key: cord-0292241-eu7cggrz authors: Caruso, Icaro Putinhon; dos Santos Almeida, Vitor; Juliani do Amaral, Mariana; de Andrade, Guilherme Caldas; de Araújo, Gabriela Rocha; de Araújo, Talita Stelling; de Azevedo, Jéssica Moreira; Barbosa, Glauce Moreno; Bartkevihi, Leonardo; Bezerra, Peter Reis; dos Santos Cabral, Katia Maria; Lourenço, Isabella Otênio; Malizia-Motta, Clara L. F.; de Luna Marques, Aline; Mebus-Antunes, Nathane Cunha; Neves-Martins, Thais Cristtina; de Sá, Jéssica Maróstica; Sanches, Karoline; Santana-Silva, Marcos Caique; Vasconcelos, Ariana Azevedo; da Silva Almeida, Marcius; de Amorim, Gisele Cardoso; Anobom, Cristiane Dinis; Da Poian, Andrea T.; Gomes-Neto, Francisco; Pinheiro, Anderson S.; Almeida, Fabio C. L. title: Structure insights, thermodynamic profiles, dsDNA melting activity, and liquid-liquid phase separation of the SARS-CoV-2 nucleocapsid N-terminal domain binding to DNA date: 2021-07-22 journal: bioRxiv DOI: 10.1101/2021.07.21.453232 sha: ab0db60b4f94cf41b8fe4866417f6e8336dd5c7f doc_id: 292241 cord_uid: eu7cggrz The SARS-CoV-2 nucleocapsid protein (N) is a multifunctional promiscuous nucleic acid-binding protein, which plays a major role in nucleocapsid assembly and discontinuous RNA transcription, facilitating the template switch of transcriptional regulatory sequences (TRS). Here, we dissect the structural features of the N protein N-terminal domain (N-NTD), either with or without the SR-rich motif (SR), upon binding to single and double-stranded TRS DNA, as well as their activities for dsTRS melting and TRS-induced liquid-liquid phase separation (LLPS). Our study gives insights on specificity for N-NTD/N-NTD-SR interaction with TRS, including an unfavorable energetic contribution to binding along with hydrogen bonds between the triple-thymidine (TTT) motif in the dsTRS and β-sheet II due to the defined position and orientation of the DNA duplex, a well-defined pattern (ΔH > 0 and ΔS > 0 for ssTRS, and ΔH < 0 and ΔS < 0 for dsTRS) for the thermodynamic profile of binding, and a preference for TRS in the formation of liquid condensates when compared to a non-specific sequence. Moreover, our results on DNA binding may serve as a starting point for the design of inhibitors, including aptamers, against N, a possible therapeutic target essential for the virus infectivity. grown at 37 °C until optical density ~0.7 at 600 nm. Protein expression was induced with 0.2 mM IPTG for 16 h at 16 °C. For the production of 15 N and 15 N/ 13 C-labeled protein, expression was induced in M9 minimal medium containing either 15 Simultaneously, the constructs were cleaved with TEV protease (1:30 TEV:protein molar ratio) to remove the His 6 tag. After dialysis, samples were reapplied to a HisTrap FF column, using the same purification buffers. Fractions containing the proteins of interest were pooled, dialyzed against buffer containing 20 mM sodium phosphate (pH 6.5) and 50 mM NaCl, and concentrated. Protein concentration was determined spectrophotometrically using the molar extinction coefficient of 26,930 M -1 ·cm -1 at 280 nm. In this study, we used a 11-nucleotide DNA sequence corresponding to a 7nucleotide conserved sequence of the positive-sense ssTRS followed by a CGCG generic segment (ssTRS(+), 5'-TCTAAACCGCG-3'). This CGCG sequence increased the melting temperature of the corresponding TRS duplex and thus enabled us to work at room temperature. The negative-sense ssTRS contained the complementary strand (ssTRS(-), 5'-CGCGGTTTAGA-3'). As a control, we used the single and doublestranded non-specific (NS) oligonucleotides containing the 7-nucleotide DNA sequences used by Dinesh and cols. (2020) (15) followed by the CGCG generic segment (ssNS(+), 5'-CACTGACCGCG-3'; and ssNS(-), 5'-CGCGGTCAGTG-3'). All DNA oligonucleotides were purchased from GenScript (Piscataway, USA). Equimolar amounts of positive and negative sense ssDNAs were dissolved in buffer and annealed by heating to 50 °C for 10 min and slowly cooling to room temperature. The concentrations of ssTRS(+), ssTRS(-), ssNS(+), and ssNS(-) were determined spectrophotometrically using the molar extinction coefficients of 103000, 