Optimized Propagation and Purification Protocols for Large-Scale Production of Rhinovirus C

TL;DR

This paper describes optimized methods to grow and purify large amounts of Rhinovirus C, which could help in developing treatments and vaccines.

Contribution

The study introduces a robust system for large-scale RV-C production and recommends a lipase-based purification method for high-quality virus preparations.

Findings

RV-C15a yields were significantly lower in suspension cultures compared to adherent cultures.

Lipase treatment followed by ultracentrifugation produced highly pure and concentrated virus preparations.

Infection timing and calcium supplementation improved virus production in adherent cultures.

Abstract

Background: Rhinovirus C (RV-C) is one of three species of rhinoviruses (RVs), which cause the common cold, preschool wheezing illnesses and exacerbations of asthma. RV-C types are more virulent, especially in children, but progress in developing treatments is limited by difficulties in generating high-titer virus preparations. The goals of this study were to optimize methods for large-scale production and purification of RV-C to facilitate structure and immune response studies. Methods: We optimized protocols for the propagation and purification of RV-C15a, a clinical isolate adapted to HeLa-E8 cells stably expressing virus receptor CDHR3. We compared virus yields in adherent and suspension cultures, evaluated the effects of calcium supplementation and infection timing, and tested multiple purification strategies, including ultracentrifugation, dialysis, and lipase treatment. Results: RV-C15a yields were significantly lower in suspension vs. adherent cultures despite comparable virus binding and entry, suggesting post-entry replication limitations in suspended cells. In adherent cultures, infecting soon after cell seeding and calcium supplementation reduced the time of virus production and modestly improved virus progeny yields. Surface CDHR3 expression declined over time, potentially restricting viral spread. Among purification methods, lipase treatment of infected cell lysates followed by ultracentrifugation produced highly pure and concentrated virus preparations suitable for structural and immunological applications, with high yields. Conclusions: We present a robust system for large-scale RV-C15a production in adherent HeLa-E8 cells and recommend a lipase-based purification method as a rapid and effective approach for producing high-quality viral preparations. These advances will support structural studies and accelerate the development of RV-C-targeted therapeutics and vaccines.

Figures

Full text

1. Introduction

The common cold is highly prevalent in the U.S. and industrialized nations, with adults experiencing 2–4 illnesses yearly. Respiratory infections are especially common in young children; a recent surveillance study found that children under age 2 years have an average of 9.4 infections per year [1]. This ailment causes significant work and school absenteeism, amounting to 40% of job time lost and 30% of school days missed [2]. Rhinoviruses (RVs) cause over half of respiratory illnesses [1,2]. RVs belong to a family of small single-stranded, positive-sense RNA viruses known as picornaviruses that include over 160 genetically assigned types classified into three species: RV-A, RV-B, and RV-C. Of the three species, infections with RV-B usually cause asymptomatic or mild illness [3,4]. RV-A and RV-C cause upper respiratory illnesses as well as wheezing illnesses and exacerbations of chronic airway diseases such as asthma. RV-C is most likely to cause wheezing in preschool children and provoke exacerbations of asthma [4,5,6,7,8].

Cadherin-related family member 3 (CDHR3) is a transmembrane protein expressed by ciliated airway epithelial cells that mediates RV-C binding to cells and enables virus entry and replication [9]. The CDHR3 extracellular domains or ectodomains 1–6 (EC1-EC6) retain a long, rod-like structure with calcium ions bound to acidic clusters at the domain junctions [10,11]. Calcium binding is required for proper folding and rigid structure of CDHR3 and for virus binding to recombinant CDHR3 protein domains [12].

