Melt Curves and Standard Curve PDF

Journal of Microbiological Methods 132 (2017) 139–147 Contents lists available at ScienceDirect Journal of Microbiological Methods...

Journal of Microbiological Methods 132 (2017) 139–147 Contents lists available at ScienceDirect Journal of Microbiological Methods journal homepage: www.elsevier.com/locate/jmicmeth Development of a sensitive and false-positive free PMA-qPCR viability assay to quantify VBNC Escherichia coli and evaluate disinfection performance in wastewater effluent Richard J. Kibbee, Banu Örmeci ⁎ Department of Civil and Environmental Engineering, Carleton University, 1125 Colonel By Drive, Ottawa, ON, Canada a r t i c l e i n f o a b s t r a c t Article history: The detection and quantification of viable Escherichia coli cells in wastewater treatment plant effluent is very im- Received 30 August 2016 portant as it is the main disinfection efficacy parameter for assessing its public health risk and environmental im- Received in revised form 3 December 2016 pact. The aim of this study was to develop a sensitive and false-positive free propidium monoazide-quantitative Accepted 3 December 2016 polymerase chain reaction (PMA-qPCR) assay to quantify the viable but non-culturable (VBNC) E. coli present in Available online 6 December 2016 secondary wastewater effluent after chlorine disinfection. The qPCR target was the E. coli uidA gene, and native Keywords: Taq was used to eliminate false positives caused by the presence of contaminant E. coli DNA in recombinant qPCR Taq polymerase reagents. Due to issues with qPCR inhibitors in wastewater, this study explored several pre- PMA DNA extraction treatment methods for qPCR inhibitor removal. PMA-qPCR validation was done using salmon tes- VBNC tes DNA (Sketa DNA) as an exogenous control added directly to the wastewater samples and amplified using a E. coli separate qPCR assay. After disinfection of secondary effluent with 2 ppm chlorine at the plant, the mean Log10 Wastewater CFU reduction in E. coli was 2.85 from a mean CFU of 3.48/10 mL compared to 0.21 Log10 CCE mean reduction Disinfection of the uidA gene from a mean CCE of 3.16/10 mL. The VBNC cell concentrations were calculated as 2.32 Log10/ 10 mL by subtracting the colony forming units (CFU) obtained from membrane filtration from the calculated CFU equivalent (CCE) values obtained from PMA-qPCR. These results demonstrate the effective use of a PMA- qPCR method for the quantification of the E. coli uidA gene and indicate there are high numbers (2.01 × 103 CCE/100 mL) of VBNC E. coli cells leaving the wastewater treatment plant in the final effluent after chlorine treatment. VBNC bacterial cells are of concern as they have the potential to resuscitate and grow, regain virulence, affect natural microbiome in the discharge sites, and pass on antimicrobial resistant genes to other microorganisms. © 2016 Elsevier B.V. All rights reserved. 1. Introduction (Oliver et al., 2005). In the VBNC state, bacteria maintain a low metabol- ic activity and do not divide, but they can resuscitate and regrow later The detection and quantification of viable Escherichia coli (E. coli) when the environmental conditions are favourable (Pinto et al., 2011). cells in wastewater effluent is used as an indicator of disinfection effica- It was reported that the VBNC state produces antibiotic-tolerant popula- cy at wastewater treatment plants (WWTP). The methods commonly tions (Ayrapetyan et al., 2015) since antibacterials would not be as ef- used for E. coli detection are culture-based, and these methods only de- fective on VBNC bacteria due to their low metabolic rate (Li et al., tect cells that are capable of growing in culture media. As a result, using 2014a). The presence of VBNC cells can also influence the native culture-based quantification methods in assessing the treatment or dis- microbiome and contribute to the distribution of antimicrobial resistant infection performance result in the underestimation of total viable cells. genes (ARGs) within the microbiome at effluent discharge sites (Marti It is well established that bacteria, including E. coli, can go into a vi- et al., 2013), where higher levels of ARGs have been reported (Li et al., able-but-non-culturable (VBNC) state to cope with environmental 2014a; Taskin et al., 2011). Additionally, pathogens have been stressors. Chlorine disinfection, for example, was reported to cause shown to regain their virulence after exiting the VBNC state (Du et VBNC state in E. coli, especially in the presence of high organic matter al., 2007), which poses another concern. It is important, therefore, to develop methods that can accurately quantify the population of VBNC after water and wastewater disinfection. These methods should also ideally have the sensitivity to detect very low numbers ⁎ Corresponding author. of VBNC cells and be able to assist in determining and mitigating E-mail address: [email protected] (B. Örmeci). public health risks. http://dx.doi.org/10.1016/j.mimet.2016.12.004 0167-7012/© 2016 Elsevier B.V. All rights reserved. 