Drug-induced phospholipidosis (PLD) is an adaptive histologic alteration that is seen with various marketed drugs and often encountered during drug development. Various in silico and in vitro cell-based methods have been developed to predict the PLD-inducing potential of compounds. These methods rely on the inherent physicochemical properties of the molecule and, as such, tend to overpredict compounds as PLD inducers. Recognizing that the distribution of compounds into tissues or tissue accumulation is likely a key factor in PLD induction, in addition to key physicochemical properties, we developed a model to predict PLD in vivo using the measures of basicity (pK(a)), lipophilicity (ClogP), and volume of distribution (V(d)). Using sets of PLD inducers and noninducers, we demonstrate improved concordance with this method. Furthermore, we propose a screening paradigm that includes a combination of various methods to predict the in vivo PLD-inducing potential of compounds, which may be especially useful in lead identification and optimization processes in drug discovery.

Lysosomes are membrane-bound subcellular organelles involved in the degradation of macromolecules and pathogens in diverse processes, including endocytosis, phagocytosis, and autophagy. A red fluorescent probe was developed that is selectively sequestered in acidic organelles. U20s cells pretreated with 64 muM chloroquine for as little as 5 h show a dramatic increase in lysosome-like vesicle number and volume. The probe can be employed for highlighting lysosome-like organelles under conditions wherein cells produce vacuoles that contain most of the degradative enzymes of the lysosome but are not as acidic as the parent organelle. Using a conventional fluorescence microplate reader, the half-maximal effective concentration (EC50) of chloroquine was estimated. The high Z' score obtained using the assay demonstrated excellent signal-to-noise ratios. The fluorescence microplate assay was successfully employed to screen a small-molecule compound library for agents that increase lysosomal volume and number. One potential application of the new assay is in the toxicology portion of preclinical drug safety assessment (ADME-Tox) workflows, using in vitro cell culture models to aid in the drug development process.(Journal of Biomolecular Screening XXXX:xx-xx).

The liver is the most important target for toxicity caused by drugs. This vulnerability is a consequence of the functional features of the liver and their role in the metabolic elimination of most drugs. Therefore, evaluation of potential hepatotoxicity represents a critical step in the development of new drugs. The liver is very active in metabolising foreign compounds and, although biotransformation reactions generally parallel detoxification processes, the formation of reactive metabolites is relatively frequent. Thus, drug-induced hepatotoxicity can be due to the administered compound itself or to metabolites formed by hepatic metabolism. The most important systems to study hepatotoxicity and metabolic activity in vitro are liver slices, isolated liver cells in suspensions or in primary cultures including co-culture methods and special 3D techniques, various subcellular fractions and hepatic cell lines. These models can be used for cytotoxicity and genotoxicity screening, and also to identify the mechanisms involved in drug-induced hepatotoxicity. Assessment of current cytotoxicity and hepatic-specific biochemical effects are limited by the inability to measure a wide spectrum of potential mechanistic changes involved in the drug-induced toxic injury. A convenient selection of end-points allows a multiparametric evaluation of drug toxicity. In this regard, omic (cytomic, metabonomic, proteomic and toxicogemic) approaches help defining patterns of hepatotoxicity for early identification of potential adverse effects of the drug to the liver. The development of robust in vitro-based multiparametric screening assays covering a wider spectrum of key effects will heighten the predictive capacity for human hepatotoxicity, and accelerate the drug development process.

Macrophage foam cells formed during uptake of atherogenic lipoproteins are a hallmark of atherosclerotic lesion development. In this study, human macrophages were incubated with two prototypic atherogenic LDL modifications enzymatically degraded LDL (E-LDL) and oxidized LDL (Ox-LDL) prepared from the same donor LDL. To detect differences in macrophage lipid storage, fluorescent high-content imaging was used. Lipid droplets were stained using Bodipy 493/503, and the fluorescent phospholipid probe NBD-PE was used to detect endolysosomal phospholipidosis in high-content imaging assays. The phospholipidosis assay was validated using phospholipidosis-inducing cationic amphiphilic drugs. In addition, neutral lipids and phospholipidosis were determined using LipidTOX. Images of 96-well cell culture microtiter plates were captured with multichannel laser-based high-content confocal microscopy, and subsequently cell- and well-based data were analyzed. E-LDL-loaded macrophages show increased intensity of Bodipy 493/503 and LipidTOX-Green neutral lipid droplet staining and a greater mean area and number of lipid droplets per cell compared to Ox-LDL-loaded and M-CSF-differentiated control macrophages. In contrast, Ox-LDL-loaded macrophages show increased intensity of NBD-PE and LipidTOX-Red detectable phospholipidosis in the endolysosomal compartment compared to E-LDL-loaded and M-CSF-differentiated macrophages. Treatment with the peroxisome proliferator-activated receptor-gamma agonist pioglitazone leads to lipid droplet induction depending on the lipid loading state of the macrophages. These results indicate that E-LDL preferentially induces lipid droplets, while Ox-LDL provokes endolysosomal phospholipidosis in human macrophages representing two different lipid storage principles. Therefore, fluorescent high-content imaging is a useful tool to discriminate between and quantify lipid storage compartments in macrophages also in response to drugs affecting cellular lipid metabolism.(c) 2009 International Society for Advancement of Cytometry.

