lable at ScienceDirect

Water Research 132 (2018) 61e70

Contents lists avai

Water Research

journal homepage: www.elsevier .com/locate/watres

Pepper mild mottle virus as a process indicator at drinking watertreatment plants employing coagulation-sedimentation, rapid sandfiltration, ozonation, and biological activated carbon treatments inJapan

Ryuichi Kato a, Tatsuya Asami a, Etsuko Utagawa a, Hiroaki Furumai b,Hiroyuki Katayama a, c, *

a Department of Urban Engineering, Graduate School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japanb Research Center for Water Environment Technology, School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japanc Vietnam - Japan University, Luu Huu Phuoc Str., My Dinh I Ward, Nam Tu Liem Dist., Hanoi, Viet Nam

a r t i c l e i n f o

Article history:Received 21 September 2017Received in revised form13 December 2017Accepted 23 December 2017Available online 27 December 2017

Keywords:Virus reductionQuantitative PCRDrinking water treatment plantPepper mild mottle virusIndicator

* Corresponding author. Department of Urban EngEngineering, The University of Tokyo, 7-3-1 Hongo,Japan.

E-mail address: katayama@env.t.u-tokyo.ac.jp (H.

https://doi.org/10.1016/j.watres.2017.12.0680043-1354/© 2017 Elsevier Ltd. All rights reserved.

a b s t r a c t

To assess the potential of pepper mild mottle virus (PMMoV) as a viral process indicator, its reductionthrough coagulationesedimentation (CS) and rapid sand filtration (RSF) were compared with those ofEscherichia coli, previously used viral indicators, and norovirus genotype II (NoV GII; enteric virusreference pathogen) in a bench-scale experiment. PMMoV log10 reductions in CS (1.96 ± 0.30) and RSF(0.26± 0.38) were similar to those of NoV GII (1.86 ± 0.61 and 0.28 ± 0.46). PMMoV, the most abundantviruses in the raw water, was also determined during CS, RSF, and advanced treatment processes at twofull-scale drinking water treatment plants under strict turbidity management over a 13-month period.PMMoV was concentrated from large-volume water samples (10e614 L) and quantified by Taqman-based quantitative polymerase chain reaction. The PMMoV log10 reduction in CS (2.38 ± 0.74, n¼ 13and 2.63 ± 0.76, n¼ 10 each for Plant A and B) and in ozonation (1.91± 1.18, n¼ 5, Plant A) greatlycontributed to the overall log10 reduction. Our results suggest that PMMoV can act as a useful treatmentprocess indicator of enteric viruses and can be used to monitor the log10 reduction of individual treat-ment processes at drinking water treatment plants due to its high and consistent copy numbers in sourcewater.

© 2017 Elsevier Ltd. All rights reserved.

1. Introduction

Microbial safety management of drinking water has beenchanged after Cryptosporidium caused several outbreaks worldwidein the 1990s (Mackenzie et al., 1994) due to its tolerance to chlorine.In Japan, the Interim Guideline for Cryptosporidium Treatment inthe Water Supply was established in 1996 and reformed in 2007,where a strict turbidity control policy was employed as a criticalcontrol parameter not to exceed 0.1� (approximately 0.14 NTU)after filtration treatment. Careful operation has been in practice,including minute control of the coagulant dosage, re-addition of a

ineering, Graduate School ofBunkyo-ku, Tokyo 113-8656,

Katayama).

small amount of coagulant after sedimentation/before sand filtra-tion, and initiation at a lower flow rate after backwashing of thesand filter. No outbreaks of Cryptosporidium via the water supplyhave been reported in Japan since implementation of the guideline;yet, its efficiency at protecting the drinking water supply fromenteric viruses, which may more easily pass through coagulation/sedimentation and filtration process due to their small size, re-mains uncertain.

As a means of managing the risk of illness due to public con-sumption of drinking water, quantitative microbial risk assessment(QMRA) has been conducted (Masago et al., 2006; Smeets et al.,2009) and is also included in the Water Safety Plan recom-mended by the World Health Organization (WHO, 2011). In QMRAprocesses, log10 reduction values (LRVs) corresponding to virusesreduced in each treatment process are a key parameter in assessingthe virus concentration in treated water using pathogen levels

R. Kato et al. / Water Research 132 (2018) 61e7062

determined from raw water (RW) (Teunis et al., 1997). Althoughseveral studies have evaluated virus reduction in pilot or bench-scale studies using test water spiked with viruses (Abbaszadeganet al., 2007, 2008; Boudaud et al., 2012; Shin and Sobsey, 2015;Shirasaki et al., 2010), virus reduction in the laboratory may notfully represent reduction in full-scale DWTPs (Medema et al.,2006). In a QMRA in The Netherlands, plant scale data werepreferred, followed by pilot plant scale data, over data from labo-ratory experiments (Schijven et al., 2011).

Studies associatedwith virus reduction at full-scale DWTPs havebeen limited so far due to the challenges of detecting low con-centration of viruses in water (Albinana-Gimenez et al., 2009;Asami et al., 2016), where a large volume of water samples mustbe concentrated. During virus concentration procedures, polymer-ase chain reaction (PCR) inhibiting substances are co-concentratedwith viruses and interfere with the following molecular methods:RNA extraction, reverse transcription (RT) and quantitative PCR(qPCR). Many researchers have dealt with this problem (Schraderet al., 2012), and PCR inhibition was found to be caused mainlyby low molecular weight organic fractions (Hata et al., 2015a).Although some procedures have effectively removed or mitigatedPCR inhibition (Gibson et al., 2012; Hata et al., 2011), a commonstrategy that mitigates PCR inhibition has not been widely estab-lished. Recently, Asami et al. (2016) quantified the reduction effi-ciency of viruses at a full-scale DWTP located in Bangkok, Thailand,using indigenous pepper mild mottle virus (PMMoV) and JC poly-omavirus (JC PyV) as treatment process indicators. However, virusreduction during water treatment in Japan may be different fromthat of other countries given the previously mentioned strictturbidity control policy for Cryptosporidium safety management.Therefore, it is necessary to monitor virus reduction for Japanesewater treatment systems specifically.

To observe the fate of enteric viruses in water, indicator virusesare often targeted due to their higher relative concentrationcompared with low concentration indigenous enteric viruses in thesource and treated water (Haramoto et al., 2004; Kittigul et al.,2012). Recently, a number of studies have compared the preva-lence of PMMoV to enteric viruses in feces, wastewater and envi-ronmental waters (Rosario et al., 2009; Zhang et al., 2006; Asamiet al., 2016; Hamza et al., 2011; Haramoto et al., 2013; Kurodaet al., 2015; Rachmadi et al., 2016). PMMoV showed greaterpersistence than human enteric viruses in surface water (Symondset al., 2016), wastewater (Kitajima et al., 2014b; Schmitz et al., 2016;Symonds et al., 2015, 2014), in wetland treatment (Rachmadi et al.,2016), and in membrane filtration (Shirasaki et al., 2017); thissuggest PMMoV is useful as an indicator representing the efficiencyof virus removal via a treatment process. However, the efficiency ofa drinking water treatment process at reducing PMMoV in com-parison with enteric viruses (including human noroviruses, one ofthe major causes of waterborne nonbacterial gastroenteritis[Maunula et al., 2005; Parshionikar et al., 2003]) has not yet beenexamined. Since PMMoV and enteric viruses are different from amorphological viewpoint, it is necessary to evaluate whetherPMMoV can act as an indicator of the presence of enteric viruses.