107500, 99100, and 103000 M -1 ·cm -1 at 260 nm, respectively. Protein intrinsic fluorescence quenching measurements were performed using a PC1 steady-state spectrofluorimeter (ISS, Champaign, IL, USA) equipped with a quartz cell of 1.0 cm optical path length and a Neslab RTE-221 thermostat bath. Excitation and emission bandwidths were set to 1.0 and 2.0 mm, respectively. To solely excite tryptophan, the excitation wavelength (λ ex ) was set to 295 nm. The emission spectrum the accessibility of the fluorophores (W52, W108, and W132) by the quencher (DNAs). Two of the three fluorophores (W52 and W108) are located at the putative nucleic acidbinding site in N-NTD, while the third (W132) is oppositely oriented to this binding site and thus presents a non-specific linear contribution for the fluorescence titration curves. To remove the non-specific contribution, F 0 /F versus [L T ] curves for each protein:DNA titration (duplicate or triplicate) was fitted using Equation (1) plus a linear term (constant·[L T ]). After that, F 0 /F linear contributions were calculated and removed manually ( Figure S1 ). Fluorescence measurements were recorded in 20 mM Bis-Tris buffer (pH 6.5) in the absence or presence of 100 mM NaCl, and in 20 mM sodium phosphate buffer (pH 6.5) containing 50 and 100 mM NaCl. The FRET efficiency was calculated from 1 0 Chemical shift perturbations (CSP) were monitored by a series of 2D [ 1 H, 15 N] HSQC spectra, recorded on a Bruker 600 MHz spectrometer at 20 °C, after the addition of increasing concentrations of ssDNAs (TRS and NS). N-NTD and N-NTD-SR were dissolved in 20 mM sodium phosphate buffer (pH 6.5), 50 mM NaCl, 500 µM PMSF, 3 mM sodium azide, 3 mM EDTA, and 5.5% (v/v) of D 2 O, at a final concentration of 70 µM. The ssDNAs were titrated into N-NTD and N-NTD-SR for the following ssDNA:protein molar ratios: 0.14, 0. 34 (36) . Periodic boundary conditions were used, and all simulations were performed in NPT ensemble, keeping the system at 25 °C and 1.0 bar using Nose-Hoover thermostat (τ T = 2 ps) and Parrinello-Rahman barostat (τ P = 2 ps and compressibility = 4.5×10 -5 ·bar -1 ). A cutoff of 12 Å for both Lennard-Jones and Coulomb potentials was used. The long-range electrostatic interactions were calculated using the particle mesh Ewald (PME) algorithm. In every MD simulation, a time step of 2.0 fs was used and all covalent bonds involving hydrogen atoms were constrained to their equilibrium distance. A conjugate gradient minimization algorithm was used to relax the superposition of atoms generated in the box construction process. Energy minimizations were carried out with the steepest descent integrator and conjugate gradient algorithm, using 1,000 kJ·mol -1 ·nm -1 as maximum force criterion. One hundred thousand steps of molecular dynamics were performed for each NVT and NPT equilibration, applying force constants of 1,000 kJ·mol −1 ·nm −2 to all heavy atoms of N-NTD:DNA complexes. At the end of preparation, 2 μs MD simulations of each molecular system (N-NTD bound to ssTRSs, dsTRS, ssNS, and dsNS) were carried out for data acquisition. Following dynamics, the trajectories of each molecular system were firstly concatenated and analyzed according to the root mean square deviation (RMSD) for the backbone atoms of protein and DNA, number of contacts for distances lower than 0.6 nm between pairs of atoms of N-NTD and DNA, and number of protein-DNA hydrogen bonds with cutoff distance (heavy atoms) of 3.5 Å and maximum angle of 30°. The percentages of protein-DNA hydrogen bond persistence were obtained from plot_hbmap_generic.pl script (37) . The number of protein-DNA hydrogen bonds with persistence greater than 10% was counted with respect to amino acid and nucleotide residues for each molecular system. The contributions of N-NTD residues to the DNA- PCA scatter plots were generated, and conformational motions were filtered from the eigenvectors of the first and second principal components (PC1 and PC2, respectively). The conformational space was quantified by fitting an elliptical shell with 95% (confidence) of the density for each scatter plot and making its extent