Jan 31, 1986 - Several methodological variables were critical in two commonly used electron transport activity assays. The dehydrogenase assay based on ...
Vol. 51, No. 5
APPLIED AND ENVIRONMENTAL MICROBIOLOGY, May 1986, p. 931-937
0099-2240/86/050931-07$02.0010 Copyright © 1986, American Society
for Microbiology
Variables Affecting Two Electron Transport System Assays G. ALLEN BURTON, JR.,t* AND GUY R. LANZA Environmental Sciences Program, University of Texas at Dallas, Richardson, Texas 75080-0688 Received 19 August 1985/Accepted 31 January 1986
Several methodological variables were critical in two commonly used electron transport activity assays. The dehydrogenase assay based on triphenyl formazan production exhibited a nonlinear relationship between formazan production (dehydrogenase activity) and sediment dilution, and linear formazan production occurred for 1 h in sediment slurries. Activity decreased w;th increased time of sediment storage at 40C. Extraction efficiencies of formazan from sediment varied with alcohol type; methanol was unsatisfactory. Phosphate buffer (0.06 M) produced higher activity than did either U.S. Environmental Protection Agency reconstituted hard water or Tris buffer sediment diluents. Intracellular formazan crystals were dissolved within minutes when in contact with immersion oil. Greater crystal production (respiration) detected by a tetrazolium salt assay occurred at increased substrate concentrations. Test diluents containing macrophyte exudates produced greater activity than did phosphate buffer, U.S. Environmental Protection Agency water, or ultrapure water diluents. Both assays showed decreases in sediment or bacterial activity through time.
nutrient amendments (10, 30). Tabor and Neihof (29) developed a modified counting technique which eliminates immersion oil contact and membrane filter interference with the detection of small crystals. This method was also recently modified, reportedly decreasing background fluorescence and increasing cell detectability (20). Past studies of microbial activity in aquatic systems have frequently ignored critical methodological variables. In the present study, variables affecting the DHA and INT activity assays of intact sediment consortia and bacteria isolated from sediments were investigated. Our studies were not intended to be a statistical evaluation of the degree of influence of each variable on a test or a direct comparison between the two ETS assays, but rather a qualitative identification of trends in effects for each test variable.
Electron transport system (ETS) activity provides a relative determination of microbial community metabolic activity (14, 22, 33). Various forms of respiration used by aerobic, facultative, and anaerobic organisms share similar ETS reactions (1). While different terminal electron acceptors are used, similar dehydrogenases are part of the preceding transfer steps (16). A commonly used ETS assay involves the use of triphenyltetrazolium chloride (TTC) reduction as a measure of dehydrogenase activity (DHA) and electron transfer capacity, which is proportional to oxygen consumption (17). TTC is reduced by electrons transported by dehydrogenases to and through the electron transport chain(s) common to aerobic and anaerobic respiration (16). Slight modifications to the assay have been reported, including different buffers, substrate concentrations, incubation periods, and volumes (8, 13, 17, 18, 23, 27, 28; G. A. Burton, Jr., Ph.D. dissertation, University of Texas at Dallas, Richardson, 1984). Studies of ETS activity in soils, sludges (10, 28, 30), and sediments (7, 13, 21, 23, 31, 33, 35; Burton, Ph.D. dissertation) have been conducted. Another assay involves the use of the tetrazolium salt 2-iodophenyl-3-phenyl-5-nitrophenyl tetrazolium chloride (INT) (36). Intracellular reduction of INT produces opaque purple formazan crystals. By a count of the bacteria containing crystals, the respiring portion of the community can be differentiated from the dormant, nonrespiring portion. The crystal method has been primarily used in studies of water (3, 36), although a recent test measured sediment activity (19). Variable activity reported in different ETS studies may result from several factors. Droste and Sanchez (11) suggested that variability is due to adverse effects of blending, high dilution, and filter matrix interference with enumeration. Zimmermann et al. (36) alluded to a slight methodological problem whereby immersion oil contact during microscopical counting slightly dissolved the formazan crystals. Factors affecting tetrazolium salt reduction include pH, temperature, incubation period, oxygen concentration, and
MATERIALS AND METHODS Sample collection. The sample site location, characterization, and type tested are given in Table 1. Water collected for sediment assay diluent was placed in sterile, acid-rinsed plastic bottles. Sediments from Texas and Arkansas were taken with a petite ponar dredge (Wildco) and sealed in plastic buckets. Sediments from Colorado were collected by scraping surface material into sterile, acid-rinsed glass jars. All samples were immediately iced until laboratory refrigeration at 4°C. Isolation and culture. Sediments from Colorado sites and Lake Lavon were diluted in 0.06 M phosphate buffer (PB) and spread plated onto casein-peptone-starch (CPS) agar plates (9) for the enumeration of heterotrophic bacteria. Random colonies were picked from these plates, and several gram-negative bacterial strains were isolated and purified by repetitive transfers. Cultures were grown for 24 h on CPS agar plates at room temperature and inoculated into the test diluent at an approximately density of 109 cells per ml. These isolates were tested for ETS activity by a modified method of Zimmermann et al. (36). General methods for assays. Stoppered 125-ml Erlenmeyer flasks containing sediment slurries were used in DHA and formazan crystal production activity (INT) assays. Prior to assay, test sediments were mixed for 3 min with a Fasco 5VA mixer. Assays were conducted with and without gyra-
Corresponding author. t Present address: Biological Sciences Department, Wright State University, Dayton. OH 45435. *
931
932
BURTON AND LANZA
APPL. ENVIRON. MICROBIOL.