Noroviruses, which are found in water and cause waterbornedisease, are the most common cause of pediatric gastroenteritis(Katayama and Vinje, 2017); for this reason, norovirus genotype II(NoV GII) was used as a reference pathogen for comparison withPMMoV in this study instead of using other surrogates(Abbaszadegan et al., 2008; Bae and Schwab, 2008; Boudaud et al.,2012). For the bench-scale experiments in this study, severalreference indicators were adopted to compare their reductionduring the treatments with that of PMMoV: murine norovirus(MNV), often used as a human norovirus surrogate; aichivirus (AiV)as one of the most prevalent virus in water environment; MS2 and

Qb, the most widely used bacteriophages used as virus indicators inwater treatment studies; and E. coli, which is used as a fecalindicator.

To overcome the difficulty in measuring low concentration ofindigenous viruses in water, we needed to investigate how tomeasure, remove, or mitigate PCR inhibition. Cucumber greenmottle mosaic virus (CGMMV), which, like PMMoV, belongs to theTobamovirus group, was introduced as amolecular control due to itsphylogenetical and morphological similarity to the target virus (i.e.,PMMoV).

Our study was composed of two main parts: a bench-scaleexperiment using coagulationesedimentation (CS) and rapid sandfiltration (RSF), as well as a field survey at full-scale DWTPs. The aimof the bench-scale experiment was to evaluate the behavior ofPMMoV as a treatment process indicator for enteric viruses,including human norovirus. The aims of the full-scale DWTP surveywere (1) to investigate whether the indigenous PMMoV copynumber in RW was high enough to evaluate the stepwise virusreduction efficiency in DWTPs; (2) to evaluate the moleculardetection inhibition caused by co-concentrated organic matter andto verify methods tomitigate this inhibition; and (3) to evaluate thevirus reduction efficiency of CS, ozone, biological activated carbon(BAC), and RSF process at two full-scale DWTPs under strictturbidity control in Japan.

2. Material and methods

2.1. Physicochemical water quality parameters

Water temperature, pH and electrical conductivity (EC) weremeasured immediately after collecting the samples using an HI98129 water tester (HANNA, Japan). Turbidity was measured with aDR/890 portable colorimeter (HACH, Japan).

2.2. Quantification of microbes

2.2.1. Quantification of indicator bacteriaWater samples containing chlorine were collected in plastic

bags in which sodium thiosulfate was added in advance fordechlorination. Concentrations of E. coli and total coliforms werequantified within 6 h of sample collection for all water samplescollected at DWTPs using a filter unit (37-mm 192 MONITOR UNIT;37-mm diameter, 0.45-mm pore size, ADVANTEC) and m-Col-iBlue24® Broth (HACH). In bench-scale experiment, E. coli wasquantified within 6 h of sample collection using Chromocult®

Coliform Agar, (Merck Millipore, Japan) following the manufac-turer's instructions.

2.2.2. Virus quantification by RT-qPCRViral RNA was extracted from 140 mL of concentrated samples

and reverse-transcribed following previously established methods(Asami et al., 2016). The primers, probes and positive controls areshown in Table A1, further details of PCR shown in the supple-mental information.

2.2.3. Evaluation of viral RNA extraction and RT-qPCR efficiencyIn order to evaluate RNA extraction and RT-qPCR efficiency,

CGMMV, provided by the National Institute of Agrobiological Sci-ences (NIAS, Japan), was added as a molecular control based on itsphylogenetic and morphological similarity to the target virus(PMMoV). Although MNV has been widely introduced as a molec-ular control in previous studies (Asami et al., 2016; Kitajima et al.,2014a), it is phylogenetically and morphologically different fromPMMoV; thus, CGMMV,which has higher similarity to PMMoV, waschosen as a more appropriate molecular control for this study.

R. Kato et al. / Water Research 132 (2018) 61e70 63

Briefly, 1.0 mL of CGMMV was added into all concentrated watersamples as wll as a control sample with 140 mL of DNase/RNase-freedistilled water to obtain a final copy number of approximately 105

copies per sample, as described previously (Hata et al., 2013). TheCGMMV RT-qPCR recovery efficiency was determined bymeasuring the copies in each water sample and comparing it to thecontrol using qPCR and the following equation:

MME¼ C/C0� 100

MME: Molecular method (RNA extraction plus RT-qPCR)efficiencyC: detected CGMMV copy numbers obtained from eachconcentrated water sampleC0: detected CGMMV copy numbers obtained from the control

AnMME value of�10% was used as a criterion to adopt the viruscopy number in awater sample; if the sample was <10% then it wassubjected to a further sample purification process (describedbelow). The MME values in the bench-scale experiment were largerthan 10% throughout the study. MME values were used only toevaluate sample inhibition tendency and were not used to back-calculate virus copy numbers.

2.2.4. Calculation of virus log10 reduction and statistical analysisThe reduction of viruses, indicator bacteria, and turbidity during

each treatment process in the bench-scale experiment and in full-scale DWTPs were calculated as LRVs only when their copynumbers were quantified before and after treatment. Statisticalanalysis (t-test) was performed using Microsoft Excel 2013 toevaluate the significant difference between log10 reductionscomparing the tested parameters.

2.3. Bench-scale experiment

2.3.1. Preparation of microbes for the spike experimentOne milliliter of E. coli (K12 IFO3301, Institute for Fermentation,

Japan) was inoculated into 9mL of lysogeny broth (Invitrogen, USA)to obtain approximately 10mL of cultured solution. This culturewas centrifuged at 6000 rpm for 10min to pellet the bacteria, thesupernatant was removed and the pelleted bacteria were resus-pended in 10mL of phosphate buffered saline. This wash processwas repeated three times.

Bacteriophage Qb (ATCC 23631) and MS2 (ATCC 15597) wereobtained from the American Type Culture Collection (Manassas,Virginia, USA) andwere cultivated using E. coli K12A/l (Fþ) as a host(Furuse et al., 1978). MNV (strain S7-PP3) was provided courtesy ofDr. Tohya (Nihon University, Japan) and propagated in RAW 264.7(ATCC TIB-71) cells. AiV (strain A846/88) was provided courtesy ofDr. Yamashita (Syubun University, Japan) and propagated in buffalogreen monkey cells provided courtesy of Dr. Yano (Tokyo Metro-politan Institute of Public Health, Japan). NoV GII (strain GII/6) wasprovided by Dr. Utagawa (University of Tokyo, Japan) from the fecalsample of an infected patient, purified by ultracentrifugation withCsCl density gradient. PMMoV and its host Nicotiana benthamianawere provided by the National Agriculture and Food Research Or-ganization, Japan.