proportional to the area (S el ) of this shell. The structural representations of the constructed models were displayed using PyMOL (42) . Cleaning of microscopy glasses followed by coating of Knittel coverslips with 0.5% (w/v) bovine serum albumin (BSA, Sigma-Aldrich A2153) was carried out to System, Thermo Fisher Scientific, USA) with a 40× apo objective. To quantify the number of condensates, five areas of 100 µm 2 were analyzed for each sample followed by Fourier bandpass filtering. Next, images had their threshold adjusted for liquid droplets correct recognition, followed by mask creation and the "fill holes" (fills droplets that were poorly delimited) and "watershed" operations (separate droplets undergoing fusion). The condensates number (above 0.5 µm condensates were counted) and size (area in µm 2 ) were determined using the "analyze particles" plugin. All quantification and image processing steps were performed in Fiji. The DAPI fluorescence images were obtained with 1% excitation intensity, apart from the crystal sample formed by 1:1 N-NTD-SR:dsNS which was excited with 20% laser power. We first measured the dsDNA melting activity of N-NTD, as well as N-NTD containing the SR-rich motif (N-NTD-SR), using fluorescence resonance energy transfer (FRET) from the 5' Q570-labeled positive-sense TRS (Q570-ssTRS(+)) to the 3' Q670-labeled negative-sense TRS (Q670-ssTRS(-)) ( Figure 1A ). The DNA was used as a N-NTD/N-NTD-SR ligand due to its higher stability over RNA, while maintaining a relative structural similarity. As it will be further described, the use of DNA is supported by the chemical shift profiles observed for dsDNA (either TRS or NS), which are largely similar to those obtained with dsRNA (9) . The study with DNA is complementary to the available data on the N protein interaction with RNA (7, 15, 45) . In the absence of N-NTD, the FRET pair Q570/Q670 is in proximity, displaying FRET efficiency for dsTRS. Addition of N-NTD or N-NTD-SR results in an increase in fluorescence emission at 570 nm with a concomitant decrease at 670 nm, leading to a decay in FRET efficiency, indicating DNA duplex melting ( Figure 1 ). N-NTD and N-NTD-SR shows similar dsTRS melting activity. We also measured N-NTD-SR melting activity toward dsTRS RNA, which presented similar activity ( Figure S2 ). We and salt concentration (0, 50, or 100 mM NaCl) ( Figure S2 ). The protein:DNA binding affinities are at the nanomolar range (Table S7 ). The affinity for ssTRS(-) is higher than that for ssTRS(+), being 4-5-fold for N-NTD and 15-16-fold for N-NTD-SR. This tendency was also observed for the NS DNA, albeit with a smaller magnitude. This observation is remarkable since ssTRS(-) is the strand transferred to TRS-L during template switch (4). The thermodynamic parameters are similar for both N-NTD and N-NTD-SR binding to TRS. For ssTRSs, the interaction is enthalpically unfavorable (ΔH > 0) and entropically driven (ΔS > 0), while for dsTRS, it is enthalpically driven (ΔH < 0) and entropically unfavorable (ΔS < 0). The analysis of these enthalpic and entropic contributions reveals that hydrophobic contacts are important for N-NTD/N-NTD-SR binding to ssTRSs, while hydrogen bonds (and salt bridges) and van der Waals interactions play a major role in stabilizing the interaction with dsTRS (46) . Based on this observation, we hypothesize that the destabilization of dsTRS Watson-Crick hydrogen bonds induced by N-NTD or N-NTD-SR binding would be entropically driven. It would also be enthalpically unfavorable, which might contribute to further dissociation of ssTRSs. In general, we observed a higher affinity for TRS than for NS in the interaction with N-NTD and N-NTD-SR, except for ssTRS(+) ( Table S7 ). For N-NTD-SR, the affinity is even higher, especially for ssTRS(-), suggesting that the arginine residues in the SR-rich motif may contribute to binding with non-specific electrostatic interactions. Unlike the thermodynamic pattern observed for N-NTD/N-NTD-SR binding to the specific sequences (TRSs), we did not observe a regular pattern for the enthalpic and entropic contributions to the interaction with NS DNAs ( Figure 2D