Effects of Airborne Particulate Matter in Biomass Treatment Plants on the Expression of DNA Repair and IL-8 Genes

Featured Application

In vitro models are valuable tools for evaluating the genotoxicity and oxidative stress risks associated with PM₁₀ emissions from organic waste treatment.

Abstract

Biogas plants for sewage and organic waste treatment are rapidly expanding. While these facilities provide valuable benefits, such as renewable energy production and the promotion of circular economy practices, they also emit airborne particles of biological origin, which may pose potential health risks. This study aims to evaluate, by in vitro assay, the cytotoxic and genotoxic potential of PM₁₀ sub-fractions (0.49–10 µm and <0.49 µm) generated in eight different plants, also assessing the endotoxin component using the Limulus Amebocyte Lysate (LAL) assay. Human embryonic lung fibroblasts (HELF) were exposed to organic extracts of particulate matter (PM). Cytotoxic effects (XTT assay) were analyzed, along with the modulation of gene expression involved in DNA repair (ERCC1, XRCC1, XPA, and XPF) and IL-8 production as a marker of inflammatory response. PM₁₀ and endotoxin concentrations varied significantly among the plants (ANOVA, p < 0.01), with PM₁₀ levels ranging from 14 to 18,000 µg/m³ and endotoxin content from 1 to 138 EU/m³. Exposure significantly increased ERCC1 and IL-8 expression by 25% and 53%, respectively (paired t-test, p < 0.01). IL-8 expression correlated with endotoxin exposure (Spearman’s rho = 0.35; p < 0.01). A deeper understanding of the biological component of airborne PM₁₀ can enhance risk assessments for occupational and nearby resident communities’ safety.

1. Introduction

Particulate matter (PM) pollution is one of the most critical issues in air quality assessment. Over time, the WHO has progressively lowered the guideline values for PM due to its health risk potential, with the most recent update published a few years ago [1,2].

The primary sources of air pollution include both anthropogenic activities and natural emissions. Air quality is influenced by chemical pollutants as well as microbiological contaminants, and their combined effects on human health are modulated by interactions occurring in the atmosphere.

Genotoxic and epigenetic effects of PM exposure have been widely documented in the literature through both experimental models and epidemiological studies [3,4]. In 2013, the IARC classified outdoor PM₁₀ as a human carcinogen [5]. Gene expression analyses following exposure to urban fine PM have revealed significant alterations in the transcriptomic profile, particularly in pathways related to oxidative stress and inflammation [6,7,8,9].

Although PM generally contains only minimal amounts of microorganisms and airborne endotoxins, studies have demonstrated that inhalable dust in composting and biogas plants can serve as a reliable marker of bioaerosol exposure [10,11,12]. This is supported by the strong correlation between dust levels and the concentration of transported microorganisms [13,14].

Genotoxic effects are well documented for urban PM_2.5 [7,15], yet the specific contribution of biological particles—whether in total suspended matter, PM₁₀, or finer fractions—remains unclear.

This study aims to evaluate the effects of particulate matter from different biomass treatment plants on human cell cultures, assessing its potential toxicity. The PM samples analyzed are characterized by a predominantly biological composition. The study endpoints include cytotoxicity, the modulation of gene expression involved in DNA repair, and the activation of inflammatory responses triggered by oxidative stress.

2. Materials and Methods

2.1. Samplings and Gravimetric Analysis

The sampling involved eight full-scale biomass treatment plants placed in Northern Italy (food waste, wastewater sludge, lignocellulosic, and agro-livestock residues) during the warm season in the same year. In particular, we sampled three composting plants processing the organic fraction of municipal solid waste (OFMSW) and other plant-based byproducts, three plants treating agro-livestock matrices, including both swine and bovine slurry and agricultural processing residues, one facility dedicated to the treatment of wastewater sludge, and one facility processing lignocellulosic waste.

PM₁₀ sub-fractionated sampling was conducted in different areas of the plants, including areas where waste materials were received, treatment areas, and finally the areas where refinement of the effluents took place (some examples are shown in Figure 1). A high-volume cascade impactor sampler (High Volume Sampler, HVS; Airflow PM₁₀, Analitica Strumenti, Pesaro, Italy) was used, operating with a suction flow rate of 1.24 m³/min, in compliance with UNI EN 12341 standards [16]. The sampler employs a multistage impact mechanism, allowing airborne particles to pass through progressively narrower slits, thereby fractionating them based on their aerodynamic diameter onto sequentially arranged filters. The collected PM₁₀ was divided into six sub-fractions of equivalent aerodynamic diameter: 10–7.2 μm, 7.2–3.0 μm, 3.0–1.5 μm, 1.5–0.95 μm, 0.95–0.49 μm, and <0.49 μm. Slotted glass fiber filters (A/E type, 8 × 10, Gelman Sciences, Bloomfield Twp, MI, USA; labeled as A) were used to collect PM_0.49-10, while A4-sized filters (203 × 254 mm, Pall Corporation, New York, NY, USA; labeled as B) were used for PM_0.49 [17].

