Background

The interaction between external and internal environmental factor on living organisms has grown wide interest [9, 16, 20]. Till now, scientists have not reached a consensus on how living organisms reacting to this complicated circumstance, such as magnetic field, solar activity indices, temperature, humidity, atmospheric pressure, geomagnetic conditions, etc. [4, 33, 58, 61, 66, 72, 76]. Hence, model experiments make it possible to trace certain patterns of such interactions, which may both have a general physical content [66] and be of interest applications [58]. It was demonstrated that some chemicals could induce the hormesis in vitro by very low-dose environmental factors such as low-dose-rate gamma rays and X-rays [62, 82].

Hormesis, a dose–response relationship characterized by a low-dose stimulation and a high-dose inhibition [11] has drawn increasing interests on dose–response studies, as it has either a harmful, beneficial or indifferent effect in an environment. In recent decades, a growing body of evidence has accumulated on hormetic effects of several chemicals for a number of biological endpoints of cellular systems and organisms [1, 11, 13, 18, 19, 57]. Hormetic effects were typically represented in graphs as an inverted U- or J-shaped dose–response, depending on the endpoint measured [9]. This dose–response represents an evolutionarily conserved process of adapting to changing environments, potentially beneficial responses to low doses of a stressor agent/condition [10] which would make the current ecosystem more balanced [15].

On the other hand, studies on the reaction of biological systems to magnetic fields (MF) are of great interest for not only fundamental science (ecology and biophysics) but also applied science, as some aquatic ecosystems become subject to electromagnetic pollution due to an increase in output of anthropogenic sources of low-frequency electromagnetic fields (EMF) [8]. Weak magnetic field (WMF) is of special interest among weak environmental factors that influence living systems, such as weak low-frequency EMF from electrical equipment. In terms of thermodynamics, the energy obtained in such action is insufficient to reach the thermal noise limit but results in clear repeatable bioeffects [33]. At present, the research on WMF in the field of environmental science (has focused on or) focus on water treatment technology. Application of zerovalent iron (ZVI) has been proved to be an environmentally friendly approach for heavy metal removal from water due to its low cost, simplicity in handling, and scalability in aquatic ecosystems [32, 39]. It has been proved that ZVI corrosion could be promoted by WMF over a wide pH range, resulting in more rapid release of some heavy metals [37, 63], such as the effect of WMF on metal (such as Cu [28, 31], Sb [35, 70], Cr, As [63, 64], Se [37]) removal from water by ZVI. But the ultimate application of these technologies is oriented to the aquatic ecosystem, although not at present. Nevertheless, the mechanisms of how WMF and low-frequency EMF influence on living systems are seldom discussed as well as no commonly accepted models regarding to the biological effect [7, 45, 64, 68, 71, 80]. Therefore, it is of great significance to explore whether WMF affects the dose–response relationship of environmental pollutants and its toxic effects.

Polymyxin B sulfate (POL), an antibiotic widely used in the medical profession and the general public health, which is mainly used for the infections caused by sensitive bacteria and the urinary tract infection caused by pseudomonas aeruginosa, meningitis, sepsis, burn infection and skin mucosal infection, and so on [5, 23, 24, 34]. It was found that POL has weak time-dependent hormetic dose/concentration–response profile on Vibrio qinghaiensis sp.-Q67 (Q67) [25], which reminds us to think whether the WMF can influence the hormetic relationship of POL on Q67 or not. The main purpose of this paper is to examine the effect of WMF on the hormesis of POL by means of setting up different MF treatments in the time-dependent toxicity tests.

Materials and methods

Chemicals

Polymyxin B sulfate was purchased from TRC (Canada, CAS: 1405-20-5, 95% Purity). The stock solution concentration of POL was 1.521E−05 mol/L. All solutions were prepared with Milli-Q water and stored in darkness at 4 °C before the test. The molecular structure of POL is shown in Fig. 1.

Fig. 1
figure 1

The molecular structure of polymyxin B sulfate (POL) [25]

Full size image

Q67 culture

The luminescent bacterium Q67 (Beijing Hamamatsu Corp., Ltd., Beijing, China) was incubated in liquid culture medium with a shaking speed of 120 rpm at 22 ± 1 °C for 9–10 h, and the bacteria in the logarithmic growth phase were used in experiments [36, 41, 78, 84]. Then, the culture medium containing Q67 was mixed with an equal amount of the twofold concentrated medium [67, 73]. Based on the growth curve in the mixed culture medium, Q67 was further incubated for about 40–60 min.