140 R.J. Kibbee, B. Örmeci / Journal of Microbiological Methods 132 (2017) 139–147 PMA-qPCR (propidium monoazide-quantitative polymerase chain forms of E. coli (survive at 44.5 °C) to b200 CFU/100 mL in the final ef- reaction) methods have been employed for the differentiation of viable fluent as cultured on selective agar for E. coli. and VBNC cells from non-viable fecal indicator bacteria in food, recrea- All WW and WWD samples were collected into autoclave-sterilized tional water, and WWTP samples including sludge, treated biosolids 5-liter carboys. Experiments were performed on the day the samples and effluents (Varma et al., 2009; van Frankenhuyzen et al., 2011; Li et were collected. All samples were stored at 4 °C. al., 2014c). qPCR cannot differentiate the DNA from viable and nonviable cells since DNA can linger for some time after cell death (Masters et al., 1994). PMA functions by entering a compromised bacterial cell wall 2.2. Pre-treatments for qPCR inhibitor removal and intercalating with its DNA, rendering it unamplifiable by qPCR. This enables the differentiation between intact and compromised cells. Three procedures for removal of humic acids and related substances Intact cells are regarded as viable (including culturable and non- from wastewater were tested. For each procedure, 1 mL of a 20 h culture culturable cells), and those that are not, are dead. In order to eliminate of E. coli was washed in PBS, then spiked into 400 mL fresh WW, and the false positives in the qPCR assay for the detection of the uidA gene mixed by inversion for 30 s. in E. coli, Kibbee et al. (2013) used native Taq due to the presence of con- taminant E. coli DNA in recombinant Taq polymerase reagents, and were Procedure 1: Three replicates of 1, 10 and 100 mL WW samples were able to detect as low as 5 copies of the uidA gene compared with 10– concentrated by centrifugation at 8000 × g for 5 min 10,000 gene copy numbers reported in the literature. (samples represent 1×, 10× and 100× concentrates), su- Application of PMA-qPCR for detection of E. coli in secondary pernatant discarded and pellets resuspended in 450 μL of wastewater effluent requires the effluent first to be concentrated. phosphate buffer (pH 6.6) and to this, 50 μL of 100 mM This has been shown to be problematic. The methods used to concen- aluminum sulfate (Al2SO4) solution was added to give a trate E. coli increase the negative effects of the inherent qPCR inhib- final concentration of 10 mM. The samples were then itors, mainly humic acids (Li et al., 2014b). These humic acids can vortexed for 2 min followed by a pH adjustment to 8.0 interfere with DNA extraction as well by competing for binding by adding 150 μL of 1 M sodium hydroxide NaOH as de- sites on silica beads or particles used in DNA extraction kits. Various scribed by Dong et al., 2006. The samples were then cen- physical and chemical methods were reported in the literature to re- trifuged for 5 min at 8000 ×g and the pellets used for DNA duce or remove the PCR inhibitors prior to sample concentration and extraction. DNA extraction (Fatima et al., 2014; Braid et al., 2003; Dong et al., Procedure 2: 25, 50 and 100 mM solutions of aluminum ammonium 2006; Lakay et al., 2007). sulfate (AlNH4(SO4)2) were prepared in accordance to Traditional PMA validation methods include comparison of the Braid et al., 2003. 200 μL of each were added separately qPCR's Cq values of cell membrane compromised E. coli using heat, to three replicate sets of 1×, 10× and 100× concentrated isopropanol or freeze-thaw methods to compromised cell membranes. WW sample pellets in their respective DNA extraction kit The difference in Cq is equal to the sample's non-viable concentrations lysis tubes, samples were briefly vortexed to incorporate when uncompromised cell membranes are considered viable. These the chemistry, and DNA extractions carried out. methods showed varying levels of bacterial cell membrane damage Procedure 3: Three replicate sets of 1, 10 and 100 mL WW samples and therefore varying PMA effectiveness with Δ Cq's of no N 7 (~ 2.2 were concentrated by centrifugation at 8000 × g for Log10) (Barbau-Piednoir et al., 2014). Alternatively, the PMA intercala- 5 min (1×, 10× and 100× concentrates), the sample su- tion of DNA in the sample matrix can be validated using salmon testes pernatants removed and the pellet resuspended in PBS DNA (Sketa DNA). This exogenous DNA control, recommended by the for a single wash. The wash consisted of a 15 s vortex in U.S. Environmental Protection Agency (USEPA Method 1611, 2012) PBS followed by centrifugation at 8000 × g for 5 min to has been used in recent studies to quantify PCR inhibition (Cao et al., obtain a pellet ready for DNA extraction. 2012; Gentry-Shields et al., 2013), as well as DNA extraction efficiency (Haugland et al., 2010). The objective of this study was to develop a PMA-qPCR assay, to be For the three procedures, 1 μL DNA templates were assessed by used in conjunction with culture-based methods, to be able to enumer- qPCR using the uidA gene qPCR protocol as described herein to ate VBNC E. coli after wastewater disinfection. A previously published determine inhibition by comparing the Cq values of the treated and false-positive free E. coli qPCR assay (Kibbee et al., 2013) was modified untreated concentrates. All pre-treatment experiments were run to include PMA for detecting VBNC cells. Several pre-treatment methods with untreated controls using PBS. In the case of the third procedure, were tested and controls were introduced to achieve an accurate and no wash was performed on the WW concentrates before DNA sensitive method that is free from false-positives and that can quantify extraction. low numbers of VBNC cells. The method was used to determine the Removal of the supernatant with a PBS wash was done to remove number of viable and VBNC E. coli cells that survive chlorine disinfection extracellular DNA as well as humic acids. This increased the effective- in wastewater and to evaluate the true disinfection performance. ness of the PMA treatment and allowed the DNA extraction kit chemis- try to perform to specifications. 