APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Oct. 1979, p. 688-693 0099-2240/79/10-0688/06$02.00/0
Vol. 38, No. 4
Procedure for the Recovery of Airborne Human Enteric Viruses During Spray Irrigation of Treated Wastewater BARBARA E. MOORE,* BERNARD P. SAGIK, AND CHARLES A. SORBER Center for Applied Research and Technology, The University of Texas at San Antonio, San Antonio, Texas 78285 Received for publication 13 April 1979
Because of the relatively low number of indigenous enteric viruses recovered from secondary wastewater effluents, their presence in air (aerosols) as a result of wastewater spray irrigation requires extensive sampling. Methodology to allow the recovery of indigenous enteroviruses from aerosols generated at an operational wastewater irrigation site was tested under both laboratory and field conditions. In recent years, both Federal legislation and economic considerations have revitalized interest in land disposal of treated domestic wastewaters. The practice of spray irrigation of such wastewaters inevitably leads to questions regarding the potential health risks associated with aerosolization of indigenous bacteria and viruses. Although some published data are available on the occurrence and viability of bacterial aerosols at spray irrigation sites (6, 7, 13, 14), little analogous information has been obtained for indigenous enteroviruses. Methodology has been developed and successfully used to detect indigenous coliphage aerosolized during secondary wastewater treatment (3) and seeded f2 bacteriophage aerosolized at a wastewater spray irrigation site (H. T. Bausum, S. A. Schaub, and C. A. Sorber, Abstr. Annu. Meeting Am. Soc. Microbiol. 1976, Q16, p. 193). In these studies, both Anderson stacked sieve samplers and large-volume air samplers with recirculation devices were used. Attempts by Fannin and co-workers (2) to recover animal viruses by directly assaying collection fluids from large-volume air samplers (Litton Systems, Inc.) were unsuccessful. Recently, Teltsch and Katzenelson (14) were able to detect human viruses at an effluent irrigation site in Israel. Using a large-volume liquid scrubber (Aerojet-General), echovirus 7 was recovered in 4 of 12 runs at a point 40 m downwind from a single sprinkler. These viruses were isolated by inoculating a portion of each aerosol collection fluid directly into cell cultures to screen for the presence of indigenous animal viruses. Environmental monitoring of indigenous enteroviruses in the United States is hampered by comparatively low virus concentrations, especially in secondarily treated wastewaters. These
initially low concentrations, coupled with the dilution of organisms in air after aerosolization, make quantitative detection of viruses difficult. This premise was substantiated during an intensive aerosol-monitoring program carried out at Pleasanton, Calif., in 1976. Although indigenous coliphage were detected as far as 100 m downwind from the wet line edge, only two aerosol samples yielded confirmed enteroviruses by direct plaque assay of aerosol collection fluid (6). Since coliphage levels in composite effluent samples were 3 to 4 logio units greater than enterovirus levels (as assayed on HeLa cells), it became obvious that in order to detect aerosolized viruses reproducibly, extremely large air volumes would have to be sampled, and additional concentration of aerosol collection fluids would be necessary. During the Pleasanton study, high-volume air samplers manufactured by Litton Systems, Inc., were operated, using brain heart infusion broth (BHI; Difco Laboratories) containing 0.1% Tween 80 as the collection fluid. The greatest obstacles to successful recovery of viruses from aerosol sampler collection fluids were the high concentrations of both particulates (including microorganisms) and soluble organic compounds. Conventional concentration techniques used in the recovery of viruses from surface waters were not applicable. Rather, methodologies directed to the recovery of viruses from cell culture fluids were considered. Parameters considered in the selection of methodology for fluid testing included ease of operation, time required for concentration, volume reduction, virus recovery, and reproducibility of results. The work reported here details the selection and application of viral concentration procedures to allow the recovery of human viruses from aerosol collection fluids. 688
VOL. 38, 1979
PROCEDURE FOR RECOVERING ENTERIC VIRUSES
MATERIALS AND METHODS Virus techniques. Poliovirus 1 (Chat) and coxsackievirus B-3 (Nancy) were used in laboratory studies. Both viruses were grown on HeLa cell monolayers in 150-cm2 flasks. Stock suspensions were prepared by infecting at a multiplicity of infection of approximately 10 plaque-forming units per cell. After a 30-min attachment period, the virus inoculum was removed and maintenance medium, Eagle minimum essential medium (modified) with Hanks basal salt solution containing 2% heat-inactivated calf serum, was added to the infected cells. The flasks were incubated at 37°C in a 5% C02-in-air incubator for either 8 to 10 h (poliovirus) or 20 to 24 h (coxsackievirus). Viruses were harvested from infected cells at these times by three cycles of rapid freezing and thawing. Cellular debris was removed by centrifugation at 12,100 x g for 15 min. Supernatant fluids containing viruses were titrated by plaque assay and held at 4°C until use. Plaque assays for laboratory studies were performed with confluent HeLa cell monolayers. Virus inocula were adsorbed for 45 min at 37°C with periodic rocking of plates. Poliovirus-infected cells were overlaid with Eagle minimum essential medium (modified) containing 1% agar, 8% calf serum, 100 U of penicillin per ml, and 50 ,ug of streptomycin per ml. Coxsackievirusinfected cells were overlaid with the same medium containing only 2% fetal calf serum with the addition of 250 yg of diethylaminoethyl dextran (Pharmacia Fine Chemicals). After incubation at 37°C in 5% CO2 for 2 days, monolayers were stained with 2% neutral red in Hanks balanced salt solution, and plaques were
