Predictive, integrative, and regulatory aspects of AI-driven computational toxicology - Highlights of the German Pharm-Tox Summit (GPTS) 2024.

The 9th German Pharm-Tox Summit (GPTS) and the 90th Annual Meeting of the German Society for Experimental and Clinical Pharmacology and Toxicology (DGPT) took place in Munich from March 13-15, 2024. The event brought together over 700 participants from around the world to discuss cutting-edge developments in the fields of pharmacology and toxicology as well as scientific innovations and novel insights. A key focus of the conference was on the rapidly increasing role of computational toxicology, artificial intelligence (AI), and machine learning (ML) into the field, marking a shift away from traditional methods and allowing the reduction of animal testing as primary tool for toxicological risk assessment. Tools such as Toxometris.ai showcased the potential of AI-based risk assessments for predicting carcinogenicity, offering more ethical and efficient alternatives. Additionally, computer-driven models like computer-aided pattern analysis (C@PA) for drug toxicity prediction were presented, emphasizing the growing role of chem- and bioinformatic applications in computational sciences. Throughout the summit, there was a strong focus on the need for regulatory innovation to support the adoption of these advanced technologies and ensure the safety and sustainability of chemical substances and drugs.

Objectives and achievements of the HUMN project on its 26th anniversary.

Micronuclei (MN) are a nuclear abnormality that occurs when chromosome fragments or whole chromosomes are not properly segregated during mitosis and consequently are excluded from the main nuclei and wrapped within nuclear membrane to form small nuclei. This maldistribution of genetic material leads to abnormal cellular genomes which may increase risk of developmental defects, cancers, and accelerated aging. Despite the potential importance of MN as biomarkers of genotoxicity, very little was known about the optimal way to measure MN in humans, the normal ranges of values of MN in healthy humans and the prospective association of MN with developmental and degenerative diseases prior to the 1980's. In the early 1980's two important methods to measure MN in humans were developed namely, the cytokinesis-block MN (CBMN) assay using peripheral blood lymphocytes and the Buccal MN assay that measures MN in epithelial cells from the oral mucosa. These discoveries greatly increased interest to use MN assays in human studies. In 1997 the Human Micronucleus (HUMN) project was founded to initiate an international collaboration to (i) harmonise and standardise the techniques used to perform the lymphocyte CBMN assay and the Buccal MN assay; (ii) establish and collate databases of MN frequency in human populations world-wide which also captured demographic, lifestyle and environmental genotoxin exposure data and (iii) use these data to identify the most important variables affecting MN frequency and to also determine whether MN predict disease risk. In this paper we briefly describe the achievements of the HUMN project during the period from the date of its foundation on 9th September 1997 until its 26th Anniversary in 2023, which included more than 200 publications and 23 workshops world-wide.

Genotoxicity of selected cannabinoids in human lymphoblastoid TK6 cells.

Natural non-psychoactive cannabinoids such as cannabigerol (CBG), cannabidiol (CBD), cannabichromene (CBC), cannabidivarin (CBDV), and cannabinol (CBN) are increasingly consumed as constituents of dietary products because of the health benefits claims. Cannabinoids may reduce certain types of pain, nausea, and anxiety. Anti-inflammatory and even anti-carcinogenic properties have been discussed. However, there are insufficient data available regarding their potential (geno-)toxic effects. Therefore, we tested CBG, CBD, CBC, CBDV, and CBN for their genotoxic potential and effects on mitosis and cell cycle in human lymphoblastoid TK6 cells. The selected cannabinoids (except CBDV) induced increased micronuclei formation, which was reduced with the addition of a metabolic activation system (S9 mix). CBDV induced micronuclei only after metabolic activation. Mitotic disturbances were observed with all tested cannabinoids, while G1 phase accumulation of cells was observed for CBG, CBD and CBDV. The genotoxic effects occurred at about 1000-fold higher concentrations than are reported as blood levels from human consumption. However, the results clearly indicate a need for further research into the genotoxic effects of cannabinoids. The mechanism of the mitotic disturbance, the shape of the dose-response curves and the possible effects of mixtures of cannabinoids are aspects which need clarification.

Obesity-associated non-oxidative genotoxic stress alters trophoblast turnover in human first-trimester placentas.