Each sampling session lasted at least 4 h and was generally conducted twice a day (morning and afternoon), except in specific cases where only one sampling per day was performed due to logistical and organizational constraints. Each filter underwent a pre- and post-sampling conditioning process for 48 h in a desiccator with silica gel to minimize any interference humidity might introduce into weight measurements. All membranes underwent a gravimetric analysis of particulate matter by pre- and post-sampling weighing with a five-decimal analytical balance (±0.00001 g), including a blank filter as a control to assess potential weight changes not attributable to sampling. The PM concentration on each filter was calculated considering the particulate mass and the volume of air passing through the filter. The PM concentration was expressed as μg per m³ of sampled air.

The sampled filters were used to perform water-based and organic-based extractions, respectively, assess endotoxin levels and expose human embryonic lung fibroblasts (HELF) in order to evaluate the in vitro modulation of gene expression involved in DNA repair (XRCC1, ERCC1, XPA, XPF) [18] and IL-8 production [19].

2.2. Extraction and Quantification of Endotoxins

To extract endotoxins, portions of the original filters were cut (half of the A filters and 1/4 or 1/8 of the B filters), further sectioned into small strips, and placed in a sterile 50 mL tube with 15 mL of RPMI-1640 supplemented with 0.025% Tween-20 (Sigma-Aldrich, St. Louis, MI, USA). The tubes were then subjected to ultrasonic bath treatment for 10 min and vortexed for 30 s. These steps were repeated three times. Subsequently, the samples were centrifuged at 5000 rpm for 10 min to remove glass fibers. The supernatant was then collected and then used for endotoxin determination [20,21].

For endotoxin quantification, the chromogenic endpoint LAL (Limulus Amebocyte Lysate) assay (QCL-1000™ 50-648 U, Lonza, Walkersville, MD, USA) was used, along with 96-well transparent polystyrene microplates (Costar^® 3596, Corning Incorporated, Corning, NY, USA). This method is based on the activation of an enzymatic cascade in Limulus triggered by endotoxins, leading to the activation of a coagulation enzyme. When mixed with a chromogenic substrate and incubated, the activated enzyme cleaves the substrate, releasing p-nitroaniline, which produces a yellow coloration.

The intensity of the yellow color was measured by spectrophotometric reading at 407 nm and 37 °C (TECAN Infinite^® M200 Pro, Tecan, Seestrasse, Switzerland, I-Control™ software 1.10) and was directly proportional to the amount of endotoxin in the sample. Absolute quantitative evaluation was performed, following the supplier’s instructions, using a standard curve (five points: 1, 0.5, 0.25, 0.1, and 0.05 EU/mL) obtained from a stock solution of lyophilized endotoxins reconstituted in 1 mL of LAL Reagent Water.

The endotoxin concentration was expressed as EU (Endotoxin Unit) per 1 m³ of sampled air for each of the six PM sub-fractions. For A filters, corresponding to five sub-fractions, results were aggregated by summing the determinations from each sub-fraction.

2.3. Organic Extraction of PM_100.49 and PM_0.49

An organic extraction was then performed on pooled A filters and on B, (half of the A filters and 1/4 or 1/8 of the B filters), obtaining 64 samples (of which 24 were from OFMSW, 24 from agro-zootechnic, 8 from WWTP, and 8 from lignocellulosic implants). Organic extracts were obtained through 80 Soxhlet cycles with acetone, followed by solvent evaporation using a rotary evaporator (Rotavapor R-300, BUCHI, Cornaredo, Italy). Finally, the sample was resuspended in a volume of dimethyl sulfoxide (DMSO) to achieve a final concentration of 0.2 m³/μL [18,22].

2.4. Exposure and Cytotoxicity Assay

Human embryonic lung fibroblasts (HELF) were obtained from the pathology section of the Regina Margherita Hospital, Turin with the collaboration of Prof. G. Gribaudo and Dr. A. Luganini, University of Turin [18]. Such cells were maintained in culture using Eagle’s Minimum Essential Medium (MEM, 51411C Sigma-Aldrich) supplemented with 10% fetal bovine serum (FBS, F0804 Sigma-Aldrich), 1 μg/mL ampicillin/streptomycin (P4083-100 ML), and 2 mM sodium pyruvate (S8636 Sigma-Aldrich). The cells were incubated at 37 °C in a humidified atmosphere with 4–5% CO₂.

Cytotoxicity was assessed using the XTT assay with the Cell Proliferation Kit II (XTT, Sigma-Aldrich). This colorimetric assay relies on the ability of mitochondrial succinate dehydrogenase in viable cells to reduce tetrazolium salt (XTT) into formazan, a soluble, orange-colored compound. The intensity of the coloration, measured by spectrophotometry, is directly proportional to the number of metabolically active cells.

