Next we examined whether one of the cell-free synthesized Fab chains is protected against protease digestion and whether such a protection is only possible if the chain has been synthesized with signal peptide. A protection only for signal peptide bearing Fab in the disulfide insect system would give evidence that it has been specifically translocated into microsomes where it is inaccessible for the protease.
The H-chain of anti-lysozyme Fab synthesized with signal peptide using the template pIX5.0-Mel-HaLys and without signal peptide using pIX5.0-HaLys was subjected to the proteinase K protection assay. To force digestion, vesicles of control reaction samples were lysed by pre-treatment with detergent as described in Materials and methods.
Our results show clearly that the H-chain of anti-lysozyme Fab is protected against protease digestion when synthesized with a signal peptide and completely digested if produced without a signal peptide. In addition, the protection of the signal peptide bearing H-chain is disabled when the vesicles have been lysed by pretreatment with detergent. We conclude that the synthesized protein is specifically translocated into the microsomal vesicles (Figure 4).
Furthermore, to exclude that the microsomes have a nonspecific positive effect on Fab activity, we synthesized the Fab chains without signal peptides and measured Fab activity. As expected, anti-lysozyme Fab expressed without encoded signal peptide is not active in the disulfide insect system (data not shown). In summary, our results indicate that both redox conditions and translocation into microsomes are required for the synthesis of Fab with high activity.
In cell-free expression systems published to date, synthesis of Fab takes place in prokaryotic cell lysates in an environment corresponding to the cytosol of a living cell. Chaperones and PDI necessary for the synthesis of active Fab are supplemented to these systems. In the eukaryotic system described here, the antibody fragment encoding templates contain a sequence for a signal peptide. By means of the signal peptide, the antibody fragments are cotranslationally translocated by a natural mechanism into microsomal vesicles derived from the ER, which are highly active because of gentle preparation of the cell lysate. This kind of synthesis is more similar to the natural synthesis of antibodies in B cells than a cytosolic synthesis. Most likely the microsomal vesicles already contain all the components such as chaperones and disulfide isomerase required for the maturation of antibody fragments, so there is no need for a balanced supplementation of all these components.
In current systems, a certain redox potential can only be maintained for a limited period of time because of enzymatic processes generating a more reducing environment that is disadvantageous for disulfide bond formation. Hence, in order to inhibit these enzymes, such systems are pretreated chemically, for instance, with an excess of oxidized glutathione or iodoacetamide. In the system described here, the redox potential seems to be sufficient for the almost complete conversion of monomeric immunoglobulin chains to dimeric Fab. A possible reason for that seems to be the natural capability of the ER to maintain its redox potential. We assume that this potential is still present in the microsomal vesicles.
Present prokaryotic systems contain endotoxins. Thus, antibody fragments produced in such systems cannot be used directly for applications involving endotoxin sensitive eukaryotic cells, as cytotoxic effects interfere with the measurement of the specific impact of antibodies on cells. Using the eukaryotic endotoxin-free system presented here, a purification step after protein synthesis for removal of endotoxins can be avoided.
So far, mRNA or circular DNA are mostly used as templates for cell-free protein synthesis. With the system described here, linear template DNA produced by PCR can be applied even without prior purification. During the process of template generation, sequences encoding the signal peptide required for translocation and affinity tags are easily added to the open reading frames of the immunoglobulin chains. Together with the avoidance of chemical pretreatment of cell lysates and the lack of necessity for the removal of endotoxins, the potential to use linear templates without the need of time-consuming plasmid construction is an ideal basis for the automized production of antibody fragments and other disulfide proteins for screening purposes in a high throughput manner.
Although prokaryotic cell-free systems usually yield up to one milligram de novo synthesized protein per milliliter reaction, to date the production of only 30 µg active Fab per milliliter has been reported (24). The consequence of a low percentage of active Fab for applications that may be negatively affected by the presence of inactive portions, is that the inactive portion has to be removed. Such additional purification steps complicate automation of antibody fragment production markedly. Although the yield of synthesized Fab is comparatively low, the specific activity of antibody fragments produced with the expression system described here is much higher compared with present prokaryotic systems since Fab is produced predominantly in active form. We conclude that purification of Fab from the disulfide insect system in order to remove inactive immunoglobulin chains is obsolete.
This study was supported by a grant from the Bundesministerium für Wirtschaft und Technologie (BMWi, No. ZIM EP091944) and partially by the Bundesministerium für Bildung, Wissenschaft, Forschung und Technologie (BMBF, No. 0313965A). We like to thank IBA Göttingen for kindly providing the anti-CD4 (13B.8) Fab sequence.
The authors declare no competing interests.
Address correspondence to Helmut Merk, RiNA Netzwerk RNA-Technologien GmbH, Volmerstraße 9, D-12489, Berlin, Germany. Email: