Improving the Thermostability of the Qβ Bacteriophage Coat Protein Through Single-Site Mutation Based on Molecular Dynamics

Abstract

Norovirus is a major cause of acute viral gastroenteritis in humans. Molecular biology-based detection methods play a pivotal role in ensuring accurate and specific diagnosis. The inclusion of Qβ phage particles as armored positive controls in these assays can further enhance their reliability and specificity. Herein, we discuss rational design strategies to improve the stability of Qβ bacteriophage capsid proteins armored with RNA using Discovery Studio 2019 protein design software. Amino acid mutation sites were deter-mined based on changes in folding free energy differences (ΔΔGmut). These single-site mutations were subsequently evaluated using molecular dynamics simulations. Wild-type and mutant recombinant expression plasmids were constructed and transformed into Escherichia coli BL21 (DE3) for cloning and expression. The stability of Qβ virus-like particles (VLPs) was assessed using real-time fluorescence RT-qPCR. The results showed that structurally intact and uniformly distributed wild-type and single-site mutant VLPs were successfully obtained. Stability analyses indicated that at 4 °C, 25 °C, 37 °C, 45 °C, and 60 °C, the single-site mutant exhibited a significantly lower rate of degradation than the wild-type. In conclusion, rational design enables the generation of single-site mutant VLPs with enhanced stability, providing a safer and more stable standard reference material for the molecular detection of foodborne viruses.

1. Introduction

Foodborne disease is a significant contributor to the global disease burden. Outbreaks and illnesses caused by foodborne microbial pathogens place a heavy burden on public health. Each year, unsafe food causes an estimated 600 million cases of food-borne disease and 420,000 deaths worldwide [1]. Human noroviruses (NoVs), hepatitis A virus, human rotavirus, and hepatitis E virus are among the most frequent causes of foodborne infections. NoV is one of the most prevalent infectious agents, capable of causing nonbacterial gastroenteritis and responsible for numerous foodborne outbreaks associated with the consumption of contaminated food [2,3]. Real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) is currently the gold standard for sensitive and accurate detection of these pathogens and serves as a critical tool in outbreak prevention and control [4]. However, low viral concentrations, high genetic diversity among viruses, and the complexity of food matrices containing various PCR enzyme inhibitors make viral detection in food samples challenging [5,6], which can lead to false-negative results. Therefore, the inclusion of a positive control or reference material is essential for monitoring molecular assay performance and ensuring that tests are conducted correctly. To date, various biological materials, including live viruses, inactivated viruses, and naked RNA, have been used as controls for the molecular diagnosis of RNA viruses. However, loss of accuracy with these positive controls is often associated with RNA or cDNA degradation, cross-contamination, biosafety risks, and human errors occurring during multistep extraction or during reverse transcription, PCR, or one-step RT-PCR procedures [7,8]. To address these limitations, researchers have developed armored RNA (AR) technology. AR is a complex of MS2 bacteriophage coat protein and RNA produced in Escherichia coli through induction of an expression plasmid encoding the coat protein and a target RNA sequence [9]. AR offers several advantages, including noninfectivity, ability to assess the entire testing process, resistance to RNase degradation, ease of storage, quantifiability, ease of preparation, and suitability for transportation [10]. Thus, AR plays a key role as a positive control in ensuring the accuracy and reliability of foodborne virus detection methods. This technology has been successfully used to con-struct quality controls for NoV [11], hepatitis A virus [12], and rotavirus [13]. AR is commonly prepared using bacteriophages, with MS2 and Qβ bacteriophages being the most widely used. Although AR based on MS2 bacteriophage performs well in certain applications, AR based on Qβ bacteriophage is considered more stable as a positive control due to the presence of disulfide bonds [14,15], a finding that is also supported by our previous study [11].

