Patent Application
20180371408
The inventor has found that use of immortalized cell in a cell culture model described herein results in an unexpected increase in the reproducibility and sensitivity compared to the same cell culture model based on primary cells (Yu et al., 2005, Toxicol Sci, 84(2):378-93). Cytotoxicity (LC50) data generated from the co-culture system described herein are more consistent with the in vivo animal's reproductive toxicity as compared to a primary cell culture system (Yu et al., 2005, Toxicol Sci, 84(2):378-93). The use of immortalized cells significantly reduced source of variations in obtaining primary cells, and reduced the operation time, which significantly increases the power of sensitivity in detecting the potential toxicity of compounds.
Provided herein is a composition including cells and a protein matrix. The cells include spermatogonial cells, Sertoli cells, and Leydig cells. In one embodiment, the cells are immortalized. The spermatogonial cells can be present at 70-90%, the Sertoli cells can be present at 10-20%, and the Leydig cells can be present at 1-10%, where the spermatogonial cells, Sertoli cells, and Leydig cells add up to at least 100% of the cells in the composition.
The protein matrix includes a protein mixture that can represent an extracellular matrix. In one embodiment, extracellular matrix molecules can include collagen, fibronectin, laminin, vitronectin, tenascin, entactin, thrombospondin, elastin, gelatin, fibrillin, merosin, anchorin, chondronectin, link protein, bone sialoprotein, osteocalcin, osteopontin, epinectin, hyaluronectin, undulin, epiligrin, Kalinin, glycosaminoglycans, proteoglycans, a chemoattractant, a cytokine, a growth factor, or a combination thereof.
The spermatogonial cells, Sertoli cells, and Leydig cells can be rodent cells. An example of a spermatogonial cell is a C18-4 cell, and example of a Sertoli cell is a TM3 cell, and an example of a Leydig cell is a TM4 cell.
Also provided is a method for producing a cell culture. In one embodiment, the method includes combining cells and a protein matrix in a container to result in a cell culture, and incubating the cell culture under conditions suitable for maintaining viability of the cells. The cells include spermatogonial cells, Sertoli cells, and Leydig cells. In one embodiment, the cells are immortalized. In one embodiment, the spermatogonial cells are present at 70-90%, the Sertoli cells are present at 10-20%, the Leydig cells are present at 1-10%, and the spermatogonial cells, Sertoli cells, and Leydig cells add up to 100% of the total cells.
The protein matrix includes a protein mixture that can represent an extracellular matrix. In one embodiment, extracellular matrix molecules can include collagen, fibronectin, laminin, vitronectin, tenascin, entactin, thrombospondin, elastin, gelatin, fibrillin, merosin, anchorin, chondronectin, link protein, bone sialoprotein, osteocalcin, osteopontin, epinectin, hyaluronectin, undulin, epiligrin, Kalinin, glycosaminoglycans, proteoglycans, a chemoattractant, a cytokine, a growth factor, or a combination thereof.
Further provided are methods of using the composition. In one embodiment, the composition is present in a container, such as a well of a multi-well plate. The method includes contacting the cells with a compound to form a mixture, and incubating the mixture under conditions suitable for maintaining viability of the cells in the absence of the compound. The method also includes determining the status of cells.
In one embodiment, the status includes changes of cell cycle, DNA damage, nuclear shape, cell proliferation, and/or cytoskeleton. In one embodiment, the status includes cell viability. The method can further include determining whether the compound affects that status of the spermatogonial cells, the Sertoli cells, the Leydig cells, or a combination thereof. A compound can alter the status of cells, e.g., reduce cell viability, or have no detectable effect. In one embodiment, determining includes measuring neutral red uptake capacity of the cells.
In another embodiment, the method including contacting a composition described herein with a compound and analyzing viability of the cells, where a reduction of viability indicates the compound is a toxic compound.
The term “and/or” means one or all of the listed elements or a combination of any two or more of the listed elements.
The words “preferred” and “preferably” refer to embodiments of the disclosure that may afford certain benefits, under certain circumstances. However, other embodiments may also be preferred, under the same or other circumstances. Furthermore, the recitation of one or more preferred embodiments does not imply that other embodiments are not useful, and is not intended to exclude other embodiments from the scope of the disclosure.
The term “comprises” and variations thereof do not have a limiting meaning where these terms appear in the description and claims.
It is understood that wherever embodiments are described herein with the language “include,” “includes,” or “including,” and the like, otherwise analogous embodiments described in terms of “consisting of” and/or “consisting essentially of” are also provided.
Unless otherwise specified, “a,” “an,” “the,” and “at least one” are used interchangeably and mean one or more than one.
Also herein, the recitations of numerical ranges by endpoints include all numbers subsumed within that range (e.g., 1 to 5 includes 1, 1.5, 2, 2.75, 3, 3.80, 4, 5, etc.).
Reference throughout this specification to “one embodiment,” “an embodiment,” “certain embodiments,” or “some embodiments,” etc., means that a particular feature, configuration, composition, or characteristic described in connection with the embodiment is included in at least one embodiment of the disclosure. Thus, the appearances of such phrases in various places throughout this specification are not necessarily referring to the same embodiment of the disclosure. Furthermore, the particular features, configurations, compositions, or characteristics may be combined in any suitable manner in one or more embodiments.
In the description herein particular embodiments may be described in isolation for clarity. Unless otherwise expressly specified that the features of a particular embodiment are incompatible with the features of another embodiment, certain embodiments can include a combination of compatible features described herein in connection with one or more embodiments.
For any method disclosed herein that includes discrete steps, the steps may be conducted in any feasible order. And, as appropriate, any combination of two or more steps may be conducted simultaneously.
Statistical analysis was conducted by one-way ANOVA followed by Tukey-Kramer multiple comparison (*p≤0.05 and **p≤0.001). The shaded area reflects the 95% confident intervals.
The following detailed description of illustrative embodiments of the present disclosure may be best understood when read in conjunction with the following drawings.
Provided herein are compositions that include testicular cells and methods for making and using the compositions. A composition includes spermatogonial cells, Sertoli cells, and Leydig cells. The cells are in vitro (e.g., the cells are present in an artificial environment, such as a test tube or culture dish). The cells useful herein are cells that are capable of long term culture in tissue culture medium. A population of cells having a uniform genetic makeup, e.g., one type of cell in a culture, is also referred to herein as a cell line. Typically, a cell useful herein has the characteristic of being immortalized, and can be maintained in long term cultures.
Spermatogonial cells, also referred to as gonocytes, can have morphological features of type A spermatogonia, and express germ cell-specific genes such as GFRA1, Dazl and Ret, and stem cell specific genes such as piwi12 and prame11. In one embodiment, whether a cell is a spermatogonial cell can be determined using typical markers of mouse germ cell nuclear antigen (GCNA1). Spermatogonial cells can be obtained from any mammal. In one embodiment the cells are obtained from a rodent, such as a mouse or rat. Methods for obtaining spermatogonial cells include the STAPUT method that utilizes gravity sedimentation on a 2%-4% bovine serum albumin (BSA) gradient (Dym et al., 1995, Biol Reprod., 52:8-19; Hofmann et al., 2005, Stem Cells, 23(2):200-210). In one embodiment a spermatogonial cell is immortalized. As used herein an immortalized cell is one that does not undergo senescence and instead can continue to undergo division. Methods for immortalizing a cell, such as a spermatogonial, Sertoli, or Leydig cell, are known in the art and include, but are not limited to expression of certain viral genes. An example of a useful spermatogonial cell line is the C18-4 cell line, a cell line established by transfecting mouse spermatogonial stem cells with a plasmid allowing the expression of the SV40 large T antigen under the control of a ponasterone A-driven promoter (Hofmann et al., 2005, Stem Cells, 23(2):200-210).
Sertoli cells express follicle stimulating hormone, androgen receptor, and progesterone receptor. In one embodiment, whether a cell is a Sertoli cell can be determined by identifying the presence of SOX9, a Sertoli cell specific nuclear protein. For instance, anti-SOX9 antibody can be used to determine if SOX9 is present. In one embodiment the cells are obtained from a rodent, such as a mouse or rat. Methods for obtaining Sertoli cells are known (Chang et al., 2011, Biotechniques, 51(5): 341-344). In one embodiment a Sertoli cell is immortalized. Methods for immortalizing a Sertoli cell are known (Hofmann et al., 1992, Exp Cell Res. 201(2):417-35). An example of a useful Sertoli cell line is the TM4 cell line, available from the ATCC® as CRL-1715™.
Leydig cells express androgen receptor and progesterone. In one embodiment, whether a cell is a Leydig cell can be determined using steroidogenic acute regulatory protein (StAR) or identifying the presence of 3β-HSD, a Leydig cell specific protein. For instance, anti-3β-HSD antibody can be used to determine if 3β-HSD is present. In one embodiment the cells are obtained from a rodent, such as a mouse or rat. Methods for obtaining Leydig cells are known (Chang et al., 2011, Biotechniques, 51(5): 341-344). Methods for immortalizing a Leydig cell are known (Hofmann et al., 1992, Exp Cell Res. 201(2):417-35). An example of a useful Leydig cell line is the TM3 cell line, available from the ATCC® as CRL-1714™.
The spermatogonial cells in a composition make up from 70% to 90% of the cells in the composition (e.g., 70%, 71%, 72%, 73%, 74%, 75%, 76%, 77%, 78%, 79%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, or 90%). The Sertoli cells in a composition make up from 10% to 20% of the cells in the composition (e.g., 10%, 11%, 12%, 13%, 14%, 15%, 16%, 17%, 18%, 19%, or 20%). The Leydig cells in a composition make up from 1% to 10% of the cells in the composition (e.g., 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, or 10%). In one embodiment, the spermatogonial cells, Sertoli cells, and Leydig cells are present in the composition at fractions that mimic mouse testis at approximately 5 days postnatal, for instance 80%, 15%, and 5%, respectively (e.g., at a ratio of 16:3:1). In one embodiment, the spermatogonial cells, Sertoli cells, and Leydig cells make up essentially all the cells in the composition, thus the fractions of these three cell types add up to 100%. In one embodiment, the amount of each cell type in a composition can mimic the fractional percent present in the mouse testis at 5 days postnatal, e.g., 80% spermatogonial cells, 15% Sertoli cells, and 5% Leydig cells. In one embodiment, the composition can include one or more additional cells types. The one or more additional cells types can be present at, for instance, no more than 0.1% to 1% of the total number of cells.