counted. Concentrated field samples were plaque assayed as described above on both HeLa and Buffalo green monkey monolayers. Since the only positive virus isolates from aerosol samples were made on HeLa cells, only results from this cell line are presented. In other studies conducted by our laboratory this inhouse HeLa cell line consistently has yielded superior recoveries of indigenous viruses identified as polioviruses 1, 2, and 3, coxsackieviruses B-3, B-4, and B-5, and echoviruses 11 and 21 when compared with the Buffalo green monkey line (passages 115 through 150). Inocula were adsorbed for 45 to 60 min, at which time the monolayers were washed with prewarmed phosphate-buffered saline containing 500 U of penicillin per ml and 250 jig of streptomycin per ml. After 15 min, this wash was removed by aspiration, and the monolayers were overlaid with the medium described above containing 8% calf serum and additional antibiotics: 25 ,ug of gentamicin (Schering Laboratories) and 0.5 jig of amphotericin B (E. R. Squibb & Sons) per ml. One-half of the plates were scored for plaques at 3 days postinfection; the remainder were scored at 5 days. Suspected plaques were isolated and subjected to verification as viral isolates by two 5-day passages in homologous cell tube cultures. Confirmed virus isolates were identified by using the Lim-Benyesh-Melnick enterovirus pools (8). Concentration methods. A two-phase aqueous polymer system, based on the mechanism of liquidliquid partitioning, was originally applied to the recov-
689
ery of T2 virus, echovirus 7, and poliovirus from harvested cell culture medium (1). Necessary components obtained from Pharmacia Fine Chemicals were prepared on a weight/weight basis (in sterile distilled water) at the following stock concentrations: sodium dextran sulfate (Na-DS), 20%; dextran T-2000, 20%; and dextran T-500, 25%. Polyethylene glycol (PEG) 6000 (Union Carbide Corp.) was prepared as a 30% (wt/wt) stock solution, and sodium chloride was prepared as a 5 M solution. Appropriate volumes of dextran, PEG, and NaCl were added to the BHI containing 0.1% Tween 80 to achieve final stated concentrations. The mixture was placed in a separatory funnel, shaken vigorously, and held at 4°C to allow phase separation. PEG hydroextraction was used to dewater BHI containing 0.1% Tween 80. Dialysis tubing (2-cm diameter) with an average pore size of 2.4 nm was filled with test volumes, packed in PEG 6000, and held at 4°C for 12 to 24 h. After the concentrated volume was collected, the tubing was rinsed with 3% beef extract in 0.1 M tris(hydroxymethyl)aminomethane buffer at pH 9.0. Both dewatered and rinsed volumes were assayed for viral infectivity. Wastewater sampling. During each aerosol collection study, composite samples of secondarily treated, nondisinfected wastewater entering the spray line were collected and concentrated for indigenous enteroviruses. Two-liter volumes were mixed with expanded bentonite (100 mg/liter) and calcium chloride (final concentration, approximately 0.01 M) at pH 6.0 for 30 min. Suspended solids were collected by lowspeed centrifugation and subsequently eluted by suspension in 20 ml of tryptose phosphate broth (Difco). After 15 min of vigorous intermittent mixing, the solids were recentrifuged at 12,100 x g for 15 min. Supematant volumes were assayed for infectivity on HeLa cells. Minimally, 10% of all plaques counted were passaged by tube culture, with 100% being confirmed as viruses. However, no attempt was made to identify all of these isolates. Control volumes of 0.5 liter of the same wastewater were seeded with poliovirus 1 (Chat) and concentrated as above to allow determination of relative recovery efficiency vis-a-vis poliovirus. Aerosol sampling. High-volume air samplers (Litton model M, Litton Systems, Inc.) were used to collect airborne viruses. The samplers are designed to collect airborne particles and concentrate them into a thin, moving film of liquid. The range of airflow rates of these samplers is 400 to 1,200 liters/min, and collection fluid flow rates may be up to 8 ml/min. Basically, the sampler is an electrostatic precipitator. Air is drawn into the unit through a converging nozzle and passes through the center of a high-voltage plate. It then flows radially between this plate and a lower rotating collection disk. An electric potential of 15,000 V, maintained across a spacing between the plate and disk, creates two effects: a corona is emitted, exposing particles to air ions so that they acquire an electrical charge, and the electrical field provides a driving force, which results in the precipitation of charged particles onto the lower disk. Collection fluid is pumped onto the center of the collection disk and, due to centrifugal force resulting from the spinning disk, forms a thin moving film over the entire disk
690
APPL. ENVIRON. MICROBIOL.
