Optimal Conditions for In Vitro Assembly of Respiratory Syncytial Virus Nucleocapsid-like Particles

The nucleocapsids (NCs) of the respiratory syncytial virus (RSV) can display multiple morphologies in vivo, including spherical, asymmetric, and filamentous conformations. Obtaining homogeneous ring-like oligomers in vitro is significant since they structurally represent one turn of the characteristic RSV NC helical filament. Here, we analyzed and optimized conditions for forming homogenous, recombinant nucleocapsid-like particles (NCLPs) of RSV in vitro. We examined the effects of modifying the integrated RNA length and sequence, altering incubation time, and varying buffer parameters, including salt concentration and pH, on ring-like NCLPs assembly using negative stain electron microscopy (EM) imaging. We showed that high-quality, homogeneous particles are assembled when incubating short, adenine-rich RNA sequences with RNA-free N associated with P (N0P). Further, we reported that a co-incubation duration greater than 3 days, a NaCl concentration between 100 mM and 200 mM, and a pH between 7 and 8 are optimal for N-RNA ring assembly with polyadenine RNA sequences. We believe assembling high-quality, homogeneous NCLPs in vitro will allow for further analysis of RSV RNA synthesis. This work may also lend insights into obtaining high-resolution nucleocapsid homogeneous structures for in vitro analysis of antiviral drug candidates against RSV and related viruses.


Introduction
Respiratory syncytial virus (RSV) is a significant viral pathogen most commonly responsible for inducing lower respiratory tract diseases in infants, older adults, and immunocompromised individuals globally [1][2][3][4]. In infants, RSV primarily causes bronchiolitis and pneumonia [5], and pathology manifests more generally as immunological hypersensitivity in the respiratory tract associated with increased vulnerability to the development of allergy and asthma [6,7]. Currently, there are no approved vaccines for immunization against RSV for the pediatric or elderly population. RSV is classified as an enveloped nonsegmented negative sense (NNS) RNA virus of the Pneumoviridae family [8]. The genome of a commonly studied RSV A2 strain comprises 15,222 nucleotides and contains ten sequential genes that encode eleven different proteins (of which nine are structural proteins and two are non-structural proteins) [9]. From 3' to 5', the genes encode the following proteins: non-structural proteins 1 and 2 (NS1 and NS2), the nucleoprotein (N), the phosphoprotein (P), the matrix protein (M), the small hydrophobic protein (SH), the glycoprotein (G), the fusion protein (F), M2-1 and M2-2, and a large protein (L, an RNA-dependent RNA polymerase) [10]. Each gene is situated between gene start (GS) and gene end (GE) sequences. The RSV genome also contains two extragenic regions: a 44 nucleotide (nt) leader (Le) sequence and a 155 nt trailer (Tr) sequence at the 3' and 5' ends, respectively, and the Le sequence functions as a single promoter for both transcription and replication. During viral replication, the RSV antigenome (genome complementary sequence) uses the Tr complementary (TrC) sequence as the replication promoter [11]. demonstrated that the efficient assembly of homogeneous RSV N-RNA ring particles in vitro is favored by co-incubating adenine-rich RNA sequences greater than 6 nt in length with RNA-free N 0 P NTD for greater than 3 days, a buffer NaCl concentration between 200 and 400 mM, and a pH value of between 7 and 8. We believe that the formation of stable, homogeneous NCLPs in vitro can be further used to construct a high-resolution structure through cryo-EM and serve as a platform for exploring both of the mechanisms involved in RSV RNA synthesis and elucidating methods for antagonizing the NC complex for therapeutic benefit.

N 0 -P 1-126 Protein Expression and Purification
The N 0 -P 1-126 plasmid was synthesized by SynbioTechnologies (NJ, USA), which allowed for the expression of the N 0 -TEV-P 1-126 -10xHis fusion protein. The E. coli BL21 (DE3) chemically competent strain was utilized for producing N 0 -P  . Cells were cultured at 37 • C in Miller's Luria Broth medium until the A 600 absorbance value reached 0.8. Expression was induced after lowering the temperature to 16 • C and incubating with 0.5 mM of isopropyl β-d-1-thiogalactopyranoside (IPTG) overnight. Cell lysis was conducted through sonication in lysis buffer A (50 mM of sodium phosphate at pH 7.4, 500 mM of NaCl, 5 mM of imidazole, 10% glycerol, and 0.2% NP-40 detergent) and subsequent centrifugation. The supernatant was then loaded onto a cobalt column equilibrated with lysis buffer. The cobalt column (GoldBio, St Louis, MO, USA) was washed with buffer B (50 mM of Tris-HCl at pH 7.4, 1 M of NaCl, 10% glycerol, and 5 mM of imidazole) and buffer C (50 mM of Tris-HCl at pH 7.4, 500 mM of NaCl, 10% glycerol, and 5 mM of imidazole). The protein was eluted from the cobalt column with buffer D (50 mM of Tris-HCl at pH 7.4, 500 mM of NaCl, and 400 mM of imidazole). Further protein purification was performed via Q column (HiTrap Capto Q ImpRes, 5 × 5 mL, Cytiva, Marlborough, MA, USA), followed by TEV (tobacco etch virus) cleavage with dialyzation against buffer E (25 mM of Tris-HCl at pH 8.0, 300 mM of NaCl, 10% glycerol, 1 mM of DTT). The TEV enzyme was removed by re-running the Q column (HiTrap Capto Q ImpRes, 5 × 5 mL, Cytiva, Marlborough, MA, USA). Finally, purification was completed using a Superdex 200 gel-filtration column equilibrated with buffer F (20 mM of HEPES at pH 7.4 and 200 mM of NaCl). The purified protein N 0 -P 1-126 complex was concentrated using a centrifugal concentrator (MilliporeSigma, Burlington, MA, USA) with a molecular weight cut-off (MWCO) of 10 kDa.

