United States Environmental Protection Agency Municipal Environmental Research, Laboratory Cincinnati OH 45268 Research and Development EPA-600/S2-81-014 Apr. 1981 Project Summary Firefly Luciferase ATP Assay Development for Monitoring Bacterial Concentrations in Water Supplies Grace L Picciolo, EmmettW. Chappell, Jody W. Deming, Richard R.Thomas, D. A. Nibley, and Harold Okrend This research program was initiated to develop a rapid, automatable system for measuring total viable microorga- nisms in potable drinking water supplies using the firefly luciferase adenosine triphosphate (ATP) assay. The ATP assay was adapted to an' automatable flow system that, in less than 2 minutes, provided assays with sensitivity comparable with establish- ed methodology (105 bacteria/mL). Quality controls for required reagents were established. To achieve the sensi- tivity necessary for bacterial measure- ments in water, the sample must be concentrated before assay. Filtration systems were evaluated for ability to concentrate bacteria from large volume samples rapidly, efficiently, and with- out damage to the organisms. Results indicated most filtration systems test- ed had a limited capability to meet project criteria. Promising results (200- to 600-fold concentration and up to 88% recovery of bacteria) were obtained using hollow-fiber concentra- tion systems modified to incorporate repeated backwash steps. This Project Summary was devel- oped by EPA's Municipal Environmen- tal Research Laboratory, Cincinnati, OH, to announce key findings of the research project that is fully docu- mented in a separate report of the same title (see Project Report ordering information at back). Introduction Current methods for examining bac- teriological quality of potable water supplies depend on time-consuming culture procedures. Efficient control of contamination breakthroughs into fin- ished waters or detection of water quality deterioration in distribution networks is delayed until results from the coliform test and standard plate counts become available. A method for rapid measurement of bacterial concen- tration in potable water would greatly enhance water quality maintenance and control. Criteria for an automated assay include sensitivity comparable to established methodology, minimum mechanical manipulations, real-time analysis, and quality controlsfor required reagents. Intracellular adenosine triphosphate (ATP) extracted from microorganisms can be detected and quantified with the use of the firefly luciferase ATP assay. This assay is a rapid, "wet chemical" reaction that is easily automated. Rela- tive ATP concentrations could be used to monitor changes in the levels of microbial populations as well as detect contamination breakthroughs into fin- ished water supplies or water quality deterioration in network distribution systems in "real time." If the majority of microbes in the water supply are assumed to be bacteria, ATP levels could be ------- further used to estimate the total number of bacteria present. Experimental data indicate that the deviation of ATP con- tent of bacteria, accounting for differ- ences between species and variation during the growth cycle, does not exceed one order of magnitude. The sensitivity limit of the ATP assay with pure cultures of bacterial strains is about 10s bacteria/mL Assuming 500 bacteria/mL as an acceptable upper limit for the standard plate count popu- lation in drinking water, contaminating levels of bacteria must be concentrated by a factor of at least 200 to permit detection by an ATP assay. For example, a 20-L sample volume containing 500 bacteria/mL would have to be reduced to 100 mL with complete retention of bacteria to provide a concentration equal to the sensitivity limit of the assay. To ensure operation above the minimum sensitivity limit, bacterial concentration greater than 105 cells/mL would be desirable. This study focused on development of an automated ATP assay and a suitable technique for concentrating low levels of bacteria from large sample volumes. The concentration technique had to: (1) be compatible with automation, (2) provide a high concentration factor, (3) have a rapid concentration rate, and (4) yield complete recovery of intact, undam- aged bacterial cells in a reduced volume that could be diverted to an automated ATP assay system. Automated A TP Assay The established method for bacterial ATP assay generally includes adding ATP extractant by pipet to the bacterial sample in a polypropylene test tube, diluting by pipet, mixing, distributing the enzyme mixture into individual cuvettes, and finally, injecting 0.1 mL extracted, diluted sample into a