Di-(2-ethylhexyl)phthalate (bis(2-ethylhexyl) benzene-1,2-dicarboxylate; DEHP), di-isobutylphthalate (bis(2-methylpropyl) benzene-1,2-dicarboxylate; DiBP), di-isononylphthalate (bis(7-methyloctyl) benzene-1,2-dicarboxylate; DiNP), di-(2-ethylhexyl) adipate (bis(2-ethylhexyl) hexanedioate; DEHA), and di-(2-ethylhexyl)terephthalate (bis(2-ethylhexyl) benzene-1,4-dicarboxylate; DEHT; DEHTP) were obtained from Sigma-Aldrich, St. Louis, Missouri, USA. Di-isononyl-1,2-cyclohexanedicarboxylic acid (bis(7-methyloctyl)-cyclohexane-1,2-dicarboxylate; Hexamoll^® DINCH; DINCH) was obtained from BASF, Ludwigshafen am Rhein, Germany. All chemicals were dissolved in dimethyl sulfoxide (DMSO) (Sigma-Aldrich, St. Louis, Missouri, USA), and stored at -20°C. For mixture exposures, an equimolar six-component mixture was prepared by combining each compound at identical molar proportions (DEHP: DiBP: DiNP: DEHA: DEHT: DINCH = 1: 1: 1: 1: 1: 1). The final mixture concentration applied to cells reflects the sum of all six components.

NCI-H295R cells were cultured in Dulbecco’s modified Eagle’s medium/F-12 medium (Gibco, Burlington, Ontario, USA), including 10% fetal bovine serum (Sigma-Aldrich, StLouis, Missouri, USA), and 1% insulin-transferrin-selenium (Sigma-Aldrich, St. Louis, Missouri, USA). Cells were kept at 37°C and 5% CO₂ in 175 cm² flasks (Greiner bio-one, Frickenhausen, Germany), medium was renewed every second day.

Once a week, given sufficient cell density assessed by microscopic control, cells were harvested using phosphate-buffered saline (Sigma-Aldrich, St. Louis, Missouri, USA), and EDTA-trypsin (Sigma-Aldrich, St. Louis, Missouri, USA). For cell counting the Countess II FL (Life Technologies, Carlsbad, California, USA) was used, and if not needed for cell treatments, remaining cells were split into new cell flasks, sticking to a 1:3 ratio.

NCI-H295R cells) were disseminated in 96-well black plates (Corning Incorporated, Corning, New York, USA) with optically clear flat bottoms. A seeding density of 5x10⁴/well was used in a total volume of 100 µL of complete medium. After 24 h of incubation at 37°C and 5% CO_2, cells were treated with the vehicle DMSO (1%), DEHP, DiBP, DiNP, DEHA, DEHT, or DINCH, of 0.001, 0.05, 0.1, 0.25, 0.5, 1, 10, 25, 100, 250, and 1000 µM. For the combinatory treatment, the six-component-mixture was used in the same concentrations as in the single-substance treatments. Working solutions were obtained using the DMSO-dissolved stocks and diluted in complete medium, reaching a final DMSO concentration of 1%. In all experiments, concentrations are reported and interpreted on a nominal basis. A total volume of 100 µL was added to each well. Cells were exposed to the treatment for 72 h. After the end of the incubation, 150 µL supernatant of eight identically treated wells was collected, pooled, centrifuged at 800 rpm at 23°C, and stored at -20°C for later steroid quantification. Each treatment was performed in three independent experiments. Concentration selection and overall experimental framework were based on OECD Test Guideline 456 with minor adaptations to fit the specific objectives and constraints of the present study (43).

Remaining cells were analyzed for cell viability after treatment using CellTiter-Glo^® Luminescent assay (Promega, Madison, Wisconsin, USA) following the manufacturer’s protocol. Cells were incubated with CellTiter-Glo^® reagent for 10 min, and luminescence was determined with 1420 multilabel counter Victor³ (Perkin Elmer, Waltham, Massachusetts, USA). Each experiment was performed in three independent runs in 8-plicates.

To analyze the steroid metabolome, liquid chromatography tandem mass spectrometry (LC-MS/MS) using a Sciex 6500+ QTRAP (SCIEX, Framingham, USA) MS-system linked with an Agilent 1290 HPLC-system (G4226A, autosampler, infinityBinPump, G1316C column-oven, G1330B thermostat; Santa Clara, USA) was utilized. Using the MassChrom-Steroids in Serum Plasma^® IVDR conform kit (Chromsystems^®, Gräfelfing, Germany) (44) 15 steroids (aldosterone, androstenedione, corticosterone, cortisol, cortisone, dehydroepiandrosterone (DHEA), dehydroandrosterone-sulfate (DHEAS), 11-deoxycorticosterone, 11-deoxycortisol, 21-deoxycortisol, dihydrotestosterone (DHT), estradiol, 17-hydroxyprogesterone (17-OHP), progesterone, and testosterone) were quantified. After off-line solid phase extraction of 500 µl medium supernatant, 15 µL were used for analysis. With Analyst Software (1.6.3), concentrations were calculated via six-point calibration and 1/x weighing. Commercial quality controls and periodic participations in ring trails ensured the correctness of the described measurements. Absolute steroid values and limits of quantification per analyte are provided in Supplementary Tables S6, S7.

NCI-H295R cells were seeded in 12-well plates (Corning Incorporated, Corning, NY, USA) at a density of 1 × 10⁶ cells/well and exposed to 1% DMSO or the combinatory mixture of all six test substances, each applied at concentrations of 0.25, 0.5, 1, 2.5, 5, 10, and 25μµM. After 72μhours of exposure, cells were harvested using PBS and EDTA-trypsin, centrifuged, and washed with PBS. Total RNA was extracted using the Maxwell^® RSC simplyRNA Tissue Kit (Promega, Madison, WI, USA) according to the manufacturer’s instructions and stored at −80°C until further use.

A total of 1 µg RNA per sample was reverse transcribed using the QuantiTect^® Reverse Transcription kit (Qiagen, Venlo, Netherlands) in a final volume of 40 µL according to manufacturer’s protocol. cDNA was diluted by 1:2 in nuclease-free water and stored at -20°C for further analysis. NRT controls were included to confirm the absence of genomic DNA contamination.

Duplex real-time PCR was conducted to simultaneously quantify β-actin (ACTB) as a housekeeping gene and various genes of interest (StAR, CYP11B1, CYP11B2, CYP17A1, CYP21A2, HSD3B2, AGTR1, MC2R, SF-1). Primer/probe interactions (FAM- and CY5-labeled) were evaluated in both single and duplex reactions to confirm specificity and amplification efficiency. Primer sequences and TaqMan Assay IDs are provided in the Supplementary Tables S2. Reactions were performed using a C1000 Touch Thermal Cycler and CFX96 Real Time System (Bio-Rad, Hercules, CA, USA) in a final volume of 20 µL, containing 6µL nuclease-free water, 10 µL TaqMan^® Gene Expression Master Mix (Thermo Fisher Scientific, Waltham, MA, USA), 1 µL of CY5-labeled Actb primer/probe mix, 1 µL of FAM-labeled gene-specific primer/probe mix (Thermo Fisher Scientific) and 2 µL of diluted cDNA per reaction. Amplification was conducted under the following conditions: initial hold at 50°C for 2 minutes, followed by enzyme activation and initial denaturation at 95°C for 10 minutes. Then, 50 amplification cycles, each consisting of denaturation at 95°C for 15 seconds, and annealing/extension at 60°C for 1 minute, during which fluorescence data were collected.

mRNA expression was quantified using the comparative Ct (2^−ΔΔCt) method. Each condition was analyzed in three independent biological replicates and three technical. Results are provided in Supplementary Table S5.

Statistical analysis was performed with GraphPad Prism Software (version 9.01 for Windows, GraphPad Software, Inc.). Data regarding cell viability assays, steroid quantification, calculated substrate-product ratios, and mRNA expression levels in response to chemical treatment were calculated as fold changes compared to vehicle-treated controls. All data are expressed as mean ± standard deviation (SD) unless otherwise indicated. For statistical analysis, one-way ANOVA followed by Dunnett’s post hoc test was performed using vehicle-treated controls as reference. A p-value <0.05 was considered statistically significant.

To assess potential alterations in enzyme-dependent steroidogenic steps, substrate-to-product ratios were calculated for each conversion and normalized to vehicle controls using the following formula, indicating relative activity:

Deviations indicate alterations in steroidogenic enzyme activity. Specifically, values >1 suggest an inhibition of the respective enzymatic activity, while values <1 indicate a potential enhancement.

Chemical structures were designed in ChemSketch (FreeWare) 2024.1.4 (ACD/Labs, Toronto, Canada).