In Vitro Assembly of N-RNA Complex
RNA oligos were purchased from Integrated DNA Technologies and saved as stock at a concentration of 100 µM. In order to assemble the NCLPs in vitro, the protein N 0 -P 1-126 complex and RNA were co-incubated with a molar ratio of 1:1.5. The sample was used directly to make the negative stain grids for screening.

RNase A Digestion
For RNase A (MilliporeSigma, Burlington, MA, USA) treatment, the N 0 -P 1-126 complex was preincubated in the presence of RNA at 4 • C overnight, then incubated for one hour at room temperature in the presence of RNase A (final concentration 100 µg/mL).

Negative Stain Electron Microscopy Imaging of N-RNA Complexes
First, 3 µL of each of the samples was applied to the clean side of freshly glowdischarged grids and subsequently stained with 1% (w/v) uranyl formate. Negative stain EM imaging was performed using an FEI Talos L120C electron microscope (Thermo Fisher Scientific, Waltham, MA, USA) operating at 120 keV and equipped with an FEI Ceta 4k × 4k charge-coupled device camera. Micrographs were collected at nominal magnifications of 73,000 (1.97 Å/pixel). The images were acquired at a defocus value of −1.2~−2.0 µm and electron doses of −25 e − /Å 2 .

Results
In this study, we describe the effects of varying sample incubation time, buffer salt concentration, buffer pH, and the length and sequence of encapsidated RNA on in vitro RSV NCLPs assembly. Further, we show an RNase A digestion analysis of NCLPs assembled with varying lengths of short RNA sequences to assess N-RNA complex stability qualitatively.
NCLPs were assembled by incubating RNA-free N (N 0 ) with a fixed molar ratio of RNA. For these complexes, a fusion protein comprised of full-length N (N 1-391 ) and the N-terminal domain of P (P 1-126 ) was utilized, separated by a TEV cleavage domain with a C-terminal 10-His tag. A fixed molar ratio of protein-to-RNA was utilized in this study to achieve saturation of N with RNA when assembling the NCLPs, and protein concentrations were held constant between samples to allow for qualitative comparison. The peak corresponding to the NP 1-126 complex in the gel filtration profile showed an A 260nm /A 280nm ratio of less than 1, implying no RNA contamination in the sample ( Figure 1A,B). A brief overview of the purification protocol to generate the N 0 P 1-126 complex has been outlined ( Figure 1C). We generally observed baseline NCLPs assembly when N and RSV-specific RNA molecules were co-incubated, which reinforced the notion that the polymerization of RSV N and RSV-specific RNA into the ring-like structure is favored and stabilizing. NCLPs were assembled by incubating RNA-free N (N 0 ) with a fixed molar ratio of RNA. Negative stain EM imaging was used to analyze the NCLPs ring assembly. To qualitatively describe the particle quality, we use the term "efficiency", referring to the proportion of particles within a sample that resemble a complete ring-like N-RNA oligomeric assembly. To qualitatively define the energetic favorability of oligomerization, we use the term "propensity", referring to the raw proportion of the specific type of oligomeric structure observed within the sample, including complete N-RNA ring assemblies, incomplete N-RNA ring assemblies, and particle aggregates.

