| U CUAGA G U U UC GUUC U |||| u CAAG A U CA U U A A A A U U U CUGUGUGGCUG CACUCGGC G |||.|||. || |||| .|| C GACGCACUCAC GUGAUUCG A U U A U A A U U A A U A A C U A A U U A C U G AC UAA G UCGUUGA CAGG ACGAG CUC UCUAUCUU ||||||| |||| ||.|| ||| ||.||||| AGCAACU GUCC UGUUC GAG AGGUAGAA UUG C C D G G G G U U A A A A U GGUUU UACCU C ||||| ||||| CCAAA AUGGA C A CA C A C C A A C U U U C C GAUCU U ||||| U CUAGA G U U UC GUUC U |||| u CAAG A U CA U U A A A A U U U CUGUGUGGCUG CACUCGGC G |||.|||. || |||| .|| C GACGCACUCAC GUGAUUCG A U U A U A A U U A A U A A C U A A U U A C U G AC UAA G UCGUUcc CAGG ACGAG CUC UCUAUCUU ||||| |||| ||.|| ||| ||.||||| AGCAACU GUCC UGUUC GAG AGGUAGAA UUG C C G G G U U A A A A U GGUUU UACCU C ||||| ||||| CCAAA AUGGA C A CA C A C C A A C U U U C C GAUCU U ||||| U CUAGA G U U UC GUUC U |||| A CAAG u U CA U U A A A A U U U CUGUGUGGCUG CACUCGGC G |||.|||. || |||| .|| C GACGCACUCAC GUGAUUCG A U U A U A A U U A A U A A C U A A U U A C U G AC UAA G UCGUUGA CAGG ACGAG CUC UCUAUCUU ||||||| |||| ||.|| ||| ||.||||| AGCAACU GUCC UGUUC GAG AGGUAGAA UUG C C G G G U U A A A A U GGUUU UACCU C ||||| ||||| CCAAA AUGGA C A CA C A C C A A C U U U C C GAUCU U ||||| U CUAGA G U U UC GUUC U |||| A CAAG u U CA U U A A A A U U U CUGUGUGGCUG CACUCGGC G |||.|||. || |||| .|| C GACGCACUCAC GUGAUUCG A U U A U A A U U A A U A A C U A A U U A C U G AC UAA G

Wildtype +dsRNAb +dsRNAd +ssRNAa -ssRNA Restore pairing

3.2µM N-protein 25nM RNA 18 hours 

G G G U U A A A A U GGUUU UACCU C ||||| ||||| CCAAA AUGGA C A CA C A C C A A C U U U C C GAUCU U ||||| U CUAGA G U U UC GUUC U |||| A CAAG A U CA U U A A A A U U UNano G G G U U A A A A U GGUUU UACCU C ||||| ||||| CCAAA AUGGA C A CA C A C C A A C U U U C C GAUCU U ||||| U CUAGA G U U UC GUUC U |||| A CAAG A U CA U U A A A A U U U CUGUGUGGCUG CACUCGGC G |||.|||. || |||| .|| C GACGCACUCAC GUGAUUCG A U U A U A A U U A A U A A C U A A U U A C U G AC

Only +dsRNA (teal) mutants enhance LLPS in the context of the 5′UTR fragment. All other mutations do not significantly alter LLPS. 3.2uM N-protein (green) 25nM RNA 18 hours of incubation. H20 is water only control. (B) Design of luciferase fusion to the 5′UTR of SARS-CoV-2 constructs. Wildtype or a more structured 5′UTR (+dsRNAb) was fused to nano luciferase (orange). Arrows show the approximate length of nano luciferase as compared to 5′end 1-1000 RNA. (C) Only +dsRNAb UTR: Nano Luciferase undergoes LLPS at the highest tested RNA concentration (25nM/3.2µM N-protein (green)) (D) A280 absorbance of the remaining protein in the diffuse phase from Fig. 4C . Error bars mark standard deviation for the three replicates and * indicate significance students T test (** <0.01, ns not significant) with brackets showing comparison for the indicated statistical test. (E) In vitro translation assay results for nano luciferase wildtype or more structured fusion constructs. 20-minute incubation with 3.2µM Nprotein prior to in vitro translation is sufficient to completely repress translation of nano luciferase. Error bars mark standard deviation for the three replicates and * indicate significance students T test (** <0.01, ns not significant) with brackets showing comparison for the indicated statistical test. (F) Presence of N-protein LLPS promoting RNA structures is associated with reduced translation in diffuse phase conditions. Normalized luminescence for nano luciferase constructs (no protein control fluorescent signal is set to 1). Nano luciferase +dsRNAb has a much greater reduction in normalized signal as compared to wildtype. 

(A) At 3.6µM N-protein (green) and 15nM RNA following 18 hours of incubation, mutations in the context of 1-1000nt which increase the dsRNA (+dsRNA teal) of principal site 1 only, 2 only, or 1, and 2 lead to enhanced LLPS in comparison to wildtype regardless of whether the anti-sense sequence to the principal site was inserted on the 5′ or 3′ side. Enhancement of LLPS may be length dependent. H20 in indicates equivalent volume of added water only control. (B) Location or sequence of ssRNA (+ssRNA orange) insertion leads to equally negligible levels of LLPS enhancement. 3.6µM N-protein, 50nM RNA 18hours of incubation in the 500-1000 sequence context. (C-E) Phase diagrams for equivalent RNA and N-protein concentrations (3µM, 1µM, 0.3µM) for 1-1000 wildtype, +dsRNAa for 5nM, 15nM, or 25nM RNA. At 3µM, +dsRNAa leads to more LLPS relative to Wild-type 1-1000 for all tested RNA concentrations. At 1µM +dsRNAa shifts the phase boundary to the left relative to wildtype. 0.3µM N-protein does not drive LLPS for any sequence at any tested RNA concentration. (E) A280 absorbance of for 3µM N-protein concentration for panels S1C, D. For all tested RNA concentrations relative to wildtype, +dsRNAa has less protein in solution following 16 hours incubation as measured by A280 nanodrop. Error bars mark standard deviation for the three replicates and * indicate significance students T test (*** p<0.001, ** p<0.01, ns not significant) with brackets showing comparison for the indicated statistical test. (G) 1-1000 RNA and its mutants (purple -ssRNA, teal +dsRNA, oranges -ssRNA, -dsRNA gray, Restore pairing blue) lead to equivalent LLPS of Whi3 protein (green). For all images scale bar indicates 10µm all experiments show representative images from at least 3 replicates.

Structure model of SL5 for SARS-CoV-2 with the location of the start codon of non-structural protein 1 (NSP1) in orange text. AU base pair of the start codon is replaced with a GU wobble pair (ATG à gTG) to eliminate NSP1 translation while sparing RNA structure. (B) Wildtype and Start Mutant RNA result in similar levels of LLPS cell free. 3.2µM N-protein (green) and 25nM RNA following 2.5 hours of incubation. (C) NSP1 protein sequence of mutants tested in S2E-G. destroy NSP1 production. (D) N: GFP protein signal (FIRE blue low signal white high signal) key and 5′end overexpression plasmids design. (E) Representative HEK293T cells co-transfected with N: GFP and the indicated 5′end fragment at 24 hours (left 3 panels) or 48 hours (right 3 panels). +dsRNAa mutant produces more puncta at 24 hours (4-5 per cell) compared to wildtype, start, or empty (2-3 per cell). Difference is reduced at 48 hours (4-5 puncta in all three 5′end containing cells). (F) Quantification of the number of droplets per Micron^2 at 24 hours. +dsRNAa produces significantly more puncta per unit area then the Start mutant. * indicate significance students T test (*** p<0.001, ns not significant) with brackets showing comparison for the indicated statistical test. (G) Quantification of the mean intensity of N: GFP signal at 24 hours. Analyzed cells have similar GFP signal distribution. No comparisons are significant (ns). (H) Addition of dsRNA (teal) (+dsRNAa-d) enhances N-protein LLPS. Representative images from the two-hour incubation timepoint for panels 1F. N-protein signal is shown in green. (I and J) Representative images from the two-hour incubation timepoint for panels Figure 1D (1-500) and 1E (500-1000). For all images scale bar indicates 10µm. (A) Structure of the 5′end 1-1000 RNA adapted from Iserman et al. . Blue squares indicate the location of reduced N-protein crosslinking following Y109A mutation which destroys RBD1 (see B). Dark Blue and light blue circles indicate locations of the perfect TRS loop sequence (CUAAAC dark blue) verses (YYAAAY light blue) showing that peak reduction is often adjacent to TRS or TRS-like motifs. (B) RNP-map signal for wildtype N-protein and Y109A mutant N-protein (which is predicted to destroy RBD1. Blue boxes (lower case letters) are equivalent between S3A and B. Note that crosslinking of N-protein to RNA appears to occur more readily at single-stranded than double-stranded nts, despites N-protein's ability to bind both double-and single-stranded nts. (see Figure 1) . Green text indicates location of principal sites. (C-I) Density of TRS-Loop-like motif (YYAAAY) across the genome other than CUAAAC (Fig 2C and D) . Y axis is the relative density of the indicated sequence across the SARS-CoV-2 genome. Fig. S4A . Only 37°C shows a difference in signal consistent with previous results (Fig 3B) . (E) Alignment of the predicted structure for SARS-CoV-2 (orange ribbon), the crystal structure of MERS-CoV (magenta ribbon), to the crystal structure of SARS-CoV-1 (blue ribbon) RBD2 dimerization domain. Proteins may adopt similar folds. (F) A280 measurement of remaining N-protein in the solution for Fig. 3G .

