Improvement of the Bag-Mediated Filtration System for Sampling Wastewater and Wastewater-Impacted Waters
- 494 Downloads
Environmental surveillance of poliovirus (PV) plays an important role in the global program for eradication of wild PV. The bag-mediated filtration system (BMFS) was first developed in 2014 and enhances PV surveillance when compared to the two-phase grab method currently recommended by the World Health Organization (WHO). In this study, the BMFS design was improved and tested for its usability in wastewater and wastewater-impacted surface waters in Nairobi, Kenya. Modifications made to the BMFS included the size, color, and shape of the collection bags, the filter housing used, and the device used to elute the samples from the filters. The modified BMFS concentrated 3–10 L down to 10 mL, which resulted in an effective volume assayed (900–3000 mL) that was 6–20 times greater than the effective volume assayed for samples processed by the WHO algorithm (150 mL). The system developed allows for sampling and in-field virus concentration, followed by transportation of the filter for further analysis with simpler logistics than the current methods. This may ultimately reduce the likelihood of false-negative samples by increasing the effective volume assayed compared to samples processed by the WHO algorithm, making the BMFS a valuable sampling system for wastewater and wastewater-impacted surface waters.
KeywordsEnvironmental surveillance Poliovirus Enterovirus Pathogens BMFS Wastewater
Environmental surveillance is crucial for monitoring and preventing the spread of pathogens, which are present in wastewater, surface waters impacted by wastewater, recreational waters, and drinking water sources (Betancourt et al. 2014; Deboosere et al. 2011; Fuhrimann et al. 2015) that humans and animals regularly contact. Environmental surveillance can highlight epidemiological patterns, especially for diseases with high subclinical infection rates, where pathogens silently circulate. For example, although less than 1 in 100 poliovirus infections results in acute flaccid paralysis, the virus is shed in the stool of all infected persons (Nathanson and Kew 2010). Environmental surveillance of wastewater can detect poliovirus from these asymptomatic carriers, resulting in early pathogen detection, focused vaccine efforts, and eventual eradication certification (Hovi et al. 2012).
As pathogens are typically present in low numbers in wastewater-impacted surface waters, detection can be challenging. The likelihood of detection can be improved through concentration of the environmental samples, which results in an increase in the effective volume assayed. Typically, after grab sample collection, concentration occurs in the laboratory by chemical and/or physical mechanisms. For example, when sampling for poliovirus from wastewater and wastewater-impacted surface waters, the two-phase grab sample method commonly used by the World Health Organization (WHO) involves concentrating 500 mL down to 10 mL by a two-phase separation method using polyethylene glycol (PEG) and dextran (World Health Organization 2015). For physical concentration, vacuum-driven or pump-driven filtration is often used (Gantzer et al. 1997; Karim et al. 2009; Katayama et al. 2002). These types of filtration methods require concentration to be completed in the laboratory, which requires transportation and storage of samples prior to processing. In low-resource settings, such sampling methodology can be expensive and logistically challenging to perform.
On-site filtration could be used to concentrate large sample volumes, resulting in a high effective volume assayed and an increased likelihood of pathogen detection. Filtration at the field site is advantageous because it reduces the need to transport large volumes of potentially biohazardous materials, allows for sampling of larger volumes without requiring transportation or shipment of heavy samples, and reduces the storage footprint in space-limited laboratories. However, in-field filtration typically requires the use of an electric-powered pump (Corsi et al. 2014). This can make filtration in the field prohibitive as pumps are expensive, require a continual power source, and are large and heavy. Portable, low-cost considerations for in-field filtration are needed.
Materials and Methods
The basic BMFS consists of a 6- to 12-L collection bag, filter, collapsible tripod stand, personal protective equipment, and decontamination supplies. All supplies are transportable in an insulated backpack.