A composition disclosed herein also includes a protein matrix. In one embodiment, a protein matrix represents an extracellular matrix. As used herein, the terms “extracellular matrix” and “ECM” are used interchangeably and refer to a natural or artificial composition that includes extracellular molecules typically secreted by cells and providing structural and/or biochemical support for cells. In some embodiments, a protein matrix includes proteins from solubilized basement membrane extracted from a tissue, such as testicular tissue. Examples of ECM molecules include, but are not limited to, collagen, fibronectin, laminin, vitronectin, tenascin, entactin, thrombospondin, elastin, gelatin, fibrillin, merosin, anchorin, chondronectin, link protein, bone sialoprotein, osteocalcin, osteopontin, epinectin, hyaluronectin, undulin, epiligrin, Kalinin, glycosaminoglycans, proteoglycans, chemoattractants, cytokines, growth factors, and a combination thereof (Ingber et al., US Published Patent Application 20170158997). One example of a useful protein matrix is a gelatinous protein mixture secreted by Engelbreth-Holm-Swarm (EHS) mouse sarcoma cells. This gelatinous protein mixture is available from Corning Life Sciences and BD Biosciences under the trade name MATRIGEL (Hughes et al., 2010, Proteomics, 10(9):1886-1890), and from Trevigen, Inc. under the trade name CULTREX BME.
The amount of protein matrix present in a composition can vary. In one embodiment, an amount of protein matrix used is sufficient to cover the surface of a container holding the composition. In another embodiment, an amount of protein matrix used is sufficient to produce a three-dimensional gel. Examples of concentrations of a protein matrix include, but are not limited to, at least 10 micrograms/ml (μg/ml), at least 50 μg/ml, or at least 75 μg/ml, and no greater than 200 μg/ml, no greater than 175 μg/ml, no greater than 150 μg/ml, or no greater than 125 μg/ml. In one embodiment, the amount of protein matrix used is at least 75 μg/ml to no greater than 125 μg/ml, such as 100 μg/ml.
In one embodiment, when present in a container, e.g., a well of a multi-well plate, the composition can include F-actin connecting cells in a three-dimensional structure. In one embodiment, after incubation for 48 hours a composition can include bundles of F-actin stretching out to other cells and forming cord-like structures.
Also provided are methods for making a composition described herein. A composition can be produced by combining cells and a protein matrix in a suitable container and incubating the cell culture under conditions suitable for maintaining the viability of the cells. Suitable conditions include, but are not limited to, 37° C., 5% CO2, and a standard medium for culturing rodent cells. In one embodiment, the three types of cells (spermatogonial cells, Sertoli cells, and Leydig cells) are combined to result in a composition described herein and then added to a suitable container. The protein matrix described herein can be added to the container before, with, or after addition of the cells. Examples of a suitable container includes, but is not limited to, standard cell culture containers such as a multi-well plate, e.g., a 96-well plate, and other containers used in tissue culture. The total number of cells added to a container can be adjusted to result in 70% to 80% confluence after overnight incubation. For instance, 1.5×104 cells can be used to inoculate a well of a 96-well plate. A cell culture can be used in a method described herein after overnight incubation, but other incubation times before use in a method are possible.
Further provided are methods for using a composition described herein. In one embodiment, a method includes providing a composition of cells described herein, and contacting cells in the composition with a compound, also referred to as a test compound, to form a mixture. The cells are typically present in a container, such as a well of a multi-well plate. At the time the compound is added the cells can be present in the container at a level of confluence of at least 50% (50 percent of the surface is covered by the cells), at least 60%, at least 70%, at least 80%, at least 90%, or at least 100% (the surface is completely covered by the cells) confluence. In one embodiment, the cells are at 70% to 80% confluence.
The compound can be one that is known to be or suspected of being a toxicant. As used herein, a “toxicant,” “testicular toxin,” and “reproductive toxicant” are used interchangeably and refer to a compound that can be toxic to a male reproductive system. A compound can be a chemical compound, including, for instance, an organic compound, an inorganic compound, a protein, or a metal. The compound can be a member of a library of many compounds, or the method can be used to specifically test one or more specific compounds or specific classes of compounds. Examples of libraries include, but are not limited to, libraries from the National Toxicology Program or the Environmental Protection Agency. In one embodiment, a compound is one that does not require metabolic activation.
In one embodiment, a compound is assessed at concentrations that are physiologically relevant to a mammalian system. Such concentrations can be determined by the skilled person for each specific compound examined. In one embodiment, the compound is assessed in concentrations routinely found, or expected to be found, in mammalian tissue or systems (e.g., from routine exposure). In another embodiment, the compound is assessed in concentrations routinely found in the environment. In one embodiment, the compound is assessed at a concentration of 1 nM to 100 mM. In one embodiment, the concentration of the compound is assessed at from 1 μM to 100 mM. In one embodiment, the compound is at a concentration of at least 1 nM, at least 10 nM, at least 100 nM, at least 1 μM, at least 5 μM, at least 10 μM, at least 20 μM, or at least 50 μM. In one embodiment, the compound is at a concentration of no greater than 80 μM, no greater than 100 μM, no greater than 200 μM, no greater than 500 μM, no greater than 750 μM, no greater than 0.1 mM, no greater than 0.5 mM, no greater than 1 mM, no greater than 5 mM, no greater than 10 mM, no greater than 20 mM, no greater than 50 mM, no greater than 75 mM, or no greater than 100 mM. A compound can be assessed at different concentrations, e.g., different doses. Higher and lower concentrations may also be useful, depending upon the specific compound being assessed. Higher concentrations can be used for compounds that have limited to poor solubility. In one embodiment, an initial dose-response curve can be established and used to adjust a dose range of a compound. The compound can be dissolved in a solvent, e.g., an aqueous or organic solvent, prior to use.
The method also includes incubating the mixture under conditions suitable for maintaining viability of the cells in the absence of the compound. Such conditions include the use of standard media and methods for maintaining cultured rodent cells, such as cultured mouse cells, including TM3, TM4, and C18-4 cells.
The status of cells in the mixture is determined a suitable period of time after the compound is added, such as 12 hours, 24 hours, 48 hours, or 72 hours after addition of the compound. In one embodiment, the viability of the cells is determined. Cell viability can be determined using essentially any method. In one embodiment, cell viability is determined by measuring neutral red uptake capacity of cells. Neutral red is retained inside the lysosomes of viable cells, while the dye is not retained by dead cells. Dye retention is proportional to the number of viable cells, and can be measured based on a neutral red absorbance value. Cells treated with solvent alone can be used as a control group with cell viability set as 100%. After incubation with the compound, e.g., 12 hours, 24 hours, 48 hours, or 72 hours, the medium can be replaced with a medium containing neutral red at an appropriate concentration. Following incubation with the dye, cells can be washed, neutral red eluted with a suitable solution, such as a combination of acetic acid and ethanol, and absorbance values measured at 540 nm. Cell viability can be expressed as a percentage of the mean of solvent controls after subtracting a background control. In one embodiment, a compound is considered to be a toxicant for cells of the cell culture when there is a statistically significant reduction of cell viability compared to the solvent control. In one embodiment, a compound is considered to be a toxicant for cells of the cell culture when the fold change over the control is at least 0.1, at least 0.2, at least 0.3, at least 0.4, at least 0.5, at least 0.6, at least 0.7, at least 0.8, or at least 0.9. Optionally, the inhibitory concentration (IC) of a compound can be determined, including the IC20, IC50 (half maximal inhibitory concentration), and/or IC85.
In another embodiment, the cell status evaluated can be changes of cell cycle, DNA damage, nuclear shape, cell proliferation, and/or cytoskeleton. Cytoskeleton can include F-actin cytoskeletal structure, such as intensity and distribution of actin fibers. Methods for evaluating these and other variables include, but are not limited to, single-cell based high-content analysis (Liang et al., 2017, Toxicol Sci, 155(1), 43-60).
A compound can affect one cell type in a composition, two cell types in a composition, or all cell types in a composition. Thus, in one embodiment, the method further includes determining whether the compound affects spermatogonial cells, Sertoli cells, Leydig cells, or a combination thereof.
A compound that alters the status of one or more cell types in the composition, e..g, reduces viability, is considered a possible toxicant, and optionally can be further tested using other model systems available to further evaluate whether the compound is a male reproductive toxicant. It is also expected that in some embodiments compounds will be identified that do not affect cell viability.
The present invention is illustrated by the following examples. It is to be understood that the particular examples, materials, amounts, and procedures are to be interpreted broadly in accordance with the scope and spirit of the invention as set forth herein.
Primary testicular cell co-culture model has been used to evaluate testicular abnormalities during development, and was able to identify the testicular toxicity of phthalates. However, the primary testicular cell co-culture model has disadvantages in employing animals for the isolation of testicular cells, and the complicated isolation procedure leads to inconsistent results. We developed an in vitro testicular co-culture model from rodent testicular cell lines, including spermatogonial cells, Sertoli cells, and Leydig cells with specified cell density and Extracellular
Matrix (ECM) composition. Using comparative high-content analysis of F-actin cytoskeletal structure between the co-culture and single cell culture models, we demonstrated a three-dimensional structure of the co-culture, which created an in vivo-like niche, and maintained and supported germ cells within a three-dimensional environment. We validated this model by discriminating between reproductive toxicants and non-toxicants among 32 compounds in comparison to the single cell culture models. Furthermore, we conducted a comparison between the in vitro (IC50) and in vivo reproductive toxicity testing (lowest observed adverse effect level on reproductive system, rLOAEL) We found the in vitro co-culture model could classify the tested compounds into four clusters, and identify the most toxic reproductive substances, with high concordance, sensitivity, and specificity of 84%, 86.21%, and 100%, respectively. We observed a strong correlation of IC50 between the in vitro co-culture model and the in vivo testing results. Our results suggest that this novel in vitro co-culture model may be useful for screening testicular toxicants and prioritize chemicals for further assessment in the future.
