**4. Purification**

Downstream purification is the major determinant of cost of goods and monoclonal antibody therapies are currently regarded as some of the most expensive drugs in the world [45]. For SIgA to enter the immunotherapy market, the cost of downstream processing must be comparable to that of IgG. However, there are hurdles associated with SIgA purification which must be addressed beforehand.

Firstly, there is no established gold standard purification resin for SIgA or IgA antibodies. Protein A (the well-established resin used for industry scale IgG purification) cannot be used for SIgA due to the lack of suitable binding sites on the antibody complex. Alternative purification resins that can be used for circumventing this are jacalin [46], SSL7 [47] or generic anti-kappa affinity resins such as protein L [48]; each method is associated with different limitations.

Jacalin is an O-linked glycan-specific lectin derived from jackfruit which is used to purify monoclonal IgA1, and can demonstrate purification efficiency similar to IgG and protein A [49]. Specifically binding to α-D-galactose on the antibody surface, jacalin will not bind to IgA2 or IgA1 with modified glycosylation (e.g. without the O-glycosylated hinge), which may be a benefit or a hindrance depending on the desired product [49]. Inconveniently, Jacalin will also bind to host cell proteins exhibiting O-linked glycosylation, which presents a problem if the target antibody is not present at high concentrations [46].

Protein L is a bacterial cell wall molecule with high binding affinity to certain kappa light chain sequences [48]. This can purify any IgA subtype provided the target sequences are present in the light chain. Unfortunately, protein L will also bind SIgA assembly intermediates and fragments, single light chains, along with monomers and dimeric IgA. This complicates downstream processing since the eluted sample must be further processed (by size exclusion chromatography, for example) to isolate fully formed SIgA. Other purification options, such as CaptureSelect® using llama antibody Fab fragments against single alpha or kappa chain, also suffer from this limitation [50].

Antibody engineering efforts directed at facilitating IgA purification have highlighted the sequences required for affinity purification binding. For example, engineering SIgA light chains to gain protein L binding ability has been described [48]. In addition, murine IgA was made purifiable by SSL7 (another IgA binding protein derived from bacteria) using a 2-amino acid mutation in the Fc region [51], showing that engineering novel purification methods of IgAs are feasible and potentially able to simplify downstream processing.