Modifications to the bag-mediated filtration system kit
30 × 76 cm (10 L)
A pre-screen mesh on the collection bag inlet prevented coarse sediment and debris from entering the bag
Detachable brass ring
Stainless steel rolled ring clamp
The stainless steel rolled ring clamp held the pre-screen mesh onto the collection bag inlet
The clamp had a side loop for rope attachment
The clamp had a second side angled loop that facilitated easy placement of the filled collection bag on the tripod stand
Sample collection bag
Quick-disconnect check valve port
Located 4 cm above the bag’s bottom
Open ¼-inch barbed tubing adapter port
Located 10 cm above the bag’s bottom
Due to clogging in the first-generation bag, a straight barbed tubing adapter replaced the quick-disconnect check valve
The catchment area below the outlet valve increased from 4 to 10 cm for capture of settled solids. Settleable solids accumulated below the outlet prior to filtration, thus minimizing solids entering and clogging the filter
Open ¼-inch barbed tubing adapter port
Located at the bottom of the bag
Heat-sealed 15.5° slope angles toward the outlet
The lower sediment catchment area of the bag was ineffective as settled solids resuspended during the rolling of the bag when pressure was applied, released, and reapplied. Therefore, the catchment area was removed
A 15.5° slope was heat sealed onto the bottom of the bag. This enabled bleeding of settled solids to a second vessel prior to filtration of the remaining clarified sample
Sample collection bag
30 cm × 76 cm (10 L)
29 cm × 81 cm (10 L)
The bag height increased due to the switch from the detachable brass ring to the stainless steel rolled ring clamp. The top of the bag folds over the metal clamp for attachment, requiring a greater bag height
28 cm × 64 cm (6 L)
28 cm × 94 cm (12 L)
Typically, field technicians processed less than 4-L wastewater or wastewater-impacted surface waters in Nairobi, making collection of a 10-L water sample unnecessary. A 6-L bag enabled bleeding and collection of an initial 1 L settled volume, and filtration of the remaining clarified 5 L
A 12-L bag facilitated use of a new protocol (not discussed in this work). The large bag enabled bleeding and collection of an initial 2 L settled volume, and filtration of the remaining clarified 10 L through two filters
Sample collection bag
Drab green with white screen printing
Volumetric demarcations: 0.5-L increments
Bright yellow bags attracted attention from on-lookers. Therefore, the color was changed to drab green to reduce attention
Volumetric demarcations were added, enabling measurement of the filtered volume
Drab green with white screen printing
Volumetric demarcations: 0.25-L increments
The volumetric demarcations were changed to 0.25-L gradations to facilitate easier and more accurate measurement of the filtered volume
Second-generation collection bags held 10 L, were drab green with white screen printing that demarcated 0.5-L volume increments, and had 1-cm heat-sealed seams (Fig. 1b). Samples were collected at the top, which was held open by a stainless steel rolled ring clamp with a spring form mechanism. The inner and outer rings were attached by a lanyard. The clamp held a pre-screen mesh on the inlet of the bag, and the mesh had a uniform pore size. Several different mesh pore sizes were tested, ranging from 249 µm to 1.2 mm, and a final pore size of 249 µm was selected. The clamp also included a loop for rope attachment and an angled loop for easy placement on the tripod stand. Samples flowed through an open ¼-inch-diameter barbed tubing adapter port located 10 cm above the bag’s bottom. An 8-cm section of tubing with a tubing clamp was attached to the port, which was clamped while the bag was filling.
Third-generation collection bags held 6 or 12 L, were drab green with white screen printing that demarcated 0.25-L volume increments, and had 1-cm heat-sealed seams (Fig. 1c). Sample collection was completed the same as for the second-generation bag with the clamp and pre-screen mesh. The port was moved to the bottom of the bag and a 15.5° slope was heat sealed onto the bottom of the bag to angle sediments toward the outlet port so they could be collected prior to filtration.
Sample Collection and Filtration
Laboratory Testing of Filters
For laboratory testing, 10 L influent wastewater (after bar screens) grab samples were collected from the West Point Wastewater Treatment Plant in Seattle, WA, USA. Samples were collected by bucket and rope, poured into 20-L carboys, and transported back to the laboratory where they were stored at room temperature until processing (within 7 days of collection). Samples were subsequently filtered through 5-cm ViroCap™ filters (Scientific Methods, Inc., Granger, IN, USA) using a peristaltic pump.