2.3.2. Bench-scale CS experimentSixty liters of raw water (RW) were collected monthly from

August to December 2016 at Plant B on the same day of full-scaleDWTP sampling, transferred to a laboratory at the University ofTokyo, Japan, and subjected to bench-scale water treatment in the

laboratory within the sampling day.The CS laboratory experiment was conducted using 40 L of the

collected plant B RW. The following microbes were added into theRW after equilibrating to room temperature (15e25 �C); 10mL ofE. coli (IFO3301), up to 2ml each of PMMoV, bacteriophage Qb, MS2,AiV, MNV (final concentration >103 copy/ml), and NoV GII. PACl(basicity: 54%) provided by Plant B was used for the coagulationexperiments; the pH was adjusted to approximately 7.0 usingH2SO4 while mixing with a stirrer. The dose rate of H2SO4 and PACl(17e40mg/L) followed the actual treatment condition of Plant B onthe same sampling day. CS was conducted in accordance with thejar-test condition of Plant B, with slight modification of the settlingtime because the test water volume in our study was different fromthe jar test in Plant B. Briefly, after coagulant addition the test waterwas mixed rapidly for 2min (120 rpm) and slowly for 10min(60 rpm), and then left for 15min for floc settling. Supernatantswere collected after settling, of which 2000mL, 50mL, and 350mLwere used to quantify the virus, detect E. coli, and analyze basicwater quality, respectively.

2.3.3. Bench-scale RSF experimentThe sand filtration column and treatment condition were

designed in accordance with design criteria for waterworks facil-ities in Japan. Standardized sand (effective diameter: 0.6mm, uni-formity coefficient: 1.5, Nihon Genryo, Japan) was used as a filtermedia in the RSF process to obtain a 65-cm filter depth in a 8-cmdiameter tubular tower with a gravel layer basement. Flow ratewas adjusted to 120e150m/day (z0.13e0.18 cm/s in this experi-mental condition) using a peristaltic pump. The sand filtrationcolumn was backwashed with 60 L of tap water after each experi-ment to reset the media condition and was saturated with tapwater. After resetting the media condition, the virus copy numberin filtered tap water was confirmed to be below the detection limitor at a negligible level compared to the virus concentration of thesample water (data not shown).

To evaluate the step-wise reduction of microbes by CS and RSFin a similar fashion to a full-scale system, water from the finishedCS experiment was used for the influent of the sand filtration col-umn. Filtered water samples were collected after 15 L of superna-tant was filtered in order to avoid the effect of the tap water beingsaturated in the column prior to the experiment.

2.3.4. Sample collectionWater samples were collected at each step of RW treatment

after each of the following steps: addition of microbes (n¼ 5), CS(n¼ 5), and RSF (n¼ 5). Sample concentration steps are explainedbelow. Concentrated sample volumes were 500mL for RW and1000mL for both post CS and post RSF. Indigenous virus copynumber in the RW before addition of experimental viruses wasdetermined for every sampling occasion and was found to benegligible compared with that of the inoculated viruses (data notshown).

2.3.5. Virus concentration methodWater samples were concentrated using two different primary

concentration methods: (1) an acid rinseealkaline elution methodusing an electronegative membrane filter, as previously described(Katayama et al., 2002), and (2) a slightly modified method. Briefly,for method (1), MgCl2 was added to the water sample to obtain afinal concentration of 25mmol/L and the samples were passedthrough an electronegative membrane (HA, 90-mm diameter, 0.45-mm pore, Millipore, Tokyo, Japan). Magnesium ions were thenremoved by passing 200mL of 0.5mmol/L H2SO4 through themembrane (acid rinse). Viruses adsorbed on the filter were elutedusing 10mL of 1.0mmol/L NaOH (alkaline elution). The eluate was

R. Kato et al. / Water Research 132 (2018) 61e7064

recovered in a plastic tube containing 50 mL of 0.5mol/L H2SO4 and100 mL of 100� Tris-EDTA buffer (pH 8.0) for neutralization andfurther concentrated using a Centriprep YM-50 filter unit (MerckMillipore, Japan) to obtain a final volume of approximately 700 mL.

For method (2), a slight modification for the primary virusconcentration step was performed. A previous study reported thatit was sometimes difficult to elute viruses adsorbed on membranefilters simply through filtration with NaOH (Ahmed et al., 2015).Specifically in the current study, it is possible that the coagulantremaining after CS could interfere with the virus concentration,especially the alkaline elution process. Therefore, instead of addingMgCl2, pH levels of the water samples were adjusted to acid con-dition (pH 3.5) using 0.5mol/L H2SO4 in order to dissolve thealuminum coagulant, and the water samples were then passedthrough an electronegative membrane. The subsequent procedurewas the same as that of the original method, and resulted in a finalvolume of approximately 700 mL. The virus copy number of eachsample was determined using both primary concentrationmethods; the highest virus copy number obtained in each of thetwo methods was used for subsequent analysis.

2.4. Full-scale DWTP sampling

2.4.1. Water samples testedWater samples were collected monthly from December 2015 to

November 2016 at Plant A, and fromMarch 2016 to January 2017 atPlant B, both of which are located in an urban area of Japan. TheseDWTPs receive river water and produce yearly averages of7.9� 105m3 (Plant A) and 9.4� 105m3 (Plant B) of drinking waterper day.

The characteristics of these DWTPs as well as sampling pointsare shown in Fig. 1. Chlorine was injected before CS from Decemberto March at both DWTPs. Polyaluminum chloride (PACl) was usedas a coagulant at both plants at a range of 17e40mg/L. A smallamount of coagulant (�1mg/L) was re-added before RSF at Plant Aand before secondary RSF at Plant B to enhance the treatment ef-ficiency of RSF.

At Plant A, samples were collected once a month from RW(n¼ 13), from post CS (n¼ 13), from post ozone (n¼ 13), from postBAC (n¼ 13), and from post RSF (n¼ 13) to evaluate the reductionefficiency of viruses through thewhole treatment process including

Fig. 1. Process flow and sample volume of two drink

advanced treatment processes. At Plant B, samples were collectedfrom RW (n¼ 11), from post CS (n¼ 11) and from post primaryRSF(n¼ 11) to evaluate the virus reduction efficiency. Samples werecollected sequentially in the order of the treatment sequence fromRW to after treatments on the same day, but with shorter intervalsthan the hydraulic retention time of each treatment.

2.4.2. Onsite virus concentration methodLarge-volume water samples (20 L for RW,100 L for post CS, and

500 L for treated water) were concentrated by the acid rinse-ealkaline elution method with a cartridge-type electronegativefilter on site, as described previously (Hata et al., 2015b), and storedat 4 �C for further analysis. For the RW samples, the 20 L watersample was collected in a gamma-sterilized container and MgCl2was added directly to obtain a final concentration of 25mmol/L. Forother samples, water samples weremixedwithMgCl2 continuouslyby gravity flow to obtain a final concentration of 25mmol/L and thiswas passed through an electronegative membrane filter (Asamiet al., 2016). Viral RNA was extracted from concentrated samples,subjected to RT and quantified by qPCR, as described in sections2.2.2.

