Apr 17, 1992 - incorporation into DNA was independent of thymidine concentration. Elevated ... 20 times faster than incorporation of label into DNA. We found ...
APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Nov. 1992, p. 3614-3621 0099-2240/92/113614-08$02.00/0 Copyright © 1992, American Society for Microbiology
Vol. 58, No. 11
Assessment of [3H]Thymidine Incorporation into DNA as a Method To Determine Bacterial Productivity in Stream Bed Sediments LOUIS A. KAPLAN,1,2* THOMAS L. BOTT,1'2 AND JOHN K. BIELICKI1 Stroud Water Research Center, Academy of Natural Sciences of Philadelphia, 512 Spencer Road, Avondale, Pennsylvania 19311-9516, 1* and Department of Biology, University of Pennsylvania, Philadelphia, Pennsylvania 19104 Received 17 April 1992/Accepted 21 August 1992
We performed several checks on the underlying assumptions and procedures of the thymidine technique applied to stream bed sediments. Bacterial production rates were not altered when sediments were mixed to form a slurry. Incubation temperature did affect production rates. Controls fixed and washed with formaldehyde had lower backgrounds than trichloroacetic acid controls. DNA extraction by base hydrolysis was incomplete and variable at 25°C, but hydrolysis at 120°C extracted 100%l of the DNA, of which 84% was recovered upon precipitation. Production rates increased as thymidine concentrations were increased over 3 orders of magnitude (30 nM to 53 FM thymidine). However, over narrower concentration ranges, thymidine incorporation into DNA was independent of thymidine concentration. Elevated exogenous thymidine concentrations did not eliminate de novo synthesis. Transport of thymidine into bacterial cells occurred at least 5 to 20 times faster than incorporation of label into DNA. We found good agreement between production rates of bacterial cultures based upon increases in cell numbers and estimates based upon thymidine incorporation and amount of DNA per cell. Those comparisons emphasized the importance of isotopic dilution measurements and validated the use of the reciprocal plot technique for estimating isotopic dilution. Nevertheless, the thymidine technique cannot be considered a routine assay and the inability to measure the cellular DNA content in benthic communities restricts the accuracy of the method in those habitats.
slowly lifted the base and collar out of the water, and decanted the overlying water off the sediments. Plastic rings (11 mm [inside diameter] by 4 mm [depth]) were used to delimit 0.4-cm3 cores of sediments, and these subsamples were transferred intact to caps of the same diameter. We prepared sediment slurries by scraping sediments to the same depth as the cores, collecting these scrapings in a beaker, and mixing them without dilution. We subsampled the slurry with the plastic rings used for cores. Thymidine solutions (100 ,ul) of different specific activities were added with an Eppendorf pipette directly to the surface of cores. Coring resulted in a slight loss of interstitial water, so the thymidine solutions readily penetrated the cores. After isotope addition, caps containing the cores were lowered with forceps into 10 ml of stream water in sterile Oak Ridge centrifuge tubes for incubation, and the tubes were not vortexed. Slurries were added directly to tubes containing stream water and thymidine, and the tubes were vortexed gently. The cores were incubated immediately at in situ temperatures, whereas slurries were either incubated immediately at in situ temperatures, held at in situ temperatures for 1 h prior to isotope addition and incubated at in situ temperatures, or incubated immediately at an elevated tem-
In recent years, the production rates of aquatic heterotrophic bacteria have been estimated from measurements of [3H]thymidine incorporation. The thymidine procedure has enjoyed widespread acceptance because it purports to measure heterotrophic bacterial production to the exclusion of other organisms. Critical assumptions of the thymidine technique are as follows: (i) the rate of thymidine incorporation into DNA is independent of concentration, (ii) the rate of thymidine incorporation into DNA is not limited by the rate of thymidine transport into cells, (iii) dilution of radiolabeled thymidine can be estimated or eliminated, and (iv) thymidine incorporation rates can be accurately converted to rates of bacterial cell production (35). Our study addressed each of these assumptions as well as several checks of the [3H]thymidine methodology when applied to stream bed sediments. Checks included sample handling prior to incubation, incubation temperature and duration, adsorption of thymidine and reduction of background associated with unincorporated radioactivity, hydrolysis conditions for the extraction and recovery of DNA, determination of isotopic dilution, and conversion of thymidine incorporation rates into bacterial production rates.
