TY - JOUR
AU1 - Wakizaka,, Yoshikazu
AU2 - Itoi,, Takayuki
AU3 - Takano,, Masayo
AU4 - Kato,, Eiko
AU5 - Sato,, Yusei
AU6 - Morita,, Satoshi
AU7 - Enjoji,, Takaharu
AB - Abstract Background PixeeMo™ is a compact instrument that enables bacterial cell counting using microfluidic chips instead of counting of colonies on culture media. Chips containing electrodes, based on fluid, electric filtering and sorting technology (FES), allow the selection of bacterial cells from other components in the sample. In the United States (US), surface water or ground water affected by surface water must be treated to reduce the total microbial load to less than 500 CFU/mL. In Japan, drinking water regulations limit the total bacterial load to 100 CFU/mL. Objective To validate the PixeeMoTM aerobic bacteria method based on the Japanese regulation in the range of 30–300 CFU/mL in drinking water. Method PixeeMoTM aerobic bacteria method was compared to the Standard Method for the Examination of Water and Wastewater (SMEWW) 9215B (2017) using naturally contaminated drinking water. Results The maximum repeatability standard deviation of the PixeeMoTM method was 14.8%. The difference of mean log10 values between the PixeeMoTM and SMEWW 9215B methods ranged from −0.015 to 0.258. Similar results were obtained in the independent laboratory study. Conclusions The PixeeMoTM method is equivalent to that of the SMEWW 9215B methods. The product consistency and stability study demonstrated no significant difference within the expiration date. The robustness study confirmed that there was no effect within the expected range. The instrument variation study also demonstrated no significant difference among the data of three PixeeMoTM instruments. Highlights Total counts of bacteria in drinking water can be determined accurately within 1 h with PixeeMoTM. Scope of Method Analyte(s).—Aerobic bacteria. Matrix.—Drinking water. Summary of validated performance claims.—Performance of the PixeeMoTM method is equivalent to that of the Standard Method for the Examination of Water and Wastewater (SMEWW) 9060, Samples (1) and SMEWW 9215, Heterotrophic Plate Count, Part B—Pour Plate Method (2) for enumeration of aerobic bacteria in drinking water. Definitions Repeatability (sr).—Standard deviation of replicates for each analyte at each concentration of each matrix for each method. Mean difference between candidate and peference methods.—Mean difference between candidate and reference method transformed results with 95% confidence interval for each analyte at each concentration of each matrix. Principle The PixeeMoTM method consists of feeding a liquid sample to a microchannel chip, capturing bacteria contained in the sample, observing microscopically, and counting the captured bacteria. The measurement flow diagram of the PixeeMoTM method is shown in Figure 1. Fluid, electric filtering and sorting technology (FES) is used for capturing bacteria. FES is a selection technology that integrates electrical measurement and fluid control. A sample is fed into an appropriately designed microchannel, an AC electric field is generated by the electrode in the microchannel, and the electric force acts to attract bacteria when passing over the electrode. This method uses the intracellular ion homeostasis of the bacteria, therefore an intact cellular membrane is required for capture on the electrodes, implying only viable organisms will be captured. To perform the method, the water sample is first centrifuged, then the water replaced with buffer, and the sample is then loaded into a syringe, which is placed on the instrument and brought into contact with the chip. The software controls the settings and operation of the instrument for drinking water analysis. While the syringe plunger expels liquid at a constant flow rate for a specified time period, the voltage supply is turned on and microbes are captured on the electrodes. At the end of the collection period, digital images are taken and analyzed by the software to determine the microbial concentration. By collecting the bacteria on the electrode, it is possible to directly observe and count the bacteria within one or two fields of view. With this method, the bacterial concentration of a sample is determined by counting the captured bacteria from 1 mL of sample fed through the microchip. Feeding time is approximately 17 min, therefore it is possible to determine the bacterial concentration of a sample within 1 h. Figure 1. Open in new tabDownload slide Measurement flow diagram of PixeeMoTM method. Figure 1. Open in new tabDownload slide Measurement flow diagram of PixeeMoTM method. General Information The United States (U.S.) (3), European Union (EU) (4), and World Health Organization (WHO) (5) do not have regulations for total counts in drinking water, only limits for coliforms and Escherichia coli. In the U.S., surface water or ground water affected by surface water must be treated to reduce the total microbial load to less than 500 CFU/mL (3). In Japan, aerobic plate count in drinking water is regulated at less than 100 CFU/mL (6). With the PixeeMoTM method, it is possible to measure bacterial concentrations correlated with aerobic plate counts. This method enables the measurement of bacterial concentration of drinking water rapidly and easily without culture. This paper reports on the AOAC validation of the PixeeMoTM method. Materials and Methods Test Kit Information Kit name.—PixeeMoTM chips, package of 100. Catalog No.—ELC121. Ordering information.—AFI Corp., Shin-nihon Umeshin Building 2F, 6-7-2 Nishitenma, Kita-ku, Osaka 530-0047, Japan. e-mail: info@afi.co.jp. Tel: +81-6-6360-9010. Additional Supplies and Reagents PixeeMoTM buffer.—Cat. No. ELB100N (AFI Corp.), 1 L. Waste tank.—(AFI Corp.). Syringe.—10 mL disposable, Cat. No. SS-10SZ (Terumo Corp., Tokyo), or equivalent. Centrifuge tubes.—50 mL with screw cap. Pipets.—25 mL. Sodium thiosulfate.—Cat. No. 197-03605 (FUJIFILM Wako Pure Chemical Corp., Osaka), or equivalent. Standard method agar.