Using ATP Bioluminescence Technique for Monitoring Microbial Activity in Sludge C. P. Chu,1 D. J. Lee,1 Bea-Ven Chang,2 C. S. Liao2 1
Chemical Engineering Department, National Taiwan University, Taipei 106, Taiwan; telephone: 886-2-2362-5632; fax: 886-2-2362-3040; e-mail:
[email protected] 2 Department of Microbiology, Soochow University, Taipei, Taiwan Received 20 June 2000; Accepted 19 July 2001
Abstract: This study demonstrates, for the ®rst time, that the presence of suspended solids in waste-activated sludge interferes with adenosine triphosphate (ATP) bioluminescence tests. The sludge subject to acid/alkaline treatment represented the test sample. Without consideration of the effect of solid concentrations, one would erroneously estimate the density levels of heterotrophic bacteria in the sludge using ATP data. A light blockage model was proposed to evaluate the luminescence reading without the interference of suspended solids. ã 2001 John Wiley & Sons, Inc. Biotechnol Bioeng 75: 469±474, 2001.
Keywords: ATP; bioluminescence; light blockage; solids
INTRODUCTION Living cells contain adenosine triphosphate (ATP), which serves as the universal energy donor for metabolic reactions. ATP bioluminescence has been proposed since 1960 as an on-line technique for estimating microbial activity in a biological process. Owing to its simplicity and rapid measurement, this technique has been applied successfully in a variety of industrial applications, including breweries, dairy plants, and fruit juice processing. Many types of instruments are available for ATP bioluminescence tests. Griths (1996) and Hawronskyj and Holah (1997) reviewed the components and the measuring principles of these devices. The principle of the ATP bioluminescence technique is as follows. An enzyme±substrate complex, luciferase±luciferin, converts the chemical energy associated with ATP into light by a stoichiometric reaction as follows: ATP Luciferin
Mg2
Luciferin:AMP PPi Oxyluciferin AMP Light
1
Correspondence to: D. J. Lee
ã 2001 John Wiley & Sons, Inc.
The amount of light emitted is thereby proportional to the concentration of ATP present. If all the emitted light could be received by a receiver, the ATP, and hence the microbial activity of the sludge, could be estimated quantitatively using certain correlation. If, on the other hand, part of the emitted light does not arrive at the receiver, the estimated ATP concentration would be underestimated. ATP bioluminescence detects both the microorganisms and the product residue so that it cannot currently provide a detailed indication of the presence (or absence) of speci®c pathogens but only the total level of ATP contamination in the samples (Griths, 1996; Hawronskyj and Holah, 1997; Lasko and Wang, 1996; Tanaka et al., 1997). The ATP bioluminescence technique has been applied to monitor the amount of active biomass in biological wastewater treatment. Patterson et al. (1970) estimated the eects of environmental factors, such as pH value and ionic concentration, on the density levels of viable cell in the activated sludge. Several research groups employed fATP as an index to monitor the viability and growth dynamics of biomass in activated sludge (Chiu et al., 1973; Kao et al., 1973; Weddle and Jenkins, 1971). Jorgensen et al. (1992) compared the bacterial biomass in wastewater measured using the ATP bioluminescence technique and on hydrolysis of ¯uorescein-diacetate (FDA) is dierent physiological states. Kim et al. (1994) investigated the inhibition of chlorine and hydrogen peroxide on ®lamentous, pure cultures and activated sludge by measuring the ATP content, INT-dehydrogenase activity (DHA), and oxygen uptake rate (OUR). A satisfactory linear correlation was obtained for all three tests, although ATP tests showed the lowest sensitivity. Kahru et al. (1996) utilized extracted ATP from activated sludge to characterize the toxicity of ash-heap water. Reliable ATP measurements depend on ecient luminescent light collection from the enzyme±substrate complex. The vast amount of suspended solids appearing in a sludge stream might block out the emitted light and thereby hinder reliable measurement of ATP