MATERIALS AND METHODS Effects of sampling, physical disturbance, and temperature on bacterial community production in stream bed sediments. Surface sediments were removed from the stream bed with minimal disturbance. We slid a 2-mm-thick plastic rectangular base under the sediments to a depth of approximately 15 mm, placed a rectangular collar on top of the sediments, *
perature. Sediment particle size distributions were determined by wet sieving a preweighed subsample through a series of 12 sieves (28). Dry mass and ash-free dry mass of the sediment samples were measured after drying at 60°C and combustion at 550°C, respectively. Chlorophyll a was extracted with 90% acetone and quantified by spectrophotometric analysis (30). Epifluorescence microscopic counts of total and active bacteria were made with 4',6-diamidino-2-phenylindole (DAPI) (45) and rhodamine 123 (33) as the respective fluo-
Corresponding author. 3614
VOL. 58, 1992
[3H]THYMIDINE METHOD APPLIED TO STREAM BED SEDIMENTS
rochromes. The rhodamine 123 staining was performed prior to formalin fixation, and then bacteria were separated from sediments by sonication (53) followed by centrifugation in glycerol (5). [3H]thymidine incorporation into DNA. Stream water sampled with the sediments was filtered (glass fiber Gelman GF/C), dispensed into sterile Oak Ridge centrifuge tubes, and held in a water bath. Sediments and 20 ,uCi of [methyl3H]thymidine (2.22 to 3.33 TBq/mmol) amended with 0 to 550 nmol of unlabeled thymidine were added to the incubation tubes. Incubation, typically for 0.5 to 0.75 h, was stopped by the addition of formalin (0.5 ml). Preliminary experiments were performed to determine the time course of thymidine incorporation. Formalin was added to controls 0.5 h prior to the addition of isotope. Unincorporated thymidine was removed from the formalin-fixed samples by centrifugation (27,000 x g, 10 min) and three washes in 10.0 ml of 5.0% formaldehyde. To hydrolyze RNA, remove low-molecularweight intracellular compounds, and solubilize DNA and protein, 5.0 ml of a solution of 25 mM EDTA and 0.1% sodium dodecyl sulfate (SDS) in 0.6 N NaOH (16) was added and sediments were autoclaved for 0.5 h (37). Samples were cooled rapidly and centrifuged (2,500 x g, 6 min), and 4.0 ml of the supematant fluid was transferred to sterile 15-ml Corex tubes. DNA was precipitated in the Corex tubes after the addition of 100 ,J of a 5.0-mg/ml concentration of herring sperm DNA, 100 Fl of a 5.0-mg/ml concentration of thymidine, 500 ,ul of humus extract, and 150 ,ul of concentrated HCI in 5% trichloroacetic acid. Humus extract was prepared by autoclaving forest floor leaf litter and soil in 0.6 M NaOH for 0.5 h and neutralizing it (37). Samples were cooled rapidly and then centrifuged (3,090 x g, 10 min). DNA was separated from protein in the precipitate by hydrolysis in 2.0 ml of 5% trichloroacetic acid heated to 100°C for 0.5 h. The tubes were centrifuged (4,000 x g, 10 min), and a 1.0-ml aliquot was removed and placed into scintillation fluor prior to counting in a Beckman 3801 liquid scintillation counter. DNA recovery was determined from the addition of Escherichia coli (thymidine-[methyl-14C]DNA; New England Nuclear) to the base hydrolysis mixture. DNA recovery data and isotopic dilution were used to calculate the quantity of thymidine incorporated into DNA. Isotopic dilution was determined by the reciprocal plot method (18). Measurements of thymidine incorporation in culture experiments, described below, were handled similarly. Measurement of total thymidine uptake. Thymidine incorporation into low-molecular-weight pools and all macromolecules was determined from the base hydrolysate described above. Radioactivity in a 100-,ul subsample of this cell extract was measured as described above. Optimization of the base hydrolysis procedure. The temperature, incubation time, and normality of NaOH were varied to ascertain the optimal conditions for extraction and recovery of DNA from stream bed sediments. The temperatures used included 25, 40, 55, and 120°C, and the concentrations of NaOH included 0.3 and 0.6 N. Incubation time for the 120°C treatments (autoclave) was 0.5 h, and for all other temperatures the incubation time was 12 h. The efficacy of DNA extraction was determined by repeating the hydrolysis step two or three times. Adsorption of thymidine to sediments. Formalin-killed controls provided a measure of the thymidine adsorbed after washes, hydrolysis, and precipitation. To assess whether adsorption during incubation significantly reduced thymidine available for uptake, sterile sediments were tested with
3615
[methyl-3H]thymidine. Stream bed sediments (50 g) were autoclaved in stream water for 1 h and cooled, and subsamples were transferred to Oak Ridge centrifuge tubes containing 10.0 ml of filter-sterilized autoclaved stream water and 10 ,uCi of [3H]thymidine. The samples were vortexed, incubated for 0.5 h at room temperature, and then centrifuged (15,000 x g, 10 min). Radioactivity of [3H]thymidine in the supernatant fluid was determined as described above. Culture studies. Separate estimates of bacterial production rates based on increases in cell numbers and thymidine incorporation into DNA were derived from batch cultures of Bacillus megatenum (ATCC 14581) and from continuousflow cultures of White Clay Creek (WCC) isolates. B. megatenum was grown on 0.1% tryptone-0.1% yeast extract (TYE) in Nephelo flasks kept at 20°C with a water bath shaker. Population growth was monitored by determining optical density at 665 nm. During log phase, subsamples of culture were removed from the flasks for the measurement of [3H]thymidine incorporation, cell size, cell number, and amount of DNA per cell. We estimated cell densities from epifluorescent microscopic counts (22). Cell dimensions were determined under phase-contrast microscopy with an eyepiece reticle calibrated with a stage micrometer. The formula for a prolate spheroid was used to calculate volume. The specific growth rate, k, was calculated from cell densities as k = (ln N/NO)It, where No is the initial density and N is the density at time t (11). The product of the specific growth rate and the initial population density is the production rate (kNo). Subsamples of culture were incubated in thymidine concentrations ranging from 1 to 216 ,uM. Rates of thymidine incorporation into B. megaterium, corrected by the reciprocal plot method, were converted into cell production rates by using a theoretical conversion factor, N/mol T = [318 x (1Ip)]IW, where N is the number of cells produced, mol T is the moles of thymidine incorporated into DNA, 318 is the average molecular weight of the four nucleotides in DNA, p is the proportion of thymine nucleotides in DNA, and W is the weight of DNA per cell (35). In the case of B. megaterium, G+C is between 36 and 38 mol% (7), so A+T is 63 mol% and P is 0.315. The amount of DNA per cell was measured in extracts of the B. megatenum culture by spectrofluorometric detection with bisbenzimide (Hoechst product no. 33258). Subsamples of culture frozen after collection were thawed, transferred to Corex centrifuge tubes, washed twice in sterile phosphate buffer (0.1 M, pH 7), resuspended in 5.0 ml of SSC buffer (0.154 M NaCl, 0.015 M trisodium citrate [pH 7]), autoclaved for 20 min, cooled, and sonicated in an ice-cooled tube for 7.5 min (6.5 min at 50 W and 1.0 min at 75 W). The sonicated material was centrifuged (20,200 x g, 10 min), and the supernatant fluid was removed for analysis. To measure DNA concentrations in the extracts and standards, 0.4 ml of extract was mixed with 2.0 ml of phosphate-buffered saline high-salts buffer (2.0 M NaCl in 0.05 M phosphate buffer, pH 7.4) and 0.1 ml of bisbenzimide (20 ,ug of Hoechst 33258 per ml of H20), the mixture was incubated for 1.5 h at room temperature, and the fluorescence was measured (excitation, 356 nm; emission, 458 nm; slit width, 4 nm; model 430 spectrofluorometer, G. K. Turner Associates). Continuous-flow stream water cultures were prepared from sediments, a bacterial isolate, and stream water from WCC. The isolate was a gram-negative rod capable of incorporating thymidine and attaining a generation time of 0.75 h in 0.1% TYE at 25°C. Stream bed sediments were wet sieved (355-,um mesh), and 155 g (wet weight) was placed
3616
APPL. ENVIRON. MICROBIOL.