—Cat. No. 05618 (Nissui Pharmaceutical Co., Tokyo), for reference method. Apparatus PixeeMoTM.—ELS-002 (AFI Corp.) including dedicated 64-bit PC with Windows 10 Pro installed with PixeeMoTM counter software (Version 2.3 R4). Centrifuge.—With 50-mL conical tube rotor capable of 8000 × g. Vortex mixer. Incubator.—Capable of maintaining 35 ± 1°C, for reference method. Water bath.—Capable of maintaining 45 ± 1°C, for reference method. Reference Material Escherichia coli.—American Type Culture Collection (ATCC) strain 25922 (Manassas, Virginia, USA). General Precautions Read this supplied document thoroughly and carefully follow the specified operating procedure and precautions during use. It is not possible to guarantee the reliability of items that have exceeded their specified expiry dates and these products should not be used. Electrode chips should be used once only. It is not possible to guarantee the reliability of chips that have been exposed multiple times and they should not be used. Do not use electrode chips that show obvious signs of abnormalities such as damage to their containers, contamination, discoloration, or absorption of moisture. Place any remaining electrode chips in their aluminum bag, seal the bag with tape and store it so the chips will not be exposed to moisture or light. The chips should be used as soon as possible. Take care as the electrodes are particularly prone to oxidation by ambient oxygen. Operating Precautions It is important to prevent falling microorganisms or suspended particulate matter such as fine particles or dust from being mixed into the solution and affecting measurement. When handling the syringe or PixeeMoTM chip, ensure contaminants do not enter the end of the syringe or sample supply port of the chip. Also ensure contaminants are not transferred from your fingers. When inserting the PixeeMoTM chip into a slot on the stage, align the left edge of the chip with the left edge of the stage and insert it to the back end. If it is not fully inserted into the back end, it may cause an electrical contact failure. When capturing measurement images, adjust the focus to ensure the lines of the electrodes can be seen clearly. It will be difficult to perform an accurate count unless the focus is sharp. Be sure to check them before capturing the 1st and 2nd measurement images. If there are false detections, manually correct after automatic analysis. It is important to accurately measure. Suspended particles other than microbes may be collected on the electrodes. In this case, manually correct to exclude the non-microbial particles in the count. Safety Precautions If medium or reagent enters your eyes or mouth, immediately wash the area with plenty of water, and consult a physician for medical advice. Handling of microorganisms always involves a certain risk of laboratory-acquired infection. They should only be handled under the supervision of a skilled specialist after taking appropriate biohazard prevention measures. Any laboratory equipment or medium that has touched a sample should be regarded as infectious and handled appropriately. Sample Preparation Collect a 25 mL drinking water sample in a 50 mL centrifuge tube. Centrifuge at 8000 × g for 10 min. Carefully remove 22.5 mL of supernatant. Add 22.5 mL of PixeeMoTM buffer and vortex mix. System Startup Turn on the power switch on the front of the PixeeMoTM main unit. Turn on the power to the dedicated PC, start Windows, and log in. Start up the PixeeMoTM counter software (Version 2.3 R4). Sample Setup Open the aluminum bag containing the PixeeMoTM chips and take one out. Insert the PixeeMoTM chip in a slot on the stage. Align the left edge of the chip with the left edge of the stage and insert it into the back end. Vortex mix the sample/buffer solution and draw 3 mL into the 10 mL syringe. Note that the minimum sample volume required for the measurement is 1 mL + 0.2 mL (corresponding to the dead volume in the chip); however, 3 mL are loaded into the syringe to ensure ample volume. Expel any residual air from inside the syringe. Raise the plunger of the syringe piston on the main unit and set the syringe containing the sample in the syringe holder. Lower the setting lever to bring the syringe and PixeeMoTM chip in contact. Attach the waste tank to the waste liquid port. Turn the condenser turret to select phase contrast (Ph1). Use the light control knob to adjust the illumination to an appropriate level for observation. Measurement Operation In the PixeeMoTM counter software (Version 2.3 R4), fill in the product name and lot name fields. Choose the “drinking water” preset. Click once the “syringe continuous down” button. The syringe will automatically move to the proper position of measurement start and preparation for inspection will be completed. Turn the adjustment knob while watching the screen and align the position with the measurement electrode area. Display one image of six slits in line from the left side of Channel 2 (CH2). Adjust the focus until the lines of the electrodes can be seen clearly. Readjust the position if the measurement position shifts during focus adjustment. Press the “start button” to begin the measurement. The liquid feeding and the capture of the bacteria will start, and a first photo is taken automatically. Turn the adjustment knob while watching the screen and align the position with the next measurement electrode area. Display one image of five slits in line from the right side of CH2 and take a photo. At the same time, readjust the focus if necessary, to ensure the lines of the electrodes can be seen clearly. When focused well, the electrode edge looks sharp and the edge shines slightly. After taking the second photo, click the completion button. Captured bacteria are automatically counted by image analysis. Manual Correction and Report Output After automatic analysis, manually correct false detection as necessary. There are three types of false