Drug candidates under development by industry frequently show phospholipidosis as a side-effect in pre-clinical toxicity studies. This study sets up a cell-based assay for drug-induced phospholipidosis (PLD) and its performance was evaluated based on the in vivo PLD potential of compounds in 2-week toxicity studies in rats. When HepG2 cells were exposed simultaneously to PLD-inducing chemicals and a phospholipid having a fluorophore, an accumulation of phospholipids was detected as an increasing fluorescent intensity. Amiodarone, amitriptyline, fluoxetine, AY-9944, and perhexiline, which are common PLD-inducing chemicals, increased the fluorescent intensity, but acetaminophen, ampicillin, cimetidine, famotidine, or valproic acid, which are non-PLD-inducing chemicals, did not. The fluorescent intensity showed concordance with the pathological observations of phospholipid lamellar bodies in the cells. Then to confirm the predictive performance of the in vitro PLD assay, the 32 proprietary compounds characterized in 2-week toxicity studies in rats were evaluated with this in vitro assay. Because this in vitro assay was vulnerable to cytotoxicity, the innate PLD potential was calculated for each compound. A statistically significant increase in the in vitro PLD potential was seen for the compounds having in vivo PLD-inducing potential in the rat toxicity studies. The results suggest that the in vitro PLD potential could be appropriate to detect the appearance of PLD as a side effect in pre-clinical toxicity studies in rats.

Background and purpose: Tissue deposits of the anti-arrhythmic drug amiodarone are a major source of side effects (skin discoloration, etc.). We addressed the mechanism of the concentration of amiodarone in cells, and characterized the resulting vacuolar cytopathology and its evolution towards macroautophagy. Experimental approach: Sequestration of amiodarone in human cells (macrophages, smooth muscle cells, HEK 293a cells) was evaluated using its violet fluorescence and cytopathology using GFP-conjugated subcellular markers. Autophagic signalling was probed by immunoblotting for the effector protein LC3. A patient biopsy of amiodarone-induced blue-gray skin discoloration was investigated for the presence of macroautophagy (immunofluorescence for LC3). Key results: Most of the amiodarone (1-20 microM, 4-24 h) captured by cultured cells (macrophages were most avid) was present in enlarged vacuoles. The specific vacuolar ATPase (V-ATPase) inhibitors, bafilomycin A1 or FR167356, prevented vacuolization and drug uptake. Vacuoles in HEK 293a cells were positive for markers of late endosomes and lysosomes (GFP-Rab7,-CD63) and for an effector of macroautophagy, GFP-LC3. The vacuoles accumulated endogenous LC3 and filled with lipids (Nile red staining) following longer amiodarone treatments (>/=24 h). The electrophoretic mobility of both GFP-LC3 and endogenous LC3 changed, showing activation in response to amiodarone. Paraffin tissue sections of the pigmented skin exhibited granular LC3 accumulation in superficial dermis macrophages. Conclusion and implications: Vacuolar sequestration of amiodarone occurs at concentrations close to therapeutic levels, is mediated by V-ATPase and evolves towards persistent macroautophagy and phospholipidosis. This cytopathology is not cell type specific, but tissue macrophages appear to be particularly susceptible.

Phospholipid bilayers represent a complex, anisotropic environment fundamentally different from bulk oil or octanol, for instance. Even "simple" drug association to phospholipid bilayers can only be fully understood if the slab-of-hydrocarbon approach is abandoned and the complex, anisotropic properties of lipid bilayers reflecting the chemical structures and organization of the constituent phospholipids are considered. The interactions of drugs with phospholipids are important in various processes, such as drug absorption, tissue distribution, and subcellular distribution. In addition, drug-lipid interactions may lead to changes in lipid-dependent protein activities, and further, to functional and morphological changes in cells, a prominent example being the phospholipidosis (PLD) induced by cationic amphiphilic drugs. Herein we briefly review drug-lipid interactions in general and the significance of these interactions in PLD in particular. We also focus on a potential causal connection between drug-induced PLD and steatohepatitis, which is induced by some cationic amphiphilic drugs.

