A typical application of filter binding assays is the discovery of unspecific plasma protein binding to novel nucleic acid therapeutics. It is assumed that these binding events could block the biological activity of the RNA drug, or – lets say – at least it could influence its physico-chemical properties within the blood compartment, which transports the drug to the target organ. There is a common agreement among the nucleic acid biotech community that it does play a role, whether the nucleic acid therapeutic is administered in its naked form (single or double-stranded), or if it´s packed into liposomal vehicles for intracellular uptake. Certainly any chemical modification of the bases or polynucleotide backbone could also influence the binding of plasma components (small molecules, metabolites, proteins, lipids, or whatever…). And so does the addition of chemical modifications to the polynucleotide. A nice example for this observation is the lipophilic cholesterol, which often is added to the terminal ends e.g. of aptamers or siRNA therapeutics. Cholesterol anchors the nucleic acid drug not only to the liposome shuttle, but also to any plasma component capable to bind to cholesterol – which in turn influences biodistribution and plasma clearance of the nucleic acid drug.
Well, lets focus on the naked (which means unpacked) nucleic acid therapeutics. Ultrafiltration binding assays have been published to study plasma protein binding both in vivo and in vitro (1). The nucleic acid therapeutic is incubated with plasma, and the mixture is centrifuged through a filter with a defined molecular weight cutoff. The (unbound) parent compound should pass the filter, while bound compounds should remain in the supernatant. In many cases a radioactively labelled component allows for a semi-quantitative analysis of the procedure – to address the question of how much compound remains unbound, and therefore probably bioactive in vivo. Alternative binding assays were published (2) studying famous G3139 antisense binding to lipophilic plasma components (both naked and lipoplexed). High density lipoproteins (HDL) and low density lipoproteins (LDL) components were separated by density gradient centrifugation, and filtered afterwards by equilibrium dialysis through a molecular weight cutoff membrane.
In most cases the nucleic acid therapeutic is radiolabelled (e.g. with [3H] tritium), which is a rather cheap method for labelling, but produces incalculable follow-up costs by means of safety precautions in the laboratory. For this reason we feel confident that plasma protein binding assays should only be performed with label-free nucleic acid parent compound.
Of course these ligand binding assays can also be applied to tissue lysates of any kind, from any species which is included in the preclinical drug development process. Only to make this short collection essay on filter assays complete, plasma protein binding to small molecule drugs is also studied by filter assays, but with LC-MS/MS bioanalytics as final readout (3).
(1) Watanabe TA et al (2006), Oligonucleotides 16: 169-80.
(2) Wasan EK et al (2002), Int J Pharm 241(1): 57-64.
(3) Larsen HC et al (2011), Bioanalysis 8: 843-52.