KAPLAN ET AL.
TABLE 1. Influence of sediment handling on bacterial production and other sediment parameters' Date
Treat-
ment
February 1986C Core Slurry May 1986f
Core Slurry
Production (107 cells g [dry wt] of
sediment-' h-') 8.81 + 3.99d 8.42 t 3.75d 37.0 ± 15.3g 10.7 t 1.29
Total cells (109 cells g [dry wt] of
(108 sediment-' h-1) 1.90 t 0.97 (21.6)e 4.23 t 1.16 (4.8)e 1.81 ± 0.37 (21.5)e 5.05 t 0.24 (6.0)e
sediment-')
4.69 ± 3.16 (12.7)e 2.29 t 0.49 (21.4)e
Active cells
cells g [dry wt] of
Sed.ent organic aChlorophyll (L.g/g [dry content (%) wt])
Median grain size Skewnessb
(A.m)
5.1 t 1.4 4.1 t 0.8
-
340 300
0.84 0.97
8.1 t 2.2 4.3 ± 0.6
81.3 ± 58.1 32.0 ± 5.4
620 660
1.39 0.73
a Data are expressed as x + SD. -, no data. Skewness is a measure of the asymmetry of the particle size distribution curve, expressed as quartile (Q) measures: [(Ql)(Q3)/(Q2)21. c Values of n for each parameter are as follows: production, 9; total cells, 10; active cells, 5; organic content, 17. d No significant difference in production rates; t-test P > 0.82. e Generation time (in hours), shown in parentheses, is calculated as the quotient of cell density (dividend) and production rate (divisor). f Values of n for each parameter are as follows: production, 12; total cells, 10; organic content, 15; chlorophyll a, 5. g Production rates are significantly different; t-test P > 0.00. b
into a 4-liter Erlenmeyer flask modified with an overflow tube 150 mm from the bottom. We added enough stream water to cover the sediments, stoppered the flask with cotton, and autoclaved it for four successive 1-h periods separated by 24 h at room temperature. The sterility of the sediments and later the purity of the culture were checked with spread plates on TYE agar incubated at room temperature. The bacterial isolate was grown in nutrient broth. A turbid culture (5.0 ml) was used to inoculate the sterile sediments and 2.4 liters of autoclaved stream water. After an initial 24-h colonization period, a 20-liter reservoir containing sterile stream water was connected to the flask via a BBL cam pump operated at a flow rate of 1.2 liters h-1. Over a 14-day period, the flask was agitated slowly on a reciprocal shaker and the stream water volume was turned over every 2 h. Sediments in the flask were sampled aseptically through the top opening by using a long glass tube as a coring device. The shaker was turned off, and the sediments were allowed to settle for 2 h prior to sampling. After 14 days, a step-up experiment was initiated. The flow rate of stream water through the system was doubled and amended with 50 ,ug of C per liter of TYE. Two days later, the sediments were sampled for bacterial numbers, cell size, and thymidine incorporation as described above except that cell size was determined from photomicrographs of DAPI-stained cells. The sampling was repeated 10.5 h later. RESULTS Influence of sediment handling prior to and during incubation. In general, the bacterial production rates measured with slurries were similar to the rates estimated from intact cores. The data for February 1986 in Table 1 illustrate this point. The values for cell densities, sediment organic matter, median grain size, and skewness also were similar for the two treatments, observations which confirm that the slurry accurately represented the composition of the cores. The second set of data in Table 1 (May 1986) illustrates the potential problem of preparing slurries so that they are representative of the intact sediments, especially when a discontinuity such as an enriched surface layer exists. In this particular case, only the values for median grain size agreed between treatments, while all other parameters indicated an unequal distribution of the microbiota and sediment organic matter. Incubation temperature had a pronounced effect on production rates. There were no significant differences between
rates for a slurry incubated without delay at the in situ temperature and a slurry held for 1.0 h at that temperature before incubation. However, a significant 2.5-fold increase occurred when the incubation temperature was increased from 12.5 to 26.0°C (Tukey's Studentized Range Test, a = 0.05). Data from two different experiments with incubation periods ranging from 0.083 to 3.0 h and temperatures ranging from 5 to 20°C showed a broad linear range for thymidine uptake by the stream bed community. Assessment of hydrolysis conditions. Base hydrolysis at 25°C yielded incomplete and variable extraction on three different occasions, with efficiencies ranging from 83% + 12% to 56% + 3% (ic ± standard deviation [SD], n = 5). However, hydrolysis at 120°C extracted virtually all of the DNA, 99% + 0% (n = 6) and 100% + 0% (n = 10) (x + SD), with little variation and in a fraction of the time. The temperature of the base hydrolysis markedly affected the ability to recover DNA by precipitation. DNA recovery equaled 93% +± 3% at 25°C and continued to decline with increasing temperature to a low of 32% + 10% at 55°C before increasing dramatically to 86% + 4% at 1200C (.1 + SD, n = 3). Recovery of [14C]DNA was improved by the use of humus extract (37), increasing from 39% ± 11% (n = 3) to 77% + 2% (n = 4) in a single test (C + SD). Adsorption of thymidine to sediments. Only 3% + 2% (x ± SD, n = 5) of added thymidine was adsorbed by autoclaved sediments, indicating that most of the thymidine incubated with the stream bed sediments would be available for uptake by the bacterial community. With formalin-fixed control samples that were carried through the entire DNA isolation procedure, a 20-min desorption step, which consisted of holding the samples in the 10-ml formaldehyde washes prior to centrifugation, reduced the background by as much as 3 orders of magnitude. Formaldehyde washes of the sediments were also found to be more effective than washes with cold 5% trichloroacetic acid in reducing the background. In a direct comparison with stream bed sediments, three 10-ml formaldehyde washes followed by hydrolysis and precipitation resulted in a background of 378 + 65 dpm (n = 6), while substituting cold trichloroacetic acid for the washes gave a background of 3,071 ± 667 dpm (n = 5). The amount of [3H]thymidine adsorbed in killed controls was independent of the specific activity. When the amount of labeled thymidine was kept at a constant 20 ,uCi but the specific activity ranged over 3 orders of magnitude from 73 to 0.038 p,Ci/nmol, radioactivity in the formalin controls ranged from 321 to 1,738 dpm/g (dry weight) with no relationship between the two parameters.