Enhancing stability is one of the crucial issues in the preparation and development of AR [16], as it is essential for ensuring the accuracy and reliability of foodborne virus detection and serves as an important strategy for cost reduction. In addition to screening for more stable bacteriophages such as Qβ, another powerful approach is the directed evolution of the Qβ capsid protein to enhance its thermostability and thereby improve the stability of AR prepared using Qβ bacteriophage. Traditionally, directed evolution relies on random mutagenesis and in vitro recombination; however, the degeneracy of the genetic code can skew and restrict library design. Over the years, substantial advances in semi-rational and computational engineering have enabled more intelligent navigation of sequence space, and quantum mechanics (QM) and molecular dynamics (MD) calculations, along with machine-learning algorithms, have become invaluable tools [17]. In a broader context, rational design enhances protein stability through precise modifications based on a deep molecular understanding, allowing tailored molecular adjustments for specific performance improvements and operational efficiency. Contrary to trial-and-error approaches, this strategy minimizes unnecessary experiments and reduces resource usage [18,19].

In this study, a rational design strategy was adopted to achieve the directed evolution of Qβ AR and to screen for Qβ AR mutants with enhanced thermostability. The thermo-stabilities of Qβ AR mutants were first estimated computationally using MD simulations. Subsequently, two Qβ AR mutants with improved thermostability were constructed following confirmation of site-directed mutagenesis. The final mutant plasmids were ex-pressed in E. coli BL21 (DE3). The thermal stability of the two mutants was evaluated via PCR after incubation at 4 °C, 25 °C, 37 °C, 45 °C, and 60 °C. The Qβ AR mutants exhibited higher thermal stability than the wild-type (WT) at all tested temperatures, suggesting that a rational design strategy can be effectively applied to improve the thermal stability of AR for the detection of foodborne viruses.

2. Results

2.1. Engineering of the Qβ Phage Coat Protein Using Rational Strategies

Rational strategies were used to extract evolutionary information from homologous proteins to identify nonconserved amino acid residues [20,21]. In total, 29 nonconserved amino acid sites were identified as potential mutation targets. Based on these nonconserved residues listed in Table 1, the Calculate Mutation Energy (Stability) and Predict Stabilizing Mutations modules were applied to perform single-point mutations. Changes in mutation energy before and after mutation were calculated, and mutations were classified as stabilizing when ΔΔGmut ≤ 0.5 kcal/mol. The five mutations with the lowest predicted mutation energies were selected as the top candidates, as shown in Table 2.

Table 1.

Selection results of nonconserved amino acid residues.

Number	Amino Acid Site	Abbreviation	Number	Amino Acid Site	Abbreviation
1	Leu3	L3	16	Lys67	K67
2	Thr5	T5	17	Asn70	N70
3	Thr7	T7	18	Cys74	C74
4	Ile11	I11	19	Cys80	C80
5	Gly15	G15	20	Ser83	S83
6	Lys16	K16	21	Val84	V84
7	Val20	V20	22	Arg86	R86
8	Arg24	R24	23	Tyr89	Y89
9	Asn30	N30	24	Ser95	S95
10	Gln37	Q37	25	Phe96	F96
11	Thr49	T49	26	Tyr99	Y99
12	Ser51	S51	27	Thr101	T101
13	Arg59	R59	28	Ile122	I122
14	Lys60	K60	29	Asn129	N129
15	Gln65	Q65

Table 2.

Mutation energy prediction for single-site mutations (top five).

Number	Single Mutation Sites	Name	Mutation Energy (kcal/mol)	Result
1	SER51.TYR	S51Y	−3.26	Stabilizing
2	SER51.PHE	S51F	−2.39	Stabilizing
3	GLN65.TRP	Q65W	−2.17	Stabilizing
4	LYS16.CYS	K16C	−2.14	Stabilizing
5	SER83.PHE	S83F	−2.08	Stabilizing