2.4.3. Large-volume, primary virus concentration method efficiencyDue to the fact that the target DWTPs used PACl as a coagulant,

which may interfere with the electrostatic nature of the virusconcentration method, it was important to determine the virusconcentration efficiency (recovery yield) for the large-volume pri-mary virus concentration method used in this study. However, theconcentration method for DWTP samples requires hundreds of li-ters of water, leading to difficulties performing the virus-spikingexperiment for direct evaluation of the recovery yield. Therefore,a one-hundredth-scale laboratory experiment was conducted toevaluate the efficiency of the primary virus concentration methodfrom March 2016 to January 2017. The same virus concentrationmethod described in the previous sectionwas used in this recoveryexperiment. Water samples were collected on each samplingoccasion in a sterilized plastic bag and transported to the laboratoryin a cooler with ice packs. PMMoV was added into the watersamples to obtain a final copy number of approximately 106e107

copies per sample. Sample volumes in the recovery experimentwere calculated as one-hundredth of the actual DWTP sample

ing water treatment plants and sampling points.

R. Kato et al. / Water Research 132 (2018) 61e70 65

volumes. Similarly, electronegative membrane filter used for therecovery experiment (HA, 47-mm diameter, 0.45-mm pore, Milli-pore, Japan) had an approximately 100-fold reduced area (9.6 cm2)compared to the cartridge-type filter used for the DWTP samples(1000 cm2

filtration area). Subsequent processes were completed asdescribed previously (section 2.3.5, method (1)) to obtain 10mL ofthe primary concentrate. Recovery yield (Y) was calculated asfollows:

Y¼X/(C0� V)� 100

Y: Recovery yield of the primary concentration methodX: Recovered PMMoV copy number (copies)C0: Stock PMMoV copy number added into test water (copy/mL)V: Added stock PMMoV volume (mL)

Finally, the geometric mean of Y was calculated to evaluate therecovery tendency of the primary virus concentration.

2.4.4. Evaluation of viral RNA extraction and RT-qPCR efficiencyViral RNA extraction and RT-qPCR efficiency for the full scale

plant samples were evaluated as described in section 2.2.3. Whenthe MME value was lower than 10%, the water sample was sub-jected to a subsequent sample purification process describedbelow.

2.4.5. Sample purification and dilution to mitigate the inhibition ofviral detection

Dilution or sample purificationwas applied to inhibited samples(MME <10%) prior to the molecular methods. According to a pre-vious study (Gibson et al., 2012), sample dilution prior to RT-qPCRcan effectively mitigate inhibition. In addition, a purificationmethod used to remove humic acids from nucleic acid extractsusing the polymeric adsorbent Supelite™ DAX-8 was employed,which was developed previously (Schriewer et al., 2011) andmodified by Dr. Akihiko Hata (private communication). Briefly, a

Fig. 2. Flow chart depicting determination of virus concentration in samples corresponDil: dilution.a MME represents recovery efficiency of CGMMV as a molecular control during RNA extracb Cp represents indigenous pepper mild mottle virus concentration in the sample.

hydrophobic resin (DAX-8) was purified following the methoddescribed by (Thurman and Malcolm, 1981) and stored in an equalvolume of 99.5% ethanol. The mixture of DAX-8/ethanol wasintroduced to the RNA extraction process, as described in a previousstudy (Asami et al., 2016). The mixture of DAX-8/ethanol wascombined at a ratio of 1:1 with 140 mL of the concentrated sampleand mixed well. The DAX-8/sample mixture was filtered using anULTRAFREE® centrifugal filter unit (PVDF, 0.45-mm pore size, Mil-lipore). This primary filtrate was then combinedwith 560 mL of lysisbuffer (Buffer AVL) provided with the RNA extraction kit mentionedpreviously. In addition, 560 mL of Buffer AVL was mixed with theDAX-8 resin retained in the filter unit and incubated for 10min atroom temperature (15e25 �C), from which secondary filtrate wasrecovered in the filter unit. This secondary filtrate was then com-bined with the primary filtrate/Buffer AVL mixture. The combinedprimary and secondary filtrate was subjected to RNA extraction, asdescribed above. The extracted RNA was additionally subjected togel filtration chromatography using illustra™ Microspin™ S-300HR columns (exclusion limit, 118 bpz 70 kDa; GE Healthcare) toremove low molecular weight organic matter following the man-ufacturer's protocol.

A flow chart depicting the determination of the virus copynumber in samples corresponding to the extent of inhibition isshown in Fig. 2. First, the virus copy number in the sample and theMME value with and without 10� dilution prior to RT-qPCR weredetermined (Step 1). When the MME value of the sample wasgreater than 10%, the higher value for either condition (with orwithout 10� dilution) was adopted as an indigenous PMMoV copynumber for analysis, even if the indigenous PMMoV copy numberwas negative or below the quantification limit. In cases where bothMME values with and without 10� dilution were greater than 10%,the higher PMMoV copy number was used for analysis. If bothMMEvalues were less than 10%, extracted RNA was subjected to100� dilution prior to RT-qPCR and judged in the same manner asStep 1 (Step 2). If MME values remained less than 10% for100� diluted samples, these were then subjected to DAX-8 treat-ment with and without 10� dilution prior to RNA extraction andjudged in the same manner as Step 1 (Step 3). In cases where all of

ding to poor molecular method efficiency.

tion and RT-qPCR.

Fig. 3. Log10 reduction efficiencies for viruses, E. coli, and turbidity in bench-scalecoagulationesedimentation, and rapid sand filtration processes.Reduction efficiency values at sampling occasion immediately after typhoon wereseparated from normal conditions.AiV: Aichi virus; CS: coagulationesedimentation; MNV: murine norovirus; RSF: rapidsand filtration.* Average log10 reduction of Qb at RSF in normal conditions was negative.n¼ 4 for normal and n¼ 1 for typhoon (NoV GII in typhoon period was not evaluated).Error bar represents standard deviation.

R. Kato et al. / Water Research 132 (2018) 61e7066

the MME values remained less than 10%, the highest indigenousPMMoV copy number obtained was used for analysis.

3. Results and discussion

3.1. Bench-scale experiment

The ranges of temperature, pH, turbidity, EC and E. coli in the testwaters (Table A2) showed fluctuations in all water quality param-eters during the sampling period, likely due to seasonality andrainfall. This experiment was conducted from summer (August) towinter (December) in Japan, including one sampling day immedi-ately after a typhoon (August).