Repeat of LCST experiment with wildtype RNA at lower N-protein (1µM) and RNA (15nM) concentration. H20 alone results in less LLPS indicative for some RNA dependence for N-protein RBD-2 Del. (H) A280 measurements for (Fig. S4G) 

Recombinant Protein Expression and Purification: For protein purification, full-length Nprotein was tagged with an N-terminal 6-Histidine tag (pET30b-6xHis-TEV-Nucleocapsid, N-Y109A, and N-RBD2-Del) were expressed in BL21 E. coli (New England Biolabs). All steps of the purification after growth of bacteria were performed at 4°C. Cells were lysed in lysis buffer (1.5M NaCl, 20 mM Phosphate buffer pH 7.5, 20 mM Imidazole, 10mg/mL lysozyme, 1 tablet of Roche EDTA-free protease inhibitor cocktail Millipore Sigma 11873580001) and via sonication. The lysate was then clarified via centrifugation (SS34 rotor, 20,000 rpm 30 minutes) and the supernatant was incubated and passed over a HisPurTM Cobalt Resin (ThermoFisher Scientific 89965) in gravity columns. The resin was then washed with 4X 10 CV wash buffer (1.5M NaCl, 20 mM Phosphate buffer pH 7.5, 20 mM Imidazole) and protein was eluted with 4 CV Elution buffer (0.25 M NaCl, 20 mM Phosphate buffer pH 7.5, 200 mM Imidazole). The eluate was then dialyzed into fresh storage buffer (0.25 M NaCl, 20 mM Phosphate buffer) and aliquots of protein were flash frozen and stored at -80 °C. Protein was checked for purity by running an SDS-PAGE gel followed by Coomassie staining as well as checking the level of RNA contamination via Nanodrop and through running of a native agarose RNA gel. All experiments were performed with His-tagged N-protein. Whi3 was purified according to our established protocols (Langdon et al., 2018; Zhang et al., 2015) . Wildtype   ATGCACCATCATCATCATCATTCTTCTGGTGAAAACCTGTATTTTCAGGGCGTCGACATGTC  TGATAATGGACCCCAAAATCAGCGAAATGCACCCCGCATTACGTTTGGTGGACCCTCAGAT  TCAACTGGCAGTAACCAGAATGGAGAACGCAGTGGGGCGCGATCAAAACAACGTCGGCCC  CAAGGTTTACCCAATAATACTGCGTCTTGGTTCACCGCTCTCACTCAACATGGCAAGGAAG  ACCTTAAATTCCCTCGAGGACAAGGCGTTCCAATTAACACCAATAGCAGTCCAGATGACCA  AATTGGCTACTACCGAAGAGCTACCAGACGAATTCGTGGTGGTGACGGTAAAATGAAAGAT  CTCAGTCCAAGATGGTATTTCTACTACCTAGGAACTGGGCCAGAAGCTGGACTTCCCTATG  GTGCTAACAAAGACGGCATCATATGGGTTGCAACTGAGGGAGCCTTGAATACACCAAAAGA  TCACATTGGCACCCGCAATCCTGCTAACAATGCTGCAATCGTGCTACAACTTCCTCAAGGA  ACAACATTGCCAAAAGGCTTCTACGCAGAAGGGAGCAGAGGCGGCAGTCAAGCCTCTTCT  CGTTCCTCATCACGTAGTCGCAACAGTTCAAGAAATTCAACTCCAGGCAGCAGTAGGGGAA  CTTCTCCTGCTAGAATGGCTGGCAATGGCGGTGATGCTGCTCTTGCTTTGCTGCTGCTTGA  CAGATTGAACCAGCTTGAGAGCAAAATGTCTGGTAAAGGCCAACAACAACAAGGCCAAACT  GTCACTAAGAAATCTGCTGCTGAGGCTTCTAAGAAGCCTCGGCAAAAACGTACTGCCACTA  AAGCATACAATGTAACACAAGCTTTCGGCAGACGTGGTCCAGAACAAACCCAAGGAAATTT  TGGGGACCAGGAACTAATCAGACAAGGAACTGATTACAAACATTGGCCGCAAATTGCACAA  TTTGCCCCCAGCGCTTCAGCGTTCTTCGGAATGTCGCGCATTGGCATGGAAGTCACACCTT  CGGGAACGTGGTTGACCTACACAGGTGCCATCAAATTGGATGACAAAGATCCAAATTTCAA  AGATCAAGTCATTTTGCTGAATAAGCATATTGACGCATACAAAACATTCCCACCAACAGAGC  CTAAAAAGGACAAAAAGAAGAAGGCTGATGAAACTCAAGCCTTACCGCAGAGACAGAAGAA  ACAGCAAACTGTGACTCTTCTTCCTGCTGCAGATTTGGATGATTTCTCCAAACAATTGCAAC  AATCCATGAGCAGTGCTGACTCAACTCAGGCCTAA   Y109A   ATGCACCATCATCATCATCATTCTTCTGGTGAAAACCTGTATTTTCAGGGCGTCGACATGTC  TGATAATGGACCCCAAAATCAGCGAAATGCACCCCGCATTACGTTTGGTGGACCCTCAGAT  TCAACTGGCAGTAACCAGAATGGAGAACGCAGTGGGGCGCGATCAAAACAACGTCGGCCC  CAAGGTTTACCCAATAATACTGCGTCTTGGTTCACCGCTCTCACTCAACATGGCAAGGAAG  ACCTTAAATTCCCTCGAGGACAAGGCGTTCCAATTAACACCAATAGCAGTCCAGATGACCA  AATTGGCTACTACCGAAGAGCTACCAGACGAATTCGTGGTGGTGACGGTAAAATGAAAGAT  CTCAGTCCAAGATGGGCATTCTACTACCTAGGAACTGGGCCAGAAGCTGGACTTCCCTATG  GTGCTAACAAAGACGGCATCATATGGGTTGCAACTGAGGGAGCCTTGAATACACCAAAAGA  TCACATTGGCACCCGCAATCCTGCTAACAATGCTGCAATCGTGCTACAACTTCCTCAAGGA  ACAACATTGCCAAAAGGCTTCTACGCAGAAGGGAGCAGAGGCGGCAGTCAAGCCTCTTCT  CGTTCCTCATCACGTAGTCGCAACAGTTCAAGAAATTCAACTCCAGGCAGCAGTAGGGGAA  CTTCTCCTGCTAGAATGGCTGGCAATGGCGGTGATGCTGCTCTTGCTTTGCTGCTGCTTGA  CAGATTGAACCAGCTTGAGAGCAAAATGTCTGGTAAAGGCCAACAACAACAAGGCCAAACT  GTCACTAAGAAATCTGCTGCTGAGGCTTCTAAGAAGCCTCGGCAAAAACGTACTGCCACTA  AAGCATACAATGTAACACAAGCTTTCGGCAGACGTGGTCCAGAACAAACCCAAGGAAATTT  TGGGGACCAGGAACTAATCAGACAAGGAACTGATTACAAACATTGGCCGCAAATTGCACAA  TTTGCCCCCAGCGCTTCAGCGTTCTTCGGAATGTCGCGCATTGGCATGGAAGTCACACCTT  CGGGAACGTGGTTGACCTACACAGGTGCCATCAAATTGGATGACAAAGATCCAAATTTCAA  AGATCAAGTCATTTTGCTGAATAAGCATATTGACGCATACAAAACATTCCCACCAACAGAGC  CTAAAAAGGACAAAAAGAAGAAGGCTGATGAAACTCAAGCCTTACCGCAGAGACAGAAGAA ACAGCAAACTGTGACTCTTCTTCCTGCTGCAGATTTGGATGATTTCTCCAAACAATTGCAAC AATCCATGAGCAGTGCTGACTCAACTCAGGCCTAA ATGCACCATCATCATCATCATTCTTCTGGTGAAAACCTGTATTTTCAGGGCGTCGACATGTC  TGATAATGGACCCCAAAATCAGCGAAATGCACCCCGCATTACGTTTGGTGGACCCTCAGAT  TCAACTGGCAGTAACCAGAATGGAGAACGCAGTGGGGCGCGATCAAAACAACGTCGGCCC  CAAGGTTTACCCAATAATACTGCGTCTTGGTTCACCGCTCTCACTCAACATGGCAAGGAAG  ACCTTAAATTCCCTCGAGGACAAGGCGTTCCAATTAACACCAATAGCAGTCCAGATGACCA  AATTGGCTACTACCGAAGAGCTACCAGACGAATTCGTGGTGGTGACGGTAAAATGAAAGAT  CTCAGTCCAAGATGGTATTTCTACTACCTAGGAACTGGGCCAGAAGCTGGACTTCCCTATG  GTGCTAACAAAGACGGCATCATATGGGTTGCAACTGAGGGAGCCTTGAATACACCAAAAGA  TCACATTGGCACCCGCAATCCTGCTAACAATGCTGCAATCGTGCTACAACTTCCTCAAGGA  ACAACATTGCCAAAAGGCTTCTACGCAGAAGGGAGCAGAGGCGGCAGTCAAGCCTCTTCT  CGTTCCTCATCACGTAGTCGCAACAGTTCAAGAAATTCAACTCCAGGCAGCAGTAGGGGAA  CTTCTCCTGCTAGAATGGCTGGCAATGGCGGTGATGCTGCTCTTGCTTTGCTGCTGCTTGA  CAGATTGAACCAGCTTGAGAGCAAAATGTCTGGTAAAGGCCAACAACAACAAGGCCAAACT  GTCACTGGTCCAGAACAAACCCAAGGAAATTTTGGGGACCAGGAACTAATCAGACAAGGA  ACTGATTACAAACATTGGCCGCAAATTGCACAATTTGCCCCCAGCGCTTCAGCGTTCTTCG  GAATGTCGCGCATTGGCATGGAAGTCACACCTTCGGGAACGTGGTTGACCTACACAGGTG  CCATCAAATTGGATGACAAAGATCCAAATTTCAAAGATCAAGTCATTTTGCTGAATAAGCAT  ATTGACGCATACAAAACATTCCCACCAACAGAGCCTAAAAAGGACAAAAAGAAGAAGGCTG  ATGAAACTCAAGCCTTACCGCAGAGACAGAAGAAACAGCAAACTGTGACTCTTCTTCCTGC  TGCAGATTTGGATGATTTCTCCAAACAATTGCAACAATCCATGAGCAGTGCTGACTCAACTC  AGGCCTAA Dyeing of N-protein: N-protein was dyed by adding (3:1) Atto 488 NHS ester (Millipore Sigma 41698) to purified protein and incubating mix at 4°C for 1 hour with rocking. Unbound dye was removed by overnight dialysis into protein storage buffer. For LLPS percent of dyed protein was adjusted to 10% of total by dilution with undyed protein.