2.2. Analysis of MD Simulation Results

To further investigate structural alterations, MD simulations were conducted on the top five mutations identified in Section 2.1. The MD results for both WT and single-point mutants were analyzed, and comparative analyses of RMSD, RMSF, and potential energy were performed. RMSD was used as a key metric to evaluate protein structural changes, with lower RMSD values indicating enhanced structural stability [22,23]. Previous studies have indicated that the RMSF of a protein not exceeding 2 Å and the RMSD not exceeding 3 Å values is regarded as within an acceptable range [24,25]. According to the RMSD results (Figure 1A), all five mutants exhibited significantly lower RMSD values than the WT, indicating that these mutations reduced overall flexibility, enhanced rigidity, and thereby improved structural stability. The WT trajectory showed substantial fluctuations within the first 7 ns, whereas mutants Q65W and S51F exhibited consistently lower RMSD values with smaller fluctuations, stabilizing at approximately 12.5 ns. The WT RMSD plot equilibrated at approximately 18 Å, whereas the Q65W and S51F RMSD plots equilibrated at approximately 8 Å. Meanwhile, the RMSD curves were not completely identical to the final conformations of WT and the mutants. A slight downward trend was observed for the mutants, indicating modest structural deviations from the WT conformation [22]. RMSF reflects the conformational fluctuations of individual amino acid residues during an MD trajectory [26,27]. The results showed that the RMSF values of the WT at 300 K were higher than those of the single-point mutants, with differences exceeding 0.05 Å, suggesting increased rigidity in the single-point mutants (Figure 1B). A previous study demonstrated a positive correlation between increased residue rigidity and enhanced protein thermal stability [28]. Regions with higher RMSF peaks indicate less stable areas of the protein. Among the mutants, Q65W showed markedly reduced overall conformational fluctuations compared with the WT and other mutants. Changes in the proportion of α-helix and β-sheet elements in the protein secondary structure may underlie this improved thermostability [29,30]. Comparison of secondary structure content before and after simulation (Figure 2) showed that many α-helix and β-sheet structures in the WT were disrupted and converted into random coil and turn structures after simulation, whereas these elements were largely retained in Q65W and S51F. During the MD simulation of WT protein, the protein almost completely unfolded for reasons unknown. In comparison, the MD simulations of Q65W and S51F mutants display less pronounced unfolding suggesting that these mutants have a stabilizing effect on the protein structure. Based on these results, the single-point mutants Q65W and S51F were selected for confirmatory experiments to evaluate their thermostability.

Figure 1.

MD simulation results of WT and single-point mutants at 300 K for 20 ns: (A) RMSD; (B) RMSF. Wild type is shown in black, S51Y in yellow, S51F in pink, Q65W in blue, K16C in green and S83F in red.

Figure 2.

MD simulation trajectory results of WT and single-point mutants: (A) initial state; (B) WT; (C) Q65W; (D) S51F. α-Helices are shown in red, β-sheets in blue, random coils in gray and β-bridges in green.

2.3. Expression of WT and Mutant AR

Efficient expression of mutants in E. coli BL21 (DE3) is critical for engineering enhanced thermal stability. Following successful expression and purification of pET-QINoVGII and the mutants Q65W and S51F, WT and mutant AR were analyzed via SDS–PAGE after IPTG induction. The molecular weight of the target protein was approximately 14 kDa on SDS–PAGE, consistent with the theoretical value (Figure 3), confirming effective expression of pET-QINoVGII and the Q65W and S51F mutants in E. coli BL21. TEM results showed that the VLPs had a diameter of approximately 25 nm (Figure 4). The VLPs were structurally intact, uniformly sized, and morphologically similar to native bacteriophage Qβ.

Figure 3.

SDS–PAGE analysis of VLP expression. Lane M, protein molecular weight marker; lane 1, proteins of E. coli BL21; lane 2, proteins of E. coli BL21 harboring pET-28a(+); lane 3, proteins of E. coli BL21 carrying WT (A), mutant Q65W (B), and mutant S51F (C).

Figure 4.

TEM images of WT and single-point mutant VLPs: (A) WT; (B) Q65W; (C) S51F.