Placental growth is most rapid during the first trimester (FT) of pregnancy, making it vulnerable to metabolic and endocrine influences. Obesity, with its inflammatory and oxidative stress, can cause cellular damage. We hypothesized that maternal obesity increases DNA damage in the FT placenta, affecting DNA damage response and trophoblast turnover. Examining placental tissue from lean and obese non-smoking women (4-12 gestational weeks), we observed higher overall DNA damage in obesity (COMET assay). Specifically, DNA double-strand breaks were found in villous cytotrophoblasts (vCTB; semi-quantitative γH2AX immunostaining), while oxidative DNA modifications (8-hydroxydeoxyguanosine; FPG-COMET assay) were absent. Increased DNA damage in obese FT placentas did not correlate with enhanced DNA damage sensing and repair. Indeed, obesity led to reduced expression of multiple DNA repair genes (mRNA array), which were further shown to be influenced by inflammation through in vitro experiments using tumor necrosis factor-α treatment on FT chorionic villous explants. Tissue changes included elevated vCTB apoptosis (TUNEL assay; caspase-cleaved cytokeratin 18), but unchanged senescence (p16) and reduced proliferation (Ki67) of vCTB, the main driver of FT placental growth. Overall, obesity is linked to heightened non-oxidative DNA damage in FT placentas, negatively affecting trophoblast growth and potentially leading to temporary reduction in early fetal growth.

Disruptive effects of plasticizers bisphenol A, F, and S on steroidogenesis of adrenocortical cells.

Introduction: Endocrine disrupting chemicals (EDCs) are known to interfere with endocrine homeostasis. Their impact on the adrenal cortex and steroidogenesis has not yet been sufficiently elucidated. This applies in particular to the ubiquitously available bisphenols A (BPA), F (BPF), and S (BPS). Methods: NCI-H295R adrenocortical cells were exposed to different concentrations (1nM-1mM) of BPA, BPF, BPS, and an equimolar mixture of them (BPmix). After 72 hours, 15 endogenous steroids were measured using LC-MS/MS. Ratios of substrate and product of CYP-regulated steps were calculated to identify most influenced steps of steroidogenesis. mRNA expression of steroidogenic enzymes was determined by real-time PCR. Results: Cell viability remained unaffected at bisphenol concentrations lower than 250 µM. All tested bisphenols and their combination led to extensive alterations in the quantified steroid levels. The most profound fold changes (FC) in steroid concentrations after exposure to BPA (>10µM) were seen for androstenedione, e.g. a 0.37±0.11-fold decrease at 25µM (p≤0.0001) compared to vehicle-treated controls. For BPF, levels of 17-hydroxyprogesterone were significantly increased by 25µM (FC 2.57±0.49, p≤0.001) and 50µM (FC 2.65±0.61, p≤0.0001). BPS treatment led to a dose-dependent decrease of 11-deoxycorticosterone at >1µM (e.g. FC 0.24±0.14, p≤0.0001 at 10µM). However, when combining all three bisphenols, additive effects were detected: e.g. 11-deoxycortisosterone was decreased at doses >10µM (FC 0.27±0.04, p≤0.0001, at 25µM), whereas 21-deoxycortisol was increased by 2.92±0.20 (p≤0.01) at 10µM, and by 3.21±0.45 (p≤0.001) at 50µM. While every measured androgen (DHEA, DHEAS, androstenedione, testosterone, DHT) was lowered in all experiments, estradiol levels were significantly increased by BPA, BPF, BPS, and BPmix (e.g. FC 3.60±0.54, p≤0.0001 at 100µM BPF). Calculated substrate-product ratios indicated an inhibition of CYP17A1-, and CYP21A2 mediated conversions, whereas CYP11B1 and CYP19A1 showed higher activity in the presence of bisphenols. Based on these findings, most relevant mRNA expression of CYP genes were analysed. mRNA levels of StAR, CYP11B1, and CYP17A1 were significantly increased by BPF, BPS, and BPmix. Discussion: In cell culture, bisphenols interfere with steroidogenesis at non-cytotoxic levels, leading to compound-specific patterns of significantly altered hormone levels. These results justify and call for additional in-vivo studies to evaluate effects of EDCs on adrenal gland functionality.

The Comet Assay as a Tool in Human Biomonitoring Studies of Environmental and Occupational Exposure to Chemicals-A Systematic Scoping Review.