luciferase-contain- ing cuvette positioned in front of a photomultiplier tube. The firefly luciferase ATP assay was adapted to an automated flow system that greatly reduced the mechanical manipulations required (Figure 1). Buch- ler peristaltic pumps move both sample and reagents through interconnected tubes. Relative flow rates shown in Figure 1 provide the optimal concentra- tion of nitric acid extractant (0.1 N), with minimal sample dilution (50%), and 0.2 mL of luciferase enzyme solution. The luciferase enzyme is pulsed into the flow system only as the final processed sample reaches the glass coil positioned next to the photomultiplier tube. A Chem Glow photometer* (American Instrument Company) equipped with a coiled flow cell was used to measure maximum light output. Total light pro- duction was measured by coupling an Aminco Integrator-Timer to the photo- meter. Sample ATP concentrations were compared with ATP standards(0.1 fjg ATP/mL); the latter were prepared by diluting purified ATP (Sigma Chemical Company) in sterile, deionized water. ATP concentrations were converted to bacterial numbers using the conver- sion factor of 2.5 x 10~10 /ug ATP/cell, which represents the average ATP content of 19 bacterial species. For application of the flow ATP assay to potable water analysis, the ATP conver- sion factor should be determined using bacteria isolated from actual water samples. Figure 2 shows typical ATP concentra- tion curves determined for Escherichia coli in saline and in tap water using the flow ATP assay. Nitric acid (0.6 N) extraction was most suitable for bacterial ATP and was optimized for use in the flow system. Use of acid extraction requires that the final sample be buffered to pH 7.75 to optimize the luciferin-luciferase enzyme reaction. Luciferin-luciferase extracted from dessicated firefly lanterns was obtained in purified form from E. I. Dupont de Nemours and Company, Inc., or was prepared in the laboratory. Use of highly purified luciferase requires adding synthetic luciferin and results in 100- fold increase in test sensitivity. Purified luciferase is rehydrated in 0.25 M TRIS buffer containing 0.01 M MgS04 and 0.001 M Cleland's reagent (dithiothreitol). When Cleland's reagent is included, stability of rehydrated luciferase increases from 4 hours at 10°C to 8 hours at 10°C. At least 60% of functional activity can be preserved at 10°C in the dark for 48 hours by adding 0.001 M EDTA. Since luciferase is an enzyme, precaution must be taken to avoid denatuation due to overheating or exposure to marked temperature fluctu- ations. For ATP assay, the enzyme preparation should be brought to 20°C. To minimize inherent light from the luciferase mixture, rehydrated enzyme should be incubated at room tempera- ture for 30 minutes before assay.Re- maining inherent light should be less Relative Flow Rates J.20 ml/min 60 sec Residence Drain Flow Head Glass Coil Photomultiplier Tube Chem Glow Photometer 0.25 M TRIS-pH 8.2 0.01 M MgS04 0.001 M Cleland's Reagent Recorder To Drain Figure 1. Schematic of automated firefly lucifera.se flow system for detecting bacterial A TP including nitric acid extraction and subsequent dilution. "Mention of trade names or commercial products does not constitute endorsement or recommenda- tion for use. ------- Is u. •Q r Legend: O in Saline • in Tap Water 1 Figure 2. 6789 Log Bacteria/'mL by Plate Count Concentration curve ofE. coli in saline and in tap water; bacteria/mL by A TP flow vs. plate count. than 10% after 30 minutes, and an ppropriate control cuvette will indicate the inherent light correction needed. ATP standards were prepared with the use of purified chemical ATP (Sigma Chemical Co.) in 0.001 M EDTA and 0.01 MMgSCv The purified ATP is used for daily quality control and for preparing ATP standard curves for determining sample ATP concentrations. Loss of sample bacterial ATP due to hydrolysis by nitric acid extractant was minimized by 50% dilution after 60- second extraction, followed immediately by reaction with the luciferase enzyme system. for subsequent assay using the ATP flow system. Recovery in a small volume of water rather than on a membrane surface would permit measurement of the bacterial population by standard plate count, Coulter count, or other method to confirm ATP assay results. Additionally, the concentration pro- cedure must be able to process large volume samples (at least 10 L) in a time period compatible with the desired sampling frequency without damaging the cells or significantly altering the cell ATP content. Potential concentrator! techniques were tested for concentration of known densities of E. co//from sterile, deionized water. After a