RNA Template Design/Production: Template predicted structure was designed using Vienna fold (http://rna.tbi.univie.ac.at). Sequences were generated via site directed mutagenesis using overlapping oligos (IDT). DNA sequences of tested RNA fragments are as follows.

GGGTTAAAGGTTTATACCTTCCCAGGTAACAAACCAACCAACTTTCGATCTCTTGTAGATCT GTTCTCTAAACGAACTTTAAAATCTGTGTGGCTGTCACTCGGCTGCATGCTTAGTGCACTCA GGGTTAAAGGTTTATACCTTCCCAGGTAACAAACCAACCAACTTTCGATCTCTTGTAGATCT  TTTAAAATCTGTGTGGCTGTCACTCGGCTGCATGCTTAGTGCACTCACGCAGTATAATTAAT  AACTAATTACTGTCGTTGACAGGACACGAGTAACTCGTCTATCTTCTGCAGGCTGCTTACG  GTTTCGTCCGTGTTGCAGCCGATCATCAGCACATCTAGGTTTCGTCCGGGTGTGACCGAAA  GGTAAGATGGAGAGCCTTGTCCCTGGTTTCAACGAGAAAACACACGTCCAACTCAGTTTGC  CTGTTTTACAGGTTCGCGACGTGCTCGTACGTGGCTTTGGAGACTCCGTGGAGGAGGTCT  TATCAGAGGCACGTCAACATCTTAAAGATGGCACTTGTGGCTTAGTAGAAGTTGAAAAAGG  CGTTTTGCCTCAACTTGAACAGCCCTATGTGTTCATCAAACGTTCGGATGCTCGAACTGCA  CCTCATGGTCATGTTATGGTTGAGCTGGTAGCAGAACTCGAAGGCATTCAGTACGGTCGTA  GTGGTGAGACACTTGGTGTCCTTGTCCCTCATGTGGGCGAAATACCAGTGGCTTACCGCA  AGGTTCTTCTTCGTAAGAACGGTAATAAAGGAGCTGGTGGCCATAGTTACGGCGCCGATCT  AAAGTCATTTGACTTAGGCGACGAGCTTGGCACTGATCCTTATGAAGATTTTCAAGAAAACT  GGAACACTAAACATAGCAGTGGTGTTACCCGTGAACTCATGCGTGAGCTTAACGGAGGGG  CATACACTCGCTATGTCGATAACAACTTCTGTGGCCCTGATGGCTACCCTCTTGAGTGCAT  TAAAGACCTTCTAGCACGTGCTGGTAAAGCTTCATGCACTTTGTCCGAACAACTGGACTTTA  TTGACACTAAGAGGGGTGTATACTGCTGCCGTGAACATGAGCATGAAATTGCTTGGTACAC  GGAACGTTCTGGGCCCTCGA   1-1000 Add TRS 3′  GGGTTAAAGGTTTATACCTTCCCAGGTAACAAACCAACCAACTTTCGATCTCTTGTAGATCT  GTTCTCTAAACGAACTTTAAAATCTGTGTGGCTGTCACTCGGCTGCATGCTTAGTGCACTCA  CGCAGTATAATTAATAACTAATTACTGTCGTTGACAGGACACGAGTAACTCGTCTATCTTCT  GCAGGCTGCTTACGGTTTCGTCCGTGTTGCAGCCGATCATCAGCACATCTAGGTTTCGTCC  GGGTGTGACCGAAAGGTAAGATGGAGAGCCTTGTCCCTGGTTTCAACGAGAAAACACACG  TCCAACTCAGTTTGCCTGTTTTACAGGTTCGCGACGTGCTCGTACGTGGCTTTGGAGACTC  CGTGGAGGAGGTCTTATCAGAGGCACGTCAACATCTTAAAGATGGCACTTGTGGCTTAGTA  GAAGTTGAAAAAGGCGTTTTGCCTCAACTTGAACAGCCCTATGTGTTCATCAAACGTTCGG  ATGCTCGAACTGCACCTCATGGTCATGTTATGGTTGAGCTGGTAGCAGAACTCGAAGGCAT  TCAGTACGGTCGTAGTGGTGAGACACTTGGTGTCCTTGTCCCTCATGTGGGCGAAATACCA  GTGGCTTACCGCAAGGTTCTTCTTCGTAAGAACGGTAATAAAGGAGCTGGTGGCCATAGTT  ACGGCGCCGATCTAAAGTCATTTGACTTAGGCGACGAGCTTGGCACTGATCCTTATGAAGA  TTTTCAAGAAAACTGGAACACTAAACATAGCAGTGGTGTTACCCGTGAACTCATGCGTGAG  CTTAACGGAGGGGCATACACTCGCTATGTCGATAACAACTTCTGTGGCCCTGATGGCTACC  CTCTTGAGTGCATTAAAGACCTTCTAGCACGTGCTGGTAAAGCTTCATGCACTTTGTCCGA  ACAACTGGACTTTATTGACACTAAGAGGGGTGTATACTGCTGCCGTGAACATGAGCATGAA  ATTGCTTGGTACACGGAACGTTCTGTTCTCTAAACGAACGGGCCCTCGA   1-1000 A68U  GGGTTAAAGGTTTATACCTTCCCAGGTAACAAACCAACCAACTTTCGATCTCTTGTAGATCT  GTTCTCTTAACGAACTTTAAAATCTGTGTGGCTGTCACTCGGCTGCATGCTTAGTGCACTCA  CGCAGTATAATTAATAACTAATTACTGTCGTTGACAGGACACGAGTAACTCGTCTATCTTCT  GCAGGCTGCTTACGGTTTCGTCCGTGTTGCAGCCGATCATCAGCACATCTAGGTTTCGTCC  GGGTGTGACCGAAAGGTAAGATGGAGAGCCTTGTCCCTGGTTTCAACGAGAAAACACACG  TCCAACTCAGTTTGCCTGTTTTACAGGTTCGCGACGTGCTCGTACGTGGCTTTGGAGACTC  CGTGGAGGAGGTCTTATCAGAGGCACGTCAACATCTTAAAGATGGCACTTGTGGCTTAGTA  GAAGTTGAAAAAGGCGTTTTGCCTCAACTTGAACAGCCCTATGTGTTCATCAAACGTTCGG  ATGCTCGAACTGCACCTCATGGTCATGTTATGGTTGAGCTGGTAGCAGAACTCGAAGGCAT  TCAGTACGGTCGTAGTGGTGAGACACTTGGTGTCCTTGTCCCTCATGTGGGCGAAATACCA  GTGGCTTACCGCAAGGTTCTTCTTCGTAAGAACGGTAATAAAGGAGCTGGTGGCCATAGTT  ACGGCGCCGATCTAAAGTCATTTGACTTAGGCGACGAGCTTGGCACTGATCCTTATGAAGA  TTTTCAAGAAAACTGGAACACTAAACATAGCAGTGGTGTTACCCGTGAACTCATGCGTGAG  CTTAACGGAGGGGCATACACTCGCTATGTCGATAACAACTTCTGTGGCCCTGATGGCTACC  CTCTTGAGTGCATTAAAGACCTTCTAGCACGTGCTGGTAAAGCTTCATGCACTTTGTCCGA  ACAACTGGACTTTATTGACACTAAGAGGGGTGTATACTGCTGCCGTGAACATGAGCATGAA  ATTGCTTGGTACACGGAACGTTCTGGGCCCTCGA   1-1000 A69U  GGGTTAAAGGTTTATACCTTCCCAGGTAACAAACCAACCAACTTTCGATCTCTTGTAGATCT  GTTCTCTATACGAACTTTAAAATCTGTGTGGCTGTCACTCGGCTGCATGCTTAGTGCACTCA  CGCAGTATAATTAATAACTAATTACTGTCGTTGACAGGACACGAGTAACTCGTCTATCTTCT  GCAGGCTGCTTACGGTTTCGTCCGTGTTGCAGCCGATCATCAGCACATCTAGGTTTCGTCC  GGGTGTGACCGAAAGGTAAGATGGAGAGCCTTGTCCCTGGTTTCAACGAGAAAACACACG  TCCAACTCAGTTTGCCTGTTTTACAGGTTCGCGACGTGCTCGTACGTGGCTTTGGAGACTC  CGTGGAGGAGGTCTTATCAGAGGCACGTCAACATCTTAAAGATGGCACTTGTGGCTTAGTA  GAAGTTGAAAAAGGCGTTTTGCCTCAACTTGAACAGCCCTATGTGTTCATCAAACGTTCGG  ATGCTCGAACTGCACCTCATGGTCATGTTATGGTTGAGCTGGTAGCAGAACTCGAAGGCAT  TCAGTACGGTCGTAGTGGTGAGACACTTGGTGTCCTTGTCCCTCATGTGGGCGAAATACCA  GTGGCTTACCGCAAGGTTCTTCTTCGTAAGAACGGTAATAAAGGAGCTGGTGGCCATAGTT  ACGGCGCCGATCTAAAGTCATTTGACTTAGGCGACGAGCTTGGCACTGATCCTTATGAAGA  TTTTCAAGAAAACTGGAACACTAAACATAGCAGTGGTGTTACCCGTGAACTCATGCGTGAG  CTTAACGGAGGGGCATACACTCGCTATGTCGATAACAACTTCTGTGGCCCTGATGGCTACC  CTCTTGAGTGCATTAAAGACCTTCTAGCACGTGCTGGTAAAGCTTCATGCACTTTGTCCGA  ACAACTGGACTTTATTGACACTAAGAGGGGTGTATACTGCTGCCGTGAACATGAGCATGAA  ATTGCTTGGTACACGGAACGTTCTGGGCCCTCGA   1-1000 A70U  GGGTTAAAGGTTTATACCTTCCCAGGTAACAAACCAACCAACTTTCGATCTCTTGTAGATCT  GTTCTCTAATCGAACTTTAAAATCTGTGTGGCTGTCACTCGGCTGCATGCTTAGTGCACTCA  CGCAGTATAATTAATAACTAATTACTGTCGTTGACAGGACACGAGTAACTCGTCTATCTTCT  