Biomonitoring of human populations exposed to chemical substances that can act as potential mutagens or carcinogens, may enable the detection of damage and early disease prevention. In recent years, the comet assay has become an important tool for assessing DNA damage, both in environmental and occupational exposure contexts. To evidence the role of the comet assay in human biomonitoring, we have analysed original research studies of environmental or occupational exposure that used the comet assay in their assessments, following the PRISMA-ScR method (preferred reporting items for systematic reviews and meta-analyses extension for scoping reviews). Groups of chemicals were designated according to a broad classification, and the results obtained from over 300 original studies (n = 123 on air pollutants, n = 14 on anaesthetics, n = 18 on antineoplastic drugs, n = 57 on heavy metals, n = 59 on pesticides, and n = 49 on solvents) showed overall higher values of DNA strand breaks in the exposed subjects in comparison with the unexposed. In summary, our systematic scoping review strengthens the relevance of the use of the comet assay in assessing DNA damage in human biomonitoring studies.

The ameliorative effect of rosmarinic acid and epigallocatechin gallate against doxorubicin-induced genotoxicity.

Doxorubicin (Dox), an effective anticancer agent, is known for its genotoxic effects on normal cells. Phenolic compounds, renowned for their antitumor, antioxidant, and antigenotoxic properties, have gained prominence in recent years. This study investigates the individual and combined protective effects of rosmarinic acid (RA) and epigallocatechin gallate (EGCG) against Dox-induced genotoxicity using various in vitro test systems. The synergistic/antagonistic interaction of these combinations on Dox's chemotherapeutic effect is explored in breast cancer cell lines. Both RA and EGCG significantly mitigate Dox-induced genotoxicity in comet, micronucleus, and Ames assays. While Dox exhibits higher selectivity against MCF-7 cells, EGCG and RA show greater selectivity against MDA-MB-231 cells. The coefficient of drug interaction reveals a synergistic effect when RA or EGCG is combined with Dox in breast cancer cells. In conclusion, both EGCG and RA effectively reduce Dox-induced genetic damage and enhance Dox's cell viability-reducing effect in breast cancer cells.

Rosmarinic acid (RA) showed protective effect against doxorubicin-induced genotoxicity.Epigallocatechin gallate (EGCG) demonstrated pro-oxidant properties at high concentrations.EGCG and RA selectively targeted MDA-MB-231 cells.Synergistic effect was observed when EGCG or RA was administered together with Dox on breast cancer cells.

Environmental Chemical-Induced Reactive Oxygen Species Generation and Immunotoxicity: A Comprehensive Review.

Significance: Reactive oxygen species (ROS), the reactive oxygen-carrying chemicals moieties, act as pleiotropic signal transducers to maintain various biological processes/functions, including immune response. Increased ROS production leads to oxidative stress, which is implicated in xenobiotic-induced adverse effects. Understanding the immunoregulatory mechanisms and immunotoxicity is of interest to developing therapeutics against xenobiotic insults. Recent Advances: While developmental studies have established the essential roles of ROS in the establishment and proper functioning of the immune system, toxicological studies have demonstrated high ROS generation as one of the potential mechanisms of immunotoxicity induced by environmental chemicals, including heavy metals, pesticides, aromatic hydrocarbons (benzene and derivatives), plastics, and nanoparticles. Mitochondrial electron transport and various signaling components, including NADH oxidase, toll-like receptors (TLRs), NF-κB, JNK, NRF2, p53, and STAT3, are involved in xenobiotic-induced ROS generation and immunotoxicity. Critical Issues: With many studies demonstrating the role of ROS and oxidative stress in xenobiotic-induced immunotoxicity, rigorous and orthogonal approaches are needed to achieve in-depth and precise understanding. The association of xenobiotic-induced immunotoxicity with disease susceptibility and progression needs more data acquisition. Furthermore, the general methodology needs to be possibly replaced with high-throughput precise techniques. Future Directions: The progression of xenobiotic-induced immunotoxicity into disease manifestation is not well documented. Immunotoxicological studies about the combination of xenobiotics, age-related sensitivity, and their involvement in human disease incidence and pathogenesis are warranted. Antioxid. Redox Signal. 40, 691-714.

Endocrine Disruptors: Focus on the Adrenal Cortex.