Field Testing of BMFS
For field testing, 10 L surface water grab samples were collected from Lake Union in Seattle, WA, USA and from four study sites in Nairobi, Kenya. In Nairobi, samples were collected from the Motoine River (which is impacted by untreated latrine waste) bordering the Kibera informal settlement, a latrine waste stream in the Mathare informal settlement that flows to the Mathare River, and sewer conveyance lines in the Eastleigh neighborhood (two sites). These collection sites will be referred to as Kibera, Mathare, Eastleigh A, and Eastleigh B.
ViroCap Filter Housings
Commercially produced, disposable ViroCap filter housings and reusable ViroCap filter housings designed for this study were used for laboratory tests and field sampling. The reusable housings aimed to reduce the system’s environmental footprint and reduce the holdup/elution volume from 175 to 100 mL. The holdup volume is defined as the non-solid void within the filter housing, a measurement used in calculating the concentration factor of the system. The reusable housing sumps were custom-designed, injection-molded, made from clear polycarbonate (Protolabs, Maple Plain, MN, USA), and used in conjunction with a reinforced polypropylene lid (Pentek, Inc, Upper Saddle River, NJ, USA). The filter is seated in the body of the housing and the lid is tightened with the use of a customized wrench to secure the filter in place. An O-ring sits at the interface between the housing and lid to provide a water-tight, low-pressure seal. The threaded lid of the housing allows for an expandable seal to accommodate a wide range of filter heights, from 49.784 to 51.816 mm (1.96–2.04 inches). An elastomer O-ring seal prevented leakage. Polypropylene barbed tubing adapters (1/4 NPT × 0.9525 cm inner diameter) connect the housing and collection bag. All housing components can withstand autoclaving (121 °C, 124 kPa, 30 min) for at least five cycles with no loss of sealing integrity (data not shown). They can also be disinfected by soaking for 30 min in a Virkon™ (DuPont, Wilmington, DE, USA) or 0.5% HOCl (bleach) solution.
Elution and Secondary Concentration
ViroCap filters were eluted in a 1.5% beef extract and 0.05 M glycine solution, at pH 9.5. The eluent was poured into the filter housing inlet, let stand for a specified period of time, and then pumped out with a peristaltic pump. The eluate was adjusted to pH 7.0–7.5 with 5 M HCl and 5 M NaOH. Disposable housings were eluted with 175 mL and the eluent was held in the filter housing for 30 min. Reusable housings were eluted by two methods: (1) a single elution in which they were eluted with 100 mL and the eluent was held in the filter housing for 30 min, and (2) a double elution in which they were eluted two times with 100 mL and the eluent was held in the filter housing for 15 min for each elution.
Secondary concentration was performed by PEG/NaCl precipitation. PEG 8000 (14 g/100 mL) and NaCl (1.17 g/100 mL) were added to the sample and dissolved by vigorous shaking (5 min). Samples were shaken overnight (200 RPM, 4 °C) and then centrifuged (6000×g, 4 °C, 30 min). The supernatant was discarded and the pellet was resuspended in 10 mL of phosphate-buffered saline.
Water Sample Characteristics
Concentrations of total dissolved solids (TDS) were determined for the water samples collected at the Kibera, Mathare, Eastleigh A, and Eastleigh B sites. TDS were determined by Kenya Standard Method 459 (Gazette 1985).
Filter Clogging Challenge
To determine particle sizes that resulted in filter clogging in the BMFS, 10 L of tap water seeded with 25 g soil particles was filtered through the BMFS in the laboratory. Particle sizes included <850, <180, <150, <63, and <25 µm. Filters were also tested for clogging by filtering 10 L of surface water and influent wastewater samples from Seattle, WA, USA. Field testing was conducted at the Kibera, Mathare, Eastleigh A, and Eastleigh B sites in Nairobi.
