Influence of pH on the toxic effects of zinc, cadmium, and pentachlorophenol on pure cultures of soil microorganisms

Culture of microorganisms

Aspergillus niger CBS 121.49, DSM 2182 (filamentous fungi) isolated in 1947 from Dutch soil in Wageningen was obtained from the German Collection of Microorganisms and Cell Cultures DSM in Braunsweig, Germany. For storage, 0.1 ml of a spore suspension in 4.8% skim milk with 4% glutamate was pipetted in a tube with 0.5 g of dry and sterilized silica gel. The fungus was grown on nutrient broth (Oxoid CM1, Hampshire, UK) agar plates. The spores were harvested from the agar plate and suspended in water. About 12.5 ml of a spore suspension with A₅₅₀ = 1 was diluted in 2.5 L of liquid nutrient broth and incubated overnight at 28°C until fungal spheres were formed (1 mm diameter, 0.8 g dry weight/L). The spheres were harvested on a sterilized stainless-steel sieve (38 μm) and washed with the appropriate buffer for toxicity testing.

Streptomyces lividans 66 NCIMB 11416 (actinomycetes) was obtained from the national collections of industrial and marine bacteria in Aberdeen, UK. One colony with fresh white spores was inoculated in 100 ml of liquid nutrient broth, and the actinomyces was grown for 2 d at 28°C. This was used to inoculate 2.5 L of liquid nutrient broth, which was incubated for 24 h at 28°C. The spheres that passed through a 425-μm sieve were harvested and washed on a 38-μm sieve.

Pseudomonas putida MT-2 (Gram-negative bacteria) was originally isolated from Japanese soil. The strain was obtained from P.A. Williams, who stored it in the American Type Culture Collection (ATCC 3301). The closely related strain Pseudomonas putida DSM 50026, obtained from the German culture collection, was isolated from a channel in Berlin and is used as a test strain for water toxicity [14, 15].

Rhodococcus erythropolis A177 (Gram-positive bacteria) was obtained from H.H.M. Rijnaarts [16]. The Rhodococcus and Pseudomonas strains were stored at −80°C, grown on nutrient broth, and washed by centrifugation for a toxicity test.

Zinc sorption

The sorption of zinc to P. putida MT-2 and P. putida DSM 50026 was measured in a 20-mmol/L tris(hydroxymethyl) aminomethane buffer (Tris) (Sigma Chemical Co., St. Louis, USA) neutralized with HC1 (Merck, Darmstadt, Germany). Cells were washed with buffer by centrifugation at 3,300 g for 15 min, and the pellet was resuspended in buffer. For Tris buffer at pH 5.5 and pH 6, nominal zinc concentrations of 30 and 10 mg/L, respectively, were used. For Tris buffer at pH 7.8, three nominal concentrations of zinc were used, 0.3, 3, and 30 mg/L. The suspensions were incubated for 90 min at 28°C. After centrifugation the zinc concentration was measured in the supernatant and the pellet with an atomic absorption spectrophotometer (Perkin and Elmer, type 2380, Ueberlingen, Germany).

PCP sorption

The sorption of PCP (Fluka, Buchs, Switzerland) was measured in a 20-mmol/L potassium phosphate buffer (Merck). Washed microorganisms were incubated in 2.5 ml of buffer with [¹⁴C]PCP (Amersham, Buckinghamshire, UK) at a concentration of 3 μg [¹⁴C]PCP/ml with 16 Bq/ml for 1 h at 28°C. After centrifugation the PCP concentration in the supernatant and the pellet was measured with a Tricarb liquid scintillation counter (Packard, type 1500 or 1550, Downers Grove, IL, USA).

Toxicity testing

The method used to perform the toxicity tests was described in more detail previously [17]. Washed microorganisms were incubated in a closed 60-ml bottle with 20 ml of buffer containing 1 ng [¹⁴C]acetate/ml with 10 Bq/ml. The low amounts of microorganisms and substrate assure that the 40 ml of air in each bottle is not depleted. The bottles were incubated for different time intervals with different toxicant concentrations, and mineralization was stopped with 1 ml of 2 M H₂SO₄. The ¹⁴CO₂ was flushed out of the bottles with nitrogen gas, trapped, and counted with a scintillation detector. The remaining substrate was also measured with a scintillation detector.

The tests are designed to measure the toxicity under defined conditions with minimal interaction of the toxicant with the medium. For PCP a 20-mmol/L potassium phosphate buffer (KPi) was used so that potassium was the counter ion of the pentachlorophenolate anion. Pentachlorophenol was added as phenolate anion, which is more soluble, and Zn²⁺ was added as ZnC1₂. For Zn²⁺ a 20-mmol/L Tris buffer neutralized with HC1 or a 20-mmol/L bis(2-hydroxyethyl)imino-tris(hydroxymethyl) methane buffer (BisTris)(Sigma) neutralized with HC1 was used. At 25°C the pK_a of KPi, BisTris, and Tris is 7.21, 6.5, and 8.06, respectively. Because BisTris and Tris are both positively charged buffers, the counter ion of Zn²⁺ was chloride in our experiments.

