Td

Ta

b

ARRAA

KUVSMHNF

1

ovzcslaia

tabaat

0d

Journal of Virological Methods 161 (2009) 199–204

Contents lists available at ScienceDirect

Journal of Virological Methods

journa l homepage: www.e lsev ier .com/ locate / jv i romet

he effect of pre-treatment and sonication of centrifugal ultrafiltrationevices on virus recovery

.H. Jones a,∗, J. Brassard b, M.W. Johns a, M.-J. Gagné b

Agriculture and Agri-Food Canada, Lacombe Research Centre, 6000 C & E Trail, Lacombe, Alberta, Canada T4L 1W1Agriculture and Agri-Food Canada, Food Research and Development Centre, 3600 Casavant Blvd. West, St-Hyacinthe, Québec, Canada J2S 8E3

rticle history:eceived 12 December 2008eceived in revised form 8 June 2009ccepted 15 June 2009vailable online 23 June 2009

eywords:ltrafiltrationirus concentrationonication

a b s t r a c t

Enteric viruses such as norovirus (NV) and hepatitis A (HAV) are responsible for a large proportion offood and water-borne illnesses. Most human pathogenic enteric viruses cannot be cultured so theymust be detected by molecular techniques. Male specific (F+) RNA coliphages, a potential surrogatefor human enteric viruses, can be detected by culture and molecular assays. Numbers of viruses andF-RNA coliphages in contaminated food or water may be too low for direct detection. Ultrafiltrationis a general concentration method for all virus types but there is little information on the recoveryefficiency of F-RNA coliphages and enteric viruses. The recovery of F-RNA coliphage MS2 was only25% by plaque assay in initial trials. The objective was to optimize the recovery of concentrated MS2from Microsep 100K ultrafiltration devices. The mean recovery of MS2 increased significantly to 85%

S2epatitis Aoroviruseline calicivirus

by plaque assay and 65% by real-time RT-PCR when ultrafiltration devices were treated with 1% BSAbefore concentration and then ultrasonicated after concentration. The method was validated with MS2,HAV, NV and feline calicivirus (FCV) in water and spinach eluate. The recovery of MS2, HAV and NVwas significantly higher from concentrates obtained from water with treated devices than untreateddevices but not significantly different for FCV or from spinach eluate. To our knowledge, this is the firststudy to use ultrasonication as a post-treatment step to increase recovery of viruses from ultrafiltration

devices.

. Introduction

Enteric and hepatic viruses are responsible for a large proportionf food and water-borne illnesses. Most human pathogenic entericiruses such as Norovirus (NV) and hepatitis A (HAV) and potentialoonotic enteric viruses related closely to human strains, such asalicivirus, hepatitis E virus and rotavirus (RV), cannot be culturedo they must be detected by molecular techniques. However, theack of a direct relationship between the detection of viral nucleiccid and the potential infectivity of the virus particle has importantmplications, particularly when microbial inactivation treatmentsre employed (Sobsey et al., 1998).

Male specific (F+) RNA coliphages are a normal component ofhe mammalian gut flora and have been proposed as an appropri-te model or surrogate for the behaviour of human enteric viruses

ecause they are of similar size, have similar survival characteristicsnd can be readily, rapidly, and cultured economically (Havelaar etl., 1984; Dawson et al., 2005). In addition, broadly reactive quanti-ative real-time RT-PCR detection methods for molecular detection

∗ Corresponding author. Tel.: +1 403 782 8100; fax: +1 403 782 6120.E-mail address: Tineke.Jones@agr.gc.ca (T.H. Jones).

166-0934/$ – see front matter. Crown Copyright © 2009 Published by Elsevier B.V. All rioi:10.1016/j.jviromet.2009.06.013

Crown Copyright © 2009 Published by Elsevier B.V. All rights reserved.

and quantification of environmental F-RNA strains have been devel-oped (Jones et al., 2009; Wolf et al., 2008).

The levels of F-RNA phage detected in feces of swine, broilerchickens, sheep and calves are variable, ranging from 2.0 to6.0 log pfu/g (Havelaar et al., 1986) while the maximum numbers ofpresumptive F-RNA phages recovered from fresh poultry and porkor ground beef approach 2.4 or 0.3 log pfu/g, respectively (Kennedyet al., 1986). The mean numbers of F-RNA coliphages recovered fromchicken carcasses at various stages of the dressing process are low,ranging from 0.8 to −0.8 log pfu/g (Hsu et al., 2002).

