Monoclonal antibodies and derivatives are currently the fastest growing class of therapeutic molecules . Their inherent promise to minimize side effects by selectively targeting specific target cells has fuelled their development, leading to several FDA-approved antibody therapeutics so far and many more in the pipeline. Although unmodified mAbs proved their worth, the conjugation of effector molecules (like toxins, drugs, radionuclides,…) to mAbs broadened their therapeutic potential, especially in the domain of cancer therapeutics. Besides cancer, other diseases could also benefit from antibody-directed therapies, the only prerequisite being the identification of a receptor exclusively expressed on those immune cells involved in the induction of pathology. In this respect, sialic acid-binding immunoglobulin-like lectins (siglecs) are compelling candidates for therapy, as they display very restricted expression patterns on subsets of immune cells and may regulate immune cell functions. Furthermore, siglecs are endocytic receptors allowing therapeutic agents conjugated to a mAb to be carried efficiently into the cell [27, 28].
Sialoadhesin or Siglec-1 is expressed on cells of the monocyte/macrophage lineage, notably on subsets of resident tissue macrophages and inflammatory monocytes/macrophages [15, 27]. A recent report describes the expression of Sn on human mature DCs treated with LPS in vitro, suggesting that Sn may be present on mature DCs during inflammation in vivo as well. Not only have Sn+ monocytes/macrophages been described in several diseases like inflammatory and autoimmune disorders as well as viral infections, they also appear to play a role in the initiation of an adaptive immune response as recently shown by different independent research groups and nicely reviewed by Martinez-Pomares and Gordon . Together, this makes these Sn+ cells not only attractive targets for cell-directed therapies, but also an appealing target for vaccination. In our previous study, we developed immunoconjugates by the chemical linkage of the model antigen HSA or a toxin to the pSn-specific mAb 41D3 . Although these immunoconjugates proved efficient for boosting immune responses and killing pSn-expressing cells respectively, the chemical linkage of the cargo to a targeting antibody has many disadvantages. First of all, chemical coupling procedures rely on the presence and distribution of reactive groups, like e.g. primary amines on lysine residues, that can be located in or near the antigen-binding region, which upon coupling might result in partial or complete loss of the antibody’s affinity for the target antigen. Secondly, because of the large number of reactive groups present in antibody molecules, a typical distribution can be observed of zero to eight molecules per antibody [30, 31], resulting in high variation of the final conjugate. This variation is unwanted, as it may lead to a heterogeneous mixture of components with distinct affinities, stabilities, pharmacokinetics, efficacies, and safety profiles . Moreover, chemical coupling implies that both antibody and cargo to be linked are independently produced and purified, which represents a significant challenge, especially when the cargo is also a biologic. To circumvent these problems, we opted to generate a recombinant form of the pSn-specific mAb 41D3. As shown in this study, this recombinant antibody displays a comparable affinity for pSn compared to the native mAb. In addition, the recombinant mAb also induces pSn endocytosis in primary macrophages, a feature important to allow functionality of antibody-cargo constructs. As protein sequences are attached to the C-terminus of the antibody’s heavy chain, they are less likely to hinder antigen binding by the variable immunoglobulin domains. In addition, each heavy chain will contain only one cargo fused to the C-terminal end. This will result in an antibody with 2 cargos in a defined position and a high intra and inter batch consistency. Furthermore, we could purify the antibody-cargo fusion proteins using standard protein G chromatography, which represents a major advantage compared to chemical coupling in which purification is needed for both cargo and antibody before, as well as after chemical conjugation.
In this study we managed to make genetic fusion constructs of a peptide or a protein linked to our recombinant mAb. Obviously, the recombinant antibody vector does not allow to make genetic fusion constructs with chemical compounds. For vaccination strategies however, this limitation is not expected to pose any problems, as most antigens used in vaccines are protein based. One challenge however would be to ensure correct folding of the antigen upon genetic fusion to the antibody and to maintain this fold during purification procedures. Similarly, immunotoxins can be made using the Sn targeting vector. Although the production of immunotoxins in eukaryotic cells has been limited due to potential toxicity to the producing cells, several independent research groups have reported on the successful production of immunotoxins in mammalian cell lines, including HEK293T [32–34]. In case a specific application would require the chemical linkage to an antibody, e.g. when vaccines are based on glyco-epitopes, a recombinant mAb has some major advantages. It allows addition of specific amino acid modifications to the antibody, which will result in site-specific incorporation of drug molecules through chemical linkage yielding batch to batch consistency of antibody-drug conjugates. Examples of such already implemented modifications are the THIOMABTM technology of Genentech Inc  or the methodology of Axup et al. .
As our future plans include the use of the developed recombinant antibody to target antigens towards pSn-expressing macrophages in vivo, one might be concerned about the immunogenicity of mouse antibodies in pigs. Poderoso et al. previously used mouse mAbs as surrogate antigens in pigs to evaluate the role of Sn in the induction of humoral responses and noticed an enhanced anti-mouse antibody response in comparison with a non-targeting isotype control mAb . The induction of anti-mouse antibodies was however low after primary injection of the mAb, only after a booster vaccination antibody titres rose significantly. Previously, we have observed an enhanced anti-HSA antibody response after a single dose vaccination of HSA coupled to mAb 41D3 without adverse clinical effects . Therefore, in our future experiments, we will use a single dose of rec41D3-antigen to evaluate the protective efficacy of antigen targeting to pSn. If further experiments confirm the applicability of this targeting technology, ‘porcinization’ of the recombinant antibody will be examined to enable prime-booster vaccination schedules.