Statistics

All toxicity and sorption experiments were performed at least in duplicate. The percentage of substrate consumed and CO₂ formed are plotted against the logarithm of the toxicant concentration. Two logistic curves were fitted with these data, and the EC50 and EC10 were calculated as described previously [17].

In addition, a more detailed statistical analysis was performed with GraphPad Prism Software (San Diego, CA, USA) using the standard formula for sigmodial dose response:

(1)

where X = the log₁₀ of the toxicant concentration Y = the percentage CO₂ or acetate t = the value at the top of the curve (the maximal activity without inhibition) b = the value at the bottom of the curve (usually 0) and h = the slope of the dose-effect curve.

At the EC10 toxicant concentration, Y = b + (t - b)· 0.9. When this is entered into the above equation, the EC10 can be calculated from the EC50 and slope, giving

(2)

The standard error in the logEC10 is named SE10 and was calculated using the rules for the propagation of linear and multiplicative random errors [18]:

(3)

where SE50 = the standard error of log(EC50) SE_h = the standard error of h, and SQRT = square root.

The standard error in the percentage substrate consumed was larger than the standard error in CO₂ production because an increase in the percentage CO₂ from 0 to 10% is measured more accurately than a decrease in the percentage substrate from 100 to 90%. The 95% confidence intervals of the log(EC50) were calculated from the log(EC50) ± 1.96 × SE50. Results obtained with this statistical procedure were compared with those obtained using the GenStat 5 (Oxford) and Kaleidagraph 3.0.1 (Abelbeck Software) statistical packages, and all were similar.

Toxic effects of PCP on acetate mineralization by five microorganisms

The mineralization of low concentrations of substrates by microorganisms can be described by first-order kinetics. When increasing concentrations of a toxicant are added, the mineralization rate decreases.

Figure 1 shows the toxic effect of PCP on acetate mineralization by spheres of A. niger mycelium in phosphate buffer at pH 6. At low toxicant concentrations, about 54% of the [¹⁴C]acetate is converted to cell material plus 39% ¹⁴CO₂. The PCP concentration is shown on a logarithmic scale on the X axis. The control is arbitrarily drawn at 0.001 mg PCP/L because zero cannot be plotted on a logarithmic scale. At high PCP concentrations no ¹⁴CO₂ is formed, and the [¹⁴C]acetate is not mineralized during incubation (200 min). The descending ¹⁴CO₂ production curve was fitted by nonlinear regression and can be described by Equation 1 using a log(EC50) of -0.9 (with a standard deviation of 0.1) and a slope h of -2.4 (with a standard deviation of 1.2). This results in an EC50 of 0.13 mg PCP/L with a 95% confidence interval of 0.08 to 0.2 mg PCP/L and an EC10 of 0.05 mg PCP/L (using Eqn. 2) with a 95% confidence interval of 0.02 to 0.15 mg PCP/L (calculated with Eqn. 3). Note that the confidence intervals are asymmetric because of the ¹⁰log transformation of the EC50 and EC10. The EC50 and EC10 from the ascending acetate curve were calculated in a similar way. In Table 1 the data of Figure 1 are summarized at the second row from the bottom. The other EC50 and EC10 values in Table 1 were calculated from dose-effect curves that are not shown.

Table 1 gives an overview of 10 dose-effect curves describing the effect of PCP on different microorganisms of various pH values. Striking differences in sensitivity are evident between species. The Gram-negative P. putida strains are not very sensitive to PCP, whereas A. niger was the most sensitive. Pentachlorophenol becomes less toxic at high pH values with such very different microorganisms as a Gram-negative bacterium Pseudomonas, an actinomycete Streptomyces, and a fungus Aspergillus.

Uptake and sorption of PCP to Pseudomonas putida MT2

The decreased toxicity of PCP at high pH might be caused by decreased uptake by the microorganisms. Uptake into the cell is not easy to measure because it is difficult to distinguish sorption to the outside of the cell with uptake to the inside. Therefore, the sorption of PCP onto and into the cells was studied on one of the species tested. Sorption experiments were performed with [¹⁴C]PCP and P. putida MT2 in phosphate buffer. The sorption at pH 7 can be described with a logarithmically transformed Freundlich isotherm [9]:

(4)

where C_b = the concentration in the solid phase (in mg/kg) C_w = the concentration in the water phase n = the correlation factor for nonlinear sorption, and K_f = the sorption constant (in L/kg)

The lower line in Figure 2 shows logarithmic transformation of the water concentration (C_w) in milligrams of PCP per liter of phosphate buffer at pH7 and the concentration in P. putida MT2 (C_b) in milligrams of PCP per kilogram of bacteria dry weight. The line was fitted to the data points with linear regression and can be described with Equation 4 using n = 1.1 and log(K_f) = 2.3. This equation can be rewritten to C_b = K_f · C, which can be simplified to

(5)

since the nonlinearity factor n = 1.1 is close to 1. At pH 7 the total sorption ratio, K, is almost equal to K_f = 10^2.3L/kg, that is, 210 L/kg.