detection: When there are many bacteria, cells are trapped in a line so-called “pearl chain” or block. If there are visible boundaries in the pearl chain, count as one bacterium per boundary. In the case of long bacteria with no visible boundaries, count as one. Occasionally, the edge of the electrode may be falsely detected. If there is nothing, but something is falsely detected, delete it. In rare cases, clearly non-bacterial deposits may be falsely detected, delete these. Press the report output button. Results are automatically converted to the number of bacteria per 1 mL and a report is generated. There is an upper limit to the bacterial concentration that can be counted by PixeeMoTM. If the count exceeds 300 cells/mL, or if there are more than four pearl chains bridging the electrode in one screen, it is necessary to dilute the sample in PixeeMoTM buffer and repeat the analysis. After the Test Raise the setting lever to detach the syringe from the PixeeMoTM chip. Remove the PixeeMoTM chip and the waste tank from the slot. At this time, be careful not to spill the waste liquid remaining in the waste tank. Whether bacteria are detected or not, the removed PixeeMoTM chip and waste tank containing waste liquid should be disposed of properly as biohazardous materials. Validation Study This validation study was conducted under the AOAC Research Institute Performance Tested MethodSM program and the AOAC INTERNATIONAL Methods Committee Guidelines for Validation of Microbiological Methods for Food and Environmental Surfaces (7). Method developer studies were conducted in the laboratories of AFI Corp. and included a matrix study, product consistency and stability studies, instrument variation, and robustness testing. The independent laboratory study was conducted by Japan Food Research Laboratories and included a matrix study. Method Developer Studies Matrix study Methodology Drinking water was prepared according to SMEWW 9060 and evaluated by the PixeeMoTM method and by the SMEWW 9215 reference method Part B (Table 1). The study included five replicate test portions at each of three contamination levels. See Table 1 for target contamination levels. Table 1. Matrix study design Matrix . Inoculation . Target contamination level, CFU/mL . Replicate test portions per method . Reference method . Drinking water Naturally contaminated 50 5 SMEWW 9215B 500 5 5000 5 Matrix . Inoculation . Target contamination level, CFU/mL . Replicate test portions per method . Reference method . Drinking water Naturally contaminated 50 5 SMEWW 9215B 500 5 5000 5 Open in new tab Table 1. Matrix study design Matrix . Inoculation . Target contamination level, CFU/mL . Replicate test portions per method . Reference method . Drinking water Naturally contaminated 50 5 SMEWW 9215B 500 5 5000 5 Matrix . Inoculation . Target contamination level, CFU/mL . Replicate test portions per method . Reference method . Drinking water Naturally contaminated 50 5 SMEWW 9215B 500 5 5000 5 Open in new tab Analyst One prepared samples and randomized blind coded test portions. Drinking water is naturally contaminated, typically at ∼50–100 CFU/mL. Validation samples were prepared as follows to attain the desired contamination levels and residual chlorine was neutralized with sodium thiosulfate according to SMEWW 9060 (1) before analysis. To obtain the samples, cold tap water was allowed to run until the temperature stabilized before taking a 2 L sample. Three hundred milliliters were transferred to a clean container and 0.3 mL of 3% sodium thiosulfate was added and mixed well to neutralize up to 5 mg/L residual chlorine. The remaining chlorinated water was kept capped at refrigerated temperature. An initial plate count was performed on the neutralized water according to SMEWW 9215B (2). The neutralized water was allowed to sit loosely covered at 25 ± 1°C for a period of time (∼7 days) to reach ∼5000 CFU/mL as determined by plate count according to SMEWW 9215B (2). This was Material 3. The chlorine was then neutralized in the reserved water and an aliquot of Material 3 was mixed with neutralized reserved water to create a sample at ∼500 CFU/mL. This was Material 2. The neutralized reserved water was used as is for Material 1. All materials were well mixed. Five replicate test portions of 25 mL each from Materials 1, 2, and 3 were transferred into 50-mL screw cap tubes and randomized and blind coded for the PixeeMoTM method. Five replicate test portions of 5 mL each from Materials 1, 2, and 3 were transferred into screw-cap tubes and randomized and blind coded for SMEWW method 9215B. Analyst Two began all analyses of all test portions on the same day. The reference method was performed according to SMEWW 9215B, the pour plate method (2). Briefly, Plate Count Agar (PCA) was prepared and tempered to 45 ± 1°C in a water bath. Duplicate empty Petri plates were labelled for each blind coded test portion and for each 1:10 and 1:100 dilution of each blind coded test portion. Serial dilutions of each test portion were made by diluting 1 mL of test portion into 9 mL sterile water and 1 mL of the 1:10 dilution into 9 mL sterile water. One milliliter of the appropriate blind coded test portion or diluted test portion was pipetted into each plate and 12–15 mL tempered PCA was added and swirled gently to mix. The agar was allowed to harden and plates were incubated at 35 ± 1°C for 7 d (Note: Preliminary testing demonstrated that a 7 day incubation was required for sufficient colony growth.) Colony counts from duplicate plates containing 30–300 CFU were averaged and the CFU/mL of water for each blind coded test portion was calculated. Results A logarithmic transformation was performed on the reported CFU/mL using log10 [CFU/mL + (0.1)f] where f is the reported CFU/mL corresponding to the smallest reportable result. The candidate method and reference method transformed results were then plotted and regression analysis performed as shown in Figure 2. The candidate method