Environ. Sci. Technol. 2007, 41, 2750-2755
Biosorption of Nonpolar Hydrophobic Organic Compounds to Escherichia Coli Facilitated by Metal and Proton Surface Binding LIN XIAO, XIAOLEI QU, AND DONGQIANG ZHU* State Key Laboratory of Pollution Control and Resource Reuse, and School of the Environment, Nanjing University, Jiangsu 210093, P.R. China
We observed that the presence of transition metal ion, Ag+, Cu2+, or Fe3+, at a concentration of 3 mg L-1 increases sorption of two nonpolar hydrophobic organic compounds (HOCs), phenanthrene (PHEN), and 1,2,4,5-tetrachlorobenzene (TeCB) by 1.5-4 times to Gram-negative bacteria Escherichia coli. Complexation of transition metals with the deprotonated functional groups (mainly carboxyl) of bacterial cell walls neutralizes the negative charge, making the bacterial surface less hydrophilic and enhancing hydrophobic partition of HOCs. This is evidenced by the fact that the zeta potential (ζ) value of bacteria becomes less negative when a transition metal is present. Furthermore, the observed higher sorption of PHEN than TeCB at low pH (3.8) cannot be fully explained by the pHdependent hydrophobic effects. The results led us to propose two specific sorption mechanisms for π-donor compounds: cation-π interactions with protonated amines and π H-bonding with protonated carboxyls. The biosorption of PHEN was best described as π-donor compared to the biosorption of TeCB considered non-π-donor. Results of the present study highlight that the presence of coexisting transition metals and changes on pH have a major effect on the biosorption of nonpolar HOCs.
Introduction Microorganisms including bacteria, fungi, and protozoan are ubiquitous, comprising the most diversified, fundamental component of the biosphere (1). It is thus of great importance to study the molecular mechanisms and environmental factors that govern uptake and accumulation of hydrophobic organic compounds (HOCs) by microorganisms, referred to as biosorption. For instance, bacteria are recognized as major colloidal constituents with high mobility in groundwater (24), and so organic pollutants may be sorbed and transferred with the bacteria, leading to further, more extensive contamination in the aquifer. Furthermore, bacteria tend to sorb many organic pollutants strongly, including dyes, phenolics, and pesticides, and biosorption is becoming a promising alternative to existing methods of wastewater treatment (59). Because the interior plasma membrane is impermeable to organic pollutants, the bacterial cell wall is expected to be the primary component responsible for organic biosorption. * Corresponding author phone: +86 025-8359-6496; fax: +86 0258359-6496; e-mail:
[email protected]. 2750
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 8, 2007
Bacterial cell walls are composed of biphilic polymers (polysaccharide, cross-linked shot peptides, proteins, organic acids), containing both hydrophobic moieties such as biolipid alkane chains and hydrophilic moieties such as carboxyl, hydroxyl, and phosphate groups (10). In addition, the O-containing groups (mainly carboxyl and phosphate, with pKa values of 4.87 and 6.9, respectively, ref 11) can be deprotonated to accumulate negative charge at the bacterial surface. Corresponding to the heterogeneous structures of bacterial surfaces, sorption mechanisms of HOCs may include hydrophobic effects and specific molecular-level interactions (e.g., H-bonding and electrostatic forces), the type and intensity of which is dependent on the solute’s structural properties. Wu and Yu (12) showed that sorption of phenol and mono- or dichlorophenols by fungal mycelia increases with increasing solute hydrophobicity, as justified by their water solubilities (SW) and n-octanol-water partitioning coefficients (KOW). Solution chemistry, including pH and ionic strength may also impact biosorption of HOCs via effects on bacterial surface chemistry (9). For example, Ju et al. (13) found that decreasing pH or increasing ionic strength increases biosorption of lindane. Complexation of heavy metals may modify the chemical and structural properties of bacterial surfaces, and thus influence sorption of HOCs. Binding of transition metals (e.g., Cd2+ and Fe3+) to bacterial surfaces has been well characterized by equilibrium chemical reaction models to incorporate complexation with functional groups and metal speciation in aqueous solution (11, 14, 15). In these studies pH was found to be the master variable to affect metal adsorption by portioning the deprotonated surface functional groups, i.e., sorption increases with deprotonation of surface groups. However, the impact of transition metal cosolute on organic biosorption has not been adequately examined. For example, contaminated soils often contain polycyclic aromatic hydrocarbons (PAHs), organic solvents, pesticides, and heavy metals in addition to all naturally occurring chemical species (alkali and alkaline earth metals, trace metals, anions, natural organic matter) (16, 17). To date there are only a few relevant studies carried out in the field of wastewater treatment on polar, ionic organics with metals present at high concentrations that greatly surpass the levels in soil and subsurface environments. Aksu and Akpinar (18, 19) observed that the sorption of phenol to activated sludge is inhibited by Ni2+ or Cr6+ ranging from 25 to 500 mg L-1 due to competition for similar binding sites (deprotonated functional groups). O’Mahony et al. (20) reported that the presence of high-level Cd2+ (100 mg L-1) slightly decreases sorption of ionic dyes to bacteria. Previous studies of organic biosorption by bacteria are mainly focused on ionizable, polar compounds such as dyes, phenolics, and pesticides. This is probably due to attention on the relatively strong polar/ionic interactions between these solutes and microbial surfaces, and the occurrence of highlevel such chemicals in industrial wastewater. However, direct studies of low-polar organic solutes, including PAHs and chlorinated benzenes, that are also of great environmental importance are very limited in biosorption literature. We herein studied sorption of a nonpolar PAH and a nonpolar chlorinated benzene to commonly found bacteria, Escherichia coli, at different pH and with the presence of transition metals at environmentally relevant, low levels to better understand the potential effects on biosorption of HOCs in the environment. 10.1021/es062343o CCC: $37.00
2007 American Chemical Society Published on Web 03/07/2007
TABLE 1. Water Solubility (SW) and n-octanol-water Partitioning Coefficient (KOW) for Solutesa compound
abbrev.