Reproductive and developmental disorders caused by exposure to environmental chemicals are a prominent health issue worldwide (Jenardhanan et al., 2016; Leung et al., 2016). Animal testing for evaluating potential reproductive toxicity is one of the most complicated, time-consuming and expensive processes when examining complex endpoints. Testing under the current guidelines requires a large number of animals, ranging from 560-6,000 animals per chemical or drug (OECD, 2016). The implementation of the new European Registration, Evaluation, Authorization and Restriction of Chemical (REACH) program requires toxicological information to be submitted for about 30,000 existing chemicals. Although REACH promotes limiting vertebrate animal testing as far as possible, the lack of suitable alternatives will probably increase the use of animals (Hareng et al., 2005; Hartung and Rovida, 2009; Hofstetter et al., 2013; Luijten et al., 2007; Parks Saldutti et al., 2013; Sauer, 2004; Scialli and Guikema, 2012). Moreover, every year, approximately 700 new chemicals are introduced into the market, which imposes a great burden on reproductive and developmental toxicity testing. The Interagency Coordinating Committee on the Validation of Alternative Methods (ICCVAM) developed recommendations for minimum procedural standards and testing methods for the validation of in vitro estrogen receptor (ER) and androgen receptor (AR) binding and transcriptional activation (TA) assays (Casey, 2016; ICCVAM, 2012). So far, there are no validated alternative tests that would cover different aspects of the reproductive cycle. Thus, it has become increasingly important to develop an in vitro test that can serve as an equally effective alternative to animal testing for reproductive toxicity. In 2007, the U.S. Environmental Protection Agency (EPA) launched a large-scale program, ToxCast, to investigate high-throughput, in vitro assays to prioritize substances for further in-depth toxicological evaluation, identify mechanisms of action, and develop predictive models for in vivo biological response (Houck et al., 2009). The focus of the ToxCast program was to generate an in vitro bioactivity profile for each chemical, and correlate this profile with the toxicity data from in vivo animal studies (Auerbach et al., 2016; Karmaus et al., 2016; Kavlock et al., 2012; Paul Friedman et al., 2016). There are more than 500 assays across nine assay technology platforms, including cell-free high throughput screening assays and cell-based assays in multiple human and rodent primary and derived cell lines. Although a small number of assays were found to be associated with the identified reproductive toxicants, and showed a certain level of predictive power for reproductive toxicity (Leung, et al., 2016; Martin et al., 2011; Reif et al., 2010), there is no in vitro model in the ToxCast program designed specifically for detecting reproductive toxicity.
Currently, in vitro reproductive screening models for testicular development and spermatogenesis are actively being developed (Hareng, et al., 2005; Hofmann et al., 2005a; Luijten, et al., 2007; Parks Saldutti, et al., 2013; Yu et al., 2008; Yu et al., 2009; Yu et al., 2005). In vitro culture systems have been used to evaluate testicular changes during normal development (Bilinska, 1989; Chapin et al., 1988; Gray, 1986; Griswold, 1998; Hadley et al., 1985; Lejeune et al., 1998). Sertoli/gonocytes co-cultures (SGC) are used to examine cell-cell interactions, as well as effects of hormones and growth factors on spermatogonia survival and proliferation in vitro (Mather et al., 1990; Orth and Jester, 1995). Human fetal testis xenografted into the renal subcapsular space was developed to study the effects of toxicants on human tissues (Spade et al., 2014a; Spade et al., 2014b). Production of post-meiotic spermatids in explained pieces of testis were observed in both mouse and rat models (Brannen et al, 2016; Sato et al., 2011). These advances provide evidence that Sertoli and Leydig cells, as well as peritubular myoid or endothelial cells, are all essential in supporting and maintaining spermatogenesis in the testis. Self-renewal and progeny production of spermatogonia are controlled by the neighboring differentiated cells and extracellular matrix (ECM), known as substrate in vivo niches, while Sertoli cells are required for successful differentiation of germ cells in vitro culture systems (Griswold, 1998). The ECM Matrigel-based primary testicular cell model was reported to form a testicular-like multilayered architecture that mimics in vivo characteristics of seminiferous tubules (Harris et al., 2015; Wegner et al., 2013; Wegner et al., 2014; Yu, et al., 2008; Yu, et al., 2009; Yu, et al., 2005). As previously demonstrated, this model differentially responds to various phthalate exposures with changes in several cellular pathways, showing sensitive response to testicular toxicity (Harris, et al., 2015; Yu, et al., 2009). However, this in vitro primary testicular cell co-culture model has the disadvantage of employing animals for the isolation of testicular cells, and the complicated isolation procedure leads to inconsistent results (Wegner, et al., 2013). Therefore, in this study, we developed an in vitro testicular cell co-culture model from rodent testicular cell lines using spermatogonial cells (C18-4), Sertoli cells (TM4), and Leydig cells (TM3). We tested this animal-free in vitro testicular co-culture model with 32 compounds and compared their cytotoxicities with any single cell culture of spermatogonia, Sertoli cell or Leydig cells, and further conducted a comparison between the in vitro (IC50 of cell viability) and in vivo reproductive toxicity testing (lowest observed adverse effect level on the reproductive system). We observed that the in vitro co-culture model could classify the tested compounds into four clusters, and identified the most toxic reproductive substances, which had high concordance, sensitivity, and specificity values of 84%, 86.21%, and 100%, respectively. We observed a strong correlation of IC50 between this in vitro testicular co-culture model and the in vivo testing results. We have demonstrated that this novel in vitro co-culture model may be useful in screening testicular toxicants in a wide concentration range, and will help prioritize chemicals for future assessment.
Dulbecco's modified Eagle's medium (DMEM), antibiotics (penicillin and streptomycin), fetal bovine serum (FBS), 0.25% trypsin/EDTA, and ethanol were purchased from GE Healthcare Life Sciences (Logan, Utah.). Nu-Serum culture supplement (Nu-serum) and Extracellular Matrix (ECM) Matrigel were from BD BioScience (Redford, Mass.). Glacial acetic acid was obtained from Merck (Darmstadt, Germany).
Both recognized reproductive toxicants and non-reproductive toxic compounds were selected for testing, as listed in Table 1. We selected 32 compounds, and obtained their in vivo toxicities by manually searching the literature and public sources, such as the LOAEL values provided in the ToxCast database (Chapin and Stedman, 2009; CIRM, 2008; Moorman et al., 2000). Most of compounds were purchased from Sigma-Aldrich (St. Louis, Mo.), Fisher Scientific (Gaithersburg, Md.), Chem Service (West Chester, Penn.) and GFS Chemicals (Powell, Ohio.), as indicated. Some compounds were graciously provided by the National Toxicology Program (NTP).
Mouse Leydig cells (TM3) and Sertoli cells (TM4) were purchased from ATCC. These cells were isolated from pre-pubertal mouse gonads (Mather, 1980; Mather and Phillips, 1984). TM3 cells specifically express androgen receptor and progesterone. TM4 cells specifically express follicle stimulating hormone, androgen receptor and progesterone receptor (Mather, 1980; Mather and Phillips, 1984). The mouse spermatogonial cell line C18-4 was established from germ cells isolated from the testes of 6-day-old Balb/c mice. This cell line shows morphological features of type A spermatogonia, and expresses germ cell-specific genes such as GFRA1, Dazl and Ret, and stem cell specific genes such as piwi12 and prame11. It proved to be an ideal cell model for studying the early phase of spermatogenesis, although the functional transplantations were not conducted to prove the stem cell nature (Hofmann, et al., 2005a; Hofmann et al., 2005b). The spermatogonial cells were maintained in DMEM medium composed of 5% FBS, and 100 U/ml streptomycin and penicillin in a 33° C., 5% CO2 humidified environment in a sub-confluent condition with passaging every 3-4 days. Leydig cells and Sertoli cells were cultured in DME/F12 medium supplemented with 100 U/ml streptomycin and penicillin, 5% horse serum, and 2.5% FBS, and maintained at 37° C. with 5% CO2. For both the single cell culture models and testicular co-culture model, the cells were inoculated with 1.5×104 per well in a 96-well plate, cultured overnight at 70-80% confluence, and treated with various concentrations of the testing compounds. For the testicular cell co-culture model, the cellular composition of fractions mimicked mouse testis around 5 days postnatal, and was 80%, 15% and 5% for gonocyte, Sertoli cell, and Leydig cell, respectively. Spermatogonial cells, Sertoli cells, and Leydig cells were mixed in this defined proportion and seeded into a 96-well plate in a DMEM/high glucose medium at 33° C., supplemented with 5% Nu-serum. ECM Matrigel (Corning, N.Y.) was added to each well for a final concentration of 100 μg/ml, and the plates were gently swirled to ensure dispersal of Matrigel after addition. Cells were cultured overnight and treated with various concentrations of testing compounds in the doses and time periods indicated.