Results
In this study, we describe the effects of varying sample incubation time, buffer salt concentration, buffer pH, and the length and sequence of encapsidated RNA on in vitro RSV NCLPs assembly. Further, we show an RNase A digestion analysis of NCLPs assembled with varying lengths of short RNA sequences to assess N-RNA complex stability qualitatively.
NCLPs were assembled by incubating RNA-free N (N 0 ) with a fixed molar ratio of RNA. For these complexes, a fusion protein comprised of full-length N (N1-391) and the Nterminal domain of P (P1-126) was utilized, separated by a TEV cleavage domain with a Cterminal 10-His tag. A fixed molar ratio of protein-to-RNA was utilized in this study to achieve saturation of N with RNA when assembling the NCLPs, and protein concentrations were held constant between samples to allow for qualitative comparison. The peak corresponding to the NP1-126 complex in the gel filtration profile showed an A260nm/A280nm ratio of less than 1, implying no RNA contamination in the sample ( Figure 1A,B). A brief overview of the purification protocol to generate the N 0 P1-126 complex has been outlined ( Figure 1C). We generally observed baseline NCLPs assembly when N and RSV-specific RNA molecules were co-incubated, which reinforced the notion that the polymerization of RSV N and RSV-specific RNA into the ring-like structure is favored and stabilizing. NCLPs were assembled by incubating RNA-free N (N 0 ) with a fixed molar ratio of RNA. Negative stain EM imaging was used to analyze the NCLPs ring assembly. To qualitatively describe the particle quality, we use the term "efficiency", referring to the proportion of particles within a sample that resemble a complete ring-like N-RNA oligomeric assembly. To qualitatively define the energetic favorability of oligomerization, we use the term "propensity", referring to the raw proportion of the specific type of oligomeric structure observed within the sample, including complete N-RNA ring assemblies, incomplete N-RNA ring assemblies, and particle aggregates.

Within 7 nt RNA Sequences, Adenine Is the Preferred Ribonucleotide for Efficient N-RNA Ring Assembly
The atomic resolution crystal structure of the RSV ring-like NCLPs showed that each N subunit coordinates with 7 ribonucleotides and assembles as a core region [14]. In comparing 7 nt polyadenine (polyA) sequences with randomized sequences consisting of either adenine (A) and cytosine (C) ( Figure 2B) or an A/C/G/U randomized sequence ( Figure 2C), complete N-RNA ring NCLPs were only obtained with the poly7A sequence ( Figure 2A). When N 0 P was incubated with the poly7N RNA sequence that contained a random 7 nt sequence of the 4 ribonucleotides, only particle aggregates could be observed; distinct ring-like structures could not be observed from negative-stain EM imaging ( Figure 2C). Further, more particle aggregation was observed with the A/C randomized sequence than the A/C/G/U randomized sequence, suggesting that with 7 nt RNA sequences, adenine increased the propensity for N-RNA ring assembly and added stability favoring the efficient formation of higher-order morphological structures.

Within 7 nt RNA Sequences, Adenine is the Preferred Ribonucleotide for Efficient N-R Ring Assembly
The atomic resolution crystal structure of the RSV ring-like NCLPs showed tha N subunit coordinates with 7 ribonucleotides and assembles as a core region [14]. In paring 7 nt polyadenine (polyA) sequences with randomized sequences consisting ther adenine (A) and cytosine (C) ( Figure 2B) or an A/C/G/U randomized sequence (F 2C), complete N-RNA ring NCLPs were only obtained with the poly7A sequence (F 2A). When N 0 P was incubated with the poly7N RNA sequence that contained a ran 7 nt sequence of the 4 ribonucleotides, only particle aggregates could be observed; di ring-like structures could not be observed from negative-stain EM imaging (Figur Further, more particle aggregation was observed with the A/C randomized sequence the A/C/G/U randomized sequence, suggesting that with 7 nt RNA sequences, ad increased the propensity for N-RNA ring assembly and added stability favoring th cient formation of higher-order morphological structures. Figure 2. The poly7A sequence was preferred for N-RNA ring assembly over randomized 7 quences with either an A and C or A/C/G/U randomized sequence. Two representative ne stain images are included for a statistically representative depiction. (A) poly7A; (B) poly7M nine/cytosine randomized sequence); (C) poly7N (adenine/cytosine/guanine/uracil randomiz quence). Markers have been utilized to denote representative particle morphologies in A. O rectangles denote incomplete ring assemblies, blue rectangles denote particle aggregation, an rectangles denote complete ring assemblies. Scale bars within the original images correspond nm, and scale bars within the enlarged images correspond to 20 nm.

N-RNA Ring Assembly Propensity Increases when Encapsulating polyA RNA over 6 Ribonucleotides in Length
With varying lengths of polyadenine sequences, we observed that the minimal l required for well-defined N-RNA ring assembly was 7 since no particles were obs for poly5A or poly6A ( Figure 3A,B). An increased propensity for N-RNA oligomeriz was observed when using poly7A; at this length, the first well-defined N-RNA rings seen ( Figure 3C). However, when N 0 P was incubated with poly8A, NCLPs ring form propensity seemed to have increased even further at the detriment of efficiency, le Figure 2. The poly7A sequence was preferred for N-RNA ring assembly over randomized 7 nt sequences with either an A and C or A/C/G/U randomized sequence. Two representative negative stain images are included for a statistically representative depiction. (A) poly7A; (B) poly7M (adenine/cytosine randomized sequence); (C) poly7N (adenine/cytosine/guanine/uracil randomized sequence). Markers have been utilized to denote representative particle morphologies in A. Orange rectangles denote incomplete ring assemblies, blue rectangles denote particle aggregation, and red rectangles denote complete ring assemblies. Scale bars within the original images correspond to 100 nm, and scale bars within the enlarged images correspond to 20 nm.

