A recombinant antibody toolbox for Dictyostelium discoideum

Objective The amoeba Dictyostelium discoideum has been a valuable model organism to study numerous facets of eukaryotic cell biology, such as cell motility, cell adhesion, macropinocytosis and phagocytosis, host–pathogen interactions and multicellular development. However, the relative small size of the Dictyostelium community hampers the production and distribution of reagents and tools, such as antibodies, by commercial vendors. Results For the past 5 years, our laboratory has worked to promote an increased use of recombinant antibodies (rAbs) by academic laboratories. Here we report our efforts to ensure that Dictyostelium researchers have access to rAbs. Using hybridoma sequencing and phage display techniques, we generated a panel of recombinant antibodies against D. discoideum antigens, providing a useful and reliable set of reagents for labelling and characterization of proteins and subcellular compartments in D. discoideum, accessible to the entire Dictyostelium community.


Introduction
For the past seven decades, the social amoeba Dictyostelium discoideum has been used as a model organism to elucidate the molecular mechanisms underlying numerous eukaryotic cellular processes, notably multicellular development, cell motility, chemotaxis, macropinocytosis and phagocytosis, endocytic vesicle traffic, cell adhesion, caspase-independent cell death, host-pathogen interactions, and microbial infections [1,2]. D. discoideum is a haploid organism, with a relatively small genome (34 Mb) containing around 12,000 predicted proteins. From an evolutionary perspective, D. discoideum diverged from the animal lineage before fungi and yeasts, but after plants. Many cellular functions, absent in plant cells, appeared at this stage, in particular cell motility, chemotaxis, phagocytosis and cell adhesion. Due to the higher rate of evolutionary changes inside the Fungi group, the evolutionary distance between Dictyostelium and human is shorter than that between human and yeast [3]. Many gene products lost in fungi are maintained in D. discoideum, including orthologs of human genes responsible for a variety of diseases [1,4].
Despite its relevance as a model organism, the Dictyostelium scientific community is relatively small, and this can hamper the development of new tools and techniques. One of the most critical tools in cell biology research are antibodies, extensively used for detection and characterization of proteins and cellular components. The mono-and polyclonal antibodies against D. discoideum antigens currently in use have been mostly developed during the 1980s; due to the small size of the community, they are not commercially available [1]. Because mono-and polyclonal antibodies are finite resources, many of these anti-Dictyostelium tools are either not widely available or have already been lost forever. This is a critical situation, and the solution is twofold. On the one hand, the existing monoclonal antibodies produced by hybridomas need to be secured by determining their aminoacid sequences. On the other hand, the past 15 years have seen an explosion of techniques and tools for discovery of recombinant antibodies (rAbs), notably the phage display technique. These new tools allow the community to enrich the list of rAbs against D. discoideum targets. Each time a laboratory engages in the sequencing of a hybridoma or in the discovery and characterization of a new recombinant antibody, it facilitates future work for the whole Dictyostelium research community.
For the past 5 years, we have used hybridoma sequencing and phage display technologies to ensure that more and more Dictyostelium researchers have access to recombinant antibodies [5]. In this study, we describe a panel of recombinant antibodies against D. discoideum antigens, providing a toolbox for detection, characterization and isolation of proteins and subcellular compartments in D. discoideum.
Following the cDNA synthesis reaction, the heavy (V H ) and light (V L ) variable domain sequences were amplified by PCR using degenerate primer sets (500 nM, Integrated DNA Technologies; Additional file 1: The PCR products were column-purified and doubledigested with HindIII-Hf and SacII (NEB #R3104 and #R0157). V L products were additionally treated with BfuI and/or BseRI (ThermoFisher #ER1501 and NEB #R0581), to eliminate the myeloma aberrant light chain from the myeloma fusion partner. Digested products were gel-purified (300-400 bp) and cloned into pBSKvector (GenBank #X52330.1, HindIII-Hf/SacII digested and dephosphorylated). 6-10 clones with the correct fragment size were sequenced to obtain the consensus sequences of the V H and V L domains; definition of the heavy and light chain boundary sequences, based on alignment with germline sequences, was done using the VBASE2 server [7].

Antibody conversion into an scFv-Fc format
A recombinant antibody can be made in several formats and flavors; our format of choice is an scFv (single chain Fv) linked to an Fc moiety by a small AAARS linker (Additional file 1: Fig. S1). The V H and V L sequences, joined by a peptide linker (GGGGS) 3 , were synthesized by Invitrogen GeneArt (ThermoFisher). The synthesized fragments were double-digested with NotI-HF and NcoI-HF (NEB #R3189 and #R3193) and cloned into homemade expression vectors, containing the Fc moieties of mouse (IgG2A, UniProt #P01867) or rabbit (IgG, UniProt #P01870).

