Evaluation of Toxicity and Genotoxicity of Concrete Cast with Steel Slags Using Higher Terrestrial Plants

The potential impact of concrete mixtures containing steel slag (SS) as a partial replacement of natural aggregates (NA) on the terrestrial ecosystem was assessed using a battery of plant‐based bioassays. Leaching tests were conducted on four concrete mixtures and one mixture containing only NA (reference concrete). Leachates were tested for phytotoxicity using seeds of Lepidium sativum, Cucumis sativus, and Allium cepa. Emerging seedlings of L. sativum and A. cepa were used to assess DNA damage (comet test). The genotoxicity of the leachates was also analyzed with bulbs of A. cepa using the comet and chromosome aberration tests. None of the samples caused phytotoxic effects. On the contrary, almost all the samples supported the seedlings; and two leachates, one from the SS‐containing concrete and the other from the reference concrete, promoted the growth of C. sativus and A. cepa. The DNA damage of L. sativum and A. cepa seedlings was significantly increased only by the reference concrete sample. In contrast, the DNA damage in A. cepa bulbs was significantly enhanced by the reference concrete but also by that of a concrete sample with SS. Furthermore, all leachates caused an increase in chromosomal aberrations in A. cepa bulbs. Despite some genotoxic effects of the concrete on plant cells, the partial replacement of SS does not seem to make the concrete more hazardous than the reference concrete, suggesting the potential use of SS as a reliable recycled material. Environ Toxicol Chem 2023;42:2193–2200. © 2023 The Authors. Environmental Toxicology and Chemistry published by Wiley Periodicals LLC on behalf of SETAC.


INTRODUCTION
Steel slags (SS) are significant by-products of the steel industry.European production in 2018 amounted to 16.3 million tons (Harder, 2020).Steel slags are used as recycled materials in many applications such as road base, asphalt mixtures, and in the construction industry.The use of SS as a partial replacement for natural aggregates (NA) makes it possible to obtain concretes with good workability and mechanical properties (i.e., compressive strength and modulus of elasticity) compatible with their use in civil engineering works (Diotti et al., 2021).Not least, the reuse of SS reduces their final disposal in landfills.
However, the reuse of this material can lead to a potential release of harmful compounds to the environment and humans (Primavera et al., 2016).Few studies have analyzed the toxicity of SS following their recovery/disposal.To address this concern, which has not yet been clearly defined from a regulatory perspective, previous studies have been developed examining the impact of granular SS and NA leachates on living organisms (bacteria, animals, plants), using an integrated chemicalbiological approach (Alias et al., 2021;Benassi et al., 2019).The study of the effects of concrete cast with a partial replacement of NA with SS is a step forward.The direct contact of concrete products with soil raises many ecotoxicological and toxicological concerns.Concrete materials are known to release toxic substances (heavy metals, organic compounds, etc.) into the aquatic compartment.Moreover, building materials are often treated with biocides against microbial colonization and biodeterioration (Reiß et al., 2021).
Soil is a complex matrix, with strong interactions with the other environmental matrices (air and water).In particular, urban soils are the first interface of contamination of several harmful compounds (Modabberi et al., 2018).
The application of short-term toxicity and genotoxicity assays, together with chemical analysis, has gained importance in the characterization of a wide range of materials, such as nanomaterials (De Marchi et al., 2018) and construction additives (Baderna et al., 2015).
The multiplicity of endpoints and/or of organisms is a feature of toxicological investigation.Tests are routinely combined to reduce uncertainty (Bierkens et al., 1998).Different organisms belonging to almost three trophic levels are required for the ecotoxicity and genotoxicity assessment of several environmental matrices, such as waste, contaminated soil, and drinking water.
Plant-based assays are part of these analytical batteries and are often included because of their good correlations with other plant and animal systems, including humans (Reis et al., 2017).For these reasons, the results of such batteries have been combined for the construction of risk matrices (Marmiroli et al., 2022) and environmental footprint maps (Bertanza et al., 2021;Pedrazzani et al., 2020).
Terrestrial plants are the largest interface between the environment and the biosphere, and plant-based assays have several advantages: high sensitivity, ecological pertinence, ease of execution, and low cost.Therefore, plants are useful tools to thoroughly investigate the environmental degradation caused by the most conventional soil pollutants, like pesticides (Menzyanova et al., 2022) and urban and industrial wastewater (Chowdhary et al., 2022;Singh et al., 2022), as well as the emerging ones, including plastic particles (Bouaicha et al., 2022) and carbon quantum dots (Vijeata et al., 2022).Furthermore, the combined effect of different pollutants (bisphenol A and titanium dioxide nanoparticles) on agricultural environments has been investigated from the toxicological to the proteomic level (Huang et al., 2022).
Among others, Allium cepa is an excellent in vivo model, suitable for both toxicity and genotoxicity biomonitoring (Tedesco & Laughinghouse, 2012) and readily available in two forms, seeds and bulbs.The size and number of Allium cepa chromosomes (2n = 16) allow for (relatively) easy analysis of microscopic parameters such as mitotic index (MI), chromosomal aberrations (CAs; including c-mitosis, bridges, fragments, buds), and micronuclei (Iqbal et al., 2019).Furthermore, Allium cepa is the most widely used higher plant for the investigation of early damage by single-cell gel electrophoresis (or comet) assay (Alias, Zerbini, & Feretti, 2023).Because of these characteristics, Allium cepa is widely applied to determine the cytotoxicity and genotoxicity of soil contaminants such as pesticides (Rosculete et al., 2019), veterinary drugs (de Souza et al., 2022), biosorbents (Pantano et al., 2021), as well as the phytotoxicity and genotoxicity of different sludges (Santos et al., 2022).
In the present study, an integrated toxicological and genotoxicological assessment is proposed to evaluate the impact of reference concrete mixtures cast with NA and concrete cast with SS as a partial replacement of NA using different plantbased assays.In particular, the same Lepidium sativum and Allium cepa seedlings were subsequently used to study toxicity and genotoxicity in a wide battery of plant tests.