All cultures were viewed with an Olympus inverted microscope equipped with phase-contrast optics (Olympus, Tokyo, Japan) at intervals during the culture to assess their general appearance. Images were captured and digitized with a Nikon Camera. To further examine the morphology of the cultured cells, a multi-parametric high-content analysis (HCA) was applied to quantify the F-actin cytoskeleton, nuclear shape, and cell proliferation. The spermatogonial cells were identified using typical markers of mouse germ cell nuclear antigen (GCNA1) (Developmental Studies Hybridoma Bank, http://dshb.biology.uiowa.edu, Iowa, IA)(Enders and May, 1994). Leydig cells were identified using steroidogenic acute regulatory protein (StAR)(Clark et al., 1994) (kindly gifted by Dr. Douglas M. Stocco, Texas Tech University). Sertoli cells were stained with SOX9 (Developmental Studies Hybridoma Bank, Iowa, IA). After culture for 48 and 72 h, cells were washed with phosphate buffered saline (PBS), fixed with 4% formaldehyde for 20 min at room temperature, and then washed with PBS three times. After permeabilization in Triton/PBS solution containing 0.1% Triton X-100 (TX-100) for 15 min, cells were blocked for 30 min in 3% bovine serum albumin (BSA, Sigma) in PBS with 0.1% TX-100, and then incubated with primary phospho-histone H3 antibody (1:200, Thermo Scientific) or phospho-γH2AX (γH2AX, Millipore-Sigma) in PBS and 0.1% TX-100 over night at 4° C. After washing with PBS/Tween-20 twice, cells were incubated with goat anti-mouse Dylight 650, goat anti-rabbit Dylight 550 (Thermo Scientific, MA), and Hoechst 33342 (Molecular Probes, OR) in PBS/BSA solution for 90 min at room temperature. Prior to image acquisition, cells were stained for 30 min at room temperature with Alexa Fluor 488 Phalloidin (Cell Signaling, MA) for F-actin staining. The multichannel images were automatically captured using an Arrayscan VTI HCS reader with a Studio 2.0 Target Activation BioApplication module (Thermo Scientific, MA). Thirty-six fields per well were acquired at 40× magnification using a Hamamatsu ROCA-ER digital camera in combination with 0.63× coupler and Carl Zeiss microscope optics for automatic image acquisition. Channel 1 (Ch1) applied the BGRFR 386_23 for Hoechst 33342, which was used for an auto-focus channel, and the objects (nuclei) were identified. For F-actin and γH2AX or Phospho-Histone H3 staining, Ch2 applied the BGRFR 485_20 for F-actin and Ch3 applied the BGRFR 650_13 for γH2AX or Phospho-Histone H3.
A four-channel assay was conducted to assess the cell-type specific marker staining for GCNA1 and StAR. Channel 1 was applied the BGRFR 386_23 for Hoechst 33342, Ch2 applied the BGRFR 485_20 for F-actin, Ch3 applied the BGRFR 650_13 for GCNA1, and Ch4 for StAR. For negative control, the primary antibody was omitted, and was stained with the secondary antibody only, indicating that the primary antibodies used were due to primary antibody specificity and not to unspecific binding of the secondary antibody to the cells.
Single-cell based High-content Analysis (HCA) provided multi-parameter phenotypic profiling characterization, including number, nuclear area, shape, and intensity, as previously reported (Liang et al., 2017). Nuclear shape measurement included P2A and LWR parameters. P2A, which evaluates nucleus smoothness, is a shape measurement based on the ratio of the nuclear perimeter squared to 4π*nucleus area (perimeter2/4π*nucleus area). LWR (length-width ratio), which evaluates nucleus roundness, measures the ratio of the length to the width of the nucleus. Total intensity was defined as the total pixel's intensities within a cell in the respective channel; the average intensity was defined as the total pixels' intensities divided by the area of a cell in the respective channel. With 36 fields of each well, at least 4000 cells were analyzed per well, and single-cell based data for each channel were exported for further statistical analysis. The experiments were performed with at least four biological replicates and repeated at least twice.
For quantitative analysis of three-dimensional F-actin cytoskeleton, we applied the Cytoskeletal Re-arrangement Assay in the Morphology Explorer Bioapplication (ThermoScientific, MA) to determine the number, dimensions and alignment—as well as the texture—of actin fibers based on the manufacturer's guideline (ThermoFisher, 2010). Briefly, for identification of fibers, pixels with high intensities were selected, and the threshold level controlled by the Assay Parameter Detection was set at a value of 1 (
Cell viability was determined by measuring the capacity of cells to take up NR (Repetto et al., 2008b; Yu, et al., 2009; Yu, et al., 2005). NR is retained inside the lysosomes of viable cells, while the dye cannot be retained if the cells die. Dye retention is proportional to the number of viable cells, and can be measured based on NR absorbance value. Cells were seeded to 96-well plates, and treated with different compounds at four doses with five replicates. Cells treated with the vehicle (0.05% DMSO) were used as the background group with cell viability set as 100%. After 24 h or 48 h, the medium was replaced with a medium containing NR dye (50 μg/ml, 200 μl per well). Following 3 h incubation, the supernatants were removed, the cells were washed with PBS twice, and NR was eluted with 100 μl of a 0.5% acetic acid/50% ethanol solution. The plate was gently rocked on a plate shaker, and absorbance values were measured at 540 nm with a Synergy HT microplate reader (BioTek, VT). Cell viability was expressed as a percentage of the mean of vehicle controls after subtracting the background control. The initial testing concentrations of these compounds were determined based on the published cytotoxicity data. After examination of the initial dose-response curve from the co-culture model, the dose ranges were adjusted. For those compounds with less cytotoxicity, the highest concentrations tested were 50 mM and 5 mM in the water-soluble or DMSO vehicle, respectively.
The U.S. EPA's ToxCast program reviewed the in vivo animal studies and established the Toxicity Reference Database (ToxRefDB) (available on the workd wide web at actor.epa.gov/actor/home.xhtml). Reproductive lowest observed adverse effect levels (rLOAEL) from in vivo studies were generated, and reflected reproductive toxicities (Martin, et al., 2011). The endpoints for determining rLOAEL of in vivo studies include, but are not limited to, primary fertility, early offspring survival, offspring weight, longer-term offspring survival, and other systemic toxicities of offspring. As previously reported, in vivo reproductive toxicants were defined as having achieved an rLOAEL lower than 500 mg/kg/day (Martin, et al., 2011). Compounds such as BPS, BPAF, and TBBPA with insufficient in vivo reproductive toxicity data were marked as “NA” for “no available information.” A concordance analysis was performed to assess the degree of agreement of chemical positive or negative response between the NR assay and in vivo evaluation, or the percentage of cell viability from the NR assay that matched the calls from those in the literature. In addition, sensitivity (%) was calculated based on the formula: 100×(the proportion of chemicals with a positive result in an NR assay that were positive based on the literature calls). Similarly, specificity (%) was derived from the formula: 100×(the proportion of chemicals with a negative result in the NR assay that were negative based on the literature).
All data obtained from the HCS Studio™2.0 BioApplication were exported, and further analysis was conducted using the JMP statistical analysis package (SAS Institute, NC). The parameters from the single-cell based imaging were quantified, and the geometric mean for each well was determined. Data were presented as geometric mean±standard deviation (SD). Statistical significance was determined using one-way ANOVA followed by a Tukey-Kramer all pairs comparison. The cell viabilities were calculated as the arithmetic mean percentages of treated versus the respective control. The data represented the average±standard deviation of five replicates. IC20, IC50, and IC85 values were calculated using the logistic 4P models in the Sigmoid Curves fit in the JMP. Non-supervised Hierarchical Clustering analyses of dose-response curves (IC20, IC50 and IC85) of tested compounds were used to determine the binary category based on the average linkage and elucidation distance correlation coefficients using MeV software (Chittenden et al., 2012). Correlation of IC50 values against the published in vivo reproductive LOAEL data was calculated. The degree of correlation was examined based on the r value and the regression coefficient (R2). The Pearson correlation coefficient between IC50 values and rLOAEL were also calculated. Statistical analysis was conducted using JMP software (SAS Inc. Cary, N.C.).
We established a testicular co-culture model using spermatogonial, Sertoli, and Leydig cells, based on the cellular composition of 5-day postnatal mouse testis. The testicular cells in a defined proportion were seeded in a 96-well plate in DMEM supplemented with 5% Nu-serum with an overlay of ECM. In order to examine whether the co-cultured cells grew into biomimetic and three-dimensional structures, we examined F-actin cytoskeleton along with cell proliferation markers (
We further quantitatively characterized F-actin cytoskeleton using a Cytoskeletal Re-arrangement algorithm to demonstrate the three-dimensional structure in the co-culture.
To assess the validity of toxicity testing using this co-culture model, we applied it to examine the effect of cadmium, a known reproductive toxicant, using a multi-parametric high-content assay.
To further identify the composition of the micro-niches of the co-culture model, we applied cell-type specific protein markers to characterize the testicular cells in the co-culture. Germ cell nuclear antigen 1 (GCNA1) is a continually expressed protein specific to spermatogonial cells. Leydig cells were identified as having an antibody against steroidogenic acute regulatory protein (StAR). SOX9 is proposed to be a specific marker for Sertoli cells, but we found it also expressed in C18-4 and TM3 cells. As shown in
As the first step to validate whether this in vitro co-culture model can be used to screen testicular toxicants, we compared the cytotoxicity with known reproductive toxicants (Table 1). We applied the NR uptake assay to examine dose-dependent responses with those selected compounds in the co-culture model, as well as a single cell culture model for 24 and 48 h, including spermatogonial cells (C18), Sertoli cells, and Leydig cells.