The influence of combinatorial chemistry and high-throughput screening (HTS) technologies in the pharmaceutical industry during the last 10 years has been enormous. However, the attrition rate of drugs in the clinic due to toxicity during this period still remained 40-50%. The need for reduced toxicity failure led to the development of early toxicity screening assays. This chapter describes the state of the art for assays in the area of genotoxicity, cytotoxicity, carcinogenicity, induction of specific enzymes from phase I and II metabolism, competition assays for enzymes of phase I and II metabolism, embryotoxicity as well as endocrine disruption and reprotoxicity. With respect to genotoxicity, the full Ames, Ames II, Vitotox, GreenScreen GC, RadarScreen, and non-genotoxic carcinogenicity assays are discussed. For cytotoxicity, cellular proliferation, calcein uptake, oxygen consumption, mitochondrial activity, radical formation, glutathione depletion as well as apoptosis are described. For high-content screening (HCS), the possibilities for analysis of cytotoxicity, micronuclei, centrosome formation and phospholipidosis are examined. For embryotoxicity, endocrine disruption and reprotoxicity alternative assays are reviewed for fast track analysis by means of nuclear receptors and membrane receptors. Moreover, solutions for analyzing enzyme induction by activation of nuclear receptors, like AhR, CAR, PXR, PPAR, FXR, LXR, TR and RAR are given.

This review summarizes the current knowledge of endolysosomal and cytoplasmic lipid storage in macrophages induced by oxidized LDL (Ox-LDL), enzymatically degraded LDL (E-LDL) and other atherogenic lipoprotein modifications, and their relation to the adapter protein 3 (AP-3) dependent ABCA1 and ABCG1 cellular lipid efflux pathways. We compare endolysosomal lipid storage caused either through drug induced phospholipidosis, inheritable endolysosomal and cytosolic lipid storage disorders and Ox-LDL or E-LDL induced phagosomal uptake and cytosolic lipid droplet storage in macrophages. Ox-LDL is resistant to rapid endolysosomal hydrolysis and is trapped within the endolysosomal compartment generating lamellar bodies which resemble the characteristics of phospholipidosis. Various inherited lysosomal storage diseases including sphingolipidosis, glycosphingolipidosis and cholesterylester storage diseases also present a phospholipidosis phenotype. In contrast E-LDL resembling coreless unesterified cholesterol enriched LDL-particles, with a multilamellar, liposome-like structure, lead to rapid phagosomal degradation and cytosolic lipid droplet accumulation. As a consequence the uptake of E-LDL through type I and type II phagocytosis leads to increased lipid droplet formation and moderate upregulation of ABCA1 and ABCG1 while uptake of Ox-LDL leads to a rapid expansion of the lysosomal compartment and a pronounced upregulation of the ABCA1/ABCG1/AP-3 lipid efflux pathway.

A large number of cationic amphiphilic drugs (CADs) are known to cause phospholipidosis (PLD) in vivo. In the present study, we have built upon our previous findings to further qualify the use of a fluorescently labeled phospholipid-based cell-culture assay to detect PLD-inducing drugs. In this paper, we demonstrate that 12 PLD-negative compounds and 11 drugs known to cause PLD in vivo are all correctly identified by using this assay. Interestingly, we found that in cells treated with certain CADs, the fluorescent phospholipid was sequestered in a very specific punctate pattern, which overlapped strongly with the staining pattern seen with a lysosomal marker protein. Our data also show that false positives can be generated with the fluorescence assay when compounds are used at concentrations that cause a >30% decrease in cell number in this assay. Confocal microscopy demonstrated that the staining pattern of fluorescent phospholipids in these cases may be differentiated from those of true positives by the fact that diffuse, rather than punctuate, fluorescence is observed. These studies confirm and expand our previous results showing that the fluorescent phospholipid assay is a highly sensitive, specific tool for detecting PLD-inducing drugs, if care is taken to rule out cytotoxicity-related artifact.

Chromatin modifications are now widely accepted as being essential steps involved in activation, repression and poising of the expression of large number of genes within the genome. Not only does understanding the role of such changes provide an opportunity to elucidate mechanisms controlling gene expression but in parallel offers the ability to develop novel indicative and predictive biomarkers of disease and toxicity. In the current study we have applied the chromatin immunoprecipation assay to investigate putative changes in the chromatin environment associated with the Kidney Injury Molecule 1 (Kim1) gene upon exposure of rats to the nephrotoxicant, gentamicin. Chromatin was isolated from the kidneys of both control and gentamicin-treated animals and interrogated using specific antibodies recognizing two modifications of histone H3, acetylation of lysine 9 and tri-methylation of lysine 4, along with RNA polymerase II. Enriched chromatin fractions were analyzed by quantitative PCR using tiled primer pairs covering 4 kb of the Kim1 gene (spanning -2kb to +2kb, relative to the transcription start site). The results demonstrate a substantial increase in the presence of RNA polymerase II and both histone modifications following gentamicin treatment in regions downstream but not upstream of the transcriptional start site of the Kim1 gene. These changes were associated with a marked increase in mRNA coding for the Kim1 protein. Together these data suggest, for the first time, that the Kim1 gene is regulated in an epigenetic fashion under conditions of nephrotoxicity.

