Optical characterization of liquid bacteria samples is essential for studying the effect of pollutants on concentration of live bacteria in liquid samples. In contrast to our previous studies [16, 17] where the methods of fluorescence microscopy and optical density (OD600) were used for characterization of liquid bacteria samples, in this work, we deployed fluorescent microscopy for characterization of bacteria immobilized on the surface of screen-printed gold electrode. Fluorescence microscopy images in Fig. 1 show the effect of Pb2+ ions on Shewanella oneidensis bacteria immobilized on modified screen-printed gold electrodes where live and dead bacteria appear as green and red spots, respectively . It is clear that the exposure to 1 M solution of PbCl2 salt for 2 h reduced the number of live bacteria (green spots) and increased the dead ones (red spots). Such experiments were carried out for all three types of bacteria and all analytes used. The results of this study are presented in Table 1 as the numbers of live (green) and dead (red) bacteria on recorded images of identical dimensions.
Analysis of fluorescence microscopy data in Table 1 revealed that E. coli and M. capsulatus (Bath) are badly affected by large concentrations of Pb2+ ions, while S. oneidensis are less affected. The negative effect of atrazine is dramatic and more or less similar for all three bacteria. Hexane, however, did not affect M. capsulatus (Bath), though it inhibited both E. coli and S. oneidensis. Such a behavior of immobilized bacteria is similar to those bacteria in solution . The study of optical density (OD600) results of liquid bacteria samples in Fig. 2 shows the bacterial viability ratios (e.g., the ratios of live to dead bacteria) before and after treatment with large concentrations (1 M) of PbCl2. The results are similar to those of fluorescent microscopy; all bacteria appeared to be affected by PbCl2 though this effect was less pronounced for S. oneidensis. It has to be said that the results of optical density measurements, which are based on light scattering, could be affected by different motilities of the bacteria studied.
The most accurate account of bacteria wellbeing can be obtained from flow cytometry measurements which combine the advantages of both fluorescence microscopy and optical density methods. Typical results of flow cytometry for all three bacteria before and after treatment with 1 M solution of PbCl2 are presented in Fig. 3. In these experiments, bacteria were stained with L7012 Live/Dead Bacterial Viability Kit and appeared on the graphs in Fig. 3 as blue dots (for live bacteria) and orange dots (for dead bacteria). The increase in the dead bacteria counts after exposure to PbCl2 salt (1 M concentration for 2 h) is visually apparent for all three types of bacteria studied.
In addition to that, after PbCl2 treatment, dead E. coli and M. capsulatus (Bath) bacteria appear mostly in the bottom-left quadrant of the graph in Fig. 3a and c, indicating the increase in the bacteria size is most likely due to hyper atrophy of cell membrane or rapture of cell walls. On the contrary, the size of S. oneidensis bacteria was affected much less by PbCl2; dead bacteria appeared slightly enlarged since they were shifted to the bottom-left in Fig. 3b.
Flow cytometry tests were carried out for the other two pollutants, e.g., atrazine and hexane, and the results are summarized in Table 2 as the percentage of live and dead bacteria.
Analysis of these data allowed us to conclude that E. coli bacteria are strongly inhibited by all three pollutants. S. oneidensis bacteria are less affected by Pb2+ ions as compared to the strong inhibition effect of atrazine and hexane. M. capsulatus (Bath) bacteria are badly affected by Pb2+ ions and atrazine, while hexane/ethanol mixture stimulates their growth.
Among the three optical methods used to determine the live and dead bacteria percentage, flow cytometry appeared to be the most reliable and not affected by different motilities of E. coli, M. capsulatus (Bath), and S. oneidensis bacteria. The dead bacteria are not motile and tend to sediment which may affect the results of static fluorescent microscopy and optical density measurements. Nevertheless, the results of optical characterization of bacteria samples provided a background for further study using much simpler electrochemical method.
Electrochemical Study of Bacteria in Solution and Immobilized Bacteria Samples
In this work, the effect of Pb2+ ions, atrazine, and hexane on cyclic voltammograms (CVs) of all three bacteria, in both bacteria solutions and immobilized bacteria, was studied. Typical series of CVs recorded on E. coli, S. oneidensis, and M. capsulatus (Bath) samples are shown in Fig. 4. The graphs of CV in Fig. 4 are almost featureless in the selected voltage range from − 0.5 to + 0.5 V, which was chosen deliberately in order to avoid electrochemical reactions on the electrodes, with both cathodic and anodic currents just beginning to rise. The values of both cathodic and anodic currents at − 0.5 V and + 0.5 V, respectively, depend on the bacteria concentration in solution [16, 17]; however, the effect on anodic current is more pronounced and it is therefore used for analysis in this work. The experiments are repeated several (3 to 5) times and show similar results.
In Fig. 4, CV cycles appear to shift upwards upon increasing the pollutant concentration from 0 (untreated bacteria) to 0.1 mM, 1 mM, 10 mM, 100 mM, and 1 M. The characteristic parameter in this study, e.g., the value of anodic current at + 0.5 V, increases with the increase in pollutant concentration for all three bacteria in both liquid and immobilized forms. This means that the electrical conductivity is controlled by bacteria adsorbed on the surface of screen-printed gold electrodes and acting as an insulating layer reducing the current. The correlation between bacteria concentration and the electric current (or conductivity) values is very important for further studying the effect of pollutants, and such measurements were always carried out first [16, 17]. The presence of pollutants (Pb2+ ions, atrazine, and hexane in our case) causes the damage of bacteria cells, and therefore bacteria became less insulating, in turn leading to the increase in the anodic current, which is observed in Fig. 4.