Endocrine-disrupting chemicals (EDCs) are exogenous substances known to interfere with endocrine homeostasis and promote adverse health outcomes. Their impact on the adrenal cortex, corticosteroids and their physiological role in the organism has not yet been sufficiently elucidated. In this review, we collect experimental and epidemiological evidence on adrenal disruption by relevant endocrine disruptors. In vitro data suggest significant alterations of gene expression, cell signalling, steroid production, steroid distribution, and action. Additionally, morphological studies revealed disturbances in tissue organization and development, local inflammation, and zone-specific hyperplasia. Finally, endocrine circuits, such as the hypothalamic-pituitary-adrenal axis, might be affected by EDCs. Many questions regarding the detection of steroidogenesis disruption and the effects of combined toxicity remain unanswered. Not only due to the diverse mode of action of adrenal steroids and their implication in many common diseases, there is no doubt that further research on endocrine disruption of the adrenocortical system is needed.

Differential Modulation of Markers of Oxidative Stress and DNA Damage in Arterial Hypertension.

Patients with arterial hypertension have an increased risk of developing tumors, particularly renal cell carcinoma. Arterial hypertension is linked to DNA damage via the generation of oxidative stress, in which an upregulated renin-angiotensin-aldosterone system plays a crucial role. The current study investigated surrogates of oxidative stress and DNA damage in a group of hypertensive patients (HypAll, n = 64) and subgroups of well (HypWell, n = 36) and poorly (HypPoor, n = 28) controlled hypertensive patients compared to healthy controls (n = 8). In addition, a longitudinal analysis was performed with some of the hypertensive patients. Markers for oxidative stress in plasma (SHp, D-ROM, and 3-nitrotyrosine) and urine (8-oxodG, 15-F2t-isoprostane, and malondialdehyde) and markers for DNA damage in lymphocytes (γ-H2AX and micronuclei) were measured. In HypAll, all markers of oxidative stress except malondialdehyde were increased compared to the controls. After adjustment for age, this association was maintained for the protein stress markers SHp and 3-nitrotyrosine. With regard to the markers for DNA damage, there was no difference between HypAll and the controls. Further, no significant differences became apparent in the levels of both oxidative stress and DNA damage between HypWell and HypPoor. Finally, a positive correlation between the development of blood pressure and oxidative stress was observed in the longitudinal study based on the changes in D-ROM and systolic blood pressure. In conclusion, we found increased oxidative stress in extensively treated hypertensive patients correlating with the level of blood-pressure control but no association with DNA damage.

Datasets Construction and Development of QSAR Models for Predicting Micronucleus In Vitro and In Vivo Assay Outcomes.

In silico (quantitative) structure-activity relationship modeling is an approach that provides a fast and cost-effective alternative to assess the genotoxic potential of chemicals. However, one of the limiting factors for model development is the availability of consolidated experimental datasets. In the present study, we collected experimental data on micronuclei in vitro and in vivo, utilizing databases and conducting a PubMed search, aided by text mining using the BioBERT large language model. Chemotype enrichment analysis on the updated datasets was performed to identify enriched substructures. Additionally, chemotypes common for both endpoints were found. Five machine learning models in combination with molecular descriptors, twelve fingerprints and two data balancing techniques were applied to construct individual models. The best-performing individual models were selected for the ensemble construction. The curated final dataset consists of 981 chemicals for micronuclei in vitro and 1309 for mouse micronuclei in vivo, respectively. Out of 18 chemotypes enriched in micronuclei in vitro, only 7 were found to be relevant for in vivo prediction. The ensemble model exhibited high accuracy and sensitivity when applied to an external test set of in vitro data. A good balanced predictive performance was also achieved for the micronucleus in vivo endpoint.

DNA strand break levels in cryopreserved mononuclear blood cell lines measured by the alkaline comet assay: results from the hCOMET ring trial.

The comet assay is widely used in biomonitoring studies for the analysis of DNA damage in leukocytes and peripheral blood mononuclear cells. Rather than processing blood samples directly, it can be desirable to cryopreserve whole blood or isolated cells for later analysis by the comet assay. However, this creates concern about artificial accumulation of DNA damage during cryopreservation. In this study, 10 laboratories used standardized cryopreservation and thawing procedures of monocytic (THP-1) or lymphocytic (TK6) cells. Samples were cryopreserved in small aliquots in 50% foetal bovine serum, 40% cell culture medium, and 10% dimethyl sulphoxide. Subsequently, cryopreserved samples were analysed by the standard comet assay on three occasions over a 3-year period. Levels of DNA strand breaks in THP-1 cells were increased (four laboratories), unaltered (four laboratories), or decreased (two laboratories) by long-term storage. Pooled analysis indicates only a modest positive association between storage time and levels of DNA strand breaks in THP-1 cells (0.37% Tail DNA per year, 95% confidence interval: -0.05, 0.78). In contrast, DNA strand break levels were not increased by cryopreservation in TK6 cells. There was inter-laboratory variation in levels of DNA strand breaks in THP-1 cells (SD = 3.7% Tail DNA) and TK6 reference sample cells (SD = 9.4% Tail DNA), whereas the intra-laboratory residual variation was substantially smaller (i.e. SD = 0.4%-2.2% Tail DNA in laboratories with the smallest and largest variation). In conclusion, the study shows that accumulation of DNA strand breaks in cryopreserved mononuclear blood cell lines is not a matter of concern.

