DNA Heat Treatment for Improving qPCR Analysis of Human Adenovirus in Wastewater

PCR inhibitory substances in complex sample matrices can cause false negatives or under-estimation of target concentration. This study assessed DNA heat treatment for reducing inhibition during qPCR analysis of human adenovirus (HAdV) in wastewater samples. Inhibition was reduced by heat treating DNA, where mean HAdV concentration was increased by 0.71 log10 GC/L (and up to 3.04 log10 GC/L in one case), and replicate variability and false negatives were reduced. DNA heat treatment should be further investigated for improving reliability of HAdV concentration estimates in water, which can support more accurate assessment of health risks associated with viral pathogen exposure. Electronic supplementary material The online version of this article (doi:10.1007/s12560-017-9294-4) contains supplementary material, which is available to authorized users.


DNA extraction and treatment
The virus concentrate from water samples was DNase treated prior to extraction, to remove contaminating bacterial DNA and DNA from non-encapsidated viruses, which ensures the qPCR signal is related to detection of intact viral particles only (Fongaro et al. 2013). For DNase treatment, 20 µL of 10X Turbo DNase buffer and 4 units of Turbo DNAse (Ambion, Life Technologies, Carlsbad, CA, USA) were added to 200 µL of the virus concentrate and incubated at 37 o C for 30 min. DNA was extracted from DNase treated virus concentrate (222 µL) using a Qiagen Blood and Tissue extraction kit (Qiagen, Hilden, Germany) according to manufacturer instructions, and resuspended in 100 µL buffer AE.

Quantitative molecular standards
A hexon gene fragment from human adenovirus 41 (Ad41) was used as a molecular standard.

Quantitative PCR
HADV specific qPCR was performed using the primers and Taqman probe described by (Heim et al. 2003)  A standard curve for each run was generated using a dilution series of 10 5 -10 2 and 20 gene copies (GC)/reaction of Ad41 plasmid in duplicate reactions. Cycle threshold (C t ) and sample GC/reaction, were determined automatically using Rotor-Gene Q Series Software 2.3.1. The limit of detection (LOD) was determined as 20 GC/reaction (C t = 40). QPCR inhibition was initially assessed for a subset (n=4) of unheated samples, where the qPCR reaction was spiked with 10 3 GC of Ad41 plasmid. Inhibition was considered to be indicated where the subsequent HAdV GC observed was <60% of the expected value (spiked GC + endogenous sample GC).
HAdV concentration and qPCR inhibition were further assessed in all samples using 5-fold dilutions of DNA (1:1, 1/5 and 1/10). A given dilution was repeated in triplicate where signal was diluted to below the LOD, inhibition was indicated and/or when results were not consistent in the initial dilution series. For samples where HAdV was not detected, 1/10 diluted DNA was spiked with HAdV control plasmid (10 3 GC/reaction), to determine whether the negative result was due to inhibition.

Statistical analysis
The aim of the statistical analysis was to assess whether there was a significant difference between the mean HAdV concentrations using either heated (5 min The mean HAdV concentration (GC/L) for each sample was determined from triplicate technical replicates (for heated and unheated DNA). Log 10 transformed data were then used to estimate mean differences for HAdV concentration determined using heated (5 min) and unheated DNA for: 1) all samples (pooled data, n=22); and 2) inlet and outlet samples analysed separately (samples H1-H20, inlet n=10 and outlet n=10). Log 10 transformed data were also used to estimate mean differences for HAdV concentration determined using heated (30 min) and unheated DNA for a subset of samples (pooled data, n=16). A paired ttest was used to assess significance of differences between the population means (P value < 0.05).
Additionally, the log 10 reduction value (LRV) in HAdV concentration due to WSP treatment was determined for heated (5 min) compared to unheated DNA (samples H1-H20). LRVs were initially calculated as point estimates, as recommended in the Australian Guidelines for Water Recycling (AGWR 2006): LRV = average(log 10 (inlet GC/L) -average(log 10 (outlet GC/L)).
Statistical Package for the Social Sciences (SPSS Version 23, IBM) was used for all statistical analyses.