10) where R(pCa) represents the column of the measured ratios, and
5. If the readout needs to be compared with data from the xCELLigence system, the corresponding values from the
Simin ệz et al.
References
1. Giaever I, Keese CR (1984) Monitoring fibro- blast behavior in tissue culture with an applied electric field. Proc Natl Acad Sci U S A 81:3761–3764
2. Giaever I, Keese CR (1993) A morphological bio- sensor for mammalian cells. Nature 366:591–592 3. Kirstein SL, Atienza JM, Xi B et al (2006) Live
cell quality control and utility of real-time cell electronic sensing for assay development. Assay Drug Dev Technol 4:545–553
4. Thedinga E, Kob A, Holst H et al (2007) Online monitoring of cell metabolism for studying pharmacodynamic effects. Toxicol Appl Pharmacol 220:33–44
5. Giaever I, Keese C (1991) Micromotion of mammalian cells measured electrically. Proc Natl Acad Sci U S A 88:7896–7900
6. Lo CM, Keese CR, Giaever I (1995) Impedance analysis of MDK cells measured by electric cell- substrate impedance sensing.
Biophys J 69:2800–2807
7. Keese CR, Wegener J, Walker SR et al (2004) Electrical wound-healing in vitro. Proc Natl Acad Sci U S A 101:1554–1559
8. Clark PR, Kim RK, Pober JS et al (2015) Tumor necrosis factor disrupts claudin-5 endo- thelial tight junction barriers in two distinct NF-kappaB-dependent phases. PLoS One 10(3):e0120075
9. Rees MD, Thomas, SR (2015) Using cell- substrate impedance and live cell imaging to measure real-time changes in cellular adhesion and de-adhesion induced by matrix modifica- tion. Vis Exp (96):e52423
10. Bilandzic M, Stenvers KL (2014) Assessment of ovarian cancer spheroid attachment and invasion of mesothelial cells in real time. J Vis Exp (87). doi:10.3791/51655
11. ệz S, Maercker C, Breiling A (2013) Embryonic carcinoma cells show specific dielectric resistance profiles during induced differentiation. PLoS One 8:e59895
12. Angstmann M, Brinkmann I, Bieback K et al (2011) Monitoring human mesenchymal stromal cell differentiation by electrochemi- cal impedance sensing. Cytotherapy 13:
1074–1089
275
Daniel F. Gilbert and Oliver Friedrich (eds.), Cell Viability Assays: Methods and Protocols, Methods in Molecular Biology, vol. 1601, DOI 10.1007/978-1-4939-6960-9_22, © Springer Science+Business Media LLC 2017
Chapter 22
Nano-QSAR Model for Predicting Cell Viability of Human Embryonic Kidney Cells
Serena Manganelli and Emilio Benfenati
Abstract
Traditional Quantitative Structure-Activity Relationships (QSAR) models based on molecular descriptors as translators of chemical information show some drawbacks in predicting toxicity of nanomaterials due to their unique properties and to their nonhomogeneous structure.
This chapter provides instructions on how to use CORAL, freely available software for building nano- QSAR models. CORAL makes use of descriptors based on “quasi-SMILES” representing physicochemical features and/or experimental conditions as an alternative to traditional SMILES encoding chemical struc- ture to build up predictive nano-QSAR models for cytotoxicity.
Key words Nanoparticles, Cell viability, Nano-QSAR, CORAL, Quasi-SMILES
1 Introduction
Nanoparticles (NPs), nano-objects with all three dimensions at the nanoscale, are commonly included in a larger matrix or substrate referred to as nanomaterial (NM) that measures less than 100 nm at least one dimension [1, 2].
The past decade has been characterized by an increase of NMs applications in several areas of science and technology [3]. These materials are commonly employed in consumer products (cosmet- ics and sunscreens, stain-resistant clothing, self-cleaning windows), in biomedical fields (drug delivery agents, biosensors, or imaging contrast agents), and also in various electronics systems and space technology [3, 4].
However, information concerning the possible toxicological implications derived from the use of these materials for the human and environmental health is still few [5]. Considering their wide range of applications, the impact of NMs on human health and the environment is of great interest. Nanotoxicology has been proposed as a new field of toxicology necessary to fill the gap in knowledge
related to the adverse health effects likely to be caused by exposure to NMs [4].
As already mentioned a very small dimension characterizes NPs. Thanks to the small sizes, these substances have highly desir- able properties from mechanical, electrical, and chemical points of view. Interestingly this same property is also one of the main rea- sons of toxicological reactivity [6]. However, the NPs size is not the only factor that can lead to adverse effects for human health. In particular, nanoscale materials have far larger surface area than larger-scale materials of similar masses: this is an important feature affecting bio-reactivity. As surface area per mass of a material increases, a greater amount of the material can be exposed to sur- rounding environment, thus affecting reactivity [7]
Also size distribution, agglomeration state, shape, crystal struc- ture, chemical composition, surface area, surface chemistry, surface charge, and porosity may influence the biological activity of these substances [8].
The human body can be exposed to NPs through different routes such as inhalation, ingestion, injection, and dermal expo- sure. After absorption, these substances can be transported to blood leading to the possibility to cause adverse reactions in several organs. Among the organs involved, the kidney could be a target since it receives high blood supply from the total organism and it has an active role in elimination of xenobiotics [9].
