Tributyltin chloride, varied from 3 μM to 10 mM as a final concentration, was added into the cell suspension, and the number of the Sn atom originated in TBT adsorbed by the resting cells was measured with the accelerator (Figure 1). The total number of the Sn atom linearly increased with an increase in the concentrations of TBTCl externally added up to 3 mM, where the averaged value was to be 10^8.2 cell⁻¹ (n=2). The value became saturated in the presence of more than 3 mM of TBTCl. Since the isolate could not degrade TBT [31], the entire Sn atoms adsorbed are originated in TBT.

In the presence of less than 3 mM of TBTCl, the numbers of Sn atom actually measured with PIXE were smaller than those calculated under a hypothesis that all of TBT added were adsorbed by the cells suspended in the solution. For example, the number of the Sn atom originated in TBT was actually measured to be 10^{7.9 ± 0.13} cell⁻¹ in the presence of 1 mM of TBTCl (n=4). The value on the dotted line should be 10^8.5 cell⁻¹ to the addition of 1 mM of TBTCl externally. This value was about 4 times larger than that obtained experimentally. The difference between the measured and the calculated values was observed for any concentration of TBTCl externally added. The same phenomenon has also been observed for the samples obtained from growing cells [31]. As of now, it is difficult to explain why the values actually measured are smaller than those calculated. A possible explanation is that a certain number of TBT might be desorbed from the cells to reach a new equilibrium condition through the washing process.

Figure 2 shows the changes in the total number of the Sn atom originated in TBT adsorbed by lysozyme-treated cells in relation to the concentrations of TBTCl externally added. All the values, except for the one obtained in the presence of 3 mM of TBTCl, were lower than those on the regression curve shown in Figure 1. It might be available to think that the amount of TBT adsorbed by lysozyme-treated cells reduced by about four-fifth comparing with that by native cells.

Since addition of too much lysozyme into the cell suspension and a long period of enzyme reaction with the cells cause cell lysis, we checked the number of colony-forming cells and morphological changes after the cells were treated with lysozyme (1%) for 1 h at 30°C. After the reaction, no reduction of the number of colony-forming cells was observed. As for the morphological changes, cells became spherical shape (Figure 5B) from rod one (Figure 5A). This morphological change indicates the weakness of cell-wall by the lysozyme treatement. The reduction of the numbers of the Sn atom originated in TBT adsorbed by lysozyme-treated cells seems to be closely related to the loss of cell-wall components.

We compared the numbers of Sn atom originated in TBT adsorbed by the cells suspended in pH 3.0 with those in pH 7.8 (Figure 3). Slight reduction of the values in pH 3.0 was observed comparing with those in pH 7.8 on the regression curve shown in the figure. The maximum reduction, 92.1%, was obtained from the value in the presence of 0.3 mM of TBTCl. In acidic pH, a negatively charged cell surface seems to be shifted to a neutral surface because, for example, a carboxylic group in cell-wall components is protonated in the proton rich environment. The surface charge, however, was not affected by an ionic strength of an electrolyte solution [33]. Therefore, the negative charge on the cell surface plays a minor role for the adsorption of TBT⁺ onto the cell surface in pH 7.8.

We also examined the contribution of a hydrophobic interaction of three alkyl groups originated in TBT to the adsorption onto the cells in pH 7.8 by comparing with that of inorganic SnCl₂ in pH 3.0. When SnCl₂ was added into the cell suspension adjusted to pH 3.0, the numbers of the Sn atom adsorbed were very close to those on the regression curve obtained in pH 7.8, indicating that the contribution of the hydrophobic nature for those butyl groups is weak or negligible to the adsorption of TBT onto the cell surface.

Hoch et al. [34] have examined TBT adsorption onto a mineral, which has ionizable hydroxyl groups on the surface, in pH values varied from 4.0 to 8.0. In this experiment, pH-dependent adsorption of TBT was observed, and the maximum adsorption was found in the range of pH between 6.0 and 7.0. The cation exchange capacity on the surface of the clay was suggested to be closely related to the efficient adsorption of TBT. For the adsorption of TBT by the isolate, Pseudoalteromonas sp. TBT1, the number of TBT adsorbed was slightly affected by the exposure to acidic pH, indicating that the isolate has the capacity of cation exchange on the surface of the cells. As another possibility, a receptor protein which has high affinity to a certain substance might recognize Sn atom itself as the substance.

We used heat-treated cells for the adsorption experiment of TBT (Figure 4A). Throughout the given concentrations of TBTCl up to 3 mM, the numbers of the Sn atom adsorbed by the cells were hardly reduced in comparison with those on the regression curve obtained from native cells (see the solid line in Figure 4). As for the morphological changes of the cells, rod shape remained, but new string-like substances were observed (Figure 5C), which look like proteins leached from the cells. In general, heat-treated cells have lost activities of cell-associated proteins. Therefore, it might be concluded that native proteins in the cell do not function to adsorb TBT.

Ethanol-treated cells were exposed to given concentrations of TBTCl when the cells were suspended in distilled water and the solution which was used for the preparation of resting cells (see Experimental Section) (Figure 4B).