Treatment with drugs from multiple classes induces vascular injury with medial necrosis, hemorrhage, endothelial damage, and inflammation. Previous research has suggested early events might be occurring well in advance of the full lesions that appear forty-eight to seventy-two hours after dosing with SCH 351591, a PDE IV inhibitor. This study was performed to study early events in detail. Rats were dosed with 20 mg/kg of drug by gavage and sacrificed at times between fifteen and 240 minutes after dosing. Tissues were collected for histopathological analysis and gene expression studies. Serum was collected for biomarker analysis. The data from biomarker analysis showed a three-part response with an early phase that was maximal at fifteen to thirty minutes, a second phase from forty-five to 180 minutes, and the third phase that was starting to rise at four hours. The first phase included increases in lymphocytes, serum histamine, and serum nitrite. The second phase shows continued elevation of serum nitrite. The third phase was marked by an increase in serum GRO/CINC-1. At fifteen minutes, histopathology showed activation of mast cells, but not degranulation. Increases in endothelial activation and perivascular inflammatory cells were first apparent at thirty minutes and increased through 240 minutes.

Carcinogenicity of chemicals can currently only be evaluated in 2-year rodent bioassays. Therefore, the development of early biomarkers for carcinogenesis would result in substantial savings in time and expense. The current study investigates whether early changes in gene expression may be developed as markers for cancer. Animals were treated for 1 or 5 days with either non-genotoxic carcinogens or non-carcinogens and gene expressionwas analyzed by quantitative PCR (qPCR).We tested two gene signatures previously reported to detect non-genotoxic carcinogens. Using one gene signature itwas confirmed that 3/3 nongenotoxic carcinogens and 2/2 non-carcinogens are correctly identified with data from1 or 5 days of dosing. In contrast an alternative signature correctly identified 0/3 and 2/3 nongenotoxic carcinogens at 1 and 5 days of treatment, respectively and 2/2 non-carcinogens at both time-points. Additionally, we evaluated a novel panel of putative biomarker genes, fromthe literature,many of whichhave roles in cell growthand division, includingmyc, cdc2 and mcm6. These genes were significantly induced by non-genotoxic carcinogens and not by non-carcinogens. Using the average fold-induction across this panel, 2/3 non-genotoxic carcinogens were detected on both day 1 and day 5. These data support the idea that acute changes in gene expression may provide biomarkers for non-genotoxic carcinogenesis but also highlight interesting differences in the sensitivities of distinct gene signatures.

Non-genotoxic carcinogenicity of chemicals is currently routinely evaluated in 2-year rodent bioassays. Therefore, the development of early biomarkers for non-genotoxic carcinogenesis would result in substantial savings in time and expense. The current study investigates whether early changes in gene expression may be developed as markers for cancer. Animals were treated for 1 or 5 days with either non-genotoxic carcinogens (NGTCs) or non-carcinogens and gene expression was analyzed by quantitative PCR (qPCR). We tested two gene signatures previously reported to detect non-genotoxic carcinogens. Using one gene signature it was confirmed that 3/3 non-genotoxic carcinogens and 2/2 non-carcinogens are correctly identified with data from 1 or 5 days of dosing. In contrast an alternative signature correctly identified 0/3 and 2/3 non-genotoxic carcinogens at 1 and 5 days of treatment, respectively and 2/2 non-carcinogens at both time-points. Additionally, we evaluated a novel panel of putative biomarker genes, from the literature, many of which have roles in cell growth and division, including myc, cdc2 and mcm6. These genes were significantly induced by non-genotoxic carcinogens and not by non-carcinogens. Using the average fold-induction across this panel, 2/3 non-genotoxic carcinogens were detected at both 1 and 5 days. These data support the idea that acute changes in gene expression may provide biomarkers for non-genotoxic carcinogenesis but also highlight interesting differences in the sensitivities of distinct gene signatures.

Drug-induced renal injury is a common finding in the early preclinical phase of drug development. But the specific genes responding to renal injury remain poorly defined. Identification of drug-induced gene changes is critical to provide insights into molecular mechanisms and detection of renal damage. To identify genes associated with the development of drug-induced nephrotoxicity, a literature survey was conducted and a panel of 48 genes was selected based on gene expression changes in multiple published studies. Male Sprague-Dawley rats were dosed daily for 1, 3 or 5 days to the known nephrotoxicants gentamicin, bacitracin, vancomycin and cisplatin, or the known hepatotoxicants ketoconazole, 1-naphthyl isothiocyanate and 4,4-diaminodiphenylmethane. Histopathological evaluation and clinical chemistry revealed renal proximal tubular necrosis in rats treated with the nephrotoxicants, but not from those treated with the hepatotoxicants. RNA was extracted from the kidney, and RT-PCR was performed to evaluate expression profiles of the selected genes. Among the genes examined, 24 genes are confirmed to be highly induced or repressed in rats treated with nephrotoxicants; further investigation identified that 5 of the 24 genes were also altered by hepatotoxicants. These data led to the identification of a set of genomic biomarker candidates whose expression in kidney is selectively regulated only by nephrotoxicants. Among those genes displaying the highest expression changes specifically in nephrotoxicant-treated rats were kidney injury molecule 1 (Kim1), lipocalin 2 (Lcn2), and osteopontin (Spp1). The establishment of such a genomic marker set offers a new tool in our ongoing quest to monitor nephrotoxicity.