This simplification facilitates calculation of PCP sorption to the bacterial cells at a nominal concentration of 3 mg/L at different pH values. The total sorption ratio K is equal to the sum of the sorption of the neutral K_n and anionic form K_i of PCP [19]:

(6)

where the undissociated fraction

(7)

Figure 3 shows the effect of pH on the sorption of PCP to P. putida MT 2. The curve describes the K value calculated with Equation 6 and fitted with nonlinear regression with a fixed pK_a of 4.75.K_n = 2,300 L/kg dry weight with a standard error of 103, and K_i = 150 L/kg dry weight with a standard error of 40.

Because the sorption constant is now known at all the test pH values, the “internal dose” sorbed onto or into the bacteria can be calculated from the effect concentrations. This internal dose is the sum of the amount of toxicant that is sorbed onto the cell and the intracellular dose. The PCP concentration that is present in the water is referred to as the “external dose.” Table 2 shows the calculation of the internal dose of PCP that inhibits acetate mineralization by 50%. Columns 2 through 4 show the EC50s from Table 1 converted to the “internal” doses using K_f values from Figure 3. The “minimum” and “maximum” columns indicate the 95% confidence intervals that were calculated from the nonlinear regression of the dose-effect curves. The internal EC50 (in mg PCP/g cell dry weight) is not constant for all pH values. Therefore, the differences in the sensitivity to PCP of P. putida MT 2 at various pH values cannot be attributed solely to differences in sorption.

The internal EC50 concentrations of PCP in P. putida can be compared with the internal EC50 concentration in goldfish (0.27 mg PCP/g dry weight [8], assuming that a goldfish contains 80% water). This value is in the same range as the internal EC50 of PCP in P. putida (see Table 2).

Toxic effect of zinc on acetate mineralization by five species of microorganisms

A cationic buffer like Tris or BisTris was used for the experiments with Zn²⁺ to avoid the strong interactions like precipitation and complexation that can occur between an anionic buffer and zinc. The counter anion was chloride for both the buffer and the toxicant. A cationic buffer might show some competition with Zn²⁺ on the uptake system of the cells, and therefore it is not possible to select a buffer without any possible interaction with Zn²⁺. Table 3 shows the toxic effects of zinc on acetate mineralization by five microorganisms in Tris or BisTris buffer at different pH values. Large differences in sensitivity to zinc are evident between the different species. Pseudomonas putida MT2 was the most sensitive to zinc, whereas A. niger was not sensitive at all.

At higher pH the toxicity of zinc increased considerably. This is in agreement with earlier observations that zinc was more toxic to microorganisms in alkaline soils (see Introduction). Acid soils are known to contain more fungi that are relatively tolerant to metals [20]. Both the decreased toxicity and the increased proportion of fungi might add to the relative insensitivity of the microflora in acid soils to zinc.

Table Table 4.. Toxic effect of zinc and cadmium on glucose mineralization of Pseudomonas putida MT-2 at different pH values

						EC50 (mg/L)		EC10 (mg/L)
	pH	Biomass (mg/L)	Half-life (min)	Rate (L/μg min−1)	Time (min)	CO	Acetate	CO2	Acetate
Zn	6	6	406	0.28	110	11b	11	3d	<1
Zn	7	6	315	0.37	110	<1	<1	<0.3	<0.3
Zn	8	6	741	0.16	110	0.09c	<0.3	0.02e	<0.1
Cd	6	6	406	0.28	110	122d		43f
Cd	7	6	315	0.37	110	5.9e	3.6	2.9g	2.9
Cd	8	6	741	0.16	110	0.1e	0.3	0.03e	0.2

• ^a The EC50 for CO₂ was derived from the ¹⁴CO₂ production from [¹⁴C]glucose.
• ^b-g Values with the same letters have overlapping 95% confidence intervals.

Toxic effect of zinc and cadmium on glucose mineralization by P. putida MT2

Acetic acid has a pK_a value of 4.75 [21], which might also influence the acetate mineralization and the toxic effect of a metal like zinc at different pH values. Tables 1 and 3 show that the half-lives of the acetate mineralization increase at higher pH with P. putida MT2 and A. niger but not with P. putida DSM 50026 or R. erythropolis. An indirect effect of the cation Zn²⁺ on the mineralization of the anion acetate could not be excluded from the above experiments. Therefore, the toxicity tests were also performed with [¹⁴C]glucose as a substrate for P. putida MT2.