We and others have propagated multiple RV-C clinical isolates in primary cultures of airway epithelial cells differentiated at air–liquid interface (ALI) [13,14,15,16,17], and in immortalized ALI cultures of bronchial epithelial cells HBEC3-KT [18]. However, these cultures are impractical for large-scale virus production due to limited supply of primary cells, and the high costs and several weeks required for cell differentiation. Since CDHR3 is not expressed by cells commonly used to culture RV-A and B, such as WI-38, WisL, and HeLa cells, we developed a transduced HeLa cell line derivative (HeLa-E8) that stably expresses the receptor and enables propagation of RV-C isolates in this continuous cell line [9]. RV-C15a was adapted from a clinical isolate by serial passaging to optimize replication in HeLa-E8 cells [19]. While RV-C15a grows to high titers in adherent cultures of HeLa-E8, the yields of virus progeny in suspension cell cultures remain below those achieved with HeLa-adapted strains of RV-A or RV-B. Propagating these RV species in suspension cultures allows for infection of many cells in a single Erlenmeyer flask at high concentration (2–8 × 10^8^ cells/150 mL), which simplifies virus purification steps [20]. Similar large-scale production of purified RV-C has not been achieved, but would support studies of viral 3D structure, receptor binding, and replication mechanisms, as well as the development of monoclonal antibodies and reference sera in animal models. It would also facilitate efforts to develop RV-C vaccines and antivirals.

The objectives of this study were to optimize protocols for large-scale propagation and purification of adapted RV-C strains in HeLa-E8 cells. To achieve this goal, we tested RV-C binding and replication in HeLa-E8 cells in suspension and adherent cultures, optimized culture conditions, and developed purification methods that are both efficient and suitable for various downstream applications.

2. Materials and Methods

2.1. Viruses

RV-C15a [19], adapted to HeLa-E8 cells from clinical isolate W10 (GenBank accession number GU219984), was used at passage 18. Two major receptor group RV-A types with different adaptation levels to HeLa cells served as controls: RV-A16 (GenBank accession number L24917), a highly adapted laboratory strain, and RV-A34 (GenBank accession number PX726324), a clinical isolate serially passaged eight times in HeLa cells. All viruses were propagated in adherent HeLa-E8 cultures and purified by ultracentrifugation through a sucrose cushion (100,000× g, 4 h, 10 °C).

2.2. Reverse Transcription and Quantitative PCR (RT-qPCR)

Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Hilden, Germany) and reverse transcribed with TaqMan Reverse Transcription Reagents (Life Technologies, Carlsbad, CA, USA). Viral RNA concentration was assessed by qPCR using the Power SYBR Green PCR master mix (Applied Biosystems, Waltham, MA, USA), as previously described [21]. RV RNA levels were determined using a standard curve derived from serial 10-fold dilutions of virus with known titer and expressed as PFU equivalents (PFUe).

2.3. Infection of HeLa-E8 Suspension Cultures

HeLa-E8 cells were grown in suspension culture using medium B (S-MEM, Gibco, Grand Island, NY, USA [Cat. No. 11380-037] with non-essential amino acids, L-glutamine, Penicillin/Streptomycin, Pluronic F68 and 10% serum [6% FBS and 4% NCS]) and passaged when they reach 4–5 × 10^5^ cells/mL. Cells were infected using published protocols [20] with some modifications. For RV infection, 5 × 10^7^ cells were pelleted (500× g, 5 min, room temperature), washed (5 mL PBS with Ca and Mg), gently resuspended, and pelleted again. The cell pellet was resuspended in 4 mL PBS with Ca and Mg, and inoculated with 1 mL of clarified virus lysate (multiplicity of infection [MOI] of 10 PFUe/cell, room temperature, 1 h) in a 15 mL conical tube. After virus attachment, the cell suspension was transferred to a 125 mL Erlenmeyer flask containing 45 mL of pre-warmed Medium B, and the culture was adjusted for pH (7.2–7.4), if needed, by CO_2_ gassing. The culture was incubated in an incubator-shaker (130 rpm, 34 °C, 24 h) and frozen at −80 °C before virus purification.

To investigate the effects of calcium concentration on virus binding, 10^6^ cells were pelleted (500× g, 5 min, room temperature) in 1.5 mL Eppendorf tubes, washed (0.5 mL PBS with Ca and Mg), resuspended in medium A (1.8 mM calcium), medium B (no calcium) or PBS (2 mM calcium) and inoculated (MOI = 10, room temperature, 2 h) with gentle agitation. Cells were washed (3 times, PBS with Ca and Mg) to remove unattached virus, pelleted, and lysed by adding the RLT lysis buffer (Qiagen) for total RNA extraction and viral RNA quantification by RT-qPCR.