VOL. 58, 1992
[3H]THYMIDINE METHOD APPLIED TO STREAM BED SEDIMENTS a
1000
3617
200 1
.
150.
x 100-
100 50-
>m
0
0
o
-20
20
40
60
2
3
4
0
E
-9;
10
b
200-
&150o0
40 E
100-
a
50
0.01
10
0.1
30-
m 2010-
100
b
S(uM) 40
LIU
v
FIG. 1. Influence of substrate concentration on velocity of [3H]thymidine uptake (@) and incorporation into DNA (0). Datum points are the means of five observations.
Vv
S (pM) C
60
0
E 4
The thymidine technique-incorporation of label into DNA. We examined two critical assumptions of the thymidine technique: (i) the uptake of thymidine into cells occurs at a rate much faster than its incorporation into DNA, and (ii) the incorporation into DNA is independent of thymidine concentration. The radioactivity of the base hydrolysate presumably represents net uptake into the cells; gross uptake would include label excreted as [3H]H20. The data in Fig. 1 show that when slurries of WCC stream bed sediments were exposed to thymidine concentrations ranging from 0.031 to 53 ,uM, the velocity of uptake into the cells was at least 7 to 20 times greater than the rates of incorporation into DNA. This experiment was repeated, and again rates of uptake were at least 5 to 14 times greater than incorporation rates. The specific activities of thymidine pools for both rates were assumed to be equivalent to the specific activity of the incubation water. No isotopic dilution by de novo synthesis or exogenous thymidine was included. When the thymidine incorporation data from Fig. 1 were expressed in reciprocal plot form, three linear ranges were observed (Fig. 2). These results are representative of results we obtained with stream bed sediments on two different occasions. Increasing the concentration of unlabeled thymidine present during incubation did not eliminate de novo synthesis. Plots of the reciprocals of radioactivity in DNA versus thymidine concentration did not pass through the origin. The three linear ranges represent three estimates of isotopic dilution (Fig. 2a, 28.84 ,uM; 2b, 0.72 ,uM; 2c, 0.08 ,uM). Moreover, when the rates of DNA incorporation were corrected with the estimates of isotopic dilution, it was clear that concentration influenced the rates of incorporation into DNA, but that within specific ranges of thymidine concentration, the rates were independent of concentration (Fig. 3). Similar results also were obtained numerous times with stream bed sediments when unlabeled thymidine concentrations in the narrower range of 0.029 to 3.33 ,uM were used. The data presented in Fig. 4 are representative of this pattern. The six points on the reciprocal plot were subjected to regression analyses consisting of comparisons of r2 values from all possible combinations of adjacent points. The best fit was achieved by separating the six points into two linear regions of three points each (?2 = 0.999), but this was not a statistically significant (a = 0.05) improvement over one line
Z 2-
0.1
-0.1 S
0.2
0.3
(OM)
FIG. 2. Influence of thymidine concentrations spanning 3 orders of magnitude on the determination of isotopic dilution by the reciprocal plot technique. Plots on the right are linear regressions, and the letters and symbols correspond to regions of the plot on the left.
(P = 0.10). The two linear regions resulted in two estimates of isotopic dilution that differed by a factor of 5.5 (0.146 and 0.797 ,M). The productivity estimates calculated from the estimates of isotopic dilution increased 2.4 fold from 7.23 x 107 to 1.77 x 108 cells g (dry weight) of sediment-' h-1 at thymidine concentrations above 0.5 FM. Validation of the reciprocal plot method for the estimation of isotopic dilution. In batch cultures of B. megatenum, concentrations of added unlabeled thymidine ranged from 180 . 0
150
120 m
90' E
-l
.-,
6030 *
0.01
.
0.1
10
100
S (MM)
FIG. 3. Velocity of [3H]thymidine incorporation into DNA from Fig. 1 corrected for isotopic dilution.
3618
APPL. ENvIRON. MICROBIOL.
KAPLAN ET AL. 1201
the amount of DNA per cell was calculated from measurements of cell size and assumptions that the amount of DNA per cell is proportional to cell size and equal to 3.5% of the dry weight, that bacterial density is 1.1 x 10-12 g/,um3, and that the cells are 80% water (6, 32). A value of 0.25 was used
for the proportion of thymine nucleotides in DNA. Values used in the calculation of k and the amount of DNA per cell are shown in Table 2. Sediments in stream water culture had three linear regions and therefore three estimates of isotopic dilution and production. The isotopic dilution value from the middle range provided the best agreement with the estimate from cell densities.
E
'O
60-
40-
00 20-
DISCUSSION 0
-1.0
-0.5
0.0
0.5
1.5
1.0 S
2.0
2.5
3.0
3.5
(MM)
FIG. 4. Influence of thymidine concentrations within a narrow concentration range on the estimation of isotopic dilution (0.146 puM [-] and 0.797 ,uM [0]). Datum points are the means of five observations.