MOORE, SAGIK AND SORBER
surface. Particles collected on the film are transported across the rotating disk to a collection funnel, from where the fluid is punmped back to the collection
TABLE 1. Virus concentration with two-phase aqueous polymer systems
reservoir.
System
Poliovirus recovery
Concn factorb
Aerosol sampler placement and operation. During preliminary aerosol sampling at the Pleasanton site, it became obvious that due to the low levels of enteroviruses in the wastewater being irrigated, the normal sampling procedures would be inadequate to obtain statistically significant numbers of viruses from the air. Normal sampler operation involved collecting approximate 30 m3 of air (1,000 liters/min for 30 min) in 100-ml volumes of collection fluid. Based on the available data, it was estimated that approximately 1,600 m3 of air would be required to detect significant numbers of enteroviruses at 50 m downwind of the spray-wetted edge. To accomplish this goal, eight high-volume samplers were set close to each other at the appropriate downwind distance. The samplers were operated simultaneously for six to eight consecutive 30-min periods. At the end of each of these operating periods, the 100 ml of collection fluid (BHI containing 0.1% Tween 80) from each sampler was pooled in a common sterile container, and fresh sterile collection fluid was provided each sampler for the next operating period. These procedures resulted in sampling of the large quantities of air desired; however, they yielded large volumes of collection fluid, therefore necessitating concentration.
0.2% Na-DSc 132 54 6.5% PEG 6000 0.15 M NaCl 0.2% Na-DSc 153 18 6.5% PEG 6000 0.3 M NaCl 0.2% dextran T-2000d 131 28 6.5% PEG 6000 0.3 M NaCl 0.25% dextran T-500e 84 26 7.0% Peg 6000 0.15 M NaCl 0.25% dextran T-500e 95 23 7.0% PEG 6000 0.3 M NaCl aPercentage of input virus detected in the lower phase. bConcentration factor = original volume (in milliliters)/lower-phase volume (in milliliters). 'After Philipson et al. (9). d After Shuval et al. (12). 'After Grindrod and Cliver (5).
RESULTS
TABLE 2. Recovery of viruses by PEG
(%a
hydroextraction Laboratory experiments. Parallel laboraVirus recovery (%) tory tests to evaluate a series of previously pubConcn faclished two-phase aqueous polymer systems DeVirus Rinse Total tora watered based on dextran and PEG were conducted. In the initial experimental system, poliovirus 1 vol Vol (Chat) was suspended in BHI containing 0.1% Poliovirus 1 100%, directed toward optimizing two-phase polymer with 3- to 25-fold volume reductions. The degree concentration for detection of enteric viruses in of sample reduction, a time-dependent function, aerosol collection fluids. appeared to affect recovery of viral infectivity. An important consideration for field applicaFor poliovirus, observed recoveries decreased as tion of detection methodology is the relative dewatered volumes became smaller. viral stability during shipping. Test viruses were Of the methodologies tested, two-phase poly- inoculated into BHI containing 0.1% Tween 80 mer concentration gave a better volume reducand mixed to facilitate particle dispersion. Voltion while allowing viral recoveries in the range umes of 100 ml were removed, held at 4°C, and of 50%. Although hydroextraction resulted in assayed for infectivity at intervals indicated in comparable recoveries at lower dewatering lev- Table 3. The remaining BHI was concentrated a
PROCEDURE FOR RECOVERING ENTERIC VIRUSES
VOL. 38, 1979
TABLE 3. Viral stability in sampler fluid and concentrated volumes Virus recovery (% of time 0) Time
(h)
BHI with 0.1% Tween 80 CoxPolio- sackievirus virus 1 B-3
Lower phase'
Poliovirus 1
Coxsackievirus B-3
100 100 100 0 100 86 12 100 100 100 86 100 25 100 100 50 72 44 Lower phase resulting from concentration of BHI + 0.1% Tween 80 using Na-DS (0.2%), PEG (6.5%), and NaCl (0.15 M).