Materials
Four samples of SS from electric furnaces were collected from four steel plants located in northern Italy.Samples were coded as A, B, C, and D. Particle size ranged from 1 to 5 cm, whereas density ranged from 2.5 to 3.3 g/cm 3 .Samples were stored by the factories in open, unprotected areas and subjected to a maturation process for at least 3 months after production, which resulted in the natural carbonation of the alkaline residue.A sample of NA was collected from a quarry in northern Italy.
Four concrete mixtures, referred to as SS concrete, were cast with Portland cement (13% w/w) and a partial replacemet (30%) of NA with SS (CSS); moreover, a reference concrete mixture was cast with only natural aggregates (CNA) and the same amount of Portland cement.The cement-to-water ratio was 0.48, and the concrete density was approximately 2470 kg/m 3 .The casts were naturally dried to a constant weight before the analysis (∼30 days).

Leaching test
Leaching tests were performed on CSS and CNA following the standard procedure for construction products (European Committee for Standardization, 2014) only for the first two extraction steps for a total duration of 24 h, which reasonably represented the most critical situation for the release of pollutants.Tests were performed on concrete blocks constantly soaked in demineralized water (leachant) with a liquid-tosurface area ratio of 8 mL/cm 2 , resulting in an amount of 1728 mL of leachant per block.Tests were conducted at room temperature, and the leachant was renewed after 6 and 18 h of contact (CSS-6h, CSS-18h, CNA-6h, CNA-18h).Leachates were filtered (membrane pore size 0.45 μm) and stored at 4 °C.

Lepidium sativum, Cucumis sativus, and Allium cepa seed germination and root elongation tests
The assays were performed following the Italian Environmental Agency guidelines (APAT, 2004).Briefly, seeds of Lepidium sativum, C. sativus, and Allium cepa not treated with fungicides were subjected to a preliminary viability check in distilled water in the dark at 25 ±1 °C (germination rates >90%).Leachate solutions were tested without any dilution, and distilled water was used as negative control.Three replicates per treatment were arranged by wetting a Whatman filter paper no. 1 using 2 mL solution each.Ten seeds for each replicate were spread on the filter.The three dishes of each replicate were packed in a tightly closed plastic bag and incubated at 25 ± 1 °C in the dark for 72 h.At the end of the incubation time, complete sprouts (≥1 mm) and root lengths were assessed.Results were expressed as mean germination index ± standard error.Values <80% describe inhibition effects, values >120% indicate germination supporting effects (namely biostimulation), and values between 80% and 120% describe no effects.