GCAGGCTGCTTACGGTTTCGTCCGTGTTGCAGCCGATCATCAGCACATCTAGGTTTCGTCC  GGGTGTGACCGAAAGGTAAGATGGAGAGCCTTGTCCCTGGTTTCAACGAGAAAACACACG  TCCAACTCAGTTTGCCTGTTTTACAGGTTCGCGACGTGCTCGTACGTGGCTTTGGAGACTC  CGTGGAGGAGGTCTTATCAGAGGCACGTCAACATCTTAAAGATGGCACTTGTGGCTTAGTA  GAAGTTGAAAAAGGCGTTTTGCCTCAACTTGAACAGCCCTATGTGTTCATCAAACGTTCGG  ATGCTCGAACTGCACCTCATGGTCATGTTATGGTTGAGCTGGTAGCAGAACTCGAAGGCAT  TCAGTACGGTCGTAGTGGTGAGACACTTGGTGTCCTTGTCCCTCATGTGGGCGAAATACCA  GTGGCTTACCGCAAGGTTCTTCTTCGTAAGAACGGTAATAAAGGAGCTGGTGGCCATAGTT  ACGGCGCCGATCTAAAGTCATTTGACTTAGGCGACGAGCTTGGCACTGATCCTTATGAAGA  TTTTCAAGAAAACTGGAACACTAAACATAGCAGTGGTGTTACCCGTGAACTCATGCGTGAG  CTTAACGGAGGGGCATACACTCGCTATGTCGATAACAACTTCTGTGGCCCTGATGGCTACC  CTCTTGAGTGCATTAAAGACCTTCTAGCACGTGCTGGTAAAGCTTCATGCACTTTGTCCGA  ACAACTGGACTTTATTGACACTAAGAGGGGTGTATACTGCTGCCGTGAACATGAGCATGAA  ATTGCTTGGTACACGGAACGTTCTGGGCCCTCGA   1-1000 Rare Y  GGGTTAAAGGTTTATACCTTCCCAGGTAACAAACCAACCAACTTTCGATCTCTTGTAGATCT  GTTCTCCAAACGAACTTTAAAATCTGTGTGGCTGTCACTCGGCTGCATGCTTAGTGCACTC  ACGCAGTATAATTAATAACTAATTACTGTCGTTGACAGGACACGAGTAACTCGTCTATCTTC  TGCAGGCTGCTTACGGTTTCGTCCGTGTTGCAGCCGATCATCAGCACATCTAGGTTTCGTC  CGGGTGTGACCGAAAGGTAAGATGGAGAGCCTTGTCCCTGGTTTCAACGAGAAAACACAC  GTCCAACTCAGTTTGCCTGTTTTACAGGTTCGCGACGTGCTCGTACGTGGCTTTGGAGACT  CCGTGGAGGAGGTCTTATCAGAGGCACGTCAACATCTTAAAGATGGCACTTGTGGCTTAGT  AGAAGTTGAAAAAGGCGTTTTGCCTCAACTTGAACAGCCCTATGTGTTCATCAAACGTTCG  GATGCTCGAACTGCACCTCATGGTCATGTTATGGTTGAGCTGGTAGCAGAACTCGAAGGC  ATTCAGTACGGTCGTAGTGGTGAGACACTTGGTGTCCTTGTCCCTCATGTGGGCGAAATAC  CAGTGGCTTACCGCAAGGTTCTTCTTCGTAAGAACGGTAATAAAGGAGCTGGTGGCCATAG  TTACGGCGCCGATCTAAAGTCATTTGACTTAGGCGACGAGCTTGGCACTGATCCTTATGAA  GATTTTCAAGAAAACTGGAACACTAAACATAGCAGTGGTGTTACCCGTGAACTCATGCGTG  AGCTTAACGGAGGGGCATACACTCGCTATGTCGATAACAACTTCTGTGGCCCTGATGGCTA  CCCTCTTGAGTGCATTAAAGACCTTCTAGCACGTGCTGGTAAAGCTTCATGCACTTTGTCC  GAACAACTGGACTTTATTGACACTAAGAGGGGTGTATACTGCTGCCGTGAACATGAGCATG  AAATTGCTTGGTACACGGAACGTTCTGGGCCCTCGA  1-1000 Common Y  GGGTTAAAGGTTTATACCTTCCCAGGTAACAAACCAACCAACTTTCGATCTCTTGTAGATCT  GTTCTTTAAATGAACTTTAAAATCTGTGTGGCTGTCACTCGGCTGCATGCTTAGTGCACTCA  CGCAGTATAATTAATAACTAATTACTGTCGTTGACAGGACACGAGTAACTCGTCTATCTTCT  GCAGGCTGCTTACGGTTTCGTCCGTGTTGCAGCCGATCATCAGCACATCTAGGTTTCGTCC  GGGTGTGACCGAAAGGTAAGATGGAGAGCCTTGTCCCTGGTTTCAACGAGAAAACACACG  TCCAACTCAGTTTGCCTGTTTTACAGGTTCGCGACGTGCTCGTACGTGGCTTTGGAGACTC  CGTGGAGGAGGTCTTATCAGAGGCACGTCAACATCTTAAAGATGGCACTTGTGGCTTAGTA  GAAGTTGAAAAAGGCGTTTTGCCTCAACTTGAACAGCCCTATGTGTTCATCAAACGTTCGG  ATGCTCGAACTGCACCTCATGGTCATGTTATGGTTGAGCTGGTAGCAGAACTCGAAGGCAT  TCAGTACGGTCGTAGTGGTGAGACACTTGGTGTCCTTGTCCCTCATGTGGGCGAAATACCA  GTGGCTTACCGCAAGGTTCTTCTTCGTAAGAACGGTAATAAAGGAGCTGGTGGCCATAGTT  ACGGCGCCGATCTAAAGTCATTTGACTTAGGCGACGAGCTTGGCACTGATCCTTATGAAGA  TTTTCAAGAAAACTGGAACACTAAACATAGCAGTGGTGTTACCCGTGAACTCATGCGTGAG  CTTAACGGAGGGGCATACACTCGCTATGTCGATAACAACTTCTGTGGCCCTGATGGCTACC  CTCTTGAGTGCATTAAAGACCTTCTAGCACGTGCTGGTAAAGCTTCATGCACTTTGTCCGA  ACAACTGGACTTTATTGACACTAAGAGGGGTGTATACTGCTGCCGTGAACATGAGCATGAA  ATTGCTTGGTACACGGAACGTTCTGGGCCCTCGA   1-1000 Rescue WT  GGGTTAAAGGTTTATACCTTCCCAGGTAACAAACCAACCAACTTTCGATCTCTTGTAGATCT  GTTCTCTAAACGAACTTTAAAATCTGTGTGGCTGTCACTCGGCTGCATGCTTAGTGCACTCA  CGCAGTATAATTAATAACTAATTACTGTCGTTCCCAGGACACGAGTAACTCGTCTATCTTCT  GCAGGCTGCTTACGGTTTCGTCCGTGTTGCAGCCGATCATCAGCACATCTAGGTTTCGTCC  GGGTGTGACCGAAAGGTAAGATGGAGAGCCTTGTCCCTGGTTTCAACGAGAAAACACACG  TCCAACTCAGTTTGCCTGTTTTACAGGTTCGCGACGTGCTCGTACGTGGCTTTGGAGACTC  CGTGGAGGAGGTCTTATCAGAGGCACGTCAACATCTTAAAGATGGCACTTGTGGCTTAGTA  GAAGTTGAAAAAGGCGTTTTGCCTCAACTTGAACAGCCCTATGTGTTCATCAAACGTTCGG  ATGCTCGAACTGCACCTCATGGTCATGTTATGGTTGAGCTGGTAGCAGAACTCGAAGGCAT  TCAGTACGGTCGTAGTGGTGAGACACTTGGTGTCCTTGTCCCTCATGTGGGCGAAATACCA  GTGGCTTACCGCAAGGTTCTTCTTCGTAAGAACGGTAATAAAGGAGCTGGTGGCCATAGTT  ACGGCGCCGATCTAAAGTCATTTGACTTAGGCGACGAGCTTGGCACTGATCCTTATGAAGA  TTTTCAAGAAAACTGGAACACTAAACATAGCAGTGGTGTTACCCGTGAACTCATGCGTGAG  CTTAACGGAGGGGCATACACTCGCTATGTCGATAACAACTTCTGTGGCCCTGATGGCTACC  CTCTTGAGTGCATTAAAGACCTTCTAGCACGTGCTGGTAAAGCTTCATGCACTTTGTCCGA  ACAACTGGACTTTATTGACACTAAGAGGGGTGTATACTGCTGCCGTGAACATGAGCATGAA  ATTGCTTGGTACACGGAACGTTCTGGGCCCTCGA   1-1000 Rescue A68U  GGGTTAAAGGTTTATACCTTCCCAGGTAACAAACCAACCAACTTTCGATCTCTTGTAGATCT  GTTCTCTATACGAACTTTAAAATCTGTGTGGCTGTCACTCGGCTGCATGCTTAGTGCACTCA  CGCAGTATAATTAATAACTAATTACTGTCGTTCCCAGGACACGAGTAACTCGTCTATCTTCT  GCAGGCTGCTTACGGTTTCGTCCGTGTTGCAGCCGATCATCAGCACATCTAGGTTTCGTCC  GGGTGTGACCGAAAGGTAAGATGGAGAGCCTTGTCCCTGGTTTCAACGAGAAAACACACG  TCCAACTCAGTTTGCCTGTTTTACAGGTTCGCGACGTGCTCGTACGTGGCTTTGGAGACTC  CGTGGAGGAGGTCTTATCAGAGGCACGTCAACATCTTAAAGATGGCACTTGTGGCTTAGTA  GAAGTTGAAAAAGGCGTTTTGCCTCAACTTGAACAGCCCTATGTGTTCATCAAACGTTCGG  ATGCTCGAACTGCACCTCATGGTCATGTTATGGTTGAGCTGGTAGCAGAACTCGAAGGCAT  TCAGTACGGTCGTAGTGGTGAGACACTTGGTGTCCTTGTCCCTCATGTGGGCGAAATACCA  GTGGCTTACCGCAAGGTTCTTCTTCGTAAGAACGGTAATAAAGGAGCTGGTGGCCATAGTT  ACGGCGCCGATCTAAAGTCATTTGACTTAGGCGACGAGCTTGGCACTGATCCTTATGAAGA  