Table 4 shows that the toxicity of both Zn²⁺ and Cd²⁺ to glucose mineralization also increases at higher pH. The sensitivities to zinc of acetate and glucose mineralization by P. putida MT2 are very similar. The toxicity of cadmium increases sharply at increasing pH. One might speculate that the increased toxicity of zinc and cadmium at higher pH is not a peculiarity of these metals but a more general phenomenon that occurs with many metals.

Sorption of Zn²⁺ to P. putida MT−2 in Tris buffer at different pH values

The decreased toxicity observed at low pH might be caused by competition of H⁺ with Zn²⁺ at the uptake systems of the microorganisms. Because uptake into cells is not measured easily, we studied the sorption of Zn²⁺. Sorption of zinc to a concentrated suspension of 2.3 g P. putida MT-2/L Tris buffer was measured after 90 min of incubation at 28°C. The sorption is described by a Freundlich equation:

(8)

where C_b = the internal concentration sorbed onto or in the bacteria (in mg Zn/kg bacteria dry weight) and C_w = the concentration in the water phase (in mg Zn/L)

The constants n = 1.6 and log(K_f) = 3.2 were derived by linear regression of the sorption data shown in Figure 2. Table 5 shows that sorption is not very dependent on pH for this Gram-negative bacterium. At pH 5.5 the sorption constant K was 1.9 L/g, while at pH 7.8 it was 3.2 (see the upper and lower rows of Table 5). The EC50 at pH 5.5 was 64 times higher than the EC50 at pH 7.8 (see Table 3). Hence, no single internal dose corresponds to the EC50 of zinc at all pH values.

Sorption and toxicity of metals to R. erythropolis A177

The sorption of zinc and cadmium on the cell walls and intact cells of R. erythropolis A177 was decreased at lower pH [22] and higher Ca²⁺ concentrations due to competition of H⁺ and Ca²⁺ with Zn²⁺ or Cd²⁺ on the bindings sites on the cell wall. Rhodococcus erythropolis A177 is a Gram-positive bacterium that has a thick cell wall that is not covered with a membrane [22]. In contrast, Gram-negative bacteria like P. putida have a thin cell wall covered with a membrane that forms a barrier for the sorption of ions. For R. erythropolis A177, the large increase of the EC50 for zinc and cadmium at lower pH could be explained by decreased sorption to the cell wall [22], whereas for P. putida MT-2 this was not the case.

Data comparison

The toxic effects of PCP, zinc, and cadmium were measured on different pure cultures of microorganisms. A detailed comparison is possible only when identical strains are used. Table 6 shows the reported toxic effects of PCP, zinc, and cadmium on the growth or respiration of pure cultures of strains similar to those used in this study. Glutamate respiration by P. putida DSM 50026 in continuous culture was five times more sensitive for PCP than [¹⁴C]acetate mineralization by the same strain (Table 1). In both tests a low substrate concentration was present, which is normal in natural environments. Growth of A. niger PC6 on agar medium was 1,000 times less sensitive for PCP than [⁴C]acetate mineralization by A. niger CBS 121.49 (Table 1). The EC50 of zinc for P. putida ranged from 0.14 to 9 depending on the pH and the strain (Table 3), whereas higher values have been reported for P. fluorescens (Table 6).

Also, the EC50 of cadmium on [¹⁴C]glucose mineralization by P. putida MT2 was lower (Table 4) than the EC50 at pH 7 for other Pseudomonas strains (Table 6). Thus, in general assessing the mineralization of low concentrations of ¹⁴C-labeled acetate or glucose in buffer seems to be a sensitive method for the measurement of toxic effects. In contrast, growth of A. flavus or A. niger on media (Table 6) was more sensitive to zinc than [¹⁴C]acetate mineralization by A. niger CBS 121.49 (Table 1). This might be due to differences in the sensitivity of the strains because the toxicity of zinc to the growth of fungi in medium decreases at higher pH because of precipitation of the Zn²⁺ with components of the medium [23]. The presence of the Tris⁺ cation in our experiments at low pH also may have led to decreased toxicity because cations may compete with metal ions for the available anionic sites on the surface of microorganisms.

Influence of acidification on toxic effects of metals for soil animals and microorganisms

The toxicity of metals to soil animals can increase with soil acidification [2], whereas an opposite effect occurs with zinc on soil microorganisms [3-5]. Three factors can contribute to this difference: (1) acid soils contain fungi that are more resistant to metals [20]; (2) the toxicity of metals like zinc and cadmium to microorganisms is decreased at low pH values, (shown in this article); and (3) soil animals can control the level of pH in their intestines, which makes metal uptake via the consumption of pore water and food less dependent on the external pH.