results were lower than the reference method results at the high concentration level; however, there were no obvious discrepancies in the results at any concentration. Figure 2. Open in new tabDownload slide Method comparison plots of the PixeeMoTM method versus the SMEWW 9215B method. Figure 2. Open in new tabDownload slide Method comparison plots of the PixeeMoTM method versus the SMEWW 9215B method. The matrix study results and statistical analyses are summarized in Table 2. The maximum repeatability (RSDr) of the PixeeMoTM method was 14.8% in the low concentration level and the maximum RSDr of the SMEWW 9215B method was 1.7% in the low concentration level. The difference of means (DOM) between the PixeeMoTM and SMEWW 9215B methods ranged from −0.015 to 0.258. However, in microbiology, a difference of <0.5 log10 is not considered a significant difference (8). Moreover, the 95% confidence interval on DOM for each concentration was within (−0.5, 0.5). Thus, there is no statistically significant difference between the two methods. Table 2. Method comparison data summary and statistics Matrix (organism) . Contamination level . N . PixeeMoTM . SMEWW 9215B . DOMa . 95% CIb . Mean . sr . RSDr, % . Mean . sr . RSDr, % . LCLc . UCLd . Drinking water (naturally contaminated) Low 5 1.761 0.26 14.8 1.704 0.028 1.7 −0.057 −0.362 0.248 Medium 5 3.006 0.19 6.3 2.992 0.025 0.8 −0.015 −0.221 0.191 High 5 3.755 0.14 3.7 4.012 0.023 0.6 0.258 0.069 0.446 Matrix (organism) . Contamination level . N . PixeeMoTM . SMEWW 9215B . DOMa . 95% CIb . Mean . sr . RSDr, % . Mean . sr . RSDr, % . LCLc . UCLd . Drinking water (naturally contaminated) Low 5 1.761 0.26 14.8 1.704 0.028 1.7 −0.057 −0.362 0.248 Medium 5 3.006 0.19 6.3 2.992 0.025 0.8 −0.015 −0.221 0.191 High 5 3.755 0.14 3.7 4.012 0.023 0.6 0.258 0.069 0.446 a DOM = Difference of means. b CI = Confidence interval for DOM. c LCL = Lower confidence limit for DOM. d UCL = Upper confidence limit for DOM. Open in new tab Table 2. Method comparison data summary and statistics Matrix (organism) . Contamination level . N . PixeeMoTM . SMEWW 9215B . DOMa . 95% CIb . Mean . sr . RSDr, % . Mean . sr . RSDr, % . LCLc . UCLd . Drinking water (naturally contaminated) Low 5 1.761 0.26 14.8 1.704 0.028 1.7 −0.057 −0.362 0.248 Medium 5 3.006 0.19 6.3 2.992 0.025 0.8 −0.015 −0.221 0.191 High 5 3.755 0.14 3.7 4.012 0.023 0.6 0.258 0.069 0.446 Matrix (organism) . Contamination level . N . PixeeMoTM . SMEWW 9215B . DOMa . 95% CIb . Mean . sr . RSDr, % . Mean . sr . RSDr, % . LCLc . UCLd . Drinking water (naturally contaminated) Low 5 1.761 0.26 14.8 1.704 0.028 1.7 −0.057 −0.362 0.248 Medium 5 3.006 0.19 6.3 2.992 0.025 0.8 −0.015 −0.221 0.191 High 5 3.755 0.14 3.7 4.012 0.023 0.6 0.258 0.069 0.446 a DOM = Difference of means. b CI = Confidence interval for DOM. c LCL = Lower confidence limit for DOM. d UCL = Upper confidence limit for DOM. Open in new tab Product consistency and stability study Methodology.– For product consistency, 3 lots of PixeeMoTM chips (3-month shelf life at 10–30°C) and 3 lots of buffer (6-month shelf life at refrigerated storage of 4–10°C) were examined. Each lot of buffer was tested with the same lot of chips and each lot of chips was tested with the same lot of buffer as in Table 3. Table 3. Product consistency study design Component . Test 1 . Test 2 . Test 3 . Test 4 . Test 5 . Buffer Lot 1 Lot 2 Lot 3 Lot 1 Lot 1 (Lot No.) (KY0419023) (KY0419026) (KY0419022) (KY0419023) (KY0419023) Chips Lot 1 Lot 1 Lot 1 Lot 2 Lot 3 (Lot No.) (C919E07) (C919E07) (C919E07) (C919E10P) (C919E24) Component . Test 1 . Test 2 . Test 3 . Test 4 . Test 5 . Buffer Lot 1 Lot 2 Lot 3 Lot 1 Lot 1 (Lot No.) (KY0419023) (KY0419026) (KY0419022) (KY0419023) (KY0419023) Chips Lot 1 Lot 1 Lot 1 Lot 2 Lot 3 (Lot No.) (C919E07) (C919E07) (C919E07) (C919E10P) (C919E24) Open in new tab Table 3. Product consistency study design Component . Test 1 . Test 2 . Test 3 . Test 4 . Test 5 . Buffer Lot 1 Lot 2 Lot 3 Lot 1 Lot 1 (Lot No.) (KY0419023) (KY0419026) (KY0419022) (KY0419023) (KY0419023) Chips Lot 1 Lot 1 Lot 1 Lot 2 Lot 3 (Lot No.) (C919E07) (C919E07) (C919E07) (C919E10P) (C919E24) Component . Test 1 . Test 2 . Test 3 . Test 4 . Test 5 . Buffer Lot 1 Lot 2 Lot 3 Lot 1 Lot 1 (Lot No.) (KY0419023) (KY0419026) (KY0419022) (KY0419023) (KY0419023) Chips Lot 1 Lot 1 Lot 1 Lot 2 Lot 3 (Lot No.) (C919E07) (C919E07) (C919E07) (C919E10P) (C919E24) Open in new tab For accelerated stability, the chips and buffer were stored at ∼20°C above normal storage temperature and tested over the course of 2 weeks to approximate the normal shelf life. See Table 4. Table 4. Accelerated stability study design Component . Normal shelf life and storage . Accelerated stability temperature, °C . Testing time points, days . Chips 3 months at 10–30°C 45 0, 1, 3, 7, 10, 14 Buffer 6 months at 4–10°C 25 0, 1, 3, 7, 10, 14 Component . Normal shelf life and storage . Accelerated stability temperature, °C . Testing time points, days . Chips 3 months at 10–30°C 45 0, 1, 3, 7, 10, 14 Buffer 6 months at 4–10°C 25 0, 1, 3, 7, 10, 14 Open in new tab Table 4. Accelerated stability study design Component . Normal shelf life and storage . Accelerated stability temperature, °C . Testing time points, days . Chips 3 months at 10–30°C 45 0, 1, 3, 7, 10, 14 Buffer 6 months at 4–10°C 25 0, 1, 3, 7, 10, 14 Component . Normal shelf life and storage . Accelerated stability temperature, °C . Testing time points, days . Chips 3 months at 10–30°C 45 0, 1, 3, 7, 10, 14 Buffer 6 months at 4–10°C 25 0, 1, 3, 7, 10, 14 Open in new tab These studies were carried out with pure culture. Escherichia coli ATCC 25922 was cultured on PCA for 24 h at 35 ± 1°C and the colonies were suspended in PixeeMoTM buffer to a high level (e.g., 103 CFU/mL), medium level (e.g., 102 CFU/mL), and a low level (e.g., 10 CFU/mL) within the quantitative range of the PixeeMoTM method. The levels were adjusted to span the method range as much as possible. Testing was conducted according