2. Material and methods A flow-chart that illustrates the experimental steps and protocols 2.3. E. coli culture preparation are illustrated in Fig. 1. A fresh 20 h culture of E. coli (ATCC 19853) grown in tryptic soy 2.1. Wastewater samples broth (TSB) was used for the spiked control experiments including the uidA gene standard curves. 1 mL of the frozen stock was added to Secondary effluent wastewater samples prior to disinfection (WW) 99 mL autoclave sterilized TSB and incubated at 37 °C for 20 h in a and after chlorine disinfection (WWD) were collected from a conven- vented Erlenmeyer flask. This yielded a consistent concentration of ap- tional activated sludge treatment plant in Ontario, Canada. At the treat- proximately 5 × 108 CFU/mL at a late-log phase. 1 mL aliquots of the ment plant, the secondary effluent was treated with 2 ppm chlorine and 20 h suspension were washed with phosphate buffered saline (PBS) dechlorinated before discharge into the environment. The WWTP main- and centrifuged at 8000 ×g for 5 min to remove the growth media be- tains monthly geometric mean compliance levels of thermally tolerant fore use.  R.J. Kibbee, B. Örmeci / Journal of Microbiological Methods 132 (2017) 139–147 141 Fig. 1. Flowchart of the experimental steps and protocols. 2.4. Culture-based detection and enumeration of E. coli 2.5. PMA treatment Detection and enumeration of E. coli in WW samples was PMA (Biotium, Inc.) was dissolved in 20% dimethyl sulfoxide achieved by serial dilutions of the WW and using 100 μL of each (Sigma) to create a 20 mM working stock solution and stored at 4 °C. dilution to establish the standard curve. For the WWD samples, 1– To each sample, 5 μL of the working stock was added to 495 μL of the 100 mL was filtered through a 47 mm diameter nitrocellulose mem- test sample to achieve a final concentration of 200 μM. Samples were brane filters (EMD-Millipore) with a pore diameter of 0.45 μm and vortexed on high for 5 s, followed by incubation at 30 °C for 10 min to placed onto separate mTEC (Criterion, Hardy Diagnostics) agar allow the PMA to enter into the cells with compromised membranes. Petri plates. The Petri plates were incubated at 44.5 °C overnight. Samples were then placed on ice, in front of a fan, 20 cm from a After incubation, filter membranes containing yellow colonies were photoactivation light (500 W halogen bulb lamp, 120 V) and exposed transferred to pads saturated with a urea-phenol solution and incu- for 10 min. Samples were mixed by inversion every 2 min. All samples bated for 15–20 min. E. coli is urease negative and remains yellow/ including those that were not treated with PMA were included in all yellow-brown in colour. Colonies that are not E. coli turn purple. All steps of the treatment. The samples were then centrifuged at yellow/yellow-brown colonies were counted and recorded in CFU/ 10,000 × g for 3 min, supernatant discarded and DNA extracted. The mL. All samples and dilutions were plated in triplicate. (U.S. EPA PMA method validation was done using Sketa DNA (Sigma-Aldrich) Method 1103.1, 2010) with the PMA concentration at 200 μM. 142 R.J. Kibbee, B. Örmeci / Journal of Microbiological Methods 132 (2017) 139–147 2.6. PMA validation recognized that the uidA gene is present in certain species of Shigella which can be found in wastewater effluent but at very low concentrations, For PMA validation, 20 ng of Sketa DNA was added to 495 μL of resus- ~1000 times lower than that of E. coli. A 3-step qPCR protocol consisted of pended PBS washed WWD with and without 5 μL of the PMA working an initial 2 min denaturation at 95.0 °C followed by 39 cycles of denatur- stock solution. The PMA treatment was done as described previously, ation at 95.0 °C for 10 s, annealing at 61.4 °C for 20 s and an elongation then the sample DNA extracted. Stock solutions of the Sketa DNA were at 72.0 °C for 15 s with a melt curve from 65.0–95.0 °C with an increment prepared by diluting the frozen 20 μg/mL, 1 in 10 (150 μL into of 0.2 °C for 5 s. Each 25 μL reaction had 2.5 μL 10× PCR buffer, 0.5 μL of 1350 μL) in molecular grade water (Sigma) as described by Haugland 10 mM deoxynucleotide triphosphates (dNTPs), 1.25 μL of 50 mM MgCl, et al. (2005). 0.625 μL (0.10 μM) of each primer, 0.5 μL (1 U/μL) native Taq polymerase, 1.25 μL of EvaGreen dye (Biotium Inc.), 1 μL of 10 μg/μL bovine serum albu- 2.7. DNA extraction min (BSA), 0.25 μL (1%) formamide and 50 ng in control samples or 10 μL of test sample genomic DNA template topped up to 25 μL with DNA-free Genomic DNA was isolated from the WWTP samples and the con- water (Sigma). trol strain of E. coli K12 (ATCC 19853) using the PowerSoil® DNA Iso- The PMA-qPCR with Sketa DNA amplified a 76 bp segment (Fig. 2) of lation Kit (MO BIO Laboratories, Inc.). The quantity of genomic DNA the internal transcribed spacer region 2 of the ribosomal RNA gene oper- from the control strain was determined using the Qubit® 3.0 Fluo- on of chum salmon, Oncorhynchus keta, (Domanico et al., 1997) by using rometer and the quality determined by analyzing the 260:280 nm a pair of 16-mer primers: forward, SketaF2 (5′-GGTTTCCGCAGCTGGG- ratio using the Varian CARY-100 BIO UV-VIS spectrophotometer; all 