Cell viability was not affected by vehicle (1% DMSO) and low-dose treatment (<100 µM) of DEHP, DiBP, DiNP, DEHA, DEHT, DINCH, and the combined mixture after 72 hours of treatment (Figure 1). However, cell viability decreased significantly below 80% at concentrations greater than 100 µM of DEHP, DiBP, DiNP, and the mixture. The calculated median lethal concentration (LC₅₀) for DiBP was 1.91 mM. Since none of the other treatments reduced viability below 50%, no representative LC₅₀ values were determined for these compounds.

The most notable alterations were observed in enzymatic steps catalyzed by steroid 11β-hydroxylase (CYP11B1): the conversion of 11-deoxycorticosterone to corticosterone was significantly enhanced, as indicated by reduced substrate-to-product ratios, e.g., 0.22 ± 0.02, 0.25 ± 0.02, and 0.16 ± 0.01 at 5μµM treatment with DEHP, DiNP, and DINCH (each p ≤ 0.0001; full data available in the Supplementary Table S4). Similarly, the conversion of 17-hydroxyprogesterone to 21-deoxycortisol (catalyzed by CYP11B1) occurred more frequently in treated cells compared to vehicle controls, with the highest observed changes at 5μµM DEHP, DEHT, and the mixture (0.13 ± 0.01, 0.46 ± 0.04, 0.17 ± 0.01; each p ≤ 0.01).

Regarding CYP17A1 activity, a significant increase in 17,20-lyase-mediated conversion of 17-hydroxyprogesterone to androstenedione was detected, particularly following DEHP and DEHT exposure (e.g., 0.68 ± 0.02 at 2.5 µM; p ≤ 0.0001), while the 17α-hydroxylase activity of the same enzyme was unaffected.

The ratio referring to the hydroxylation of 17-hydroxyprogesterone to 11-deoxycortisol (CYP21A2) decreased under different exposures (e.g., 0.56 ± 0.06 at 25μµM DEHP; p ≤ 0.001; 0.61 ± 0.07 at 10μµM DiNP; p ≤ 0.001; 0.75 ± 0.01 at 2.5μµM DEHT; p ≤ 0.0001), whereas the subsequent conversion of 21-deoxycortisol to cortisol was impaired, as shown by increased product-to-substrate ratios.

The relative activity of aromatase (CYP19A1), producing estradiol from testosterone, has been slightly increased by DEHP (e.g., 0.58 ± 0.05 at 10 µM; p = 0.1286) and DiNP (e.g., 0.61 ± 0.02 at 10 µM; p < 0.0001).

In addition, the ratio of cortisol to its inactive metabolite cortisone - reflecting HSD11B2 activity- was consistently elevated across all treatments, suggesting reduced enzymatic function. The strongest changes were observed at 5μµM DEHP (4.48 ± 0.29; p ≤ 0.0001), DiNP (4.04 ± 0.26; p ≤ 0.0001), and the mixture (4.12 ± 0.16; p ≤ 0.01). Furthermore, 5α-reductase (3-oxo-5α-steroid 4-dehydrogenase, SRD5) activity, responsible for the conversion of testosterone to the more potent dihydrotestosterone, appeared increasingly inhibited with higher concentrations of DEHP, DEHT, DINCH, and the mixture, e.g., 3.92 ± 0.90 at 25 µM of DINCH (p ≤ 0.0001).

No consistent changes were seen in the relative activity of CYP11B2, HSD3B2, SULT2A1, or HSD17B3.

Gene expression of essential mediators of steroidogenesis in response to the mixture of DEHP, DiBP, DiNP, DEHA, DEHT, and DINCH was quantified via RT-qPCR (Figure 3; full data available in the Supplementary Table S5).

Respective treatment resulted in a modest, concentration-dependent elevation of StAR mRNA expression compared with the 1% DMSO control (Figure 3A). Concentrations ≥ 2.5 µM produced a slight upregulation, e.g.1.53 ± 0.30 at 2.5 µM (p = 0.6134). mRNA levels of CYP11B1 and CYP11B2 were strongly elevated by treatment with the mixture compared to vehicle control (Figures 3B, C). For instance, treatment with 10 µM resulted in 11.82 ± 4.40-fold changes in CYP11B1 and 44.12 ± 9.77-fold changes in CYP11B2 expression (both p ≤ 0.0001). Also, expression of HSD3B2 was increased by treatment to a maximum fold change at 10 µM (4.26 ± 0.87; p ≤ 0.0001) (Figure 3F). Expression of CYP17A1 was even affected at nanomolar concentrations, e.g., 1.63 ± 0.24 at 500 nM (p ≤ 0.0001), with maximum values at 25 µM (2.10 ± 0.18; p ≤ 0.0001) (Figure 3D). mRNA levels of CYP21A2 were found to be increased by treatment, peaking at 5 µM (2.28 ± 0.42; p ≤ 0.0001) (Figure 3E).