Lepidium sativum and Allium cepa seedling comet tests
At the end of the germination and elongation test, 30 seedlings exposed to the 6-h fraction of each leachate were collected in an ice-cold dish.Negative and positive controls were performed using distilled water and methyl methanesulfonate (MMS; 10 mg/L), respectively.The tips were finely minced with a scalpel, and 500 µL of ice-cold nuclei isolation buffer (200 mM Tris, 4 mM MgCl 2 × 6H 2 O, 0.5% Triton-X) were added.The suspension was allowed to sediment on ice for several minutes.The supernatant (180 μL) was resuspended in low-melting agarose (0.35% agarose final concentration).The mixture was spread on an agarose-coated glass slide and immediately covered with a coverslip.The slides were then placed for 30 min at 4 °C, to allow the agarose to solidify.After that, the coverslips were gently removed and the samples subjected to 1 h of unwinding and 20 min of electrophoresis (0.8 V/cm and 25 V at limit) in alkaline buffer (1 mM Na 2 ethylenediaminetetraacetic acid, 300 mM NaOH, pH 12.3).Then, the slides were neutralized in distilled water, dehydrated in 100% ethanol, and air-dried overnight.Finally, the slides were stained with GelRed Nucleic Acid Gel Stain (Biotinum).Fifty nuclei per slide (two slides per experimental condition) were examined with a fluorescence microscope (Olympus CX 41RF) equipped with a BP 515-560 nm excitation filter and an LP 580 nm barrier filter.For each slide, levels of DNA damage were assessed by the median value of the comet parameter tail intensity (percentage of DNA migrated in the tail) detected by automatic image analysis software (Komet 5, Kinetic Imaging, Ltd.).The mean (±standard deviation) of tail intensity values was calculated for each experimental condition.Statistical analysis was performed using univariate analysis of variance and Dunnett's multiple comparison test, where p < 0.05 was considered significant.Each experiment was conducted in duplicate.

Allium cepa bulb toxicity test
The toxicity test was performed to determine the doses to be used in the genotoxicity test (Rank et al., 2002).Equal-sized young onion bulbs were purchased from the local market without any treatment.Twelve bulbs were exposed for 72 h in the dark to different sample solutions (100%, 50%, 10%, and 1%).Distilled water was used to dilute the samples and as a negative control.The mean root length was used to calculate the median effect concentration value (EC50) with Microsoft Excel 2019.Other toxicity parameters (turgidity, consistency, color change, and root tip shape) were also evaluated.

Allium cepa bulb comet test
Three equal-sized young onion bulbs for each sample were subjected to a 48-h pregermination period in Rank's solution, then exposed for 24 h to undiluted samples based on the results obtained in the bulb toxicity test described above.Negative and positive controls were conducted using Rank's solution and MMS (10 mg/L), respectively.Fifty meristematic root tips (5 mm long) were cut and collected in an ice-cold dish.The procedure described above for the seedlings was followed in full for the bulbs.

Allium cepa bulb CA test
The Allium cepa test for the evaluation of CAs (Rank, 2003) was performed using six onion bulbs for each sample.After 48-h pregermination in Rank's solution, the bulbs were exposed for 24 h to undiluted samples based on the results obtained in the bulb toxicity test described above.The roots were cut, fixed in 1:3 acetic acid-ethanol for 24 h, and stored in 70% ethanol.Five roots of each sample were considered for microscopic analysis after staining with 2% acetic orcein: 5000 cells were scored for the MI, the measure of the cell division rate.Following recommendations, samples with MI <1% were not included.Then, 1000 cells in mitosis were scored for the CA frequency.The different types of aberrations were classified and divided into three main categories: fragments, rings, sticky chains, and bridges as direct DNA damage (DDD); laggards, binucleation, polar slips, multipoles, and c-mitosis as mitotic spindle defects (MSD), and buds as genic amplification (GA).Rank's solution was used as a negative control.A positive control was performed using maleic hydrazide (10 mM).The experiments were performed in duplicate.The statistical analysis was performed using the χ 2 test, where p < 0.05 was considered significant.

Phytotoxicity
The germination and root elongation of seeds of three plant species (Lepidium sativum, C. sativus, and Allium cepa) were assessed (Figure 1).Both fractions of all of the samples obtained from the leaching tests (6 h and 18 h) had a nontoxic effect on the plant species, demonstrated by germination indeces of approximately 100%, and ranging from 80% to 120% (Da Ros et al., 2018).In contrast, the germination index of C. sativus treated with samples CSS-C-6h and CNA-6h were 121.3 ± 3.1% and 133.2 ± 8.9%, respectively, and the germination index of Allium cepa treated with sample CNA-18h was 124.3 ± 4.0%.