3′) and reverse Sketa22 (5′-CCGAGCCGTCCTGGTC-3′) (Haugland et al., samples had ratios N 1.8. Isolated DNA samples were stored at 2010). A 2-step qPCR protocol consisted of an initial 2 min denaturation − 20 °C prior to analysis. at 98.0 °C followed by 39 cycles of denaturation at 98.0 °C for 2 s, anneal- ing/elongation at 60.0 °C for 5 s with a melt curve from 65.0–95.0 °C with 2.8. qPCR protocol optimization and analysis an increment of 0.2 °C for 5 s. Each 25 μL reaction consisted of 12.5 μL SsoFast EvaGreen 2× Supermix (Bio-Rad), 0.9375 μL (0.15 μM) of each The uidA gene target qPCR protocol was generated using the Bio-Rad primer, 0.25 μL formamide, 1 μL of 10 μg/μL BSA and 50 ng in control sam- CFX96 software and all but one parameter was altered. Initial qPCR runs ples or 10 μL of test sample genomic DNA template topped up to 25 μL with spiked WW samples showed nonspecific amplification and exces- with DNA-free water (Sigma). All primers were desalted and shipped sive primer dimer formation. Both of these issues created falsely in- dry, resuspended in molecular grade water and stored at − 20 °C. All creased Cq values. To address this, the initial denaturation of 3 min primers were purchased from Life Technologies, (Burlington, ON). All was changed to 2 min, as well, the 96-well plate was maintained at 2– qPCR runs were performed using the Bio-Rad CFX96™ Real-Time PCR 4 °C with the aid of a custom made 96-well cold block throughout the Detection System (Mississauga, ON). Determinations of cycle threshold loading process. HotStart qPCR using anti-Taq DNA polymerase antibod- (Cq) were performed automatically by the instrument after manually ies was not investigated due to the elimination of nonspecific amplifica- adjusting the baseline subtracted relative fluorescence units to ~ 100 tion products. Other optimization steps taken were a gradient qPCR to RFU. confirm primer annealing temperature and a magnesium chloride opti- An E. coli uidA gene detection limit standard curve was run, where mal concentration confirmation by running a series of concentrations the DNA extraction of a 20 h culture was serially diluted 1 in 10 by from 1.0 to 2.5 mM. Each concentration was run in triplicate. Formamide adding 5 μL into 45 μL molecular grade water; triplicates of each dilution was tested and adopted in both the uidA and Sketa DNA protocols to help were run. The genomic DNA was measured using the Qubit 3.0, and the decrease primer-dimmers to improve the resolution of qPCR signals volume of DNA template was calculated to obtain 50 fg–50 ng for each when very low copy numbers of the target gene are present. dilution. The Sketa DNA was prepared as described in PMA Validation A 167 bp segment of the uidA gene in E. coli (Fig. 2) was amplified and serially diluted as described above and run using the Sketa qPCR by using a 21-mer forward primer, UAL1939b (5′-ATGGAATTTCG- protocol. CCGATTTTGC-3′) and a 20-mer reverse primer, UAL2105b (5′- ATTGTTTGCCTCCCTGCTGC-3′) (Heijnen and Medema, 2006). It is 2.9. qPCR, PMA-qPCR VBNC cell concentrations In all VBNC experiments, 10 mL of the WW and WWD samples were concentrated by centrifugation at 8000 ×g for 5 min, washed once with PBS, pelleted, and DNA extracted. The genomic DNA was measured and compared to the qPCR Cq values. 10 mL concentrated samples yielded 0.38–2.47 ng/μL of genomic DNA (Table 1). The VBNC cell concentra- tions were calculated by subtracting the CFU values from the CCE PMA treated values. This compares the cultured live bacterial cells to those that are intact and considered viable. Table 1 Qubit 3.0 DNA concentration compared to observed Cq values. Date Sample (10 mL) Cq Qubit (ng/μL) 04.10.2015 WW 25.37 2.47 WWD 25.11 1.87 06.18.2015 WW 26.93 0.94 WWD 29.25 1.05 07.07.2015 WW 29.92 0.64 Fig. 2. 2.5% Agarose gel of the 167 bp amplicon –uidA gene (Heijnen and Medema, 2006) WWD 32 0.38 and 76 bp amplicon -Sketa DNA-rRNA gene operon (Haugland et al., 2010); Lane 1: 50 bp ladder; Lane 2 was intentionally left empty; Lane 3: uidA amplicon from E. coli ATCC The above table represents the genomic DNA concentrations and Cq values of the 3 sets of 19853; Lane 4: Sketa DNA amplicon. WWTP samples.  R.J. Kibbee, B. Örmeci / Journal of Microbiological Methods 132 (2017) 139–147 143 3. Results and discussion a 3.1. qPCR inhibition and DNA extraction validation Humic acids and related substances have been reported as PCR in- hibitors. They can inhibit by covalently binding to DNA or interfering with the Taq polymerase, primer binding as well as co-purifying and competing with DNA for space on the silica beads used in many DNA ex- traction kits including the PowerSoil® DNA Isolation Kit used in this study (Opel et al., 2009; Lloyd et al., 2010). It has been shown that 1 μg/mL concentration of humic acids will completely inhibit a PCR reac- tion (Matthews et al., 2010). Although the kit protocol includes an in- hibitor removal step, further reductions were necessary due to high concentrations of inhibitors in WWTP samples. Secondary wastewater effluent may have as much as 10 μg/mL of humic acids and therefore, the centrifugation concentration of the 10 and 100 mL WW samples could result in inhibitor levels beyond the inhibition removal capacity b of the kit, which is ~20 μg/mL (Vakondios et al., 2014). Three qPCR inhibitor removal methods were considered in this study to improve the removal of qPCR inhibitors. The Al2SO4 treatment appeared to inhibit the qPCR reaction