Although the numbers of pathogenic viruses in contaminatedwater or most foods are generally too low for direct detection, inges-tion of even a single infectious virus particle may be sufficient tocause disease (Gerba, 2006). While 1 ml or larger volumes can beused to detect a single infectious particle with a culture based assay,the maximum volume of extracted nucleic acid used in a molecu-lar assay to detect one or more genome copies is only 2.5–5 �l.Therefore, concentration is an important and critical step for the

detection and quantification of viruses in food and water samplesby cultivation assays and even more so by molecular techniques.The major strategies for concentrating food and waterborne virusesare adsorption/elution, immunoseparation, ultracentrifugation, orentrapment, which encompasses ultrafiltration and PEG precipi-

ghts reserved.

2 logica

tirUotoRfi2cRf2ifaouaiwctfa

2

2

capsS(th

cpT(wTi(

2

Cp(wOtBtrbdt

00 T.H. Jones et al. / Journal of Viro

ation (Wyn-Jones and Sellwood, 2001). Each of the approachess based on different properties of virus particles and has beeneviewed by Croci et al. (2008) and Wyn-Jones and Sellwood (2001).ltrafiltration and PEG precipitation methods are used for primaryr secondary concentration with the advantage that they can cap-ure a wide range of viruses because the methods are not dependentn the composition of the viral coat. The recovery of HAV, NV andV from berries was higher for eluates concentrated with an ultra-ltration centrifugal device than by PEG precipitation (Butot et al.,007). There is little information on the recovery efficiency of F-RNAoliphages using ultrafiltration. The recovery efficiency of MS2, a F-NA coliphage prototype, using an ultrafiltration centrifugal device

or secondary concentration ranged from 62% to 100% (Hill et al.,007) but in our hands the recovery of MS2 was about 25% in an

nitial trial using an ultrafiltration centrifugal device. The manu-acturer instructions suggest that yields can increase when devicesre pre-treated by various means. The objective of the study was toptimize the recovery of concentrated F-RNA coliphage MS2 fromltrafiltration devices by exploring the effects of pre-treatments ondsorption and post-treatments that effectively remove F-RNA col-phage MS2 from ultrafiltration membranes. The optimized method

as then validated for concentrating MS2, HAV, NV and feline cali-ivirus (FCV) from water and spinach eluates. FCV was included inhe validation study as it can be used as a sample process controlor the detection of food and waterborne RNA viruses (Mattison etl., 2009).

. Materials and methods

.1. Virus stocks

F-RNA coliphage MS2 (ATCC 15597-B1) obtained from Ameri-an Type Culture Collection (Manassas, VA, USA) was propagatednd stored according to ISO 10705-1 protocol (Anon., 1995). A sus-ension of 1 × 104 plaque forming units (pfu)/ml of MS2 in peptonealine was used for recovery experiments, unless noted otherwise.erial 10-fold dilutions of MS2 suspensions prepared in peptone0.1%) saline (0.85%) were enumerated by plaque assay accordingo standard procedures (Anon., 1995) using S. enterica WG49 as theost strain. Each dilution was plated out in triplicate.

FCV strain F9 (ATCC VR-782), kindly provided by A. Houde (Agri-ulture and Agri-Food Canada, St. Hyacinthe, QC, Canada), wasropagated in CRFK cells as described by Mattison et al. (2009).he cytopathic HAV strain HM-175, kindly provided by S. BidawidBureau of Microbial Hazards, Health Canada, Ottawa, ON, Canada),as grown in FRhK-4 cells as described by Mbithi et al. (1991, 1992).

he NV genogroup II clinical specimen SH-29 was collected in 2003n Toronto and kindly provided by the Central Public Health LabToronto, Ontario).

.2. Preparation of ultrafiltration devices

Microsep 100K (Pall Life Sciences, VWR, Edmonton, Alberta,anada) devices were treated by incubation for 1 h at room tem-erature with 3.5 ml of filter sterilised 1% bovine serum albuminBSA; EMD Chemicals, VWR) in phosphate buffered saline (PBS) orith 5% polyethylene glycol 8000 (PEG; Sigma–Aldrich Canada Ltd.,akville, Ontario, Canada) in distilled water, or for 24 h at room

emperature with 3.5 ml of filter sterilised 10% glycerol (Amershamiosciences; GE Healthcare, Piscataway, NJ, USA). After incuba-