Following the kit instructions, the cells were seeded in 96-well tissue culture plates (TPP^®, Europe/Switzerland) at a density of 50,000 cells per well in a volume of 200 μL of culture medium. After seeding, the cells were incubated for approximately 24 h to allow for proper growth and attachment to the plate. Starting from a medium containing 3% organic extract, the following media with 2%, 1%, 0.5%, and 0.25% PM were prepared through appropriate dilutions. The growth medium was replaced with 100 μL of medium containing the extract, performing three technical replicates for each concentration. A blank (cells exposed to medium only) and negative controls (cells exposed to 3%, 2%, 1%, 0.5%, 0.25% DMSO) were also included. After 24 h of incubation, the XTT assay was carried out following the protocol provided by the kit. To the medium containing the extract, 50 μL of XTT Labeling Mixture was added, which had been previously prepared with 5 mL of Labeling Reagent and 0.1 mL of Electron Coupling Reagent (final XTT concentration of 0.3 mg/mL). After 4 h of incubation, the plate was read using a spectrophotometer (TECAN Infinite^® M200 Pro, Switzerland, I-Control™ software) at a wavelength of 492 nm, with a reference wavelength of 690 nm.

The concentration corresponding to 1% PM showed an acceptable average cytotoxicity (<20%), allowing for subsequent gene expression analysis. The results of the test were expressed as mitochondrial dehydrogenase activity, corresponding to the absorbance value obtained by substituting “1” for x in the regression equation, which was calculated by correlating the exposure dose with the observed absorbances at the five concentrations used. Furthermore, the percentage of inhibition of mitochondrial dehydrogenase activity was calculated for each sample by subtracting the y value of the regression line corresponding to cells exposed to the blank reading.

2.5. Extraction of Nucleic Acids

Cells were seeded at a density of 5 × 10⁵ cells/mL in 6-well plates (Corning^® Costar^® TC-Treated Multiple Well Plates) and incubated for 48 h. For each extraction, four plates were used, for a total of 10 PM samples and their respective controls (cells exposed to medium only or exposed to 1% DMSO). For each sample of the organic extract, a growth medium containing 1% PM was prepared by adding 70 μL of each extract to 7 mL of medium. After 48 h of incubation, the growth medium was replaced with 3 mL of medium containing the extract, and the cells were exposed to the extract for 24 h in the incubator.

The quantity of nucleic acids extracted was assessed using a spectrophotometer (TECAN Infinite^® M200 Pro, Switzerland, I-Control™ software), with the appropriate NanoQuant plate (Tecan Trading AG, Switzerland), reading the absorbance at 260 nm. The results were expressed as ng/μL, and the minimum extract concentration considered quantifiable and suitable for subsequent expression analyses was 1 ng/μL.

RNA mean concentrations were 118.47 ng/μL and 72.78 ng/μL of the PM_0.49 and PM_0.49-10, respectively. RNA samples were stored at −80 °C pending further reverse transcription and gene expression analysis.

2.6. Gene Expression

All RNA extracts were reverse transcribed using the iScript™ Reverse Transcription Supermix for RT-qPCR (Bio-Rad, Hercules, CA, USA), following the protocol for single cDNA synthesis reactions for each sample. The reaction mixture contained 4 μL of iScript RT Supermix and an appropriate volume of MilliQ water (autoclaved and prepared for expression analysis) to reach a final reaction volume of 20 μL.

The reactions were incubated in a T100™ Thermal Cycler (Bio-Rad) following the thermal protocol provided by the kit and including priming (5 min at 25 °C), reverse transcription (20 min at 46 °C), and enzyme inactivation (1 min at 95 °C).

The gene expression levels of XRCC1, XPA, ERCC1, XPF, and IL8 were evaluated using GAPD and HPRT1 as housekeeping genes. Quantitative Real-Time PCR was performed using the SsoAdvanced™ Universal SYBR^® Green Supermix (Bio-Rad) on the CFX96™ Real-Time System (Bio-Rad).

For each reaction, 2 μL of cDNA sample was combined with a reaction mix containing 10 μL of SsoAdvanced SYBR Green Supermix, 0.5 μL of forward and reverse primers (Table 1) (at a concentration of 4 μM), and 7 μL of MilliQ water, resulting in a final reaction volume of 20 μL in each well of the plate.

The protocol used included an initial polymerase activation step for 30 s at 98 °C, followed by 40 cycles of: denaturation at 95 °C for 10 s, annealing/extension at 60 °C for 30 s, plate reading, and finally, generating a melting curve between 65 °C and 95 °C (increasing by 0.5 °C every 2–5 s).

The primers used for each target gene expression are shown in Table 1. To calculate gene expression modulation, the ΔΔCt method was used. The results were expressed as the increase or decrease in gene expression following exposure to PM. Finally, to normalize the data distribution and perform statistical tests, the Log10 (2^−(ΔΔCt)) was calculated [18,23].

Table 1. Description of each gene target’s functions and the primers included in the gene expression analysis.