minimum of three test runs, if a concentration factor of 10 could not be achieved with 50% recovery of bacteria, further work with the tech- nique was abandoned. The percentage of bacteria recovered in the final concen- trate was determined by enumerating the bacteria before and after concentra- tion. Standard plate counts, ATP assays. Coulter counts, luminol iron porphyrin assays, or a combination of these methods was used to determine cell concentration. Four basic concentration techniques were tested; for some techniques, more than one type of unit was evaluated. The techniques tested were: centrifugation; direct in-line filtration, flat-surface membrane filtration, and hollow fiber filtration. Centrifugation Continuous flow centrifugation pro- vided large sample volume processing with about 75% bacterial recovery, but concentration factors were very low. Mechanical manipulations involved in collecting the bacterial concentrates were not suitable for automation, so work on this procedure was discontinued. Direct In-Line Filtration Sample under pressure was passed through a 0.45 /urn cellulose acetate filter (Swinnex filter) unit to concentrate bacteria. After sample filtration, attempts to backwash with small volumes of deionized water to resuspend cells for the ATP assay were unsuccessful. Adequate seals to prevent leakage could not be made using Nucleopore polycar- bonate filters. With 10-fold concentration of bacter- ial cells, attempts to assay ATP using direct extraction of ATP after filtration resulted in an average of only 18% recovery (Table 1). Bacterial Concentration Procedures To detect contaminating levels of microorganisms in potable water using the firefly luciferase ATP assay, the organisms must first be concentrated to achieve at least the minimum sensitivity cell concentration of 105 cells/mL. For samples containing 500/mL, 200-fold or greater concentration must be achieved. If concentrated by membrane filtration, the cells must be recovered intact and undamaged in a small volume of water Table 1. Test Results Using Mill/pore In-Line Filter Concentration with In-Line Extraction In-Line Extraction Sample fag A TP/mL) External Extraction fag ATP/mL) % Recovery with In-Line Extraction 1 2 3 4 5 3.3 x 1Q-* 3.5 x W* 2.5 x W4 8. 1 x 10~* 1.0x10-* 1.9 x 10~3 1.85 x 10'3 1.8 x 10~3 1.08 x 10'3 3.3 x W* 17 19 14 7 30 ------- Flat-Surface Membrane Filtration Sample concentration by membrane filtration systems that use flat surfaces and high sample flow velocities parallel to the membrane filter surface was explored. The sample is concentrated by volume loss through the filter, but particles are retained in a decreased sample volume (the retentate) (Figure 3). Sample flow parallel to the membrane surface reduces particle buildup on the membrane surface. Three systems utilizing this principle were tested; these were Uni-pore Radial and Stirred Flow Cells (Biorad Laboratories), Sartor- ius Ultrafiltration System (Sartorius Corporation), and the Pel I icon Cassette Molecular Filter (Millipore Corporation). All flat-surface systems yielded poor results when tested with bacterial suspensions, and recoveries were less than 50%. A backwash step was added in an attempt to improve bacterial recoveries from such systems. Back- washing alone proved to be insufficient to recover the bacteria, and the volume of backwash required often negated any concentration effect. Adding Triton X- 100 (a nonionic detergent) or 0.1% Rhozyme (proteases and glucosidases mixture isolated from Aspergillus oryzae) to samples as a filtration aid did not consistently improve bacterial recovery with any of the flat-plate systems, and concentration factors greater than 10 could not be achieved (Table 2). Hollow-Fiber Membrane Filtration Hollow-fiber membrane filters also incorporate sample flow parallel to the membrane surface, but the membrane configuration is tubular and has pores of controlled dimensions (Figure 4). The sample solution is pumped through a bundle of the hollow fibers, and particles larger than the pore size of the particular hollow-fiber filter in use are retained. Particle concentration occurs when low molecular weight solutes and water pass through the membranes into the filtrate. The sample volume is recircu- lated to achieve further volume reduction and consequent sample concentration. Backwashing was also necessary with the hollow-fiber membrane filter sys- tems. Two hollow-fiber member filter systems, the Bio-Fiber 80 Mini-plant (Biorad Laboratories) and the Diaflow Hollow Fiber Concentrator (Amicon Corporation), were tested. The Bio-Fiber 80 Mini-plant yielded up to 83% bacterial recovery at concen- tration