and Table S1 ). Although a thermodynamic profile similar to that of TRS was observed for N-NTD binding to NS DNAs, a different energetic profile was observed for N-NTD-SR. To map the residues involved in N-NTD and N-NTD-SR interaction with DNA, single and double-stranded DNA oligonucleotides were titrated into 15 N-labeled protein samples and amide chemical shifts were measured from 2D [ 1 H, 15 N] HSQC spectra ( Figure S4 ). Residues displaying statistically significant chemical shift perturbation (CSP) values (higher than the average plus one standard deviation (SD)) are located in well-defined regions (Figure 3 and S5). Using the experimental CSP data, we modeled the interaction of N-NTD with the DNAs (Figures 3G-3I ). The structural models were fundamental to discriminate the interacting amino acid residues from the ones that seem to undergo allosteric effects upon DNA binding. The binding site includes: (i) the flexible N-terminal region (N47, N48, T49), (ii) the palm formed by β-sheet I (β2/β3/β4) (Y86, A89, R107, W108, Y109, Y111, I131, W132) and the N-terminal 1 8 residues (A50, S51, F53, T54, A55), (iii) the finger (β2/β3 loop) (A90, T91, R92, R93, I94, R95, G96, K100, K102, D103 , L104, S105, R107), (iv) the β-sheet II (β1/β5) (L56, T57, Q58, H59, G170, Y172, A173, E1 74), (v) the thumb (α2/β5 loop) (R149, N150, A152, A156, I157), and (vi) the C-terminal region (G175, S176, R177). We also observed significant CSPs for residues L64, K65, F66, G164, T165, T166, and L167, which are located in a remote region from the principal binding site ( Figure 3G and 3H), suggesting that these residues form either a secondary binding site, interacting with DNA directly, or an allosteric site, undergoing indirect conformational changes due to DNA binding. In addition, residues at the SR motif (N192, S193, N196, and S197) displayed significant CSPs, suggesting that this region engages in DNA binding. The observed CSPs are strikingly similar (even for the remote region) to those observed by To find a consensus amino acid sequence for the interaction of N-NTD and N-NTD-SR with ssDNAs and dsDNAs, we aligned the residues with CSPs larger than the average plus one SD for all titrations ( Figure S6 ). To identify which residues are common to all interactions and which are unique, we grouped the different titration results and plotted them as an intersecting set of residues represented by circles containing the CSP information for each titration (Figure 4 ). When we analyzed the set of intersections for N-NTD and N-NTD-SR with single and duplex TRS, as well as with the non-specific (NS) sequences ( Figure 4A , 4B, 4C, and 4D), the following residues stood out: (i) A152 in the thumb is present in all 4 intersections; (ii) Y111 in the interface between the palm and the β-sheet II, and T166 in the remote region from the main binding site are present in 3 intersections; (iii) N47 in the N-terminus, K65 in the remote region, R95 and K102 in the finger, A156 in the thumb, and Y172 and A173 in the β-sheet II are present in 2 intersections. Note that N47 seems to be unique for NS, while K65 for TRS. The regions mapped by the residues at the intersections involve the finger, the two elements of the palm, and the flexible thumb of N-NTD ( Figure 4E ). These residues may be key for N-NTD binding to several nucleic acids, independent of the sequence specificity. Furthermore, they comprise not only positively charged residues, responsible for electrostatic interactions, but also hydrophobic residues (Y111, A156, T166, Y172, and A173). We used the CSP as a function of DNA concentration to estimate the K d for the interaction with N-NTD (Table S8 ). In contrast to the nanomolar affinities measured in the absence of NaCl and inorganic phosphate (Table S7) , we observed apparent dissociation constants in the order of micromolar, and yet, despite the difference in affinities the protein is active in all tested condition ( Figure S2 ). The MD trajectories may report, not all, but important interactions related to the binding stability and specificity. Thus, we determined protein-DNA hydrogen bonds from the 2 μs MD trajectories