The first formal qualification of safety biomarkers for regulatory decision making marks a milestone in the application of biomarkers to drug development. Following submission of drug toxicity studies and analyses of biomarker performance to the Food and Drug Administration (FDA) and European Medicines Agency (EMEA) by the Predictive Safety Testing Consortium's (PSTC) Nephrotoxicity Working Group, seven renal safety biomarkers have been qualified for limited use in nonclinical and clinical drug development to help guide safety assessments. This was a pilot process, and the experience gained will both facilitate better understanding of how the qualification process will probably evolve and clarify the minimal requirements necessary to evaluate the performance of biomarkers of organ injury within specific contexts.

Application of any new biomarker to support safety-related decisions during regulated phases of drug development requires provision of a substantial data set that critically assesses analytical and biological performance of that biomarker. Such an approach enables stakeholders from industry and regulatory bodies to objectively evaluate whether superior standards of performance have been met and whether specific claims of fit-for-purpose use are supported. It is therefore important during the biomarker evaluation process that stakeholders seek agreement on which critical experiments are needed to test that a biomarker meets specific performance claims, how new biomarker and traditional comparators will be measured and how the resulting data will be merged, analyzed and interpreted.

Drug resistant tumor "side-populations," enriched in cancer stem cells and identified by reduced accumulation of Hoechst 33342 under ABCG2-mediated efflux, may compromise therapeutic outcome. Side-population cells have predicted resistance to minor groove ligands, including the DNA topoisomerase I poison topotecan. We have used a stable Hoechst 33342-resistant murine L cell system (HoeR415) to study resistance patterns, removing the need for SP isolation before microarray analysis of gene expression and the tracking of cell cycle dynamics and cytotoxicity. The majority of HoeR415 cells displayed a side-population phenotype comparable with that of the side-population resident in the ABCG2 over-expressing A549 lung cancer cell line. Photo-crosslinking showed direct protection against minor groove ligand residence on DNA, driven by ABCG2-mediated efflux and not arising from any binding competition with endogenous polyamines. The covalent minor-groove binding properties of the drug FCE24517 (tallimustine) prevented resistance suggesting a mechanism for overcoming SP-related drug resistance. Hoechst 33342-resistant murine cells showed lower but significant crossresistance to topotecan, again attributable to enhanced ABCG2 expression, enabling cells to evade S-phase arrest. Hoechst 33342/TPT-resistant cells showed limited ancillary gene expression changes that could modify cellular capacity to cope with chronic stress including over-expression of Aldh1a1 and Mgst1, but under-expression of Plk2 and Nnt. There was no evidence to link the putative stem cell marker ALDH1A1 with any augmentation of the TPT resistance phenotype. The study has implications for the patterns of drug resistance arising during tumor repopulation and the basal resistance to minor groove-binding drugs of tumor side-populations.(c) 2009 International Society for Advancement of Cytometry.

The 3-hydroxy-3-methylglutaryl (HMG)-coenzyme A reductase inhibitor, lovastatin (lova), has been reported to both sensitize to, and protect against, the toxic effects of the antitumor anthracycline doxorubicin (dox) in cellular and in vivo systems. The mechanism by which these effects occur has not yet been determined. In the present study, lova is shown to enhance the genotoxicity of dox in the V79 cell in vitro micronucleus assay and to do so, most likely, via noncovalent interaction with DNA adjacent to sites of dox binding. These studies confirm and extend the experimental evidence strongly suggesting the importance of noncovalent drug/DNA interactions in cellular responses to genotoxic stimuli.

The genotoxicity testing battery is highly sensitive for detection of chemical carcinogens. However, it features a low specificity and provides only limited mechanistic information required for risk assessment of positive findings. This is especially important in case of positive findings in the in vitro chromosome damage assays, since chromosome damage may be also induced secondarily to cell death. An increasing body of evidence indicates that toxicogenomic analysis of cellular stress responses provides an insight into mechanisms of action of genotoxicants. To evaluate the utility of such a toxicogenomic analysis we evaluated gene expression profiles of TK6 cells treated with four model genotoxic agents using a targeted high density real-time PCR approach in a multi-laboratory project coordinated by the HESI Committee on the Application of Genomics in Mechanism-based Risk Assessment. We show that this gene profiling technology produced reproducible data across laboratories allowing us to conclude that expression analysis of a relevant gene-set is capable of distinguishing compounds that cause DNA adducts or double strand breaks from those that interfere with mitotic spindle function or that cause chromosome damage as a consequence of cytotoxicity. Furthermore, our data suggest that the gene expression profiles at early time points are most likely to provide information relevant to mechanisms of genotoxic damage and that larger gene expression arrays will likely provide richer information for differentiating molecular mechanisms of action of genotoxicants. Although more compounds need to be tested to identify a robust molecular signature, this study confirms the potential of toxicogenomic analysis for investigation of genotoxic mechanisms.