2.4. WT and Mutant AR Quantification

RNA encapsulated in the VLPs was extracted and analyzed using RT-qPCR. Purified AR samples consistently showed no amplification after 40 cycles, whereas the positive control containing the recombinant plasmid produced strong and distinct amplification signals. These results indicate that the purified WT, Q65W mutant, and S51F mutant AR preparations were free of residual plasmid DNA (Figure 5). RT-qPCR was then performed to quantify the concentrations of the AR samples. Based on serial dilution gradients and corresponding Ct values, standard curves were generated as follows: y = −3.565x + 54.35 (R² = 0.998, efficiency = 0.908), y = −3.298x + 46.54 (R² = 0.9992, efficiency = 1.010), and y = −3.545x + 48.77 (R² = 0.9992, efficiency = 0.915) for WT and AR, the Q65W mutant, and the S51F mutant, respectively. Both R² values and amplification efficiencies were within the optimal ranges [31]. The concentrations of target RNA in the WT, Q65W mutant, and S51F mutant working solutions were 2.3 × 10⁹, 6.9 × 10⁷, and 9.2 × 10⁷ copies/μL, respectively.

Figure 5.

Real-time fluorescence PCR results for residual plasmid detection in WT and mutant AR: 1. Positive control; 2. WT; 3. Q65W; 4. S51F; 5. Blank control.

2.5. Stability Tests

The stability of AR under both conventional and extreme conditions relevant to storage and transportation was evaluated. The stability of the WT and the two mutants at different temperatures and over various durations was analyzed and compared (Figure 6). Under extreme conditions at 60 °C, WT AR exhibited a degradation rate of 48.60% after 20 days, whereas the Q65W and S51F mutants showed degradation rates of 28.2% and 31.5%, respectively. Under another extreme condition at 45 °C, after 30 days of storage, degradation rates were 26.59% for the WT, 18.91% for Q65W, and 23.45% for S51F. Similarly, at 37 °C, the WT exhibited a degradation rate of 25.56% after 30 days, whereas the Q65W and S51F mutants showed degradation rates of 18.38% and 17.5%, respectively. Under conventional conditions at 25 °C, after 30 days of storage, the WT showed a degradation rate of 23.68%, whereas Q65W and S51F showed degradation rates of 13.18% and 17.92%, respectively. Finally, under conventional storage at 4 °C, after 60 days, the WT exhibited a degradation rate of 24.14%, whereas Q65W and S51F showed markedly lower degradation rates of 4.36% and 16.68%, respectively. Overall, the Q65W and S51F mutants demonstrated superior stability compared with the WT under both conventional (4 °C and 25 °C) and extreme (37 °C, 45 °C, and 60 °C) conditions.

Figure 6.

Effect of temperature on the stability of WT and mutant AR.

3. Discussion

Illnesses and deaths resulting from food contamination represent a persistent threat to public health and constitute a significant impediment to global socioeconomic development. In 2010, foodborne viruses were estimated to contribute 33 million disability-adjusted life years worldwide. Among these, enteric pathogens, particularly NoVs, are the most prevalent causes of foodborne disease [32,33]. Molecular detection technologies are critical for the rapid identification of viral pathogens because of their high sensitivity and efficiency. AR technology overcomes many of the limitations associated with the manufacture and use of naked RNA as standards or controls in clinical diagnostic assays. Improving the stability of AR is therefore essential to ensure the robustness and reproducibility of molecular detection processes. The conformational stability of a protein is thermodynamically defined by changes in free energy [34]. Targeted mutation of “hot spot” amino acid residues identified through computational screening and sequence analysis has emerged as an effective strategy for enhancing protein stability. For improving the thermostability of the Qβ phage coat protein through site-directed mutation, rational se-lection of candidate residues is a critical and essential step. MD simulations were employed to guide purposeful modifications of the AR amino acid sequence with the aim of enhancing thermal stability. Previous studies by Oliver [35], Wang [36], Sawitri [37], and others have demonstrated, using MD simulations, that mutant proteins can exhibit greater rigidity than WT, which is closely associated with improved thermal stability. Following prediction of mutation sites through MD analysis, mutant variants were obtained via sequence synthesis. Compared with traditional PCR-based point mutation strategies, synthetic gene synthesis offers clear advantages in terms of success rate and time efficiency. Additionally, the experimental costs of gene synthesis are comparable to those of PCR-based mutagenesis and subsequent subcloning. These advantages have contributed to the increasing adoption of synthetic sequencing as a preferred approach for generating mutant genes, particularly for relatively small gene fragments, over conventional point mutation PCR methods [38].