and calculated their percentages of persistence along the 1 simulations ( Figure S8 ). Protein-DNA hydrogen bonds with persistence higher that 10% were counted with respect to amino acid and nucleotide residues ( Figure 5 , Table S1 -S6). It is noteworthy that protein-DNA hydrogen bonds also report the presence of salt bridges between arginine or lysine residues and DNA phosphate groups ( Figure 5G The positive-sense strands of dsDNA (middle, Figure 5A and 5B) and ssDNA ( Figure 5C and 5D) are mainly hydrogen bonded to N-NTD through their 5' termini (nucleotide 1 to 4). Conversely, the negative-sense strands of dsDNA are mainly recognized by N-NTD through hydrogen bonds with their 3' termini (nucleotides 4 to 9) (bottom, Figure 5A and 5B), which include the specific TTT motif for TRS. It is interesting to note that binding of N-NTD to positions 1 to 5 of the positive-sense strand and 4 to 9 of the negative-sense strand is maintained independently of the nucleotide sequence (TRS or NS), suggesting that the orientation of dsDNA with respect to the positively charged cleft is conserved and sequence independent. We observed consistently that the total number of protein-DNA persistent hydrogen bonds with ssDNA(-) ( Figure 5E and 5F) is significantly larger than that with ssDNA(+) ( Figure 5C and 5D). These results are in agreement with the observation that ssDNA(-) shows higher affinity to N-NTD and N-NTD-SR than ssDNA(+) ( Table S7 ). The same tendency is observed for the negative-sense strands of dsDNAs ( Figure 5A and 5B). Due to the high content of charged residues and the above-mentioned electrostatic contribution (salt and phosphate dependence, Figure 2B and 2C) to DNA recognition, we decided to compute the theoretical Gibbs free energy of binding ‫ܩ∆(‬ ௧ ) using Poisson-Boltzmann Surface Area (PBSA). This method enables to evaluate and discriminate the main protein residues and nucleotides that contribute to ∆ ‫ܩ‬ ௧ . The most significant contributions come from the charged residues ( Figure 6A to 6F) distributed throughout the protein ( Figure 6G and 6H). Positively charged arginine and lysine residues contribute favorably, while negatively charged aspartate and glutamate residues contribute unfavorably. It is important to consider the unfavorable contributions because they may be responsible for the DNA/RNA duplex melting activity of N-NTD. It is also important to analyze the contributions to ∆ ‫ܩ‬ ௧ coming from the nucleic acid. For dsDNA, the contribution is mostly favorable for the positive-sense strand. For the negative-sense strand, the contribution is favorable at the 5' and 3' termini, but for positions 4 to 8 it varies according to the nucleotide sequence, being less favorable or near-zero for dsNS and unfavorable for dsTRS. For dsTRS, the stretch of nucleotides at positions 4 to 8, which contains the specific TTT motif, is hydrogen bonded to the β-sheet II (β1/β5) ( Figure 5A and 5G). The same is not observed for dsNS, because it is tilted ~25° away from β-sheet II (β1/β5) ( Figure 3I ). In this protein region, we observed charged residues that contribute unfavorably to ∆ ‫ܩ‬ ௧ , such as E174, which was identified in our previous study with dsRNA (14). For ssDNAs, both positive and negative-sense strands contributed similarly to ∆ ‫ܩ‬ ௧ , being slightly more favorable for the positive-sense strands. Interacting as a duplex, the positive-sense strands (TRS and NS) showed favorable contribution all along its sequence ( Figure 6A and 6B); however, when interacting as a ssDNA, the contribution is less favorable ( Figure 6C and 6D ). For the negative-sense strands (TRS and NS), the nucleotides at positions 4 to 8 contribute unfavorably when interacting as a duplex DNA ( Figure 6A and 6B) , while favorably as a single strand ( Figure 6E and 6F ). This is remarkable and agrees with the hypothesis that these unfavorable contributions (TTT motif and β-sheet II) may play a key role in N-NTD melting activity. The contrasting behavior for each strand in the single or double-stranded DNA may be attributed to the fact that the dsDNA has a better-defined positioning when compared to the ssDNA, due