1.68 to 512 ,uM and resulted in two distinct linear regions in the reciprocal plot, giving two estimates of isotopic dilution and thymidine incorporation (Table 2). For the B. megaterium batch cultures, estimates of production rates derived by using the higher levels of isotopic dilution were in better agreement with the estimates from cell densities than estimates derived by the lower levels of isotopic dilution (Table 2). Increases in the numbers of bacteria attached to sediments in the continuous-flow stream water culture also were used to evaluate the reciprocal plot method and production estimates from thymidine incorporation into DNA. Spread plates of the culture system showed that the initially sterile sediments were eventually contaminated and colonized by two species of bacteria in addition to the original inoculum. Each species tested positive for thymidine incorporation, and we continued working with the system as a mixed culture. The presence of procaryotes unable to grow on TYE plates or eucaryotes was not determined. Calculations of production rates from the sediment data were handled as described for B. megatenum above, except that the amount of DNA per cell was not measured directly. An estimate of
Rates of thymidine incorporation into bacterial DNA in marine sediment communities have been found to change little in response to short-term disturbance (12, 38, 39). This is advantageous because sediment slurries can reduce variability by homogenizing microhabitat patchiness (36) and help mix the radioactive precursor (39). Our data demonstrate that for aerobic stream bed sediment communities growing at 12.5 to 17°C, rates of thymidine incorporation were not affected for up to 1 h after mixing. We also found that the variance terms for many of the biological parameters in stream bed sediments were reduced with slurries (Table 1). However, preparing slurries which are representative of the intact sediments is not always straightforward. We suggest that the measurement of organic matter content or bacterial densities in an intact core of the same sediments used to prepare the slurry may be a simple and effective quality control check on the slurry preparation. Given the effect of temperature on enzymatic reaction rates, it is not surprising that temperature is usually carefully controlled during activity measurements. Thymidine incorporation into DNA is temperature sensitive. This has been demonstrated with lacustrine sediment communities (52) and bacterioplankton from a dimictic eutrophic lake (31). Our data showed a statistically significant Q10 of approximately 2 for stream bed sediment communities and underscore the importance of maintaining in situ temperatures during sediment handling. The extraction of DNA from sediments by using a base has been tried with several variations of normality, temperature, and time. These include 0.3 N NaOH for 16 h at 37°C
TABLE 2. Comparison of bacterial production estimated from thymidine incorporation and changes in cell numbers Bacterial culture
B. megaterium Expt 1
Expt 2 WCC isolate
Production ratea estimated from: Thymidine Changes in incorporation cell numbers
(0.27)b 107 (25.29)
1.18 x 107 x
1.66
6.75 x 106 (2.29)
1.02
x
107 (61.2)
3.99 x 106 (0.02) 4.76 x 106 (0.15) 4.47 x 107 (6.48)
.1 107 1.17x 9.97 x
1O6
4.72 x 107
Amt of DNA/cell (g)
Cell vol (>m3)
.2 1-14 3.27x
42 4.25
3.50 x
4491-1 4.49
1.76 x 10-15c
0.228
a Values for B. megateriun are expressed as cells per milliliter per hour. Values for the WCC isolate are expressed as numbers of cells per gram (dry weight) of sediment per hour. b Values in parentheses indicate the micromolar isotopic dilutions. cAssuming a cell density of 1.1 g (wet weight) per cm3, a water content of 80%, and a DNA content equal to 3.5% of dry weight (32).
VOL. 58, 1992
[3H]THYMIDINE METHOD APPLIED TO STREAM BED SEDIMENTS
for lake sediments (52), 0.4 M NaOH at 120°C for 0.5 h for seagrass sediments (37), and 0.3 N NaOH, 25 mM EDTA, and 0.1% SDS at 25°C for 12 h for river sediments (16). Brief extractions with a strong base at high temperatures have been shown to be effective with lake sediments (4) and superior to longer extractions with a weaker base at lower temperatures with marine sediments (37). Our data support these results and indicate a significant and variable reduction in efficiency when extractions are carried out at 25°C. Although the acid-base hydrolysis procedure has been criticized for lack of specificity (49, 50), recent studies with DNase and protease treatments have shown the procedure to be sufficiently selective for the isolation of DNA from seagrass sediments (38). In both of the studies which were critical of acid-base hydrolysis, DNA was extracted with 0.5 N NaOH at 60°C, conditions that, as we and others (16) have shown, can lead to excessive degradation of DNA. Alternative DNA extraction procedures have been suggested (50, 54), but these have not been tested on sediments. We used the reciprocal plot method (36) to measure isotopic dilution in stream bed sediments, with a particular focus on three specific areas: (i) the rate of thymidine uptake compared with the rate of thymidine incorporation into DNA, (ii) the inhibition of de novo synthesis by additions of exogenous thymidine, and (iii) the importance of adsorption of thymidine to sediments in altering the effective concentration of thymidine additions. The validity of the isotope dilution analysis rests in part on whether the velocity of thymidine incorporation into DNA is independent of the concentration of the radioactive precursor (29, 34). A related question is whether the rate of thymidine uptake into cells limits the rate of incorporation into DNA (47). Our data (Fig. 1) support the contention that bacteria in nature have thymidine uptake rates that exceed the rates of incorporation into DNA (34). Our estimate of community generation times in these studies was less than 20 h, so these results run counter to the suggestion that incorporation of thymidine into DNA is uptake limited when generation times are less than 20 h (51). The comparison of uptake rates and incorporation rates in Fig. 1 was done without correction for isotopic dilution. The presence of exogenous thymidine pools would influence both uptake and incorporation rate calculations equally, but to the extent that de novo synthesis plays an important role, the ratio of uptake to incorporation would be overestimated. However, even