This study employed a semi-rational design strategy to purposefully modify the amino acid sequence of AR with the objective of enhancing its thermal stability. Identification of nonhomologous amino acid regions helps mitigate the risk of introducing changes that could adversely affect protein function, as certain regions may play critical roles in protein folding and stability, and alterations in these areas may lead to instability or functional loss. Consequently, comparative analysis of amino acid sequences across different bacteriophage capsid proteins is essential for identifying nonconserved residues suitable for mutation [39]. Systematic analysis of MD-associated parameter trajectories after mutation showed that the RMSD of the protein backbone converged during the later stages of the simulations. WT exhibited fluctuation values of approximately 20 Å, whereas Q65W and S51F showed markedly lower fluctuation values of approximately 6.5 Å and 7 Å, respectively, indicating increased rigidity in the mutant proteins. RMSF analysis further demonstrated that the fluctuation trends of Q65W and S51F were substantially lower than those of WT and the other mutants. Analysis of the final simulation frames revealed that the completeness of secondary structure elements in Q65W and S51F exceeded that observed in WT. Based on the favorable performance of Q65W and S51F, the enhanced thermostability of these mutants may be attributed to the more favorable hydrophobic environment introduced by substitution with tryptophan and phenylalanine. Kauzmann et al. proposed that hydrophobic effects are a primary driving force for protein folding [40]. The introduction of additional hydrophobic interactions may strengthen the hydrophobic core of individual monomers and reinforce interfacial interactions, thereby improving the stability of the protein’s globular conformation [41,42]. Although the mutants in this study partially lost their original secondary structures at 300 K, their relatively compact scaffolds allowed the globular conformation to remain folded, enabling stable expression and purification of AR [34]. Notably, the initial stock concentrations of the mutants varied substantially, with the WT copy number exceeding that of the mutants by approximately 100-fold. This difference may be explained by the specific combination of mutations increasing the conformational potential energy of the mutants, weakening critical intersubunit contacts, and thereby reducing the self-assembly efficiency of AR, ultimately resulting in variable initial stock concentrations.

After predicting, designing, screening, and synthesizing theoretically more stable mutants, experimental validation was performed. Compared with WT, the Q65W and S51F mutants exhibited lower degradation rates at 60 °C, 45 °C, 37 °C, 25 °C, and 4 °C. This overall agreement with the trends predicted by MD simulations supports the effectiveness of the selected mutation sites identified through virtual amino acid substitution and simulation-based approaches. However, the changes in degradation rates of the Q65W and S51F mutants between 25 °C and 37 °C were not fully consistent with the expected trends. Across this temperature range, the degradation rate of Q65W increased from 13.18% to 18.38%. Conversely, S51F showed only minor variation, with degradation rates changing from 17.92% to 17.5%. Notably, the change observed for S51F was less pronounced than that for Q65W, and S51F even demonstrated a slight decrease in degradation rate (0.42%), suggesting that temperature variation between 25 °C and 37 °C has no significant impact on its stability. Markovic-Housley et al. [43] reported that replacement of phenylalanine with tryptophan in the IIA domain had a destabilizing effect. We favor the interpretation that the observed differences in stability are attributable to the specific nature of the mutation site. Substitution at this position may result in the loss of hydrogen-bonding interactions or weakened intersubunit interactions within the dimer, thereby reducing structural stability [44]. It remains unclear from the current structural data which specific conformational changes occur in the S51F mutant that could account for the observed effects. Moreover, given the complexity of protein sequence–structure–energy relationships, multiple factors, including the local mutation environment, hydration structure, and unfolded-state energetics, contribute to protein stability. Therefore, computational engineering approaches typically achieve only moderate success in prospectively predicting protein stability [45].