concentrations. Existing literature data have often compared the samples subjected to various treatments at the same solids concentration. However, various sludge treatments could markedly alter the solids concentration in the suspension (sedimentation or digestion process). Without taking into account the eect of solids concentration, erroneous conclusions could potentially be drawn from the ATP measurement for characterizing the microbial density level changes in the same sludge subject to various applications, or for comparing ATP readings between different sludges subject to the same treatment. No existing studies, to our knowledge, have investigated the possible interference of suspended particles with regard to accuracy of the ATP test. Most viable microorganisms in activated sludge exist in the aggregate form. In other words, the active biomass that could emit luminescent light behaves as an obstacle to block out the light as well. To avoid the possible interference of the solid phase in measurements, the ATP could be removed from the sample to the solution using certain pretreatment and extraction procedures. After certain solid±liquid separation procedures, and assuming no destruction of ATP during pretreatment/extraction, luminescence measurement could be made on the supernatant without the interference of the solid phase (Holm-Hansen and Booth, 1996). Various pretreatment/extraction procedures have been proposed (Lahdes et al., 1987; Somiya et al., 2000). Van Esbroeck et al. (1984) proposed an automatic ATP monitoring system with sample pretreatment and extraction; however, incorporation of these steps was shown to markedly increase the complexity of the automatic monitoring systems. Moreover, ``optimal'' pretreatment/extraction combination depends on the nature of the system considered and the solution is not universal. Direct measurement on ATP content without the pretreatment/extraction steps, if feasible, is preferable for achieving on-line monitoring and easy control of activated sludge processes (Kucnerowicz and Verstraete, 1979). In this regard, the possible interference of the solid phase in the sludge sample on the luminescence tests must be taken into account. This study demonstrates, for the ®rst time, that the presence of suspended solids interferes with ATP bioluminescence tests. Using the activated sludge subject to acid/alkaline treatment as the testing sample, we propose a method to estimate luminescence reading without interference from suspended solids. EXPERIMENTAL The Sample Waste-activated sludge was obtained from a wastewater treatment plant (Presidential Enterprise Co., Taoyuan, Taiwan). All tests started within 2 after sampling to 470
prevent subsequent sludge changes. The total solids content was 7910 mg/L. An Accupyc Pycnometer 1330 (Micromeritics) measured the true density of dried solids in sludge, giving a result 1453 kg/m3, with a relative deviation of 1 indicate the addition of inert solids (clay powders) to the suspension.
BIOTECHNOLOGY AND BIOENGINEERING, VOL. 75, NO. 4, NOVEMBER 20, 2001
Microbial Enumeration Heterotrophic bacteria were enumerated by using heterotrophic plate count (HPC) and R2A agar (APHA, 1992). The R2A agar contains yeast extract (0.5 g/L), proteose peptone (0.5 g/L), casamino acids (0.5 g/L), glucose (0.5 g/L), soluble starch (0.5 g/L), K2HPO4 (0.3 g/L), MgSO4 á 7H2O(0.05 g/L), sodium pyruvate (0.3 g/ L), and agar (15 g/L). A dilution series (10)3 to 10)5) of each sample was prepared by serial transfer of a 1-mL quantity of sludge. Plates were incubated at 35°C for 5 days (APHA, 1992). All media ingredients were purchased from Difco Laboratories (Detroit, MI). The heterotrophic bacteria density level for the original sludge was measured as 3.9 ´ 107 colony-forming units (CFU)/mL, which is equivalent to 4.9 ´ 109 CFU/g dried solids (DS). RESULTS AND DISCUSSION Effects of pH Value Figure 1 demonstrates the aM versus heterotrophic bacteria density level data (®lled circles and solid line). Heterotrophic bacteria density is expressed in units of CFU/g DS. It is notable that, despite certain data scattering, the trend in ATP luminescence and that in heterotrophic bacteria density level correlated with one another with respect to pH adjustment. Clearly, the bacteria was more resistant to an acidic than to an alkaline environment, which correlates with literature data (Mukherjee and Levine, 1992). The linear correlation in the ®gure is merely for demonstrating the feasibility of using ATP data to represent the microbial
Figure 1. The ATP luminescence measurement (aM) versus heterotrophic bacteria density levels at various pH values at a solids concentration of 7910 mg/L.