IC50 values of the compounds treated in the co-culture or single cell culture models at 24 and 48 h are summarized in Table 2. As indicated there, IC50 values for 24 h treatment were generally higher than those for 48 h treatment. The IC50 values for Cd, ZEA, As, HEP, DES, TBBPA, TCS, HEX, and CHL were mostly ≤100 μM. The IC50 values of chemicals in the second and third groups (
Cell type-specific difference of cytotoxicity was observed by comparing IC50 values from different cell types treated with the same compound using a radar plot, a multivariate graphical method, as shown in
To compare the testicular toxicity among all tested chemicals using our model, a non-supervised two-dimensional hierarchical cluster analysis of IC20, IC50, and IC85 values was employed using the co-culture model for 48h treatment (
Correlation Between IC50 Values from in Vitro Toxicity and in Vivo Reproductive LOAEL Values
The rLOAEL values were extracted from the published literature and indicate the lowest effective doses for reproductive toxicity in vivo. Among 32 tested compounds, 17 of them had reported in vivo rLOAEL values, and the correlation between IC50 values from in vitro and reproductive LOAEL values was examined, and determined to what extent an in vitro IC50 could predict an in vivo rLOAEL. As shown in
The concordance, sensitivity, and specificity were calculated for the co-culture model. Among 32 tested compounds, BPS, BPAF, and TBBPA—the analogues of BPA, tested positive in vitro and were extrapolated to be reproductive toxicants in vivo. Since there were no available in vivo data regarding their reproductive toxicity, they were not included in the calculation. Based on the results of cluster analysis, compounds in the cluster A, B and C were defined as positive reproductive toxicants, and Cluster D was defined as negative. TCP, BA, CYC, and ME in Cluster D were negative in vitro but positive in vivo models. At 48 h, the co-culture model displayed high concordance, sensitivity, and specificity, with values of 84%, 86.21%, and 100%, respectively (Table 4).
Sertoli cells, Leydig cells, and peritubular myoid or endothelial cells all play critical roles in supporting and maintaining spermatogenesis in the testis. Damage of any type to these cells will result in testicular dysfunction. Single cell culture models of germ cells, Sertoli cells, and Leydig cells have been previously used to examine testicular toxicity (Bilinska, 1989; Chapin, et al., 1988; Gray, 1986; Griswold, 1998; Hadley, et al., 1985; Lejeune, et al., 1998; Mather, et al., 1990; Orth and Jester, 1995; Yang et al., 2003). Spermatogonial C18-4 germline cells exhibited the morphological features of spermatogonial cells and expressed germ cell-specific proteins (Hofmann, Braydich-Stolle et al. 2005). This cell line was used as an in vitro cell model to evaluate testicular toxicity (Hofmann, et al., 2005a; Hofmann, et al., 2005b; Kokkinaki et al., 2009; Liang, et al., 2017; Oatley and Brinster, 2008), determine testicular signaling pathways (Golestaneh et al., 2009; He et al., 2008; Zhang et al., 2013), and characterize the molecular mechanisms of reproductive toxicity of nanoparticles (Braydich-Stolle et al., 2005; Braydich-Stolle et al., 2010; Lucas et al., 2012). In our previous study, we developed automated multi-parametric high-content analysis (HCA) using this cell line as an in vitro model to examine the effects of Bisphenol A and its analogs on changes of cell cycle, DNA damage, and cytoskeleton. The testis has a diverse cell population, which researchers are increasingly finding useful in generating the in vitro models needed to capture the complexity of in vivo conditions. Therefore, the reconstituted co-culture models with different somatic and germ cells have become increasingly important. Primary testicular cell co-culture models have been used to evaluate testicular abnormalities during development, and have been able to identify the testicular toxicity of phthalates. However, the disadvantage of the primary testicular cell co-culture model is in employing animals for the isolation of testicular cells, and the complicated isolation procedure leads to inconsistent results. Therefore, in this study, we developed an in vitro testicular co-culture model from rodent testicular cell lines, including spermatogonial cells, Sertoli cells, and Leydig cells with specified cell density and Extracellular Matrix (ECM) composition.
Cytoskeletal proteins are known to have numerous roles, such as determination of cell shape, cell motility, maintenance of cell junctions, and intracellular trafficking to maintain normal function and morphology (Fletcher and Mullins, 2010). Spermatogenesis is a complicated process, resulting in the production of mature sperm from primordial germ cells. During spermatogenesis, significant structural and biochemical changes take place in the seminiferous epithelium of the testis, and testis-specific actin cytoskeleton plays an important role in the acquisition of mature sperm functionality during spermatogenesis and motility during fertilization (Lie et al., 2012; Lie et al., 2010). It has been shown that Sertoli cells promote the development of germ cells (Griswold, 1998; Miryounesi et al., 2013), and co-culture of germ cells with Sertoli cells in vitro could induce germ cell differentiation. Through qualitative and quantitative comparison of F-actin cytoskeletal structure between the co-culture model and single cell culture models, we found that addition of Sertoli and Leydig cells to C18-4 spermatogonia cells significantly altered their in vitro cellular structures (
As the first step to validate this in vitro co-culture model, we selected 32 compounds and applied a simple neutral red uptake assay to examine whether this in vitro co-culture model could identify the reproductive toxicants. The NR uptake assay is one of the most sensitive and reliable cytotoxicity tests (Ceridono et al., 2012; Repetto et al., 2008a). In addition, we compared the cytotoxicity from the individual cell lines, including spermatogonia C18-4, TM3, and TM4, to examine whether the in vitro co-culture model was more closely correlated with the in vivo results. We utilized the in vivo animal studies conducted by NIEHS/NTP, in which a group of experts reviewed the R/D toxicity of chemicals and identified 45 compounds as reproductive and developmental toxicants (Moorman, et al., 2000). We also drew upon the EPA's ToxCast program, which compiled the Toxicity Reference Database (ToxRefDB) including thousands of studies using a standardized approach (Auerbach, et al., 2016; Karmaus, et al., 2016; Kavlock, et al., 2012; Paul Friedman, et al., 2016). Among the 32 compounds selected for testing in the present study, 24 were confirmed in vivo reproductive toxicants and 4 of them were confirmed negative controls. Additionally, three compounds that are structure analogs of BPA, BPAF, BPS and TBBPA, were also included without in vivo information regarding their reproductive toxicity.
Using non-supervised cluster analysis of cytotoxicity, we found that the in vitro co-culture model was able to discriminate compounds into distinguishing clusters and allow the ranking of chemical toxicity. Compared with published in vivo studies, our in vitro results were consistent overall, with in vivo toxicity data with a concordance of 86.2% and specificity of 83.3%. The reproductive LOAEL values of Cd, ZEA, As, DES, and HEP in rats were 0.088, 1, 8, 10, and 3 mg/kg/day, respectively (Martin, et al., 2011; Seiler and Spielmann, 2011). DEP, DOTP, and DMP—the least toxic group—are well known as developmental non-toxic phthalates (Gray et al., 2000; Yu, et al., 2009). SAC is a sweetener and was used as a negative reference compound. By comparing the current model with our previous primary cells co-culture model (Yu, et al., 2009), we found that both our cell line model and the primary co-culture model can distinguish the toxic phthalates (DEHP, DBP, BBP, and DPP) from the non-toxic phthalates (DEP, DMP and DOTP). We observed a higher cell death in the current cell line model than that of the previous primary cell model at the same concentrations for cadmium treatment (Yu, et al., 2005). Cluster analysis can help to predict the toxicity level of new chemicals based on the cluster in which the chemical resides. TBBPA, BPS, and BPAF are analogues of BPA, and emerged as alternatives for BPA. So far, there is still insufficient in vivo toxicological data regarding their reproductive toxicity. Based on the results of their positions in the cluster analysis, as well as their structure similarity to BPA, these compounds are predicted to be reproductive toxicants. This prediction was reinforced by the current ongoing in vivo study from the National Toxicology Program (NTP), which showed that BPAF exposure uniquely impaired pregnancies and sexual development in rats (Sutherland. et al., 2017).
Through comparing the cytotoxicity data from the co-culture model and the single cell culture models, we found the co-culture model had the highest correlation with the in vivo data. Our results suggest that the co-culture model thus offered a better predictive power than the single cell culture models. Different cell types displayed different sensitivities to the same chemical, as indicated by IC50 values (Table 2 and
Although our current in vitro co-culture provided valuable information on the potential toxicity of chemicals, it also demonstrated the limitations commonly shared by in vitro cell based assays. For example, chemicals that require metabolic activation, such as TCP, BA, CYC, and ME, will not be predicted as reproductive toxicants using this model. ME has been reported to induce testicular atrophy and disrupt spermatogenesis via the metabolism of Ethylene glycol monomethyl ether (EGME) (Foster et al., 1984; Foster et al., 1983; Starek-Swiechowicz et al., 2015; van der Laan et al., 2012). It was found that ME undergoes metabolic activation to appropriate methoxyacetic acid (MAA) via EGME (Takei et al., 2010). MAA primarily affects tissues with rapidly dividing cells and high rates of energy metabolism in the testes, leading to apoptosis of primary spermatocytes. Spermatogenesis is a multi-step complex process. Our current in vitro co-culture model captured an early stage of spermatogenesis and will probably not capture all the toxicologically-vulnerable processes in the testis. Chemicals such as boric acid would likely cause male reproductive toxicity through nonmolecular interactions, and lead to damage of developing spermatids (Chapin and Ku, 1994; Jewell et al., 1998). Similarly, VIN treatment from gestation day 15 to postnatal week 4 at the concentration of 7.2 and 72 mg/kg/day has induced abnormal spermatozoa with nuclear and acrosomal defects (Veeramachaneni et al., 2006). Therefore, further efforts are needed in order to differentiate spermatogonia in the co-culture model to produce post-meiotic spermatids as observed in explanted pieces of testis (Brannen, et al., 2016; Sato, et al., 2011). Thus, it is unlikely that any individual in vitro model is sufficient as a final decision point for male reproductive toxicity; rather, a tiered approach using a series of models containing multiple endpoint analysis and concentration response curves is essential to building a robust screening platform to improve the prediction accuracy of in vitro assays.