4. Materials and Methods

4.1. Plasmids and Reagents

The pET28a(+) plasmid was used as the expression vector, and E. coli BL21 (DE3) was used as the host strain for protein expression. All plasmids, including pET-QINoVGII (WT) and its mutant derivatives, were synthesized by Sangon Biotech Co., Ltd. (Shanghai, China). Molecular biology reagents, including 2× Premix Ex Taq (probe qPCR) and the One Step PrimeScript™ RT-PCR Kit (Perfect Real Time), were purchased from TaKaRa Bio Inc. (Dalian, China) and used for real-time PCR assays. All PCR primers listed in Table 3 were synthesized by Sangon Biotech Co., Ltd. (Shanghai, China). All chemicals used in this study were of analytical reagent grade.

Table 3.

Primers and TaqMan probes used in the real-time PCR system.

Name	Sequence (5′-3′)	Product Length	Purpose
QNIF2d	ATGTTCAGRTGGATGAGRTTCTCWGA	89 bp	Identification of GII norovirus detection target
COG2R	TCGACGCCATCTTCATTCACA
QNIFS	FAM- AGCACGTGGGAGGGCGATCG- TAMARA

4.2. Multiple Sequence Alignment and Homology Modeling

The complete genomes of Escherichia phage Qβ (GenBank No. AB971354.1), Marinomonas phage (GenBank No. NC_018269.1), Cellulophaga phage (GenBank No. NC_021795.1), Enterobacteria phage (GenBank Nos. NC_008720.1 and NP_891732.1), and Xenorhabdus bovienii (GenBank No. NC_013892.1) were downloaded from the NCBI GenBank database (https://www.ncbi.nlm.nih.gov/, accessed on 21 June 2023). MEGA 11 software (version 11.0.8) was used for multiple sequence alignment and phylogenetic analysis using the maximum-likelihood (ML) method, and nonconserved amino acid exhibiting variation in all 6 capsid proteins were identified and selected as potential sites for mutagenesis [46]. Homology models were generated using Discovery Studio 2020 based on the crystal structure with PDB code 5VLY (chain B) obtained from the Protein Data Bank (https://www.rcsb.org/, accessed on 5 July 2023) as the template (Sequence identity: 100%) [47]. The quality of the mutant models was evaluated using the SAVES v6.0 scoring program (https://saves.mbi.ucla.edu, accessed on 5 July 2023) [20,48]. Protein structures were visualized and analyzed using Discovery Studio 2019 (BIOVIA, San Diego, CA, USA) and the PyMOL (v3.1) Molecular Graphics System (DeLano Scientific, San Carlos, CA, USA) [49]. Based on predicted folding free energy changes (ΔΔG), 10 candidate mutations were identified through energy- and evolution-based calculations, and the final candidate mutants were selected from the top five mutations.

4.3. MD Simulations

To further evaluate the stability of the WT and mutant proteins, MD simulations were performed at 300 K using Discovery Studio 2019 (BIOVIA, San Diego, CA, USA). The simulation protocol included system preparation using the “Prepare Protein” module. All MD simulations were performed with explicit TIP3P solvent. The CHARMM36 force field was used with a 2 fs time step and SHAKE constraints on hydrogen bonds. Long-range electrostatics were treated with the Particle Mesh Ewald (PME) method, and van der Waals interactions were truncated at 12 Å with a smooth switching function between 10 and 12 Å [50,51]. Steepest descent energy minimization was first performed to eliminate steric clashes and reduce maximum forces, followed by an equilibration step [52]. A stepwise heating protocol was employed to gradually raise the system temperature from 50 K to 300 K (10 ps per 50-K step). Subsequently, a 20 ns NPT (constant pressure and temperature) production simulation was conducted. Finally, a 100 ns MD simulation was performed in the NPT ensemble without any restraint. During the simulations, key structural parameters, including root mean square deviation (RMSD) and root mean square fluctuation (RMSF), were analyzed to assess protein stability [53].

