ATP measurements for online monitoring of microbial drinking quality - evaluating potential Óluva K. Vang Measuring microbial drinking quality today? Current methods: late results (2 to 3 days) contamination spread in the distribution system consumed Low frequency sampling contaminations not detected Supervisors: Hans-Jørgen Albrechtsen, DTU Environment Claus Tilsted Mogensen, Grundfos Continuous monitoring concept Sensor application Conc. (pg ATP/mL) 8 6 2 Continuous monitoring i (e.g. ATP) Routine monitoring (grab samples)..8 2.2 2.6 3 3. 3.8( year.2period).6 5 Time Detection of a pulse contamination through continuous monitoring between regulatory controls (grab sampling) Possible contamination scenarios: Breach in hygienic barrier (treatment) at the works Intrusion of contaminant (e.g. dirt, surface ) when renovating the distribution network Intrusion of contaminant (e.g. bird feces, surface ) into tower/reservoir Sensitive consumers (e.g. hospitals, pharmaceutical and food industry) Monitoring points: Water works outlet Major pipe connections/branches Water tower/reservoir outlet Inlet to consumer
Novel method Principle of ATP-method ATP for continuous monitoring of microbial drinking quality??? Advantages: Simple measurement procedure Results within minutes (real-time analysis) Measurement for all active cells Small sample volumes firefly luciferase ATP D luciferin O Mg2 AMP PPi oxyluciferin CO light f 2 i 2. Extraction of ATP (cell lysis) 2. Reaction with luciferine/luciferase (production of light) 3. Light emission measurement by a photomultiplier (relative light units) Aim To optimize and further develop the ATP method for measuring microbial drinking quality Quantitative measurements on microbial ATP Solving problems: matrix effects (e.g. turbidity, color) and internal standard Stability and storage of reagents Cell lysis techniques (detergents, heat) Handling and storage of samples during analysis Background for lab experiments Sensor perspectives: maintenance e.g. once a month Focus on storage conditions and stability/activity of reagents Effect of various contamination sources (waste, surface etc.) in different experimental set-ups Validation of method on a sensor platform (field investigations) 2
Activity of luciferine/luciferase (LL) Activity of luciferine/luciferase (LL) How does activity of luciferine/luciferase reagent decrease with time? Fluorescence of LL decreased with time Reduction in activity (rlu) after 33 days: Aerob -2 o C : ~% Aerob + o C : ~6% Anaerob + o C : ~3% Same reduction was observed for all tested standards (, 5 and ) Standard (-standard ) 3, 25, 2, [rlu] 5,, 5, Aerobe -2C Aerobe +C Anaerobe +C 5 5 2 25 3 35 Time [days] Activity of luciferine/luciferase (LL) Stability of ATP standards Acceptable loss in activity of luciferine/luciferase Calibration by addition of internal ATP standard (IS) Stability of IS? Reduction in activity: Roche standard in tap : 2% Roche in tricine: i 5% Roche in MilliQ : 7% Celsis in buffer: 3% (rlu) 8 6 2 8 6 2 Standard (- standard '') Celsis (buffer) -8C Celsis (buffer) +C Celsis (tap ) -8C Celsis (tap ) +C Roche (tricine) -8C Roche (tricine) +C Roche (MilliQ) -8C Roche (MilliQ) +C Roche (tap ) -8 C Roche (tap ) +C Linear (Celsis (tap ) -8C) Linear (Celsis (tap ) +C) Linear (Roche (MilliQ) -8C) Linear (Roche (MilliQ) +C) Linear (Roche (tricine) -8C) Linear (Roche (tricine) +C) Linear (Roche (tap ) -8 C) Linear (Roche (tap ) +C) Linear (Celsis (buffer) -8C) Linear (Celsis (buffer) +C) 5 5 2 25 3 Time (days) Celsis standard in sterile filtrated autoclaved tap was stable throughout a week period at +ºC 3
Cell lysis efficiency Commercial ATP reagent kits Quantitative analysis for microbial ATP depends on an efficient cell lysis Complete lysation high h output t and accurate measurement Challenges: low quantities of ATP in drinking no loss of ATP after extraction no significant inhibition on the luciferin/luciferase reaction rlu Promicol: 3 enzymes reagents and 2 extraction reagents ATP standard curves total ATP 3, Enzyme & EX reagent Enzyme & EX reagent 5 25, Enzyme 2 & EX reagent Enzyme 2 & EX reagent 5 2, Enzyme 3 & EX reagent Enzyme 3 & EX reagent 5 5,, Investigation of 9 commercial reagents in order to identify: sensitivity measurement stability cells lysis efficiency 5, 2 6 8, Conc. (pg ATP/ml) Rlu signal: Enzyme 3 > enzyme 2 > enzyme Different slopes (quenching of light signal): extraction reagent 5 gives a higher quenching than extraction reagent Commercial ATP reagent kits Commercial ATP reagent kits rlu Linearity and y-axis intersection: 2,5 2,,5, 5 Enzyme & EX reagent Enzyme & EX reagent 5 Enzyme 2 & EX reagent Enzyme 2 & EX reagent 5 Enzyme 3 & EX reagent Enzyme 3 & EX reagent 5 ATP standardcurves 2 6 8 Conc. (pg ATP/ml) Enzyme lowest sensitivity Drinking samples microbial ATP Enzyme Enzyme 2 Enzyme 3 Sample ID Extraction Extraction 5 Extraction Extraction 5 Extraction Extraction 5 (pg ATP/ml) (pg ATP/ml) (pg ATP/ml) (pg ATP/ml) (pg ATP/ml) (pg ATP/ml) 5 3.7 3.7.5..8 3.5 237. 3. 3. 6.7. 8. 28 3.6 3.7.8 3.7.7 3.3 326. 3.7 2..3 3.2 5. 32 3.5 6. 2.6 3.6 3.5 22.9 2.5 2. 6.3 5. 6. 2.9 Lyngby WW.6 2.8 6.. 6.3.3 YE in DW (^) 369.7 93.6 339.9 876. 38.2 792.2 YE in DW (^ ) 28. 6.8 36.3 6. 33.8 56.3 YE in DW (^ 2 ).8.3 3. 5.6 3. 5.8 Negative values indicate too low sensitivity (enzyme ) Higher yield with extraction reagent 5 a better cell lysis