To analyze the effect of pollutants on electrical properties of immobilized bacteria, the values of anodic current (IA) at + 0.5 V from CV measurements were normalized by the currents values of uncoated electrodes in PBS with the addition of a particular pollution of particular concentrations (IA0) to construct the values of relative changes of anodic current ΔIA/IA0 = (IA − IA0)/IA0. For example, for S. oneidensis bacteria treated with 1 mM solution of PbCl2 (Fig. 4f), the reference was recorded on uncoated electrodes in PBS containing 1 mM of PbCl2.
The relative changes in anodic current are presented in Fig. 5 for all three bacteria studied as concentration dependences of the three pollutants. As one can see, the effects of PbCl2, atrazine, and hexane on S. oneidensis, M. capsulatus (Bath), and E. coli are completely different. E. coli appeared to be affected by PbCl2, atrazine, and hexane even at low concentrations since the ΔIA/IA0 values increase monotonically in Fig. 5a, b, and c, respectively. This means that E. coli is equally inhibited by all three pollutants and becoming less electrically resisting. In contrast, S. oneidensis is almost unaffected by PbCl2 at low concentrations of all pollutants up to 10 mM, and then ΔIA/IA0 started to increase at high concentrations of 100 mM and 1 M. Such a behavior of immobilized E. coli and S. oneidensis bacteria is similar to those free in liquid as reported in . M. capsulatus (Bath) respond to PbCl2 (Fig. 5a) and atrazine (Fig. 5b) similarly to the other two bacteria studied though the changes in ΔIA/IA0 are more pronounced at high concentrations, particularly for atrazine. However, M. capsulatus (Bath) bacteria are not affected by hexane (see Fig. 5c) even at high concentration; moreover, an overall trend to small decrease in ΔIA/IA0 is observed. Such a behavior was expected since M. capsulatus (Bath) consume some hydrocarbons .
The results presented in Fig. 5 show a possibility of pattern recognition of the effect of the three pollutants studied. An attempt of pattern recognition has been done by presenting the relative responses of the three channels, e.g., three bacteria (E. coli, M. capsulatus (Bath), and S. oneidensis) immobilized on three screen-printed electrodes, to the three pollutants (PbCl2, atrazine, and hexane) in a pseudo-3D plot in Fig. 6.
The experimental points for PbCl2, atrazine, and hexane in concentrations up to 100 mM shown in different colors are well-separated in this 3D graph. This is a clear indication that pattern recognition principles can be applied for identification of pollutants using different types of bacteria. The concentration of pollutants could be evaluated too using the appropriate calibration and data extrapolation.
Discussion of the Results of Optical and Electrochemical Study
The observed effects of the above pollutants on the three selected bacteria are somehow expected. In general terms, different chemicals of both organic and inorganic origin may affect microorganisms in two possible ways, e.g., acting as either catalyzers enhancing bacterial metabolism or as inhibitors having an opposite effect of reducing bacteria metabolism and even damaging bacteria membranes and causing their death.
In our case, E. coli is obviously inhibited by the pollutants used. This results in the reduction of live bacteria concentration which was confirmed by optical study. Consequently, the increased number of damaged or dead bacteria reduces their insulating properties, thus causing an increase in both anodic and cathodic currents.
Shewanella oneidensis bacteria are known to be tolerant to heavy metals in low concentration, which may have even growth-stimulating (catalytic) effects , which can be used in water treatment ; high concentrations of heavy metals are damaging. This explains the observed immunity of S. oneidensis to heavy metals at low concentrations, while other pollutants are still acting as inhibitors. M. capsulatus (Bath), in contrast, are known by their abilities to use some organic chemicals (hydrocarbons, alcohols) as food , and therefore are used in sewage treatment . In other words, M. capsulatus (Bath) bacteria are catalyzed by some petrochemicals, while heavy metals and pesticides are still acting as inhibitors. The optical and electrochemical study of both S. oneidensis and M. capsulatus (Bath) showed the characteristic changes, respectively, in the live bacteria concentration and anodic current in line with their expected catalytic inhibition patterns.
Combining the above three types of bacteria in a sensor array was logical and therefore enabled the array to identify the type of pollutants. This could be achieved using optical methods with flow cytometry being perhaps the most suitable method for this task. However, very simple electrochemical measurements of anodic current could do a similar job at a substantially reduced cost. Modified screen-printed electrodes with immobilized bacteria can be prepared in advance and kept active for few weeks when stored in a fridge. Such electrical tests can be used for quick preliminary analysis of water samples; the samples indicating a presence of certain pollutants can be passed to specialized laboratories further for more detailed and accurate testing. The overall cost and time of analysis will be substantially reduced as a result.
The sensor stability depends on the activity of immobilized bacteria. We found that bacteria were still alive and active after 24-h storing in the fridge (4 °C), and after 48 h, the live bacteria concentration slightly (10–15%) reduced, and after 72 h, reduced further to over 50%. Therefore, we can conclude that currently the sensor stability is limited by 24 h. Ideally, the electrodes with freshly immobilized bacteria have to be used for sensing.