Animal studies could provide credible evidence to predict the likely effect of chemical exposure on human outcomes. However, the toxicity testing from the animal studies predicted toxicity in humans with only about 50-70% accuracy (Chapin et al., 2013; Olson et al., 2000). Our current results represent an improvement over previous attempts to predict reproductive toxicity responses. The doses selected for these compounds were initially determined based on the literature and adjusted to ensure derivation of the IC values from the dose-response curves (
In summary, by utilizing testicular cell lines we constructed a testicular cell co-culture model, demonstrated the formation of a three-dimensional cytoskeleton structure, and were able to distinguish testicular toxic compounds from non-toxic chemicals. Moreover, the toxicity in the co-culture model at 48 h was found to have the highest correlation with rLOAEL in vivo. The calculation of concordance, sensitivity, and specificity further supported the reliability of this model. Our results suggest that our in vitro co-culture model may be useful in screening testicular toxicants in a wide concentration range and prioritizing chemicals for further assessment. Furthermore, the exploitation of high-content imaging and quantitative techniques provides deep insight into the molecular and cellular mechanisms. In future research, we will include more compounds to further validate this in vitro co-culture model, and establish and examine more endpoints that correlate with different adverse outcomes of in vivo reproductive toxicity.
Auerbach, S., Filer, D., Reif, D., Walker, V., Holloway, A. C., Schlezinger, J., Srinivasan, S., Svoboda, D., Judson, R., Bucher, J. R., et al. (2016). Prioritizing Environmental Chemicals for Obesity and Diabetes Outcomes Research: A Screening Approach Using ToxCast High Throughput Data. Environ Health Perspect doi: 10.1289/ehp.1510456.
Bilinska, B. (1989). Interaction between Leydig and Sertoli cells in vitro. Cytobios 60(241), 115-26.
Brannen, K. C., Chapin, R. E., Jacobs, A. C., and Green, M. L. (2016). Alternative Models of Developmental and Reproductive Toxicity in Pharmaceutical Risk Assessment and the 3Rs. ILAR J 57(2), 144-156.
Braydich-Stolle, L., Hussain, S., Schlager, J. J., and Hofmann, M. C. (2005). In vitro cytotoxicity of nanoparticles in mammalian germline stem cells. Toxicological Sciences 88(2), 412-419.
Braydich-Stolle, L. K., Lucas, B., Schrand, A., Murdock, R. C., Lee, T., Schlager, J. J., Hussain, S. M., and Hofmann, M. C. (2010). Silver Nanoparticles Disrupt GDNF/Fyn kinase Signaling in Spermatogonial Stem Cells. Toxicological Sciences 116(2), 577-589.
Casey, W. M. (2016). Advances in the Development and Validation of Test Methods in the United States. Toxicol Res 32(1), 9-14.
Ceridono, M., Tellner, P., Bauer, D., Barroso, J., Alépée, N., Corvi, R., De Smedt, A., Fellows, M. D., Gibbs, N. K., Heisler, E., et al. (2012). The 3T3 neutral red uptake phototoxicity test: Practical experience and implications for phototoxicity testing—The report of an ECVAM—EFPIA workshop. Regulatory Toxicology and Pharmacology 63(3), 480-488.
Chapin, R. E., Boekelheide, K., Cortvrindt, R., van Duursen, M. B. M., Gant, T., Jegou, B., Marczylo, E., van Pelt, A. M. M., Post, J. N., Roelofs, M. J. E., et al. (2013). Assuring safety without animal testing: The case for the human testis in vitro. Reproductive Toxicology 39, 63-68.
Chapin, R. E., Gray, T. J., Phelps, J. L., and Dutton, S. L. (1988). The effects of mono-(2-ethylhexyl)-phthalate on rat Sertoli cell-enriched primary cultures. Toxicol Appl Pharmacol 92(3), 467-79.
Chapin, R. E., and Ku, W. W. (1994). The reproductive toxicity of boric acid. Environ Health Perspect 102 Suppl 7, 87-91.
Chapin, R. E., Phelps, J. L., Burka, L. T., Abou-Donia, M. B., and Heindel, J. J. (1991). The effects of tri-o-cresyl phosphate and metabolites on rat Sertoli cell function in primary culture. Toxicol Appl Pharmacol 108(2), 194-204.
Chapin, R. E., and Stedman, D. B. (2009). Endless possibilities: stem cells and the vision for toxicology testing in the 21st century. Toxicological sciences : an official journal of the Society of Toxicology 112(1), 17-22.
Chittenden, T. W., Howe, E. A., Taylor, J. M., Mar, J. C., Aryee, M. J., Gomez, H., Sultana, R., Braisted, J., Nair, S. J., Quackenbush, J., et al. (2012). nEASE: a method for gene ontology subclassification of high-throughput gene expression data. Bioinformatics 28(5), 726-8.
CIRM (2008). Stem Cells in Predictive Toxicology CIRM Workshop Report.
Clark, B. J., Wells, J., King, S. R., and Stocco, D. M. (1994). The purification, cloning, and expression of a novel luteinizing hormone-induced mitochondrial protein in MA-10 mouse Leydig tumor cells. Characterization of the steroidogenic acute regulatory protein (StAR). J Biol Chem 269(45), 28314-22.
Enders, G. C., and May, J. J., 2nd (1994). Developmentally regulated expression of a mouse germ cell nuclear antigen examined from embryonic day 11 to adult in male and female mice. Dev Biol 163(2), 331-40.
Erdos, Z., Pearson, K., Goedken, M., Menzel, K., Sistare, F. D., Glaab, W. E., and Saldutti, L. P. (2013). Inhibin B response to testicular toxicants hexachlorophene, ethane dimethane sulfonate, di-(n-butyl)-phthalate, nitrofurazone, DL-ethionine, 17-alpha ethinylestradiol, 2,5-hexanedione, or carbendazim following short-term dosing in male rats. Birth defects research. Part B, Developmental and reproductive toxicology 98(1), 41-53.
Fletcher, D. A., and Mullins, R. D. (2010). Cell mechanics and the cytoskeleton. Nature 463(7280), 485-92.
Forgacs, A. L., Ding, Q., Jaremba, R. G., Huhtaniemi, I. T., Rahman, N. A., and Zacharewski, T. R. (2012). BLTK1 murine Leydig cells: a novel steroidogenic model for evaluating the effects of reproductive and developmental toxicants. Toxicological sciences : an official journal of the Society of Toxicology 127(2), 391-402.
Foster, P. M., Creasy, D. M., Foster, J. R., and Gray, T. J. (1984). Testicular toxicity produced by ethylene glycol monomethyl and monoethyl ethers in the rat. Environ Health Perspect 57, 207-17.
Foster, P. M., Creasy, D. M., Foster, J. R., Thomas, L. V., Cook, M. W., and Gangolli, S. D. (1983). Testicular toxicity of ethylene glycol monomethyl and monoethyl ethers in the rat. Toxicol Appl Pharmacol 69(3), 385-99.
Golestaneh, N., Beauchamp, E., Fallen, S., Kokkinaki, M., Uren, A., and Dym, M. (2009). Wnt signaling promotes proliferation and stemness regulation of spermatogonial stem/progenitor cells. Reproduction 138(1), 151-162.
Gray, L. E., Jr., Ostby, J., Furr, J., Price, M., Veeramachaneni, D. N., and Parks, L. (2000). Perinatal exposure to the phthalates DEHP, BBP, and DINP, but not DEP, DMP, or DOTP, alters sexual differentiation of the male rat. Toxicological sciences : an official journal of the Society of Toxicology 58(2), 350-65.
Gray, T. J. (1986). Testicular toxicity in vitro: Sertoli-germ cell co-cultures as a model system. Food Chem Toxicol 24(6-7), 601-5.
Griswold, M. D. (1998). The central role of Sertoli cells in spermatogenesis. Semin Cell Dev Biol 9(4), 411-6.
Hadley, M. A., Byers, S. W., Suarez-Quian, C. A., Kleinman, H. K., and Dym, M. (1985). Extracellular matrix regulates Sertoli cell differentiation, testicular cord formation, and germ cell development in vitro. J Cell Biol 101(4), 1511-22.
Hareng, L., Pellizzer, C., Bremer, S., Schwarz, M., and Hartung, T. (2005). The integrated project ReProTect: a novel approach in reproductive toxicity hazard assessment. Reprod Toxicol 20(3), 441-52.
Harris, S., Hermsen, S. A., Yu, X., Hong, S. W., and Faustman, E. M. (2015). Comparison of toxicogenomic responses to phthalate ester exposure in an organotypic testis co-culture model and responses observed in vivo. Reprod Toxicol 58, 149-59.
Hartung, T., and Rovida, C. (2009). Chemical regulators have overreached. Nature 460(7259), 1080-1.
He, Z. P., Jiang, J. J., Kokkinaki, M., Golestaneh, N., Hofmann, M. C., and Dym, M. (2008). GDNF upregulates c-fos transcription via the Ras/ERK1/2 pathway to promote mouse spermatogonial stem cell proliferation. Stem Cells 26(1), 266-278.
Hofmann, M. C., Braydich-Stolle, L., Dettin, L., Johnson, E., and Dym, M. (2005a). Immortalization of mouse germ line stem cells. Stem Cells 23(2), 200-10.
Hofmann, M. C., Braydich-Stolle, L., and Dym, M. (2005b). Isolation of male germ-line stem cells; influence of GDNF. Dev Biol 279(1), 114-124.
Hofstetter, E., Spencer, J. W., Hiteshew, K., Coutu, M., and Nealley, M. (2013). Evaluation of recommended REACH exposure modeling tools and near-field, far-field model in assessing occupational exposure to toluene from spray paint. Ann Occup Hyg 57(2), 210-20.
Houck, K. A., Dix, D. J., Judson, R. S., Kavlock, R. J., Yang, J., and Berg, E. L. (2009). Profiling bioactivity of the ToxCast chemical library using BioMAP primary human cell systems. J Biomol Screen 14(9), 1054-66.
ICCVAM (2012). ICCVAM recommends in vitro test method for endocrine-disruptors. Altern Lab Anim 40(1), 11.