by using the previously described Na-DS and PEG system, the lower phase was collected, and possible loss of viral infectivity with time was monitored at 4°C. Both viruses were more stable in the lower-phase test volume through a 3-day holding period (Table 3). Therefore, on-site reduction of aerosol collection fluid volumes was used in the field studies. Such handling also allowed air shipment of a smaller sample volume to the laboratory and limited sharply chances for laboratory virus contamination of the aerosol sample. The time required for completion of phase separation was monitored by collecting and measuring the lower-phase volume separating over a 24-h period at 4°C. Lower-phase separation could be considered 90% complete by 15 h and 95% complete within 20 h for this BHI-Tween 80 test system (Fig. 1). Field studies. The protocol adopted for field concentration of viruses from aerosol sampler fluids was based on a two-phase polymer system composed of 0.2% Na-DS, 6.5% PEG, and 0.15 M sodium chloride. Sterile stock reagents, sterile glassware, and disposable sterile plasticware were provided to field personnel trained in the concentration procedure. Lower-phase volumes were collected 18 to 20 h after addition of specified reagents and shipped at 4°C along with effluent grab samples to the laboratory for analysis. Immediately upon receipt, the concentrated aerosol sample was diluted 1:3 with sterile Hanks balanced salt solution containing penicillin and streptomycin. Then this sample was assayed for enteroviruses as described above in a segregated assay room. After the aerosol samples were assayed, indigenous animal viruses were concentrated from the effluent sample in a separate laboratory work area. Finally, a concentration efficiency was determined for each wastewater sample.
691
Two separate, paired composite aerosol and effluent samples from the Pleasanton, Calif., wastewater irrigation site were collected. Indigenous animal viruses were recovered from effluent samples with concentration efficiencies for the control virus of 79 and 100% for runs I and II, respectively. In addition to being tested for human viruses, wastewater effluent samples and aerosol collection fluids were assayed directly for coliphage by a conventional soft-agar overlay technique with Escherichia coli K13 as host organism. Results for the recovery of viruses from these runs are presented in Table 4. As has been noted, coliphage levels in the secondary effluent were 3 to 4 logio units greater than indigenous enterovirus levels. To detect significant numbers of viruses, large volumes of air were sampled. Run I involved the sampling of 1,440 m3 of air and resulted in a total of 4,716 ml of collection fluid from the ganged high-volume air samplers. During run II, 2,340 m3 of air was sampled, resulting in 7,820 ml of sampler collection fluid. (Such extremely large volumes of aerosol sampler collection fluid illustrate the need for effective enterovirus concentration methods.) Coliphage were detected consistently 50 m downwind perpendicular from the wet line edge. Concentrated volumes of aerosol collection fluid yielded potential viral isolates (as plaque-forming units) in both runs (see Table 4). Tube culture passage confirmed a total of four viruses from run I and seven viruses from run II. Poliovirus 2 was recovered during the first run. Isolates from the second run were identified as poliovirus 1 and coxsackievirus B-3.
DISCUSSION Although both methods tested for viral recovery from BHI containing 0.1% Tween 80 gave 100-t -0 0
/
0
/
'a E w
/ / /
50
/ /
/
// 0
E
5
/ 5
15 10 Time (hours)
20
FIG. 1. Phase separation as a function
25
of time.
692
APPL. ENVIRON. MICROBIOL.