The skin sensitizing potential of chemicals is an important concern for public health and thus a significant endpoint in the hazard identification process. To determine skin sensitizing capacity, large research efforts focus on the development of assays which do not require animals. As such, an in vitro test has previously been developed based on the differential expression of CREM and CCR2 transcripts in CD34(+) progenitor-derived dendritic cells (CD34-DC) which allows to classify chemicals as skin (non-)sensitizing. However, skin sensitization is not an all-or-none phenomenon and up to now the assessment of relative potency can only be derived using the in vivo local lymph node assay (LLNA). In our study we analyzed the feasibility to predict the sensitizing potency, i.e. the LLNA EC3 values, of 15 skin sensitizers using in vitro data from the CD34-DC-based assay. Hereto we extended the in vitro generated gene expression dataset by an additional source of information, the concentration of the compound that causes 20% cell damage (IC20) in CD34-DC. We statistically confirmed that this IC20 is linearly independent from the gene expression changes, but that it does correlate with LLNA EC3 values. In a further analysis we applied a robust linear regression with both IC20 and expression changes of CREM and CCR2 as explanatory variables. For 13 out of 15 compounds, a high linear correlation was established between the in vitro model and the LLNA EC3 values over a range of 4 orders of magnitude, i.e. from weak to extreme sensitizers.

ABSTRACT Phospholipidosis is the excessive intralysosomal accumulation of phospholipids and is induced in humans and animals by the chronic administration of cationic amphiphilic drugs. To identify compounds that may induce phospholipidosis early in the discovery process, we have developed a predictive fluorescent cell-based assay amenable to automated high content screening using the 2-(4,4-difluoro-5-methyl-4-bora-3a,4a-diaza-s-indacene-3-dodecanoyl)-1-hexadecanoyl-sn-glycero-3-phosphocholine (ss-BODIPY C(12)-HPC) dye and primary rat hepatocytes. ss-BODIPY C(12)-HPC localized to lysosomes that accumulate phospholipids and not to lipid droplets, indicating the selectivity for phospholipid-containing granules. Accumulation of ss-BODIPY C(12)-HPC was monitored in primary rat hepatocytes plated onto 96-well plates and 24 h after exposure to increasing concentrations of 13 drugs known to induce phospholipidosis and four negative compounds. Fluorescent images were captured and analyzed using the Discovery-1 automated cellular imaging system. Eleven out of the 12 selected positive compounds and all negative compounds were properly assigned as positive and negative inducers of phospholipidosis, respectively, indicating the high degree of sensitivity and specificity of this assay. The ability of ss-BODIPY C(12)-HPC to detect and quantify phospholipidosis is similar to that of the well-established probe, N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)-dipalmitoylphosphatidylethanolamine (NBD-PE).

ABSTRACT Drug-induced phospholipidosis (PL) is a condition characterized by the accumulation of phospholipids and drug in lysosomes, and is found in a variety of tissue types. PL is frequently manifested in preclinical studies and may delay or prevent the development of pharmaceuticals. This report describes the construction of a database of PL findings in a variety of animal species and its use as a training data set for computational toxicology software. PL data and chemical structures were compiled from the published literature, existing pharmaceutical databases, and Food and Drug Administration (FDA) internal reports yielding a total of 583 compounds suitable for modeling. The database contained 190 (33%) positive drugs and 393 (77%) negative drugs, of which 39 were electron microscopy-confirmed negative compounds and 354 were classified as negatives due to the absence of positive reported data. Of the 190 positive findings, 76 were electron microscopy confirmed and 114 were considered positive based on other evidence. Quantitative structure-activity relationship (QSAR) models were constructed using two commercially available software programs, MC4PC and MDL-QSAR, and internal cross-validation (10 x 10%) experiments were performed to assess their predictive performance. Performance parameters for the MC4PC model were specificity 92%, sensitivity 50%, concordance 78%, positive predictivity 76%, and negative predictivity 78%. For MDL-QSAR, predictive performance was similar: specificity 80%, sensitivity 76%, concordance 79%, positive predictivity 65%, and negative predictivity 87%. By combining the output of the two QSAR programs, the overall predictive performance was vastly improved and sensitivity could be optimized to 81% without significant loss of specificity (79%). Many of the structural alerts and significant molecular descriptors obtained from the QSAR software were found to be associated with parts of active molecules known for their cationic amphiphilic drug (CAD) properties supporting the hypothesis that the endpoint of PL is statistically correlated with chemical structure. QSAR models can be useful tools for screening drug candidate molecules for potential PL.

