Minimizing Bias in Virally Seeded Water Treatment Studies: Evaluation of Optimal Bacteriophage and Mammalian Virus Preparation Methodologies.

One key assumption impacting data quality in viral inactivation studies is that reduction estimates are not altered by the virus seeding process. However, seeding viruses often involves the inadvertent addition of co-constituents such as cell culture components or additives used during preparation steps which can impact viral reduction estimates by inducing non-representative oxidant demand in disinfection studies and fouling in membrane assessments. The objective of this study was therefore to characterize a mammalian norovirus surrogate, murine norovirus (MNV), and bacteriophage MS2 at sequential stages of viral purification and to quantify their potential contribution to artificial oxidant demand and non-representative membrane fouling. Our results demonstrate that seeding solvent extracted and 0.1 micron filtered MNV to ~10 PFU/mL in an experimental water matrix will result in additional total organic carbon (TOC) and 30 min chlorine demand of 39.2 mg/L and 53.5 mg/L as Cl, respectively. Performing sucrose cushion purification on the MNV stock prior to seeding reduces the impacts of TOC and chlorine demand to 1.6 and 0.15 mg/L as Cl, respectively. The findings for MNV are likely relevant for other mammalian viruses propagated in serum-based media. Thus, advanced purification of mammalian virus stocks by sucrose cushion purification (or equivalent density-based separation approach) is warranted prior to seeding in water treatment assessments. Studies employing bacteriophage MS2 as a surrogate virus may not need virus purification, since seeding MS2 at a concentration of ~10 PFU/mL will introduce only ~1 mg/L of TOC and ~1 mg/L as Cl of chlorine demand to experimental water matrices.