to the PixeeMoTM package insert for drinking water (Version 1.1). Results Product consistency study results are shown in Figure 3. The mean value of each lot of chips and buffer were plotted. The mean value of each combination was less than three times the standard deviation (sr) calculated from 5-replicate data of Lot 1 of the chips and Lot 1 of the buffer at each concentration level, and consequently no significant differences were observed (9). Figure 3. Open in new tabDownload slide Product consistency study results for chips and buffer. (a) Low level bacterial concentration, (b) Medium level and (c) High level. Figure 3. Open in new tabDownload slide Product consistency study results for chips and buffer. (a) Low level bacterial concentration, (b) Medium level and (c) High level. Accelerated stability results are shown in Figure 4. The time course of the mean value is plotted. The data show no significant time slope and the mean values of all testing time points were less than three times the standard deviation (sr) of calculated from 5-replicate data of the zero time point chips and buffer data. Therefore, no significant differences were observed (9). Figure  4. Open in new tabDownload slide Accelerated stability study results for chips and buffer. (a) Low level bacterial concentration, (b) Medium level and (c) High level. Figure  4. Open in new tabDownload slide Accelerated stability study results for chips and buffer. (a) Low level bacterial concentration, (b) Medium level and (c) High level. Robustness study Methodology.– Three parameters were varied (see Table 5). Each of the parameters is varied only in the direction that could possibly affect the results. During sample preparation (exchange of water for PixeeMoTM buffer), the method performance could only be affected by less centrifugal force, not more, so this parameter is varied only downward. The sample conductivity can affect the method results by interfering with capture of bacteria on the electrodes, but only higher conductivities than recommended could have an effect. The minimum conductivity is 20 µS/cm because of the PixeeMoTM buffer composition. The voltage is preset by the software, but various factors could cause the voltage to decline and only less voltage could possibly affect the results. Table 5. Robustness parameters varied Parameter . Low value . Nominal value . High value . Centrifugation of sample, × g 7000 8000 8000 Sample conductivity, µS/cma 20 20 30 Output voltage, Channel 2, Vppb 18 20 20 Parameter . Low value . Nominal value . High value . Centrifugation of sample, × g 7000 8000 8000 Sample conductivity, µS/cma 20 20 30 Output voltage, Channel 2, Vppb 18 20 20 a µS/cm = Micro siemens/centimeter. b Vpp = Peak-to-peak voltage. Open in new tab Table 5. Robustness parameters varied Parameter . Low value . Nominal value . High value . Centrifugation of sample, × g 7000 8000 8000 Sample conductivity, µS/cma 20 20 30 Output voltage, Channel 2, Vppb 18 20 20 Parameter . Low value . Nominal value . High value . Centrifugation of sample, × g 7000 8000 8000 Sample conductivity, µS/cma 20 20 30 Output voltage, Channel 2, Vppb 18 20 20 a µS/cm = Micro siemens/centimeter. b Vpp = Peak-to-peak voltage. Open in new tab A full factorial design of experiment was employed, incorporating the high and low parameter values (see Table 6). All other parameters remained unchanged. In this case, treatment combination No. 6 includes the recommended parameter values for the method. E. coli ATCC 25922 was cultured on PCA for 24 h at 35 ± 1°C and the colonies were suspended in PixeeMoTM buffer to a high level (e.g., 103 CFU/mL), medium level (e.g., 102 CFU/mL) and a low level (e.g., 10 CFU/mL) within the quantitative range of the PixeeMoTM method. The levels were adjusted to span the range as much as possible. Testing was conducted according to the PixeeMoTM package insert for drinking water (Version 1.1) with the parameter variations as described in Table 5. Standard deviations (sr) at each concentration were compared to determine if any treatment combination or parameter value causes more than a three-fold increase in sr. Table 6. Factorial design for robustness Treatment combination . Sample centrifugation, × g . Sample conductivity, µS/cm . Output voltage, Channel 2, Vppa . 1 7000 20 18 2 7000 20 20 3 7000 30 18 4 7000 30 20 5 8000 20 18 6b 8000 20 20 7 8000 30 18 8 8000 30 20 Treatment combination . Sample centrifugation, × g . Sample conductivity, µS/cm . Output voltage, Channel 2, Vppa . 1 7000 20 18 2 7000 20 20 3 7000 30 18 4 7000 30 20 5 8000 20 18 6b 8000 20 20 7 8000 30 18 8 8000 30 20 a Vpp = Peak-to-peak voltage. b Nominal treatment combination. Open in new tab Table 6. Factorial design for robustness Treatment combination . Sample centrifugation, × g . Sample conductivity, µS/cm . Output voltage, Channel 2, Vppa . 1 7000 20 18 2 7000 20 20 3 7000 30 18 4 7000 30 20 5 8000 20 18 6b 8000 20 20 7 8000 30 18 8 8000 30 20 Treatment combination . Sample centrifugation, × g . Sample conductivity, µS/cm . Output voltage, Channel 2, Vppa . 1 7000 20 18 2 7000 20 20 3 7000 30 18 4 7000 30 20 5 8000 20 18 6b 8000 20 20 7 8000 30 18 8 8000 30 20 a Vpp = Peak-to-peak voltage. b Nominal treatment combination. Open in new tab Results Robustness study results are shown in Table 7. The sr values of all treatment combinations were less than three times the sr of treatment combination 6 (nominal treatment combination) at the high level, medium level, and low level, and consequently no significant differences were observed. Therefore, the PixeeMoTM method is robust for sample centrifugation of 7000–8000 × g, sample conductivity of 20–30 µS/cm, and output voltage of Ch 2 of 18–20Vpp. Table 7. Robustness testing with E. coli ATCC 25922 Treatment combination . Sample centrifugation, × g . Sample conductivity, µS/cm . Output voltage, Ch 2, Vpp . Na . High level: 103 CFU/mL . Medium level: 102 CFU/mL . Low level: 101 CFU/mL . Mean detected value, cells/mL . Srb . Mean detected value, cells/mL . Srb . Mean detected value, cells/mL . Srb . 