TABLE 1. Descriptions of freshwater sample sites Test site
Lavon (Texas)
Features of greater region
Vol
Watershed, agriculture (with grazing and herbicide runoff) 380,000 acre-feet
Elevation (in)
Depth (in)
151
11
122
60
DeGray (Arkansas) Watershed, grazing, woods
13,400 acre-feet
Como (Colorado)
Subalpine terrain
2.3 surface hectares
3,581
3
Blue (Colorado)
Alpine terrain
1.8 surface hectares
3,627
1.8
tory shaking at 150 rpm. At test completion, overlying water in the flask was decanted, minimizing the resuspension of sediment. The remaining sediment was mixed with a magnetic stirring bar, and test aliquots were removed with a sterile pipette or syringe. Dry weight determinations were made concurrently with all assays by drying the sediment for 24 h at 105°C. Diluents. Diluent types tested in slurries to determine their effect on assay results were (i) 0.06 M PB (0.81% Na2HPO4, 0.11% KH2PO4), (ii) U.S. Enviornmental Protection Agency (EPA) reconstituted hard water (2), (iii) Lake Lavon site water, (iv) CPS without agar (9), (v) laboratory aquarium water, (vi) distilled water, (vii) ultrapure water (distilled and passed through a Millipore Q system), (viii) standard methods stock PB (2), and (ix) modified EPA water and PB. Modified EPA water and PB had two aquatic macrophytes, duckweed (Lemmna sp.) and an aquatic fern (Ceratopteris sp.), added for 24 h to add organic exudates to the diluent. All diluents except modified EPA water and PB were sterilized by autoclaving prior to testing. Modified EPA water and PB were filter sterilized. Ionic strengths and osmolarities of test diluents were measured with a model 31 conductivity bridge (Yellow Springs Instrument Co.) and an Osmette model 2007 osmometer (Precision Systems, Inc.), respec-
tively. Activity descriptions. Enzyme activity is normally expressed as a rate measurement, i.e., enzyme units per unit of substrate per unit of time. DHA (enzyme activity) in these studies refers to the analyate produced (formazan) per gram (dry weight) of sediment per hour of incubation time. INT activity is described as the percentage of the total number of cells containing formazan crystals. Variation is calculated as coefficients of variation. Statistical comparisons were made by using the paired Student t test and analysis of variance, with values significant at the 95% level (P s 0.05) (34). DHA was determined by the method of Rous (27), with modifications. Assays were conducted in either duplicate or triplicate sterile 50-ml polypropylene centrifuge tubes, which were purged for I min with nitrogen gas after all ingredients were added. After incubation with TTC at 30°C in the dark, triphenyl-formazan was separated from the assay mixture by alcohol extraction, followed by filtration or centrifugation. Formazan production was assayed spectrophotometrically at 485 nm. Triphenyl-formazan standards were run to allow conversion of A485 values to micrograms of formazan. DHA variables evaluated included (i) sediment dilution, (ii) incubation time, (iii) centrifugation versus filtration, (iv) formazan extraction, (v) slurry diluents, (vi) activity over time, and (vii) sediment age. (i) Sediment dilutions were tested by mixing stored sediment samples for 3 min and by removing portions for immediate dilution with various amounts of PB. Dilutions ranged from 5 to 100% (vol/vol). (ii) A time course of DHA was determined for a Lake Lavon
sediment. Twenty-two replicates were prepared from a mixed sediment, and DHA was determined at seven time intervals (three replicates per time interval). Activity was terminated on triplicate tubes with propanol, and absorbance was determined. (iv) Extracted formazan was separated from sediment either by filtration through a Whatman no. 42 cellulose filter or by centrifugation at 4,424 x g at 4°C for 15 min. (v) Various volumes of methanol (J. T. Baker Chemical Co.) were used to extract formazan from test assays. Propanol (Baker), ethanol (Mallinckrodt, Inc.), and tetrachloroethylene-acetone (2:3) (Baker) were compared with methanol for extraction efficiency. (vi) When slurries were incubated in flasks, various types of overlying waters (diluents such as reservoir site water, EPA water, PB, Tris buffer, and aquarium water) were examined for effects on DHA. (vii) DHA in slurries was determined over time by measuring the activity initially and at various time intervals up to several days. (viii) Aliquots of the samples stored at 4°C were tested at various time periods from time zero to over 1 year. DHA sample variability was measured by collecting 10 replicate ponar dredge samples from one station at Lake Lavon. Each dredge sample was -placed in a separate polyethylene bucket and stored at 4°C. DHA was determined in quadruplicate from each of the 10 mixed-sediment samples (n = 40) within 24 h, and coefficients of variation were calculated. Bacterial respiration (ETS activity) was measured by using a modification of the method of Zimmermann et al. (36). INT was added to test solutions and incubated at ambient temperatures for 30 min, and activity was terminated with Formalin. An aliquot from the test solution was stained for 2 min with 0.04% acridine orange, filtered onto a 0.1-,um irgalan-stained (CIBA-GEIGY Corp.) polycarbonate filter (Nuclepore Corp.), and viewed at a magnification of x 1,500 with a Balpan microscope (Bausch & Lomb, Inc.) with fluorescent attachments (tungsten-halogen lamp, BG-12 exciter filter, and an OG530 barrier filter). Active cells were counted by using UV-stimulated fluorescence and transmitted light to view formazan crystals. A minimum of 200 cells per filter were counted. ETS activity of sediment consortia was measured as follows: 2.0 ml of mixed test sediment slurry was added to a flask with diluent for a total volume of 20 ml, including INT and all amendments. Activity was terminated with Formalin, and sediments were processed for counting by being diluted in PB and homogenized for 1 min before being mounted and stained. Sediment suspensions were mounted by a modified (Burton, Ph.D. dissertation) method of Tabor and Neihof (29). The gelatin-cover slip filter mount was replaced with a 0.1% agar-cover slip mount, and the filter was not peeled away after drying. Bacteria were suspended in molten 0.1% agar, mixed, and spread over 1 cm2 of a glass slide (32). After being dried, the agar was
VOL. 51, 1986
VARIABLES IN TWO ELECTRON TRANSPORT SYSTEM ASSAYS
stained for 10 min with 0.04% acridine orange and destained for 6 min each in 1 M citrate buffers (pHs 6.6, 5, and 4). After destaining, the agar slide was rinsed in distilled water, covered with a cover slip, and examined to determine bacterial counts. Variables investigated for their effects on INT activity in sediments included (i) INT substrate concentration, (ii) diluent type, (iii) activity over time, and (iv) variation in counts.