ion with a pre-treatment solution, the ultrafiltration devices wereinsed and subsequently washed twice with 3 ml sterile watery centrifugation at 7500 × g for 20 min. Untreated ultrafiltrationevices were washed twice with 3 ml sterile water by centrifuga-ion at 7500 × g for 20 min. Untreated and unwashed ultrafiltration

l Methods 161 (2009) 199–204

devices were used as controls. A 2.5 ml aliquot of MS2 suspen-sion was concentrated in each ultrafiltration device to a finalvolume of ≤80 �l by centrifugation at 2500 × g for 30 min. After theconcentrate was removed, some devices were subjected to a post-treatment to remove adhering MS2 particles by rinsing membraneswith 60 �l of peptone saline or by sonicating ultrafiltration deviceswith 60 �l of peptone saline in an ultrasonic cleaning bath (VWRSignature Ultrasonic cleaner B1500A-DTH, VWR) for 4 min on thehighest power setting (50 W at 42 kHz). For enumeration of MS2by plaque assay, recovered concentrates and rinse solutions werecombined and reconstituted with saline peptone to a total volumeof 2.5 ml.

2.3. Preparation of spinach eluate

Batches of spinach eluate were prepared by combining 100 g ofspinach with 250 ml of 2.9% tryptose phosphate 6% glycine broth,pH 9.5 in a stomacher bag equipped with a strainer bag. The mixturewas shaken by hand and then agitated on a platform for 1 h at roomtemperature. The eluate was removed and the pH was adjusted to7.0 with 1 M HCl. Aliquots of spinach eluate and reverse osmosiswater were artificially inoculated with MS2, FCV, HAV or NV andconcentrated with an ultrafiltration device.

2.4. Taqman real time RT-PCR assays

For real time RT-PCR assays, recovered concentrates and rinsesolutions were combined and reconstituted with saline peptoneto a final volume of 140 �l before viral RNA was extracted witha QIAamp® Viral RNA Mini kit (Qiagen Inc., Missisauga, Ontario,Canada). The nucleic acid sequences for primers and probes usedfor the real-time RT-PCR assays are summarized in Table 1. For MS2,real time RT-PCR was performed using the LV1 assay described inJones et al. (2009) and was carried out in a 25 �l total volume with aStratagene MX3005P QPCR thermocycler (Stratagene, La Jolla, Cal-ifornia, USA) using the Quantitect® Multiplex no ROX RT-PCR kit(Qiagen). Each 25 �l reaction contained 200 nM of forward primerLV1, reverse primer LV1 and LV1 probe, 0.25 �l of Quantitect Multi-plex RT mix, 0.03 �M of ROX reference dye and 2.5 �l of nucleic acidin 1 × Quantitect Multiplex RT-PCR Master Mix. Thermal cyclingconditions were one cycle at 50 ◦C for 20 min, then one cycle at 95 ◦Cfor 15 min followed by 45 cycles of 94 ◦C for 45 s and 60 ◦C for 75 s.Real time RT-PCR assays for FCV, HAV and NV were carried out withthe 1-step Brilliant® II QRT-PCR Core Reagent Kit (Brilliant® II kit,Stratagene). For FCV, each 25 �l reaction contained 800 �M of dNTPmix, 0.625 U of Stratascript reverse transcriptase, 1.25 U of SureStartTaq DNA polymerase in 1 × core RT-PCR buffer, 0.03 �M of ROX ref-erence dye and 300 nM FCV forward primer, 300 nM FCV reverseprimer, 200 nM FCV TaqMan probe and 5 mM MgCl2, as describedby Mattison et al. (2009). Thermal cycling conditions were one cycleat 45 ◦C for 30 min, then one cycle at 95 ◦C for 10 min followed by45 cycles of 95 ◦C for 15 s and 60 ◦C for 1 min. DNA standard curveswere constructed from serial 10-fold dilutions of 105–1 genomeequivalents of purified DNA plasmids that contained the appro-priate cDNA fragment in a 5 ng/ml salmon sperm DNA solution(Gibco BRL, Invitrogen Canada Inc., Burlington, Ontario, Canada).The real-time RT-PCR assays for HAV and Norovirus were carriedout as described by Houde et al. (2006, 2007).

2.5. Data analysis and statistics

The % recovery for viruses was obtained by dividing the num-ber of pfu or genome copies per milliliter recovered from thereconstituted concentrate by the respective number per milliliterin the suspension before concentration and then multiplying thevalue by 100. The Anderson–Darling test for normal distribution

T.H. Jones et al. / Journal of Virological Methods 161 (2009) 199–204 201

Table 1Primer and probe nucleotide sequences.