1 7000 20 18 5 835 135.0 98 10.4 18 13.5 2 7000 20 20 5 884 118.8 172 18.9 73 38.0 3 7000 30 18 5 804 47.2 140 57.0 20 13.7 4 7000 30 20 5 888 112.8 139 23.8 29 14.7 5 8000 20 18 5 835 287.4 100 16.2 27 16.0 6* 8000 20 20 5 986 127.4 111 30.9 25 17.0 7 8000 30 18 5 1149 290.8 115 30.0 35 30.2 8 8000 30 20 5 890 203.7 95 7.9 37 10.4 Treatment combination . Sample centrifugation, × g . Sample conductivity, µS/cm . Output voltage, Ch 2, Vpp . Na . High level: 103 CFU/mL . Medium level: 102 CFU/mL . Low level: 101 CFU/mL . Mean detected value, cells/mL . Srb . Mean detected value, cells/mL . Srb . Mean detected value, cells/mL . Srb . 1 7000 20 18 5 835 135.0 98 10.4 18 13.5 2 7000 20 20 5 884 118.8 172 18.9 73 38.0 3 7000 30 18 5 804 47.2 140 57.0 20 13.7 4 7000 30 20 5 888 112.8 139 23.8 29 14.7 5 8000 20 18 5 835 287.4 100 16.2 27 16.0 6* 8000 20 20 5 986 127.4 111 30.9 25 17.0 7 8000 30 18 5 1149 290.8 115 30.0 35 30.2 8 8000 30 20 5 890 203.7 95 7.9 37 10.4 a N = Number of replicates. b Sr = Standard deviation of replicates. Open in new tab Table 7. Robustness testing with E. coli ATCC 25922 Treatment combination . Sample centrifugation, × g . Sample conductivity, µS/cm . Output voltage, Ch 2, Vpp . Na . High level: 103 CFU/mL . Medium level: 102 CFU/mL . Low level: 101 CFU/mL . Mean detected value, cells/mL . Srb . Mean detected value, cells/mL . Srb . Mean detected value, cells/mL . Srb . 1 7000 20 18 5 835 135.0 98 10.4 18 13.5 2 7000 20 20 5 884 118.8 172 18.9 73 38.0 3 7000 30 18 5 804 47.2 140 57.0 20 13.7 4 7000 30 20 5 888 112.8 139 23.8 29 14.7 5 8000 20 18 5 835 287.4 100 16.2 27 16.0 6* 8000 20 20 5 986 127.4 111 30.9 25 17.0 7 8000 30 18 5 1149 290.8 115 30.0 35 30.2 8 8000 30 20 5 890 203.7 95 7.9 37 10.4 Treatment combination . Sample centrifugation, × g . Sample conductivity, µS/cm . Output voltage, Ch 2, Vpp . Na . High level: 103 CFU/mL . Medium level: 102 CFU/mL . Low level: 101 CFU/mL . Mean detected value, cells/mL . Srb . Mean detected value, cells/mL . Srb . Mean detected value, cells/mL . Srb . 1 7000 20 18 5 835 135.0 98 10.4 18 13.5 2 7000 20 20 5 884 118.8 172 18.9 73 38.0 3 7000 30 18 5 804 47.2 140 57.0 20 13.7 4 7000 30 20 5 888 112.8 139 23.8 29 14.7 5 8000 20 18 5 835 287.4 100 16.2 27 16.0 6* 8000 20 20 5 986 127.4 111 30.9 25 17.0 7 8000 30 18 5 1149 290.8 115 30.0 35 30.2 8 8000 30 20 5 890 203.7 95 7.9 37 10.4 a N = Number of replicates. b Sr = Standard deviation of replicates. Open in new tab Instrument variation study Methodology This study examined three PixeeMoTM instruments (Lot No. M22-002, M22-003, and M22-004) to ensure repeatability of instrument manufacture. E. coli ATCC 25922 was cultured on PCA for 24 h at 35 ± 1°C and the colonies were suspended in PixeeMoTM buffer to a high level (e.g., 103 CFU/mL), medium level (e.g., 102 CFU/mL), and a low level (e.g., 10 CFU/mL) within the quantitative range of the PixeeMoTM method. The levels were adjusted to span the range as much as possible. Testing was conducted on five replicates at each concentration according to the PixeeMoTM package insert for drinking water (Version 1.1) on three PixeeMoTM instruments. Results The 5-replicate raw data for each of three PixeeMoTM instruments were plotted in Figure 5. Table 8 shows the summary of one-way analysis of variance for each concentration. As a result of analysis of variance, the F critical value is larger than the F value, therefore it is concluded that there is no significant difference among the data of three PixeeMoTM instruments for each concentration level. Figure 5. Open in new tabDownload slide Plots of the instrument variation study data. Figure 5. Open in new tabDownload slide Plots of the instrument variation study data. Table 8. Summary of one-way analysis of variance for each concentration . (a) Low level: 10 CFU/mL . (b) Medium level: 102 CFU/mL . (c) High level: 103 CFU/mL . . ANOVA: single factor . ANOVA: single factor . ANOVA: single factor . SUMMARY . Groups . Count . Sum . Average . Variance . Count . Sum . Average . Variance . Count . Sum . Average . Variance . M22-002 5 7.835 1.567 0.015 5 12.382 2.476 0.007 5 16.197 3.239 0.001 M22-003 5 7.739 1.548 0.011 5 12.166 2.433 0.005 5 16.375 3.275 0.007 M22-004 5 6.457 1.291 0.184 5 12.494 2.499 0.006 5 16.305 3.261 0.002 . (a) Low level: 10 CFU/mL . (b) Medium level: 102 CFU/mL . (c) High level: 103 CFU/mL . . ANOVA: single factor . ANOVA: single factor . ANOVA: single factor . SUMMARY . Groups . Count . Sum . Average . Variance . Count . Sum . Average . Variance . Count . Sum . Average . Variance . M22-002 5 7.835 1.567 0.015 5 12.382 2.476 0.007 5 16.197 3.239 0.001 M22-003 5 7.739 1.548 0.011 5 12.166 2.433 0.005 5 16.375 3.275 0.007 M22-004 5 6.457 1.291 0.184 5 12.494 2.499 0.006 5 16.305 3.261 0.002 ANOVA . Source of variation . SS . df . MS . F . P-value . F crit . SS . df . MS . F . P-value . F crit . SS . df . MS . F . P-value . F crit . Between groups 0.237 2 0.118 1.686 0.226 3.885 0.011 2 0.006 0.905 0.431 3.885 0.003 2 0.002 0.478 0.631 3.885 Within groups 0.843 12 0.070 0.074 12 0.006 0.040 12 0.003 Total 1.079 14 0.085 14 0.044 14 ANOVA . Source of variation . SS . df . MS . F . P-value . F crit . SS . df . MS . F . P-value . F crit . SS . df . MS . F . P-value . F crit . Between groups 0.237 2 0.118 1.686 0.226 3.885 0.011 2 0.006 0.905 0.431 3.885 0.003 2 0.002 0.478 0.631 3.885 Within groups 0.843 12 0.070 0.074 12 0.006 0.040 12 0.003 Total 1.079 14 0.085 14 0.044 14 Open in new tab Table 8. Summary of one-way analysis of variance for each concentration . (a) Low level: 10 CFU/mL . (b) Medium level: 102 CFU/mL . (c) High level: 103 CFU/mL . . ANOVA: single factor . ANOVA: single factor . ANOVA: single factor . SUMMARY . Groups . Count . Sum . Average . Variance . Count . Sum . Average . Variance . Count . Sum . Average . Variance . M22-002 5 7.835 1.567 0.015 5 12.382 2.476 0.007 5 16.197 3.239 