MOORE, SAGIK AND SORBER
TABLE 4. Results of virus isolations from wastewater and aerosol samples Wastewater concn (PFU/liter)a
Organism
Aerosol concn
PFU/ml of collection fluid Run I
Run II
Run I
Run II
4.7 x 105 3.7 x 105 0.12 Coliphageb 0.44 Enteroviruses 3 day 23 260 1.1 x 10-3c
Vol. 38, No. 4
Procedure for the Recovery of Airborne Human Enteric Viruses During Spray Irrigation of Treated Wastewater BARBARA E. MOORE,* BERNARD P. SAGIK, AND CHARLES A. SORBER Center for Applied Research and Technology, The University of Texas at San Antonio, San Antonio, Texas 78285 Received for publication 13 April 1979
Because of the relatively low number of indigenous enteric viruses recovered from secondary wastewater effluents, their presence in air (aerosols) as a result of wastewater spray irrigation requires extensive sampling. Methodology to allow the recovery of indigenous enteroviruses from aerosols generated at an operational wastewater irrigation site was tested under both laboratory and field conditions. In recent years, both Federal legislation and economic considerations have revitalized interest in land disposal of treated domestic wastewaters. The practice of spray irrigation of such wastewaters inevitably leads to questions regarding the potential health risks associated with aerosolization of indigenous bacteria and viruses. Although some published data are available on the occurrence and viability of bacterial aerosols at spray irrigation sites (6, 7, 13, 14), little analogous information has been obtained for indigenous enteroviruses. Methodology has been developed and successfully used to detect indigenous coliphage aerosolized during secondary wastewater treatment (3) and seeded f2 bacteriophage aerosolized at a wastewater spray irrigation site (H. T. Bausum, S. A. Schaub, and C. A. Sorber, Abstr. Annu. Meeting Am. Soc. Microbiol. 1976, Q16, p. 193). In these studies, both Anderson stacked sieve samplers and large-volume air samplers with recirculation devices were used. Attempts by Fannin and co-workers (2) to recover animal viruses by directly assaying collection fluids from large-volume air samplers (Litton Systems, Inc.) were unsuccessful. Recently, Teltsch and Katzenelson (14) were able to detect human viruses at an effluent irrigation site in Israel. Using a large-volume liquid scrubber (Aerojet-General), echovirus 7 was recovered in 4 of 12 runs at a point 40 m downwind from a single sprinkler. These viruses were isolated by inoculating a portion of each aerosol collection fluid directly into cell cultures to screen for the presence of indigenous animal viruses. Environmental monitoring of indigenous enteroviruses in the United States is hampered by comparatively low virus concentrations, especially in secondarily treated wastewaters. These
initially low concentrations, coupled with the dilution of organisms in air after aerosolization, make quantitative detection of viruses difficult. This premise was substantiated during an intensive aerosol-monitoring program carried out at Pleasanton, Calif., in 1976. Although indigenous coliphage were detected as far as 100 m downwind from the wet line edge, only two aerosol samples yielded confirmed enteroviruses by direct plaque assay of aerosol collection fluid (6). Since coliphage levels in composite effluent samples were 3 to 4 logio units greater than enterovirus levels (as assayed on HeLa cells), it became obvious that in order to detect aerosolized viruses reproducibly, extremely large air volumes would have to be sampled, and additional concentration of aerosol collection fluids would be necessary. During the Pleasanton study, high-volume air samplers manufactured by Litton Systems, Inc., were operated, using brain heart infusion broth (BHI; Difco Laboratories) containing 0.1% Tween 80 as the collection fluid. The greatest obstacles to successful recovery of viruses from aerosol sampler collection fluids were the high concentrations of both particulates (including microorganisms) and soluble organic compounds. Conventional concentration techniques used in the recovery of viruses from surface waters were not applicable. Rather, methodologies directed to the recovery of viruses from cell culture fluids were considered. Parameters considered in the selection of methodology for fluid testing included ease of operation, time required for concentration, volume reduction, virus recovery, and reproducibility of results. The work reported here details the selection and application of viral concentration procedures to allow the recovery of human viruses from aerosol collection fluids. 688
VOL. 38, 1979
PROCEDURE FOR RECOVERING ENTERIC VIRUSES
MATERIALS AND METHODS Virus techniques. Poliovirus 1 (Chat) and coxsackievirus B-3 (Nancy) were used in laboratory studies. Both viruses were grown on HeLa cell monolayers in 150-cm2 flasks. Stock suspensions were prepared by infecting at a multiplicity of infection of approximately 10 plaque-forming units per cell. After a 30-min attachment period, the virus inoculum was removed and maintenance medium, Eagle minimum essential medium (modified) with Hanks basal salt solution containing 2% heat-inactivated calf serum, was added to the infected cells. The flasks were incubated at 37°C in a 5% C02-in-air incubator for either 8 to 10 h (poliovirus) or 20 to 24 h (coxsackievirus). Viruses were harvested from infected cells at these times by three cycles of rapid freezing and thawing. Cellular debris was removed by centrifugation at 12,100 x g for 15 min. Supernatant fluids containing viruses were titrated by plaque assay and held at 4°C until use. Plaque assays for laboratory studies were performed with confluent HeLa cell monolayers. Virus inocula were adsorbed for 45 min at 37°C with periodic rocking of plates. Poliovirus-infected cells were overlaid with Eagle minimum essential medium (modified) containing 1% agar, 8% calf serum, 100 U of penicillin per ml, and 50 ,ug of streptomycin per ml. Coxsackievirusinfected cells were overlaid with the same medium containing only 2% fetal calf serum with the addition of 250 yg of diethylaminoethyl dextran (Pharmacia Fine Chemicals). After incubation at 37°C in 5% CO2 for 2 days, monolayers were stained with 2% neutral red in Hanks balanced salt solution, and plaques were