Virus Nucleotide Sequence (5′–3′) Reference

MS2 LV1 forward CCAGCATCCGTAGCCTTATTGG Jones et al. (2009)LV1 reverse GTTGCTTGTTCAGCGAACTTCTT Jones et al. (2009)LV1 probe TAAGGCGCTGCATCCTGCAACTTGTGC Jones et al. (2009)

FCV SH-FCV3-Q-A GACACCTCCGACGAGTTATGC Mattison et al. (2009)SH-FCV3-Q-1 CCGGGTGGGACTGAGTGG Mattison et al. (2009)SH-FCV3-P CGCCTTACGGATATGAGCAGCCACATTAAC Mattison et al. (2009)

HAV SH-Poly A GARTTTACTCAGTGTTCAATGAATGT Houde et al. (2007)SH-Poly Q TCTCCAAAACGCTTTTAGAAAGAGTCC Houde et al. (2007)SH-Poly 1 AATTTTCCTGCAGCTATGCC Houde et al. (2007)

CNATGTTYAGRTGGATGAG Kageyama et al. (2003)CATCTTCATTCACA Kageyama et al. (2003)GCGATCGCAATCT Kageyama et al. (2003)

wteMP

3

vobbmd(wn(wruMuf

fipuSafirw

Table 3The effect of post-treatment on the recovery of MS2 by plaque assay from pre-treatedultrafiltration devices.

Pre-treatment Post-treatment Recovery (%)a

Meanb SDc

Wash Rinse 30.3 A 9.9Sonication 60.4 B 7.3

1% BSA Rinse 68.3 B 20.4Sonication 84.5 C 9.5

NV COG2F CARGARBCOG2R TCGACGCProbe RING2-TP TGGGAGG

as applied to each data set for each treatment and values forhe means were separated by the Tukey test of the general lin-ar model procedure or compared with the Student’s T test withinitab version 14 statistical software (Minitab Inc., State College,

A, USA).

. Results

In initial trials of concentrating MS2 with Microsep 100K to smallolumes for the purpose of molecular detection by RT-PCR, only 25%f MS2 were recovered from the concentrate and 1% from the filtratey plaque assay. When selected strategies to reduce non-specificinding to the ultrafiltration membranes were used according to theanufacturer suggestions, the recovery of MS2 from Microsep 100K

evices pre-treated with 10% glycerol or 1% BSA was significantlyP < 0.05) higher than from untreated devices or those pre-treatedith 5% PEG while the recovery from washed devices was not sig-

ificantly (P < 0.05) different from untreated or pre-treated devicesTable 2). Ultrafiltration with Microsep 100K devices pre-treatedith 10% glycerol was inefficient as ≥500 �l of the concentrate

emained after centrifugation. Despite washing or pre-treating theltrafiltration devices with 1% BSA, the highest mean recovery ofS2 achieved was less than 40%. Therefore, post-treatments after

ltrafiltration were explored as a means to improve the yield of MS2rom ultrafiltration devices.

The mean recovery of MS2 increased to 68% when the ultra-ltration devices pre-treated with 1% BSA were rinsed witheptone saline after ultrafiltration but the mean recovery remainednchanged when the devices were pre-treated by washing (Table 3).onicating ultrafiltration devices in an ultrasonic cleaning bath

fter ultrafiltration increased the recovery of MS2 from the ultra-ltration devices pre-treated by washing, however, the highestecovery of MS2 was obtained when a pre-treatment with 1% BSAas combined with sonication. When this treatment combina-

Table 2The effect of pre-treatment of ultrafiltration devices on the recovery ofMS2 by plaque assay.

Treatment Recovery (%)a

Meanb SDc

Untreated 25.6 A 7.6Washed 35.0 A,B 10.110% glycerol 37.4 B 6.05% PEG 25.3 A 4.41% BSA 39.4 B 9.8

a N = 3, analyzed in triplicate.b Values followed by the same letter are not significantly different at

P = 0.05.c SD = standard deviation.

6

a N = 9, analyzed in triplicate.b Values followed by the same letter are not significantly different at P = 0.05.c SD = standard deviation.

tion was tested with samples that contained approximately 100-or 1000-fold lower numbers of MS2, the mean recovery was notsignificantly different from solutions containing higher numbers(Table 4). For solutions with an initial concentration of 8 pfu/ml,MS2 were recovered from all 9 ultrafiltration devices but the recov-ery was more variable, ranging from 40% to 92% with a meanrecovery of 71%.