Drug candidates under development by industry frequently show phospholipidosis as a side-effect in pre-clinical toxicity studies. This study sets up a cell-based assay for drug-induced phospholipidosis (PLD) and its performance was evaluated based on the in vivo PLD potential of compounds in 2-week toxicity studies in rats. When HepG2 cells were exposed simultaneously to PLD-inducing chemicals and a phospholipid having a fluorophore, an accumulation of phospholipids was detected as an increasing fluorescent intensity. Amiodarone, amitriptyline, fluoxetine, AY-9944, and perhexiline, which are common PLD-inducing chemicals, increased the fluorescent intensity, but acetaminophen, ampicillin, cimetidine, famotidine, or valproic acid, which are non-PLD-inducing chemicals, did not. The fluorescent intensity showed concordance with the pathological observations of phospholipid lamellar bodies in the cells. Then to confirm the predictive performance of the in vitro PLD assay, the 32 proprietary compounds characterized in 2-week toxicity studies in rats were evaluated with this in vitro assay. Because this in vitro assay was vulnerable to cytotoxicity, the innate PLD potential was calculated for each compound. A statistically significant increase in the in vitro PLD potential was seen for the compounds having in vivo PLD-inducing potential in the rat toxicity studies. The results suggest that the in vitro PLD potential could be appropriate to detect the appearance of PLD as a side effect in pre-clinical toxicity studies in rats.

Drugs are capable of inducing hepatic lipid accumulation. When fat accumulates, lipids are primarily stored as triglycerides which results in steatosis and provides substrates for lipid peroxidation. An in vitro multiparametric flow cytometry assay was performed in HepG2 cells by using fluorescent probes to analyze cell viability (propidium iodide, PI), lipid accumulation (BODIPY493/503), mitochondrial membrane potential (tetramethyl rhodamine methyl ester, TMRM) and reactive oxygen species generation (ROS)(2',7'-dihydrochlorofluorescein diacetate, DHCF-DA) as functional markers. All the measurements were restricted to live cells by gating the cells that excluded PI or those that exhibited the typical forward and side scatter features of live cells. The assay was qualified by analyzing a number of selected model drugs with a well documented induction of steatosis in vivo using different mechanisms as positive controls and several non steatosic compounds as negative controls. For the cytometric screening assay, the concentrations tested were up to the corresponding IC(10) value determined by the MTT assay. Among the parameters analyzed, increased BODIPY fluorescence was the most sensitive and selective marker of drug-induced steatosis. However, a more consistent predictive approach was the combination of two endpoints: lipid accumulation and ROS generation. The assay correctly identified 100% of steatosis-positive and steatosis-negative compounds, and a high steatosis risk was predicted for amiodarone, doxycycline, tetracycline and valproate treatments at therapeutic doses. The results suggest that this cell-based assay may be a useful approach to identify the potential of drug candidates to induce steatosis.

Drug-induced phospholipidosis (PLD) is characterized by the excessive accumulation of phospholipids in lysosomes. It is accompanied by intracellular retention of drug that could be associated with increased cytotoxicity. Drug-induced PLD is recognized as a significant challenge for drug development, depending on the severity of the effect it could be reversible or caused cell death. Therefore, the identification at early stages of drug discovery of the potential to induce PLD can be advantageous for selecting improved development candidates. PLD has commonly been associated with cationic amphiphilic drugs (CADs) composed by a hydrophobic ring structure and a hydrophilic side chain with a charged amine group. 4(1H)-pyridone derivatives are a family of antimalarial agents that act as potent selective inhibitors of Plasmodium falciparum mitochondrial function and according to their chemical structure might be considered to be CADs. In the present study, the potential of 4(1H)-pyridone derivatives to induce PLD in vitro and their general cytotoxicity properties were investigated. A cell-based fluorescence assay using the fluorescent phospholipid probe NBD-PE [N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)-1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine, triethylammonium salt] was established. Five PLD-inducing reference compounds and six negative reference compounds were evaluated in vitro in HepG2 cell line. The pyridones tested were ranked by using a chloroquine-equivalent scale (chloroquine constituting a well-known antimalarial drug that acts as a potent inducer of lysosomal storage of phospholipids in both cell cultures and in vivo studies). The present findings indicate that these novel chemical antimalarial compounds are not PLD inducers despite to be considered structurally as CADs. Furthermore, none of the compounds tested showed significant cytotoxicity at their maximum solubility.