Fig. 3 shows LRVs of viruses, E. coli, and turbidity in CS and RSFrespectively in the bench-scale experiment. Data from one sam-pling day (after the typhoon) are shown in a separate bar due to thegreat differences observed in RW quality (turbidity> 100 NTU).Average virus LRVs by CS, with the exception of the typhoon period,were as follows: PMMoV: 1.96± 0.30 log10; AiV: 1.83± 0.27 log10;MNV: 1.86± 0.37 log10; NoV GII: 1.86± 0.61 log10; Qb: 1.42± 0.27log10; MS2: 1.44± 0.41 log10 (mean± standard deviation (SD), n¼ 4for each), and were higher than those values from the typhoonperiod for all viruses. Average virus LRVs in RSF, with the exceptionof the typhoon period, were as follows: PMMoV: 0.26± 0.38 log10;AiV: 0.62± 0.49 log10; MNV: 0.40± 0.63 log10; NoV GII: 0.28± 0.46log10; Qb: �0.10± 0.32 log10; MS2: 0.99± 0.45 log10 (mean± SD,n¼ 4). In the August sample affected by the typhoon, less removalvia CS but greater removal via RSF was observed. The turbidity inRW was much different from that of other samples, which mayhave caused the change in interaction between coagulants andviruses, despite the fact that coagulant dosage was adjusted basedon the jar test. It is possible that the reduction tendency in eachtreatment process was affected by the RW quality. Hereinafter, theanalysis was conducted using all data except that affected by thetyphoon.

The removal of NoV GII in CS in this study was greater than thatof NoV GI (1.5 log10) using alum (Shin and Sobsey, 2015) or re-combinant norovirus virus-like particles (approximately 1 log10)using PACl, alum or FeCl3 (Shirasaki et al., 2010). The range of

PMMoV LRVs was quite similar to those of enteric viruses tested,including NoV GII. The LRVs for PMMoV and NoV GII were notsignificantly different for CS and RSF (t-test, p > .05). Positive cor-relations between the PMMoV and NoVGII LRVs in CS and RSF werealso found (r ¼ 0.72, y ¼ 1.45 x � 0.98 for CS; r ¼ 0.86, y ¼ 1.03x þ 0.02 for RSF; x represents PMMoV LRVs, y represents NoV GIILRVs). The log10 reductions in AiV, MNV, Qb, MS2 as well as NoV GIIin CS and RSF, were more similar to that of PMMoV than to that ofE. coli or turbidity (except for Qb log10 reductions in RSF which wasnot determined). These results suggest that PMMoV may be auseful treatment process indicator of NoV GII in CS and RSFprocesses.

3.2. Full-scale DWTPs experiment

3.2.1. Indicator bacteria and physicochemical parameters at full-scale DWTPs

The ranges of temperature, pH, turbidity, EC, E. coli, and totalcoliforms in full-scale DWTPs are shown in Table A3. As describedin section 3.1, fluctuations were observed in all water qualitiesduring the sampling period, especially in terms of concentrationranges of indicator bacteria. Turbidity, which is adopted as a criticalcontrol parameter for Cryptosporidium management for treatmentprocesses in Japan, was reduced greatly in CS at both plants A and B.The turbidity after RSF at Plant A was less than 0.14 NTU on allsampling occasions, whereas that of Plant B sometimes exceeded0.14 NTU. In the latter case, further treatment such as advancedtreatment and secondary RSF are performed; the effluent fromsecondary RSF is maintained below 0.14 NTU to meet the criteria.

3.2.2. Recovery yield of primary virus concentration for DWTPsamples

The recovery yields of primary virus concentration from DWTPsamples (one-hundredth-scale laboratory experiment) are shownin Fig. 4. Most recovery yields were larger than 10%, showing thatthe method employed functioned efficiently.

RFS at Plant A showed a remarkably low recovery yield; this wassuspected to be due to a high concentration of residual chlorine. Toassess this further, post RSF samples collected at Plant A in August,October, and November 2016 were dechlorinated with sodiumthiosulfate and provided for recovery experiment. The recoveryyield of the dechlorinated samplewas 101.7±0.28% (n¼ 3), whichwasmuch higher compared with that of the chlorinated control sample(10�0.26±0.53%, n¼ 3). Because the recovery yield of CS at Plant Bwas also less than 10%, residual chlorine was likely the main causeof low recovery; thus, degradation of the virus by chlorinationduring the process of virus concentration should be consideredwhen assessing the virus recovery yield in the laboratory. Becausevirus degradation by chlorine does not hinder virus recovery duringthe process of virus concentration, primary virus concentrations foractual treatment plant samples were assumed to have performedwell.

3.2.3. Viral RNA extraction and RT-qPCR efficiency using molecularcontrol CGMMV

In order to evaluate the efficiency of RNA extraction and RT-qPCR, a known amount of CGMMV was added to concentratedsamples as amolecular control. PMMoV copy number in the sample(Cp) was determined in cases where the MME value was higherthan 10% (recovery of added CGMMV). A total of 98 samples weresubjected to RNA extraction and RT-qPCR with and without10� dilution before the RT process (step 1). Cp values were deter-mined in 81 samples (82.7%), with either the diluted or undilutedsamples showing MME values of >10%. Among the 81 samples, Cpwas determined in 19 untreated samples and in 62 samples with

Fig. 4. Average recovery yield of PMMoV in one-hundredth-scale of onsite primary concentration.BAC: biological activated carbon; CS: coagulationesedimentation; PMMoV: pepper mild mottle virus; RSF: rapid sand filtration; RSF*: RSF sample with dechlorinated treatment;RW: raw water.n¼ 9 for Plant A (n¼ 8 for RSF, n¼ 3 for RSF*) and n¼ 11 for Plant B.Error bar represents standard deviation.

R. Kato et al. / Water Research 132 (2018) 61e70 67

10� dilution (Table 1). The remaining 17 samples were subjected toRNA extraction and RT-qPCR with 100� dilution before the RTprocess (step 2). However, inhibition was mitigated in only twosamples (2/17, 11.8%) with this treatment. This result suggests thatsample dilution before RT can effectively mitigate inhibition butmitigationwas limitedwhen the dilution factor was increased up to100� . The remaining 15 inhibited samples were subjected to DAX-8 treatment with RNA extraction followed by RT-qPCR with andwithout 10� dilution before the RT process (step 3), which wasable to improve MME values to >10% for all the samples. Cp wasdetermined in 5 samples with 10� dilution and 10 sampleswithout dilution with the DAX-8 treatment before RT. According toour results, inhibition was effectively removed by the DAX-8treatment when it was not simply mitigated via sample dilutionafter RNA extraction, which may indicate that the RNA extractionstep was inhibited (i.e., rather than the RT step or the PCR).

In summary, the evaluation of extraction/RT-qPCR efficiency viaCGMMV molecular control demonstrated that inhibition wasgreatly mitigated by sample dilution before RT. When inhibitionwas not mitigated by 10� dilution, DAX-8 treatment combinedwith the RNA extraction step was sufficient to mitigate inhibition.As a result, inhibitionwas mitigated sufficiently in all samples, thusenabling better quantification of target virus copy number.

3.2.4. Quantification of PMMoV in water samples at DWTPsPMMoV copy number in a total of 98 samples was determined

by RT-qPCR. Positive ratio and quantified ratio are summarized in

Table 1Inhibition tendency in RNA extraction and RT-qPCR using cucumber green mottle mosai

Step 1

No treatment 10 times dil.c

MMEa < 10% 53/98 19/98MMEa � 10% 45/98 79/98Remain undetermined 17/98Adopted for Cpb 19/98 62/98

Dil.: dilution.a MME represents recovery efficiency of CGMMV as a molecular control during RNA eb Cp represents pepper mild mottle virus concentration in sample.c Sample dilution before reverse transcription.