Approximately 65% and 44% of genomic copies of MS2 wererecovered from the concentrated solutions that contained 3.8 × 105

or 3.8 × 103 genome copies/ml, respectively, before concentrationbut not in any of the 9 concentrates from solutions that containedlower levels (Table 5). MS2 RNA was recovered at low levels from5 out of 6 filtrates from solutions that contained 3.8 × 105 genomecopies/ml before concentration but not from any of the filtratesfrom solutions that contained lower levels of MS2.

The % of pfu of MS2 recovered from water was significantlyhigher when treated ultrafiltration devices were used but therewas no significant difference between the % pfu of MS2 recov-ered from spinach eluate using treated or untreated ultrafiltration

devices (Table 6). The % of pfu of MS2 recovered from concen-trated spinach eluate was 72% or higher but the % pfu recoveredfrom concentrated water was 22% with untreated ultrafiltrationdevices. The % of genome copies (GC) of MS2 recovered from the

Table 4The effect of initial concentration on the recovery of MS2 by plaque assay fromsonicated ultrafiltration devices pre-treated with 1% BSA.

Initial concentration (pfu/ml) Recovery (%)a

Meanb SDc

900 84.5 A 9.567 81.5 A 14.8

8 71.4 A 32.6

a N = 9, analyzed in triplicate.b Values followed by the same letter are not significantly different at P = 0.05.c SD = standard deviation.

202 T.H. Jones et al. / Journal of Virological Methods 161 (2009) 199–204

Table 5The recovery of genome copies of MS2 by realtime RT-PCR from sonicated ultrafiltration devices pre-treated with 1% BSA.

Before concentration (GCa/ml) Concentrateb Filtratec

Expected GC/RT-PCR reactiond % recovered % recovered

Mean SDe Range Mean SD Range

3.8 × 105 2.9 × 104 64.5 22.1 36.4–96.9 1.4 1.4 0–3.73.8 × 103 2.9 × 102 43.8 22.6 3.3–72.2 ndf

3.8 × 102 2.9 × 101 nd nd

a GC = genome copies.b N = 9, analyzed in duplicate.c N = 6, analyzed in duplicate.d Based on 2.5 �l extracted RNA per RT-PCR reaction.e SD = standard deviation.f nd = not detected.

Table 6The recovery of plaque forming units (pfu) or genome copies (GC) of MS2 or feline calicivirus (FCV) from water or spinach eluates after concentration with untreated ortreated ultrafiltration devices.

Virus Assay Matrix Input (ml) Expected / RT-PCR reactiona % recovered

Untreated Treatedb P

Meanc SDd Mean SD

MS2 pfu Water 1.3 × 103 22.4 22.4 67.8 22.5 0.001Spinach 6.4 × 104 88.8 20.1 87.1 11.4 0.859

8.2 × 103 72.3 11.1 75.6 11.6 0.5549.0 × 102 75.8 5.6 80.2 6.8 0.154

GC Water 1.2 × 105 6.7 × 102 20.0 17.8 49.0 10.0 0.0013.3 × 103 1.9 × 101 27.2 11.3 56.1 23.2 0.007

Spinach 5.2 × 104 3.1 × 102 26.6 12.1 36.4 12.7 0.1239.4 × 103 5.5 × 101 37.4 14.0 28.2 4.3 0.094

FCV GC Water 7.7 × 105 4.5 × 103 15.7 4.4 16.8 8.6 0.7331.6 × 105 9.1 × 102 19.7 7.1 26.2 8.4 0.097

Spinach 5.2 × 106 3.0 × 104 30.3 7.6 30.7 4.7 0.9132.8 × 105 1.7 × 103 13.5 4.1 14.0 7.1 0.833

n.

setfi2ufltTN

4

thufitdHtpacfe

Table 7The effect of treating ultrafiltration devices on real time RT-PCR detection of virusesextracted from water or spinach eluate.