and showed an increase in Cq of the 1, 10 and 100 mL concentrates of 1.14, 0.6 and 1.4 respectively com- pared to the PBS control (Fig. 3a, b). This increase in Cq represents a re- duction in Log10 genomic units detected in the concentrates of 0.35, 0.18 a Fig. 4. a. This uidA gene qPCR Cq plot representing the mean calculated values shows the shift in Cq between the spiked AlNH4(SO4)2 treated WW (green) and untreated WW (blue) samples. Positive control Cq 11.63, SD 0.22 (Red) and NTC Cq 33.49, SD 0.03 (Purple). Sample sets from left to right represent 100, 10 and 1 mL, treated and untreated concentrates with Cq values of 13.33, 13.87; 15.94.16.07; 19.26, 19.55 respectively, with Δ Cq values of 0.56, 0.13 and 0.29. b. Aluminum ammonium sulfate treated and untreated 1, 10 and 100 mL WW concentrates uidA gene qPCR results show a slight improvement in Cq with the 100 mM treatment (yellow bars) with decreases in Cq of 0.29, 0.13 and 0.56 respectively, (n = 3). and 0.43 respectively. The AlNH4(SO4)2 treatment did reduce the Cq values slightly of the 1, 10 and 100 mL concentrates by 0.29, 0.13 and b 0.56 respectively compared to the PBS control with the 100 mM solu- tion, as seen in Fig. 4a, resulting in an increased detection of the Log10 genomic units of 0.09, 0.04 and 0.17, respectively. The 25 and 50 mM treatments showed less of an effect; results are shown in Fig. 4b. The third method, the E. coli spiked WW PBS-washed samples, showed an increase in Cq for the 1 mL WW concentrate sample of 0.93 and a de- crease of 0.15 and 0.08 for the 10 and 100 mL WW concentrate samples compared to the spiked PBS samples (Fig. 5) resulting in a decreased de- tection of the Log10 genomic units of 0.28 for the 1 mL concentrate and an increased detection of the Log10 genomic units of the 10 and 100 mL concentrates of 0.05 and 0.03, respectively. These data show that there is measurable inhibition of the qPCR reaction. However it is quite low, and therefore this method was used to help remove the qPCR inhibitors from the concentrated wastewater samples. The DNA extraction kit was validated by measuring the total geno- Fig. 3. a. This Cq plot, representing the mean calculated values shows the shift in Cq mic DNA of E. coli spiked WW and PBS with the Qubit 3.0 Fluorometer, between the spiked untreated–Cq 16.24, 19.95 and 23.97 (Blue) and Al2SO4 treated- HS (high sensitivity) genomic DNA reagent kit. The quantity of DNA in E. 17.64, 20.54 and 25.14 (Green) 100, 10 and 1 mL concentrated WW samples coli spiked PBS and WW (1, 10 and 100 mL) samples yielded linear respectively. The qPCR reaction showed an increase in Cq of the 100, 10 and 1 mL concentrates of 1.4, 0.6 and 1.16 respectively compared to the PSB control. Positive trend R2 values N0.98 (n = 9) even with the 1 mL sample sets which control Cq 12.79, SD 0.31 (Red) and NTC 30.71, SD 0.2 (Purple). b. The Al2SO4 treated showed less efficient DNA extraction, likely due to loss by adsorption WW concentrates (red bars) showed a slight increase in Cq. onto test tube walls (Gaillard and Strauss, 1998). The mean of all the 144 R.J. Kibbee, B. Örmeci / Journal of Microbiological Methods 132 (2017) 139–147 human feces and wastewater matrices. Taskin et al., 2011, developed a PMA-qPCR method to detect the uidA gene in biosolids which might be used to quantitatively assess VBNC cells. In the study by Li et al., 2014c, PMA-qPCR was used to target and quantify the 23S rRNA of En- terococcus and the uidA gene of E. coli throughout the WWTP process in- cluding the secondary effluent samples. Genomic copies of these targets were much lower than expected with ~2–3Log10 genomic copies for the influent and sludge samples, and no detection from secondary effluent samples. It could be postulated that there were DNA extraction interfer- ences due to the sample matrix, especially the 400 mL concentrate of the secondary effluent. It was noted that the sample matrix may have adverse effects on the performance of PMA which was also found by Varma et al., 2009 and Taskin et al., 2011. These studies have all con- cluded that the use of PMA-qPCR for the quantification of viable cells in various wastewater matrices is promising and have brought to light factors influencing the method; these include TSS, DNA contaminants Fig. 5. The above representative qPCR Cq plot shows the 100 mL concentrated spiked WW in qPCR reagents and PCR inhibitors. In this study, we were able to ad- (orange triangle) sample compared to the 100 mL spiked PBS sample (orange X) with a dress PMA performance issues by increasing the concentration from mean Cq value of 13.49 and 13.40 respectively (ΔCq = 0.08); 10 mL concentrated spiked WW (blue triangle) sample compared to the 10 mL spiked PBS sample (blue X) 100 mM to 200 mM per reaction and incorporating a PBS wash pre- with Cq values of 16.11 and 15.95 respectively (ΔCq = 0.15); and 1 mL concentrated treatment of the sample concentrates. We showed that TSS was not spiked WW (dark green triangle) sample compared to the 1 mL spiked PBS sample an issue with 10 mL concentrated secondary WW samples and elimi- (dark green X) with Cq values of 19.95 and 20.87 respectively (ΔCq = 0.93); light green nated the chance of false positives in the qPCR non-template controls is the (+) control, Cq 12.89, SD 0.28; dark red are the NTC's Cq 30.59, SD 0.54. from DNA contaminated qPCR reagents by adopting a false-positive free qPCR method which uses native Taq polymerase. The use of Sketa DNA has