N-RNA Ring Assembly Propensity Increases when Encapsulating polyA RNA over 6 Ribonucleotides in Length
With varying lengths of polyadenine sequences, we observed that the minimal length required for well-defined N-RNA ring assembly was 7 since no particles were observed for poly5A or poly6A ( Figure 3A,B). An increased propensity for N-RNA oligomerization was observed when using poly7A; at this length, the first well-defined N-RNA rings were seen ( Figure 3C). However, when N 0 P was incubated with poly8A, NCLPs ring formation propensity seemed to have increased even further at the detriment of efficiency, leading to Viruses 2023, 15, 344 6 of 14 a greater proportion of incomplete ring assemblies and particle aggregates but a greater total density of the particles captured ( Figure 3D). Images of N-RNA with the poly11A sequence depicted a high density of the well-defined N-RNA ring structures, suggesting that there was now a high-efficiency assembly of N-RNA rings without detriment to the ring formation propensity ( Figure 3E). These findings highlight the requirement of a minimum sufficient polyadenine sequence length such that N-RNA ring assembly efficiency may be balanced with the propensity for oligomerization. ses 2023, 15, 344 6 of to a greater proportion of incomplete ring assemblies and particle aggregates but a grea total density of the particles captured ( Figure 3D). Images of N-RNA with the poly11 sequence depicted a high density of the well-defined N-RNA ring structures, suggesti that there was now a high-efficiency assembly of N-RNA rings without detriment to t ring formation propensity ( Figure 3E). These findings highlight the requirement of a m imum sufficient polyadenine sequence length such that N-RNA ring assembly efficien may be balanced with the propensity for oligomerization. Markers have been used to denote representative particle morphologies. Orange rectangles den incomplete ring assemblies, blue rectangles denote particle aggregation, and red rectangles den complete ring assemblies. Scale bars within the original images correspond to 100 nm, and sc bars within the enlarged images correspond to 20 nm.

Increasing Incubation Time Increases the Proportion of Complete RSV NCLPs Assemblies with poly7A RNA
When N 0 P was incubated with poly7A RNA, and the particle sets were analyzed greater number of closed-ring particles were identified when the samples were incubat for longer periods at 4 °C. Longer incubation times served to generate a larger number closed-ring N-RNA assemblies with defined shapes. At the 1 h time point, no appreciab NCLPs formation was observed ( Figure 4A). With incubation of the sample overnig ( Figure 4B), an observable proportion of particles captured consisted of unstable, inco plete rings. This proportion decreased with increasing incubation time, while ring fo mation propensity increased ( Figure 4C,D). At the 3-day incubation point, a high propo tion of particle aggregation was observed, seemingly due to the increased propensity b subpar assembly efficiency ( Figure 4C). At the 15-day incubation point, ring formati efficiency was the highest, with minimal particle aggregation, incomplete N-RNA ri assembly, and a high proportion of well-defined, complete N-RNA rings. Markers have been used to denote representative particle morphologies. Orange rectangles denote incomplete ring assemblies, blue rectangles denote particle aggregation, and red rectangles denote complete ring assemblies. Scale bars within the original images correspond to 100 nm, and scale bars within the enlarged images correspond to 20 nm.

Increasing Incubation Time Increases the Proportion of Complete RSV NCLPs Assemblies with poly7A RNA
When N 0 P was incubated with poly7A RNA, and the particle sets were analyzed, a greater number of closed-ring particles were identified when the samples were incubated for longer periods at 4 • C. Longer incubation times served to generate a larger number of closed-ring N-RNA assemblies with defined shapes. At the 1 h time point, no appreciable NCLPs formation was observed ( Figure 4A). With incubation of the sample overnight ( Figure 4B), an observable proportion of particles captured consisted of unstable, incomplete rings. This proportion decreased with increasing incubation time, while ring formation propensity increased ( Figure 4C,D). At the 3-day incubation point, a high proportion of particle aggregation was observed, seemingly due to the increased propensity but subpar assembly efficiency ( Figure 4C). At the 15-day incubation point, ring formation efficiency was the highest, with minimal particle aggregation, incomplete N-RNA ring assembly, and a high proportion of well-defined, complete N-RNA rings.  15-day incubation period. The first image in each column represents negative controls (N 0 P1-126 incubated without RNA), followed by two representative negative stain images. Markers have been added to the figure series to denote distinct and representative particle morphologies. Orange rectangles denote incomplete ring assemblies, blue rectangles denote particle aggregation, and red rectangles denote complete ring assemblies. Negative controls depict that RNA is essential for RSV N oligomerization and that the N 0 P protein complex cannot independently form highly ordered ring particles. Scale bars within the original images correspond to 100 nm, and scale bars within the enlarged images correspond to 20 nm.