James, R. W., Heywood, R., and Crook, D. (1980). Quantitative aspects of spermatogenesis in rats and dogs after repeated hexachlorophene treatment. Toxicol Lett 5(6), 405-12.
Jenardhanan, P., Panneerselvam, M., and Mathur, P. P. (2016). Effect of environmental contaminants on spermatogenesis. Semin Cell Dev Biol 59, 126-140.
Jewell, W. T., Hess, R. A., and Miller, M. G. (1998). Testicular toxicity of molinate in the rat: metabolic activation via sulfoxidation. Toxicol Appl Pharmacol 149(2), 159-66.
Karmaus, A. L., Filer, D. L., Martin, M. T., and Houck, K. A. (2016). Evaluation of food-relevant chemicals in the ToxCast high-throughput screening program. Food and Chemical Toxicology 92, 188-96.
Kavlock, R., Chandler, K., Houck, K., Hunter, S., Judson, R., Kleinstreuer, N., Knudsen, T., Martin, M., Padilla, S., Reif, D., et al. (2012). Update on EPA's ToxCast Program: Providing High Throughput Decision Support Tools for Chemical Risk Management. Chemical Research in Toxicology 25(7), 1287-1302.
Kokkinaki, M., Lee, T. L., He, Z. P., Jiang, J. J., Golestaneh, N., Hofmann, M. C., Chan, W. Y., and Dym, M. (2009). The Molecular Signature of Spermatogonial Stem/Progenitor Cells in the 6-Day-Old Mouse Testis. Biology of Reproduction 80(4), 707-717.
Lejeune, H., Sanchez, P., and Saez, J. M. (1998). Enhancement of long-term testosterone secretion and steroidogenic enzyme expression in human Leydig cells by co-culture with human Sertoli cell-enriched preparations. Int J Androl 21(3), 129-40.
Leung, M. C., Phuong, J., Baker, N. C., Sipes, N. S., Klinefelter, G. R., Martin, M. T., McLaurin, K. W., Setzer, R. W., Darney, S. P., Judson, R. S., et al. (2016). Systems Toxicology of Male Reproductive Development: Profiling 774 Chemicals for Molecular Targets and Adverse Outcomes. Environ Health Perspect 124(7), 1050-61.
Liang, S., Yin, L., Shengyang Yu, K., Hofmann, M. C., and Yu, X. (2017). High-Content Analysis Provides Mechanistic Insights into the Testicular Toxicity of Bisphenol A and Selected Analogues in Mouse Spermatogonial Cells. Toxicol Sci 155(1), 43-60.
Liang, S., Yin, L., Shengyang Yu, K., Hofmann, M. C., and Yu, X. (2016). High-Content Analysis Provides Mechanistic Insights into the Testicular Toxicity of Bisphenol A and Selected Analogues in Mouse Spermatogonial Cells. Toxicological sciences: an official journal of the Society of Toxicology doi: 10.1093/toxsci/kfw178.
Lie, P. P., Cheng, C. Y., and Mruk, D. D. (2012). The biology of interleukin-1: emerging concepts in the regulation of the actin cytoskeleton and cell junction dynamics. Cell Mol Life Sci 69(4), 487-500.
Lie, P. P., Mruk, D. D., Lee, W. M., and Cheng, C. Y. (2010). Cytoskeletal dynamics and spermatogenesis. Philos Trans R Soc Lond B Biol Sci 365(1546), 1581-92.
Lucas, B. E., Fields, C., Joshi, N., and Hofmann, M. C. (2012). Mono-(2-ethylhexyl)-phthalate (MEHP) affects ERK-dependent GDNF signalling in mouse stem-progenitor spermatogonia. Toxicology 299(1), 10-9.
Luijten, M., de Vries, A., Opperhuizen, A., and Piersma, A. H. (2007). Alternative methods in reproductive toxicity testing: state of the art. RIVM report, 340720002.
Martin, M. T., Knudsen, T. B., Reif, D. M., Houck, K. A., Judson, R. S., Kavlock, R. J., and Dix, D. J. (2011). Predictive model of rat reproductive toxicity from ToxCast high throughput screening. Biol Reprod 85(2), 327-39.
Mather, J. P. (1980). Establishment and characterization of two distinct mouse testicular epithelial cell lines. Biol Reprod 23(1), 243-52.
Mather, J. P., Attie, K. M., Woodruff, T. K., Rice, G. C., and Phillips, D. M. (1990). Activin stimulates spermatogonial proliferation in germ-Sertoli cell cocultures from immature rat testis. Endocrinology 127(6), 3206-14.
Mather, J. P., and Phillips, D. M. (1984). Establishment of a peritubular myoid-like cell line and interactions between established testicular cell lines in culture. J Ultrastruct Res 87(3), 263-74.
Miryounesi, M., Nayernia, K., Dianatpour, M., Mansouri, F., and Modarressi, M. H. (2013). Co-culture of Mouse Embryonic Stem Cells with Sertoli Cells Promote in vitro Generation of Germ Cells. Iran J Basic Med Sci 16(6), 779-83.
Moorman, W. J., Ahlers, H. W., Chapin, R. E., Daston, G. P., Foster, P. M., Kavlock, R. J., Morawetz, J. S., Schnorr, T. M., and Schrader, S. M. (2000). Prioritization of NTP reproductive toxicants for field studies. Reprod Toxicol 14(4), 293-301.
Oatley, J. M., and Brinster, R. L. (2008). Regulation of spermatogonial stem cell self-renewal in mammals. Annu Rev Cell Dev Biol 24, 263-86.
OECD (2016). OECD Guidelines for the Testing of Chemicals.
Olson, H., Betton, G., Robinson, D., Thomas, K., Monro, A., Kolaja, G., Lilly, P., Sanders, J., Sipes, G., Bracken, W., et al. (2000). Concordance of the toxicity of pharmaceuticals in humans and in animals. Regul Toxicol Pharmacol 32(1), 56-67.
Orth, J. M., and Jester, W. F., Jr. (1995). NCAM mediates adhesion between gonocytes and Sertoli cells in cocultures from testes of neonatal rats. J Androl 16(5), 389-99.
Parks Saldutti, L., Beyer, B. K., Breslin, W., Brown, T. R., Chapin, R. E., Campion, S., Enright, B., Faustman, E., Foster, P. M., Hartung, T., et al. (2013). In vitro testicular toxicity models: Opportunities for advancement via biomedical engineering techniques. ALTEX 30(3), 353-77.
Paul Friedman, K., Watt, E. D., Hornung, M. W., Hedge, J. M., Judson, R. S., Crofton, K. M., Houck, K. A., and Simmons, S. O. (2016). Tiered High-Throughput Screening Approach to Identify Thyroperoxidase Inhibitors Within the ToxCast Phase I and II Chemical Libraries. Toxicological Sciences 151(1), 160-80.
Reif, D. M., Martin, M. T., Tan, S. W., Houck, K. A., Judson, R. S., Richard, A. M., Knudsen, T. B., Dix, D. J., and Kavlock, R. J. (2010). Endocrine profiling and prioritization of environmental chemicals using ToxCast data. Environ Health Perspect 118(12), 1714-20.
Repetto, G., del Peso, A., and Zurita, J. L. (2008a). Neutral red uptake assay for the estimation of cell viability/cytotoxicity. Nature protocols 3(7), 1125-31.
Repetto, G., del Peso, A., and Zurita, J. L. (2008b). Neutral red uptake assay for the estimation of cell viability/cytotoxicity. Nature Protocols 3(7), 1125-1131.
Sato, T., Katagiri, K., Gohbara, A., Inoue, K., Ogonuki, N., Ogura, A., Kubota, Y., and Ogawa, T. (2011). In vitro production of functional sperm in cultured neonatal mouse testes. Nature 471(7339), 504-7.
Sauer, U. G. (2004). Avoidance of animal experiments in the new EU Chemicals Regulation—opportunities and problems from the point of view of animal welfare. Altex 21(1), 9-14.
Scialli, A. R., and Guikema, A. J. (2012). REACH and reproductive and developmental toxicology: still questions. Syst Biol Reprod Med 58(1), 63-9.
Seiler, A. E., and Spielmann, H. (2011). The validated embryonic stem cell test to predict embryotoxicity in vitro. Nat Protoc 6(7), 961-78.
Spade, D. J., Hall, S. J., Saffarini, C. M., Huse, S. M., McDonnell, E. V., and Boekelheide, K. (2014a). Differential response to abiraterone acetate and di-n-butyl phthalate in an androgen-sensitive human fetal testis xenograft bioassay. Toxicological sciences: an official journal of the Society of Toxicology 138(1), 148-60.
Spade, D. J., McDonnell, E. V., Heger, N. E., Sanders, J. A., Saffarini, C. M., Gruppuso, P. A., De Paepe, M. E., and Boekelheide, K. (2014b). Xenotransplantation models to study the effects of toxicants on human fetal tissues. Birth defects research. Part B, Developmental and reproductive toxicology 101(6), 410-22.
Starek-Swiechowicz, B., Szymczak, W., Budziszewska, B., and Starek, A. (2015). Testicular effect of a mixture of 2-methoxyethanol and 2-ethoxyethanol in rats. Pharmacological reports: PR 67(2), 289-93.
Sutherland., V., McIntyre., B., Pelch., K., Waidyanatha., S., Conley., J. M., Gray., L. E., and Foster, P. M. A Comparison of In Vivo Reproductive and Developmental Toxicity (DART) Endpoints for Bisphenol AF and Bisphenol A2017, Baltimore, Md.
Takei, M., Ando, Y., Saitoh, W., Tanimoto, T., Kiyosawa, N., Manabe, S., Sanbuissho, A., Okazaki, O., Iwabuchi, H., Yamoto, T., et al. (2010). Ethylene glycol monomethyl ether-induced toxicity is mediated through the inhibition of flavoprotein dehydrogenase enzyme family. Toxicological sciences: an official journal of the Society of Toxicology 118(2), 643-52.