to the presence of secondary structure in the duplex. The difference of ~25° in orientation of dsTRS and dsNS ( Figure 3I ) may serve as an indicative of specificity of the TRS, contributing to the positioning β-sheet II, E174 for instance, close to TTT motif. This difference in positioning seems to be more relevant for the specificity than the difference in binding affinities between dsTRS and dsNS. We investigated the ability of N-NTD and N-NTD-SR to form biomolecular condensates by liquid-liquid phase separation (LLPS) as a consequence of their interaction with specific or non-specific nucleic acids. Indeed, full-length N has been shown to form liquid droplets induced by RNA interaction (20) (21) (22) (23) . Interestingly, N from several coronaviruses is predicted to undergo LLPS by the catGRANULE algorithm ( Figure S11A Figure S13A and S13C). Specifically, an excess of RNA (1:2 N-NTD-SR:RNA) resulted in 2 5 305 ± 7 condensates in a 100 µm 2 area (top, Figure S13B ). This number was only 1.5 times higher than that observed at critical concentration for RNA-driven N-NTD-SR LLPS, i.e. at the 2:1 N-NTD-SR:RNA stoichiometry. The 1:2 N-NTD-SR:RNA ratio showed the highest dispersion on condensates size, consonant to large condensates formed by fusion ( Figure S13B , bottom graph). Since nucleic acid structure and sequence can finely tune the phase behavior, we followed this phenomenon in the presence of ssTRS and dsTRS DNAs (Figure 7 ). dsTRS ( Figure 7A and Figure 7D Interestingly, in buffer containing 20 mM sodium acetate at pH 5.5, droplet formation was higher, and condensates had marked circular morphology ( Figure S14 ). Specifically, the droplet number increased by 32% compared to pH 7.5 (203.2 ± 8.2 droplets at pH 7.5 versus 298.6 ± 16.8 droplets at pH 5.5), suggesting that acidic pH induces nucleic acid-driven N-NTD-SR LLPS. To investigate the role of sequence specificity in promoting the N-NTD-SR LLPS, we followed the condensation process using the NS DNAs ( Figure S15 ). In agreement with dsTRS, the relevant condition for LLPS was at 1:1 protein:DNA stoichiometry. In addition, excess of dsNS dissolved the condensates (127.2 ± 13.9 condensates/µm 2 at 1:1 versus 18.4 ± 3.6 condensates/µm 2 at 1:2). This behavior was confirmed by turbidity measurements ( Figure S15E ). However, condensates were not homogeneously spherical, as the ones formed with specific dsTRS, and most of them wetted the coverslip surface. Curiously, when observing the entire cover of the glass slide from 1:1 N-NTD-SR:dsNS samples, we observed a few crystals. Since all images were obtained after 30 minutes incubation, we sought to understand whether crystal formation would be enhanced with prolonged incubation. After 2 hours of incubation, we observed crystals presenting DAPI staining ( Figure S15D , inset). Condensates are supersaturated and estimated to be 10 to 300 times enriched in macromolecules compared to the diffuse phase (54), thus crystallization is thermodynamically favored. The buffer with addition of 10% PEG-4000 did not show any artifact ( Figure S12B The nucleocapsid N protein is well characterized for its interaction with RNA, which is essential to understand two important biological processes for the viral cycle: (i) the assembly of the helical ribonucleoprotein (RNP) complex, and (ii) the discontinuous transcriptional mechanism. N's dsRNA melting activity enables template switch during discontinuous transcription (7, 10) . In addition, its ability to phase separate creates a membraneless compartment (liquid-like condensates) that regulates transcription and replication. Previous studies have shown that N binds to single and double-stranded DNA as RNA mimetics (55, 56) . Takeda Structural and binding studies revealed that solvent-exposed charged residues and electrostatic interactions are the main driving forces for N:nucleic acid complex formation (10, 47, 55, 57, 58) . In line with that, we observed that formation of the N-NTD:DNA complex is NaCl and inorganic phosphate dependent ( Figure 2B and 2C ). In addition, dissociation constant increases from nanomolar in 20 mM Bis-Tris buffer (no salt and