Gene Function	Primers	5’-3′ Sequence	Ref.
Hypoxanthine phosphoribosyltransferase 1. It allows cells to recycle purines	HPRTI_f HPRTI_r	TGACACTGGCAAAACAATGCA GGTCCTTTTCACCAGCAAGCT	[24]
Glyceraldehyde-3-phosphate dehydrogenase	GAPD_f GAPD_r	TGCACCACCAACTGCTTAGC GGCATGGACTGTGGTCATGAG	[24]
DNA excision repair protein ERCC-1 is involved in nucleotide excision repair (NER). It creates a cut at the 3′ end of the damage site and forms an enzymatic complex with XPF	ERCC1_f ERCC1_r	GGGAATTTGGCGACGTAATTC GCGGAGGCTGAGGAACAG	[25]
Together with ERCC1, it creates a cut on both sides of the damage site leading to the removal of a piece of DNA	XPF_f XPF_r	TGACCCTAGCAAGCCTTTCTCTCT GGGGCACCAGAGAATCCGTAGTC	[25]
It recognizes damage and functions as a scaffold for other proteins, NER involved	XPA_f XPA_r	GAGTTTCAACCCAGGTCTTCTCATTTT AATGGTAACAGTGAGTAGTAGTAGTGC	[25]
Involved in the repair of single-strand breaks in DNA due, for example, to exposure to ionizing radiation and alkylating agents. Involved in BER	XRCC1_f XRCC1_r	GCAAACCCCGAGGAGAAGGCA ACTGCTGGAACCTGGCCCTGC	[26,27,28]
Chemotactic factor for neutrophils, often associated with inflammation and oxidative stress in the lungs	IL-8_f IL-8_r	TGCCAAGGAGTGCTAAAG CTCCACAACCCTCTGCAC	[19]

Table 1. Description of each gene target’s functions and the primers included in the gene expression analysis.

Gene Function	Primers	5’-3′ Sequence	Ref.
Hypoxanthine phosphoribosyltransferase 1. It allows cells to recycle purines	HPRTI_f HPRTI_r	TGACACTGGCAAAACAATGCA GGTCCTTTTCACCAGCAAGCT	[24]
Glyceraldehyde-3-phosphate dehydrogenase	GAPD_f GAPD_r	TGCACCACCAACTGCTTAGC GGCATGGACTGTGGTCATGAG	[24]
DNA excision repair protein ERCC-1 is involved in nucleotide excision repair (NER). It creates a cut at the 3′ end of the damage site and forms an enzymatic complex with XPF	ERCC1_f ERCC1_r	GGGAATTTGGCGACGTAATTC GCGGAGGCTGAGGAACAG	[25]
Together with ERCC1, it creates a cut on both sides of the damage site leading to the removal of a piece of DNA	XPF_f XPF_r	TGACCCTAGCAAGCCTTTCTCTCT GGGGCACCAGAGAATCCGTAGTC	[25]
It recognizes damage and functions as a scaffold for other proteins, NER involved	XPA_f XPA_r	GAGTTTCAACCCAGGTCTTCTCATTTT AATGGTAACAGTGAGTAGTAGTAGTGC	[25]
Involved in the repair of single-strand breaks in DNA due, for example, to exposure to ionizing radiation and alkylating agents. Involved in BER	XRCC1_f XRCC1_r	GCAAACCCCGAGGAGAAGGCA ACTGCTGGAACCTGGCCCTGC	[26,27,28]
Chemotactic factor for neutrophils, often associated with inflammation and oxidative stress in the lungs	IL-8_f IL-8_r	TGCCAAGGAGTGCTAAAG CTCCACAACCCTCTGCAC	[19]

2.7. Statistical Analysis

For statistical analysis, the IBM SPSS Statistics Package, version 29.0.1.0 (IBM Corp., Endicott, NY, USA) was used. First, a general descriptive analysis of the obtained data was conducted. Subsequently, a comparison of means was performed using Student’s t-test for independent or paired samples when applicable.

ANOVA was used for multivariate analysis, assuming equal variances, followed by Tukey’s post-hoc test for multiple comparisons. To evaluate correlations between variables, the Spearman’s rho coefficient was calculated. Differences between means and correlations were considered statistically significant at a p-value < 0.05 and highly significant with a p-value < 0.01. For the statistical analyses, the different sampling areas were grouped into two categories: one including reception, selection, composting, refining, and tunnel (influent group) and the second including sludge, methanization, and output (treated group). Similarly, the different plants were grouped into four groups based on similarities: composting plants treating organic fraction municipal waste (group 1), plants treating agro-livestock matrices (group 2), the plant treating lignocellulosic waste (group 3), and the plant treating wastewater sludge (group 4). As for the season, samples taken from March 21 to September 20 were categorized as “spring-summer,” while those from September 21 to March 20 were categorized as “autumn-winter.”

3. Results

3.1. PM₁₀, PM_0.49 and Endotoxins

From the descriptive analysis shown in Figure 2, a significant variability in PM (Figure 2A) and endotoxin (Figure 2B) concentrations across the different samples is observed. The mean PM concentrations were found to be 574.7 μg/m³ for PM_0.49-10 and 564 μg/m³ for PM_0.49. For endotoxins, the concentrations range from low levels (few EU/m³) to around 90 EU/m³, with an average of 13.6 EU/m³ for PM_0.49-10 and 11.7 EU/m³ for PM_0.49.

When calculating the total endotoxin mass, it is observed that despite high PM concentrations, the total endotoxin amount is less than one millionth of the total PM mass in each filter. However, the levels of PM and endotoxins are significantly positively correlated (Spearman’s rho = 0.654; p < 0.01) (

Table 2. Spearman’s correlations among the changing of expression level of the different gene target, with respect to the control, and the PM₁₀ and endotoxin levels. Asterisks show significant and highly significant correlations.