Table A4. In our study, when the copy number was higher than thequantification limit of one copy per well, the value was regarded asquantified, and when it was lower than the limit then it wasregarded as positive but unquantified which was not used in thesubsequent analysis. All RW and post CS samples at Plant A werequantified, however the quantified positive ratio was largelydecreased after the ozone process. This finding suggests a sub-stantial contribution of ozone in reducing virus copy number. Onthe other hand, all RW, post CS, and post RSF samples were quan-tifiable at Plant B, with the exception of one post CS sample.

PMMoV copy number for each treatment process in two DWTPswas summarized in Fig. 5. PMMoV copy numbers (mean± SD) wereas follows: 105.35±0.48 copies/L (Plant A, n¼ 13) and 105.33±0.34

copies/L (Plant B, n¼ 11) for RW samples; 102.97±0.65 copies/L (PlantA, n¼ 13) and 102.71±0.63 copies/L (Plant B, n¼ 10) for post-CSsamples, 100.91±0.84 copies/L (Plant A, n¼ 5) for the post-ozonesample; 100.99±0.75 copies/L (Plant A, n¼ 7) for the post-BAC sam-ple; and 101.23±0.42 copies/L (Plant A, n¼ 4) and 102.32±0.84 copies/L(Plant B, n¼ 11) for post-RSF samples. PMMoV copy number in RWsamples were stable and high enough to evaluate stepwise reduc-tion efficiency throughout the year in both plants.

Although the turbidity of the sand filtered water at Plant A wascontrolled to less than 0.14 NTU on all sampling occasions, PMMoVwas nevertheless detected in 61.5% (8/13) of the filtrated samples,indicating that the virus particles were not completely removeddespite strict turbidity control. Therefore, the ability to guaranteethe elimination of viruses in tap water via simple regulation of

c virus as a molecular control in each inhibition removal or mitigation treatment.

Step 2 Step 3

100 times dil.c DAX-8 DAX-8 with 10 times dil.c

15/17 0/15 0/152/17 15/15 15/1515/17 0/152/98 5/98 10/98

xtraction and RT-qPCR.

Fig. 5. PMMoV concentration of each treatment process in two drinking water treatment plants approximately over a year.BAC: biological activated carbon; CS: coagulationesedimentation; PMMoV: pepper mild mottle virus; RSF: rapid sand filtration; RW: raw water.Unfilled points represent unquantified results.Crosses represent undetected results.

R. Kato et al. / Water Research 132 (2018) 61e7068

turbidity of filtered water is limited in cases where RW has anextremely high virus concentration. Purified water is furtherchlorinated in Japan to achieve a high level of safety against entericviruses.

3.2.5. Reduction efficiency of PMMoV, indicator bacteria, andturbidity

The PMMoV, indicator bacteria, and turbidity reduction effi-ciencies are shown in Fig. 6, whose number of samples used areshown in Table A5, and their average copy numbers are shown inTable A6.

A previous study reported that the reduction efficiencies ofPMMoV in CS treatment were 0.40± 0.17 log10 (rainy season) and1.61 log10 (dry season) at a full-scale DWTP in Bangkok (Asami et al.,2016); these values are much lower than those found in the currentstudy. The difference in reduction efficiency of CS can be explainedby the use of different coagulant types (PACl in this study andaluminum sulfate in the Bangkok study), as well as higher RWturbidity in Bangkok. According to previous studies, PACl more

Fig. 6. Average log10 reduction values of PMMoV, E. coli, total coliforms, and turbidityBAC: biological activated carbon; CS: coagulationesedimentation; PMMoV: pepper mild mError bar represents standard deviation. Number of sample pairs (n) of each plot are show

effectively removed recombinant norovirus virus-like particles andbacteriophage Qb in CS treatment than did aluminum (Shirasakiet al., 2010, 2014). Moreover, the treatment plants in our studyused chlorine before the CS process, leading to a greater virusreduction compared to the study conducted in Bangkok.

A significant log10 reduction was observed during the ozoneprocess for PMMoV, likely due to the strong potential for degra-dation of viral RNA through oxidation. A previous study reportedthe virus reduction efficiency of the ozone process at a full-scaleDWTP in Spain (Albinana-Gimenez et al., 2009), where the LRVsof human adenovirus and JC PyV by ozone were 3.15 and 1.95 log10,respectively. Although the target virus of their study was differentfrom ours, the reduction efficiency obtained in this study was inaccordance with the range of the previous research. The virusreduction of the ozone process can be underestimated when usinga molecular detection method because it detects both intact virusand disinfected viruses when viral RNA is preserved, neglecting theeffect of inactivation. Regardless of this, for the case of ozonetreatment, virus reductions observed by RT-PCR were reported

in CS, ozone, BAC, and RSF processes.ottle virus; RSF: rapid sand filtration; RW: raw water; TC: total coliforms.n in Table A5.

R. Kato et al. / Water Research 132 (2018) 61e70 69

similar to those detected by infectivity assay in a laboratoryexperiment (Shin and Sobsey, 2003). Therefore, the virus reductionby ozone obtained in our study could theoretically underestimatethe actual amount of reduction due to disinfection, but this is notalways the case.

Reduction efficiencies of BAC and RSF were relatively lowcompared to those of CS or ozone. The PMMoV reduction effi-ciencies of RSF observed at a DWTP in Bangkok (Asami et al., 2016)were slightly higher (1.26± 0.25 log10 for rainy season, 0.78 log10for dry season) than those found in our study. This can be explainedby the fact that a greater portion of virus might be caught in smallflocs, which likely remained after the CS process at the DWTP inBangkok, leading to a higher virus reduction during the RSF pro-cess. The overall reductions of PMMoV observed in our study were3.80± 0.77 log10 for Plant A and 2.88± 0.57 log10 for Plant B, mostlikely due to different timings of ozone treatments at the twoplants. Since ozonation is applied to the first RSF filtrate at Plant B(i.e., after the sampling point of this study), further reduction ofviruses is expected.

Comparing the reduction efficiency for PMMoV with that ofother indicators, our results indicated that PMMoV was mosteffectively removed by CS, which is better than the reduction basedon turbidity and indicator bacteria would imply in both plants Aand B. On the other hand, it wasmost difficult to remove PMMoV byRSF processes compared to turbidity and indicator bacteria in bothplants, suggesting that the behavior of the virus as a particle duringtreatment processes may differ from that of particle causingturbidity and indicator bacteria. This highlights the importance ofregulating virus removal not by observing indicator bacteria orturbidity but by observing the virus themselves.

4. Conclusions

⁃ Average PMMoV log10 reduction for CS and RSF treatments weresimilar to other viral indicators, including the reference entericvirus, NoVGII, analyzed. PMMoV can therefore be a usefultreatment process indicator of NoV GII removal in CS and RSF.

⁃ The indigenous PMMoV copy number in river water was highenough to evaluate the stepwise reduction efficiency of drinkingwater treatment processes (mean± SD: 105.05±0.48 copies/L[Plant A, n¼ 13] and 105.03±0.34 copies/L [Plant B, n¼ 11]).