Long-term cryopreservation of potassium bromate positive assay controls for measurement of oxidatively damaged DNA by the Fpg-modified comet assay: results from the hCOMET ring trial.

The formamidopyrimidine DNA glycosylase (Fpg)-modified comet assay is widely used for the measurement of oxidatively generated damage to DNA. However, there has not been a recommended long-term positive control for this version of the comet assay. We have investigated potassium bromate as a positive control for the Fpg-modified comet assay because it generates many Fpg-sensitive sites with a little concurrent generation of DNA strand breaks. Eight laboratories used the same procedure for the treatment of monocytic THP-1 cells with potassium bromate (0, 0.5, 1.5, and 4.5 mM) and subsequent cryopreservation in a freezing medium consisting of 50% foetal bovine serum, 40% RPMI-1640 medium, and 10% dimethyl sulphoxide. The samples were analysed by the Fpg-modified comet assay three times over a 3-year period. All laboratories obtained a positive concentration-response relationship in cryopreserved samples (linear regression coefficients ranging from 0.79 to 0.99). However, there was a wide difference in the levels of Fpg-sensitive sites between the laboratory with the lowest (4.2% Tail DNA) and highest (74% Tail DNA) values in THP-1 cells after exposure to 4.5 mM KBrO3. In an attempt to assess sources of inter-laboratory variation in Fpg-sensitive sites, comet images from one experiment in each laboratory were forwarded to a central laboratory for visual scoring. There was high consistency between measurements of %Tail DNA values in each laboratory and the visual score of the same comets done in the central laboratory (r = 0.98, P < 0.001, linear regression). In conclusion, the results show that potassium bromate is a suitable positive comet assay control.

Visual comet scoring revisited: a guide to scoring comet assay slides and obtaining reliable results.

Measurement of DNA migration in the comet assay can be done by image analysis or visual scoring. The latter accounts for 20%-25% of the published comet assay results. Here we assess the intra- and inter-investigator variability in visual scoring of comets. We include three training sets of comet images, which can be used as reference for researchers who wish to use visual scoring of comets. Investigators in 11 different laboratories scored the comet images using a five-class scoring system. There is inter-investigator variation in the three training sets of comets (i.e. coefficient of variation (CV) = 9.7%, 19.8%, and 15.2% in training sets I-III, respectively). However, there is also a positive correlation of inter-investigator scoring in the three training sets (r = 0.60). Overall, 36% of the variation is attributed to inter-investigator variation and 64% stems from intra-investigator variation in scoring between comets (i.e. the comets in training sets I-III look slightly different and this gives rise to heterogeneity in scoring). Intra-investigator variation in scoring was also assessed by repeated analysis of the training sets by the same investigator. There was larger variation when the training sets were scored over a period of six months (CV = 5.9%-9.6%) as compared to 1 week (CV = 1.3%-6.1%). A subsequent study revealed a high inter-investigator variation when premade slides, prepared in a central laboratory, were stained and scored by investigators in different laboratories (CV = 105% and 18%-20% in premade slides with comets from unexposed and hydrogen peroxide-exposed cells, respectively). The results indicate that further standardization of visual scoring is desirable. Nevertheless, the analysis demonstrates that visual scoring is a reliable way of analysing DNA migration in comets.

Inter-laboratory variation in measurement of DNA damage by the alkaline comet assay in the hCOMET ring trial.