	XRCC1	ERCC1	XPF	XPA	IL-8	EU/m3	PM10 µg/m3
XRCC1	1.000	0.505 *	−0.352	−0.623 **	0.222	0.618 **	0.036
		0.019	0.118	0.003	0.333	0.003	0.876
ERCC1			0.016	−0.251	0.531 *	0.519 *	0.043
			0.947	0.273	0.013	0.016	0.854
XPF				0.552 **	−0.318	−0.370	−0.229
				0.009	0.160	0.099	0.319
XPA					−0.096	−0.221	0.204
					0.679	0.336	0.375
IL-8						0.662 **	0.325
						0.001	0.151
EU/m3							0.654 **
							<0.001

).

The organic waste treatment plant showed the highest level of both PM and endotoxins, and when the concentrations of the two PM sub-fractions (PM_0.49-10 and PM_0.49) are summed, the maximum values are quite high, with approximately 18,000 μg/m³ of PM and 138 EU/m³ of endotoxins, exceeding the recommended exposure limits for particulates not otherwise regulated, total and respirable dust (PNOR) (15 mg/m³ and 5 mg/m³), and for endotoxins (90 EU/m³) [29].

Finally, when comparing the two PM sub-fractions (PM_0.49-10 and PM_0.49) using the paired t-test, no significant differences in PM and endotoxin concentrations were found, indicating that these contaminants are evenly distributed across the investigated environments (p > 0.05).

3.2. XTT Assay

The enzyme activity—considering an exposure dose equivalent to approximately 0.2 m³ of sampled air and expressed in terms of absorbance—ranged from 1.11 to 3.27, with a mean value of 2.35 ± 0.52. These values were compared with the viability measured in control cells, and the percentage of dehydrogenase activity inhibition following exposure (Δ dehydrogenase activity/0.2 m³) was calculated.

Overall, as illustrated in Figure 3, most samples induced an enzyme activity decrease ranging from 0% to 20%, which is consistent with subsequent gene expression analyses. In addition to this decrease, an increase in mitochondrial activity (up to 27.5%) was observed in eight PM_0.49 samples.

Considering samples from different types of plants (Figure 4), the delta of bioreductase activity varies significantly. Among these, composting plants exhibited a significantly higher cytotoxic particulate activity compared to agro-zootechnical matrix and lignocellulosic matrix plants (ANOVA, p < 0.05). However, when comparing sampling areas within the plants (grouped into groups 1 and 2), no significant difference in the delta of dehydrogenase activity was detected.

Analyzing the two PM sub-fractions separately, a marked but non-significant difference was observed for PM_0.49. Specifically, the dehydrogenase activity reduction was 15% in samples collected from biomass entry and treatment areas, compared to only 3% in samples from product exit areas. This non-significant difference might be influenced by the lower inhibition of dehydrogenase activity due to most of the PM_0.49 samples having been collected from areas corresponding to product exit points.

No significant variations in bio reductase activity were observed based on season or sampling time (t-test, p > 0.05). However, a significant direct correlation was found between PM levels (μg/m³) and the delta of dehydrogenase activity (Spearman’s rho = 0.260; p < 0.05), indicating that cytotoxicity per unit volume of air increases with rising PM concentrations.

3.3. Gene Expression Analysis

From the gene expression analysis, it was observed that exposing cells to PM induces a modulation in treated cells compared to the control. Specifically, there is a significant increase (paired t-test, p < 0.01) in the expression of ERCC1 and IL-8, which are upregulated by 25% and 53%, respectively (Figure 5). A variability in expression is observed across all genes, with some showing a decrease in treated cells compared to the control, while others exhibit an increase. The gene displaying the highest variability is XPA, with a maximum decrease in expression of 2.75-fold and a maximum increase of 3.59-fold compared to the control.

The expression of the analyzed genes does not show significant differences based on the effect of the different PM fractions. Only IL-8 expression exhibits a difference, borderline to statistical significance (paired t-test, p = 0.051). The increase in IL-8 expression is predominantly induced by the PM_0.49 fraction, whereas the PM_0.49-10 fraction contributes more significantly to the average upregulation of DNA repair genes.

By evaluating the effect of the samples collected from different types of plants significant differences were observed in the expression of ERCC1, XPF, and IL-8 (Figure 6). Specifically, ERCC1 expression was significantly increased (ANOVA, p < 0.01) by PM derived from the lignocellulosic matrix plant compared to FORSU composting plants. The lignocellulosic matrix plant also showed a distinct pattern in XPF expression, which was significantly reduced (ANOVA, p < 0.01) compared to organic waste composting plants and agro-zootechnical matrix plants.

Additionally, IL-8 expression differed significantly (ANOVA, p < 0.05) among composting plants, where it was markedly overexpressed, and agro-zootechnical matrix plants. No significant differences were observed in gene expression modulation based on plant area, season, or sampling time. A significant correlation (Spearman’s rho = 0.668; p < 0.01) was found between IL-8 upregulation and the endotoxin concentration in each sample. However, no significant correlation was observed between PM concentration and the expression of any of the analyzed genes. Furthermore, IL-8 expression correlated only with ERCC1 gene expression (Table 2) among the DNA repair genes, suggesting that such biological mechanisms likely do not converge.