⁃ RNA extraction/RT-qPCR efficiency was evaluated using CGMMVas a process control. For samples with less than 10% efficiency,CGMMV quantification and thus RNA extraction/RT-qPCR effi-ciency was improved by either (1) diluting the sample prior toRT or (2) DAX-8 treatment during RNA extraction.

⁃ Based upon PMMoV reductions, CS and ozone treatmentscontributed the most to virus reduction at full-scale DWTPs,with reductions of 2.38± 0.74 log10 (Plant A, n¼ 13) and2.63± 0.76 log10 (Plant B, n¼ 10) for CS, and 1.91± 1.18 log10(Plant A, n¼ 5) for ozone.

Funding sources

This research was supported by the Tokyo metropolitan water-works and the Health Labor Sciences Research Grant (H28-Kenki-Ippan-005) by the Ministry of Health, Labour and Welfare, Japan.

Declaration of interest

Conflicts of interest: none.

Acknowledgments

We thank Dr. Uehara-Ichiki at the Genetic Resources Center andDr. Omura at the National Agriculture and Food Research Organi-zation, Japan, for kindly providing PMMoV and N. benthamiana.

Appendix A. Supplementary data

Supplementary data related to this article can be found athttps://doi.org/10.1016/j.watres.2017.12.068.

References

Abbaszadegan, M., Mayer, B.K., Ryu, H., Nwachuku, N., 2007. Efficacy of removal ofCCL viruses under enhanced coagulation conditions. Environ. Sci. Technol. 41(3), 971e977.

Abbaszadegan, M., Monteiro, P., Nwachuku, N., Alum, A., Ryu, H., 2008. Removal ofadenovirus, calicivirus, and bacteriophages by conventional drinking watertreatment. J. Environ. Sci. Health - Part A 43 (2), 171e177.

Ahmed, W., Harwood, V.J., Gyawali, P., Sidhu, J.P.S., Toze, S., 2015. Comparison ofconcentration methods for quantitative detection of sewage-associated viralmarkers in environmental waters. Appl. Environ. Microbiol. 81 (6), 2042e2049.

Albinana-Gimenez, N., Miagostovich, M.P., Calqua, B., Huguet, J.M., Matia, L.,Girones, R., 2009. Analysis of adenoviruses and polyomaviruses quantified byqPCR as indicators of water quality in source and drinking-water treatmentplants. Water Res. 43 (7), 2011e2019.

Asami, T., Katayama, H., Torrey, J.R., Visvanathan, C., Furumai, H., 2016. Evaluation ofvirus removal efficiency of coagulation-sedimentation and rapid sand filtrationprocesses in a drinking water treatment plant in Bangkok, Thailand. Water Res.101, 84e94.

Bae, J., Schwab, K.J., 2008. Evaluation of murine norovirus, feline calicivirus,poliovirus, and MS2 as surrogates for human norovirus in a model of viralpersistence in surface water and groundwater. Appl. Environ. Microbiol. 74 (2),477e484.

Boudaud, N., Machinal, C., David, F., Freval-Le Bourdonnec, A., Jossent, J., Bakanga, F.,Arnal, C., Jaffrezic, M.P., Oberti, S., Gantzer, C., 2012. Removal of MS2, Qbeta andGA bacteriophages during drinking water treatment at pilot scale. Water Res. 46(8), 2651e2664.

Furuse, K., Sakurai, T., Hirashima, A., Katsuki, M., Ando, A., Watanabe, I., 1978.Distribution of ribonucleic acid coliphages in south and east Asia. Appl. Environ.Microbiol. 35, 995e1002.

Gibson, K.E., Schwab, K.J., Spencer, S.K., Borchardt, M.A., 2012. Measuring andmitigating inhibition during quantitative real time PCR analysis of viral nucleicacid extracts from large-volume environmental water samples. Water Res. 46(13), 4281e4291.

Hamza, I.A., Jurzik, L., Uberla, K., Wilhelm, M., 2011. Evaluation of pepper mildmottle virus, human picobirnavirus and Torque teno virus as indicators of fecalcontamination in river water. Water Res. 45 (3), 1358e1368.

Haramoto, E., Katayama, H., Ohgaki, S., 2004. Detection of noroviruses in tap waterin Japan by means of a new method for concentrating enteric viruses in largevolumes of freshwater. Appl. Environ. Microbiol. 70 (4), 2154e2160.

Haramoto, E., Kitajima, M., Kishida, N., Konno, Y., Katayama, H., Asami, M., Akiba, M.,2013. Occurrence of pepper mild mottle virus in drinking water sources inJapan. Appl. Environ. Microbiol. 79 (23), 7413e7418.

Hata, A., Katayama, H., Furumai, H., 2015a. Organic substances interfere withreverse transcription-quantitative PCR-based virus detection in water samples.Appl. Environ. Microbiol. 81 (5), 1585e1593.

Hata, A., Katayama, H., Kitajima, M., Visvanathan, C., Nol, C., Furumai, H., 2011.Validation of internal controls for extraction and amplification of nucleic acidsfrom enteric viruses in water samples. Appl. Environ. Microbiol. 77 (13),4336e4343.

Hata, A., Kitajima, M., Katayama, H., 2013. Occurrence and reduction of human vi-ruses, F-specific RNA coliphage genogroups and microbial indicators at a full-scale wastewater treatment plant in Japan. J. Appl. Microbiol. 114 (2), 545e554.

Hata, A., Matsumori, K., Kitajima, M., Katayama, H., 2015b. Concentration of entericviruses in large volumes of water using a cartridge-type mixed cellulose estermembrane. Food Environ. Virol. 7 (1), 7e13.

Katayama, H., Shimasaki, A., Ohgaki, S., 2002. Development of a virus concentrationmethod and its application to detection of enterovirus and norwalk virus fromcoastal seawater. Appl. Environ. Microbiol. 68 (3), 1033e1039.

Katayama, H. and Vinj�e, J. 2017. Norovirus and other Calicivirus. In: J.B. Rose and B.Jim�enez-Cisneros, (eds) Global Water Pathogens Project. http://www.waterpathogens.org (J.S. Meschke, and R. Girones (eds) Part 3 Viruses) http://www.waterpathogens.org/book/adenoviruses Michigan State University, E. Lansing,MI, UNESCO.

Kitajima, M., Iker, B.C., Pepper, I.L., Gerba, C.P., 2014a. Relative abundance andtreatment reduction of viruses during wastewater treatment processes -identification of potential viral indicators. Sci. Total Environ. 488, 290e296.

Kitajima, M., Iker, B.C., Rachmadi, A.T., Haramoto, E., Gerba, C.P., 2014b. Quantifi-cation and genetic analysis of salivirus/klassevirus in wastewater in Arizona,USA. Food Environ. Virol. 6 (3), 213e216.

