A robust and efficient alluvial filtration method for the clarification of adeno-associated viruses from crude cell lysates

Viruses are not only pharmaceutically relevant for the production of classical vaccines (Wolff and Reichl, 2011). They are also essential as vectors for gene therapy because they have perfected their mechanisms for introducing genetic material into host cells during evolution (Nonnenmacher and Weber, 2012). One of the most promising viruses in gene therapy is the adeno-associated virus (AAV) (Berns and Muzyczka, 2017, Dhungel et al., 2021). To date, 263 gene therapy clinical trials are ongoing using AAV as a vector (John Wileys and Sons, 2022). With Luxturna, Zolgensma and Hemgenix, three AAV-based gene therapeutics have already been approved by the U.S. Food & Drug Administration (FDA).

AAV was discovered in the 1970s by Atchison et al. (1965) and was named adeno-associated virus as it could only replicate in the presence of an adenovirus (AdV). With a diameter of ~25 nm (Grieger and Samulski, 2012), AAV is a small non-enveloped (Balakrishnan and Jayandharan, 2014) single-stranded DNA (ssDNA) virus (Srivastava, 2016, Wang et al., 2019). It belongs to the Parvoviridae family (Ayuso, 2016) and the genus of Dependoparvovirus (Galibert and Merten, 2011). As of today, 13 human serotypes (AAV1-13) can be distinguished (Beloukhova et al., 2022). They differ in their capsid proteins (Weinberg et al., 2013) and exhibit tropisms for a wide variety of tissues and cell types (Balakrishnan and Jayandharan, 2014, Wang et al., 2019). Of all serotypes, AAV2 is the most common and the first to be used for gene therapy applications (Arnold et al., 2006, Lu, 2004).

Recombinant AAV (rAAV) vectors show the same capsid composition as the respective wild type AAV (wtAAV) serotype. However, besides the inverted terminal repeat (ITR) sequences, the rAAV genome does not encode any AAV protein-coding sequences (Wang et al., 2019). Usually, rAAV vectors lack the ability to integrate their genome into that of the host cell (Naso et al., 2017); integration into the genome is therefore only a rare event (Schultz and Chamberlain, 2008). The viral genome introduced is consequently present in episomal form in the host cell, thereby increasing safety when rAAV vectors are used on patients (Grimm and Kay, 2003). Furthermore, a major benefit is the non-pathogenicity and non-toxicity of AAV as no human disease caused by AAV is known so far (Grieger et al., 2016, Kantor et al., 2014).

Besides other methods, rAAV vectors are mostly produced by transfection of human embryonic kidney 293 (HEK293) cells with plasmids, consequently providing the cells with all the required information for packaging of a transgene into rAAV particles (Blessing et al., 2019, Liu et al., 2000, Sena-Esteves and Gao, 2020). Depending on the serotype, AAV is more or less expressed intracellularly (Vandenberghe et al., 2010), which is why the transfected HEK293 cells need to be lysed for AAV harvest (Wright, 2008).

A major bottleneck is downstream processing (DSP) of AAV for efficient purification of AAV from large volumes of cell lysate (Hebben, 2018, Marichal-Gallardo et al., 2021). DSP can account for a large part of production costs (Rooij et al., 2019). A typical DSP procedure for purification of AAV consists of a clarification step, generally by filtration, followed by capture chromatography, such as affinity (AC) or ion exchange chromatography (IEC). Empty AAV capsids containing no vector genome are subsequently removed by IEC or, at small laboratory scale, by density gradient centrifugation. Final steps are concentration by tangential flow filtration (TFF) and sterile filtration using membranes with a pore size of ≤0.22 µm (Hebben, 2018).

Compared to other techniques, such as centrifugation, filtration is often the method of choice for clarification of cell culture lysates. It is the most straightforward and economical process (Raghavan et al., 2019) due to lower personnel costs. Filtration separates particulate contaminants, such as cells, cell fragments and cell organelles from the crude lysate. AAV passes through membrane pores and is thus recovered in the clarified permeate. However, single-stage filtration with 0.2 µm membranes is quite challenging since AAV is, depending on the serotype, largely expressed intracellularly and not secreted into the medium (Vandenberghe et al., 2010). As such, AAV harvest, along with some other viral vectors (Merten et al., 2014), requires cell lysis and results in a difficult-to-filter cell lysate (Srivastava et al., 2021). Instead, filtration is generally performed using depth filters, membranes with a pore size of ≥0.45 µm, as two-stage filtration (Raghavan et al., 2019) or, at laboratory scale, by centrifugation with subsequent microfiltration (Marichal-Gallardo et al., 2021).

As clarification of large volumes by centrifugation or methods coupled with centrifugation are associated with high investment costs, other techniques are preferred for this process step (Besnard et al., 2016). Membrane filtration, on the other hand, is more suitable for scale-up (Besnard et al., 2016), although the membranes quickly become clogged when used for clarification of cell lysates during single-stage filtration (Labisch et al., 2021).

A possibility for preventing filter clogging and increasing filter capacity during clarification of AAV by single-stage filtration is to use filter aids, such as diatomaceous earth (DE). DE is a porous powder made from diatom skeletons (Ivanov and Belyakov, 2008, Wang et al., 2012). It is usually used for clarification of cell-culture-derived antibody harvests and for fractionation of blood plasma (van der Meer et al., 2014), as well as in the production of beverages (Martinovic et al., 2006). Furthermore, it is used as a component in depth filters (Besnard et al., 2016) and was recently shown to improve the clarification of lentiviruses (LV) (Labisch et al., 2021). Mixing DE with cell lysate and subsequent filtration, known as alluvial or body feed filtration, creates a permeable filter cake that retains the majority of cell fragments and prevents rapid filter clogging (van der Meer et al., 2014). Compared to LV, which range from 80 nm to 120 nm (Merten et al., 2016) and are enveloped RNA viruses (Clements and Zink, 1996), AAV is much smaller with about 25 nm, and is a non-enveloped ssDNA virus (Balakrishnan and Jayandharan, 2014, Srivastava, 2016). Consequently, in consideration of size and surrounding structure, AAV cannot be clarified by filtration with DE under the same process conditions as LV. Based on our knowledge to date, clarification of AAV by single-stage filtration with 0.22 µm membranes using DE as a filter aid has not yet been investigated or described.

Therefore, this study investigates the impact of DE as a filter aid during clarification of AAV2, AAV5 and AAV8 to develop a robust, simplified and effective clarification process for AAV crude cell lysates. Various AAV serotypes were studied as they differ in their aggregation and adsorption characteristics (Hebben, 2018). For this reason, we compared the turbidity reduction, loss of AAV, filter capacities and manual handling time of diverse clarification techniques.

留言 (0)

沒有登入
gif