2.4. Infection of HeLa-E8 in Adherent Cultures

Cells were initially grown in suspension and then plated (4 × 10^5^ cells/well) in a 12-well plate using medium A (MEM, Gibco, Grand Island, NY, USA [Cat. No. 11095-080] with non-essential amino acids, penicillin, streptomycin, and 10% FBS) and allowed to attach for either 1, 2 or 4 h before infection. For the 24 h time point, we plated 2 × 10^5^ cells to compensate for cell growth over 24 h. Subconfluent cell monolayers in duplicate wells were infected with virus (1 × 10^6^ PFUe/well, 2 h), then washed (3 times, PBS with Ca and Mg) to remove unattached virus. After washing, we added RLT lysis buffer (Qiagen) to the 2 h wells to extract RNA and quantify virus binding by RT-qPCR at 2 h post-infection (hpi). The 72 h wells received 1 mL of medium A and were incubated for an additional 70 h to allow virus replication. Cells were lysed using RLT lysis buffer and virus progeny yields were measured at 72 hpi by RT-qPCR.

2.5. Trypsin Treatment of Infected HeLa-E8 Cells

Trypsin treatment was employed to remove virus particles bound to the cell surface but not internalized. HeLa-E8 cells were inoculated at a MOI of 10 for 2 h at either 4 °C or 34 °C prior to treatment. For suspension cultures, HeLa-E8 cells (1 × 10^6^ cells) were pelleted (600× g, 3 min), resuspended in 0.5 mL of trypsin-EDTA (Gemini Bio-Products, West Sacramento, CA, USA, Cat. No.400-151), and incubated for 5 min at 37 °C. For adherent cultures (~4 × 10^5^ cells per well in a 12-well plate), the virus inoculum was aspirated, and 0.5 mL of trypsin-EDTA was added per well. Cells were incubated for 5 min at 37 °C. Following trypsin treatment, cells were washed (3×) with PBS, pelleted, and lysed by adding RLT lysis buffer (Qiagen) for total RNA extraction and RT-qPCR.

2.6. Immunohistochemistry

We determined CDHR3 expression in HeLa-E8 cells cultured in 12-well glass-bottom plates (Cellvis, Mountain View, CA, USA). To visualize surface CDHR3 expression, we fixed the cells at 24 h, 48 h and 72 h after plating (Hema 3 fixative, ThermoFisher, Waltham, MA, USA) and blocked non-specific binding with 10% goat serum in PBS with 0.1% Tween-20 for 30 min at room temperature. We then incubated cells overnight (4 °C) with anti-CDHR3 monoclonal antibody (anti-EC1, clone 2H5, 1/200 dilution) [12] and washed them (3×, 10 min, PBS with 1% goat serum). Next, we stained with the secondary antibody (anti-mouse-Alexa Fluor 633, 1/1000 dilution, 2 h in the dark), washed, and imaged the cells using Eclipse Ti Microscope (Nikon, Tokyo, Japan, 20×/0.75 magnification) and NIS-Elements software, version 5.02.01 (Nikon). We used ImageJ, version 1.8.0_322 [22] to quantify CDHR3 expression using a cell edge detection method to identify positive signal. First, we adjusted the brightness and contrast of the NIS-Elements multi-dimensional image files (.nd2 file format) by using the auto-adjust feature to enhance visibility. Next, we applied the “Find Edge” tool in ImageJ to delineate cell boundaries. We then set a standard threshold to define the cell edges and maintained this threshold consistently across all images. After edge detection, we measured the mean relative fluorescence intensity across the entire image, repeating this process for multiple regions within each well.