0.001 M22-003 5 7.739 1.548 0.011 5 12.166 2.433 0.005 5 16.375 3.275 0.007 M22-004 5 6.457 1.291 0.184 5 12.494 2.499 0.006 5 16.305 3.261 0.002 . (a) Low level: 10 CFU/mL . (b) Medium level: 102 CFU/mL . (c) High level: 103 CFU/mL . . ANOVA: single factor . ANOVA: single factor . ANOVA: single factor . SUMMARY . Groups . Count . Sum . Average . Variance . Count . Sum . Average . Variance . Count . Sum . Average . Variance . M22-002 5 7.835 1.567 0.015 5 12.382 2.476 0.007 5 16.197 3.239 0.001 M22-003 5 7.739 1.548 0.011 5 12.166 2.433 0.005 5 16.375 3.275 0.007 M22-004 5 6.457 1.291 0.184 5 12.494 2.499 0.006 5 16.305 3.261 0.002 ANOVA . Source of variation . SS . df . MS . F . P-value . F crit . SS . df . MS . F . P-value . F crit . SS . df . MS . F . P-value . F crit . Between groups 0.237 2 0.118 1.686 0.226 3.885 0.011 2 0.006 0.905 0.431 3.885 0.003 2 0.002 0.478 0.631 3.885 Within groups 0.843 12 0.070 0.074 12 0.006 0.040 12 0.003 Total 1.079 14 0.085 14 0.044 14 ANOVA . Source of variation . SS . df . MS . F . P-value . F crit . SS . df . MS . F . P-value . F crit . SS . df . MS . F . P-value . F crit . Between groups 0.237 2 0.118 1.686 0.226 3.885 0.011 2 0.006 0.905 0.431 3.885 0.003 2 0.002 0.478 0.631 3.885 Within groups 0.843 12 0.070 0.074 12 0.006 0.040 12 0.003 Total 1.079 14 0.085 14 0.044 14 Open in new tab Independent Laboratory Studies The PixeeMoTM method was compared with SMEWW 9215B (2) [hereinafter referred to as “the reference method”]. This comparison study was conducted with five replicates using three contamination levels of naturally contaminated drinking water. Methodology: preparation of drinking water Tap water (2 L) was collected from a faucet (in Suita-City, Osaka, Japan). From the collected water, 300 mL was taken and treated with 0.3 mL 3% sodium thiosulfate solution to neutralize the chlorine according to SMEWW 9060 (1). The neutralized water was stored at room temperature (20–28°C) for 5 days until the bacterial count reached about 5000 to 10 000 CFU/mL. This water was used as Material 3 (high contamination level). The remaining tap water was stored in a refrigerator without neutralization. On the testing day, the reserved water was neutralized with 1.7 mL 3% sodium thiosulfate solution. With this neutralized water, Material 3 was diluted to approximately 500 CFU/mL, which was used as Material 2 (medium contamination level), and to approximately 50 CFU/mL, which was used as Material 1 (low contamination level). Five portions of 25 mL of Material 1, Material 2 and Material 3 were separately transferred into five tubes and tested by the PixeeMoTM method as previously described. Similarly, five portions of 5 mL of each material were transferred into five tubes and tested by the reference method as previously described. Methodology: enumeration of bacteria The test portions were randomized and blind-coded. The bacterial counts of the test portions were measured by the PixeeMoTM method and by the reference method as previously described. The PixeeMoTM method was performed according to the package insert for drinking water (Version 1). SMEWW 9215B—Heterotrophic plate count: pour plate method (the reference method) A 1 mL portion of the test portion was diluted with 9 mL of sterile water to prepare a 1:10 dilution. Next, 1 mL of the 1:10 dilution was diluted with 9 mL of sterile water to prepare a 1:100 dilution. The test portion and the 1:10 and 1:100 dilutions (1 mL each) were separately pipetted into duplicate empty plates. To each plate, 15 mL of PCA tempered at 45°C in a water bath was added and swirled gently to mix. The plates were incubated at 35 ± 1°C for 7 days. After incubation, the colonies on each plate were counted and the number was calculated as CFU/mL in the test portion. Results Figure 6 shows the scatter plot created from the results of the comparison study of bacterial count and Table 9 shows the summary and statistics from the comparison data. Figure 6. Open in new tabDownload slide Scatter plot of the results of the comparison study of bacterial count. Figure 6. Open in new tabDownload slide Scatter plot of the results of the comparison study of bacterial count. Table 9. Summary and statistics from the comparison data Matrix (organism) . Cont. level . N . PixeeMoTM method . SMEWW 9215B . . 95% CIb . Mean . sr . RSDr, % . Mean . sr . RSDr, % . DOMa . LCLc . UCLd . Drinking water (naturally contaminated) Low 5 1.806 0.039 2.2 2.205 0.077 3.5 −0.399 −0.506 −0.292 Medium 5 2.989 0.106 3.5 3.142 0.111 3.5 −0.153 −0.343 0.039 High 5 3.937 0.091 2.3 4.119 0.172 4.2 −0.182 −0.423 0.059 Matrix (organism) . Cont. level . N . PixeeMoTM method . SMEWW 9215B . . 95% CIb . Mean . sr . RSDr, % . Mean . sr . RSDr, % . DOMa . LCLc . UCLd . Drinking water (naturally contaminated) Low 5 1.806 0.039 2.2 2.205 0.077 3.5 −0.399 −0.506 −0.292 Medium 5 2.989 0.106 3.5 3.142 0.111 3.5 −0.153 −0.343 0.039 High 5 3.937 0.091 2.3 4.119 0.172 4.2 −0.182 −0.423 0.059 a DOM = Difference of means. b CI = Confidence interval for DOM. c LCL = Lower confidence limit for DOM. d UCL = Upper confidence limit for DOM. Open in new tab Table 9. Summary and statistics from the comparison data Matrix (organism) . Cont. level . N . PixeeMoTM method . SMEWW 9215B . . 95% CIb . Mean . sr . RSDr, % . Mean . sr . RSDr, % . DOMa . LCLc . UCLd . Drinking water (naturally contaminated) Low 5 1.806 0.039 2.2 2.205 0.077 3.5 −0.399 −0.506 −0.292 Medium 5 2.989 0.106 3.5 3.142 0.111 3.5 −0.153 −0.343 0.039 High 5 3.937 0.091 2.3 4.119 0.172 4.2 −0.182 −0.423 0.059 Matrix (organism) . Cont. level . N . PixeeMoTM method . SMEWW 9215B . . 