Poliovirus Culture and Recovery Experiments
The vaccine strain of poliovirus type 1 (PV1) stocks were prepared by confluent lysis of buffalo green monkey kidney (BGMK) cell monolayers (Sobsey and Jones 1979). PV1 was extracted using Vertrel XF (E. I. du Pont de Nemours and Company, Wilmington, DE, USA) and purified stocks were stored at −80 °C (Mendez et al. 2000).
To determine the poliovirus recovery efficiency using the modified BMFS, laboratory-seeded experiments were completed. Influent wastewater samples were seeded with ~10 PFU/mL PV1 prior to filtration through ViroCap filters in either disposable or reusable housings. Samples were then eluted using a single elution (175 mL) with the disposable housing, and either a single elution (100 mL) or a double elution (100 mL each) with the reusable housing. Samples were then concentrated to 10 mL by PEG/NaCl and analyzed for PV1 by plaque assay. Viruses were enumerated on 95% confluent BGMK cells as previously described and modified to include an Avicel RC-581 (FMC Corporation, Philadelphia, PA, USA) overlay (Sobsey and Jones 1979; Matrosovich et al. 2006). Assays were performed in triplicate onto 9.5 cm2 wells using 200-µL aliquots of relevant dilutions. Infected cells were incubated at 37 °C and 5% CO2 for 40–44 h and then stained with 2% crystal violet in 20% methanol. Plaques were counted for infectious virus enumeration. Viral recovery was calculated by dividing the recovered viral count by the seeded viral count.
Unpaired Student’s or Welch’s t tests were used to compare PV recoveries between filtration through a disposable versus reusable filter housing, while using a single or double elution.
Results and Discussion
Modifications of the BMFS Design
Collection Bag Design
For the first generation of the BMFS, laboratory experiments suggested that collection and processing up to 10 L of primary or secondary wastewater samples was feasible when using the ViroCap filter (Fagnant et al. 2014). During field evaluations, however, high amounts of sediments clogged the system after filtration of 2–3 L of wastewater and wastewater-impacted surface waters from several sites in Nairobi. The clogging occurred in either the ViroCap filter (Kibera site) or collection bag (Eastleigh B site). It was thought that the clogging of the filter was due to fine suspended particles in the water samples. Clogging of the collection bag likely occurred inside of the outlet check valve due to an accumulation of celluloid material observed in the water, which prevented liquid from exiting the bag.
Water constituents for sampling sites in Nairobi
Total dissolved solids (mg/L)
Typical medium strength wastewatera
The third generation of the BMFS collection bag was modified based on these new observations (Table 1; Fig. 1c). This new design enabled a 15-min clarification/settling period, followed by bleeding of the settled solids to a second vessel prior to filtration of the remaining clarified sample. Initial settling occurred while the user performed other tasks; typically, 15 min passed while the bag hung on the tripod stand. After settling, 0.5–1.0 L of settled sediments were collected in a Whirl–Pak® bag (Nasco, Fort Atkinson, WI, USA), and the clarified sample was filtered until the filter clogged. If the full clarified volume was filtered, the collected sediments were poured back into the sample collection bag for filtration.
Disposable ViroCap filter housings were used during the initial development of the BMFS. The filter design was modified based on observations from field tests that indicated slow filtration. Reusable housings were designed to reduce the system’s environmental footprint and reduce the holdup volume from 175 to 100 mL.
Reduction of the holdup/elution volume from 175 mL with the disposable housing to 100 mL with the reusable housing improves the likelihood of pathogen detection. This is due to an increase in the concentration factor from 1:57 for the disposable housing to 1:100 for the reusable housing when a 10-L sample is filtered. Therefore, if the eluate is assayed directly, the effective volume assayed and subsequent likelihood of detection is increased by 75% for the reusable housing over the disposable housing. A smaller eluate volume also reduces the processing time in the laboratory required for secondary concentration. Secondary concentration may be desirable to increase the effective volume assayed for viruses expected to be present in low concentration or for assays requiring low sample volumes.
The addition of the extension tube in the filter housing further decreased the dead volume in the reusable housing (Fig. 6b). During elution, the liquid is extracted from the bottom of the filter housing to the outlet, reducing the amount of eluate left in the housing after the elution procedure from 20 to 7 mL.
