Rapid AAV Concentration Determination Using Size-Exclusion Chromatography with Fluorescence and UV Dual Detection

Importance of the topic

Quantifying capsid concentration in adeno-associated virus (AAV) preparations is critical for gene therapy research, process development and quality control. Traditional ELISA and UV denaturing methods can be time-consuming, require significant sample volumes and reagents, and exhibit limited precision. A rapid, sensitive and low-volume chromatographic approach can accelerate workflows, improve data consistency and support high-throughput screening of AAV batches.

Objectives and overview of the study

This application note describes a two-minute size-exclusion chromatography (SEC) method with dual UV (260/280 nm) and intrinsic protein fluorescence (FLR) detection for direct AAV8 capsid concentration determination. Key goals were minimal sample preparation, low injection volume (1 µL), broad dynamic range (2.6 × 10¹¹ to 2.0 × 10¹² Cp/mL), and correction for empty/full capsid ratios using on-line UV absorbance data.

Methodology and principles

The approach employs a short guard column packed with 1.7 µm BEH particles (125 Å) to buffer-exchange AAV into mobile phase and exclude capsids from pore volume, minimizing dispersion. Samples of AAV8 with known empty (AAV8-Empty) and full (AAV8-Full) capsid fractions were mixed over a range of mole fractions (ΧFull). SEC-UV peak areas at 260 nm and 280 nm were used to estimate ssDNA content (empty/full ratio) via the A260/A280 ratio. Intrinsic tryptophan fluorescence (excitation 280 nm, emission 350 nm) provided capsid protein signal that is less sensitive to DNA content and interfering chromophores. Empirical response factors (RF/E) for UV and FLR were derived from linear fits of normalized peak areas vs. ΧFull, enabling correction of measured signals for mixed sample composition.

Used instrumentation

• ACQUITY UPLC H-Class PLUS Bio system
• ACQUITY UPLC Protein BEH SEC Guard Column, 4.6 × 30 mm, 1.7 µm, 125 Å
• ACQUITY UPLC PDA detector, 260 nm & 280 nm, 10 pts/sec
• ACQUITY UPLC FLR detector, Ex 280 nm/Em 350 nm, 10 pts/sec
• Empower 3 chromatography software

Main results and discussion

• Calibration was linear for capsid concentrations from 2.6 × 10¹¹ to 2.0 × 10¹² Cp/mL with both UV and FLR detection.
• The FLR response factor (RF/E,FLR ≈ 0.875) is closer to unity than UV RF/E,280 (2.98) and RF/E,260 (6.81), reducing dependence on precise empty/full estimation.
• FLR exhibited ~10-fold higher signal-to-noise versus UV, enabling lower detection limits.
• On-line A260/A280 absorbance ratios correlated with ΧFull (R² > 0.999), allowing empirical, concentration-independent estimation of DNA content.
• Corrected FLR measurements yielded more accurate and precise capsid concentrations (average duplicate variance 0.67%) than corrected UV at 280 nm (1.27%).

Benefits and practical applications

• Rapid, two-minute analysis with minimal sample (1 µL) and preparation
• Sensitivity down to ~2 × 10¹¹ Cp/mL
• High throughput suitable for QC and process monitoring
• Reduced interference from free DNA, surfactants and other chromophores
• On-line empty/full ratio estimation without extra assays

Future trends and potential applications

• Extension to other AAV serotypes by determining specific RF/E values
• Integration with automated sample preparation and purification workflows
• Use of longer or larger-pore SEC columns to resolve co-eluting impurities
• Application in multivariate monitoring of viral vector manufacturing
• Adaptation to other viral or nanoparticle quantification challenges

Conclusion

The SEC-UV-FLR method delivers a fast, low-volume, dual-detection workflow for accurate AAV8 capsid quantification across a wide concentration range. Intrinsic fluorescence detection offers superior sensitivity and lower dependence on DNA content, while on-line UV absorbance ratios enable empirical correction for empty/full capsid mixtures. This approach supports high-throughput, precise analytics in gene therapy development and manufacturing.

References

1. Werling NJ, Satkunanathan S, Thorpe R, Zhao Y. Human Gene Ther Methods 2015;26(3):82–92.
2. Andreasson U, Perret-Liaudet A, et al. Front Neurol 2015;6:179.
3. Sommer JM, Smith PH, et al. Mol Ther 2003;7(1):122–128.
4. Runnels H. CMC Strategy Forum 2017.
5. Wang C, Mulagapati SHR, et al. Mol Ther Methods Clin Dev 2019;15:257–263.
6. Noble JE. Methods Enzymol 2014;536:17–26.