Silicon dioxide, also known as silica (SiO2), is one of the most commonly used NMs [10, 11]. SiO2 can be divided into types:
crystalline and noncrystalline (amorphous) silica. Amorphous silica can be further divided into naturally occurring amorphous silica and synthetic forms. The latter represents the SiO2 intentionally manufactured which has found widespread applications [12]. Silica NPs are employed in chemical mechanical polishing and as addi- tives to drugs, cosmetics, printer toners, varnishes, and food stuffs;
in biomedical and biotechnological fields, such as biosensors for the simultaneous assay of glucose, lactate, l-glutamate, and hypo- xanthine in rat striatum; and biomarkers for leukemia cell identifi- cation using optical microscopy imaging, cancer therapy, DNA delivery, drug delivery, and enzyme immobilization [13]. Since SiO2 found many applications in the everyday life, special attention should be paid to their potential toxic effects [10, 11]. Traditionally, toxicity studies on NPs often focused on lung damage, since these substances can easily disperse into the air due to their very low density and so can be inhaled. However, SiO2 (45 nm) was found also in liver, urinary bladder, and kidney after intravenous injection in mice [14]. Moreover, it was demonstrated that SiO2 NPs with a size ranging from of 50 and 100 nm were eliminated with the urine, proving the possibility of exposure to these substances also for other organs than lungs [15]. The mechanism of SiO2 toxicity
277 is still poorly understood. However, recent studies on SiO2 NPs toxicity demonstrated that these substances induce oxidative stress and pro-inflammatory responses in the rodent model and in several types of cultured mammalian cell lines [12, 15]. Oxidative stress, mediated by reactive oxygen species (ROS) production, is a key event in the mechanism of toxicity for many NPs. A healthy human organism is able to develop a series of cellular defense mechanism in order to decrease the ROS level. However, this is not always the case. Indeed, when the ROS production is high, many events such as lipid peroxidation occur that can finally lead to cellular dysfunc- tion or death [15, 16].
Since the production of synthetic NPs is increasing, the expo- sure to these substances is supposed to increase consequently in the coming years. In this context, the challenge for scientists is to develop new knowledge and approaches for the safety evaluation of NPs to be used for the assessment of the safety of NMs [6, 17].
In silico methods can be used for toxicological characterization of NPs since they have many advantages in terms of cost, time- effectiveness, and ethical implications [18]. Among in silico appro- aches, quantitative structure-activity relationships (QSAR) seem to be the most promising. Basically, QSAR are predictive computa- tional models that aim at defining a mathematical equation between the variance in molecular structures (molecular descriptors) and the variance in a given property (endpoint) for a set of compounds [19]. Aside from toxicological activity, this approach can be employed for predicting various physical–chemical properties as well, and, in such cases, it is named as QSPR (quantitative struc- ture–property relationship) [20].
Besides in silico models, a number of cell-based assays have been widely employed in toxicological studies to determine the viability and/or cytotoxicity after exposure to NPs [21]. The lac- tate dehydrogenase (LDH) leakage assay, the neutral red, and the 3-[4, 5-dimethylthiazol-2-yl]-2, 5 diphenyl tetrazolium bromide (MTT) assay are the most commonly employed tests and they use colorimetric or fluorescent dyes as markers to determine cell viabil- ity through assessment of membrane integrity or cell metabolism [21, 22].
In particular, the MTT assay is based on the protocol described for the first time by Mossmann [23]. MTT is a yellow water- soluble tetrazolium dye that succinate dehydrogenase, present within the mitochondria, can convert into insoluble purple formazan by cleavage of the tetrazolium ring. Since formazan production can only occur in metabolically active cells, the level of activity is a measure of the cells viability (Fig. 1) [22]. The amount of MTT- formazan produced is measured spectrophotometrically. However, a limitation of MTT assay is that it is not able to discriminate between a cytotoxic (cell-killing effects) and a cytostatic (reduced growth rate) effect [24].
Nano-QSAR for Cell Viability Predictions
The MTT enters the cells and passes into the mitochondria where it is reduced to an insoluble, purple colored formazan prod- uct. Since reduction of MTT can only occur in metabolically active cells, the level of this activity is a measure of the viability of the cells.
In order to assess in vitro renal toxicity, the human embryonic kidney cell line (HEK293) is a widely used model [25]. This cell line is well characterized for its relevance to the toxicity model in human and it has been previously used for toxicological assess- ments [5, 26–28].
To provide an overview of one of recently introduced stra- tegy of QSAR modeling for NPs (nano-QSAR), in the below sub- sections of this chapter we propose to illustrate the different steps involved in model building using the CORAL software. Specifically, we describe how we used this software to construct a QSAR model for SiO2 NPs [29] on a dataset of in vitro data derived from cyto- toxicity studies (MTT assay) conducted on HEK293 cell line [5].
The algorithm we propose makes use of descriptors, which are mathematical function of size, concentration, and exposure time.
The so-called optimal descriptors are based on “quasi-SMILES”.
Differently from “Simplified Molecular Input Line Entry specifica- tion” (SMILES), which are strings of characters encoding molecu- lar structures, quasi-SMILES are strings representing particular conditions, not of the molecular structure. Numbers and/or alphabet letters can specify each of these features [29].