shown to be an effective way to demonstrate the PMA treat- ment in the WWD samples is capable of binding to accessible DNA and prevent amplification. In this study, the addition of PMA was able Cq values of the spiked WW and PBS samples were within 2 standard to reduce the amplification of Sketa DNA by 3.13 Log10 genomic units deviations (95%), showing that humic acids were sufficiently decreased from a starting concentration of 4.73. This starting concentration is ~1 and the extracted E. coli DNA did not inhibit the uidA gene qPCR ampli- Log10 higher than the uidA genomic units found in the 10× concentrated fication as seen in Table 2. The Log10 CCE variation between the 1× and WW and WWD samples. A higher starting concentration, 6.97 Log10 ge- 10× concentrates was calculated using the baseline sample (100× con- nomic units of the Sketa DNA in molecular grade water, was also chal- centrated E. coli spiked PBS) with a CCE of 6.97 Log10 and showed a lenged with 200 mM PMA. Amplification was reduced by 4.41 Log10 mean standard deviation of 0.22. (Fig. 6). This is a significant control as it has been reported that waste- water matrices have an inhibitory effect on PMA performance contain- 3.2. PMA-qPCR and PMA-qPCR control ing extracellular DNA which may quench a portion the PMA and high levels of suspended solids which are presumed to interfere with light PMA-PCR/qPCR has been used by researchers to distinguish viable penetration into the sample preventing intercalation of the PMA with from non-viable bacterial cells for microbial source tracking, human the DNA. It has also been demonstrated that the use of shorter DNA tar- pathogen detection and monitoring of indicator organisms in a variety get sequences offer a more stringent control as they require very effi- of matrices including wastewater effluent since its development by cient PMA intercalation to hinder qPCR amplification (Martin et al., Nocker et al., 2007. Of these research studies, few have been published 2013). For optimal PMA performance, there must be saturation of acces- investigating pathogens or indicator organisms in wastewater and sible DNA with the dye as well as sufficient light exposure to promote more specifically, secondary effluent. Bae and Wuertz, 2009, developed intercalation. The Sketa DNA control target sequence used in this a PMA-qPCR method for microbial source tracking of Bacteroides fragilis study was 76 bp in length which is 91 bp shorter than the uidA gene tar- which was successful in its ability to discriminate viable and non-viable get sequence of 167 bp. We can theorize that the PMA efficiency would cells as well as extracellular DNA which was able to show a correlation need to be over 2 times higher to achieve the same level of qPCR inhibi- between total suspended solids (TSS) and performance of PMA in tion of the uidA gene based on the target sequence. This is based on the premise that amplifying longer target sequences require optimal perfor- mance of the PMA intercalation to all available DNA as noted by Contreras et al., 2011. Table 2 Mean Log10 CCE/mL uidA gene comparisons; WW and PBS controls with minimal to no 3.3. qPCR standards and limit of detection inhibition. uidA gene qPCR Inhibition: Log10 CCE/mL The standard curves in this study have a linear range of quantifica- samples- spiked Mean Log10 CCE dilution variation from PBS [100×] tion from 101 to 107 genomic units/reaction for the E. coli uidA gene or E. coli Cq CCE/mLa corrected baseline 50 fg to 50 ng with an R2 = 0.9983 (Fig. 7). The Sketa DNA has a linear PBS [1×] 20.87 4.71 6.71 0.26 range of 101 to 106 genomic units/reaction or 50 fg to 5 ng with an R2 = WW [1×] 19.93 4.99 6.99 −0.02 0.9962 (Fig. 8). PBS [10×] 15.95 6.2 7.2 −0.23 WW [10×] 16.11 6.15 7.15 −0.18 3.4. E. coli CFU and uidA gene CCE of WW and WWD samples PBS [100×] 13.4 6.97 0 WW [100×] 13.19 6.94 0.03 a The Cq conversion to CCE was based on the mTEC culture results Log10 CCE/mL conversion based on the 100× PBS concentrate with a Log10 CFU/mL of 6.97 as determined by mTEC agar culture. This table shows good correlation between the where 1 uidA Gene copy is equal to 1 E. coli bacterial cell (Taskin et al., Mean Cq and Log10 CFU/mL recovery on mTEC agar of both spiked PBS-washed WW and 2011). CCE was calculated based on 10 μL template addition of the E. PBS samples with the 10 and 100 mL concentrates. coli positive control of 1.86 × 107 CFU/mL (7.27 Log10) in all  R.J. Kibbee, B. Örmeci / Journal of Microbiological Methods 132 (2017) 139–147 145 Fig. 6. The Sketa DNA amplification reduction after 200 mM PMA treatment showed a 9.53 Δ Cq (3.13 from 4.73 Log 10 genomic units) for the WWD samples and a 13.53 Δ Cq (4.41 from 6.97 Log 10 genomic units) for the high titer control sample; NTC samples showed primer dimers at a Cq of 31.08 setting the detection limit to 1.48 Log10 genomic units. experiments. This control was done as an amplification control as well 4. Conclusions as starting titer Cq value in which all test Cq values were compared be- fore CCE calculated with the uidA gene mean standard curve slope inter- This study developed a sensitive and false-positive free PMA-qPCR cept. Concentrations of estimated VBNC E. coli cells in the WWTP assay to enumerate VBNC E. coli after wastewater disinfection. DNA ex- samples ranged from 1.05–3.06 Log10 CCE/10 mL; this is a comparison traction pre-treatments were investigated to remove qPCR inhibitors, of viable CFU to intact cells (CCE) as