https://doi.org/10.3390/ijms232112715
“Ion-exchange chromatography (IEC) provides an alternative separation method, which is frequently used for the determination of empty and full AAV capsids. IEC allows more parameters to be optimized in order to enhance the chromatographic resolution of AAV populations compared to SEC, such as buffer medium, pH, temperature, flow rate, salt concentration and composition, gradient slope and column properties. The separation principle is based on the interaction of a positively charged (anion-exchanger) or negatively charged (cation-exchanger) stationary phase with complementarily charged AAV capsids [21].”
”
For the development of an IEC-MALS method, AAV8 was the serotype of choice as it has been reported by Lock et al., to be separatable into filled and empty capsid populations using ion-exchange chromatography [35]. However, to determine whether an anion-exchange (AEX) or cation-exchange (CEX) column was needed, the isoelectric point of the selected serotype was measured using a capillary isoelectric focusing (cIEF) technique. Because the measured pI was 7.4 (net capsid), a CIMac AAV full/empty analytical column (anion-exchanger) was used for the development of the IEC assay. For sample binding and elution, buffers containing 20 mM Tris (pH 8.5, buffer B) and 20 mM Tris + 120 mM MgCl2 (pH 8.5, buffer E), respectively, were selected. The chromatographic separation was carried out using the gradient described in Supplementary Table S1. Due to the difference in their overall negative charge attributed to the encapsidated nucleic acid, empty AAV capsids elute earlier from the column than filled AAV capsids when increasing the salt concentration of the buffer. For sample detection, the AEX column was coupled to a UV detector, a static and a dynamic light scattering detector, which allowed the determination of the absolute molar masses of the protein and nucleic acid, the hydrodynamic radius, the radius of gyration and the polydispersity of the afore-separated empty and filled AAV capsid fractions. The capsid titer and the full-to-total ratio of the sample are additionally assessed. A schematic overview of an IEC-MALS method is given in Figure 1. Unlike SEC-MALS, which uses the UV absorption and the differential refractive index (dRI) detection for the calculation of the above-mentioned parameters, IEC-MALS demands dual wavelength UV-absorption detection, as an RI detector cannot be used when applying salt gradients due to a change in the refractive index with increasing salt concentration (dn/dc).
Figure 1. Schematic illustration of the IEC-MALS method. The AAV sample is loaded onto the anion-exchange column, eluted with a salt gradient containing MgCl2 and detected with multi-angle light scattering and UV detectors prior to data analysis using ASTRA software.
2.1. IEC-MALS Development
Like SEC-MALS, IEC-MALS requires the calibration and normalization of the MALS detector prior to sample analysis. Toluene was used as the standard for the calibration of the detector at 90°, while the remaining photodiode detectors were normalized to the 90° detector using bovine serum albumin (BSA), a monodisperse, isotropic scatterer [28]. Because the UV-Vis and light scattering detectors were operated in series, the resulting chromatograms showed shifts in the retention times as the sample is not detected simultaneously. As the sample progresses through the detectors, it becomes more diluted, and broader peaks are observed. To correct for these variations, an alignment and band broadening correction of the UV and LS signals were performed.
”
”
Because AUC is used as the standard analytical technique for the quantification of empty and filled capsids as well as other AAV subspecies, results obtained by IEC-MALS were compared to AUC data regarding the full/empty (F/E) ratio [17,19]. Unlike IEC-MALS, AUC can resolve AAV capsids containing a partial genome from empty and full ones; however, it is a more time-consuming technique with low sample throughput. Another drawback is the need for large sample volumes and high capsid titers [17]. Prompted by this, we developed an IEC-MALS assay which provides a faster and simpler alternative for the determination of the F/E ratio with the advantage of receiving additional information (hydrodynamic radius, radius of gyration, polydispersity and absolute molar mass of protein and nucleic acid) about both AAV populations in one single measurement. Therefore, two AAV8 samples comprising mostly empty (meC) and mostly filled AAV capsids (mfC), respectively, were mixed at different ratios to obtain fractions of various F/E content ranging from 28% to 96% F/E (capsid titers: 1.0 × 1013 cp mL−1). In Figure 2, an excellent linear correlation between data obtained by IEC-MALS (measured %filled) and data generated by AUC (expected %filled) is observed, with a coefficient of determination (R2) of 0.9968, suggesting that IEC-MALS can be used alternatively for the determination of the F/E ratio. Because ion-exchange chromatography does not provide any information on subpopulations due to a lack of chromatographic resolution, data from AUC for partially-filled and filled particles were added up for the comparison with IEC-MALS data.
Figure 2. Linear correlation of measured % filled AAV capsids using IEC-MALS and expected % filled AAV capsids by AUC.
“