Elution Device Design
In the initial work, a peristaltic pump was used for elution of the samples from the ViroCap filter. While functional, the pump did not enclose the system, which was a safety concern as there was the potential for aerosolization of biologicals from the sample and cross-contamination due to the system not being fully enclosed (Walls et al. 2014). Also, peristaltic pumps are expensive, not easily transferable internationally due to electrical socket and voltage requirements, and complicated to fix with internal and electronic components. Therefore, an elution apparatus was developed to replace the peristaltic pump and for deployment of the self-contained unit to international collaborators. This apparatus does not use electrical power, has a compact footprint (0.093 m2), and is compatible with autoclave systems and biosafety cabinets available at the international collaborators’ facilities. In combination with the modified filter housing, the elution device reduced the potential for cross-contamination and increased technician safety, while maintaining similar elution times. A low-cost ($20–$40) manual bilge pump created a vacuum, which exhibited low resistance and did not require significant force to perform the elution procedure. This Alfa Marine pump is not specific for the elution apparatus and so pumps from different manufacturers are compatible with the system, making replacement simple if needed. Use of the system was intuitive and required minimal training (typically less than 30 min). In addition, the elution device encloses the system outlet to reduce the possibility of cross-contamination in the laboratory. The elution device collected the sample into an enclosed sample collection cup; in combination with the optimized filter housing, the possibility of escape of pathogen-containing aerosols from the system was removed (Fig. 3d). If a small amount of bubbles or foam is present in the collection after the elution procedure, the user can wait less than 5 min for the foam to settle in the enclosed container before replacing the ported lid of the elution apparatus with the solid lid for storage and transport of the sample. All components that come in contact with the samples are autoclavable. Finally, the active personnel time required for a single elution using this pump (13 min, n = 6) was similar to that using the peristaltic pump (13 min, n = 6).
Virus Recovery Efficiency and Concentration Factor of the BMFS
In addition to modifying the BMFS design to improve usability and performance, the sensitivity of the BMFS was also determined. PV1 recovery measured 39.6 ± 9.6% (n = 18) for BMFS samples eluted from ViroCap filters in disposable housings using a single elution (eluate volume 175 mL). With the switch to the reusable housings, the PV1 recovery was 9.7 ± 2.0% (n = 21) using a single elution (eluate volume 100 mL). This recovery efficiency was significantly lower than the PV1 recovery with the disposable housing (p = 3.9 × 10−6, Student’s t test). Due to the PV1 recovery decrease and a decrease in the eluate pH, the elution protocol was modified. The new double elution method included two elution steps, each with a hold time of 15 min rather than 30 min and an eluent volume of 100 mL. Using this double elution (total eluate volume 200 mL) and the reusable housings, the PV1 recovery measured 33.1 ± 8.8% (n = 11). The PV1 recovery efficiency from the reusable housing using a double elution was not statistically different than the recovery efficiency from the disposable housing using a single elution (p = 0.33, Student’s t test).
Concentration factors and effective volume plated achieved by use of the BMFS ViroCap reusable filter housing (RH) and disposable filter housing (DH) with two volumes filtered (10 and 3 L) compared to a 0.5-L two-phase grab sample with 10 mL final concentrate volume
Volume filtered (vf, L)
Volume after primary concentration (v1, mL)
Volume after secondary concentration (v2, mL)
Volume assayed (va, mL)
Primary concentration factor (c1 = vf/v1)
Secondary concentration factor (c2 = vf/v2)
Effective volume assayed (ve = va × c2, mL)
BMFS with RH
BMFS with DH
Two-phase grab sampled
This compact and transportable field sampling system for enteric viruses requires no external power source, either in the field or in the laboratory. Transitioning from laboratory samples to field samples presented new challenges with water matrix effects, making field testing vital during kit development for troubleshooting these practical challenges to maximize concentrated volumes. Ultimately, use of the BMFS routinely achieved a concentration factor five times greater than that of the currently recommended 0.5-L grab samples processed by the WHO algorithm. With multiple product design iterations, the BMFS now allows for larger filtered volumes in the field, eases sample processing in the laboratory, and may ultimately reduce the likelihood of false-negative samples compared to samples processed by the WHO algorithm, making it a valuable sampling system for surveillance of poliovirus and other enteroviruses in wastewater and wastewater-impacted surface waters.