Virus Matrix n Log GCa detected

Untreated Treatedb P

Mean SDc Mean SD

MS2 Water 9 3.2 0.4 3.8 0.1 0.0049 1.9 0.2 2.2d 0.3 0.029

Spinach 9 3.1 0.2 3.3 0.2 0.1509 2.6d 0.1 2.4 0.1 0.040

FCV Water 9 4.1 0.1 4.1 0.2 0.9599 3.5 0.2 3.6 0.1 0.070

Spinach 9 5.2 0.1 5.2 0.1 0.7879 3.6 0.1 3.6 0.2 0.971

HAV Water 9 1.7 0.4 2.3d 0.4 0.0043 0.7 0.2 1.5 0.1 0.0053 0.2 0.1 0.4 0.3 0.408

Spinach 3 1.9 0.1 2.0 0.1 0.6583 1.3 0.1 1.1 0.0 0.185

NV Water 9 2.5 0.2 2.9 0.2 0.0009 1.8 0.2 1.9 d 0.5 0.430

Spinach 3 4.0 0.1 4.1 0.0 0.1143 3.4 0.1 3.5 0.0 0.205

a Based on 2.5 �l extracted RNA per RT-PCR reaction.b Ultrafiltration devices pre-treated with 1% BSA and sonicated after concentratioc N = 9, analyzed in duplicate.d SD = standard deviation.

pinach eluate was lower than the % of pfu but the % of GC recov-red from concentrated water was similar to the % of pfu. Whilehe mean % of GC of MS2 recovered from water with treated ultra-ltration devices was 49%, the mean % recovered for FCV was only6% at similar initial genome concentrations. The effect of treatingltrafiltration devices on the recovery of MS2, FCV, HAV and NV

rom water and spinach eluate is illustrated in Table 7. The meanog GC recovered of MS2, HAV, and NV in water was higher withreated devices than with untreated devices but not for spinach.he difference was as much as 0.8 log GC for HAV and 0.4 log GC forV.

. Discussion

Research efforts have focused on optimizing primary concen-ration methods specific to the virus of interest but little attentionas been paid to optimizing virus recovery with a centrifugalltrafiltration device for secondary concentration of viruses. Ultra-ltration is a general concentration method that can be applied

o all virus types but the efficiency of virus recovery varies forifferent viruses. The recoveries, determined by cultivation, ofAV and poliovirus were 100% and 15%, respectively, without pre-

reatment of the ultrafiltration device but increased to 100% for

oliovirus when membranes were pre-treated with beef extractnd washed with PBS after ultrafiltration (Divizia et al., 1989). Inontrast, Haramoto et al. (2004) reported a mean recovery of 74%or poliovirus when using an untreated ultrafiltration device. Butott al. (2007) suggested that the recovery of HAV, NV and RV from

a GC = genome copies.b Ultrafiltration devices pre-treated with 1% BSA and sonicated after concentra-

tion.c SD = standard deviation.d Data not normally distributed.

logica

btape7oit2gdu(

ainoicpe(fvhfiidcipaatuwwwtmpwcwcttmtuPeepogc

vbnaui

T.H. Jones et al. / Journal of Viro

erries, determined by real time RT-PCR, increased when ultrafil-ration devices were rinsed with elution buffer after ultrafiltration,nd although the recovery by ultrafiltration was higher than by PEGrecipitation, mean yields were ≤14%. A single report on the recov-ry efficiency of F-RNA coliphages concentrated with a Centricon0-Plus ultrafiltration device demonstrated that an average of 82%f MS2 were recovered when the concentrator was used accord-

ng to manufacturer’s directions (Hill et al., 2007). In contrast, inhis study, recovery of MS2 suspended in saline peptone was about5% in an initial trial using a Microsep 100K ultrafiltration centrifu-al device according to manufacturer directions. It is not known ififferences are due to differences in processing volumes or man-facturer or if additional processing steps were used by Hill et al.2007).

The findings in this study suggested that MS2 remained associ-ted with the membrane of the Microsep 100K device. Membranesn ultrafiltration devices are specifically constructed to minimizeon-specific binding, however, it is recognized that the recoveryf dilute protein solutions may not be quantitative. Manufacturer

nstructions suggest and studies have shown that increased yieldsan be obtained by preventing virus adsorption to membranes byre-treatment with beef extract or other protein solutions (Diviziat al., 1989; Soule et al., 2000; Winona et al., 2001) or by elutingWinona et al., 2001) or rinsing (Butot et al., 2007) bound virusesrom the membrane. A study on the recovery of MS2 from largeolumes of water samples concentrated with ultrafiltration usingollow fiber filters showed that recovery of MS2 increased whenlters were backflushed (Polaczyk et al., 2008). Although backflush-