pH 6.5, Table S7 ) to the micromolar range in 20 mM sodium phosphate buffer (pH 6.5, Table S8 ) containing 50 mM NaCl, while maintaining melting activity ( Figure S2 ). This is in agreements with kinetic simulations of the melting activity that suggests the activity does not depends on the binding affinity so long K d < 10 -1 M -1 (14). It is worth mentioning that the low micromolar K d values are similar to those described for . Accordingly, Dinesh and cols. (2020) showed that R92E and R107E mutations, located at the binding cleft, lead to a decrease in N-NTD-RNA binding affinity, while E174R promotes an increase (15) . We also observed from PBSA analysis that short and long-range contributions of charged residues to ∆ ‫ܩ‬ ௧ play a key role in protein-DNA binding affinity and perhaps in dsDNA melting activity. We provide the first evidence that N-NTD and N-NTD-SR are active upon binding to DNA. Both constructs show dsDNA melting activity, and N-NTD-SR forms liquid condensates in the presence of DNA in a crowded physiological buffer. It is noteworthy that these activities were reported for RNA only (7, 10, 20, 21, 60) , and that Unlike homeodomains found in eukaryotic transcription factors that possess well-characterized sequence-specific DNA binding (61) , CoV N and its CTD are described as non-specific nucleic acid-binding proteins as they bind to RNA and DNA, both single and double-stranded (10, 55, 59) . This characteristic corroborates the multifunctional role played by N during the viral cycle (11). Our results for N-NTD/N-1 NTD-SR binding to DNA reinforce the importance of non-specific charged interactions, such as those observed for RNA (7, 10, 62) . In general, we noticed modest differences comparing K d values for TRS and NS DNAs, as well as for ssDNAs and dsDNA, which might suggest a sequence-specific recognition of TRS with respect to NS and an affinity preference for ssDNA(-) over ssDNA(+) and dsDNA (Table S7 and the preference for TRS over NS DNAs in the formation of liquid condensates. We observed LLPS for ssTRSs but not for ssNS, which might be explained by the difference in affinity to ssNSs (Table S7 and Nucleocytoplasmic transport of nucleocapsid proteins of enveloped RNA viruses. β-sheet II (β1/β5), and α-helices (α1 and α2) colored in dark green, light orange, and magenta, respectively. The thumb (residues I146-V158 in α2/β5) is colored in yellow. A Review of Coronavirus Disease-2019 (COVID-19) Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and coronavirus disease-2019 (COVID-19): The epidemic and the challenges Review of the Clinical Characteristics of Coronavirus Disease 2019 (COVID-19) Continuous and Discontinuous RNA Synthesis in Coronaviruses Sequence Motifs Involved in the Regulation of Discontinuous Coronavirus Subgenomic RNA Synthesis The Architecture of SARS-CoV-2 Transcriptome Specifically Binds the Transcriptional Regulatory Sequence (TRS) and Melts TRS-cTRS RNA Duplexes Determination of host proteins composing the microenvironment of coronavirus replicase complexes by proximity-labeling Structural Insight Into the SARS-CoV-2 Nucleocapsid Protein C-Terminal Domain Reveals a Novel Recognition Mechanism for Viral Transcriptional Regulatory Sequences The SARS coronavirus nucleocapsid protein -Forms and functions The coronavirus nucleocapsid is a multifunctional protein Coronavirus Nucleocapsid Protein Facilitates Template Switching and Is Required for Efficient Transcription A Contemporary View of Coronavirus Transcription Dynamics of the SARS-CoV-2 nucleoprotein N-terminal domain triggers RNA duplex destabilization Structural basis of RNA recognition by the SARS-CoV-2 nucleocapsid phosphoprotein A Molecular Grammar Governing the Driving Forces for Phase Separation of Prion-like RNA Binding Proteins Germline P granules are liquid droplets that localize by controlled dissolution/condensation Biomolecular condensates at the nexus of cellular stress, protein aggregation disease and ageing Measles virus nucleo-And phosphoproteins form liquid-like phase-separated compartments that promote nucleocapsid assembly SARS CoV 2 nucleocapsid protein phase separates with RNA