The comet assay is a simple and versatile method for measurement of DNA damage in eukaryotic cells. More specifically, the assay detects DNA migration from agarose gel-embedded nucleoids, which depends on assay conditions and the level of DNA damage. Certain steps in the comet assay procedure have substantial impact on the magnitude of DNA migration (e.g. electric potential and time of electrophoresis). Inter-laboratory variation in DNA migration levels occurs because there is no agreement on optimal assay conditions or suitable assay controls. The purpose of the hCOMET ring trial was to test potassium bromate (KBrO3) as a positive control for the formamidopyrimidine DNA glycosylase (Fpg)-modified comet assay. To this end, participating laboratories used semi-standardized protocols for cell culture (i.e. cell culture, KBrO3 exposure, and cryopreservation of cells) and comet assay procedures, whereas the data acquisition was not standardized (i.e. staining of comets and image analysis). Segregation of the total variation into partial standard deviation (SD) in % Tail DNA units indicates the importance of cell culture procedures (SD = 10.9), comet assay procedures (SD = 12.3), staining (SD = 7.9) and image analysis (SD = 0.5) on the overall inter-laboratory variation of DNA migration (SD = 18.2). Future studies should assess sources of variation in each of these steps. On the positive side, the hCOMET ring trial demonstrates that KBrO3 is a robust positive control for the Fpg-modified comet assay. In conclusion, the hCOMET ring trial has demonstrated a high reproducibility of detecting genotoxic effects by the comet assay, but inter-laboratory variation of DNA migration levels is a concern.

Maternal Smoking in the First Trimester and its Consequence on the Early Placenta.

Smoking during pregnancy increases the risk of adverse pregnancy outcomes, such as stillbirth and fetal growth restriction. This suggests impaired placental function and restricted nutrient and oxygen supply. Studies investigating placental tissue at the end of pregnancy have revealed increased DNA damage as a potential underlying cause, which is driven by various toxic smoke ingredients and oxidative stress induced by reactive oxygen species (ROS). However, in the first trimester, the placenta develops and differentiates, and many pregnancy pathologies associated with reduced placental function originate here. Therefore, we determined DNA damage in a cohort of first-trimester placental samples of verified smokers and nonsmokers. In fact, we observed an 80% increase in DNA breaks (P < .001) and shortened telomeres by 5.8% (P = .04) in placentas exposed to maternal smoking. Surprisingly, there was a decrease in ROS-mediated DNA damage, ie, 8-oxo-guanidine modifications, in placentas of the smoking group (-41%; P = .021), which paralleled the reduced expression of base excision DNA repair machinery, which restores oxidative DNA damage. Moreover, we observed that the increase in placental oxidant defense machinery expression, which usually occurs at the end of the first trimester in a healthy pregnancy as a result of the full onset of uteroplacental blood flow, was absent in the smoking group. Therefore, in early pregnancy, maternal smoking causes placental DNA damage, contributing to placental malfunction and increased risk of stillbirth and fetal growth restriction in pregnant women. Additionally, reduced ROS-mediated DNA damage along with no increase in antioxidant enzymes suggests a delay in the establishment of physiological uteroplacental blood flow at the end of the first trimester, which may further add to a disturbed placental development and function as a result of smoking in pregnancy.

Fate of micronuclei and micronucleated cells after treatment of HeLa cells with different genotoxic agents.

Although micronuclei are well-known biomarkers of genotoxic damage, the biological consequences of micronucleus induction are only poorly understood. To further elucidate these consequences, HeLa cells stably expressing histone 2B coupled with green fluorescent protein were used for long-term live cell imaging to investigate the fate of micronuclei and micronucleated cells after treatment of cells with various genotoxic agents (doxorubicin (20, 30 and nM), tert-butyl hydroperoxide (tBHP, 50, 100 and 150 µM), radiation (0.5, 1 and 2 Gy), methyl methanesulfonate (MMS, 20, 25 and 30 µg/ml) and vinblastine (1, 2 and 3 nM)). Most micronuclei persist for multiple cell cycles or reincorporate while micronucleated cells were more prone to cell death, senescence and fatal mitotic errors compared to non-micronucleated cells, which is consistent with previous studies using etoposide. No clear substance-related effects on the fate of micronuclei and micronucleated cells were observed. To further investigate the fate of micronuclei, extrusion of micronuclei was studied with treatments reported as inducing the extrusion of micronuclei. Since extrusion was not observed in HeLa cells, the relevance of extrusion of micronuclei remains unclear. In addition, degradation of micronuclei was analysed via immunostaining of γH2AX, which demonstrated a high level of DNA damage in micronuclei compared to the main nuclei. Furthermore, transduction with two reporter genes (LC3B-dsRed and LaminB1-dsRed) was conducted followed by long-term live cell imaging. While autophagy marker LC3B was not associated with micronuclei, Lamin B1 was found in approximately 50% of all micronuclei. While degradation of micronuclei was not observed to be a frequent fate of micronuclei, the results show impaired stability of DNA and micronuclear envelope indicating rupture of micronuclei as a pre-step to chromothripsis.

