Background

Estrogens are biologically active hormones that are derived from cholesterol and released by the adrenal cortex, testes, ovary and placenta in humans and animals [1]. Natural estrogens including estrone (E1), 17β-estradiol (E2), estriol (E3), and 17α-estradiol (17α) [2], are mainly derived from human and livestock excretion [36]:

$${EEQ}_{bio}=\frac{{EC}_{20,ref}}{{EC}_{20,sample}},$$

where EC20,sample is the concentration corresponding to the sample when the estrogenic activity is equal to the estrogenic activity at the EC20 concentration of E2.

Results

Cytotoxicity

Cytotoxicity is a disturbance that can affect the assessment results of estrogenic activity [37]. As can be seen from Additional file 1: Fig. S1, the samples from the plantation village did not exhibit cytotoxicity to H4IIE cells, with cell survival rates amounted above 80%. In the industrial village, samples FYS1, FYS2, FYS4, and FYW3 induced slightly cytotoxic effects, with cell survival rates of 74.79%, 78.92%, 77.62%, and 72.73%, respectively (Additional file 1: Fig. S2). Among the samples collected from the farming village, only GMS3 and GMW1 showed significantly cytotoxicity (Additional file 1: Fig. S3). The cell viability sharply decreased starting from an exposure concentration of 2.5 g/L and reaching below 50% at an exposure concentration of 20 g/L in GMW1. With GMS3, the cell viability slightly drop below 80% only at the highest exposure concentration (20 g/L). Overall, cytotoxic effects on H4IIE cells were detected at the highest exposure concentration, leading to cell survival rates of 77.05% and 33.87%, respectively.

Estrogenic activities

The estrogenic activity of soil extracts was assessed through the YES test employing two-hybrid yeast screen systems. The concentration of soil extract varied from 0.39 g/L to 100 g/L. Concentration–effect curves were generated, and the samples’ estradiol equivalents (EEQ) were computed based on the EC20 of E2 [38]. As shown in Fig. 2, estrogenic activity was detected in all samples collected from the plantation village, with EEQ ranging from 0.21 to 2.24 μg/kg dry weight (Additional file 1: Table S2). Sample DLS2 exhibited the highest estrogenic effect, while DLS3 demonstrated the lowest estrogenic effect, with the former’s EEQ being ten times more than that of the latter.

Fig. 2
figure 2

Concentration–response curves in the YES assay for the soil in the plantation village (black closed circles), and the E2 standard (red closed circles). Concentration values on the x-axis refer to the actual exposure of the sample in the experiment. PC: positive control, DLS: Daling Summer

Full size image

Concerning the industrial village, the mean EEQ values were 0.73 μg/kg d.w. in winter samples and 0.52 μg/kg d.w. in summer samples, with no significant difference. FYS6 exhibited the strongest estrogenic effect, while FYW3 and FYW5 did not show significantly estrogenic effects (Fig. 3).

Fig. 3
figure 3

Concentration–response curves in the YES assay for the soil in the industrial village (black closed circles), and the E2 standard (red closed circles). Concentration values on the x-axis refer to the actual exposure of the sample in the experiment. PC: positive control, FYS: Fengyuan Summer, FYW: Fengyuan Winter

Full size image

Regarding the farming village samples, GMS1 demonstrated the highest estrogenic activity (EEQ = 11.3 μg/kg dry weight), followed by GMS2, which is closer to the farms. Conversely, samples of GMS3, GMS4, and GMS5 showed mild estrogenic activity and were collected from the riverbank along this village (Fig. 4). The research findings of Song et al. indicate that the concentration of E1 in the soil near a farm in Shenyang, China, reached 15.15 μg/kg d.w., which is similar to the results of this study [39].

Fig. 4
figure 4

Concentration–response curves in the YES assay for the soil in the farming village (black closed circles), and the E2 standard (red closed circles). Concentration values on the x-axis refer to the actual exposure of the sample in the experiment. PC: positive control, GMS: Goumen Summer, GMW: Goumen Winter

Full size image

The EEQ of the summer samples from the farming village is significantly greater than those from the plantation and industrial villages. The average EEQ between the industrial village and the plantation village did not show significant difference (Fig. 5).