TTTTCAAGAAAACTGGAACACTAAACATAGCAGTGGTGTTACCCGTGAACTCATGCGTGAG  CTTAACGGAGGGGCATACACTCGCTATGTCGATAACAACTTCTGTGGCCCTGATGGCTACC  CTCTTGAGTGCATTAAAGACCTTCTAGCACGTGCTGGTAAAGCTTCATGCACTTTGTCCGA  ACAACTGGACTTTATTGACACTAAGAGGGGTGTATACTGCTGCCGTGAACATGAGCATGAA  ATTGCTTGGTACACGGAACGTTCTGGGCCCTCGA   1-1000 Rescue A69U  GGGTTAAAGGTTTATACCTTCCCAGGTAACAAACCAACCAACTTTCGATCTCTTGTAGATCT  GTTCTCTATACGAACTTTAAAATCTGTGTGGCTGTCACTCGGCTGCATGCTTAGTGCACTCA  CGCAGTATAATTAATAACTAATTACTGTCGTTCCCAGGACACGAGTAACTCGTCTATCTTCT  GCAGGCTGCTTACGGTTTCGTCCGTGTTGCAGCCGATCATCAGCACATCTAGGTTTCGTCC  GGGTGTGACCGAAAGGTAAGATGGAGAGCCTTGTCCCTGGTTTCAACGAGAAAACACACG  TCCAACTCAGTTTGCCTGTTTTACAGGTTCGCGACGTGCTCGTACGTGGCTTTGGAGACTC  CGTGGAGGAGGTCTTATCAGAGGCACGTCAACATCTTAAAGATGGCACTTGTGGCTTAGTA  GAAGTTGAAAAAGGCGTTTTGCCTCAACTTGAACAGCCCTATGTGTTCATCAAACGTTCGG  ATGCTCGAACTGCACCTCATGGTCATGTTATGGTTGAGCTGGTAGCAGAACTCGAAGGCAT  TCAGTACGGTCGTAGTGGTGAGACACTTGGTGTCCTTGTCCCTCATGTGGGCGAAATACCA  GTGGCTTACCGCAAGGTTCTTCTTCGTAAGAACGGTAATAAAGGAGCTGGTGGCCATAGTT  ACGGCGCCGATCTAAAGTCATTTGACTTAGGCGACGAGCTTGGCACTGATCCTTATGAAGA  TTTTCAAGAAAACTGGAACACTAAACATAGCAGTGGTGTTACCCGTGAACTCATGCGTGAG  CTTAACGGAGGGGCATACACTCGCTATGTCGATAACAACTTCTGTGGCCCTGATGGCTACC  CTCTTGAGTGCATTAAAGACCTTCTAGCACGTGCTGGTAAAGCTTCATGCACTTTGTCCGA ACAACTGGACTTTATTGACACTAAGAGGGGTGTATACTGCTGCCGTGAACATGAGCATGAA ATTGCTTGGTACACGGAACGTTCTGGGCCCTCGA GGGAGAGCGGCCGCCAGATCTTCCGGATGGCTCGAGTTTTTCAGCAAGATTGGCTGTAGT  TGTGATCAACTCCGCGAACCCATGCTTCAGTCAGCTGATGCACAATCGTTTTTAAACGGGT  TTGCGGTGTAAGTGCAGCCCGTCTTACACCGTGCGGCACAGGCACTAGTACTGATGTCGT  ATACAGGGCTTTTGACATCTACAATGATAAAGTAGCTGGTTTTGCTAAATTCCTAAAAACTA  ATTGTTGTCGCTTCCAAGAAAAGGACGAAGATGACAATTTAATTGATTCTTACTTTGTAGTTA  AGAGACACACTTTCTCTAACTACCAACATGAAGAAACAATTTATAATTTACTTAAGGATTGTC  CAGCTGTTGCTAAACATGACTTCTTTAAGTTTAGAATAGACGGTGACATGGTACCACATATA  TCACGTCAACGTCTTACTAAATACACAATGGCAGACCTCGTCTATGCTTTAAGGCATTTTGA  TGAAGGTAATTGTGACACATTAAAAGAAATACTTGTCACATACAATTGTTGTGATGATGATTA  TTTCAATAAAAAGGACTGGTATGATTTTGTAGAAAACCCAGATATATTACGCGTATACGCCA  ACTTAGGTGAACGTGTACGCCAAGCTTTGTTAAAAACAGTACAATTCTGTGATGCCATGCG  AAATGCTGGTATTGTTGGTGTACTGACATTAGATAATCAAGATCTCAATGGTAACTGGTATG  ATTTCGGTGATTTCATACAAACCACGCCAGGTAGTGGAGTTCCTGTTGTAGATTCTTATTAT  TCATTGTTAATGCCTATATTAACCTTGACCAGGGCTTTAACTGCAGAGTCACATGTTGACAC  TGACTTAACAAAGCCTTACATTAAGTGGGATTTGTTAAAATATGACTTCACGGAAGAGAGGT  TAAAACTCTTTGACCGTTATTTTAAATATTGGGATCAGACATACCACCCAAATTGTGTTAACT  GTTTGGATGACAGATGCATTCTGCATTGTGCAAACTTTAATGTTTTATTCTCTACAGTGTATC  TTT   Nucleocapsid RNA  GATTAAAGGTTTATACCTTCCCAGGTAACAAACCAACCAACTTTCGATCTCTTGTAGATCTG  TTCTCTAAACGAACAAACTAAAATGTCTGATAATGGACCCCAAAATCAGCGAAATGCACCCC  GCATTACGTTTGGTGGACCCTCAGATTCAACTGGCAGTAACCAGAATGGAGAACGCAGTG  GGGCGCGATCAAAACAACGTCGGCCCCAAGGTTTACCCAATAATACTGCGTCTTGGTTCAC  CGCTCTCACTCAACATGGCAAGGAAGACCTTAAATTCCCTCGAGGACAAGGCGTTCCAATT  AACACCAATAGCAGTCCAGATGACCAAATTGGCTACTACCGAAGAGCTACCAGACGAATTC  GTGGTGGTGACGGTAAAATGAAAGATCTCAGTCCAAGATGGTATTTCTACTACCTAGGAAC  TGGGCCAGAAGCTGGACTTCCCTATGGTGCTAACAAAGACGGCATCATATGGGTTGCAAC  TGAGGGAGCCTTGAATACACCAAAAGATCACATTGGCACCCGCAATCCTGCTAACAATGCT  GCAATCGTGCTACAACTTCCTCAAGGAACAACATTGCCAAAAGGCTTCTACGCAGAAGGGA  GCAGAGGCGGCAGTCAAGCCTCTTCTCGTTCCTCATCACGTAGTCGCAACAGTTCAAGAAA  TTCAACTCCAGGCAGCAGTAGGGGAACTTCTCCTGCTAGAATGGCTGGCAATGGCGGTGA  TGCTGCTCTTGCTTTGCTGCTGCTTGACAGATTGAACCAGCTTGAGAGCAAAATGTCTGGT  AAAGGCCAACAACAACAAGGCCAAACTGTCACTAAGAAATCTGCTGCTGAGGCTTCTAAGA  AGCCTCGGCAAAAACGTACTGCCACTAAAGCATACAATGTAACACAAGCTTTCGGCAGACG  TGGTCCAGAACAAACCCAAGGAAATTTTGGGGACCAGGAACTAATCAGACAAGGAACTGAT  TACAAACATTGGCCGCAAATTGCACAATTTGCCCCCAGCGCTTCAGCGTTCTTCGGAATGT (Langdon et al., 2018) . Orf1ab templates were synthesized (IDT) and cloned into pJet (ThermoFisher Scientific K1231) using blunt end cloning. Directionality and sequence were confirmed using Sanger sequencing (GENEWIZ). Plasmid were linearized using PCR (iProof Bio-Rad 1725310). 5 μl of PCR product was loaded onto an agarose gel to determine size and purity. If the PCR product was pure then the sample was PCR purified (QIAGEN 28106) if the band was impure, it was gel purified (QIAGEN 28706) (PCR impurity was most often was a problem for the ultrastructured mutants of principal site 2). 100 ng of gel or PCR purified DNA was used as a template for in vitro transcription (NEB E2040S) carried out according to the manufacturer's instructions with the addition of 0.1μl of Cy3 (Sigma PA53026) or Cy5 (Sigma PA55026) labeled UTP to each reaction. Following incubation at 37°C for 18 hours, in vitro transcription reactions were treated with DNAseI (NEB M0303L) according to the manufacturer's instructions. Following DNAse treatment, reactions were purified with 2.5M LiCl precipitation. Purified RNA amounts were quantified using nanodrop and verified for purity and size using a denaturing agarose gel and Riboruler RNA ladder (Thermo Scientific SM183).