if we were to assume that our estimates of isotopic dilution were completely the result of de novo synthesis, and thus increased the rates of incorporation while having no effect on rates of uptake, uptake rates would still exceed incorporation rates by at least a factor of 2. Additionally, the production of [3H]H20 has been reported as an important pathway for thymidine in bacterial cells (23), and this would lead to an underestimation of uptake rates measured as radioactivity in the base hydrolysate. De novo synthesis occurs when bacterial cells are unable to satisfy their thymidine requirements from exogenous sources (34). Various concentrations of unlabeled thymidine have been used to inhibit de novo synthesis, ranging from 1 to 450 nM in water samples (3, 9, 19, 46) and from 300 nM to 20 ,uM in sediments (15, 16, 37). In our studies with stream bed sediment communities, concentrations of unlabeled thymidine as high as 60 ,uM did not eliminate de novo synthesis (Fig. 2). In practice, this means that estimates of isotopic dilution have to be performed with each assay. A similar inability to inhibit de novo synthesis or to saturate rates of thymidine incorporation into DNA has been reported for
3619
studies with Spartina alterniflora detritus and thymidine concentrations exceeding 100 ,uM (14) and for work with planktonic bacteria in mesotrophic lakes and thymidine concentrations of 50 nM (3) or 60 nM (41). Fallon and Newell (14) suggested that multiphasic enzyme systems or diffusion resistance could be responsible for their findings; probably both phenomena are occurring as well as the entry of thymidine into DNA by alternate metabolic pathways. The adsorption of thymidine to clay, organic matter, and hydrous oxides in sediments has been suggested as a mechanism capable of reducing the actual concentrations of thymidine that cells are exposed to during incubation. Adsorption would thwart attempts to inhibit de novo synthesis (44). More recently, the effect of different amounts of seagrass sediment on thymidine incorporation into DNA has been demonstrated and adsorption has been implicated as the mechanism (38). The suggestion is that more sediment, and especially clay, presents more sites for adsorption, thus effectively reducing the concentration of thymidine available for uptake. In contrast, our data indicate that for WCC stream bed sediments with 0.2 to 2% clays and 1 to 4% organic matter content, adsorption during incubation did not significantly reduce thymidine exposures. Similar conclusions have been reached for Hudson River water and sus-
pended matter (17). The siliceous sands of seagrass beds (Moreton Bay, Queensland, Australia) have approximately a 2- to 20-fold greater clay content than WCC sediments, and the presence of anaerobic conditions just 1 to 2 cm below the surface of the seagrass beds suggests that the organic content may also be considerably higher than that in WCC (38). These conditions may explain the differences in the adsorption phenomena reported for the two habitats. Our data that demonstrate no influence of isotopic dilution on the background in formalin-treated controls are not inconsistent with low adsorption of label (3%) during incubation. We only measured the adsorption of labeled thymidine. Presumably, the total amount of thymidine adsorbed increased with increasing thymidine concentrations. We kept the radioactivity constant, so solutions of lower specific activity would have higher thymidine concentrations and should result in higher total adsorption. Our observation that the percentage of label adsorbed remained constant simply means that the thymidine concentrations used were all within the linear region of the adsorption isotherm and sites for sorption were not limiting. Under these conditions, the more thymidine present, the more would be adsorbed, but the percentage adsorbed, albeit small, would remain constant.
The importance of accurately estimating isotopic dilution Accounting for isotopic dilution can increase estimates of bacterial productivity from 2- to 20-fold (9, 17, 20). The reciprocal plot technique for estimating isotopic dilution method has been validated with chemostat studies (44), comparisons with phospholipid biosynthesis (39), and comparisons to bacterial production measured by the direct enumeration of increases in cell numbers or biomass (2). Our data also validate the reciprocal plot technique for a gram-positive bacterium that probably is transported to streams from soils (B. megatenium) and three gram-negative isolates from WCC. Nevertheless, the isotopic dilution technique remains time-consuming, prone to large variance, and at times difficult to interpret (38, 47). In fact, presently there is no consensus on an appropriate technique for estimating isotopic dilution, and evidence has been presented that indicates that the currently available techniques are inadequate (13, 26). We would agree.
cannot be overstated.
3620
APPL. ENVIRON. MICROBIOL.
KAPLAN ET AL.
With the broad range of thymidine concentrations used in studies, reciprocal plots generated two or three linear regions and objective criteria are needed to decide which concentration of isotopic dilution is most accurate. The two lower ranges in the isotopic dilution curve (Fig. 2b and c) have a biochemical explanation. These curves most likely correspond to concentrations of thymidine where initially thymidine kinase (Fig. 2c) and then DNA polymerase (Fig. 2b) are the rate-limiting steps (34). Use of the former value would lead to an underestimate of bacterial production (Fig. 4). The explanation of the upper range (Fig. 2a) is not clear. At concentrations of thymidine exceeding 10 FLM, uptake by eucaryotes, either algae (48), protozoa (34), or fungi (42), may become quantitatively important even with incubation periods of less than 1 h. Whatever the phenomenon, we can eliminate the exceedingly large isotopic dilution values and the associated estimates of DNA synthesis on physiological grounds; no one has reported generation times ranging from 3 to < 1 h for bacterial communities in nature growing at 12.5 to 17°C. These criteria are ultimately unsatisfying and illustrate the need for a better understanding of the biochemistry of thymidine incorporation in complex microbial communiour
ties.