determined by PMA treatment. primarily humic acids, and related substances. The results of this study The CFU/10 mL Log10 reductions ranged from 2.2–3.64. The CCE PMA showed that the chemical immobilization methods were less or as effec- treated Log10 reductions ranged from 0 to 0.41; this is a comparison of tive as a simple pre-DNA extraction PBS wash which showed no qPCR the WW intact cells (CCE) to the WWD intact cells (CCE) (Table 3). inhibition at the 10 × and 100 × concentration of the WW and WWD There is quite a discrepancy between the CFU and CCE PMA treated samples. The PMA validation showed that the amplification of the Log10 reductions. These data are representative of VBNC cells in the Sketa DNA in 10 × concentrated WWD samples was successfully WWD samples even without directly analyzing the efficiency of the inhibited by the PMA treatment, reducing the genomic unit concentra- PMA to enter into the compromised cells and observing its effects on tion by 3.13 Log10 from a starting concentration of 4.73 which is 1.08 the Sketa DNA. Log10 higher than the average uidA gene concentration of 3.65 Log10 in Fig. 7. The above standard curve plot of the uidA gene qPCR data shows linearity from Fig. 8. The above standard curve plot of the Sketa DNA amplicon qPCR data showing 50 ng – 50 fg of E. coli genomic DNA; where 50 fg is equal to 50 CFU/genomic units. linearity from 5 ng to 50 fg of Sketa DNA. 146 R.J. Kibbee, B. Örmeci / Journal of Microbiological Methods 132 (2017) 139–147 Table 3 Log10 CFU/CCE in 10 mL concentrated wastewater samples before and after chlorine disinfection. Sampling date Sample CFUa CCEb CCE PMA treated Estimated VBNC/intactc CFU reductiond CCE PMA treated reductione 04.10.2015 WW 3.31 4.34 3.92 0.61 2.71 0.26 WWD 0.6 4.42 3.66 3.06 06.18.2015 WW 4.07 3.96 3.25 0 (−0.82) 3.64 0 (−0.04) WWD 0.43 3.25 3.29 2.86 07.07.2015 WW 3.05 3.29 2.31 0 (−0.74) 2.2 0.41 WWD 0.85 2.65 1.9 1.05 Averages WW 3.48 3.86 3.16 0.2 2.85 0.21 WWD 0.63 3.44 2.95 2.32 a Colony forming units as determined by mTEC agar culture. b Calculated CFU equivalent based on the E. coli 50 ng positive control Cq in each experiment. c VBNC estimates calculated by subtracting the CFU from CCE PMA treated. d CFU reduction calculates by subtracting WWD CFU from WW CFU. e CCE PMA treated reduction calculated by subtracting the WWD CCE from the WW CCE. the WW and WWD samples. This control is reliable and showed using excitation-emission matrix spectroscopy and PARAFAC. Water Res. 47 (10), 3467–3476. a better reduction in amplification compared to other methods Haugland, R.A., Siefring, S.C., Wymer, L.J., Brenner, K.P., Dufour, A.P., 2005. Comparison of which use heat, freeze-thaw or chemicals to compromise cell mem- Enterococcus measurements in freshwater at two recreational beaches by quantita- brane integrity. tive polymerase chain reaction and membrane filter culture analysis. Water Res. 39, 559–568. The PMA-qPCR assay developed was able to show that there were Haugland, R.A., Varma, M., Sivaganesan, M., Kelty, C., Peed, L., Shanks, O.C., 2010. Evalua- high numbers of VBNC E. coli cells leaving the WWTP in the final ef- tion of genetic markers from the 16S rRNA gene V2 region for use in quantitative de- fluent post-chlorine treatment which passed undetected using cul- tection of selected bacteroidales species and human fecal waste by qPCR. Syst. Appl. Microbiol. 33 (6), 348–357. ture methods. The target quantity of E. coli in the final effluent of Heijnen, L., Medema, G., 2006. Quantitative detection of E. coli, E. coli O157 and other the WWTP is b200 CFU/100 mL to meet local regulations, and it shiga toxin producing E. coli in water samples using a culture method combined was shown in our culture findings that the WWD samples had a with real-time PCR. J. Water Health 4 (4), 487–498. Kibbee, R., Linklater, N., Ormeci, B., 2013. Eliminating false positives in a qPCR assay mean concentration of 46 CFU/100 mL. However, the PMA-qPCR for the detection of the uidA gene in Escherichia coli. J. Water Health 11 (3), CCE data showed an average of 2.01 × 103 CCE/100 mL in the final ef- 382–386. fluent; an order of magnitude higher than the recommended upper Lakay, F.M., Botha, A., Prior, B.A., 2007. Comparative analysis of environmental DNA ex- limit. traction and purification methods from different humic acid-rich soils. J. Appl. Microbiol. 102, 265–273. Li, L., Mendis, N., Trigui, H., Oliver, J.D., Faucher, S.P., 2014a. The importance of the vi- able but non-culturable state in human bacterial pathogens. Front. Microbiol. 5, Acknowledgements 258. Li, W.T., Chen, S.Y., Xu, Z.X., Li, Y., Shuang, C.D., Li, A.M., 2014b. Characterization of dis- solved organic matter in municipal wastewater using fluorescence PARAFAC analysis This work was supported by the Natural Sciences and Engineering and chromatography multi-excitation/emission scan: a comparative study. Environ. Council of Canada under the Discovery Grant program (RGPIN 326978). Sci. Technol. 48 (5), 2603–2609. Li, D., Tong, T., Zeng, S., Lin, Y., Wu, S., He, M., 2014c. Quantification of viable bacteria in wastewater treatment plants by using propidium monoazide combined with quanti- References tative PCR (PMA-qPCR). J. Environ. Sci. 26 (2), 299–306. Lloyd, K.G., Macgregor, B.J., Teske, A., 2010. Quantitative PCR methods for RNA and DNA in Ayrapetyan, M., Williams, T.C., Oliver, J.D., 2015. Bridging the gap between viable but non- marine sediments: maximizing yield while overcoming inhibition. FEMS Microbiol. culturable and antibiotic persistent bacteria. Trends Microbiol. 