Corrigendum to: Purification of therapeutic & prophylactic mRNA by affinity chromatography

In the version of this article initially published, the stated binding capacity of the CIMmultus Oligo dT column was incorrectly written as 0.18mg_mRNA.mL_support^-1. It has since been brought to the authors’ attention that the correct binding capacity of the CIMmultus Oligo dT column is more than 20× higher.

Table 2 has been corrected to remove the last row that stated the incorrect binding capacity data.

Furthermore, the paragraphs under the heading ‘Oligo deoxythymidine’ that referred to the binding capacity of the CIMmultus Oligo dT column have also been corrected as follows:

The main options for industrially appropriate oligo (dT) products are the Praesto Jetted (dT)18-DVB, Poros (dT)25 and CIMmultus (dT)18 monolith. The first two are resin technologies that utilise a divinyl benzene base matrix and affix the (dT) ligand to the surface with a proprietary linker. Purolite have released a binding capacity of 2 mg.mL_resin^-1 of 200 nt Poly(A) compared to 0.62 mg.mL_resin^-1 of 40 nt Poly(A) on a Poros (dT)25 certificate of analysis [40,41] The Poros (dT) displays 10% breakthrough values of 4 and 3 mg.mL_resin^-1 capacity for 2,000 and 3,000 nt mRNAs, respectively, whilst 1,000 nt mRNAs show a 5% breakthrough of 4 mg.mL_resin^-1 [42]. A clear correlation between mRNA size and capacity is observed, with Poros (dT)25 having lower capacity for larger mRNAs. This indicates that surface crowding is preventing the full utilisation of the bound (dT) ligand. Despite the reduced capacity for larger mRNAs, the resin can be reused for 10 cycles with only a marginal drop in yield.

Commercially available monoliths include the CIMmultus (dT)18 range from BIA Separations. These are Poly glycidyl methacrylate-co-ethylene dimethacrylate monoliths where (dT)18 is immobilised with a C6 or C12 linker chain. The product exhibits a ligand density of 0.5 mg_{Oligo (dT)}.mL_{wet support}^-1. There is currently no available data for capacity with any length of mRNA [43,44]. However, a 1 mL CIMmultus™ Oligo (dT) is capable of an 80% recovery when purifying an IVT mixture containing approximately 180µg 2000 nt mRNA. Additionally, monolith separations can be completed in a shorter space of time due to the higher rates of convective flow [45].

Comparing existing products will remain challenging until capacity data for a wide range of mRNA constructs is released. The (dT)18 ligand present on the Praesto Jetted (dT)18-DVB indicates a capacity somewhat like Poros (dT)25. The comparison becomes difficult when accounting for differences in the resin and monolith technology. Each technology presents options for mRNA purification at an industrial scale. A second generation of products is required to further push the boundaries in capacity. New options could include other base materials, such as agarose. Agarose (dT)20 was prepared using NHS activated Sepharose FF. This achieved a 1.6 mg.mL_resin^-1 capacity with a 900 nt polyadenylated mRNA [46]. However, this could be indicative of the unsuitability of agarose as a base matrix at relatively large pore sizes, given that no agarose products are yet commercially available.