ng is not feasible for a centrifugal ultrafiltration device, this studyemonstrated that sonication is an effective method for mechani-al detachment of F-RNA coliphage from ultrafiltration membranesn concentration devices. Significant higher numbers of bacterio-hages are recovered from drinking water when ultrasonication ispplied to membrane filters during the elution step (Méndez etl., 2004). To our knowledge, this is the first study to use sonica-ion as a post-treatment step to increase recovery of viruses fromltrafiltration devices. The highest recovery of MS2 was obtainedhen a pre-treatment with 1% BSA of the Microsep 100K deviceas combined with sonication after ultrafiltration and >70% of MS2ere recovered when initial concentration were <10 pfu/ml in pep-

one saline. When MS2 was concentrated to levels detectable byolecular assays, the mean recovery of copies of MS2 RNA from

eptone saline was 44% or higher. Similar results were obtainedhen MS2 was concentrated from water but not when MS2 was

oncentrated from spinach. The yields of HAV and NV from waterere also higher when treated ultrafiltration devices were used for

oncentration. The elution buffer used for spinach contained 2.9%ryptose phosphate and 6% glycine. Perhaps the high concentra-ion of proteinaceous substances in the elution buffer blocked the

embranes during the ultrafiltration process, which may have con-ributed to the higher yields of pfu of MS2 observed when untreatedltrafiltration devices were used to concentrate spinach eluates.re-treating ultrafiltration membranes with 1% glycine or 3% beefxtract significantly improved the recovery of poliovirus (Diviziat al., 1989). The discrepancy observed between the recovery offu and GC for MS2 could be due to the fact that the plaque assaynly detects infectious particles while real time RT-PCR detectsenomic material from both infectious and non-infectious parti-les.

This study showed that the recovery of viruses is variable byirus type and matrix and the standard deviation varies widely

etween data sets. While pre-treatment and post-treatment mayot be necessary to improve the recovery for certain viruses fromn ultrafiltration device, the proposed method is gentle, easy tose, does not negatively affect virus recovery and can potentially

ncrease yields of a range of viruses, particularly when concen-

l Methods 161 (2009) 199–204 203

trating viruses that are suspended in solutions with a low soluteconcentration.

Acknowledgements

The authors thank Ken Li for technical support in processing ofthe samples. This research was supported by Agriculture and Agri-Food Canada Research Branch Peer Reviewed Research Projects 75and 162.

References

Anon., 1995. ISO 10705-1. Water Quality-Detection and Enumeration of Bacterio-phages. Part 1. Enumeration of F-Specific RNA Bacteriophages. InternationalOrganization for Standardization, Geneva, Switzerland.

Butot, S., Putallaz, T., Sánchez, G., 2007. Procedure for rapid concentration and detec-tion of enteric viruses from berries and vegetables. Appl. Environ. Microbiol. 73,186–192.

Croci, L., Dubois, E., Cook, N., de Medici, D., Schultz, A.C., China, B., Rutjes, S.A., Hoorfar,J., van der Poel, W.H.M., 2008. Current methods for extraction and concentra-tion of enteric viruses from fresh fruit and vegetables: towards internationalstandards. Food Anal. Methods 1, 73–84.

Dawson, D.J., Paish, A., Staffel, L.M., Seymour, I.J., Appleton, H., 2005. Survival ofviruses on fresh produce, using MS2 as a surrogate for norovirus. J. Appl. Micro-biol. 98, 203–209.

Divizia, M., Santi, A.L., Panà, A., 1989. Ultrafiltration: an efficient second step forhepatitis A virus and polio concentration. J. Virol. Methods 23, 55–62.

Gerba, C.P., 2006. Food virology: past, present and future. In: Goyal, S.M. (Ed.), Virusesin Foods. Springer, New York, NY, USA, pp. 1–4.

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 fresh water. Appl. Environ. Microbiol. 70, 2154–2160.

Havelaar, A.H., Furuse, K., Hogeboom, W.M., 1986. Bacteriophages and indicator bac-teria in human and animal faeces. J. Appl. Bacteriol. 60, 255–262.

Havelaar, A.H., Hogeboom, W.M., Pot, R., 1984. F-specific RNA bacteriophages insewage: methodology and occurence. Water Sci. Technol. 17, 645–655.

Hill, V.R., Kahler, A.M., Jothikumar, N., Johnson, T.B., Hahn, D., Cromeans, T.L., 2007.Multistate evaluation of an ultrafiltration-based procedure for simultaneousrecovery of enteric microbes in 100-liter tap water samples. Appl. Environ.Microbiol. 73, 4218–4225.