and with human hnRNPs Nucleocapsid protein of SARS-CoV-2 phase separates into RNA-rich polymerasecontaining condensates Genomic RNA Elements Drive Phase Separation of the SARS-CoV-2 Nucleocapsid The SARS-CoV-2 nucleocapsid protein is dynamic, disordered, and phase separates with RNA Large-Scale Recombinant Production of the SARS-CoV-2 Proteome for High-Throughput and Biophysical and Dynamic Characterization of Fine-Tuned Binding of the Human Respiratory Syncytial Virus M2-1 Core Domain to Long RNAs Instrumentation for Fluorescence Spectroscopy NMRPipe: A multidimensional spectral processing system based on UNIX pipes The CCPN data model for NMR spectroscopy: Development of a software pipeline The HADDOCK2.2 Web Server: User-Friendly Integrative Modeling of Biomolecular Complexes Refinement of protein structures in explicit solvent The HADDOCK web server for data-driven biomolecular docking Gromacs: High performance molecular simulations through multi-level parallelism from laptops to supercomputers Improving the Accuracy of Protein Side Chain and Backbone Parameters from ff99SB Refinement of the Cornell et al. Nucleic acids force field based on reference quantum chemical calculations of glycosidic torsion profiles Comparison of simple potential functions for simulating liquid water PROPKA3: Consistent treatment of internal and surface residues in empirical p K a predictions Scripts and Programs Electrostatics of nanosystems: Application to microtubules and the ribosome G-mmpbsa -A GROMACS tool for highthroughput MM-PBSA calculations GROMACS 4.5: A highthroughput and highly parallel open source molecular simulation toolkit Peptide Folding: When Simulation Meets Experiment The PyMOL Molecular Graphics System Liquid-liquid phase separation and fibrillation of the prion protein modulated by a high-affinity DNA aptamer Phase Separation and Disorder-to-Order Transition of Human Brain Expressed X-Linked 3 (hBEX3) in the Presence of Small Fragments of tRNA SARS CoV 2 nucleocapsid protein phase separates with RNA and with human hnRNPs Thermodynamics of Protein Association Reactions: Forces Contributing to Stability Structural basis of RNA recognition by the SARS-CoV-2 nucleocapsid phosphoprotein Interplay between Short-Range Attraction and Long-Range Repulsion Controls Reentrant Liquid Condensation of Ribonucleoprotein-RNA Complexes RNA buffers the phase separation behavior of prion-like RNA binding proteins Labeling nuclear DNA using DAPI Hexanediol: a chemical probe to investigate the material properties of membrane-less compartments Which Is Used to Disrupt Protein-Rich Liquid Droplets in the Cell Cytoplasm, Does Not Disrupt Model Cell Membranes Clusters of bacterial RNA polymerase are biomolecular condensates that assemble through liquid-liquid phase separation Stochasticity of Biological Soft Matter: Emerging Concepts in Intrinsically Disordered Proteins and Biological Phase Separation Biochemical and immunological studies of nucleocapsid proteins of severe acute respiratory syndrome and 229E human coronaviruses Biochemical characterization of SARS-CoV-2 nucleocapsid protein Structural characterization of the C-terminal domain of SARS-CoV-2 nucleocapsid protein Multiple Nucleic Acid Binding Sites and Intrinsic Disorder of Severe Acute Respiratory Syndrome Coronavirus Nucleocapsid Protein: Implications for Ribonucleocapsid Protein Packaging Phosphoregulation of Phase Separation by the SARS-CoV-2 N Protein Suggests a Biophysical Basis for its Dual Functions Homeodomain-DNA recognition Characterisation of the RNA binding properties of the coronavirus infectious bronchitis virus nucleocapsid protein amino-terminal region Detecting transient intermediates in macromolecular binding by paramagnetic NMR N= 21) and 1:2 (N= 759). Bottom inset: corresponding DAPI emission from the top graph insets images (DAPI stains condensates in presence of dsTRS and no fluorescence were observed for the ssTRSs E) Phase separation of 20 µM N-NTD-SR in 20 mM Tristhe specific DNA oligonucleotides dsTRS (top graph), ssTRS(+) (middle graph), and ssTRS(-) (bottom graph) The authors declare that no conflict of interest exists.