A highly significant direct correlation was observed between XPA and XPF expression (Spearman’s rho = 0.552; p < 0.01). Conversely, a highly significant inverse correlation was found between ERCC1 expression and XPF (Spearman’s rho = −0.251; p < 0.01).

Finally, gene expression did not correlate with changes in bio reductase activity, confirming that the induced cytotoxicity did not influence the obtained results.

4. Discussion

The uniqueness of this research can be recognized in two key aspects. First, human embryonic lung fibroblasts were used as the cellular model. These are a rarely employed model in PM toxicity studies, which more commonly rely on immortalized epithelial or tumorigenic cell lines. In a recent study, J. Lee et al. emphasized the importance of using normal human cell lines to better assess toxicological effects in a physiologically relevant context [30,31]. Second, unlike most studies, which focus on urban PM, this research examined industrial PM, specifically that derived from biomass treatment plants. These facilities are widespread and expanding across Europe and beyond, and the composition of the PM they generate is markedly different from urban PM, being primarily biological in nature.

Organic waste composting plants exhibit the highest and most significant levels of airborne PM and endotoxins. This is expected due to the intrinsic characteristics of these facilities, both in terms of the quantity and composition of the processed material and the type of treatment, which involves the continuous movement and overturning of organic matter. This process generates a substantial amount of dust, including coarse particles that are visibly detectable by personnel involved in sampling. Additionally, composting plants are predominantly indoor, whereas other plant types are semi-outdoor or outdoor, leading to differences in ventilation and contaminant dispersion. In these same plants, the cytotoxic activity of PM was significantly higher compared to other facilities, such as agro-zootechnical or lignocellulosic matrix plants. This effect is likely attributable to both the higher PM concentration—given the significant direct correlation between these two variables—and the composition of the airborne mixture in these plants.

In most cases during the sampling period, the occupational exposure limits for respirable and inhalable dust, as well as endotoxins [29], were not exceeded. However, one composting plant significantly surpassed these thresholds, while a second similar facility exceeded the EU/m³ limits during one sampling day. Additionally, data indicate higher contamination levels (PM and endotoxins) and greater cytotoxicity in the biomass entry, storage, and treatment areas compared to the product exit areas. The most polluted areas were identified as the refinement zone and the composting tunnel, which aligns with the greater biomass movement occurring during the early treatment stages.

A key biological finding emerged from the XTT assay. To assess potential biological modulation without excessive cell death, cells were exposed to 1% organic PM extract, a sub-lethal dose. Relative to the control (cells exposed only to culture medium), in most cases, a mild cytotoxic effect was observed, with cell viability decreasing between 0% and 20%. Interestingly, in some cases, PM exposure appeared to increase mitochondrial dehydrogenase activity, as indicated by negative delta values, with a maximum induction of 27.5%. This effect was exclusively observed in eight samples, all derived from the finest PM fraction (PM_0.49). The mechanisms by which PM exerts its toxicity are closely linked to mitochondrial activity, as mitochondria are the primary producers of ROS and among the most common cellular targets for oxidative damage (e.g., mtDNA damage, lipid peroxidation, and protein oxidation). Evidence from in vitro studies shows that PM generally has an inhibitory effect on mitochondrial activity, particularly reducing the functionality of certain enzymes, including succinate dehydrogenase, which is responsible for the reduction of the salt in the XTT assay, as observed for most of the samples evaluated.

On the other hand, it is important to consider the presence of a strong biological component within the mixture used in this study. As recent studies have shown, this biological component significantly influences the toxicity and oxidative potential of PM, with opposing effects depending on the microbial species present. For example, Samake et al. [32] observed, in cell-free systems, that the oxidative potential of certain chemical species in PM decreased gradually with increasing concentrations of S. epidermidis and other bacterial species, likely due to the ability of the bacterial cell wall to chelate these chemical components and inhibit their reactivity. In contrast, high concentrations of A. fumigatus spores and other fungal species increased the oxidative potential of PM.

However, it is also known that cytotoxicity assays using tetrazolium salts (e.g., XTT, MTT) are influenced by numerous factors that can perturb cellular and mitochondrial homeostasis, potentially leading to partial reprogramming of cellular metabolism, which may sometimes cause confounding effects in interpreting results. Among these factors, plant extracts (which contain polyphenols, flavonoids, etc.) are potential contributors, demonstrating the ability to directly reduce XTT salt through the presence of free -SH groups, thus overestimating cell viability. It cannot be ignored that the PM samples used in our study, due to their derivation from plants treating organic biomass, may contain plant-derived components that contributed to these effects. However, this effect was observed only in a few PM_0.49 samples, so it cannot be generalized to all of the extracts.