Phase separation assays: For in vitro reconstitution LLPS experiments, 15 μl droplet buffer (20 mM Tris pH 7.5, 150 mM NaCl) was mixed with cy3 or cy5 labeled desired RNA and DEPC treated H20 (final volume 5 μl) and 5 μl protein in storage buffer was added at desired concentration. The mix was incubated in 384-well plates (Cellvis P384-1.5H-N) for 1-20 hours at 37°C unless indicated otherwise. Droplets formed after short incubations of 20 minutes or less, however, they were initially smaller and matured into larger droplets during the overnight incubation step. Time to maturation varied based on the ratio of RNA to protein, concentration of RNA and protein and RNA sequence. Multiple conditions per mutant were tested with the most optimal conditions for differences selected for comparison. Imaging of droplets was done on a spinning disc confocal microscope (Nikon CSU-W1) with VC Plan Apo 100X/1.49 NA oil (Cargille Lab 16241) immersion objective and an sCMOS 85% QE 95B camera (Photometrics). Data shown are representative of three or more independent replicates, across 2 or more RNA preparations. Whenever possible multiple mutations were designed to disrupt the same class of feature in multiple sequence contexts.

EMSA: 65ng/μl of the indicated RNA sequence was incubated with 0, 0.75, 1.5, or 2.2 µM Y109A mutant N-protein at 25 or 37°C for 1 hour in the following buffer 10mM HEPES pH 7.5, 50µM EDTA, 10% glycerol, 1mM DTT, 5mM MgCl2, 0.1mg/ml BSA, 2.5ug Yeast TRNA, 10U RNAse inhibitor and loading dye. Samples were then loaded onto an 8% TBE gel and run at 100V for 1 hour at 4°C. Gels were then stained with SYBRgold (S11494) and imaged. Unbound RNA was quantified using ImageJ.