2.7. Virus Purification by Ultracentrifugation Through Sucrose Cushion

Infected HeLa-E8 cells were frozen and thawed 3 times and scraped from the flasks. The cell lysate was vortexed, sonicated (30 s, 30 W) using the Sonicator 3000 Ultrasonic Cell Disruptor (Misonix, Farmingdale, NY, USA) and centrifuged (10 min, 10,000× g, 4 °C) to remove cell debris. The clarified lysate was then treated with RNase A (Qiagen, 10 μg/mL, 10 min, 37 °C) to remove free RNA. Next, 1 mL of 10% N-lauroylsarcosine (Sigma-Aldrich, St. Louis, MO, USA) and 20 µL of β-mercaptoethanol (Sigma) were added per 10 mL of clarified lysate, and 11 mL of this mixture was layered over 1 mL of PBS with 30% (w/v) sucrose in an Ultra-Clear 14 mm × 89 mm centrifuge tube (Beckman, Brea, CA, USA). The virus was pelleted by centrifugation (10 °C, 4 h, 24,200 rpm, SW41 rotor, Sorvall WX 80+ ultracentrifuge), and the supernatant and sucrose were aspirated.

For use as a viral inoculum, we resuspended the virus pellet in 100 µL of cold PBS containing 0.01% bovine serum albumin (BSA). Alternatively, for virus structural studies we resuspended the pellet in 100 µL of NTE buffer (20 mM Tris, 120 mM NaCl, 1 mM EDTA, pH 8.0), which has been successfully applied in cryo-EM analyses of RV particles [23]. After adding PBS with 0.01% BSA or NTE, the tube was incubated overnight (4 °C), and the virus pellet was resuspended by pipetting up and down multiple times. The virus suspension was centrifuged (10,000× g, 3 min, 1.5 mL tube) to remove any remaining debris. Virus purification from larger volumes (>60 mL) of infected cell lysates was performed similarly, using Ultra-Clear 25 mm × 89 mm centrifuge tubes (Beckman) in an SW32 rotor. Total RNA was extracted from 2 μL of the purified viral suspension (QIAamp Viral RNA Mini Kit, Qiagen) and viral RNA concentration was determined by RT-qPCR.

2.8. Virus Filtration

We concentrated virus samples by filtration (Pierce Protein Concentrators PES 100 kDa MWCO, Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer’s protocol. Briefly, 10 mL of clarified lysate was loaded into the sample chamber and centrifuged (3000× g, 45 min) until the sample was concentrated about 50-fold. Longer centrifugation did not allow to further reduce the sample volume. Fluid retained from the bottom and center of the sample chamber was gently aspirated using a pipet tip and transferred into a 1.5 mL tube.

2.9. Virus Precipitation

We added 40% (w/v) polyethylene glycol with an average molecular weight of 6000 (PEG 6000, Sigma-Aldrich) to clarified lysate from infected cells to a final concentration of 8%. The viral particles were then precipitated (overnight, 4 °C, continual stirring), followed by centrifugation (3000× g, 15 min, 4 °C). The supernatant was removed, and the pellet was resuspended in PBS with 0.01% BSA or NTE buffer.

2.10. Virus Purification by Sucrose Gradient

A continuous gradient was prepared by overlaying 7.5% sucrose over 45% sucrose solution (5.5 mL each) in an Ultra-Clear 14 mm × 89 mm centrifuge tube (Beckman). This tube was then frozen and thawed 4 times to allow for gradual mixing. Next, the concentrated viral suspension (1 mL) was overlayed on the sucrose gradient and pelleted (24,200 rpm, SW41 rotor, Sorvall WX 80+ ultracentrifuge, 10 °C for 4 h). We collected fractions by first removing the top 3 mL of the gradient and then collecting 1 mL for fractions 1 and 2, followed by 0.5 mL for fractions 3–10, and then up to 1 mL for the final aliquot. Then the fractions 3–10 were tested by RT-qPCR to identify two fractions with the highest virus concentration. Sucrose was removed from the virus preparation by dialysis (Float-A-Lyzer G2 Dialysis Device, Spectra/Por, New Brunswick, NJ, USA) overnight at 4 °C in 2 L of NTE buffer or by 10-fold dilution in NTE buffer containing 0.1% BSA followed by centrifugation (24,200 rpm, SW41 rotor, Sorvall WX 80+ ultracentrifuge, 10 °C, 4 h).