R. Kato et al. / Water Research 132 (2018) 61e7070

Kittigul, L., Panjangampatthana, A., Pombubpa, K., Taweekate, Y., Pungchitton, S.,Diraphat, P., Siripanichgon, K., 2012. Detection and genetic characterization ofnorovirus in environmental water samples in Thailand. Southeast Asian J. Trop.Med. Publ. Health 43 (2), 323e332.

Kuroda, K., Nakada, N., Hanamoto, S., Inaba, M., Katayama, H., Do, A.T., Nga, T.T.V.,Oguma, K., Hayashi, T., Takizawa, S., 2015. Pepper mild mottle virus as an in-dicator and a tracer of fecal pollution in water environments: comparativeevaluation with wastewater-tracer pharmaceuticals in Hanoi, Vietnam. Sci.Total Environ. 506, 287e298.

Mackenzie, W.R., Hoxie, N.J., Proctor, M.E., Gradus, M.S., Blair, K.A., Peterson, D.E.,Kazmierczak, J.J., Addiss, D.G., Fox, K.R., Rose, J.B., Davis, J.P., 1994. A massiveoutbreak in milwaukee of Cryptosporidium infection transmitted through thepublic water-supply. N. Engl. J. Med. 331 (3), 161e167.

Masago, Y., Katayama, H., Watanabe, T., Haramoto, E., Hashimoto, A., Omura, T.,Hirata, T., Ohgaki, S., 2006. Quantitative risk assessment of noroviruses indrinking water based on qualitative data in Japan. Environ. Sci. Technol. 40 (23),7428e7433.

Maunula, L., Miettinen, I.T., von Bonsdorff, C.H., 2005. Norovirus outbreaks fromdrinking water. Emerg. Infect. Dis. 11 (11), 1716e1721.

Medema, G., Loret, J., Stenstr€om, T., Ashbolt, N., 2006. MICRORISK: QuantitativeMicrobial Risk Assessment in the Water Safety Plan.

Parshionikar, S.U., Willian-True, S., Fout, G.S., Robbins, D.E., Seys, S.A., Cassady, J.D.,Harris, R., 2003. Waterborne outbreak of gastroenteritis associated with anorovirus. Appl. Environ. Microbiol. 69 (9), 5263e5268.

Rachmadi, A.T., Kitajima, M., Pepper, I.L., Gerba, C.P., 2016. Enteric and indicatorvirus removal by surface flow wetlands. Sci. Total Environ. 542, 976e982.

Rosario, K., Symonds, E.M., Sinigalliano, C., Stewart, J., Breitbart, M., 2009. Peppermild mottle virus as an indicator of fecal pollution. Appl. Environ. Microbiol. 75,7261e7267. https://doi.org/10.1128/AEM.00410-09.

Schijven, J.F., Teunis, P.F.M., Rutjes, S.A., Bouwknegt, M., de Roda Husman, A.M.,2011. QMRAspot: a tool for Quantitative Microbial Risk Assessment from surfacewater to potable water. Water Res. 45, 5564e5576.

Schmitz, B.W., Kitajima, M., Campillo, M.E., Gerba, C.P., Pepper, I.L., 2016. Virusreduction during advanced bardenpho and conventional wastewater treatmentprocesses. Environ. Sci. Technol. 50, 9524e9532.

Schrader, C., Schielke, A., Ellerbroek, L., Johne, R., 2012. PCR inhibitors - occurrence,properties and removal. J. Appl. Microbiol. 113 (5), 1014e1026.

Schriewer, A., Wehlmann, A., Wuertz, S., 2011. Improving qPCR efficiency in envi-ronmental samples by selective removal of humic acids with DAX-8.J. Microbiol. Meth. 85 (1), 16e21.

Shin, G.A., Sobsey, M.D., 2003. Reduction of Norwalk virus, poliovirus 1, and

bacteriophage MS2 by ozone disinfection of water. Appl. Environ. Microbiol. 69(7), 3975e3978.

Shin, G.A., Sobsey, M.D., 2015. Removal of norovirus from water by coagulation,flocculation and sedimentation processes. Water Sci. Technol. Water Supply 15(1), 158e163.

Shirasaki, N., Matsushita, T., Matsui, Y., Oshiba, A., Marubayashi, T., Sato, S., 2014.Improved virus removal by high-basicity polyaluminum coagulants comparedto commercially available aluminum-based coagulants. Water Res. 48,375e386.

Shirasaki, N., Matsushita, T., Matsui, Y., Oshiba, A., Ohno, K., 2010. Estimation ofnorovirus removal performance in a coagulation-rapid sand filtration processby using recombinant norovirus VLPs. Water Res. 44 (5), 1307e1316.

Shirasaki, N., Matsushita, T., Matsui, Y., Murai, K., 2017. Assessment of the efficacy ofmembrane filtration processes to remove human enteric viruses and the suit-ability of bacteriophages and a plant virus as surrogates for those viruses. WaterRes. 115, 29e39. https://doi.org/10.1016/j.watres.2017.02.054.

Smeets, P.W.M.H., Medema, G.J., Dijk, J.C.v, 2009. The Dutch Secret:how to ProvideSafe Drinking Water without Chlorine in the Netherlands. CopernicusPublications.

Symonds, E.M., Cook, M.M., McQuaig, S.M., Ulrich, R.M., Schenck, R.O., Lukasik, J.O.,Van Vleet, E.S., Breitbart, M., 2015. Reduction of nutrients, microbes, and per-sonal care products in domestic wastewater by a benchtop electrocoagulationunit. Sci. Rep. 5, 9380. https://doi.org/10.1038/srep09380.

Symonds, E.M., Sinigalliano, C., Gidley, M., Ahmed, W., McQuaig-Ulrich, S.M.,Breitbart, M., 2016. Faecal pollution along the southeastern coast of Florida andinsight into the use of pepper mild mottle virus as an indicator. J. Appl.Microbiol. 121, 1469e1481. https://doi.org/10.1111/jam.13252.

Symonds, E.M., Verbyla, M.E., Lukasik, J.O., Kafle, R.C., Breitbart, M., Mihelcic, J.R.,2014. A case study of enteric virus removal and insights into the associated riskof water reuse for two wastewater treatment pond systems in Bolivia. WaterRes. 65, 257e270. https://doi.org/10.1016/j.watres.2014.07.032.

Teunis, P.F.M., Medema, G.J., Kruidenier, L., Havelaar, A.H., 1997. Assessment of therisk of infection by Cryptosporidium or Giardia in drinking water from a surfacewater source. Water Res. 31, 1333e1346.

Thurman, E.M., Malcolm, R.L., 1981. Preparative isolation of aquatic humic sub-stances. Environ. Sci. Technol. 15 (4), 463e466.

WHO, 2011. Guidelines for Drinking-water Quality, fourth ed. World Health Orga-nisation, Geneve, Switzerland, ISBN 978 92 4 154815 1.

Zhang, T., Breitbart, M., Lee, W.H., Run, J.-Q., Wei, C.L., Soh, S.W.L., Hibberd, M.L.,Liu, E.T., Rohwer, F., Ruan, Y., 2006. RNA viral community in human feces:prevalence of plant pathogenic viruses. PLoS Biol. 4 (1), e3.

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具