Temperature dependent turbidity tests: The LCST behaviors of different phase separation systems were investigated on a Cary 300 temperature-dependent ultraviolet-visible spectroscopy equipped with a multicell thermoelectric temperature controller. The samples (4 μM of N-protein with 15 nM of RNAs) were mixed and prepared in a droplet buffer (20 mM Tris pH 7.5, 150 mM NaCl) at 4 °C. Before the initiation of the heating process of the turbidity test, for the experiments shown in Fig. 3A , the samples were incubated for 1 hour at 4 °C; for the experiments shown in Fig. 3B&C , the samples were incubated for 20 min at 4 °C. A heating rate of 1°C/min was applied during the temperature ramp while the absorbance at λ = 350 nm was recorded at every 0.33 °C increment. Normalized turbidity was calculated by the absorbance at the lowest temperature point normalizes to the absorbance at the highest temperature point.

Comparison of droplet images to absorbance A280 reading in diffuse phase. The mix was incubated in 384-well plates (Cellvis P384-1.5H-N) at 25, 30, or 37°C. Following imaging. 2 μl of diffuse phase solution (taken from the top of the well) was nanodropped and absorbance A280 was recorded. Error bars indicate the A280 measurement from the 3 technical replicates. (Of note, concentrations below 3µM N-protein did not give high enough A280 absorbance to generate reliable measurements.)

Phyre Structure prediction/ Pymol structure alignment: The following SARS-CoV-2 amino acid sequence was input into Phyre TKKSAAEASKKPRQKRTATKAYNVTQAFGRRGPEQTQGNFGDQELIRQGTDYKHWPQIAQFA PSASAFFGMSRIGMEVTPSGTWLTYTGAIKLDDKDPNFKDQVILLNKHIDAYKTFP. This sequence best matched with the crystal structure of the RBD2-dimerization domain of SARS-CoV-1 (Chen et al., 2007) . The resulting structure prediction was aligned to the crystal structure of SARS-CoV-1 or MERS-CoV (Nguyen et al., 2019) using Pymol.

Mass photometry of purified N-protein: Mass photometry was performed according to established protocols (Sonn-Segev et al., 2020) . 10µL of protein storage buffer (250mM NaCl 20mM phosphate buffer pH 7.5) was used to focus followed by addition of 10µl of 40nM N-protein in protein storage buffer (wildtype or RBD2-del) for a final protein concentration of 20nM. Representative histograms from were generated from 2 minutes movies reflective of the raw detected particle molecular weight in kDAs.

In vitro translation assay: Protocol was adapted from the method described by Tsang et al. Briefly, 25nM of 5′UTR nano luciferase fusion RNA was incubated with either protein 0.3µM or 3.2µM N-protein for 20-minutes at room temperature in PCR strip tubes (8ul total volume, final buffer conditions 140mM NaCl, 4mM phosphate buffer, 12mM TRIS pH 7.5) as a control for basal luciferase RNA translation, N-protein storage buffer was added (250mM NaCl 20mM phosphate buffer pH 7.5). Following incubation, 5ul rabbit reticulocyte lysate +Met +Leu (Promega L4960), was added to the protein/RNA mixture (or RNA and buffer) and the resulting mix was incubated at 30°C for 2 hours. 2µl of in vitro translation product was then mixed with 25µl of nano luciferase assay reagents (Promega N205A). Light production was measured on a luminometer. Data depicted represents at least three replicates. Of note, similar translational repression was observed when we incubated RNA under droplet permissive conditions in plates (37°C 1-2 hours) however this was much less reproducible likely due to the difference in RNA partitioning in the well post incubation with N-protein.

Cell Culture: HEK293T, and Vero-E6 cells were originally obtained from ATCC. All cell lines were maintained in DMEM (Corning 10-013-CV) supplemented with 10% Fetal Bovine Serum (Gibco). No antibiotics were used.

Plasmid Transfection: 24 hours prior to transfection, confluent cells were split 1:5. Two hours prior to transfection, 500µl of fresh media was added to 24 well plates. 500ng of plasmid DNA for each Nucleocapsid GFP Spark (Sino biological VG40588-ACGLN) and the MSCV blast 1-1000 fragments was co-transfected using FUGENE HD. Transfections were then incubated for 24-48 hours prior to imaging.

Cell Imaging: Cells were imaged using a 40X air objective on a spinning disk confocal microscope (Nikon Ti-Eclipse, Yokogawa CSU-X1 spinning disk). Images were taken with a ANDOR camera. Representative cells are taken from 3 biological replicates.

Cell Imaging Quantification: Cells with puncta were cropped using FIJI. Experimenters were then blinded to conditions, and puncta were counted for each cell. Whole cell N:GFP signal was quantified using ImageTank (O'Shaughnessy et al., 2019) .

Transmission electron microscopy (TEM) and quantification of RNP size distribution: For negative stained TEM images used to quantify the assemblies of RNP size distribution, 5ul of 20µM protein in 250mM NaCl, 20mM phosphate buffer pH 7.5 and 5ul of 50nM RNA (FS RNA

After 2 min absorption to the carbon film, the solution was blotted and washed with 8µl of water for 10 s, blotted, stained with 8µl of 2 % uranyl acetate for 10 s, blotted, and dried. Negative stained TEM images were obtained on a FEI Morgagni microscope. Images were analyzed with ImageJ software

After confirmation of phase separation by imaging mixtures were immediately subjected to RNP-MaP treatment as described (Weidmann et al., 2021), with modifications described below. Briefly, 200 μl of mixtures were added to 10.5 μl of 200 mM SDA (in DMSO) in wells of a 6-well plate and

HCl (pH 8.0) and incubated at 37°C for 10 minutes, heated to 95°C for 5 minutes, cooled on ice for 2 minutes, and warmed to 37°C for 2 minutes. Proteinase K was then added to 0.5 mg/ml and incubated for 1 hour at 37°C, followed by 1 hour at 55°C

Briefly, 7 μl of purified modified RNA was mixed with 200 ng of random 9-mer primers and 20 nmol of dNTPs and incubated at 65°C for 10 min followed by 4°C for 2 min. 9 μl 2.22 ́ MaP buffer [1 ́ MaP buffer consists of 6 mM MnCl2, 1 M betaine, 50 mM Tris (pH 8.0), 75 mM KCl, 10 mM DTT] was added and the combined solution was incubated at 23°C for 2 min. 1 μl Superscript II Reverse Transcriptase (200 units, Invitrogen) was added and the reverse transcription (RT) reaction was performed according to the following temperature program: 25°C for 10 min, 42°C for 90 min

