An HPLC-SEC-based rapid quantification method for vesicular stomatitis virus particles to facilitate process development

Abstract

Virus particle (VP) quantification plays a pivotal role in the development of production processes of VPs for virus-based therapies. The yield based on total VP count serves as a process performance indicator for evaluating process efficiency and consistency. Here, a label-free particle quantification method for enveloped VPs was developed, with potential applications in oncolytic virotherapy, vaccine development, and gene therapy. The method comprises size-exclusion chromatography (SEC) separation using high-performance liquid chromatography (HPLC) instruments. Ultraviolet (UV) was used for particle quantification and multi-angle light scattering (MALS) for particle characterization. Consistent recoveries of over 97% in the SEC were achieved upon mobile phase screenings and addition of bovine serum albumin (BSA) as sample stabilizer. A calibration curve was generated, and the method's performance and applicability to in-process samples were characterized. The assay's repeatability variation was <1% and its intermediate precision variation was <3%. The linear range of the method spans from 7.08 × 10⁸ to 1.72 × 10¹¹ VP/mL, with a limit of detection (LOD) of 7.72 × 10⁷ VP/mL and a lower limit of quantification (LLOQ) of 4.20 × 10⁸ VP/mL. The method, characterized by its high precision, requires minimal hands-on time and provides same-day results, making it efficient for process development.

Keywords: HPLC; VSV-GP; analytical SEC; vesicular stomatitis virus; virus particle quantification.

Graphical abstract

Figure 1

SEC separation chromatogram of purified VSV-GP and impurity-rich material (A) Chromatogram of concentrated VSV-GP material (reference material) spiked into impurity-rich material (clarified harvest) on TSKgel G4000PW using optimized buffer conditions. (B) Concentrated VSV-GP material and impurity-rich material were injected separately on the SEC column using the same conditions. UV 280 nm and light-scattering signal at the 90° angle is shown.

Figure 2

Buffer screening, sample stability testing, and analytical SEC column selection for method development (A) Recovery values for virus particles upon the mobile phase optimization. Iteration rounds: I – Tris-based buffer with increasing salt content, II – Tris-based buffer with Arg and pH adjustments, III – Citrate- and phosphate-based buffer testing, IV – Tris-based buffer with combined Arg and salt content adjustment, V – Tris-based buffer with additives (sorbitol, sucrose, DMSO), VI – Tris-based buffer with Arg, salt, and DMSO content adjustments; technical replicates: n = 1 for buffers 1 to 20, n = 3 for buffer 21 (SD < 0.8%). (B) Sample stability optimization using different vials (glass and polypropylene vials) and BSA spiking. The UV 280 nm signal response is the integrated exclusion peak, identified as the virus peak. (C) SEC chromatogram comparison of BSA-spiked reference material for stainless-steel and PEEK version of G4000PW and PEEK version of G6000PW.

Figure 3

Exclusion peak characterization using BSA as sample stabilizer (A) SEC chromatogram of reference material spiked with different concentrations of BSA and non-spiked. The exclusion peaks at 10.2 min of the BSA-spiked samples are entirely overlaying. The non-spiked sample has been in the sample manager for several hours before measurement. (B) UV 260 nm and UV 280 nm absorbance for non-spiked and BSA-spiked reference material and UV260/280 nm ratio; a percentile filter of 50 was used to smooth the ratio curve. (C) RMS-radii for non-spiked and BSA-spiked reference material. Radii were only evaluable for the exclusion peak due to the low light-scattering signal for the rest of the chromatogram.

Figure 4

VP size influence on the exclusion peak and VP sizing (A) SEC exclusion peak of VSV-GP and variants of different genome sizes. Normalized light scattering and radius of gyration. (B) Radius of gyration for VSV-GP (Variant 1) and larger size variants (Variants 2 to 6) over variant genome size. The radii were measured at the peak maximum of the exclusion peak.

Figure 5

Calibration curve Diluted reference material was analyzed using BSA as the sample stabilizer to obtain measurement points for the calibration curve. The UV 280 nm integrated area of the exclusion peaks was used as signal response. Three independent runs in triplicate for eight concentration levels were obtained. (A) Exemplary exclusion peaks of the concentration levels used for the calibration are shown in comparison. (B) A 1/Y-weighted linear regression was used to fit calibration curve. Only mean values are shown due to non-displayable RSD <2% and for the lowest concentration RSD <4.5%.

Figure 6

Offline fraction VP quantification of a CEX purification step Clarified harvest material of a VSV-GP upstream process was applied in bind-and-elute mode on a CEX monolith column. Fractions of load, wash, elution, and CIP were collected and quantified offline by the presented HPLC-SEC method and qPCR. (A) Chromatogram of CEX run showing online UV 280 nm signal and online conductivity signal. The UV signal saturated at 3.0 AU. (B) Fraction quantification results for HPLC-SEC and qPCR. Due to low numbers for the load phase (and CIP for HPLC-SEC) and scale limitations, signals are plotted close to zero for these phases.