factors of 15 to 250 with the use of repeated backwashing and reconcen- tration steps. However, the unit was removed from the market and is no longer available. The Amicon Diaflow Hollow Fiber unit, with a surface area of 1,000 cm2 and 100,000-molecular weight (m.w.) cutoff, yielded nearly 100% bacterial recovery at 10-fold concentration without backwashing (Table 3). When the concentration factor was raised to 100 fold, recoveries dropped to only 32%. A dual cartridge unit with 50,000-m.w. cutoff cartridges, operated to provide 100-fold concentra- tion without backwashing, gave incon- sistent recoveries that ranged from 43% to 100%. After modification to permit backwashing, a 10,000-cm2 cartridge yielded recoveries of 53% (single back- wash) to 88% (three backwashes). Hollow-fiber concentrators modified for backwash capability proved to be the only concentration systems tested that allowed greater than 200-fold concen- tration of bacteria with adequate recov- ery of the organisms (88%). The hollow fibers of the Amicon Diaflow unit are made of noncellulosic polymers that should be durable for a period of months. For extended use, cleanup steps may be necessary to prevent the buildup of bacteria within the fibers. Although a 30-minute rinsing procedure using sterile deionized water was effective between samples during testing, use of 0.1 NaOH followed by sterile deionized water rinse may be necessary during long-term use. Manpower constraints forced Goddard Space Flight Center to discontinue projects not directly related to the space mission. They were, therefore, unable to evaluate and develop a satisfactory concentration system to go with the flow system for the ATP assay. Although preliminary results appeared promising, extended testing of the hollow-fiber concentrator (backwash modified) using Retentate Buchler Peristaltic Pump Sartor/us Ultrafiltration System Filtrate Backwash Sample Reservoir and Concentrate Collection Figure 3. Schematic of Ultrafiltration System using tangential flow and backwash. System is diagrammed in concentrating mode; valves are rotated 90° for backwashing provided by pressurized tap source of sterile, deionized water. Table 2. Test Results Using Sartorius Ultrafiltration System with Tangential Flow and Backwash Concentration % Factor* Recovery Filtration Aid 2 4 10 10 10 5 5 79 63 21 90 92 50 80 _ - _ _ With TX With Rhozyme Cellulose Acetate Filter, TX Polycarbonate Filter, TX *Each factor represents a separate test. ------- drinking water samples is needed before final recommendation can be made. If performance of the concentrator proved satisfactory with potable water samples, additional testing of the system, includ- ing the automated flow ATP assay would be necessary to establish overall system performance, sensitivity, repro- ducibility, and reliability. The full report was submitted in fulfillment of Interagency Agreement No. EPA-IAG-D6-0982 by NASA/God- dard Space Flight Center under the sponsorship of the U.S. Environmental Protection Agency Concentration Mode: Sample J_ T , .1. J ' Backwash and Collection Mode: Filtrate Concentrate . 1 — ' 1 — ' 1 A Water Backwash Figure 4. Schematic of Amicon hollow fiber cartridge modified for backwash capability. Table 3. Test Results with the Use of Amicon Hollow Fiber Cartridge with Backwash Maximum Cartridge Size Concentration Test Filter Area (cm) Factor Procedure ^Standard deviations calculated on the basis of test runs. < Recovery + o* 1,000 10.000 60 Backwash 600 Backwash Backwash, Refilter, Backwash Backwash, Refilter. Backwash, Refilter, Backwash 95 ± 5 53 ±21 83 ± 14 88 ± 12 ------- Grace L Picciolo, Emmett W. Chappell, Jody w. Darning, RichardR. Thomas, and D. A. Nibley were with the NASA/GoddardSpace Flight Center at the time this research was performed; Harold Okrend was with Howard University. All inquiries should be directed to the EPA Project Office. Donald Reasoner is the EPA Project Officer (see below). The complete report, entitled "Firefly Luciferase ATP Assay Development for Monitoring Bacterial Concentrations in Water Supplies," (Order No. PB 81-163 271; Cost: $8.00, subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield. VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Municipal Environmental Research Laboratory U.S. Environmental Protection Agency Cincinnati, OH 45268 > US GOVERNMENT PRINTING OFFICE: 1961-757-012/7075 ------- United States Center for Environmental Research Fees Paid Environmental Protection Information Environmental Agency Cincinnati OH 45268 Protection Agency EPA 335 Official Business Penalty for Private Use $300 RETURN POSTAGE GUARANTEED TNrd-Clwl 'f Bulk Rata ------- |