This research was supported by funding from the Paul G. Allen Family Foundation (Grant Number NPT.1938-603689) and by management from the Bill and Melinda Gates Foundation. We would like to thank Graciela Matrajt for her technical writing assistance and Phuong Truong and the technicians at West Point Treatment Plant (King County Metro, Wastewater Treatment Division) for their assistance in collecting wastewater.
The study was funded by the Paul G. Allen Family Foundation (Grant Number NPT.1938-603689) and was managed by the Bill and Melinda Gates Foundation.
Compliance with Ethical Standards
Conflict of interest
The authors declare that they have no conflict of interest.
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This article does not contain any studies with animals performed by any of the authors.
Informed consent was obtained from all individual participants included in this study.
- Betancourt, W. Q., Duarte, D. C., Vasquez, R. C., & Gurian, P. L. (2014). Cryptosporidium and giardia in tropical recreational marine waters contaminated with domestic sewage: Estimation of bathing-associated disease risks. Marine Pollution Bulletin, 85(1), 268–273. doi:10.1016/j.marpolbul.2014.05.059.CrossRefPubMedGoogle Scholar
- Corsi, S. R., Borchardt, M. A., Spencer, S. K., Hughes, P. E., & Baldwin, A. K. (2014). Human and bovine viruses in the Milwaukee River watershed: Hydrologically relevant representation and relations with environmental variables. Science of the Total Environment, 490, 849–860. doi:10.1016/j.scitotenv.2014.05.072.CrossRefPubMedGoogle Scholar
- Deboosere, N., Horm, S. V., Pinon, A., Gachet, J., Coldefy, C., Buchy, P., et al. (2011). Development and validation of a concentration method for the detection of Influenza A viruses from large volumes of surface water. Applied and Environmental Microbiology, 77(11), 3802–3808. doi:10.1128/aem.02484-10.CrossRefPubMedPubMedCentralGoogle Scholar
- Fuhrimann, S., Stalder, M., Winkler, M. S., Niwagaba, C. B., Babu, M., Masaba, G., et al. (2015). Microbial and chemical contamination of water, sediment and soil in the Nakivubo wetland area in Kampala, Uganda. Environmental Monitoring and Assessment. doi:10.1007/s10661-015-4689-x.PubMedPubMedCentralGoogle Scholar
- Gazette, T. K. (1985). Kenya standard specification for drinking water - part 2: Physical methods of test for the quality of drinking water. KS 05-459.Google Scholar
- Karim, M. R., Rhodes, E. R., Brinkman, N., Wymer, L., & Fout, G. S. (2009). New electropositive filter for concentrating enteroviruses and noroviruses from large volumes of water. Applied and Environmental Microbiology, 75(8), 2393–2399. doi:10.1128/aem.00922-08.CrossRefPubMedPubMedCentralGoogle Scholar
- Katayama, H., Shimasaki, A., & Ohgaki, S. (2002). Development of a virus concentration method and its application to detection of enterovirus and Norwalk virus from coastal seawater. Applied and Environmental Microbiology, 68(3), 1033–1039. doi:10.1128/aem.68.3.1033-1039.2002.CrossRefPubMedPubMedCentralGoogle Scholar
- Tchobanoglous, G., Burton, F. L., Stensel, H. D., & Eddy, M. (2003). Wastewater engineering, treatment and reuse (4th ed.). Boston: McGraw-Hill.Google Scholar
- World Health Organization. (2015). Guidelines on environmental surveillance for detection of polioviruses, working draft. http://polioeradication.org/wp-content/uploads/2016/07/GPLN_GuidelinesES_April2015.pdf. Retrieved from May 2016.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.