Increasing Buffer Salt Concentration/ionic strength Tones RSV N-RNA Ring Formation Propensity
When the incubation buffer NaCl concentration was stepped between values starting from 50 mM and extending to 2 M, a significant alteration in RSV N-RNA ring formation propensity was observed, with seemingly no qualitative change in ring formation efficiency. At 50 mM of NaCl, ring formation propensity seemed excessively high, leading to widespread aggregation and a general inability to distinguish distinct particle morphologies ( Figure 5A). Doubling the salt concentration to 100 mM reduced the degree of particle aggregation observed, and complete and incomplete N-RNA particles could now be identified ( Figure 5B). Doubling NaCl concentration again to 200 mM reduced particle aggregation further, and particle morphologies remained observable ( Figure 5C). Raising the salt concentration to 400 mM resulted in minimal to negligible aggregation and a further reduction in N-RNA ring formation propensity, reflected by the relatively lower density of particles ( Figure 5D). No ring formation was observed at NaCl concentrations of 1 M and 2 M (Figure 5E,F).  15-day incubation period. The first image in each column represents negative controls (N 0 P 1-126 incubated without RNA), followed by two representative negative stain images. Markers have been added to the figure series to denote distinct and representative particle morphologies. Orange rectangles denote incomplete ring assemblies, blue rectangles denote particle aggregation, and red rectangles denote complete ring assemblies. Negative controls depict that RNA is essential for RSV N oligomerization and that the N 0 P protein complex cannot independently form highly ordered ring particles. Scale bars within the original images correspond to 100 nm, and scale bars within the enlarged images correspond to 20 nm.

Increasing Buffer Salt Concentration/ionic strength Tones RSV N-RNA Ring Formation Propensity
When the incubation buffer NaCl concentration was stepped between values starting from 50 mM and extending to 2 M, a significant alteration in RSV N-RNA ring formation propensity was observed, with seemingly no qualitative change in ring formation efficiency. At 50 mM of NaCl, ring formation propensity seemed excessively high, leading to widespread aggregation and a general inability to distinguish distinct particle morphologies ( Figure 5A). Doubling the salt concentration to 100 mM reduced the degree of particle aggregation observed, and complete and incomplete N-RNA particles could now be identified ( Figure 5B). Doubling NaCl concentration again to 200 mM reduced particle aggregation further, and particle morphologies remained observable ( Figure 5C). Raising the salt concentration to 400 mM resulted in minimal to negligible aggregation and a further reduction in N-RNA ring formation propensity, reflected by the relatively lower density of particles ( Figure 5D). No ring formation was observed at NaCl concentrations of 1 M and 2 M (Figure 5E,F). Markers have been utilized to denote representative particle morphologies. Orange rectangles denote incomplete ring assemblies, blue rectangles denote particle aggregation, and red rectangles denote complete ring assemblies. Scale bars within the original images correspond to 100 nm, and scale bars within the enlarged images correspond to 20 nm.

N-RNA Ring Assembly Propensity is Highly Sensitive to pH
Using high RNA:N ratio samples, we observed marked particle aggregation with virtually no well-defined ring formation at both extremes of the pH range (4-9) tested ( Figure  6A,F). Notably, initial incremental pH changes appeared to alter N-RNA ring formation propensity drastically: increasing pH from 4 to 5-6 eliminated particle formation altogether, suggesting that N and RNA remained isolated ( Figure 6B,C). At pH 7, N-RNA particles were observed, albeit in aggregates, but the propensity for ring formation is moderate relative to pH 4 and pH 9 ( Figure 6D). Well-defined N-RNA rings with high efficiency and moderate propensity were observed only at pH 8 ( Figure 6E), and mass aggregation was observed at pH 9 ( Figure 6F). Due to the more gradual morphological transition between pH 7 and 8, the optimal pH for high N-RNA ring formation efficiency likely hovers around 8. This relatively narrow pH range suggests that ring assembly is susceptible to pH changes. Markers have been utilized to denote representative particle morphologies. Orange rectangles denote incomplete ring assemblies, blue rectangles denote particle aggregation, and red rectangles denote complete ring assemblies. Scale bars within the original images correspond to 100 nm, and scale bars within the enlarged images correspond to 20 nm.