95% CIb . Mean . sr . RSDr, % . Mean . sr . RSDr, % . DOMa . LCLc . UCLd . Drinking water (naturally contaminated) Low 5 1.806 0.039 2.2 2.205 0.077 3.5 −0.399 −0.506 −0.292 Medium 5 2.989 0.106 3.5 3.142 0.111 3.5 −0.153 −0.343 0.039 High 5 3.937 0.091 2.3 4.119 0.172 4.2 −0.182 −0.423 0.059 a DOM = Difference of means. b CI = Confidence interval for DOM. c LCL = Lower confidence limit for DOM. d UCL = Upper confidence limit for DOM. Open in new tab The scatter plot indicated a high correlation between the two methods. In the statistical analysis, there was no statistically significant difference at the 5% level in the high or medium contamination level. However, there was a slight difference between the two methods in the low contamination level. Discussion The SMEWW 9215B method takes 7 days to acquire results. On the other hand, the PixeeMoTM method can provide results within 1 h because incubation is unnecessary. Obtaining the results on the day of testing is a notable advantage of the PixeeMoTM method. Furthermore, the test procedure using AFI PixeeMoTM is quite simple and easy. The maximum repeatability standard deviation of the PixeeMoTM method was 14.8%. The difference of mean log10 values between the PixeeMoTM and SMEWW 9215B methods ranged from −0.015 to 0.258. Similar results have been obtained in the independent laboratory study. However, there was a slight difference only from the lower confidence limit for the difference of means between the two methods at the low contamination level by the independent laboratory. On the other hand, the scatter plot of the both methods indicated a high correlation between the two methods from the developer and the independent laboratory. Overall, there is no statistically significant difference between the PixeeMoTM and SMEWW 9215B methods. In the product consistency and stability study, chips and buffer within the expiration date are considered to have no change in performance. In the robustness study, three parameters (centrifugation of sample, sample conductivity and output voltage of Channel 2) were varied, it was confirmed that there was no effect within the expected range. As a result of comparing the measurement results of the three PixeeMoTM instruments by analysis of variance, there is no significant difference among the data of three PixeeMoTM instruments. Conclusions The PixeeMoTM method is equivalent to that of the SMEWW 9215B method for determination of aerobic bacteria in drinking water. The product consistency and stability study demonstrated no significant difference within the expiration date. The robustness study confirmed that there was no effect within the expected range. The Instrument variation study also demonstrated no significant difference among the data of three PixeeMoTM instruments. With the PixeeMoTM, the total number of bacteria in drinking waterc can be easily and accurately measured within 1 h. Acknowledgments We are grateful to Sharon Brunelle (AOAC Research Institute) for her support. Submitting Company AFI Corp., 3rd Floor, Med-Pharm Collaboration Building, Kyoto University, 46-29 Yoshida Shimoadachi-cho, Sakyo-ku, Kyoto 606-8501, Japan Independent Laboratory Japan Food Research Laboratories, 52-1 Motoyoyogi-cho, Shibuya-ku, Tokyo 151-0062, Japan Reviewers Yvonne Salfinger Independent Consultant, 2935 Parrish Dr, Tallahassee, FL 32309, USA Wendy McMahon Mérieux Nutrisciences, Silliker Food Science Center, 3600 Eagle Nest Dr, Crete, IL 60417, USA James R. Agin Q Laboratories, 1930 Radcliff Dr, Cincinnati, OH 45204, USA References 1 Standard Method for the Examination of Water and Wastewater ( 2017 ) Method 9060, Samples, American Public Health Association (Washington, D.C.), American Water Works Association (Denver, CO), and Water Environment Federation (Alexandria, VA), https://www.standardmethods.org/doi/10.2105/SMWW.2882.184 (accessed 2019) 2 Standard Method for the Examination of Water and Wastewater ( 2017 ) Method 9215, Heterotrophic Plate Count, Part B—Pour Plate Method, American Public Health Association (Washington, D.C.), American Water Works Association (Denver, CO), and Water Environment Federation (Alexandria, VA), https://www.standardmethods.org/doi/abs/10.2105/SMWW.2882.188, (accessed 2019) 3 National Primary Drinking Water Regulations ( 2016 ) Ground Water and Drinking Water, United States Environmental Protection Agency, https://www.epa.gov/sites/production/files/2016-06/documents/npwdr_complete_table.pdf (accessed 2019) 4 Council Directive 98/83/EC of 3 November 1998 on the quality of water intended for human consumption, Official Journal of the European Communities ( 2019 ) The Council of the European Union, https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:31998L0083&from=EN (accessed 2019) 5 World Health Organization ( 2017 ) Chapter 7 Microbial aspects in Guidelines for Drinking-water Quality, 4th Ed. World Health Organization, https://www.who.int/water_sanitation_health/publications/gdwq4-with-add1-chap7.pdf? ua=1 (accessed 2019) 6 Water Supply in Japan ( 2017 ) 4. Water Quality, Diagram 13 Drinking Water Quality Standards, The Ministry of Health, Labour and Welfare, Japan, https://www.mhlw.go.jp/english/policy/health/water_supply/4.html (accessed 2019) 7 Official Methods of Analysis ( 2019 ), 21st Ed., Appendix J, AOAC INTERNATIONAL, Rockville, MD, http://www.eoma.aoac.org/appendices.asp (accessed 2019) 8 ISO 16140-2:Part 2: 2016 : Protocol for the validation of alternative (proprietary) methods against a reference method, Microbiology of the food chain—Method validation, International Organization for Standardization, Brussels, Belgium 9 AOAC ( 2006 ) Final Report and Executive Summaries from the AOAC International Presidential Task Force on Best Practices in Microbiological Methodology, Appendix F, AOAC INTERNATIONAL, Gaithersburg, MD, https://www.fda.gov/media/73394/download (accessed 2019) © AOAC INTERNATIONAL 2020. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - AFI Corp. PixeeMo™ for Enumeration of Aerobic Bacteria in Drinking Water: AOAC Performance Tested MethodSM 012002
JF - Journal of AOAC INTERNATIONAL
DO - 10.1093/jaoacint/qsaa070
DA - 2020-11-01
UR - https://www.deepdyve.com/lp/oxford-university-press/afi-corp-pixeemo-for-enumeration-of-aerobic-bacteria-in-drinking-water-z8OYbJKEDd
SP - 1610
EP - 1618
VL - 103
IS - 6
DP - DeepDyve
ER -