A large number of cationic amphiphilic drugs (CADs) are known to cause phospholipidosis (PLD) in vivo. In the present study, we have built upon our previous findings to further qualify the use of a fluorescently labeled phospholipid-based cell-culture assay to detect PLD-inducing drugs. In this paper, we demonstrate that 12 PLD-negative compounds and 11 drugs known to cause PLD in vivo are all correctly identified by using this assay. Interestingly, we found that in cells treated with certain CADs, the fluorescent phospholipid was sequestered in a very specific punctate pattern, which overlapped strongly with the staining pattern seen with a lysosomal marker protein. Our data also show that false positives can be generated with the fluorescence assay when compounds are used at concentrations that cause a >30% decrease in cell number in this assay. Confocal microscopy demonstrated that the staining pattern of fluorescent phospholipids in these cases may be differentiated from those of true positives by the fact that diffuse, rather than punctuate, fluorescence is observed. These studies confirm and expand our previous results showing that the fluorescent phospholipid assay is a highly sensitive, specific tool for detecting PLD-inducing drugs, if care is taken to rule out cytotoxicity-related artifact.

Cationic amphiphilic drugs (CADs) cause massive intracellular accumulation of phospholipids, thereby resulting in phospholipidosis (PLD); however, the molecular mechanism underlying CAD-induced PLD remains to be resolved. Here, we found that treatment of normal rat kidney cells with CADs known to induce PLD caused redistribution of a mannose 6-phosphate/IGF-II receptor (MPR300) from the TGN to endosomes and concomitantly increased the secretion of lysosomal enzymes, resulting in a decline of intracellular lysosomal enzyme levels. These results enable the interpretation of why CADs cause excessive accumulation of undegraded substrates, including phospholipids in lysosomes, and led to the conclusion that the impaired MPR300-mediated sorting system of lysosomal enzymes reflects the general mechanism of CAD-induced PLD. In addition, our findings suggest that the measurement of lysosomal enzyme activity secreted into culture medium is useful as a rapid and convenient in vitro early screening system to predict drugs that can induce PLD.

Phospholipidosis (PLD) is characterized by an intracellular accumulation of phospholipids in lysosomes and the concurrent development of concentric lamellar bodies. Recently, Sawada et al.(2005) identified 17 genes as potential biomarkers of PLD in HepG2 cells. The present study was undertaken to determine if this set of genes measured by quantitative PCR could be validated in the same cell line. The objective was also to investigate the dose response relationship to further validate the assay and to select the concentrations to use for screening activities. In a first experiment (one concentration tested), out of the 17 genes, the best gene biomarkers of PLD (i.e. 11 genes) were selected for practical screening reasons. Based on these genes, 91.6 %(i.e. 11 out of 12) of the compounds known to induce PLD were identified as positive and all the negative compounds (i.e. 5 out of 5) were also confirmed. When the data obtained in the first experiment were compared to Sawada's data, the coefficient of correlation calculated was slightly higher than 75%. In the second experiment [26 compounds (all 17 compounds from the first experiment plus 9 other compounds) tested at a minimum of 3 concentrations], 93.3 %(14/15) of the compounds known to induce PLD were identified as such and all the negative controls (6 compounds) were also confirmed. Three compounds likely to induce PLD were identified as positive in our assay. Finally, two compounds for which no data are available were also tested. When both experiments 1 and 2 were compared, the coefficient of correlation for 16 compounds tested at the same concentrations reached 87.7 %. In conclusion, the present study further confirms the utility of gene expression in HepG2 cells to identify a potential to induce PLD. Finally, based on the data presented, researchers are encouraged to use a range of minimum 3 concentrations (e.g. 12.5, 25 and 50 muM) to screen for PLD in the human HepG2 cell line.

Agonists of peroxisome proliferator-activated receptor gamma (PPARgamma) are a new class of oral drugs designed to treat insulin-resistant diabetes (i.e., type 2 diabetes). However, troglitazone, the first compound in the class approved by the US Food and Drug Administration (FDA) in 1997 was found to be hepatotoxic and was withdrawn from the market after reports of severe liver failure. The mechanism of PPAR gamma agonist-induced hepatotoxicity remains unknown. In this study, we examined the hepatotoxic effects of five PPAR gamma agonists (ciglitazone, pioglitazone, rosiglitazone, troglitazone, and JTT-501) on rat primary hepatocytes and human HepG2 cells. We also compared the gene expression profiles of rat primary hepatocytes after exposure to PPAR gamma agonists by using the Rat Genome Survey Microarray system from Applied Biosystems in order to understand the mechanisms of hepatotoxicities induced by PPARgamma agonists. Consistent with the hepatotoxicity data, our results demonstrate that the gene expression profiles affected by troglitazone and ciglitazone can be clearly distinguished from those by pioglitazone and rosiglitazone. Genes that are differentially expressed between the more toxic troglitazone/ciglitazone group and the less toxic rosiglitazone/pioglitazone group are involved in necrotic, apoptotic, and cell proliferative pathways. The five compounds were also clustered based on a set of molecular descriptors. The clustering based on chemical structural information is in good agreement with the clustering of compounds based on cytotoxicity or gene expression data, indicating a strong relationship between chemical structure and biological endpoints. Our work suggests that microarray analysis together with toxicological observations can be used to rank drugs for hepatotoxicity and to evaluate the safety of new compounds.