activity of activated sludge at the same solid concentration as suggested in the literature (although the regression is far from satisfactory for large data scattering). Effects of Solids Concentration Figure 2 illustrates aM as a function of dilution ratio, with pH value as a parameter. Notably, aM depends upon the dilution ratio, which ®rst increases with dilution ratio in the low-concentration regime, and then decreases after passing through a maximum point. Such an observation clearly reveals the signi®cant role of suspended solids in the luminescence measurements. Each curve in Figure 2 exhibits a maximum point, and some crossovers among the curves are noticeable. Therefore, sludges at dierent microbial density levels may correspond to the same ATP luminescent light readings without considering the role of dilution ratio (or, equivalently, the solids concentration). Correlations must be developed for the aM versus microbial density level relationship at all dilution ratios. The ATP estimate without the interference of suspended solid is thus desirable. For better understanding the role of the solid phase, clay powders were added into the suspension with agitation. The measured aM values are shown with those in the diluted state in Figure 3. (Note: The cases with b > 1 should be the addition of biomass. Nevertheless, the data depicted in Figure 3 with greater-than-unity b-value are used to demonstrate the blockage eects of inert particles only. These data are not employed in the
Figure 2. The ATP luminescence measurement (aM) as a function of dilution ratio, b, with the pH value as a parameter. The data with b > 1 denote the cases with addition of sterilized/inert particles to block out the emitted light.
CHU ET AL.: ATP TECHNIQUE TO MONITOR MICROBIAL ACTIVITY
471
Figure 3. The dimensionless ATP luminescence measurement (aM/ aR) as functions of dilution ratio, b, diluted (®lled symbols) and added with clay powders (open symbols) (pH 7).
subsequent regression analysis.) When more clay powders were added to the sludge, the measured light would be decreased in intensity. Based on this observation, we herein propose that, although the suspended solids represent the aggregates of microorganisms and serve as the source of emitted light, they would block out the emitted light from being received by the receiver. Blockage of Emitted Luminescent Light by Suspended Solids The intrinsic quantity of interest is the ATP concentration per suspended solid mass, which is independent of the solid concentration in the suspension. The measured quantity, however, is the emitted luminescence intensity received by the receiver. We herein propose a simple way to interpret the observed trends noted in Figure 4. Assume that aR is the luminescence reading without the interference of suspended solids. Thus, the reading from the diluted sludge should be baR. In other words, the more suspended solids (bacteria) present in the sludge sample, the greater the amount of light emitted. Such an interpretation certainly could not correlate the entire range of dilution ratios observed in Figure 2. We further assume that the presence of solid phase would block the emitted light from the ATP±Luciferin complex by an amount of aB. Hence, we have: aM aR b
aB
2
where aM is the luminescence intensity measured by the receiver. Apparently, Eq. (2) could be further dierentiated to yield:
472
Figure 4. (a) Schematic drawing of aM versus b curve, which exhibits a maximum point at b =b*. (b) Derivative of aM with b. (c) Shape of the aB curve.