counted. Concentrated field samples were plaque assayed as described above on both HeLa and Buffalo green monkey monolayers. Since the only positive virus isolates from aerosol samples were made on HeLa cells, only results from this cell line are presented. In other studies conducted by our laboratory this inhouse HeLa cell line consistently has yielded superior recoveries of indigenous viruses identified as polioviruses 1, 2, and 3, coxsackieviruses B-3, B-4, and B-5, and echoviruses 11 and 21 when compared with the Buffalo green monkey line (passages 115 through 150). Inocula were adsorbed for 45 to 60 min, at which time the monolayers were washed with prewarmed phosphate-buffered saline containing 500 U of penicillin per ml and 250 jig of streptomycin per ml. After 15 min, this wash was removed by aspiration, and the monolayers were overlaid with the medium described above containing 8% calf serum and additional antibiotics: 25 ,ug of gentamicin (Schering Laboratories) and 0.5 jig of amphotericin B (E. R. Squibb & Sons) per ml. One-half of the plates were scored for plaques at 3 days postinfection; the remainder were scored at 5 days. Suspected plaques were isolated and subjected to verification as viral isolates by two 5-day passages in homologous cell tube cultures. Confirmed virus isolates were identified by using the Lim-Benyesh-Melnick enterovirus pools (8). Concentration methods. A two-phase aqueous polymer system, based on the mechanism of liquidliquid partitioning, was originally applied to the recov-
689
ery of T2 virus, echovirus 7, and poliovirus from harvested cell culture medium (1). Necessary components obtained from Pharmacia Fine Chemicals were prepared on a weight/weight basis (in sterile distilled water) at the following stock concentrations: sodium dextran sulfate (Na-DS), 20%; dextran T-2000, 20%; and dextran T-500, 25%. Polyethylene glycol (PEG) 6000 (Union Carbide Corp.) was prepared as a 30% (wt/wt) stock solution, and sodium chloride was prepared as a 5 M solution. Appropriate volumes of dextran, PEG, and NaCl were added to the BHI containing 0.1% Tween 80 to achieve final stated concentrations. The mixture was placed in a separatory funnel, shaken vigorously, and held at 4°C to allow phase separation. PEG hydroextraction was used to dewater BHI containing 0.1% Tween 80. Dialysis tubing (2-cm diameter) with an average pore size of 2.4 nm was filled with test volumes, packed in PEG 6000, and held at 4°C for 12 to 24 h. After the concentrated volume was collected, the tubing was rinsed with 3% beef extract in 0.1 M tris(hydroxymethyl)aminomethane buffer at pH 9.0. Both dewatered and rinsed volumes were assayed for viral infectivity. Wastewater sampling. During each aerosol collection study, composite samples of secondarily treated, nondisinfected wastewater entering the spray line were collected and concentrated for indigenous enteroviruses. Two-liter volumes were mixed with expanded bentonite (100 mg/liter) and calcium chloride (final concentration, approximately 0.01 M) at pH 6.0 for 30 min. Suspended solids were collected by lowspeed centrifugation and subsequently eluted by suspension in 20 ml of tryptose phosphate broth (Difco). After 15 min of vigorous intermittent mixing, the solids were recentrifuged at 12,100 x g for 15 min. Supematant volumes were assayed for infectivity on HeLa cells. Minimally, 10% of all plaques counted were passaged by tube culture, with 100% being confirmed as viruses. However, no attempt was made to identify all of these isolates. Control volumes of 0.5 liter of the same wastewater were seeded with poliovirus 1 (Chat) and concentrated as above to allow determination of relative recovery efficiency vis-a-vis poliovirus. Aerosol sampling. High-volume air samplers (Litton model M, Litton Systems, Inc.) were used to collect airborne viruses. The samplers are designed to collect airborne particles and concentrate them into a thin, moving film of liquid. The range of airflow rates of these samplers is 400 to 1,200 liters/min, and collection fluid flow rates may be up to 8 ml/min. Basically, the sampler is an electrostatic precipitator. Air is drawn into the unit through a converging nozzle and passes through the center of a high-voltage plate. It then flows radially between this plate and a lower rotating collection disk. An electric potential of 15,000 V, maintained across a spacing between the plate and disk, creates two effects: a corona is emitted, exposing particles to air ions so that they acquire an electrical charge, and the electrical field provides a driving force, which results in the precipitation of charged particles onto the lower disk. Collection fluid is pumped onto the center of the collection disk and, due to centrifugal force resulting from the spinning disk, forms a thin moving film over the entire disk
690
APPL. ENVIRON. MICROBIOL.