Time-dependent toxicity test

A time-dependent microplate toxicity analysis (t-MTA) is exactly the same as that in the previous works [55, 56, 81]. For the test chemicals, 12 diluted concentration series in 3 parallels and 12 controls were arranged in the white 96-well standard opaque plates with 12 rows and 8 columns (Corning Corp.). To avoid possible edge effects, the first and eighth rows, together with the first column and the last column (36 wells in total) were filled with 200 µL distilled water. For the remaining 60 wells, 24 wells from the second, sixth, seventh, and eleventh columns were treated as controls. The rest of the plate (36 wells) was arranged with 12 concentration series of the toxicants, and it was divided into 2 sections (a left one and a right one), both of which had 6 rows and 3 columns. The left section was arranged with 6 higher concentration series, and each concentration used 1 row with 3 wells as parallel. The right section was arranged with 6 lower concentration series. Each of the 60 wells contained 100 µL toxicants or distilled water (as controls). The prepared bacteria were added into the 60 wells to make the final test volume be 200 µL in each well [81].

To evaluate the toxic effects of POL at different times and various concentrations, the relative light unit (RLU) of every well was determined on the Power-Ware microplate spectrophotometer (American BIO-TEK Company) at 22 ± 1 °C. During exposure, readings were taken at 0.25, 3, 6, 9 and 12 h. Inhibition ratio of bioluminescence was used to characterize the toxicity or toxic effect, noted as E,

$$E = \frac{{I_{0} - I}}{{I_{0} }},$$
(1)

where I0 is the average RLU of Q67 exposed to the control groups and I the average RLU of Q67 exposed to the experimental groups.

For the monotonic concentration-effect (C-E) data in different exposure times, the concentration-effect/response curves (CRCs) are modeled by the nonlinear least squares fit [42] and for the non-monotonic C-E data, the CRCs are fitted by the least squares support vector regression (LS-SVR) procedure [53] or a method for the fitting and prediction of J- and S-shaped concentration–response curves (JSFit) program [69]. The goodness of fit is expressed as the determination coefficient (R2). The analysis of variance (ANOVA) (Origin Pro 7.5, Origin Lab Corp., USA) was carried out among the results from independent experiments, and the significance levels of 0.05 (P < 0.05) were considered statistically significant.

Weak magnetic field treatment

The static, nonuniform magnetic field was supplied by a cuboid permanent magnet of 40 × 20 × 3 mm3. Figure 2 presents a diagrammatic sketch of experimental device in the process of bacterial culture (Fig. 2a) and POL exposure (Fig. 2b). In the process of bacterial culture, six permanent magnets were set at the front (one magnet), back (one), both sides (two), and bottom (two) of the beaker flask of 100 mL in which there is 33 mL bacteria suspension, respectively (Fig. 2a). In the process of chemical exposure, six permanent magnets were also set at the top (two magnets), bottom (two) and both sides (two) of the microplates, respectively (Fig. 2b). The maximum magnetic intensity was determined to be ~ 20 mT with six permanent magnets.

Fig. 2
figure 2

The diagrammatic sketch of a laboratory-scale device with WMF where a illustrates the size and layout of the permanent magnets in exposure duration and b depicts the layout of the permanent magnets during the Q67 culture duration

Full size image

In this study, four WMF experimental groups, one without WMF (NW) in all cases, one with WMF (AW) in all cases, one with WMF only during the bacterial culture period (BW), and one with WMF only in POL exposure period (EW), were created.

Results

Concentration–response relationships of POL

The concentration-inhibition data of POL in four WMF treatments at different time points can be fitted by the LS-SVR [53]. The fitted concentration–response curves (CRCs) were subsequently used to calculate three main characterization parameters of the hormetic concentration–response curve (hCRC) such as the median effective concentrations (EC50), the minimum inhibitory effect (Emin) and its corresponding concentration (ECmin). The fitted regularization parameter (gam), kernel function parameter (sig2), determination coefficient (R2), and three characterization parameters, EC50, Emin, and ECmin are listed in Table 1.