2.11. Calcium Concentration Testing in Growth Medium

HeLa-E8 cells were grown in suspension and then plated (4 × 10^5^ cells/well, 12-well plate). One control well contained medium A (1.8 mM calcium chloride), while the experimental wells contained increasing concentrations of calcium chloride (2.5 mM, 3 mM, and 4 mM). Cells were infected (1 × 10^6^ PFUe/well, 2 h), followed by three washes with PBS. The cells were then incubated for an additional 70 h in medium A supplemented with the respective calcium concentrations. When the cells were ready for harvesting, we aspirated PBS (2 hpi) or growth medium (72 hpi) and added a lysis buffer (RLT, Qiagen) to the cells for subsequent RNA extraction and RT-qPCR.

2.12. Lipase Treatment

The clarified lysate from infected cells was treated with porcine pancreas lipase (Sigma, 0.2 units/µL, 20 min, 37 °C); then we added RNase A (Qiagen, 10 μg/mL, 10 min, 37 °C) to degrade free RNA. The sample was then subjected to the standard purification protocol, which included ultracentrifugation through a 30% sucrose cushion.

2.13. Electron Microscopy

We treated the copper grids with amyl acetate and then layered them with a 5 nm carbon film. The grids were then subjected to glow discharge treatment for 15 min to render the surface hydrophilic. We applied 3 µL of the purified RV-C15a directly onto the grid, gently blotted (Whatman filter paper, 1001-090), and incubated for 1 min. We then washed (3 µL of NTE buffer) the grid and blotted the excess buffer away, applied filtered uranyl acetate stain (3 µL, 10–30 s), blotted and rewashed. The grids were analyzed using Talos F200C TEM transmission electron microscope at room temperature with 92,000× magnification.

2.14. Statistical Analysis

All statistical analyses were performed using GraphPad Prism version 9.0.2. We used either one or two-tailed t-tests or ordinary one-way ANOVA followed by post hoc tests to assess statistical significance. p < 0.05 was considered significant. Results from ≥3 independent experiments were expressed as means ± standard deviation (s.d).

3. Results

3.1. RV-C15a Progeny Yields in HeLa-E8 Suspension Culture Are Suboptimal

RV-A and RV-B laboratory strains can grow to high yields in suspension culture of H1-HeLa cells [20]. To determine whether RV-C15a can be similarly propagated in HeLa-E8 cell suspension cultures, we compared viral progeny yields between RV-C and RV-A using a scaled-down protocol (5 × 10^7^ cells). RV-C15a progeny yields in suspension culture were 1–2–log units lower compared to HeLa-adapted RV-A laboratory strains (RV-A1a and RV-A16, [20]) and a HeLa-adapted RV-A34 clinical isolate (Figure 1A). RV-C15a infection did not result in a measurable increase in viral load, indicating a lack of productive replication.

Since calcium ions (Ca^2+^) binding at the junctions between the extracellular domains of cadherins, including CDHR3, is essential for maintaining their correct, rigid conformation, we investigated whether the calcium concentration during the virus binding step affects RV-C15a receptor binding efficiency in cell suspension. After a 2 h incubation in medium with or without calcium, comparable levels of cell-associated virus were detected, suggesting that the calcium concentration does not significantly influence RV-C15a binding (Figure 1B). Similarly, RV-A16 binding to HeLa-E8 cells was independent of calcium concentration (Figure 1B). Notably, RV-A16 exhibited a higher binding efficiency compared to RV-C15a (≥92% vs. ≥12% of input inoculum).

3.2. RV-C15a Binding and Entry in Suspension and Adherent Cultures

To determine the mechanisms for suboptimal RV-C15a progeny yields in suspension culture, we compared the receptor binding and cell entry steps of RV-C15a in suspension and adherent cultures of HeLa-E8. We inoculated cells at 4 °C to inhibit virus entry and 34 °C to enable virus entry and then treated them with trypsin to cleave any receptor-bound virus that had not yet entered the cell. The incubation temperature (4 °C vs. 34 °C) did not affect the amounts of C15a bound to cells in suspension or adherent cultures, which averaged 20–25% of the input inoculum (Figure 2A,B). When incubated at 4 °C, trypsin reduced cell-associated virus by 30-fold (adherent) and 37-fold (suspension) compared to control cells, but there was only a 2–3-fold difference at 34 °C (Figure 2). These findings suggest that most of the receptor-bound virus can enter cells in both culture types.