MOORE, SAGIK AND SORBER
surface. Particles collected on the film are transported across the rotating disk to a collection funnel, from where the fluid is punmped back to the collection
TABLE 1. Virus concentration with two-phase aqueous polymer systems
reservoir.
System
Poliovirus recovery
Concn factorb
Aerosol sampler placement and operation. During preliminary aerosol sampling at the Pleasanton site, it became obvious that due to the low levels of enteroviruses in the wastewater being irrigated, the normal sampling procedures would be inadequate to obtain statistically significant numbers of viruses from the air. Normal sampler operation involved collecting approximate 30 m3 of air (1,000 liters/min for 30 min) in 100-ml volumes of collection fluid. Based on the available data, it was estimated that approximately 1,600 m3 of air would be required to detect significant numbers of enteroviruses at 50 m downwind of the spray-wetted edge. To accomplish this goal, eight high-volume samplers were set close to each other at the appropriate downwind distance. The samplers were operated simultaneously for six to eight consecutive 30-min periods. At the end of each of these operating periods, the 100 ml of collection fluid (BHI containing 0.1% Tween 80) from each sampler was pooled in a common sterile container, and fresh sterile collection fluid was provided each sampler for the next operating period. These procedures resulted in sampling of the large quantities of air desired; however, they yielded large volumes of collection fluid, therefore necessitating concentration.
0.2% Na-DSc 132 54 6.5% PEG 6000 0.15 M NaCl 0.2% Na-DSc 153 18 6.5% PEG 6000 0.3 M NaCl 0.2% dextran T-2000d 131 28 6.5% PEG 6000 0.3 M NaCl 0.25% dextran T-500e 84 26 7.0% Peg 6000 0.15 M NaCl 0.25% dextran T-500e 95 23 7.0% PEG 6000 0.3 M NaCl aPercentage of input virus detected in the lower phase. bConcentration factor = original volume (in milliliters)/lower-phase volume (in milliliters). 'After Philipson et al. (9). d After Shuval et al. (12). 'After Grindrod and Cliver (5).
RESULTS
TABLE 2. Recovery of viruses by PEG
(%a
hydroextraction Laboratory experiments. Parallel laboraVirus recovery (%) tory tests to evaluate a series of previously pubConcn faclished two-phase aqueous polymer systems DeVirus Rinse Total tora watered based on dextran and PEG were conducted. In the initial experimental system, poliovirus 1 vol Vol (Chat) was suspended in BHI containing 0.1% Poliovirus 1 100%, directed toward optimizing two-phase polymer with 3- to 25-fold volume reductions. The degree concentration for detection of enteric viruses in of sample reduction, a time-dependent function, aerosol collection fluids. appeared to affect recovery of viral infectivity. An important consideration for field applicaFor poliovirus, observed recoveries decreased as tion of detection methodology is the relative dewatered volumes became smaller. viral stability during shipping. Test viruses were Of the methodologies tested, two-phase poly- inoculated into BHI containing 0.1% Tween 80 mer concentration gave a better volume reducand mixed to facilitate particle dispersion. Voltion while allowing viral recoveries in the range umes of 100 ml were removed, held at 4°C, and of 50%. Although hydroextraction resulted in assayed for infectivity at intervals indicated in comparable recoveries at lower dewatering lev- Table 3. The remaining BHI was concentrated a
PROCEDURE FOR RECOVERING ENTERIC VIRUSES
VOL. 38, 1979
TABLE 3. Viral stability in sampler fluid and concentrated volumes Virus recovery (% of time 0) Time
(h)
BHI with 0.1% Tween 80 CoxPolio- sackievirus virus 1 B-3
Lower phase'
Poliovirus 1
Coxsackievirus B-3
100 100 100 0 100 86 12 100 100 100 86 100 25 100 100 50 72 44 Lower phase resulting from concentration of BHI + 0.1% Tween 80 using Na-DS (0.2%), PEG (6.5%), and NaCl (0.15 M).