Regardless of the suspension medium, the numbers of the Sn atom adsorbed by the cells were hardly changed and those values were close to the ones on the regression curve. In the case of ethanol-treated cells, the membrane structure has been lost because lipids in the cell membrane have been dissolved into ethanol, indicating that the interaction of the cell membrane with TBT does not contribute to an increase in the adsorption of TBT. Rod shape of the cells was damaged and the surface became rough by the treatment of ethanol (Figure 5D). These results indicate that the function of cell surface, rather than the structure of it, plays an important role to the adsorption of TBT. When the cells, which have already adsorbed 10^7.8 Sn atom cell⁻¹ in the presence of 1 mM of TBTCl, were washed three times with ethanol, the value reduced to 10^6.1 Sn atom cell⁻¹, indicating that the affinity of TBT toward the cells is weaker than the solubility of TBT in ethanol.

We considered the adsorption structure of TBT adsorbed onto the surface of a single cell of Pseudoalteromonas sp. TBT1. We modeled cell shape of the isolate as the combination with two of the half cut of a sphere with the radius of 0.2 μm and a cylinder with the height of 1.7 μm and the bottom area of 0.13 μm² (Figure 6) based on the morphological observation with the scanning electron micrograph (Figure 5A). Following the model, an averaged cell surface area, Sc, was estimated to be 10^6.4 nm².

A molecular structure of TBTCl has been simulated using the ab initio quantum chemistry package GAMESS [35]. An equivalent radius was estimated to be 0.50 nm based on the empirical molecular orbital calculation software, Scigress Explorer and MOPAC therein (Fujitsu Co., Ltd., Japan). Accordingly, the area occupied by a TBT molecule is plausible to think as the half area of a circle with the radius of 0.50 nm.

If we assume here that the state of TBT adsorption is a single layer on the cell surface regardless of the concentrations of TBTCl externally added, then we can define a virtual area, S_o, which could be equal to an occupational area of a single TBT molecule on the cell surface; S_o = S_c / N_Sn. In general, the value of S_o decreases in relation to an increase in the number of the Sn atom, N_Sn, adsorbed by a single cell. The data on the regression curve in Figure 1 were used as the values of N_Sn to obtain S_o. As a result, the virtual area varied from 9.5 nm² for the number of 10^5.4 Sn atom cell⁻¹ at 3 μM of TBTCl to 0.013 nm² for the number of 10^8.3 Sn atom cell⁻¹ at 10 mM of TBTCl (Table 1). An appropriate equivalent area of TBT, which was obtained by using computer software, MOPAC, was to be 0.39 nm². The corresponding number of Sn atom adsorbed by a single cell was 10^6.8 cell⁻¹. Since the maximum number of TBT adsorbed by a single cell was experimentally determined to be 10^8.3 cell⁻¹, stacking of TBT molecules onto the TBT-covered cell surface seems to occur at more than 10^6.8 Sn atom cell⁻¹.

Hermosin et al. [36] examined the adsorption of monobutyltin (MBT) on clay minerals, and suggested that the MBT adsorption process is a cationic exchange on their surface attracted by electrostatic interaction and an additional adsorption of MBT is the hydrophobic interaction with previously adsorbed MBT by the lipophilic alkyl groups. For the adsorption of TBT by a Pseudoalteromonas sp. TBT1 cell, the same adsorption process might be progressed.

A TBT-resistant marine Pseudoalteromonas sp. TBT1 was isolate from the sediment in a ballast tank of a liquefied natural gas carrier (110,000 gross tons) returning to Japan from Qatar [30]. The total capacity of the ballast tanks is 55,000 m³. On the return trip from Qatar to Japan, almost all the ballast tanks were empty because liquefied natural gas was fully loaded; therefore, the sediment in the ballast tank was available to be sampled. Sediment samples were placed into an autoclaved bottle (250 mL) and kept at 4°C until arrival in Japan. After serial dilution with autoclaved seawater, the sample was spread onto an agar plate containing 5 g Bacto peptone (Difco, MD, USA) per liter of natural seawater, 1 g yeast extract (Difco, MD, USA) per liter of natural seawater, and 150 μM of TBTCl. Incubation was carried out for 5 days at 30°C, and a single colony was picked up and stored at 4°C on the plate.

Cells were grown in a medium containing 5 g Bacto peptone per liter (Difco, MD, USA), 1 g yeast extract per liter (Difco, MD, USA), 0.4 M NaCl, 10 mM KCl, 10 mM CaCl₂ 2H₂O, 53 mM MgCl₂, and 28 mM Na₂SO₄. The medium pH was adjusted to 7.8 using tetramethylammonium hydroxide (TMAH). After pre-incubation for one day at 30°C in the medium, incubation was carried out by the addition of a cell suspension to the medium to give a one-thousandth dilution.