N-RNA Ring Assembly Propensity Is Highly Sensitive to pH
Using high RNA:N ratio samples, we observed marked particle aggregation with virtually no well-defined ring formation at both extremes of the pH range (4-9) tested ( Figure 6A,F). Notably, initial incremental pH changes appeared to alter N-RNA ring formation propensity drastically: increasing pH from 4 to 5-6 eliminated particle formation altogether, suggesting that N and RNA remained isolated ( Figure 6B,C). At pH 7, N-RNA particles were observed, albeit in aggregates, but the propensity for ring formation is moderate relative to pH 4 and pH 9 ( Figure 6D). Well-defined N-RNA rings with high efficiency and moderate propensity were observed only at pH 8 ( Figure 6E), and mass aggregation was observed at pH 9 ( Figure 6F). Due to the more gradual morphological transition between pH 7 and 8, the optimal pH for high N-RNA ring formation efficiency likely hovers around 8. This relatively narrow pH range suggests that ring assembly is susceptible to pH changes.  Orange rectangles denote incomplete ring assemblies; blue rectangles denote particle aggregation; red rectangles denote complete ring assemblies. Scale bars within the original images correspond to 100 nm, and scale bars within the enlarged images correspond to 20 nm.

RNase A Treatment of N-RNA Complexes Depicts that RSV NCLPs Assembled in Vitro are Stable
In order to test the stability of the RSV NCLPs assembled in vitro, RNase A was used to digest the RNA in the NCLPs with RNA sequence lengths greater than 7 nt, reducing the sample proportion of all observed oligomer types. For the 7 nt, 8 nt, and 11 nt sequences tested, RNase A treatment of the N-RNA complexes resulted in a reduction in the sample proportion of incomplete ring oligomers and particle aggregates (Figure 7). A significant reduction in the proportion of complete, well-defined N-RNA rings was only noted with RNase treatment of the N-RNA complexes that contained poly8A ( Figure 7B) and poly11A ( Figure 7C). Orange rectangles denote incomplete ring assemblies; blue rectangles denote particle aggregation; red rectangles denote complete ring assemblies. Scale bars within the original images correspond to 100 nm, and scale bars within the enlarged images correspond to 20 nm.

RNase A Treatment of N-RNA Complexes Depicts That RSV NCLPs Assembled In Vitro Are Stable
In order to test the stability of the RSV NCLPs assembled in vitro, RNase A was used to digest the RNA in the NCLPs with RNA sequence lengths greater than 7 nt, reducing the sample proportion of all observed oligomer types. For the 7 nt, 8 nt, and 11 nt sequences tested, RNase A treatment of the N-RNA complexes resulted in a reduction in the sample proportion of incomplete ring oligomers and particle aggregates (Figure 7). A significant reduction in the proportion of complete, well-defined N-RNA rings was only noted with RNase treatment of the N-RNA complexes that contained poly8A ( Figure 7B) and poly11A ( Figure 7C). Orange rectangles denote incomplete ring assemblies; blue rectangles denote particle aggregation; red rectangles denote complete ring assemblies. Scale bars within the original images correspond to 100 nm, and scale bars within the enlarged images correspond to 20 nm.

Discussion
In this study, we show that it is possible to construct high-quality, homogeneous NCLPs assemblies within certain ranges of select buffer parameters and RNA characteristics, which can further serve as a manipulatable base unit for performing NC functional studies, such as in vitro RNA polymerase assays. An RNA-free NP fusion protein consist- Figure 7. RNase A treatment significantly reduced the sample proportion of complete and incomplete N-RNA rings and particle aggregates. Two representative negative stain images are included for a statistically representative depiction. (A) poly7A; (B) poly8A; (C) poly11A. Orange rectangles denote incomplete ring assemblies; blue rectangles denote particle aggregation; red rectangles denote complete ring assemblies. Scale bars within the original images correspond to 100 nm, and scale bars within the enlarged images correspond to 20 nm.