Fig. 5
figure 5

Box chart of estrogen equivalent (EEQ) comparison of samples from different villages in summer. FY: FengYuan village; GM: Goumen village; DL: Daling village. The different letters above the column indicate significant difference at the p < 0.05

Full size image

After fractionating sample DLS4, six fractions (G1–G6) were obtained. Estrogenic effect testing revealed that G2 exhibited estrogenic activity (EEQ of 0.28 μg/kg d.w.), while the other fractions did not demonstrate estrogenic effects in significant level. The ∑EEQ of the fractions accounted for 19.31% of the original sample’s EEQ. Nakada et al. [40] found that estrogenic activity in Fraction 3 of municipal sewage treatment plant (STP) secondary effluent accounted for 10% of the total EEQ of the original sample, which is comparable to the results of this study.

Non-target screening

Hollender et al. [41] defined the general workflow of non-targeted screening as: sampling, analysis, data pre-processing, prioritization, identification. In this study, the Soxhlet extraction method was chosen for sample processing to extract compounds from the soil. Sample analysis employed UPLC-QExactive Plus Orbitrap-HRMS, and data pre-processing utilized Thermo Fisher's Compound Discover 3.2. Estrogen-active compounds were identified as the highest priority, and all identified estrogen-active compounds were classified based on Schymanski et al. [42].

Considering sample DLS4 (EEQ of 1.45 ng/g d.w.) is in the middle of the range, and its environment is more representative of a rural environment, DLS4 was selected as the sample for the NTS. To identify the pinpoint the estrogenic effects induced compounds, the sample of DLS4 was selected and NTA was employed for DLS4 and its six fractionations. A total of 159 chemicals in sample DLS4 were identified, with detailed information available in Additional file 1: Table S3. From Additional file 1: Fig. S4, it can be observed that each fraction was screened for 13–45 organic chemicals, with the highest quantity found in G2, consistent with the estrogenic effect results. Out of these 159 chemicals, 7 estrogenic compounds, including triphenyl phosphate (TPhP) [43], bis (2-ethylhexyl) phthalate (DEHP) [44], indole [45], daidzein [46], genistein, naringenin and glycitein [47] were confirmed. The fractions detection indicated TPhP and indole were found in G2, DEHP in G4, and indole in G6. Conversely, no estrogenic active substances were identified in the remaining fractions (G1, G3, and G5).

Discussion

Potential risk of cytotoxicity

In the current study, GMW3 exhibited the strongest cytotoxicity, while GMW1, GMS1, and GMW2, which are closer to the poultry farm, did not show significant cytotoxic effects. The reason may be due to the pollutants transport along the river the longitudinal gradients. The GMW3 is relatively closer to the chicken farm and may be influenced by the wastewater and feces from the chicken farm [48]. Soil undergoes changes in physical, chemical, and biological properties during the freezing process, slowing down the migration and transformation of organic pollutants in the soil [49], leading to the accumulation of organic pollutants in the soil, consistent with the cytotoxicity results of this study.

Effects of soil environment on estrogenic activities

Chicken and duck manure contain a significant amount of natural and synthetic estrogenic compounds [50, 51], leading to estrogen pollution in farms and the surrounding soil. Hence, the primary origin of estrogenic-active compounds in GMS1 and GMS2 is predominantly livestock farming excreta. The predominant source of estrogens in livestock farming comprises excretions and blends of steroids derived from raw materials or veterinary medicine [52, 53].

The agricultural industry plays a crucial role in Liaoning province of China, encompassing the cultivation of various crops including soybeans, corn, and wheat, etc. The northeastern region of China experiences long and harsh winters [54], where low temperatures and insufficient rainfall during the winter limit the growth of crops. Therefore, this region engages in more frequent agricultural activities during the summer. Studies have indicated that certain pesticides may induce estrogenic activity [55, 56], and extensive pesticide use is involved in summer agricultural practices in this region [57]. Chemical pesticides and fertilizers used in agricultural activities may contain estrogen-active compounds. These compounds can enter water bodies, accumulate in crops, and ultimately enter the human body, posing risks to human health. Therefore, the estrogenic risks associated with pesticides and irrigation wastewater should not be overlooked.