4.4. Construction of Recombinant Plasmids of the WT and Mutants

The prokaryotic expression plasmid pET-QINoVGII, containing the WT bacteriophage Qβ genome (GenBank AB971354, nt 61–2367) and the GII NoV detection target (GenBank X86557, nt 5012–5100), was constructed in our previous study [11]. Recombinant plasmids harboring Qβ capsid protein mutation sites together with the GII NoV detection target were constructed based on the pET-QINoVGII plasmid.

4.5. Expression and Purification of WT and Mutants

Protein expression and purification of Qβ virus-like particles (VLPs) and their mutants followed the same protocol as described in our previous study [11]. Briefly, E. coli BL21 (DE3) cells harboring the recombinant prokaryotic expression plasmids were cultured in LB medium containing kanamycin (50 μg mL⁻¹). When the optical density at 600 nm reached 0.6, protein expression was induced with 1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) and continued at 37 °C for 12 h with agitation at 200 rpm. The SDS-PAGE was performed using 12% (w/v) polyacrylamide separating gels, the gels were stained with Coomassie Brilliant Blue R-250. After confirmation via SDS–PAGE, the recombinant E. coli cultures were diluted and expanded in 200 mL of LB medium under the same conditions. Cells were harvested via centrifugation at 10,000 rpm for 10 min at 4 °C, resuspended, and disrupted through sonication. The lysate was clarified by centrifugation at 12,000 rpm for 20 min at 4 °C to obtain the supernatant. To remove host genomic DNA and RNA, the supernatant was incubated with 100 U DNase I and 200 U RNase A at 37 °C for 1 h. AR particles were purified from the bacterial lysate via CsCl density gradient ultracentrifugation at 80,000 rpm for 4 h at 4 °C (CP100WX, Hitachi, Tokyo, Japan). The major fractions were collected and dialyzed against phosphate-buffered saline (pH 7.2) using sonication buffer. AR was further purified via Sephacryl S-200 size-exclusion chromatography using a BioLogic DuoFlow chromatography system (Bio-Rad Laboratories, Hercules, CA, USA). Working solutions were prepared by diluting WT samples 10-fold and mutant samples 100-fold with RNase-free ddH₂O. Subsequently, 100 μL aliquots were dispensed into multiple RNase-free Eppendorf tubes.

4.6. Transmission Electron Microscopy (TEM)

ARs were negatively stained with 2% phosphotungstic acid and visualized at a magnification of 400,000× using a transmission electron microscope (JEM-1200EX, JEOL Ltd., Akishima, Tokyo, Japan).

4.7. Residual Plasmid DNA Detection

The purity of purified ARs was assessed using real-time PCR (without reverse transcription) targeting the cloned cDNA of the NoV (GII) detection sequence with primers QNIF2d/COG2R and probe QNIFS [54,55]. Real-time PCR was performed in a total reaction volume of 20.0 μL. Each reaction mixture contained 10 μL of 2× Premix Ex Taq (Probe qPCR) (Takara Bio Inc., Dalian, China), 0.4 μL (10 μmol μL⁻¹) of QNIF2d, 0.4 μL (10 μmol μL⁻¹) of COG2R, 0.8 μL (10 μmol μL⁻¹) of QNIFS, 2.0 μL of AR, and 6.4 μL of ddH₂O. PCR amplification was performed on a Roche LightCycler 480II system under the following conditions: initial denaturation at 95 °C for 10 s, followed by 40 cycles of denaturation at 95 °C for 5 s and annealing at 60 °C for 20 s. All assays were performed in triplicate. Plasmid pET-QINoVGII was used as the positive control, and sterile water served as the blank control.