Another study evaluated the oxidative effects of tert-butylhydroquinone (t-BHQ) on Madin–Darby Bovine Kidney cells using the XTT assay [33]. The authors explain that pro-oxidant agents, such as t-BHQ, at low doses, can induce the transcription of phase II detoxifying enzymes through the Nrf2-ARE pathway (nuclear factor E2-related factor; antioxidant responsive elements), promoting cytoprotection against oxidative damage. In such studies, exposure of MDBK cells to t-BHQ at low doses induces an apparent increase in cell viability, which was not observed with the crystal violet (CV) staining.

Numerous studies have shown that the toxicity mechanisms of PM are also closely linked to the activity of the Nrf2 transcription factor mentioned above, which plays a key role in protecting the cell from oxidative damage and inflammation. There is in vitro and in vivo evidence demonstrating that organic chemical components of PM_2.5 significantly induce activation of this pathway [34,35].

The predominant component of the PM under study is of biological origin. It cannot be ignored that organic components such as PAHs (polycyclic aromatic hydrocarbons), which may also originate from the use of transport vehicles within the plants, are present in trace amounts and, along with the biological components, enhance the oxidizing nature of the sample. These effects were, furthermore, observed only in PM_0.49 samples, the finest fraction, which carries these organic components in higher amounts compared to coarse PM.

The gene expression analysis revealed that PM had a quite modulatory effect on all of the genes examined, but it had a highly significant impact on the expression of ERCC1 and IL-8, with transcript increases of 25% and 53%, respectively. In contrast, XRCC1, XPF, and XPA showed a mild increase in expression, but the biological response was highly variable. This suggests that the high heterogeneity of the pollutant compound mixture in PM₁₀ airborne particles may lead to inconsistent biological effects, sometimes even opposing ones.

The XRCC1 protein plays a crucial role as a scaffold in various enzymatic complexes for DNA repair, particularly in the processes of NER (Nucleotide Excision Repair) and BER (Base Excision Repair). The expression of XRCC1 increased by just over 10%, but this increase was very mild and, as confirmed by other studies, it can range from mild induction to actual inhibition in response to PM exposure.

The ERCC1, XPF, and XPA proteins are involved in the NER process, with XPA (Xeroderma Pigmentosum Group A) recognizing the damage and stabilizing the structure around the lesion. ERCC1 forms a heterodimer with XPF to make a cut at the 5’ site of the lesion. The expression of XPA showed a very low average increase, resulting in higher variability, making it not a good marker for damage. Other studies confirmed that PM exposure did not induce a high response for XPA [36], indicating that this protein may not be a sensitive indicator of genomic damage in this condition.

XPA and XPF are regulated similarly in expression, as expected, since both operate in the same DNA repair mechanism.

A recent study evaluated a causal link between prenatal exposure to pollutants, including PM_2.5, and the methylation of promoters of DNA repair genes, analyzing placental DNA samples. The authors observed a significant association between PM_2.5 exposure and methylation of the XPF promoter, which led to a reduction in expression, while the ERCC1 promoter remained unaffected. The authors suggest that PM_2.5 exposure might have caused a loss of stability in the endonuclease complex, which, although it could still perform the incision at the 5’ site of the damage, a task mainly performed by ERCC1, it cannot carry out optimal repair due to the lack of the essential stabilizing action of XPF. This provides a potential mechanism of PM-induced damage, which impairs proper genomic damage repair processes by altering the stability of the complexes that operate in these processes [36].

IL-8 showed the highest and most significant induction in expression, highlighting the elevated inflammatory potential of the analyzed samples. IL-8 is a chemokine that attracts neutrophils and plays a key role in innate defense against bacterial infections, making it an excellent marker of lung inflammation. Elevated levels of IL-8 are associated with the development of chronic obstructive pulmonary diseases, bronchitis, fibrosis, asthma, and allergies [37,38]. Recently, it has been shown that human lung fibroblasts exposed to various inflammatory stimuli increase IL-8 production and other inflammatory chemokines or cytokines through the overexpression and heightened sensitivity of the TRPA1 receptor (transient receptor potential channel ankyrin 1), normally expressed at the neuronal level, contributing to an allergic/asthmatic phenotype and airway hyperresponsiveness [32]. Additionally, another study found that this receptor is also associated with inflammatory lung damage induced by PM_2.5 [28]. Numerous studies show that PM (particularly PM_2.5 and PM_0.1) alters normal immune defenses at the lung level through the induction of inflammatory mechanisms, both in vitro and in vivo, with IL-8 being one of the most commonly upregulated chemokines. The production of inflammatory cytokines stimulated by PM may sometimes be the cause of increased severity in pre-existing pulmonary inflammatory diseases. Capistrano et al. [39] infected human lung fibroblasts exposed to biomass combustion PM extracts and observed significantly increased production of IL-8 compared to unexposed cells, suggesting a potential synergistic mechanism between PM effects and viral infections, leading to exacerbated inflammation through IL-8 production. This aspect is even more relevant from a health perspective in occupational settings, where exposure to potential viral, bacterial, or fungal agents is more frequent, especially in biomass treatment sectors, as shown by numerous cases of infectious diseases (e.g., aspergillosis). Therefore, IL-8 confirms its role as a crucial inflammation marker for exposure to environmental pollutants.