The final step in the thymidine incorporation procedure is the conversion of thymidine incorporation rates into cell or carbon production rates. This can be accomplished by using a conversion factor that is either calculated from the amount of DNA per cell (35) or empirically derived, usually from a change in cell densities within dilution culture (2, 10, 27). Unfortunately, both the measurement of the amount of DNA per cell and the empirical determination of conversion factors by dilution culture are not easily accomplished with sediments. In sediments, direct measurements of the amount of DNA per cell are complicated by the presence of dissolved DNA (43) and eucaryotic DNA. Additionally, it is not currently possible to distinguish DNA extracted from actively replicating bacteria versus that extracted from dormant bacteria. If literature values of the amount of DNA per cell are used, estimates can vary by as much as fourfold, within the range of 1.3 to 5 fg per cell (20, 51). Attempts to refine these estimates by measuring cell volume, converting cell volume to mass, and calculating DNA as a percentage of dry mass are subject to the variability within each of these conversion factors. For example, DNA as a percentage of dry mass has been reported to range from 3.5 to 12.7% (6, 13, 51), and water content as a percentage of volume has been reported to decrease from 82 to 46% as cells decrease in volume (51). Additional factors that diminish the accuracy of theoretical estimates of bacterial DNA content include the fact that the amount of DNA per cell varies during the cell cycle (40), increases with growth rate in culture (8, 25), and decreases with nutrient limitation or starvation (1, 24). In summary, we conclude that the estimation of bacterial production from measurements of thymidine incorporation into DNA can yield reasonable values, but it is not a routine assay. One area of concern is the inability of some bacteria, both aerobes (34) and anaerobes (21, 55), to incorporate thymidine into DNA. We have isolated several aerobes from WCC that fall into this category and thus recognize that an unknown portion of the WCC stream bed sediment community is excluded from the thymidine estimate of community productivity. Two other areas of concern include the determination of isotopic dilution and the conversion of DNA incorporation rates into cell production. A better understanding of the biochemical pathways for thymidine in
complex communities is needed to address the former and a method that would permit the isolation of DNA from actively replicating bacteria would address the latter. At present, there still appears to be a need to perform multiple assay comparisons (5) when measuring bacterial production in sediments. ACKNOWLEDGMENTS We thank S. L. Roberts and L. J. Staszak for excellent technical support, R. D. Fallon and two anonymous reviewers for comments that improved the manuscript, and J. D. Newbold for helpful discussions. This research was funded by the National Science Foundation grant DEB-8217482, the Francis Boyer Research Endowment Fund, and the Stroud Foundation. REFERENCES 1. Amy, P. S., C. Pauling, and R. Y. Morita. 1983. Starvationsurvival processes of a marine vibrio. Appl. Environ. Microbiol. 45:1041-1048. 2. Bell, R. T. 1986. Further verification of the isotope dilution approach for estimating the degree of participation of [H]thymidine in DNA synthesis in studies of aquatic bacterial production. Appl. Environ. Microbiol. 52:1212-1214. 3. Bell, R. T. 1990. An explanation for the variability in the conversion factor deriving bacterial cell production from incorporation of [3H]thymidine. Limnol. Oceanogr. 35:910-915. 4. Bell, R. T., and I. Ahlgren. 1987. Thymidine incorporation and microbial respiration in the surface sediment of a hypereutrophic lake. Limnol. Oceanogr. 32:476-482. 5. Bott, T. L., and L. A. Kaplan. 1985. Bacterial biomass, metabolic state, and activity in stream sediments: relation to environmental variables and multiple assay comparisons. Appl. Environ. Microbiol. 50:508-522. 6. Brock, T. D. 1974. Biology of microorganisms. Prentice-Hall, Inc., Englewood Cliffs, N.J. 7. Buchanan, R. E., and N. E. Gibbons (ed.). 1974. Bergey's manual of determinative bacteriology. The Williams & Wilkins Co., Baltimore. 8. Cho, B. C., and F. Azam. 1988. Heterotrophic bacterioplankton production measurement by the tritiated thymidine incorporation method. Arch. Hydrobiol. Beih. Ergeb. Limnol. 31:153162. 9. Chrzanowski, T. H. 1988. Consequences of accounting for isotopic dilution in thymidine incorporation assays. Appl. Environ. Microbiol. 54:1868-1870. 10. Coveney, M. F., and R. G. Wetzel. 1988. Experimental evaluation of conversion factors for the [3H]thymidine incorporation assay of bacterial secondary productivity. Appl. Environ. Mi-
crobiol. 54:2018-2026. 11. Dawes, E. A. 1980. Quantitative problems in biochemistry. Longman Inc., New York. 12. Dobbs, F. C., J. B. Guckert, and K. R. Carman. 1989. Comparison of three techniques for administering radiolabeled substrates to sediments for trophic studies: incorporation by microbes. Microb. Ecol. 17:237-250. 13. Ellenbroek, F. M., and T. E. Cappenberg. 1991. DNA synthesis and tritiated thymidine incorporation by heterotrophic freshwater bacteria in continuous culture. Appl. Environ. Microbiol.
57:1675-1682. 14. Fallon, R. D., and S. Y. Newell. 1986. Thymidine incorporation
by the microbial community of standing dead Spartina altemiflora. Appl. Environ. Microbiol. 52:1206-1208. 15. Fallon, R. D., S. Y. Newell, and C. S. Hopkinson. 1983. Bacterial production in marine sediments: will cell-specific measures agree with whole-system metabolism? Mar. Ecol. Prog. Ser. 11:119-127. 16. Findlay, S. E. G., J. L. Meyer, and R. T. Edwards. 1984. Measuring bacterial production via rate of incorporation of [3H]thymidine into DNA. J. Microbiol. Methods 2:57-72. 17. Findlay, S., M. L. Pace, D. Lints, J. J. Cole, N. F. Caraco, and
VOL. 58, 1992
18.
19.
20.
21.
22.
23. 24.
25. 26. 27. 28. 29. 30. 31.
32.
33. 34. 35.
36. 37.
[3H]THYMIDINE METHOD APPLIED TO STREAM BED SEDIMENTS