4.8. AR Quantification

AR quantification followed the same protocol as described in our previous study [11]. Briefly, RNA was extracted from the working solutions using TRIzol™ (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions and quantified using an absorbance-based nucleic acid quantification method with a NanoPhotometer (Implen GmbH, München, Germany). Two microliters of the extracted sample was added to 18 μL of master mix consisting of 10 μL of 2× One Step RT-PCR Buffer III, 0.4 μL (5 U/μL) of TaKaRa Ex Taq HS, 0.4 μL of PrimeScript RT Enzyme Mix II, 0.3 μL of QNIF2d, 0.3 μL of COG2R, 0.4 μL of QNIFS, and 6.2 μL of RNase-free ddH₂O. RT-qPCR was performed on a LightCycler 480 II system using the following thermal cycling conditions: one cycle at 42 °C for 5 min, followed by one cycle at 95 °C for 10 s, and 40 cycles of denaturation at 95 °C for 5 s and annealing at 60 °C for 20 s. Fluorescent signals were collected during the annealing step.

4.9. Thermostability of WT AR and Its Mutants

Thermostability assessments were performed for three AR preparations, including WT and its two mutants generated in this study. All samples were divided into five groups (20 samples per group) and incubated at 60 °C, 45 °C, 37 °C, 25 °C, and 4 °C. Stability was evaluated at different temperatures and time points as follows: 60 °C (1, 2, 3, 4, 5, 7, 10, 20, and 30 days), 45 °C (3, 4, 6, 8, 14, 20, 40, and 60 days), 37 °C (3, 4, 6, 8, 14, 20, 40, and 60 days), 25 °C (2, 4, 6, 9, 15, 30, 60, 90, and 120 days), and 4 °C (5, 15, 25, 35, 60, 90, 120, and 150 days) [11,56]. After incubation for the designated durations, samples were stored at −80 °C until completion of the experiment. For each stability study, a single batch was aliquoted into individual time point samples of 100 μL. All samples were quantified using the real-time RT-qPCR protocol described in Section 4.8.

4.10. Statistical Analyses

Cycle threshold (Ct) values of AR were measured at each time point, and statistical analyses were performed using ANOVA to evaluate differences between WT AR and its mutants under different conditions (p < 0.05) [56]. All statistical analyses were conducted using GraphPad Prism version 10.2.3 (GraphPad Software, San Diego, CA, USA; www.graphpad.com).

5. Conclusions

In this study, two mutants with high thermostability were obtained by integrating sequence and structural analyses with a computer-aided semi-rational design strategy. The Q65W and S51F mutants were stably expressed and exhibited lower degradation rates than the WT at 60 °C, 45 °C, 37 °C, 25 °C, and 4 °C. These findings provide valuable insights and demonstrate a rational and efficient approach for enhancing the thermostability of Qβ VLPs, thereby facilitating their application in industrial processes involving transport or operation across various temperatures.

Author Contributions

M.Q.: Data Curation, writing—original draft; N.L.: Software; M.L.: Data Curation; J.S.: Validation; Y.J.: Writing-review and editing; W.Z. and Y.G.: Conceptualization and methodology; D.W.: Supervision; L.Y.: Project administration and funding acquisition. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Data is contained within the article.

Conflicts of Interest

The authors declare no conflicts of interest.

Funding Statement

This research was funded by National Natural Science Foundation of China, grant number 32172292, Central Public-interest Scientific Institution Basal Research Fund, YSFRI, CAFS, grant number 20603022025007, 2023TD76, Earmarked fund for CARS, grant number CARS-49, and National Key Research and Development Program of China, grant number 2017YFC1600703.