23 (1), 7–13. Ecol. 72 (1), 143–151. Bae, S., Wuertz, S., 2009. Discrimination of viable and dead fecal bacteroidales bacteria by Marti, E., Jofre, J., Balcazar, J.L., 2013. Prevalence of antibiotic resistance genes and bacte- quantitative PCR with propidium monoazide. Appl. Environ. Microbiol. 75: rial community composition in a river influenced by a wastewater treatment plant. 2940–2944. http://dx.doi.org/10.1128/AEM.01333-08. PLoS One 8 (10), e78906. Barbau-Piednoir, E., Mahillon, J., Pillyser, J., Coucke, W., Roosens, N.H., Botteldoorn, N., Martin, B., Raurich, S., Garriga, M., Aymerich, T., 2013. Effect of amplicon 2014. Evaluation of viability-qPCR detection system on viable and dead Salmonella length in propidium monoazide quantitative PCR for the enumeration serovar enteritidis. J. Microbiol. Methods 103, 131–137. of viable cells of salmonella in cooked ham. Food Anal. Methods 6 (2), Braid, M.D., Daniels, L.M., Kitts, C.L., 2003. Removal of PCR inhibitors from soil DNA by 683–690. chemical flocculation. J. Microbiol. Methods 52 (3), 389–393. Masters, C.I., Shallcross, J.A., Mackey, B.M., 1994. Effect of stress treatments on the detec- Cao, Y., Griffith, J.F., Dorevitch, S., Weisberg, S.B., 2012. Effectiveness of qPCR permuta- tion of Listeria monocytogenes and enterotoxigenic Escherichia coli by the polymerase tions, internal controls and dilution as means for minimizing the impact of inhibition chain reaction. J. Appl. Bacteriol. 77 (1), 73–79. while measuring Enterococcus in environmental waters. J. Appl. Microbiol. 113 (1), Matthews, B., Stratton, H.M., Schreoder, S., Toze, S., 2010. Pathogen detection technologies 66–75. for wastewater and reservoirs. Urban Water Security Research Alliance Technical Re- Contreras, P.J., Urrutia, H., Soss, K., Nocker, A., 2011. Effect of PCR amplicon length on sup- port No. 32 (ISSN 1836-5566 (Online), 1836-5558 (Print)). pressing signals from membrane-compromised cells by propidium monoazide treat- Nocker, A., Sossa-Fernandez, P., Burr, M.D., Camper, A.K., 2007. Use of propidium ment. J. Microbiol. Methods 87, 89–95. monoazide for live/dead. Appl. Environ. Microbiol. 73:5111–5117. http://dx.doi.org/ Domanico, M.J., Phillips, R.B., Oakley, T.H., 1997. Phylogenetic analysis of Pacific salmon 10.1128/AEM.02987-06. (genus Oncorhynchus) using nuclear and mitochondrial DNA sequences. Can. J. Fish. Oliver, J.D., Dagher, M., Linden, K., 2005. Induction of Escherichia coli and Salmonella Aquat. Sci. 54, 1865–1872. typhimurium into the viable but non-culturable state following chlorination of waste- Dong, D., Yan, A., Liu, H., Zhang, X., Xu, Y., 2006. Removal of humic substances from soil water. J. Water Health (3), 249–257. DNA using aluminium sulfate. J. Microbiol. Methods 66 (2), 217–222. Opel, K.L., Chung, D., McCord, B.R., 2009. A study of PCR inhibition mechanisms using real Du, M., Chen, I., Zhang, X., Li, A., Li, Y., Wang, Y., 2007. Retention of virulence in a viable time PCR. J. Forensic Sci. 55 (1), 25–33. but non-culturable Edwardsiella tarda isolate. Appl. Environ. Microbiol. 73, Pinto, D., Almeida, V., Santos, M.A., Chambel, L., 2011. Resuscitation of Escherichia coli 1349–1354. VBNC cells depends on a variety of environmental or chemical stimuli. J. Appl. Fatima, F., Pathak, N., Rastogi Verma, S., 2014. An improved method for soil DNA extrac- Microbiol. 110 (6), 1601–1611. tion to study the microbial assortment within rhizospheric region. Mol. Biol. Int. Taskin, B., Gozen, A.G., Duran, M., 2011. Selective quantification of viable Escherichia coli 2014:518960. http://dx.doi.org/10.1155/2014/518960. bacteria in biosolids by quantitative PCR with propidium monoazide modification. Gaillard, C., Strauss, F., 1998. Avoiding adsorption of DNA to polypropylene tubes and de- Appl. Environ. Microbiol. 77 (13), 4329–4335. naturation of short DNA fragments. Elsevier Technical Tips Online Vol. 3. U.S. Environmental Protection Agency, 2010. Method 1103.1: Escherichia coli (E. coli) in Gentry-Shields, J., Wang, A., Cory, R.M., Stewart, J.R., 2013. Determination of specific Water by Membrane Filtration Using Membrane-Thermotolerant Escherichia types and relative levels of QPCR inhibitors in environmental water samples coli Agar (mTEC), EPA-821-R-10-002. U.S. Environmental Protection Agency,  R.J. Kibbee, B. Örmeci / Journal of Microbiological Methods 132 (2017) 139–147 147 Office of Water (4303 T) 1200 Pennsylvania Avenue, NW, Washington, DC van Frankenhuyzen, J.K., Trevors, J.T., Lee, H., Flemming, C.A., Habash, M.B., 2011. Molec- 20460. ular pathogen detection in biosolids with a focus on quantitative PCR using U.S. Environmental Protection Agency, 2012. Method 1611: Enterococci in Water by propidium monoazide for viable cell enumeration. J. Microbiol. Methods 87 (3), TaqMan® Quantitative Polymerase Chain Reaction (qPCR) Assay, EPA-821-R-12- 263–272. 008. U.S. Environmental Protection Agency, Office of Water (4303T) 1200 Pennsylva- Varma, M., Field, R., Stinson, M., Rukovets, B., Wymer, L., Haugland, R., 2009. Quantitative nia Avenue, NW, Washington, DC 20460. real-time PCR analysis of total and propidium monoazide-resistant fecal indicator Vakondios, N., Koukouraki, E.E., Diamadopoulos, E., 2014. Effluent organic matter (EfOM) bacteria in wastewater. Water Res. 43 (19), 4790–4801. characterization by simultaneous measurement of proteins and humic matter. Water Res. 63, 62–70.