Expression of angiotensin II receptor 1 (AGTR1) was significantly elevated by treatment with 2.5 µM (2.08 ± 0.30; p ≤ 0.0001) and 5 µM (1.86 ± 0.38; p ≤ 0.0001) (Figure 3G). Meanwhile, melanocortin 2 receptor (MC2R) expression was markedly increased by doses >2.5 µM, with the maximum observed fold change at 25 µM (7.06 ± 0.48; p ≤ 0.0001) (Figure 3H). mRNA levels of SF-1, however, were not significantly altered by the treatment (Figure 3I).

Epidemiological studies have shown a correlation between elevated cortisol in human samples (e.g., hair, urine) and phthalate exposure, particularly in infants, adolescents, and predominantly in females (17–19). These substances have been previously described as neuroendocrine disruptors affecting all components of the HPA-axis, including dysregulation of CRH, ACTH, cortisol, carrier proteins, peripheral steroid receptors, and deactivating enzymes (29, 30, 46). Behavioral effects have been consistently demonstrated in phthalate-treated rodent models, where HPA-axis dysregulation coincided with pronounced anxiety-like phenotypes and impaired stress response (28, 29).

In the present study, treatment with phthalates or non-phthalate substitutes increased cortisol levels while reducing its derivative cortisone. Also, the strong increase in 21-deoxycortisol, an intermediate of an alternative cortisol synthesis route, and the increase of CYP11B1 expression by 20-fold further support the activation of the glucocorticoid pathway. Consistent with the observed changes in HSD11B2 activity, previous reports indicate that phthalates can inhibit hydroxysteroid dehydrogenases (21, 47), and specifically HSD11B2 in renal, ovarian, and testicular tissue (48, 49), increasing cortisol-cortisone ratio (21, 31). HSD11B2 is a key enzyme in aldosterone and cortisol target tissues (e.g., kidney, gastrointestinal tract), converting cortisol to cortisone to prevent inappropriate mineralocorticoid receptor (MR) activation. Impaired cortisol deactivation through phthalates and their substitutes therefore results in an accumulation of active cortisol and potentially consequent aberrant MR activation. In vivo, such MR activation by cortisol driven by phthalate-induced HSD11B2 inhibition has been associated with hypertensive phenotypes (26), providing a plausible mechanistic link between the present in vitro findings and reported cardiovascular outcomes.

The present results indicate that with increased secretion of mineralocorticoids and/or by inhibition of HSD11B1/2, more active ligands on the MR are present following phthalate and substitute treatment. In epidemiological association studies, DEHP concentrations were, in fact, higher in hypertensive than normotensive infants (26). Furthermore, removing DEHP from medical equipment (e.g., IV fluids) reduced the incidence of hypertension in patients of neonatal care units, whereas reintroduction of DEHP-containing products led to an increase in cases with hypertension comparable to pre-interventional levels (50). Epidemiological evidence suggests that differences between high-molecular-weight phthalates, such as DEHP, DiNP, DiDP, and low-molecular-weight phthalates exist, as the latter have no significant effect on blood pressure [e.g., DiBP (51)], which might be reflected in the present study by their small impact on adrenal mineralocorticoids.

The hypertensive effect of DEHP was repeatedly confirmed in in vivo exposure studies, as well as epidemiological cohort and cross-sectional studies that measured urinary DEHP metabolites and found significantly higher blood pressure levels in exposed populations (26, 51–54). Moreover, enhanced expression of RAAS components, including renin (52), angiotensin-converting enzyme (ACE), angiotensin II, and AGTR1 (53, 54) has been reported. Similar effects have been detected for DiNP exposure (54). Furthermore, urine levels of DEHP were associated with altered fluid homeostasis (32), increased sodium retention, and elevated expression of renal sodium channels, including epithelial sodium channel (ENaC), and phosphorylated sodium chloride cotransporter (pNCC), being indicators of MR activation (31). Additionally, prenatal exposure studies show an inhibitory effect of DEHP on the expression of MR, while MR-associated genes (e.g., ENaC) were increased in kidney tissue (49).