3.3. Optimizing RV-C15a Propagation in Adherent Cultures

To optimize growth in adherent cultures, we tested effects of plating the cells for different periods of time before infection and increasing calcium concentration in growth medium during infection on viral binding and progeny yields. The various post-plating times all had similar amounts of cell-associated virus at 2 hpi and progeny yields 72 hpi. (Figure 3A). Increasing concentrations of calcium in the growth medium had modest effects on progeny yields, which were significantly increased (1.8-fold) at 3 mM compared to 1.8 mM concentrations (Figure 3B).

To investigate whether virus receptor expression changes over time, we analyzed surface expression of CDHR3 in non-permeabilized HeLa-E8 cells using antibodies targeting ectodomain 1 24–72 h after cells were plated (Figure 4A). We observed that surface CDHR3 expression peaked at 24 h post-seeding (p.s.) and declined significantly over time, with a 2.3-fold reduction at 48 h (p = 0.042) and 13.5-fold reduction at 72 h compared to 24 h p.s. (p = 0.008) (Figure 4B).

3.4. Optimizing RV-C Purification

We next compared three methods of RV-C concentration and purification: filtration with a protein concentrator (100 kDa MWCO), PEG-6000 precipitation, and ultracentrifugation over a 30% sucrose cushion. All three methods produced a similar viral yield (Figure 5A) but differed in the final concentration of the virus suspension. Compared to filtration (1.7 ± 0.4 × 10^6^ PFUe/µL), precipitation (4.5 ± 1.5 × 10^6^ PFUe/µL) and ultracentrifugation (3.4 ± 0.6 × 10^6^ PFUe/µL) resulted in significantly higher viral concentrations (unpaired, two-tailed t-test, p < 0.05).

We also tested ultracentrifugation through a linear sucrose gradient as a second step to obtain highly concentrated virus suitable for use as an antigen for immunization or cryo-electron microscopy (cryo-EM). Following the second ultracentrifugation, we recovered 57% of the input RV-C15a at a high concentration (5–7 × 10^6^ PFUe/µL) in 1 mL of the 25–30% sucrose fraction.

While this purified virus preparation is directly suitable for infection and immunization, sucrose must be removed for use in cryo-EM studies. We tested two methods to remove sucrose from the virus suspension: overnight dialysis in NTE buffer compared to resuspending the virus in 0.1% BSA followed by an additional ultracentrifugation through a 30% sucrose cushion. While dialysis recovered more virus particles compared to ultracentrifugation (55% of the starting amount vs. 24%, respectively), virus concentration after dialysis was lower (3.0 × 10^6^ PFUe/µL vs. 3.5 × 10^7^ PFUe/µL).

We evaluated an alternative purification protocol involving lipase treatment of clarified RV-C15a cell lysates (30 min, 37 °C) to reduce cellular debris, followed by ultracentrifugation through a 30% sucrose cushion. This streamlined approach caused only a modest 2.2-fold reduction in viral yield (Figure 5B), while improving infectivity, as evidenced by increased progeny yields at 24 hpi (1.7-fold) and 72 hpi (1.6-fold) (Figure 5C). Importantly, it also required less processing time compared to dialysis or additional rounds of ultracentrifugation. Visualization by negative-stain electron microscopy revealed that lipase treatment effectively reduced lipid membrane fragments (Figure 5D,E).

4. Discussion

Among the three RV species, RV-C remains the least characterized, with limited published literature and an incomplete understanding of its biology. There are established protocols for producing and purifying RV-C isolates via reverse genetics and ultracentrifugation [21,24], but these methods do not produce viral suspensions of sufficient concentration and purity for some downstream uses, such as structural biology. This study aimed to optimize the large-scale propagation and purification of RV-C using RV-C15a, a clinically relevant isolate adapted to a genetically modified HeLa-E8 cell line expressing the CDHR3 receptor.