Discussion
In this study, we show that it is possible to construct high-quality, homogeneous NCLPs assemblies within certain ranges of select buffer parameters and RNA characteristics, which can further serve as a manipulatable base unit for performing NC functional studies, such as in vitro RNA polymerase assays. An RNA-free NP fusion protein consisting of full-length N and the N-terminal domain of P ( Figure 8A) was used as a minimal unit for NCLPs decameric ring assembly ( Figure 8B). Our results agree with our previously published results [25] in that A is the preferred ribonucleotide for both encapsidation by N and subsequent NCLPs assembly. A total of 7 nt sequences were used since this was the minimal length required for successful encapsidation by N in vitro, maintaining consistency with both our previous findings [25] as well as recent results by Gonnin and colleagues [27]. Randomizing between A and just one other ribonucleotide, C, resulted in particle aggregation with no well-perceived N-RNA oligomeric rings. Randomizing between all four ribonucleotides resulted in particle aggregation as well. However, incubating N 0 P with the 7 nt A/C/G/U randomized sequence resulted in a seemingly lowered propensity for particle oligomerization relative to the A and C randomized RNA. This suggests that when using 7 nt RNA sequences, A has a greater binding affinity to RSV N, making N more amenable to RNA encapsidation through certain mechanisms, such as favoring P displacement from N 0 .
We also compared N-RNA ring formation propensity and efficiency when incubating N 0 P with short polyA RNA sequences of varying lengths: the first well-defined N-RNA rings were observed when using poly7A RNA. This suggests that shorter RNA sequences may not stabilize the equilibrium from the RNA-free N conformation to the RNAbound N conformation required for oligomerization. Interestingly, ring formation propensity increases, whereas formation efficiency reduces when incubating N 0 P with poly8A; ring efficiency is restored despite an even greater oligomerization propensity when incubating N 0 P with poly11A. This is potentially suggestive of different N-RNA binding modalities that may depend on encapsidated RNA length. In RSV, when using an RNA length equivalent to the length of RNA bound by each N monomer, which is 7 nt, the encapsidation conformation may follow that of other viruses in the Mononegavirales order. For instance, in the Measles virus, each N monomer binds 6 ribonucleotides such Model depicting oligomerization of the base unit N 0 P NTD with poly7A under optimal conditions. N-monomers oligomerize into a decameric ring, where P is (i) a molecular chaperone of N to produce the RNA-free N 0 P complex and (ii) is substituted with the co-incubated RNA sequence to yield the N-RNA ring.
Our results agree with our previously published results [25] in that A is the preferred ribonucleotide for both encapsidation by N and subsequent NCLPs assembly. A total of 7 nt sequences were used since this was the minimal length required for successful encapsidation by N in vitro, maintaining consistency with both our previous findings [25] as well as recent results by Gonnin and colleagues [27]. Randomizing between A and just one other ribonucleotide, C, resulted in particle aggregation with no well-perceived N-RNA oligomeric rings. Randomizing between all four ribonucleotides resulted in particle aggregation as well. However, incubating N 0 P with the 7 nt A/C/G/U randomized sequence resulted in a seemingly lowered propensity for particle oligomerization relative to the A and C randomized RNA. This suggests that when using 7 nt RNA sequences, A has a greater binding affinity to RSV N, making N more amenable to RNA encapsidation through certain mechanisms, such as favoring P displacement from N 0 .
We also compared N-RNA ring formation propensity and efficiency when incubating N 0 P with short polyA RNA sequences of varying lengths: the first well-defined N-RNA rings were observed when using poly7A RNA. This suggests that shorter RNA sequences may not stabilize the equilibrium from the RNA-free N conformation to the RNA-bound N conformation required for oligomerization. Interestingly, ring formation propensity increases, whereas formation efficiency reduces when incubating N 0 P with poly8A; ring efficiency is restored despite an even greater oligomerization propensity when incubating N 0 P with poly11A. This is potentially suggestive of different N-RNA binding modalities that may depend on encapsidated RNA length. In RSV, when using an RNA length equivalent to the length of RNA bound by each N monomer, which is 7 nt, the encapsidation conformation may follow that of other viruses in the Mononegavirales order. For instance, in the Measles virus, each N monomer binds 6 ribonucleotides such that its 15,894 nt genome is completely encapsidated and is nuclease-resistant [28,29]. Alternatively, for sequences longer than 7 nt, the RNA could be overhanging between N monomers, bridging consequent N monomers into a continuous ring, or randomly oriented within the oligomeric ring structure. However, it is also possible that some N monomers remain in the N 0 conformation within the ring and do not bind RNA. Results from the RNase A assay showed that N-RNA ring assembly propensity with poly7A RNA was impacted with RNase A treatment to a far lesser extent than for either of the longer polyA sequences tested: poly8A and poly11A. There was a reduction in the proportion of all observed morphologies regardless of polyA length, which may indicate that even with one specific RNA length, multiple binding configurations can be adopted depending on energetic favorability. This indicates that longer RNA may form overhangs that can potentially serve as nucleation sites for oligomerization, which are otherwise eliminated with RNase A treatment. However, further work needs to be performed concerning structural analyses of N-RNA interactions at the RNA-binding groove to ascertain any such possibility.