Table 1 The fitting parameters (gam, sig2 and R2) of various LS-SVR models and three characterization parameters (EC50, Emin, and ECmin) from the CRCs of POL in four WMF groups at five times
Full size table

From Table 1, R2 is higher than 0.980 (one exception), which indicates that all J-shaped hCRCs are well fitted by LS-SVR procedure. Taking EC50 as a toxicity index, apart from the toxicity of POL in four WMF treatments at the first time point (0.25 h) is less, those at the other four time points are almost the same. ECmin in all WMF treatments at all time points are around 1.00E−07 mol/L, which implies that both the WMF and exposure time do not alter the minimum inhibitory effective concentration. However, both WMF and time can alter the maximum simulative effect (Emin) of POL on Q67. The concentration-inhibition profiles of POL in four WMF treatments at different time points are shown in the three-dimensional (3D) CRCs (Fig. 3).

Fig. 3
figure 3

The three-dimensional (3D) concentration–response curves of POL at five exposure times (0.25 h, 3 h, 6 h, 9 h, and 12 h) in four WMF groups (NW, AW, BW, and EW)

Full size image

In each WMF treatment (Fig. 3), five CRCs of POL at different exposure times are different from each other, which depict the time-dependence of CRCs again. Obviously, the latter three CRCs in BW and EW groups show more significant J-shape nature than those in NW and AW, their simulative effect being greater, which demonstrates that the WMF indeed enhance the hormesis, significantly improving the simulative effect of POL.

Hormetic effects of POL

Polymyxin B sulfate shows hCRCs on Q67 at 0.25, 3, 6, 9, and 12 h in all WMF groups. Emin is from 3.6% (at 3 h in AW) to 26.8% (at 6 h in BW) and ECmin from 7.508E−08 (at 0.25 h in NW) to 7.286E−07 (at 12 h in BW) (Table 1). Notably, EW owns the maximum stimulative effect at 6 h, Emin = 26.8%; BW has the maximum stimulative effect at 6 h, Emin = 20.7%; while Emin of NW is 14.6% at 9 h (Fig. 4a). That is, stimulations of POL in BW and EW are earlier and stronger than those in NW. However, the stimulative effect enhanced by weak magnetic field did not occur in AW (Fig. 4b).

Fig. 4
figure 4

The minimum inhibition ratios of different WMF groups (NW, AW, BW, and EW) at the same time (a) and different times (0.25, 3, 6, 9, and 12 h) in the same WMF group (b)

Full size image

The Emin of POL to Q67 in BW and EW are significantly larger than those in NW and AW (P < 0.05) at 6, 9, 12 h. The Emin of POL in EW and BW are, respectively, 26.8% and 20.7% at 6 h, being 183.6% and 141.8% of the maximum stimulative effect in NW at 9 h (14.6%).

Discussion

Hormesis, ordinary or not?

Figure 1 presents the time-dependent hormesis of POL on Q67. Fan et al. [25] and Liu et al. [40] found that some antibiotics did not have hormetic CRCs on Q67. However, some ionic liquids and personal care products showed hCRCs on Q67 [55, 74, 75], implying that the different model organisms and different toxicity endpoint may result in different CRCs. For example, de Vasconcelos et al. [21] reported that some pharmaceuticals showed hCRCs on Vibrio fischeri and Desmodesmus subspicatus at low concentrations. Radak et al. [57] observed that physical exercise could evoke the hormetic curve response by the organism. Zhu et al. [85] found that some imidazolium-based ionic liquids exhibited hCRCs when took the lifespan of Caenorhabditis elegans as the toxicity endpoint. Many ionic liquids showed hormesis on different organisms, such as anaerobic Clostridium sp. and aerobic Pseudomonas putida [47], microalga Scenedesmus quadricauda [22], and firefly luciferase [27].

More recent pieces of the literature have demonstrated that hormetic effects occurred widespread with relatively high generalizability, reproducibility, frequency, and with a solid mechanistic foundations [12, 17]. Experiments showed that the damage and death of Q67 caused by gamma-ray radiation, which is the important reason for the inhibition of Q67 luminescence intensity [79]. However, the hormesis of Q67 occurred with low-dose gamma radiation [83]. Thus, hormesis is a common phenomenon. It could be found with different model organisms and different toxicity endpoints under different experimental conditions.

The omission of weak environmental factors—WMF

Table 1 shows that the toxicity of POL is basically unchanged with exposure time (except 0.25 h), but the intensity of stimulative effect increases significantly by WMF (more negative). Obviously, WFM, as a weak environmental factor, plays an important role in the enhancement of stimulative effect.