daM aR db
daB db
3
The function for aB is required for evaluating the aR in Eqs. (2) and (3). Little information is available for the blockage function, aB, which should be a complicated function of the interactions between particles in the sludge. Nevertheless, the aM versus b data depicted in Figure 2 illustrate certain characteristics for the functional form of aB. Figure 4a depicts an ``ideal'' aM versus b curve, which exhibits a maximum point at b = b and aM = aM . At the b ® 0 limit, aM would approach 0. At another extreme of b 0, both aM and its derivative (daM /db) would approach 0. (daM /db) thereby exhibits a shape like that depicted in Figure 4b. Accordingly, with the help of Eq. (3), the aB function could be schematically depicted in Figure 4c. Notably, at b = 0, (daB/db) =
BIOTECHNOLOGY AND BIOENGINEERING, VOL. 75, NO. 4, NOVEMBER 20, 2001
Table I. Parameters in Eq. (4) obtained using nonlinear regression. pH
S1(aR) (RLUs)
S0 (RLUs)
w (±)
ATP (lg/L)
3 4 5 7 9 10 11
7480 12,330 22,430 37,510 37,220 21,430 23,330
1152 7663 4447 3379 332.1 4380 377.7
0.0010 0.0094 0.0042 0.0067 0.0002 0.0027 0.00004
7.2 11.9 21.7 36.2 35.9 20.7 22.5
S0 > 0. Furthermore, at b 0, (daB/db) =S1 > S0. Clearly, S1 = aR , because (daM /db)|b 0= 0. One of the simplest relationships for aB satisfying the aforementioned characteristics could be stated: aB
b3 wb
1=S1 b2
w=S0
4
where S1 > S0 w. Three parameters exist in Eq. (4): w, S0, and S1(=aR), which were determined using nonlinear regression analysis. The Marquardt±Levenberg algorithm was adopted to seek parameter values that minimized the sum-of-squared dierences between the experimental data and model outputs. Table I lists the regressed results. The regressed results were veri®ed with the maximum point data (b = b* and aM = aM ) in Figure 2, which are derived from Eqs. (2) and (4) as follows: 2
b
aM
wS1 S0
5
r wS1 S1 S0 S0 S1 b S0 2 2
6
The estimated aM and b* values based on Eqs. (5) and (6) are close to those noted in Figure 2. Use of ATP Measurement to Monitor Microbial Density in Sludge With the estimated S1 values and the equivalence of (1 lg/L ATP) = 1036 RLUs, we could evaluate the ATP concentrations in sludge without the interference of suspended solids. The correlation between estimated ATP concentration per unit solid mass versus numbers of heterotrophic bacteria resembles that depicted in Figure 1, and is not shown here for brevity. The viable microorganisms in activated sludge exist mostly in aggregate form. The application of solid±liquid separation techniques, such as sedimentation or ®ltration, to separate the solid phase from the solution would not only remove the interference by solid phase but also largely eliminate the active biomass from the sludge. In an independent test we noted that the con-
centrations of ATP and HPC for supernatant withdrawn from the sludge were much lower than those for the whole sludge. Therefore, although the proposed scheme requires measurements at dierent dilution ratios and the employment of nonlinear regression, it remains a convenient way to estimate the microbial activity in suspensions containing a high solids concentration (like sludge). An alternative approach would be to estimate the ATP value at a very dilute limit. However, the corresponding ATP reading was relatively low and thus involved a large error. Finally, the removal of sludge particles prior to luminescence measurement, if possible, is clearly the best way to prevent sludge particles. However, if the direct measurement of ATP content in the sludge sample is desired, this work has provided a potential method for interpreting experimental data considering its eects. CONCLUSIONS The ATP bioluminescence technique was applied to monitor the microbial activity in activated sludge streams. This study illustrates that the relatively high suspended solids content in sludge would interfere with the luminescence readings. The emitted light received by the luminometer ®rst increases with the dilution ratio, then decreases after passing through a maximum point. Without considering the eect of suspended solids, ATP measurement could not serve as a unique index for assessment of microbial activity in sludge. We have proposed a light-blockage model to interpret the experimental data, considering that the suspended solids not only represent the source of emitted light, but they also block out emitted light from being received by the luminometer. A regression procedure was adopted to estimate the ATP concentration per suspended solid mass. The estimated ATP concentration per unit solid mass correlated with the number of heterotrophic bacteria. Monitoring the microbial density level in suspensions containing high concentrations of the suspended solids, such as sludge, could be achieved using the ATP bioluminescence technique at dierent dilution ratios.