933
bound red formazan after extraction with methanol. Sequential alcohol extractions of assay sediments were conducted to determine whether adsorption was occurring (Table 2). Reextracting and increasing the volume of methanol used in the extraction process allowed greater recovery of formazan (Table 2). The use of 10 ml of extractant revealed the following significant differences in formazan recovery: tetrachloroethylene-acetone > propanol > ethanol > methanol. PB diluent slurries resulted in higher DHA than did EPA water or Tris (Table 3). Activity in aquarium water was RESULTS greater than or equal to that in PB. Lake Lavon water The relationships noted among formazan production produced similar data to those for PB for the first 48 h. (DHA), sediment dilution, and sediment incubation time are DHA was observed in all test sediments to decrease provided in Fig. 1 and 2. Sediment collected from Lake significantly in sediment slurries with increasing incubation Lavon on 23 March 1983 had peak DHA at a 5% concentratime (Table 3). Usually, a sharp decrease was observed tion, while sediment collected on 21 June 1983 peaked at during the initial few hours, with a leveling off at a low 25% (Fig. 1). DHA was linear over a narrow range of activity rate thereafter. Loss of DHA was also observed with sediment concentrations and varied among sediment samsediments stored at 4°C or frozen (Table 4). ples. Analysis of 10 replicate sediment samples for DHA Data for DHA over time (incubation period) are presented showed that sample variance based on four subsamples per in Fig. 2. DHA was linear for less than 1 h, with a trend replicate was less than 28% (mean = 11%), expressed as the toward slight increases after 2 h; therefore DHA tests were coefficient of variation. The mean activity of several replisubsequently incubated for 1 h. cate samples differed significantly (analysis of variance; a. = Sediments were also tested for DHA by either centrifuging 0.05), with replicate means (± 95% confidence limits) of 6.2 or filtering alcohol-extracted samples. Filtration produced (± 2.7) to 15.4 (+ 1.9) pLg of formazan per g (dry weight). lower values, but the differences were not significant (t test, In INT assays, it was observed that crystal numbers a = 0.05; data not shown). However, the centrifuged samdecreased with oil contact time in some tests. Crystal ples tended to yield higher formazan values, i.e., 13 to 15 dissolution began immediately, with a loss of 56% of the ,ug/g compared with 6 to 12 Vig/g, and had less variation, as detectable crystals within 10 min and an 82% loss in 30 min. shown by coefficients of variation of 8.3 and 26.7%. The Other tests showed a crystal loss of 29% in 5 min, which filter method exhibited coefficients of variation of 61.5 and increased to an 83% loss after 20 min of exposure to oil. 116.7%, which were judged unacceptable. Since the rate of dissolution varied among tests owing to Some DHA tests produced data that were variable and unknown reasons, an adjustment factor could not be applied. inconsistent. On occasion, sediment appeared to contain The literature suggests the use of a final concentration of 0.02% INT for assays of activity. In these tests with bacterial cultures and sediments, increasing the INT levels up to 10-fold resulted in a trend of increased activity which was usually statistically significant. Eight