Genotoxicity of pyrrolizidine alkaloids in metabolically inactive human cervical cancer HeLa cells co-cultured with human hepatoma HepG2 cells.

Pyrrolizidine alkaloids (PAs) are secondary plant metabolites, which can be found as contaminant in various foods and herbal products. Several PAs can cause hepatotoxicity and liver cancer via damaging hepatic sinusoidal endothelial cells (HSECs) after hepatic metabolization. HSECs themselves do not express the required metabolic enzymes for activation of PAs. Here we applied a co-culture model to mimic the in vivo hepatic environment and to study PA-induced effects on not metabolically active neighbour cells. In this co-culture model, bioactivation of PA was enabled by metabolically capable human hepatoma cells HepG2, which excrete the toxic and mutagenic pyrrole metabolites. The human cervical epithelial HeLa cells tagged with H2B-GFP were utilized as non-metabolically active neighbours because they can be identified easily based on their green fluorescence in the co-culture. The PAs europine, riddelliine and lasiocarpine induced micronuclei in HepG2 cells, and in HeLa H2B-GFP cells co-cultured with HepG2 cells, but not in HeLa H2B-GFP cells cultured alone. Metabolic inhibition of cytochrome P450 enzymes with ketoconazole abrogated micronucleus formation. The efflux transporter inhibitors verapamil and benzbromarone reduced micronucleus formation in the co-culture model. Furthermore, mitotic disturbances as an additional genotoxic mechanism of action were observed in HepG2 cells and in HeLa H2B-GFP cells co-cultured with HepG2 cells, but not in HeLa H2B-GFP cells cultured alone. Overall, we were able to show that PAs were activated by HepG2 cells and the metabolites induced genomic damage in co-cultured HeLa cells.

Incidence and Risk Factors of Acute Leukemias in Armenia: A Population-Based Study.

OBJECTIVE: Leukemia represents a serious public health concern as the incidence is increasing worldwide. In this study we aimed to describe the epidemiological profile of acute lymphoblastic (ALL) and myeloid (AML) leukemia, identify disease clusters and find association with possible risk factors. METHODS: Data on leukemia cases were provided by the National Institute of Health of the Republic of Armenia for the period of 2012-2018. Age-standardized incidence rate was calculated using Segi World Population. SaTScan purely spatial analysis was applied to find leukemia clusters. To find association between leukemia and agricultural and mining activities and demographic data Poisson regression model was used. RESULTS: During the studied period 259 new cases of ALL and 478 AML were recorded. The age-standardized incidence rate was 1.5 and 1.9 per 100,000 inhabitants with male to female ratio of 0.97 and 1.1 for ALL and AML, respectively. No significant changes in ALL or AML incidence trends were found. For ALL significant cluster encompassing Shirak, Lori, Tavush and Armavir provinces of Armenia was identified, while Kotayk and Ararat was provinces with the highest incidence of AML. We found significant positive association of ALL with crop density, while no elevated risk estimates were found between AML and exposure variables. CONCLUSION: Altogether, our results suggested that acute leukemias incidence in Armenia follows the pattern described for developing countries.

Improving VAE based molecular representations for compound property prediction.

Collecting labeled data for many important tasks in chemoinformatics is time consuming and requires expensive experiments. In recent years, machine learning has been used to learn rich representations of molecules using large scale unlabeled molecular datasets and transfer the knowledge to solve the more challenging tasks with limited datasets. Variational autoencoders are one of the tools that have been proposed to perform the transfer for both chemical property prediction and molecular generation tasks. In this work we propose a simple method to improve chemical property prediction performance of machine learning models by incorporating additional information on correlated molecular descriptors in the representations learned by variational autoencoders. We verify the method on three property prediction tasks. We explore the impact of the number of incorporated descriptors, correlation between the descriptors and the target properties, sizes of the datasets etc. Finally, we show the relation between the performance of property prediction models and the distance between property prediction dataset and the larger unlabeled dataset in the representation space.