by using the previously described Na-DS and PEG system, the lower phase was collected, and possible loss of viral infectivity with time was monitored at 4°C. Both viruses were more stable in the lower-phase test volume through a 3-day holding period (Table 3). Therefore, on-site reduction of aerosol collection fluid volumes was used in the field studies. Such handling also allowed air shipment of a smaller sample volume to the laboratory and limited sharply chances for laboratory virus contamination of the aerosol sample. The time required for completion of phase separation was monitored by collecting and measuring the lower-phase volume separating over a 24-h period at 4°C. Lower-phase separation could be considered 90% complete by 15 h and 95% complete within 20 h for this BHI-Tween 80 test system (Fig. 1). Field studies. The protocol adopted for field concentration of viruses from aerosol sampler fluids was based on a two-phase polymer system composed of 0.2% Na-DS, 6.5% PEG, and 0.15 M sodium chloride. Sterile stock reagents, sterile glassware, and disposable sterile plasticware were provided to field personnel trained in the concentration procedure. Lower-phase volumes were collected 18 to 20 h after addition of specified reagents and shipped at 4°C along with effluent grab samples to the laboratory for analysis. Immediately upon receipt, the concentrated aerosol sample was diluted 1:3 with sterile Hanks balanced salt solution containing penicillin and streptomycin. Then this sample was assayed for enteroviruses as described above in a segregated assay room. After the aerosol samples were assayed, indigenous animal viruses were concentrated from the effluent sample in a separate laboratory work area. Finally, a concentration efficiency was determined for each wastewater sample.
691
Two separate, paired composite aerosol and effluent samples from the Pleasanton, Calif., wastewater irrigation site were collected. Indigenous animal viruses were recovered from effluent samples with concentration efficiencies for the control virus of 79 and 100% for runs I and II, respectively. In addition to being tested for human viruses, wastewater effluent samples and aerosol collection fluids were assayed directly for coliphage by a conventional soft-agar overlay technique with Escherichia coli K13 as host organism. Results for the recovery of viruses from these runs are presented in Table 4. As has been noted, coliphage levels in the secondary effluent were 3 to 4 logio units greater than indigenous enterovirus levels. To detect significant numbers of viruses, large volumes of air were sampled. Run I involved the sampling of 1,440 m3 of air and resulted in a total of 4,716 ml of collection fluid from the ganged high-volume air samplers. During run II, 2,340 m3 of air was sampled, resulting in 7,820 ml of sampler collection fluid. (Such extremely large volumes of aerosol sampler collection fluid illustrate the need for effective enterovirus concentration methods.) Coliphage were detected consistently 50 m downwind perpendicular from the wet line edge. Concentrated volumes of aerosol collection fluid yielded potential viral isolates (as plaque-forming units) in both runs (see Table 4). Tube culture passage confirmed a total of four viruses from run I and seven viruses from run II. Poliovirus 2 was recovered during the first run. Isolates from the second run were identified as poliovirus 1 and coxsackievirus B-3.
DISCUSSION Although both methods tested for viral recovery from BHI containing 0.1% Tween 80 gave 100-t -0 0
/
0
/
'a E w
/ / /
50
/ /
/
// 0
E
5
/ 5
15 10 Time (hours)
20
FIG. 1. Phase separation as a function
25
of time.
692
APPL. ENVIRON. MICROBIOL.
MOORE, SAGIK AND SORBER
TABLE 4. Results of virus isolations from wastewater and aerosol samples Wastewater concn (PFU/liter)a
Organism
Aerosol concn
PFU/ml of collection fluid Run I
Run II
Run I
Run II
4.7 x 105 3.7 x 105 0.12 Coliphageb 0.44 Enteroviruses 3 day 23 260 1.1 x 10-3c