B. Peierls. 1991. Weak coupling of bacterial and algal production in a heterotrophic ecosystem: the Hudson River estuary. Limnol. Oceanogr. 36:268-278. Forsdyke, D. R. 1968. Studies on the incorporation of [5-3H] uridine during activation and transformation of lymphocytes induced by phytohaemagglutinin. Biochem. J. 107:197-205. Fuhrman, J. A., and F. Azam. 1980. Bacterioplankton secondary production estimates for coastal waters of British Columbia, Antarctica, and California. Appl. Environ. Microbiol. 39:10851095. Fuhrman, J. A., and F. Azam. 1982. Thymidine incorporation as a measure of heterotrophic bacterioplankton production in marine surface waters: evaluation and field results. Mar. Biol. 66:109-120. Gilmour, C. C., M. E. Leavitt, and M. P. Shiaris. 1990. Evidence against incorporation of exogenous thymidine by sulfate-reducing bacteria. Limnol. Oceanogr. 35:1401-1409. Hobbie, J. E., R. J. Daley, and S. Jasper. 1977. Use of Nuclepore filters for counting bacteria by fluorescence microscopy. Appl. Environ. Microbiol. 33:1225-1228. Hollibaugh, J. T. 1988. Limitations of the [3H]thymidine method for estimating bacterial productivity due to thymidine metabolism. Mar. Ecol. Prog. Ser. 43:19-30. Hood, M. A., J. B. Guckert, D. C. White, and F. Deck. 1986. Effect of nutrient deprivation on lipid, carbohydrate, DNA, RNA, and protein levels in Vibrio cholerae. Appl. Environ. Microbiol. 52:788-793. Ingraham, J. L., 0. Maaloe, and F. C. Neidhardt. 1983. Growth of the bacterial cell. Sinauer Associates, Inc., Sunderland, Mass. Jeffrey, W. H., and J. H. Paul. 1988. Underestimation of DNA synthesis by [3H]thymidine incorporation in marine bacteria. Appl. Environ. Microbiol. 54:3165-3168. Kirchman, D., H. Ducklow, and R. Mitchell. 1982. Estimates of bacterial growth from changes in uptake rates and biomass. Appl. Environ. Microbiol. 44:1296-1307. Krumbein, W. C., and F. J. PettUohn. 1938. Manual of sedimentary petrography. Appleton Century Co., Inc., New York. Laws, E. A. 1983. Plots of turnover times versus added substrate concentrations provide only upper bounds to in situ substrate concentrations. J. Theor. Biol. 101:147-150. Lorenzen, C. J. 1967. Determination of chlorophyll a and pheo-pigments: spectrophotometric equations. Limnol. Oceanogr. 12:343-346. Lovell, C. R., and A. Konopka. 1985. Seasonal bacterial production in a dimictic lake as measured by increases in cell numbers and thymidine incorporation. Appl. Environ. Microbiol. 49:492500. Luria, S. E. 1960. The bacterial protoplasm: composition and organization, p. 1-34. In I. C. Gunsalus and R. Y. Stanier (ed.), The bacteria-a treatise on structure and function. Academic Press, Inc., New York. Matsuyama, T. 1984. Staining of living bacteria with rhodamine 123. FEMS Microbiol. Lett. 21:153-157. Moriarty, D. J. W. 1986. Measurement of bacterial growth rates in aquatic systems from rates of nucleic acid synthesis. Adv. Microb. Ecol. 9:245-292. Moriarty, D. J. W. 1988. Accurate conversion factors for calculating bacterial growth rates from thymidine incorporation into DNA: elusive or illusive? Arch. Hydrobiol. Beih. Ergeb. Limnol. 31:211-217. Moriarty, D. J. W., and P. C. Pollard. 1981. DNA synthesis as a measure of bacterial productivity in seagrass sediments. Mar. Ecol. Prog. Ser. 5:151-156. Moriarty, D. J. W., and P. C. Pollard. 1982. Diel variation of bacterial productivity in seagrass (Zostera capricomi) beds
38.
39.
40. 41. 42.
43. 44.
45. 46.
47.
48. 49.
50.
51. 52. 53.
54.
55.
3621
measured by rate of thymidine incorporation into DNA. Mar. Biol. 72:165-173. Moriarty, D. J. W., and P. C. Pollard. 1990. Effects of radioactive labelling of macromolecules, disturbance of bacteria and adsorption of thymidine to sediment on the determination of bacterial growth rates in sediment with tritiated thymidine. J. Microbiol. Methods 11:127-139. Moriarty, D. J. W., D. C. White, and T. J. Wassenberg. 1985. A convenient method for measuring rates of phospholipid synthesis in seawater and sediments: its relevance to the determination of bacterial productivity and the disturbance artifacts introduced by measurements. J. Microbiol. Methods 3:321-330. Morse, M. L., and C. E. Carter. 1949. The synthesis of nucleic acids in culture of Eschenchia coli, strains B and B/R. J. Bacteriol. 58:317-326. Nagata, T. 1987. Production rate of planktonic bacteria in the north basin of Lake Biwa, Japan. Appl. Environ. Microbiol. 53:2872-2882. Newell, S. Y., and R. D. Fallon. 1991. Toward a method for measuring instantaneous fungal growth rates in field samples. Ecology 72:1547-1559. Paul, J. H., W. H. Jeffrey, and M. F. Deflaun. 1987. Dynamics of extracellular DNA in the marine environment. Appl. Environ. Microbiol. 53:170-179. Pollard, P. C., and D. J. W. Moriarty. 1984. Validity of tritiated thymidine method for estimating bacterial growth rates: measurement of isotope dilution during DNA synthesis. Appl. Environ. Microbiol. 48:1076-1083. Porter, K. G., and Y. S. Feig. 1980. The use of DAPI for identifying and counting aquatic microflora. Limnol. Oceanogr. 25:943-948. Riemann, B., P. Nielsen, M. Jeppesen, B. Marcussen, and J. A. Fuhrman. 1986. Diel changes in bacterial biomass and growth rates in coastal environments, determined by means of thymidine incorporation into DNA, frequency of dividing cells (FDC), and microautoradiography. Mar. Ecol. Prog. Ser. 17:227-235. Riemann, B., and M. Sondergaard. 1986. Bacteria, p. 131-141. In B. Riemann and M. Sondergaard (ed.), Carbon dynamics in eutrophic, temperate lakes. Elsevier Science, Amsterdam. Rivkin, R. B. 1986. Incorporation of tritiated thymidine by eucaryotic microalgae. J. Phycol. 22:193-198. Robarts, R. D., R. J. Wicks, and L. M. Sepfton. 1986. Spatial and temporal variations in bacterial macromolecule labeling with [methyl-3H]thymidine in a hypertrophic lake. Appl. Environ. Microbiol. 52:1368-1373. Servais, P., J. Martinez, G. Billen, and J. Vives-Rego. 1987. Determining [3H]thymidine incorporation into bacterioplankton DNA: improvement of the method by DNase treatment. Appl. Environ. Microbiol. 53:1977-1979. Simon, M., and F. Azam. 1989. Protein content and protein synthesis rates of planktonic marine bacteria. Mar. Ecol. Prog. Ser. 51:201-213. Tobin, R. S., and D. H. J. Anthony. 1978. Tritiated thymidine incorporation as a measure of microbial activity in lake sediments. Limnol. Oceanogr. 23:161-165. Velji, M. I., and L. J. Albright. 1986. Microscopic enumeration of attached marine bacteria of seawater, marine sediment, fecal matter, and kelp blade samples following pyrophosphates and ultrasound treatments. Can. J. Microbiol. 32:121-126. Wicks, R. J., and R. D. Robarts. 1987. The extraction and purification of DNA labelled with [methyl-3H]thymidine in aquatic bacterial production studies. J. Plankton Res. 9:11591166. Winding, A. 1992. [3H]thymidine incorporation to estimate growth rates of anaerobic bacterial strains. Appl. Environ. Microbiol. 58:2660-2662.