The application of animal manure to agricultural land has been identified as a main source of estrogens in the environment [58]. In urban areas, livestock and poultry are typically raised on a large scale, and waste generated undergoes centralized management. In rural areas, however, most farming is done by individual households, and the waste is often disposed of openly or directly released into fields. While the former generates a larger quantity of waste, it generally causes minimal or no pollution to the environment after proper treatment. The latter, on the other hand, can result in more significant pollution. Rural areas predominantly consist of open soil, leading to faster migration of estrogen-active compounds between soil, surface water, and groundwater. In contrast, the presence of hardened roads in urban areas slows down the migration process.

Currently, numerous studies have confirmed that plasticizers exhibit estrogenic effects [59,60,61]. Microplastics can adsorb estrogenic compounds. The higher the crystallinity, the lower the adsorption capacity [62], thereby affecting the migration of estrogenic compounds in the environment. Agricultural cultivation involves the extensive use of plastic films. Without effective measures for disposal, aged plastics are more prone to adsorb estrogenic compounds, and accumulate in soil.

Studies have indicated that kaolin and montmorillonite have different adsorption capacities for E2 [63]. It has been reported that the wastewater discharged from mining areas contains nonylphenol, which is a chemical with estrogenic effects [64]. Mining activities can disrupt the original structure and distribution of ores, affecting the migration and transformation of estrogenic compounds in these areas. This may be a significant factor contributing to the substantial differences in estrogenic activity observed in the industrial village.

Freezing can provide a stable environment for soil and reduces the transportation of organic compounds, while creating a fluid environment in the thawed state and promoting the substances’ transportation [65], resulting in a greater concentration of estrogen in soil in summer than in winter. Soil freezing can cause soil expansion and the formation of ice lenses, resulting in soil cracking and an increase in the soil infiltration coefficient [66]. As a result, the decrease of estrogenic active contaminants in winter soils may be due to low hydrophobicity.

Identification of estrogenic active compounds

The detection of two estrogenic compounds (TPhP and indole) in fraction G2, coupled with the ability of this fraction to induce estrogenic effects, suggests that these two compounds may be the primary substances responsible for the estrogenic activity in this fraction. Many bacteria and plants produce substantial amounts of indole, and higher concentrations of indole are found in the excrement of animals such as dogs, pigs, and cattle [67]. It has been reported that derivatives of indole may also contribute to various human diseases, including bacterial infections, gastrointestinal inflammation, neurological disorders, diabetes, and cancer [68]. TPhP, as a flame retardant widely used in various everyday chemical products, is frequently detected in the environment [69, 70]. TPhP accumulates in human and animal bodies, inducing endocrine disruption. It has been reported to induce toxicity to the reproductive systems of wild fish populations at environmental concentrations, pose ecological risk [71]. Moreover, studies have found a significant correlation between the lipid content in the human body and high levels of TPhP [72]. The estrogenic pollution induced by TPhP and indole deserves attention.

Although estrogenic compounds were detected in both G4 and G6, the absence of estrogenic effects in these two fractions may be attributed to their low concentrations, which may not be sufficient to induce estrogenic effects. Some natural estrogenic compounds may have been overlooked during the pretreatment process (Soxhlet extraction), and we will strive to consider these aspects in the future to detect a wider range of estrogen-active compounds. We will conduct further research on these natural estrogenic compounds in the future to explore their effects on the ecological environment and human health.

In the future, we will investigate other identified compounds to determine if they exhibit estrogenic effects and explore the mechanistic reactions they have in comparison to the seven already established estrogen-active compounds.

Conclusions

In this study, the potential ecological risk in soil of the Northeast China was evaluated by cytotoxicity and estrogen effect, among different rural socioeconomic types and between summer and winter. The results indicated that the industrial and farming villages may be cytotoxic to H4IIE rat hepatoma cells, which the stronger cytotoxic effects were found in winter; whereas, the effects of estrogenic were found to be stronger in summer, with significantly difference observed from the farming village (0.1–11.3 EEQ μg/kg d.w.). The estrogenic active compounds were successfully identified by EDA, in which Indole and TPhP were identified from both raw sample and the fraction by NTS, with the explanation of estrogen activity accounting for 19.31% of the raw sample. Therefore, the current study is helpful for preparing measurements for estrogenic risk control.