Commercial ATP reagent kits Performance of the ATP method Promicol: Overall good measurement stability: CV <5% in most cases Various sensitivties of enzymes: enzyme 3 Various cell lysis efficiency: extraction reagent 5 6 other reagent kits were also investigated Response in ATP in different contamination scenarios surface waste Reagent significance: LuminATE vs. RapiScreen Health (Celsis) Promicol vs. Celsis Surface contamination Surface was collected from dig-outs for pipe connections Simulation of drinking contaminations 5
Surface contaminations LuminATE reagent kit Drinking contaminated with surface (LuminATE reagent kit) 58 Conc. (pg ATP/m ml) 7 35 6 3 5 25 2 3 5 2 5 2 3 3 5 5 6 6 7 7 Drinking Dilution i of surface Dilution of surface Total ATP Non microbial ATP Microbial ATP Waste contaminations LuminATE reagent kit Drinking contaminated with waste (LuminATE reagent kit) 6 5 2 Conc. (pg ATP/m ml) 8 3 6 2 2 3 5 5 6 6 7 7 Drinking Dilution of of waste Total ATP Non microbial ATP Microbial ATP -2 to -3 dilution of surface ( to L in m 3 drinking ) -3 to - dilution of waste (. to L in m 3 drinking ) Waste contamination RapiScreen Health reagent kit Drinking contaminated with waste (RapiScreen Health reagent kit) 6 Conc. Conc. (pg (pg ATP/m ATP/m ml) ml) 35 5 3 25 23 5 2 5 3 5 5 6 6 7 7 Drinking Waste contamination RapiScreen Health vs. LuminATE Improved quantification of low ATP-concentrations ti especially fraction of non-microbial ATP More stable measurements Slightly improved sensitivity Dilution of of waste Total ATP Non microbial ATP Microbial ATP - dilution of waste (. L in m 3 drinking ) 6
Traditional microbiological methods ATP compared with other methods: Total direct cell counts (DAPI) Heterotrophic t plate counts (yeast) Colilert -8 (E. coli and coliforms) Drinking contaminations Total Direct Counts (DAPI) Conc. (cells/ml) 35 3 25 2 5 5 Total direct counts DAPI Waste Surface 2 3 5 6 7 Drinking Colilert-8 (E. coli) HPC (22 /37 ) TDC (DAPI) Colilert-8 (coliforms) Dilution of contaminant Surface : L in m 3 drinking Waste : L in m 3 drinking Drinking contaminations Heterotrophic plate counts (yeast extract) 22ºC Conc. (CFU/ml) HPC 22 C Surface : L in m 3 drinking Waste :. L in m 3 drinking Waste Surface 3 5 6 7 Drinking Dilution of contaminant Drinking contamination Colilert-8 (coliforms coliforms) Coliforms in contaminated drinking >29.2 2 >29.2 2 Conc. (MPN/ ml) 3.5 2. 3.8 29.9 7. Waste Surface < < < < < 3 5 6 7 Drinking Dilution of contaminant Surface : ml in m 3 drinking Waste :. ml in m 3 drinking 7
Drinking contaminations Colilert-8 (E. coli) Conclusions Colilert-8 (E. coli) E. coli in contaminated drinking >29.2 986.3 Waste Surface 77.9 2. 2. < < < < < < < Conc. (MPN/ ml) 3 5 6 7 Drinking Dilution of contaminant ATP compared to other methods (TDC, HPC, Colilert-8) HPC and Colilert-8 more sensitive Depending on load of bacteria in contaminant t RapiScreen Health was more sensitive than LuminATE ATP can be used as a monitoring method for microbial drinking quality. Can detect -3 - dilutions of surface in drinking - -5 dilutions of waste in drinking dilutions of waste in drinking Continuous contamination at low concentrations no Pulse contaminations yes Surface : not detected (3 MPN/ ml in surface ) Waste :. ml in m 3 drinking Field testing ATP-sensor prototype Continuous monitoring Collaboration with Promicol (The Netherlands) Field investigations Continuous time series of ATP concentrations in drinking detection of pulse contaminations? Fluctuations in ATP concentrations on short term basis (minutes to hours) and long term basis (days) (In)consistency between standard microbial methods and ATP measurements ements (grab samples vs. continuous o monitoring) ing) RLU Lumatic 5,, 35, 3, 25, 2, 5,, 5, ATP sensor prototype Total ATP :33 :3 :33 2:3 2:33 3:3 3:32 Time Peak around lunch time 8
Summary Thank you Criteria for ATP as a monitoring method Luciferin/luciferase activity: ok Stable internal ATP standard: d yes Lysis efficiency: depends on extraction reagent Sensitivity: depends on enzyme reagent Questions? Performance of ATP as a monitoring method: Detection of waste and surface : yes, depends on load Continuous time series on a sensor platform: validation upcoming NOW! 9