References Allievi L, Colombi A, Calcaterra E, Ferrari A. 1994. Inactivation of fecal bacteria in sewage sludge by alkaline treatment. Biores Technol 49:25±30. APHA. 1992. Standard methods for the examination of water and wastewater, 18th ed. Washington, DC: American Public Health Association. Chiu SY, Kao IC, Erickson LE, Fan LT. 1973. ATP pools in activated sludge. J Water Pollut Control Fed 45:1746±1758. Gaudy AF Jr, Yang PY, Obayashi AW. 1971. Studies on the total oxidation of activated sludge with and without hydrolytic pretreatment. J Water Pollut Control Fed 43:40±54.
CHU ET AL.: ATP TECHNIQUE TO MONITOR MICROBIAL ACTIVITY
473
Griths MW. 1996. The role of ATP bioluminescence in the food industry: New light on old problems. Food Technol 50:62±72. Hawronskyj JM, Holah J. 1977. ATP: a universal hygiene monitor. Trends Food Sci Technol 8:79±84. Holm-Hansen O, Booth CR. 1996. The measurement of adenosine triphosphate in the ocean and its ecological signi®cance. Limnol Oceanogr 11:510±519. Jean DS, Chang Bea-Ven, Liao CS, Tsou GW, Lee DJ. 2000. Reduction of microbial density level in sewage sludge through pH adjustment and ultrasonic treatment. Water Sci Technol 42(9): 97± 102. Jorgensen PE, Eriksen T, Jensen BK. 1992. Estimation of viable biomass in wastewater and activated sludge by determination of ATP, oxygen utilization rate and FDA hydrolysis. Wat Res 26: 1495±1501. Kahru A, Kurvet M, Kulm I. 1996. Toxicity of phenolic wastewater to luminescent bacteria photobacterium phosphoreum and activated sludge. Water Sci Technol 33:139±146. Kao IC, Chiu SY, Fan LT, Erickson, LE. 1973. ATP pools in pure and mixed cultures. J Water Pollut Control Fed 45:927±931. Kim CW, Koopman B, Bitton G. 1994. Int-dehydrogenase activity test for assessing chlorine and hydrogen peroxide inhibition of ®lamentous pure cultures and activated sludge. Water Res 28: 1117±1121. Kucnerowicz F, Verstraete W. 1979. Direct measurement of microbial ATP in activated sludge samples. J Chem Technol Biotechnol 29:707±712.
474
Lahdes E, Leppanen JM, Makinen I. 1987. Comparison of extraction methods in the determination of adenosine triphosphate (ATP) in phytoplankton and sestion. Aqua Fennica 17:43±49. Lasko DR, Wang IC. 1996. On-line monitoring of intracellular ATP concentration in Escherichia coli fermentations. Biotechnol Bioeng 52:364±372. Mukherjee SR, Levine AD. 1992. Chemical solubilization of particulate organics as a pretreatment approach. Water Sci Technol 26:2289±2292. Patterson JW, Brezonik PL, Putnam HD. 1970. Measurement and signi®cance of adenosine triphosphate in activated sludge. Envir Sci Technol 4:569±575. Somiya I, Fujii S, Kishimoto N, Kim RH. 2000. Development of ATP assay as a surrogate indicator of viability of Cryptosporidium parum oocysts. Water Sci Tech 41:181±188. Tanaka H, Shinji IT, Sawada K, Monji Y, Seto S, Yajima M, Yagi O. 1997. Development and application of a bioluminescence ATP assay method for rapid detection of coliform bacteria. Water Res 31:1913±1918. Van Esbroeck H, Schram E, Vereecken J. 1984. An automatic ATP monitor for activated sludge characteristics. J Chem Technol Biotechnol 34B:76±86. Weddle CL, Jenkins D. 1971. The viability and activity of activated sludge. Water Res 5:621±640. Woodard SE, Wukasch RF. 1994. A hydrolysis/thickening/®ltration process for the treatment of waste activated sludge. Water Sci Technol 30:29±38.
BIOTECHNOLOGY AND BIOENGINEERING, VOL. 75, NO. 4, NOVEMBER 20, 2001
