A novel protocol for in‐depth analysis of recombinant adeno‐associated virus capsid proteins using UHPLC–MS/MS

Abstract

RationaleIn‐depth characterization of the three capsid viral proteins (VPs 1, 2, and 3) of adeno‐associated viruses (AAVs) is immediately needed to ensure the consistency in gene therapy products and processes. These proteins are typically present at very low concentrations in matrices containing high concentrations of excipients and salts. Thus, there is a need for convenient methods for sample preparation before proteomic analysis. The aim of this study was to meet this need by developing a fast, reliable approach for isolating VPs in a manner enabling their efficient digestion and in‐depth characterization using liquid chromatography–mass spectrometry (LC–MS).MethodsVPs from Anc80 were precipitated with different organic solvents, and the resulting precipitates were dissolved in either sodium deoxycholate (SDC) and N‐dodecyl‐beta‐D‐maltoside (DDM) or guanidine hydrochloride (Gu‐HCl). The peptides obtained by the following enzymatic digestion by either trypsin or Asp‐N were analyzed using LC–MS/MS.ResultsWe found that precipitation with chloroform/methanol/water results in fast, efficient preparation of VP samples, allowing 100% and 99.2% amino acid sequence coverage of VP1 for trypsin and Asp‐N digestion, respectively. This also allowed complete sequence confirmation of VP1, VP2, and VP3 of Anc80, as well as characterization of the amino acid sequences of the N‐ and C‐terminal regions of each VP, together with their post‐translational modifications (PTMs).ConclusionsThe presented method enables fast, reliable, and relatively cheap sample preparation for identifying AAV serotypes and characterizing the heterogeneity of capsid viral proteins, including their PTMs.

Proteomic Analysis of Adenovirus 5 by UHPLC-MS/MS: Development of a Robust and Reproducible Sample Preparation Workflow

“…The method essentially involves concentration of virus particles using a centrifugal filter with an appropriate cutoff membrane (10 kDa). This is followed by the use of organic solvents to disassemble the particles into their structural proteins and DNA, with precipitation of the VPs at the interface between organic and aqueous phases. , Sample matrix components and DNA partition into either the aqueous or organic phase and are removed prior to dryness of the protein precipitate. In this new approach, tertiary structures of the proteins are disrupted after precipitation and a further denaturation step with common chaotropic reagents (e.g., urea or guanidine hydrochloride) is not necessary.…”

“…After digestion of the proteins, it can be quite problematic to introduce a sample directly into an MS system, as a high concentration of SDC (1%) severely suppresses electrospray ionization and interferes with chromatographic separation of the peptides. Therefore, in previously published protocols, SDC is degraded after the addition of TFA and precipitates are removed by centrifugation. , This step is a major bottleneck when low volumes of material are available. Under these conditions, the supernatant cannot be fully separated from the pellet, resulting in significant sample loss (up to 21%), as shown by the results of precipitating SDC in mAb A solution, washing the pellet with water, then subjecting both the supernatant and washed pellet to LC-MS/MS analysis (Supporting Figure 1).…”

“…Digested VPs of AdV5 were analyzed according to the developed method, as illustrated in Figure and briefly described in the Material and Methods section. Due to the vast differences in relative abundances of AdV5 structural proteins (e.g., Hexon and pTP account for 59.5 and 0.1% of the total protein mass, respectively), several modifications of our previously reported LC-MS/MS setup were required . First, in silico trypsin digestion of VPs of AdV5 resulted in the generation of more than 350 peptides with four or more amino acids, so a longer gradient was needed to separate and identify the peptides.…”

“…Acetylation at N-terminals of some VPs (e.g., pentone base and fiber) cannot be quantified by the approach presented here, as these proteins have high arginine and lysine frequencies in their N-terminal amino acid sequences, so these peptides cannot be retained and detected, as discussed above. However, the use of a second protease (e.g., Asp-N) might provide the necessary information, as we previously described …”

“…VPs were precipitated with chloroform/methanol/water as previously described. , Briefly, 800 μL of methanol, 200 μL of chloroform, and 600 μL of water were sequentially added to 200 μL of concentrated sample with short vortex-mixing and fast centrifugation steps (10 s at 14,000 g ) following each addition. The protein precipitate appeared at the interface as a white layer between upper and lower phases.…”