DNA damage on Lepidium sativum and Allium cepa
Damage to the DNA was assessed on seedlings of Lepidium sativum and Allium cepa previously subjected to the germination test and on bulbs of Allium cepa using the comet test.
Because no differences were found between the two fractions of leachates (after 6 and 18 h of leaching), the comet tests were only performed on seedlings germinated with the 6-h fractions (Table 1).Damage to the DNA in Lepidium sativum seedlings was significantly increased only by the sample cast with NAs (tail intensity = 18.9 ± 0.2), whereas the four leachates from concrete cast with SS did not differ significantly from the control.The same trend was observed for Allium cepa seedlings, which were only significantly damaged by the leachate from concrete cast with NAs (tail intensity = 12.1 ± 1.6).
Given the lack of toxicity in Allium cepa bulbs in terms of root elongation (Supporting Information, Table 1S), the DNA damage was assessed on undiluted samples.The comet test on Allium cepa bulbs showed significant DNA damage in onions treated with CNA and from CSS-C leachates (Table 2).The damage was similarly induced by the two leachates derived from the leaching test (at 6 and 18 h) for both samples.

CAs on Allium cepa
Genotoxicity was also assessed in terms of CAs in Allium cepa bulbs.The frequency of CA was significantly increased by all leachates, without affecting the MIs (ranging from 10.2% to 12.2%; Table 3).In particular, the leachates from CSS-B-6h and CSS-D-6h induced the highest increase in total CA (25.8 ± 7.7   and 25.4 ± 6.6, respectively).Moreover, for both mixtures the 6-h fractions were able to induce significantly more CA than the respective 18-h fractions.Three types of CAs were identified: DDD, MSDs, and GA (Figure 2).In general, GA mechanisms accounted for less of the total aberrations (from 11.8% to 25.7%).The frequency of DDD ranged between 11.8% and 34.7%, with the only exception of cells treated with CSS-D, where direct DNA damage was >40% (40.9% and 43.3%, respectively, for both fractions of CSS-D).Finally, the predominant processes were the MSDs, with significant contributions well above 50% (p < 0.05 according to the χ 2 test) for the majority of the samples, with the few exceptions of CSS-B-18h, CSS-D-6h, and CSS-D-18h.