This study is the first to reveal that the adrenal compounds of the RAAS system, namely AGTR1, steroidogenic enzymes (StAR, HSD3B2, CYP11B1, CYP11B2), and the respective mineralocorticoids are affected by phthalates and their substitutes. Therefore, the reported direct effects on adrenal steroidogenesis, including aldosterone secretion, might explain the underlying cause of plasticizer-induced hypertension.

Additionally, phthalates may act as selective peroxisome proliferator-activated receptor gamma (PPARγ) modulators in steroidogenic tissue (55), as previously suggested for monoethylhexyl phthalate (MEHP), including selective coactivator recruitment (e.g., PPARγ coactivator 1-alpha) and potentially enhancing CYP11B1 and CYP11B2 expression, as well as aldosterone secretion (56).

Further cellular stress-mediated mechanisms such as interference with cAMP/PKA/SF-1 signaling (45), oxidative stress, or mitochondrial dysfunction may contribute to enzyme induction and altered hormone production. Phthalates have been linked to oxidative stress, which is often discussed in connection with disrupted steroidogenesis. However, available evidence indicates that phthalates induce significant oxidative stress only at concentrations exceeding 100 µM in steroidogenic cell models (23, 57–61). For instance, while DBP has been shown to induce oxidative stress concomitant with steroid alterations in H295R cells, these effects occurred only at concentrations exceeding those used in the present study (23). Therefore, while oxidative stress is a plausible mechanism, it is unlikely to account for the steroidogenic changes observed here at lower, sub-cytotoxic phthalate concentrations.

Phthalate exposure resulted in biphasic, non-monotonic dose–response curves with maximal effects between 1 and 10 µM. Non-monotonicity is well documented for several phthalates and can arise from receptor selectivity, receptor deregulation or desensitization, incipient cytotoxicity, autocrine feedback, pathway-level feedback inhibition, or dose-dependent activation of adaptive stress responses (62). In adrenal steroidogenesis specifically, low-dose phthalate exposure may enhance upstream signaling or enzyme activity (e.g., increased MC2R/ACTH responsiveness or upregulation of CYP11B1/CYP11B2), whereas higher concentrations may induce compensatory downregulation, reduced substrate availability, or early cytoprotective stress responses that attenuate steroid output. These mechanisms are consistent with the peak effects observed at 2.5–5 µM and the subsequent decline at higher concentrations.

The anti-androgenic effect of phthalates, especially DEHP, and their involvement in the development of reproductive impairment have already been described (14). Epidemiological data showed an inverse correlation of phthalate metabolites with androstenedione, DHEA, and testosterone, and a positive correlation with estradiol levels (63). The presented results confirm and extend this knowledge by showing that prominent adrenal androgens, DHEA and androstenedione, are impaired by phthalates, with the most pronounced changes observed following DEHP treatment. Additionally, relative 5α-reductase activity was significantly inhibited by DEHP, DEHT, and DINCH, reaching only ~25% of control levels. This inhibition results in decreased DHT production and may thereby represent a potential mechanism underlying the pathogenesis of phthalate-associated hypospadias (64). In contrast, non-phthalate plasticizers, namely DEHA, DEHT, and DINCH, resulted in elevated levels of androstenedione and DHEA, suggesting an androgenic effect on H295R cells.

To simulate realistic co-exposure conditions, cells were treated with an equimolar mixture of all six compounds. This combined treatment altered key steroid hormones (e.g., progesterone, corticosterone, 21-deoxycortisol, cortisol) - even at doses where individual substances showed no effect, indicating supra-additive disruption. This observation aligns with European biomonitoring data, where mixture risk assessments revealed health risks in 17% of individuals that would have been missed by single-compound evaluations (41). Thus, mixture effects should be routinely considered in EDC testing to better reflect real-world combined exposures.

These results underscore the steroidogenesis- and thus endocrine-disrupting potential of both classical phthalates and commonly used substitutes (e.g., terephthalates, cyclohexanoates). Further, the urgent need to reconsider toxicological assessment approaches and to adapt in vitro test settings, including the use of expanded steroid panels, is highlighted. If confirmed in vivo, these findings could have important implications for the understanding and management of steroid-linked disorders and environmental adrenal disruption, warranting further evaluation.