We then modified the incubation time and two buffer parameters: NaCl concentration and pH. While complete N-RNA rings were observable at 1 day, 3 days, and 15 days, the propensity for ring formation was much greater, as was the efficiency of well-defined N-RNA rings. At 15 days, there were minimal particle aggregates, suggesting that thermodynamic stability may have been attained for many N-RNA conformers. Thus, it may be the case that particle aggregates may be a result of kinetically favorable interactions, or perhaps particle aggregates are the predominant particle morphology corresponding to kinetically trapped intermediates. Our data suggest that complete N-RNA rings are the thermodynamically stable conformation for RSV NCLPs.
As elaborated on in an earlier section, RSV P functions in preventing both the selfoligomerization of N 0 and non-specific binding to random cellular RNAs [30,31], and it has been previously shown that the N-terminal domain of P is sufficient for maintaining the monomeric N 0 form [11,25]. Interestingly, previous literature has also noted that in human RSV (hRSV), the C-arm of N is directly involved in inhibiting the associative interactions between N and RNA. Specifically, a general mechanism of RNA binding inhibition within the Mononegavirales order of viruses was described as involving a dual interaction between the N C-arm and the RNA binding domain as well as the binding of P NTD to N, such that interactions with other N monomers and non-viral specific RNAs are impaired [31]. In altering buffer salt concentration, our results indicated that N-RNA ring assembly was optimal in a specific band between 200 mM and 400 mM of NaCl. It may be the case that a higher ionic strength destabilizes the salt-bridging interactions that pack the C-arm against the core of the N-monomer, favoring concurrent placement of RNA nucleotides within the RNA-binding domain and subsequent C-arm linkage with the subsequent N monomer (which is dependent on the RNA sequence length). However, elevating ionic strength above an optimum value may lead to excessive C-arm flexibility to the extent that limiting its flexibility may be entropically unfavorable despite the stabilization conferred. It may be due to monomeric inactivation through association near the RNA binding groove of its resident N monomer or polymerization via linkage to the consequent nucleoprotein monomer to form N-RNA oligomers. This may explain why at least some particle aggregation is observed at the extremes of the salt concentration tested. In relation to pH, however, while optimal N-RNA ring assembly is reserved within a small band between pH 7 and pH 8, virtually no particle formation propensity is observed between pH 5 and pH 6. N-RNA ring assembly appears to be highly sensitive to pH, implying that a significant interaction required for oligomerization may be hindered at certain pH values. Further work is necessary to elucidate the structural impact of modifying buffer ionic strength and pH on the RNA-binding and oligomerization domains of the N protein.
While our current study provides a framework of conditions for efficient in vitro N-RNA ring construction (Figure 9), further characterization of in vitro N-RNA genomic structure morphologies that may occur due to interactions with other RSV proteins remains to be conducted. A recent study conducted by Conley et al. [32] revealed that, in cells infected by RSV, N-RNA rings seem to adopt a stacked, helical, filamentous conformation within RSV virions. Additionally, while short RNA sequences were used in our analysis here, further work characterizing conditions for optimal N-RNA ring assembly when using longer sequences of RNA, such as 70 nt RNA, remains to be performed. When considering the in vitro assembly of RSV NCLPs, a noteworthy limitation is that the RNA has already been synthesized before assembling the N 0 P complex. However, within infected cells, the N 0 P complex assembly occurs concomitantly with RNA synthesis, which may impact NCLPs formation dynamics. Despite this, in vitro studies are still highly informative as they allow for greater NCLPs manipulation, enabling N to encapsidate specific RNA structures/sequences of choice. The ability to construct high-quality NCLPs in vitro and manipulate the RSV RNA synthesis machinery within this environment is, thus, a powerful method that can better inform strategies for constructing other oligomeric structures assumed by the RSV NCLPs within infected cells, such as helical filaments. Optimizing the construction of different NCLPs morphologies will ultimately assist in recreating the native RSV NCLPs structure bound to specific genomic RNA sequences in vitro, which can be used as a model for developing therapeutic interventions against RSV infection. structure morphologies that may occur due to interactions with other RSV proteins remains to be conducted. A recent study conducted by Conley et al. [32] revealed that, in cells infected by RSV, N-RNA rings seem to adopt a stacked, helical, filamentous conformation within RSV virions. Additionally, while short RNA sequences were used in our analysis here, further work characterizing conditions for optimal N-RNA ring assembly when using longer sequences of RNA, such as 70 nt RNA, remains to be performed. When considering the in vitro assembly of RSV NCLPs, a noteworthy limitation is that the RNA has already been synthesized before assembling the N 0 P complex. However, within infected cells, the N 0 P complex assembly occurs concomitantly with RNA synthesis, which may impact NCLPs formation dynamics. Despite this, in vitro studies are still highly informative as they allow for greater NCLPs manipulation, enabling N to encapsidate specific RNA structures/sequences of choice. The ability to construct high-quality NCLPs in vitro and manipulate the RSV RNA synthesis machinery within this environment is, thus, a powerful method that can better inform strategies for constructing other oligomeric structures assumed by the RSV NCLPs within infected cells, such as helical filaments. Optimizing the construction of different NCLPs morphologies will ultimately assist in recreating the native RSV NCLPs structure bound to specific genomic RNA sequences in vitro, which can be used as a model for developing therapeutic interventions against RSV infection.