Briefly, purified cDNA was added to an NEBNext second-strand synthesis reaction (NEB) at 16°C for 150 minutes. dsDNA products were purified and size-selected with SPRI beads at a 0.8 ́ ratio. Nextera XT (Illumina) was used to construct libraries according to the manufacturer's protocol, followed by purification and size-selection with SPRI beads at a 0.65 ́ ratio. Library size distributions and purities were verified (2100 Bioanalyzer, Agilent) and sequenced using 2x300 paired-end sequencing

for read alignment, mutation counting, and SHAPE reactivity profile generation. The --random-primer-len 9 option was used to mask RT primer sites with all other values set to defaults. For RNP-MaP library analysis, the protein:RNA mixture samples are passed as the --modified samples and no-protein control RNA samples as --unmodified samples. Median read depths of all SHAPE-MaP and RNP-MaP samples and controls were greater than 50

Secondary structure modeling: Secondary structure models were taken from our previous publication

2021) was used to calculate RNP-MaP "reactivity" profiles from the Shapemapper 2 "profile.txt" output. RNP-MaP "reactivity" is defined as the relative MaP mutation rate increase of the crosslinked protein-RNA sample as compared to the uncrosslinked (no protein control) sample. Nucleotides whose reactivities exceed reactivity thresholds are defined as "RNP-MaP sites". RNP-MaP site densities were calculated over centered sliding 15-nt windows to identify RNA regions bound by N-protein. An RNP-MaP site density threshold of 5 sites per 15-nt window was used to identify "N-protein binding sites

Dynamic light scattering: Dynamic light scattering (DLS) measurements were performed at 25 °C using a Wyatt DynaPro temperature-controlled Plate Reader (Wyatt Technology

Genome N-protein motif analysis: YYAAAY motifs were counted throughout the NC_045512.2 reference genome (with overlapping motifs counted separately) and the motif counts in each 1000 base pair window were plotted as a histogram. The density of doublestranded RNA was plotted using a kernel density estimation plot with smoothing parameter set to 100

All data are available upon request from C

Protein Phase Separation: A New Phase in Cell Biology

Deciphering how naturally occurring sequence features impact the phase behaviors of disordered prion-like domains

Liquid-liquid phase separation by intrinsically disordered protein regions of viruses: Roles in viral life cycle and control of virushost interactions

Accurate detection of chemical modifications in RNA by mutational profiling (MaP) with ShapeMapper 2

Phosphoregulation of Phase Separation by the SARS-CoV-2 N Protein Suggests a Biophysical Basis for its Dual Functions

The SARS coronavirus nucleocapsid protein -Forms and functions

Structure of the SARS Coronavirus Nucleocapsid Protein RNA-binding Dimerization Domain Suggests a Mechanism for Helical Packaging of Viral RNA

Liquid-liquid phase separation by SARS-CoV-2 nucleocapsid protein and RNA

LASSI: A lattice model for simulating phase transitions of multivalent proteins

Physical Principles Underlying the Complex Biology of Intracellular Phase Transitions

The SARS-CoV-2 nucleocapsid protein is dynamic, disordered, and phase separates with RNA

The coding capacity of SARS-CoV-2

PTM of CoVs

Coronavirus N Protein N-Terminal Domain (NTD) Specifically Binds the Transcriptional Regulatory Sequence (TRS) and Melts TRS-cTRS RNA Duplexes

An Optimal cis-Replication Stem-Loop IV in the 5' Untranslated Region of the Mouse Coronavirus Genome Extends 16 Nucleotides into Open Reading Frame 1

Measles virus nucleo-And phosphoproteins form liquid-like phaseseparated compartments that promote nucleocapsid assembly

Phase transitions drive the formation of vesicular stomatitis virus replication compartments

Comprehensive in vivo secondary structure of the SARS-CoV-2 genome reveals novel regulatory motifs and mechanisms

Liquid-liquid phase separation in biology

Genomic RNA Elements Drive Phase Separation of the SARS-CoV-2 Nucleocapsid

Putative cis-Acting Stem-Loops in the 5′ Untranslated Region of the Severe Acute Respiratory Syndrome Coronavirus Can Substitute for Their Mouse Hepatitis Virus Counterparts

Stem-loop 1 in the 5′ UTR of the SARS coronavirus can substitute for its counterpart in mouse hepatitis virus

Crystal structure of SARS-CoV-2 nucleocapsid protein RNA binding domain reveals potential unique drug targeting sites

The Architecture of SARS-CoV-2 Transcriptome

Phospho-dependent phase separation of FMRP and CAPRIN1 recapitulates regulation of translation and deadenylation. Science (80-. )

SARS-CoV-2 structure and replication characterized by in situ cryo-electron tomography

Insights into the secondary structural ensembles of the full SARS-CoV-2 RNA genome in infected cells

Structural Lability in Stem-Loop 1 Drives a 5′ UTR-3′ UTR Interaction in Coronavirus Replication

The SARS-CoV-2 nucleocapsid phosphoprotein forms mutually exclusive condensates with RNA and the membrane-associated M protein

In vivo reconstitution finds multivalent RNA-RNA interactions as drivers of mesh-like condensates

Valence and patterning of aromatic residues determine the phase behavior of prion-like domains

The coronavirus nucleocapsid is a multifunctional protein

Structure and oligomerization state of the C-terminal region of the Middle East respiratory syndrome coronavirus nucleoprotein

Negri bodies are viral factories with properties of liquid organelles

Phase Transition of a Disordered Nuage Protein Generates Environmentally Responsive Membraneless Organelles

Software for lattice light-sheet imaging of FRET biosensors, illustrated with a new Rap1 biosensor

SARS-CoV-2 nucleocapsid protein phase-separates with RNA and with human hnRNPs

Stem-Loop IV in the 5′ Untranslated Region Is a cis-Acting Element in Bovine Coronavirus Defective Interfering RNA Replication

Functional organization of cytoplasmic inclusion bodies in cells infected by respiratory syncytial virus

RNA contributions to the form and function of biomolecular condensates

Thermoreversible gelation in solutions of associating polymers. 2. Linear dynamics

Nucleocapsid protein of SARS-CoV-2 phase separates into RNA-rich polymerase-containing condensates

Role of spatial patterning of N-protein interactions in SARS-CoV-2 genome packaging

Thermoreversible gelation in solutions of associative polymers. 1. Statics. Macromolecules

Selective 2'-hydroxyl acylation analyzed by primer extension and mutational profiling (SHAPE-MaP) for direct, versatile and accurate RNA structure analysis

Quantifying the heterogeneity of macromolecular machines by mass photometry

In vivo structural characterization of the SARS-CoV-2 RNA genome identifies host proteins vulnerable to repurposed drugs

The viral protein NSP1 acts as a ribosome gatekeeper for shutting down host translation and fostering SARS-CoV-2 translation

Phosphoregulated FMRP phase separation models activitydependent translation through bidirectional control of mRNA granule formation

Pi-Pi contacts are an overlooked protein feature relevant to phase separation

A Molecular Grammar Governing the Driving Forces for Phase Separation of Prion-like RNA Binding Proteins

SARS-CoV-2 nucleocapsid protein undergoes liquid-liquid phase separation into stress granules through its N-terminal intrinsically disordered region

Analysis of RNA-protein networks with RNP-MaP defines functional hubs on RNA

Glycogen synthase kinase-3 regulates the phosphorylation of severe acute respiratory syndrome coronavirus mucleocapsid protein and viral replication

The structure and functions of coronavirus genomic 3' and 5' ends

Mouse Hepatitis Virus Stem-Loop

Functions as a Spacer Element Required To Drive Subgenomic RNA Synthesis

Molecular Architecture of the SARS-CoV-2 Virus

Nonstructural Protein 1 of SARS-CoV-2 Is a Potent Pathogenicity Factor Redirecting Host Protein Synthesis Machinery toward Viral RNA

Biochemical characterization of SARS-CoV-2 nucleocapsid protein

RNA Controls PolyQ Protein Phase Transitions

Energetic and structural features of SARS-CoV-2 N-protein co-assemblies with nucleic acids

GCG inhibits SARS-CoV-2 replication by disrupting the liquid phase condensation of its nucleocapsid protein

High-resolution structure and biophysical characterization of the nucleocapsid phosphoprotein dimerization domain from the Covid-19 severe acute respiratory syndrome coronavirus 2

The Short-and Long-Range RNA-RNA Interactome of SARS-CoV-2

We thank Rick Young, Phil Sharp, Alex Holehouse, Kathleen Hall, Andrea Sorrano, Ahmet Yildez and their lab members for sharing data and discussions. Alain Laederach for initial discussion on genomic sequence, Chase Weidmann for initial discussions planning