Houde, A., Leblanc, D., Poitras, E., Ward, P., Brassard, J., Simard, C., Trottier, Y.-L., 2006.Comparative evaluation of RT-PCR, nucleic acid sequence-based amplification(NASBA) and real-time RT-PCR for detection of noroviruses in faecal material. JVirol Methods 135, 163–172.

Houde, A., Guévremont, E., Poitras, E., Leblanc, D., Ward, P., Simard, C., Trottier, Y.-L.,2007. Comparative evaluation of new TaqMan real-time assays for the detectionof hepatitis A virus. J. Virol Methods 140, 80–89.

Hsu, F.-C., Carol Shieh, Y.-S., Sobsey, M.D., 2002. Enteric bacteriophages as potentialfecal indicators in ground beef and poultry meat. J. Food Protect. 65, 93–99.

Jones, T.H., Houde, A., Poitras, E., Ward, P., Johns, M.W., 2009. Development andevaluation of a multiplexed real-time TaqMan RT-PCR assay with internal ampli-fication control for detection of F-specific RNA coliphage genogroups I and IV.FAEV 1, 57–65.

Kageyama, T., Kojima, S., Shinohara, M., Uchida, K., Fukushi, S., Hoshino, F.B., Takeda,N., Katayama, K., 2003. Broadly reactive and highly sensitive assay for Norwalk-like viruses based on real-time quantitative reverse transcription-PCR. J. Clin.Microbiol. 41 (4), 1548–1557.

Kennedy Jr., J.E., Wei, C.I., Oblinger, J.L., 1986. Distribution of coliphages in foods. J.Food Protect. 49, 944–951.

Mattison, K., Brassard, J., Gagné, M.-J., Ward, P., Houde, A., Lessard, L., Simard, C.,Shukla, A., Pagotto, F., Jones, T.H., Trottier, Y.-L., 2009. The feline calicivirus as asample process control for the detection of food and waterborne RNA viruses.Int. J. Food Microbiol. 132, 73–77.

Méndez, J., Audicana, A., Isern, A., Llaneza, J., Moreno, B., Tarancón, M.L., Jofre, J.,Lucena, F., 2004. Standardised evaluation of the performance of a simple mem-brane filtration-elution method to concentrate bacteriophages from drinkingwater. J. Virol. Methods 117, 19–25.

Mbithi, J.N., Springthorpe, V.S., Sattar, S.A., 1991. Effect of relative humidity and airtemperature on survival of hepatitis A virus on environmental surfaces. Appl.Environ. Microbiol. 57, 1394–1399.

Mbithi, J.N., Springthorpe, V.S., Boulet, J.R., Sattar, S.A., 1992. Survival of hepatitis A onhuman hands and its transfer on contact with animate and inanimate surfaces.J. Clin. Microbiol. 30, 757–763.

Polaczyk, A.L., Jothikumar, N., Cromeans, T.L., Hahn, D., Roberts, J.M., Amburgey,J.E., Hill, V.R., 2008. Ultrafiltration-based techniques for rapid and simultane-ous concentration of multiple microbe classes from 100-L tap water samples. J.

Microbiol. Methods 73, 92–99.

Sobsey, M.D., Battigelli, D.A., Shin, G.A., Newland, S., 1998. RT-PCR amplificationdetects inactivated viruses in water and wastewater. Water Sci. Technol. 38,91–94.

Soule, H., Genoulaz, O., Gratacap-Cavallier, B., Chevallier, P., Liu, J.-X., Seogneurin,J.-M., 2000. Ultrafiltration and reverse transcription-polymerase chain reaction:

2 logica

W

04 T.H. Jones et al. / Journal of Viro

an efficient process for poliovirus, rotavirus and hepatitis A virus detection inwater. Water Res. 34, 1063–1067.

inona, L.J., Ommani, A.W., Olszewski, J., Nuzzo, J.B., Oshima, K.H., 2001. Efficient andpredictable recovery of viruses from water by small scale ultrafiltration systems.Can. J. Microbiol. 47, 1033–1041.

l Methods 161 (2009) 199–204

Wyn-Jones, A.P., Sellwood, J., 2001. Enteric viruses in the aquatic environment.J. Appl. Microbiol. 91, 945–962.

Wolf, S., Hewitt, J., Rivera-Aban, M., Greening, G.E., 2008. Detection and characteriza-tion of F+ RNA bacteriophages in water and shellfish: application of a multiplexreal-time reverse transcription PCR. J. Virol. Methods 149, 123–128.