Sequencing of monoclonal antibodies and conversion to recombinant antibodies
The production of monoclonal antibodies by hybridoma cell lines was first achieved in 1975 [8] and was a major technological development in biomedical research.
Despite its enormous importance, the technique is not devoid of problems, the most serious being the possibility of losing a particular hybridoma cell line-because either the cells die or they are unable to regrow. In addition, hybridoma cell lines are genetically unstable and it is not uncommon to see a cell line lose the ability to produce antibodies [9].
One easy solution is to sequence the antibody genes directly from the hybridoma cells. This guarantees a permanent, inexpensive and flexible storage, since all the information is contained in an electronic file, and/or in a DNA plasmid [10]. Once the sequence information is available, it can easily be stored and propagated, and it can also be used to produce the corresponding recombinant antibody. A recombinant antibody has the flexibility to be produced in a variety of systems (in bacteria, fungi or mammalian cells, to give only a few examples) and formats, in particular with any protein/peptide tag or Fc region (the binding site for secondary reagents) (Additional file 1: Fig. S1). As a consequence, a mouse monoclonal antibody produced by hybridoma cells can easily be turned into a rabbit-like antibody, or decorated with affinity tags or fluorochromes (such as Myc or GFP).
In our laboratory, we have a collection of more than 70 hybridoma cell lines producing antibodies against diverse Dictyostelium antigens, created by us and others over the last 30 years (for a complete list, see Additional file 1: Table S2). Using well-established techniques of hybridoma sequencing, we determined the antibody sequence for 26 of them (Table 1). Twelve were converted to recombinant antibodies. For this, the two variable regions were fused to create a single-chain variable fragment (scFv), which was then joined to a mouse or rabbit Fc moiety (Table 1 and Additional file 1: Fig. S1).
To characterize the converted rAbs, immunofluorescence and/or western blot experiments were performed, comparing the original IgG produced by the hybridoma Table 1 List of sequenced hybridomas; highlighted in bolditalic, the antibodies converted to rAbs a ABCD nomenclature (https ://web.expas y.org/abcd/); the ABCD database is a manually curated repository of sequenced antibodies [47] b "Original references" correspond to the first descriptions of a monoclonal antibody. "rAb references" describe the characterization of the converted rAbs c The recombinant versions of these antibodies were not produced efficiently in our hands  In (b), the converted rAb has a mouse Fc, and double immunolabelling could not be performed; staining patterns are identical. For more images and experimental details, please refer to the rAb references indicated in Table 1 cells and the derived rAb. Ten of the converted rAbs recapitulated the immunolabelling of the original antibody (Table 1 for references). In two cases (AK567 and AJ515), the converted rAbs had production yields too low to be usable and thus failed to generate a specific labeling.

Selection of new recombinant antibodies by phage display
Strategies to select antibodies using completely in vitro methodologies first appeared in the mid-1980s [48]. However, most academic laboratories do not have access to the phage display technology for discovery of new antibodies. Since 2013, the Geneva Antibody Facility selects, produces and characterizes recombinant antibodies for the academic community, in an open-access, not-for-profit and collaborative spirit [5].
Over the past years, we have selected 184 antibodies against 41 different D. discoideum proteins ( Table 2) using the phage display technique. The in vitro selection has been done with a wide variety of antigen molecules, including small peptides, protein domains or fragments attached to a GST carrier, full-length proteins, and even subcellular compartments and cells. All antibodies (using the already mentioned scFv-Fc format) specifically recognize the target used for phage display discovery (a peptide, a protein fragment, a full protein) by ELISA. It should however be emphasized that an antibody directed against a peptide may fail to recognize the corresponding sequence in the full folded protein. We did not characterize in depth all the antibodies discovered; further characterization should be performed and reported by the end-user scientists themselves. We strongly encourage them to publish positive as well as negative results [5]. The recognition of an epitope by an antibody is heavily influenced by the folding state of the protein; proteins are mostly folded in ELISA or immunofluorescence experiments and largely denatured in western blot or immunohistochemistry. Consequently, antibodies that perform well with one technique often perform poorly in others [49]. It is thus of paramount importance to determine the performance of each antibody in different contexts and for different techniques.

Conclusions
Given the relatively small size of the Dictyostelium scientific community, the majority of the tools and reagents are developed by the researchers themselves, and not commercially available. This is notably the case for antibodies. In addition, almost the totality of these reagents are polyclonal antibodies produced by immunizing rabbits, or monoclonal antibodies produced by mouse hybridoma cells. While polyclonal antibodies are ill-characterized reagents that should be completely phased out, monoclonal antibodies can be irretrievably lost. Since 2015, many key opinion leader scientists have published calls to employ only recombinant antibodies [9,66].
Here, we describe the efforts of the Geneva Antibody Facility to develop recombinant antibodies against a panel of Dictyostelium targets, either by sequencing existing hybridoma cell lines, or by developing new antibodies using the phage display technique. We hope that these efforts will facilitate work and increase reproducibility in this scientific community. We further hope that it will encourage others to take part in this common enterprise by (i) sequencing their own hybridomas and depositing the sequences in the ABCD database (https ://web.expas y.org/abcd/; [47]); (ii) selecting and characterizing antibodies to new Dictyostelium proteins; and (iii) publishing the results obtained with these antibodies.

Limitations
Antibodies developed by us are not always characterized in depth, as this is out of the scope of our laboratory. We strongly urge the end-user scientists themselves to characterize and publish any positive and/or negative results, to determine the antibody efficacy in different contexts and with different techniques.
Additional file 1: Fig S1. Schematic representation of the conversion of an IgG into an scFv-Fc. (A) An IgG is composed of 4 chains: 2 heavy (H) and 2 light (L) chains, made of a constant (C) and a variable (V) domain. The F V (variable fragment) consists of two chains (V H and V L ) and is the region responsible for antigen recognition and binding; thus, it is the region of interest for sequencing. (B) An scFv is made of the V H and V L variable chains joined by a peptide linker (GGGGSGGGGSGGGGS). (C) An scFv-Fc is an scFv molecule fused to an Fc region; the Fc can be of any desired species (rabbit, mouse, human, guinea pig), and it is the region where secondary reagents bind to. Table S1. List of degenerate primers used for hybridoma sequencing. Table S2. Collection of Cosson lab's hybridoma cell lines producing antibodies against Dictyostelium antigens.