concentration comparisons showed that the elevated INT concentration increased the activity from 5.0 to 33.3% (mean = 16.14%; standard deviation = 10.54%). INT activity of sediments and bacteria isolated from sediments was significantly affected by diluent composition (Table 5). Extremes in ionic concentration, e.g., ultrapure water or concentrated buffer, were inhibitory. PB produced more of a trend toward higher activity of bacterial isolates N 01 c than EPA water did, usually at a statistically significant level 0) (ax = 0.05). Organic compounds in diluents stimulated activ; ity, as seen in tests with aquarium water and EPA water in E c which Leinzna and Ceait'opteris spp. were grown. Tests of 0 U2 CPS showed a lag phase before INT activity reached 100% at c5 24 h (Table 5). Diluents of various ionic strengths and a osmolarities produced different levels of activity (Table 5). A culture (LA7) was suspended in a concentrated sucrose solution (12), concentrated PB, nutrient broth, dilute PB, EPA water, University of Texas at Dallas distilled water, and ultrapure water (Table 5). Sediment slurries varied in their TNT activity over time. In Lavon sediment slurry tests, activity significantly increased 0 o 20 30 40 50 60 70 80 90 100 10 over 24 h from 4.9 to 50.6 and from 0 to 32.1%; however, Sediment (%/c) DeGray sediment both decreased in activity (19 to 0%) and also remained stable over 24 h (data not shown). The activity FIG. 1. Relationship of formazan production (DHA) to Lake of environmental bacterial isolates generally decreased over Lavon sediment dilution (percent sediment). Formazan values are time (Table 5). the standard deviation; n 2. Sediment samples collected on 25 Cultured bacterial isolates were generally lower in variMarch 1983 ( ) and 21 June 1983 (---). =
934
APPL. ENVIRON. MICROBIOL.
BURTON AND LANZA
C l20 W
100
-
c
0
N
80
0
E
60
40 20
1
0
3
2
4
5
6
23
7
Hours FIG. 2. Relationship of formazan production (DHA) (± standard deviation; n
than sediment counts were. Higher INT activity had lower associated coefficients of variation than low INT activity levels did. This inverse relationship was evidenced by a highly significant (a = 0.001) correlation coefficient (0.7) between coefficients of variation and percent activity (based on Table 5 data). DISCUSSION The methodological variables shown to significantly affect DHA data included (i) sediment dilution, (ii) assay incubation period, (iii) separation technique of formazan extracts (i.e., centrifugation or filtration), (iv) sediment slurry diluent type, (v) sediment slurry incubation period, and (vi) preassay sediment storage period. DHA was highly dependent on sediment dilution. Traditionally, data from DHA assays have been standardized (5, 8, 18) by conversion to formazan per gram (dry weight) of soil or sediment. The shortcomings of this standardization ance
TABLE 3. Formazan production (DHA) with various microcosm diluents in Lake Lavon sedimentsa Diluent in test:
Formazan 95% confidence limit (FLg/g)
Propanol Ethanol, absolute Ethanol, 70%
Methanol
10 20
460.3 420.6
10 20
269.9 399.1
10 20
0.0
±
0.0
83.6
±
30.4
±
12.1 44.6
±
60.3
±
28.6
±
10 20
73.3
±
15.2
416.7
±
221.4
Tetrachloroethyleneacetone (2:3)
10 20
894.3 669.0
Water-INTb
20
0.0
Tetrachloroethyleneacetone-INT
10 20
1,333.5 1,286.6
±
30.4 37.1
0.0
129.5 102.7
b
Undiluted sediment. TTC replaced with INT.