SW (10-6 mol/L)
log KOW
phenanthrene 1,2,4,5-tetrachlorobenzene pentachlorobenzene
PHEN TeCB PtCB
6.31 5.89 2.63
4.57 4.72 5.18
a
Adapted from ref 24.
Experimental Section Materials. Gram-negative bacteria E. coli ATCC 25922 were initially cultured in 30 mL of LB (Luria-Bertani) broth for 12 h at 32 °C, and then transferred to 3 L of fresh LB broth and grown for another 12 h. The bacteria were separated from the nutrient medium by centrifugation, followed by three times of washing with Type I water (Milli-Q, electric conductivity >18.2 MΩ cm). The bacteria were then mixed with 0.02 M NaNO3 to prepare suspensions (1.2 × 109 cells per mil, and a solid to solution ratio of 0.29 mg L-1 on dry weight basis) for later use in batch sorption experiments. Elemental analysis on freeze-dried E. coli biomass gave: C (43.1), H (6.5), O (30.3), N (12.8), and S (0.5) on a weight basis. Test organic compounds include nonpolar phenanthrene (PHEN, Fluka), 1,2,4,5-tetrachlorobenzene (TeCB, Aldrich), and pentachlorobenzene (PtCB, Aldrich). Values of SW and KOW of the solutes are listed in Table 1. Aqueous stock solutions of Cu2+ at 320 mg L-1, Ag+ at 2.0 × 104 mg L-1, and Fe3+ at 2.1 × 104 mg L-1 were prepared from analysis grade CuCl2, AgNO3, and FeCl3, respectively. Batch Sorption. Sorption of PHEN and TeCB was carried out in PTFE-lined screw cap glass vials of capacity 22 mL with sufficient volume of bacterium suspension to eliminate headspace. Aqueous stock solution of Cu2+, Ag+, or Fe3+ was added to the suspension for dilution to the desired concentration, followed by test organic solute in methanol carrier kept below 0.1% (v:v) to minimize cosolvent effects. The initial concentration range was 2.8 × 10-4-2.3 × 10-3 mmol L-1 for PHEN, and 1.6 × 10-4-1.3 × 10-3 mmol L-1 for TeCB. The solutions were shaken by an orbital shaker incubated at 20 ( 0.5 °C for 18-24 h. This period of time is assumed sufficient to reach apparent sorption equilibrium for both organic solute and heavy metal based on sorption kinetics of PHEN (Figure S1 in the Supporting Information) and Cd2+ (11) to E. coli. Changes in bacterial activity growth curve are expected to be small during sorption experiments in the static condition (without nutrients), and the caused effects on sorption, if any, would be offset for the same batch in comparison studies. The toxicity of metal ions and organic solutes on the E. coli was analyzed using the standard plate count method. The results showed that only Ag+ at the concentration of 3 mg L-1 had the apparent germicidal effect compared to the control with no transition metal or organic solute. There were no visible changes of cell morphology caused by metal addition through the results by scanning electron microscopy (SEM) (data not shown). The pH of bacterium suspension was unadjusted except where noted with 0.1 M HCl and 0.1 M NaOH. No pH buffer solution was used to avoid possible interference with biosorption by the weak organic acid or base contained in such a solution. After centrifugation, the solute TeCB was extracted from an aliquot of the aqueous phase with hexanes and analyzed by gas chromatography (GC) with electron-capture detection (ECD) using a 60 m × 0.25 mm DB-1 capillary column (J&W Scientific). The solute PHEN was analyzed directly by highperformance liquid chromatography (HPLC) with a UV detector using a 4.6 × 150 mm HC-C18 column (Agilent).
Isocratic elution was performed under 80% methanol:20% water (v:v) with a wavelength of 254 nm. Calibration curves included at least seven standards over the test concentration range. When ECD was used, calibration curves were fit to a power law expression to account for the detector response nonlinearity. Adsorbed mass was assumed equal to the difference between added mass and mass in the solution aqueous phase. The pH of bacterium suspension at sorption equilibrium was also measured. In general, the presence of Cu2+ and Ag+ did not change the pH, but Fe3+ lowered the pH by 0.4-0.6 due to the acidic hydrolysis. Metal concentration in the aliquot from centrifugation was measured by atomic absorption (AA) spectrometry (Hitachi Z-8100). To take account of uptake by glassware, calibration curves (data not shown) were obtained from controls receiving the same treatment as the sorption samples but no bacterium adsorbent. A loss of