In general, any changes in the experimental conditions will have an impact on organisms. The report provides evidence that the greater number of concentrations employed in the low-dose zone, the better the characterization of the hormetic stimulatory response [14]. External factors, i.e. environmental parameters, are beyond the control of the experimenter (magnetic field, solar activity indices, temperature, humidity, atmospheric pressure, geomagnetic conditions, etc.), while internal factors are genetic drift, instability of genomic transposable elements, as well as other aspects of genetic variation in populations, which are also practically uncontrollable [30]. Some scholars found that power–frequency electric and magnetic fields may cause interference with cardiac pacemakers [52, 77]. Krylov et al. [33] found that developing parthenogenetic eggs of Daphnia magna exposed to a number of low-frequency electromagnetic fields with indicated parameters have shown accelerated rates of embryonic development. Izmaylov et al. [30] have demonstrated that atmospheric pressure and geomagnetic activity are significantly correlated with the parameters of lifespan distributions in Drosophila melanogaster laboratory populations.

More and importantly, the report showed that neuronal activity might be affected by magnetogenetics, such as the direction of the magnetic field and the external magnetic field applied [43]; while a magnetic protein biocompass is well created, which is a protein complex and may form the basis of magnetoreception in animals, and may lead to applications across multiple fields [54]. Hence, there may be a type of magnetoreceptor in Q67, and the magnetoreceptor may activate the bioluminescence signal pathway and show stimulate effect. But we have no idea if POL could be as a bridge of this activation due to the WMF is never considered in other chemicals that show hormetic effects.

Enhanced hormesis of POL

Weak magnetic field is of special interest among weak environmental factors that influence living systems. Energy of such action is insufficient to exceed thermal noise limit in terms of thermodynamics, but clearly repeatable bioeffects take place nevertheless [33]. However, certain mechanisms of WMF’s influence on living systems are actively discussed at present, and there are still no commonly accepted models of biological effect of this factor [6, 7, 26, 51, 65].

With the WMF treatment, it is obvious that the stimulative effect of POL increased (in BW and EW). However, there was no significant difference between the AW and NW. It means that different periods of Q67 with the same WMF treatment (especially BW and EW) have different hormesis enhancement, but the same periods of Q67 with different WMF treatment (AW and NW) have no significant difference hormesis enhancement.

At the same time point (6 h), the maximum stimulation (26.8%) occurs in EW, while the second maximum stimulation (20.7%) occurs in BW. The hormesis of POL enhanced significantly by WMF. In animal model systems, it has been reported that WMF generally stimulated ornithine decarboxylase activity and cell proliferation [38, 46].

The increase of inflexibility caused by the large ring structure of POL (20-member ring of heptapeptide [24]) may lead to some directional action (such as permeability changes [44]) during the WMF treatment. The possible mechanisms underlying the contribution of these environmental factors about the hormesis need further investigation. Based on the view of quantum mechanics, “more is different”, the complexity of the relevant mechanism will be beyond imagination [2].

Different periods by WMF treatment

Four experimental groups were constructed with different WMF treatments, including blank control and positive control. NW is blank control, AW is positive control, BW and EW are experimental groups. Research has continuously shown that it usually has a different toxic effect of chemicals on organism during different periods [3, 29, 48, 50, 59].

By comparing the results in AW, BW and EW with that in NW, hormesis enhancement is obtained during different WMF treatments. Q67 was further grown in the liquid medium when maintained for another 12 h to reach the logarithmic phase (BW), during this time, WMF may affect the metabolic activity of cells, leading more sensitive cell activities. This process is similar to domestication. Moreover, quorum sensing [49] signal pathway is modified to clarify the phenomenon when Q67 was exposed to POL in the toxicity test. However, EW may not necessarily so. It is likely that WMF has a protective effect on Q67 when exposing to low concentration of POL, thus an amplified direct stimulation [60] is expected to aim at protection. Because of little literature on this field, the true rule is still unclear, and it is difficult to explore the relevant mechanisms. It needs further investigation.

Conclusions

Taking POL as object chemical and Q67 as target organism by the method of t-MTA, with four different WMF treatments (NW, AW, BW, and EW), the results showed that with the treatment of BW and EW the Emin of POL significant increased compared that of NW, it indicated that different growth period of Q67 with the same WMF treatment has significant hormesis enhancement. In contrast, the Emin of AW has no significant difference with that of NW, it suggested that the same growth period of Q67 with different WMF have no significant hormesis enhancement. These findings prove that WMF as a neglectable factor, it actually and awfully enhanced the hormesis of POL on Q67. WMF as a complex factor during the organism growth stage, it may help unravel drivers of mechanisms of inhibition and stimulation.