Formazan ± 95% confidence limits
~ ~ (1j.g/g)
Aquarium water' PBc EPA water
0.0 0.0 0.0
5,800.0 ± 449.2 3,900.0 ± 359.4 2,500.0 ± 269.5
PB
0.0 4.5
3,290.0 ± 116.7 1,930.0 ± 628.9
EPA water
0.0 4.5
1,125.0 ± 898.5 392.0 ± 449.2
Aquarium water
0.0 4.5
3,417.0 ± 5013.5 560.0 ± 395.3
2
3
PB Lavon waterd PB Lavon water PB Lavon water
0.0 0.0 48.0 48.0 96.0 96.0
222.5 213.3 140.2 144.6 61.4 29.6
± 0.0 ± 159.0 ± ± ± ±
62.9 107.8 7.2 4.5
4 PB
0.0 5.0
1,345.0 ± 251.6 455.0 ± 80.9
EPA water
0.0 5.0
1.070.0 ± 260.6 258.0 ± 9.0
Trise
0.0 5.0
1,179.0 ± 44.9 352.0 ± 161.7
Undiluted sediment.
b From the University of Texas at Dallas aquatic toxicology laboratory.
PB is 0.06 M phosphate buffer (pH 7.5). from the Lake Lavon sediment sample station. ' Tris is 0.05 M Trizma buffer (Sigma Chemical Co.). c
a
time (h)(h) time
1-
±
Vol (ml)
2) to incubation time in Lake Lavon sediment (undiluted).
technique have been ignored; our studies show that this may give rise to misleading results. In our studies, the relationship between sediment dilution and formazan production was not linear in any test. Peak DHA varied among tests, from 5 to 25% sediment dilution. Conversion to micrograms
TABLE 2. Formazan extraction with different solvent types and volumes in Lake Lavon sedimenta Solvent
=
d Collected
VOL. 51, 1986
VARIABLES IN TWO ELECTRON TRANSPORT SYSTEM ASSAYS
(dry weight) also added error, as observed in sample variance tests. Jones and Simon (13) also noted that the relationship of ETS activity to diluted sediment was linear over a small range. Olanczuk-Neyman and Vosjan (21) found for their test sediments that INT reduction and enzyme concentration were linear up to 0.2 g (dry weight). These factors must be recognized as caveats in comparisons among DHA test data. For accurate comparison of data, tests should be precalibrated to determine the sediment dilution at which peak activity occurs. Sediment dry weight and the sediment-formazan production relationship aid in further evaluations of data. In attempts to relate DHA assay date to natural conditions, sediments should not be diluted. Incubation periods of minutes to 16 days for the DHA assay have been reported (10). Ross (27) suggested that short periods were preferred because the DHA is linear only for 2 h and test data were more indicative of in situ conditions; our data support this observation. It is advantageous to restrict the incubation period to a minimum. Extended incubation results in decreased rates and promotes changes in the microbial community. DHA was observed to be very labile, with significant decreases both during tests and during refrigerated storage of sediment. Similar observations have been reported previously (10, 13, 24, 26). Methanol is used frequently in the DHA assay to terminate activity and extract triphenyl-formazan (10). However, we observed an unacceptably large variation when using traditional methods. In some assays, sediment-bound formazan was not completely removed by repeated methanol extractions. Other alcohols were superior to various degrees as extractants. The solvent removing the greatest portion of sediment-bound formazan was tetrachloroethylene-acetone (2:3), followed by propanol, ethanol, and finally methanol. The diluent type used in sediment slurries was also shown to influence DHA. Sediments with PB or Lake Lavon water tended to produce higher levels of DHA than did diluents such as EPA water or Tris buffer. In most assays the coefficient of variation of sediment assays on replicate subsamples was relatively low (