DISCUSSION
Leachates from concrete cast with a partial SS replacement and with only NA were characterized for phytotoxicity and genotoxicity using a plant-based approach to assess their impact on the terrestrial compartment.
The phytotoxicity test on the seeds of Lepidium sativum, C. sativus, and Allium cepa did not reveal any concerning aspect in the tested samples.The germination indeces were all approximately 100% and generally between 80% and 120%, values not different from the negative control.The sensitivity of the macrophytes used was comparable, with the few exceptions of a slight maintenance of germination and root elongation (biostimulation) induced by samples CSS-C-6h and CNA-6h in C. sativus and sample CNA-18h in Allium cepa, demonstrated by germination indeces >120%.Similarly, evaluation of root elongation in Allium cepa bulbs showed the absence of toxicity.
Genotoxicity was assessed through two main endpoints: DNA strand breaks and CAs, which represent early, reparable damage and stable, inheritable defects, respectively.The DNA damage assessed on the emerged seedlings of Lepidium sativum and Allium cepa was significantly increased only by the leachate from the reference concrete (CNA).Interestingly, the response was comparable in both plant species.The same early damage assessed on the root cells of Allium cepa bulbs was significantly increased by the leachates from CNA and CSS-C.Considering the assessment of stable damage, all samples caused a significant increase in the frequency of chromosomal aberrations.In particular, the predominant mechanism of action was attributable to MSDs and only secondarily to direct DNA damage.
The different results in terms of DNA damage between seedlings and bulbs of Allium cepa are likely due to the sensitivity of the two plant forms, which represent different  physiological states of the same organism.Seeds, defined as the mature, fertilized ovules of flowers containing dormant embryos and food reserves awaiting germination (Lopez & Barclay, 2017), are protected from the external environment by several integumentary barriers.Moreover, the emerging radicle, or primary root, is an immature structure that is not yet fully capable of performing its functions.All of these aspects contribute to making the seeds less susceptible than the bulbs to external inputs.Indeed, the genotoxic damage observed in the seeds could be attributed to a mixture present in the leachate from the reference concrete that is more active toward the genome and/or more capable of interacting with the seedling structures.The analysis of DNA damage in onion bulbs confirmed these data and demonstrated the genotoxicity of a sample derived from concrete with a partial substitution of SS, because of the greater susceptibility of the bulbs.For the same reason, the detection of CAs in Allium cepa highlighted the genotoxic potential of all samples.As primary producers, plants are the first and most important link in the food chain of any ecosystem.Their direct contact with the matrix and their ability to absorb water-soluble substances enable them to interact strongly with pollutants.For these reasons, plants are useful toxicological models for environmental studies, especially in soil and water, where they can be applied using in situ, in vivo, and in vitro settings.Furthermore, different parts of plants (e.g., leaves, meristems, seeds, pollen) and several endpoints can be used to study many toxicological properties (Grant, 1994).Thus, it is evident how the integration of different biological tests offers the possibility of covering several endpoints, allowing for a detailed understanding of the effects on higher plants and an assessment of the environmental hazardousness of the tested materials.Moreover, the consistency between the results of a battery of plant assays supports the data obtained, allowing for a more certain identification of the most concerning samples.In addition, and even more importantly, the biological assays are independent from a predetermined analytical schedule, which could not fully describe a sample.Indeed, research has shown that leachates from different granular SS and NA, although chemically characterized by full compliance with the legal requirements, demonstrated toxic effects on several biological systems (Alias et al., 2021).Similarly, in another study leachates from concrete cast with NA and with a partial substitution of SS, although fully compliant with the reference limits, could have an impact on living organisms (Alias, Zerbini, Abbà, et al., 2023).
Given that the toxicological test guidelines emphasize the importance of the heterogeneity of endpoints and/or of organisms, a limitation of the present study may appear to be the lack of analysis of organisms belonging to different trophic levels, other than plants.However, the plant-based battery demonstrated a good ability to classify the toxicity of the tested samples, in terms of both phytotoxicity and genotoxicity.Thus, this plant-based approach appeared to be a reliable tool for screening aqueous samples and/or environmental extracts.
Another notable feature is the use of the same organisms to assess different endpoints.This could be a useful strategy to study the environmental impact of the different materials used, including construction materials.Such innovative toxicological workflow has recently been reported by some authors who investigated the toxicity and the genotoxicity of chemicals on plants using the same organisms, which were subsequently subjected to different assessments.Passatore et al. (2022) studied the effects of bismuth exposure on Lepidium sativum seedlings.Similarly, the physiological and genotoxic effects of phthalate on Lemna minor and Spirodela polyrhiza were investigated by Pietrini and collaborators (Pietrini et al., 2022).
A subsequent, full toxicological and genotoxicological characterization might be carried out on the most concerning samples through more analyses on different organisms and endpoints.The methodology of the present study led to an evaluation of the impact of concrete samples on higher plants to allow their potential reuse for construction applications.

CONCLUSIONS
These results demonstrated that despite some genotoxic effects of concrete on plant cells, the partial replacement of SS does not seem to make concrete any more hazardous than the reference one in terms of global assessment (toxicological, environmental, economic), suggesting the potential use of SS as a reliable recycled material.Furthermore, the present study confirmed that the plant-based approach is a useful tool for the first-line toxicological characterization of recycled materials intended for prolonged contact with soil and water.
Supporting Information-The Supporting Information is available on the Wiley Online Library at https://doi.org/10.1002/etc.5709.

FIGURE 1 :
FIGURE 1: Germination index (GI) expressed as a percentage in Lepidium sativum, Cucumis sativus, and Allium cepa treated with undiluted leachates of steel slags concrete (CSS) and reference concrete (CNA).Data are expressed as mean GI% ± standard error.

FIGURE 2 :
FIGURE 2: Frequency distribution of the three categories of chromosomal aberrations: direct DNA damage (DDD), mitotic spindle defects (MSD), and genic amplification (GA).CSS = steel slags concrete; CNA = reference concrete cast with only natural aggregates.

TABLE 1 :
Comet test of undiluted leachates of steel slags concrete and reference concrete in Lepidium sativum and Allium cepa seedlings

TABLE 2 :
Comet test of undiluted leachates of steel slags concrete and reference concrete in Allium cepa bulbs SD = standard deviation; CSS = steel slags concrete; CNA = reference concrete cast with only natural aggregates.Statistically significant versus negative control according to Dunnett's test: *p < 0.05.

TABLE 3 :
Mitotic index and chromosomal aberrations frequency of undiluted leachates of steel slags concrete and reference concrete in