Despite the successful adaptation of RV-C15a to adherent HeLa-E8 cells, viral yields in suspension cultures were significantly lower than those achieved with RV-A and RV-B strains. RV-C15a binding to HeLa-E8 cells was significantly reduced compared to RV-A16. This suggests that engineering RV-susceptible cell lines with greater CDHR3 expression could enable higher RV-C yields. RV-C binding and entry were comparable between adherent and suspension HeLa-E8 cultures, suggesting that post-entry events, such as endosomal trafficking, uncoating, or replication may limit RV-C propagation in suspension cultures, thus constraining their utility for large-scale RV-C production.

Therefore, we optimized RV-C15a propagation in adherent cultures. Our findings demonstrate that cells that are plated for just a few hours support peak virus binding and replication, offering a practical advantage for reducing time and streamlining virus production. On the other hand, prolonged (≥48 h) culture of adherent cells leads to a decline in surface CDHR3 expression thus reducing the availability of viral entry receptors. This decline likely restricts efficient viral spread in HeLa-E8 cells at later timepoints, limiting secondary rounds of infection and overall cytopathic effect progression. Additionally, increasing calcium concentrations modestly enhanced progeny yields, perhaps by supporting a CDHR3 conformation that facilitates virus binding [12]. Thus, optimizing conditions for adherent cell cultures can help boost viral yields.

We tested several purification strategies to obtain high-quality RV-C preparations suitable for various downstream applications. Filtration and precipitation methods were evaluated as alternatives to ultracentrifugation. While both approaches can concentrate virus particles, they are limited by relatively low purity and, in the case of filtration, reduced virus concentration. Despite these limitations, both methods are relatively quick and straightforward, and can produce virus suspensions suitable for use as inoculum. Comparing these two procedures, precipitation proved to be a more practical option for processing larger volumes of virus-containing lysate. Importantly, the resulting virus suspension can be further purified and concentrated by ultracentrifugation, enhancing its suitability for more demanding applications.

Concentrated suspensions of high purity are needed for use as an antigen or structural studies using cryo-EM. Based on our results, our preferred protocol is to treat clarified cell lysates with lipase to remove lipid contaminants before purification (Figure 6). Lipase treatment was previously reported to improve the purity of RV-A2 preparations without compromising virus yield and infectivity [25,26]. Lipase hydrolyzes ester bonds in phospholipids, effectively degrading lipid membranes and facilitating the removal of membrane fragments such as those derived from exosomes. Lipase treatment prior to ultracentrifugation improved RV-C15a recovery and reduced processing time, compared to serial ultracentrifugation over a sucrose cushion followed by a linear sucrose gradient (Figure 6). Lipase treatment did not negatively impact RV-C15a infectivity, consistent with previous observations for RV-A2. If sucrose removal is necessary, we prefer ultracentrifugation to dialysis because it yields higher virus concentrations.

We would like to acknowledge some limitations of our study. This study focused on a single RV-C isolate (C15a) with high binding and replication efficiency in HeLa-E8, and these methods could differ for other RV-C types with different replication kinetics or receptor affinities. However, we have adopted these protocols for ongoing experiments with RV-C41 and RV-C2, suggesting that these methods are broadly applicable for multiple RV-C types. Additionally, we did not test anion-exchange chromatography [27], which has shown promise for reducing contaminants in preparations of RV-A2, nor did we assess alternative filtration methods such as diafiltration or tangential flow filtration, which could enhance scalability and enable higher concentration factors. Future research could explore developing new RV-permissive cell lines with increased and more stable surface CDHR3 expression, which could improve RV-C binding and replication in adherent and suspension cultures.

In conclusion, our study establishes a robust system for the large-scale propagation and purification of RV-C15a using adherent HeLa-E8 cells. We found that initiating infection shortly after cell seeding can reduce incubation time, and calcium supplementation modestly improves yields in adherent cultures. Notably, the lipase-based purification method provides a rapid and effective alternative to traditional gradient ultracentrifugation, yielding highly pure virus suspensions suitable for structural and immunological applications. These protocols will facilitate deeper investigations into RV-C biology and accelerate the development of targeted therapeutics and vaccines.