PEPTIDES & PROTEINS
Figure 2 - Vapourtec VBFR & reactor controller
New concept in flow
To be able to handle internal volume changes automatically , we developed the Variable Bed Flow Reactor ( VBFR ), a new concept reactor which monitors and controls the packing density of the solid media . This ensures the resin is constantly packed , eliminating channelling of reagents and providing useful in-line data about the peptide growth . With the handling of the resin solved and using our existing approach for library synthesis , we looked at the chemical mechanism of activating an amino acid and reacting it with the solid media . In the majority of batch reactors , activation and coupling happen at the same time , usually at moderate temperatures . This is the reason why batch synthesis usually relies on very active species for activators , such as HATU . In flow , we separated both reactions by simply adding a reactor before
Figure 3 – Flow schematics for CF-SPPS
the VBFR , as shown in Figure 3 . This configuration ensures that all the amino acid is activated prior to entering the resin , leading to a more efficient use of reagents ; only active amino ester would pass through the resin beads . Thanks to this configuration , we evaluated and optimised both activation and coupling residence times . Our protocols are based on an isothermal activation time of 40-60 seconds , which gives a cycle time of ten minutes per coupled amino acid . Flow chemistry systems can easily be integrated with inline analytical techniques ( i . e . optical spectroscopy , Raman or even IR ), which provides real-time information on how the reaction is progressing . By combining UV / Vis spectroscopy with the data generated by the VBFR , we accessed information never seen before . By looking at these two datasets combined we can , for example , ‘ see ’ what is the performance of each deprotection step , in both UV and VBFR volume change . This real-time feedback can be used to monitor and take action to ensure couplings and deprotections are 100 % complete even when aggregation occurs or unnatural amino acids are used . By the end of 2018 , we successfully synthesised a variety of peptides between 10 and 40-mer at any scale between 100 mg to a couple of grams . To learn more from the experts , we collaborated with the Max Planck Institute at Germany . During our collaboration we explored more in detail the effect of different resins in different syntheses and even in the synthesis of carbohydrates via solidphase synthesis . 9-10 We also learned that aggregation is a sequence-dependent issue and that even with the best of the systems and reaction conditions , sometimes it just cannot be avoided . In these cases , we can use the real-time data to evaluate chemical alternatives to prevent such events . This was the case with JR-10 , a 10-mer sequence that completely aggregates during synthesis , yielding a low-quality crude peptide . 9 Thanks to the VBFR , we identified where , in the sequence , aggregation occurred , and evaluated the effect of using different conditions without the need to wait for an LC-MS . Within a few days , several syntheses were completed using different solvents , low loading resins and even pseudo-prolines . With the successful method in place , JR-10 was obtained in high crude purity .
Next steps
Our next step in the journey was to minimise reagent consumption . Unnatural amino acids can be hundreds of times the price of natural amino acids . We optimised solvent and reagent usage by fine-tuning the coupling step . Now we can achieve efficient synthesis with as low as two equivalents of amino acid . Glucagon-like peptide 1 is a 30-mer peptide used for the treatment of Type ‣
49