The results show a highly significant positive correlation between the IL-8 expression increase and the endotoxin concentration in each sample, while no significant correlation with the PM₁₀ mass was observed. These findings align with the literature, in which the negative impact of particulate matter on human health is more closely associated with the chemical and biological composition of PM than its mass. Endotoxins have an inherently inflammatory nature and are powerful activators of the NLRP3 inflammasome complex, which, through caspase 1 activation, is responsible for the release of pro-inflammatory cytokines and interleukins. J. Liu et al. [39] evaluated the inflammatory potential of airborne endotoxins emitted from composting plants, finding significantly elevated concentrations near the plants, as well as in surrounding areas. Cells exposed to bioaerosol extracts released high amounts of IL-8 and IL-6, which correlated significantly with endotoxin levels. When the cells were treated with Polymyxin B (a specific endotoxin inhibitor), the pro-inflammatory effect was inhibited, and the cytokine stimulation was found to be higher for airborne endotoxins than for commercially purified E. coli endotoxins.

A significant variation in gene expression was found when considering the types of plants. Normally, air and bioaerosol characterization in these environments reveals different compositions of airborne microbial communities in various plants and specific operational areas within them. This leads to heterogeneity in the overall PM composition and, consequently, in the resulting biological effects. A previous study conducted on similar plants revealed significantly higher microbial contamination in the air in organic waste treatment plants, with a prevalence of bacterial species in biomass entry and storage areas compared to plants treating different matrices, where the fungal component predominated [17,40]. Therefore, gene expression, cytotoxic activity, and PM concentration and endotoxins seem to differ significantly based on the plant type or specific operational phase, rather than the PM fraction. Specifically, IL-8 was significantly induced by PM from composting plants, which appear to produce more inflammatory PM, while XPF was significantly down-regulated by PM from lignocellulosic matrix plants, where ERCC1 significantly increased, indicating a PM that seems to impair genomic damage repair mechanisms more than others.

Although full-scale biomass treatment plants are rarely assessed for genotoxic potential, this study presents several limitations. One of the main issues is the lack of characterization of the biological and chemical components of the particulate matter. While the identification of individual agents could provide valuable insights, the objective of this study was to assess the genotoxicity of the complex, multi-component pollution mixtures typical of the waste treatment methods examined, rather than to isolate a specific microbial or chemical signature.

Another limitation is the inability to provide a mechanistic interpretation of the observed genotoxic effects, which would be desirable in future research. This study focused on organic extracts of particulate matter, which have been shown in the literature to have the strongest association with genotoxic effects. However, we cannot rule out the potential contribution of aqueous extracts, typically examined for reactive oxygen species (ROS) activity. Future research could include parallel sampling streams, one dedicated to extract-based genotoxic assays and another reserved for molecular endpoints.

Furthermore, the analytical techniques employed in this study could be expanded. Incorporating methods such as RNA sequencing or western blotting could provide complementary evidence. Additionally, validating the findings across multiple cell lines would strengthen the reliability of the data. For a genotoxicity study, it is crucial to select a genetically stable cell line, ideally derived from non-cancerous tissue, such as the HELF line, not so easy to find. Lastly, while gene expression modulation serves as an early biological response that precedes direct DNA damage—such as that detectable through the Comet assay—incorporating additional genotoxic endpoints would strengthen the overall case and provide a more comprehensive evaluation of the genotoxic risks associated with biomass treatment emissions.

5. Conclusions

This study evaluated the biological effects of exposure to organic PM extracts in terms of cytotoxicity and the modulation of specific DNA repair genes and IL-8, markers of genomic damage and inflammation.

The XTT assay results highlighted a stimulatory effect, although evident only in a few samples, which had not clearly emerged in previous toxicological studies on PM. Further analysis, perhaps with a larger number of samples, could clarify whether these are confounding effects or actual mechanisms of mitochondrial activity modulation, possibly related to the presence of biological components with variable oxidative potential influencing PM, and potentially useful as additional evaluation parameters.

Composting plants appear to pose the highest biological risk, both due to elevated concentrations of airborne pollutants and the more inflammatory potential of the PM produced.

In conclusion, this specific PM type showed toxicity characterized by a significant inflammatory potential, evident from IL-8 transcription, correlated with endotoxin levels and the generation of genomic damage, and leading to enhanced stimulation of some repair mechanisms, such as the significant increase in ERCC1. At the same time, PM might have altered its proper functioning, given the opposite regulation observed between ERCC1 and XPF, which normally function as dimers. These effects are comparable to evidence in the literature, even if related to PM of primarily biological origin, and emphasize the still under-explored contribution of bioaerosols to the genotoxic and cytotoxic potential of particulate matter.

Research on the PM biological component is still limited, and in vitro toxicological studies remain a valuable tool for highlighting early biomolecular effects to complement epidemiological evidence and to ensure proper risk assessment and management for human health. A better understanding of damage mechanisms could provide useful tools for introducing reliable and sensitive early effect biomarkers for the prevention or treatment of diseases associated with PM exposure and other pollutants, which could eventually serve as biological exposure indicators for workers’ health surveillance.