Resting cells were prepared from cells grown until the early stationary phase. After 1 mL of cell suspension was harvested by centrifugation (10,000g, 5 min), cells were washed twice with a solution containing 0.4 M NaCl, 10 mM KCl, 10 mM CaCl₂ 2H₂O, 53 mM MgCl₂, and 28 mM Na₂SO₄, where the pH was adjusted to 7.8 using 10 mM 2-[4-(2-hydroxyethyl)-1-piperazinyl] ethanesulfonic acid (HEPES) – TMAH. The cells thus obtained were resuspended in 1 mL of the same solution, unless otherwise noticed.

Resting cells suspended in the solution described above were treated with lysozyme, which hydrolyzes preferentially the β-1,4 glucosidic linkages between N-acetylmuramic acid and N-acetyl-glucosamine in the mucopeptide cell wall, for 1 h at 30°C prior to the addition of TBTCl at given concentrations. This process did not reduce the number of colony-forming cells. The lysozyme-treated cells were washed twice and resuspended in the solution described above. Then TBTCl was added to give final concentrations of up to 10 mM. After vigorous shaking for 1 min, cells were washed three times with distilled water. The samples thus obtained were stored at 4°C until use.

Acidic exposure to the resting cells was carried out for 1 h at 30° C by the addition of an acidic solution into a cell pellet. The acidic solution contained 50 mM glycine, and pH was adjusted to 3.0 by HCl. Given concentrations of TBTCl, or SnCl₂, were added into the sell suspension. After vigorous shaking for 1 min, cells were washed three times with distilled water. The samples thus obtained were stored at 4°C until use.

Resting cells suspended in the solution were heated for 1 h at 65°C. After cooling down, TBTCl was added in there to give final concentrations of up to 10 mM. After vigorous shaking for 1 min, cells were washed three times with distilled water. The samples thus obtained were stored at 4°C until use.

Ethanol-treated cells were obtained from the cells exposed to ethanol for 1 h at 30°C. Prior to the addition of TBTCl, the cells were washed twice with distilled water, or the solution described above. After the addition of TBTCl at given concentrations, they were vigorously shaken for 1 min and then washed three times with distilled water. After vigorous shaking for 1 min, the sample in which cells were suspended with the solution in the presence of 1 mM of TBTCl was washed three times with ethanol. The samples thus obtained were stored at 4°C until use.

Serial dilution of the sample was carried out with the solution described above after the cells were treated with or without lysozyme. Then cells were spread onto the agar plates containing the same composition as used in the growth medium. Plates were incubated at 30°C for 1 day, and the colonies on the plates were counted. The data are shown as colony-forming units (CFU) per mL.

Each sample was dropped onto a hollow (2 mm diameter × 0.5 mm depth) in a carbon plate (20 mm × 100 mm × 2 mm thickness) by pipetting and gently dried for about 40 min. Some repetitions of this process were required to finish loading the sample, during which process the volume of the sample was drastically reduced in the hollow [32]. A droplet containing 10 nmol of Sr(NO₃)₂ was added into each sample as an internal reference prior to loading the sample onto the hollow.

Quantitative analysis of the Sn atom originated in TBTCl molecule adsorbed by the cells was performed by the analysis of particle-induced X-ray emission (PIXE). The samples were analyzed in a vacuum chamber connected to the M30 beam line of a 5SDH-2 Pelletron accelerator under a pressure of 1.0 × 10⁻⁴ Pa. The sample target [31] was exposed to a probe beam of 3.0 MeV protons at an incident angle of 0° with a beam current of 2 nA and a beam diameter of 1 mm. Proton incidence of 3.0 μC was necessary to obtain reasonable statistical accuracy of the X-ray yields from Sn and other atomic species measured under the present condition. A Si-PIN photodiode detector was positioned at a 135° angle to the incident direction of the probe beam. A piece of aluminum foil with 90 μm thickness was mounted in front of the detector to reduce low-energy background X-rays and scattered protons. The solid angle subtended by the Si-PIN detector was 5.4 × 10⁻² sr.

Figures in the text show the relationship between the concentrations of TBTCl externally added to the cell suspension and the amounts of Sn originated in TBT adsorbed by a single cell. Major experimental error in the estimation process of the number of Sn adsorbed by a single cell seems to occur when the value of the number of Sn measured with the accelerator was divided by the number of colony-forming cells used in the experiment because, in general, the colony-counting method described above contains the experimental error of within 10^0.20 CFU mL⁻¹.

Cells were fixed in 1% (vol/vol) glutaraldehyde for 1 h at 4°C; dehydrated once for 1 h in 50, 70, 90, 95, and 100% of ethanol; and suspended in 100% of t-butyl alcohol. After being freeze-dried, they were pasted on a carbon tape and coated with Pt-Pd particles under vacuum. The samples thus obtained were observed with a scanning electron microscope (XL30 CP, FEI Company, Eindhoven, The Netherlands).

Tributyltin chloride (95%) and SnCl₂ (97%) were purchased from Wako (Osaka, Japan) and dissolved in ethanol and in 50 mM of glycine-HCl, pH 3.0, respectively. The latter compound does not dissolve in neutral to alkaline pH. Other chemicals were analytical grade.