ThermoFisher (2010). Thermo Scientific Cellomics®Morphology Explorer V4 BioApplication Guide. In (Thermo Fisher Scientific Inc. 100 Technology Drive Pittsburgh, Penn. 15219.
van der Laan, J. W., Chapin, R. E., Haenen, B., Jacobs, A. C., and Piersma, A. (2012). Testing strategies for embryo-fetal toxicity of human pharmaceuticals. Animal models vs. in vitro approaches: a workshop report. Regul Toxicol Pharmacol 63(1), 115-23.
Veeramachaneni, D. N., Palmer, J. S., Amann, R. P., Kane, C. M., Higuchi, T. T., and Pau, K. Y. (2006). Disruption of sexual function, FSH secretion, and spermiogenesis in rabbits following developmental exposure to vinclozolin, a fungicide. Reproduction 131(4), 805-16.
Wegner, S., Hong, S., Yu, X., and Faustman, E. M. (2013). Preparation of rodent testis co-cultures. Current protocols in toxicology/editorial board, Mahin D. Maines Chapter 16, Unit 16 10.
Wegner, S. H., Yu, X., Kim, H., Harris, S., Griffith, W. C., Hong, S. H., and Faustman, E. M. (2014). Effect of dipentyl phthalate in 3-dimensional in vitro testis co-culture is attenuated by cyclooxygenase-2 inhibition. JTEHS 6(8), 161-169.
Yang, J. M., Arnush, M., Chen, Q. Y., Wu, X. D., Pang, B., and Jiang, X. Z. (2003). Cadmium-induced damage to primary cultures of rat Leydig cells. Reprod Toxicol 17(5), 553-60.
Yu, X., Hong, S., and Faustman, E. M. (2008). Cadmium-induced activation of stress signaling pathways, disruption of ubiquitin-dependent protein degradation and apoptosis in primary rat Sertoli cell-gonocyte cocultures. Toxicological sciences: an official journal of the Society of Toxicology 104(2), 385-96.
Yu, X., Hong, S., Moreira, E. G., and Faustman, E. M. (2009). Improving in vitro Sertoli cell/gonocyte co-culture model for assessing male reproductive toxicity: Lessons learned from comparisons of cytotoxicity versus genomic responses to phthalates. Toxicol Appl Pharmacol 239(3), 325-36.
Yu, X., Sidhu, J. S., Hong, S., and Faustman, E. M. (2005). Essential role of extracellular matrix (ECM) overlay in establishing the functional integrity of primary neonatal rat Sertoli cell/gonocyte co-cultures: an improved in vitro model for assessment of male reproductive toxicity. Toxicological sciences : an official journal of the Society of Toxicology 84(2), 378-93.
Zhang, Z. Z., Gong, Y. H., Guo, Y., Hai, Y. N., Yang, H., Yang, S., Liu, Y., Ma, M., Liu, L. H., Li, Z., et al. (2013). Direct transdifferentiation of spermatogonial stem cells to morphological, phenotypic and functional hepatocyte-like cells via the ERK1/2 and Smad2/3 signaling pathways and the inactivation of cyclin A, cyclin B and cyclin E. Cell Commun Signal 11.
High-content analysis (HCA) has emerged as a powerful tool for chemical toxicity profiling. A multi-parametric HCA in a spermatogonial cell line was previously developed to examine the testicular toxicities of Bisphenol A (BPA) and its Analogs: bisphenol S (BPS), bisphenol AF (BPAF), and tetrabromobisphenol A (TBBPA). Due to the complexity of high-dimensional and large-scale HCA data, there is an increased demand for effective computational strategies to characterize and quantify phenotypic effects at a single-cell level. Here we describe the development of a machine learning (ML)-based HCA pipeline and explore the complex phenotypic changes. This pipeline allowed us to characterize the toxicity of BPA and its Analogs in a testicular co-culture model using spermatogonial, Leydig and Sertoli cell lines. The use of a
ML-based phenotypic analysis allowed us to observe that the treatments of BPA or its Analogs resulted in the loss of spatial arrangement of the model's three-dimensional (3D) structure and an accumulation of cells in M phase arrest in a dose- and time-dependent manner. Furthermore, BPAF induced an accumulation of multinucleated cells, which was associated with an increase of DNA damage response, and impairment in cellular actin filaments. These results showed that BPAF and TBBPA exerted toxicity at lower doses than BPA and BPS on multiple endpoints in the co-culture model. In summary, we have developed a ML-based HCA approach in the testicular cell co-culture model that reflected complex phenotypic changes and characterized the testicular toxicities of BPA and its Analogs. This approach provided an in-depth analysis of multi-dimensional HCA data and unbiased quantitative analysis of the phenotypes of interest.
With recent advances in automated fluorescence microscopy and the development of quantitative image analysis software, high-content analysis (HCA) enables measurements of unbiased multi-parametric data on a single-cell level and provides both temporal and spatial measurements of various cellular events associated with adverse health outcomes (Buchser et al., 2004; Mattiazzi Usaj et al., 2016; Zanella et al., 2010). This approach has been used to prioritize chemical toxicity for further studies, to characterize adverse outcome pathways, and to develop predictive models for toxicity evaluation in humans (Elmore et al., 2014; Merrick et al., 2015; Shukla et al., 2010). The U.S. Environmental Protection Agency (EPA) initiated the ToxCast program using in vitro high-throughput and high-content screening (HTS and HCS) to profile and predict the toxicity of thousands of environmental chemicals (Krewski et al., 2010; Martin et al., 2011; Paul Friedman et al., 2016). While HCA provides large-scale, image-based data, the analysis of these data becomes a major bottleneck as the quantification of cellular phenotypes usually depends on manual parameter adjustments, which is both time- and labor-intensive and has low reproducibility among different experiments (Sommer and Gerlich, 2013). Thus, the demand for advanced computational strategies that explore the inherent structure of multidimensional data, and provide an unbiased assessment of a variety of phenotypes in large-scale image data sets. Supervised machine learning (ML) has emerged as a powerful approach in classifying cellular heterogeneity using non-linear multi-parametric algorithms for HCA data (Altschuler and Wu, 2010; Sommer and Gerlich, 2013). These algorithms are then able to learn from the small training datasets labeled with predefined classes and then automatically infer the rules to classify full datasets.. This application of machine learning has been applied in the examiniation of the dynamic changes of the genome and the proteome in single cell imaging (Chong et al., 2015; Neumann et al., 2010), its application in toxicology, however, has not fully been explored.
In a previous study, our lab established and validated a battery of HCA assays using spermatogonial cell line C18-4 and revealed the different testicular toxicities of Bisphenol A (BPA) and its Analogs: bisphenol S (BPS), bisphenol AF (BPAF) and tetrabromobisphenol A (TBBPA) (Liang et al., 2017b). BPA is a high production-volume chemical widely used in consumer products, thermal paper, medical devices, and dental sealants (Rochester, 2013). Exposure to BPA is widespread and occurs mainly through ingestion, inhalation and dermal exposures (Kang et al., 2006; Vandenberg et al., 2007). BPA has been detected in over 90% of urine samples from the general population in the United States. (Calafat et al., 2008; Lakind and Naiman, 2011). BPA is a well-studied endocrine disruptor, as it has been determined as a reproductive and developmental toxicant in animal models (Peretz et al., 2014; Richter et al., 2007; Rochester, 2013; vom Saal et al., 2007). Due to the concern of ubiquitous exposure to BPA and its potential adverse effects on humans, the U.S. Food, and Drug Administration (FDA) and European Chemical Agency (ECHA) have placed restrictions on the use of BPA (EU., 2016; FDA., 2013; FDA., 2012). The structural Analogs of BPA have been introduced into the market as a BPA substitute and share similar manufacturing applications to BPA. Although there is a general lack of production data for BPA Analogs, the usage of these chemicals is expected to rise globally. Due to high structural similarities, the Analogs could potentially exhibit estrogenic potencies and reproductive toxicities comparable with BPA. Emerging evidence indicates that BPA Analogs have been found in food and human urine samples and interacted with various physiological receptors (Driffield et al., 2008; Kitamura et al., 2005; Liao and Kannan, 2013; Liao et al., 2012; Stossi et al., 2014). However, toxicological data concerning BPA Analogs are still limited.
In this study, we applied a testicular cell co-culture model to examine the testicular toxicities of BPA and its Analogs, which provided a more closely approximated in vivo testicular microenvironment encountered with environmental exposures. The testicular cell co-culture model was able to recapitulate the multicellular complexity and organ-like structure, and mimic the physiology relevant to vivo. Sertoli and Leydig cells play critical roles in regulating and supporting spermatogenesis and maintaining the structure and functions of the testis (Haywood et al., 2003; Lui et al., 2003; Payne, 1990). The primary testicular cell co-culture model employed germ, Sertoli, and Leydig cells with an extracellular matrix (ECM) was reported to exhibit testicle-like multilayered architecture. This model demonstrated the differential metabolic capacity, inflammatory responses, and differential toxicogenomic responses to phthalate exposures (Harris et al., 2015; Harris et al., 2016a; Harris et al., 2016b; Yu et al., 2009; Yu et al., 2005). Recently, we developed an animal-free, in vitro testicular cell co-culture model utilizing mouse spermatogonial (C18-4), Sertoli (TM4), and Leydig (TM3) cell lines, which exhibited a unique three-dimensional (3D) in vivo structure compared to single cell cultures, and enabled us to classify the reproductive toxic substances with high specificity and sensitivity (Example 1).
The purpose of this study was to develop a supervised ML-based HCA to characterize phenotypes associated with the testicular toxicities of BPA and its structural Analogs: BPS, BPAF, and TBBPA in a testicular cell co-culture model. We measured a wide spectrum of adverse endpoints, which characterized individual cells from different phenotypes in testicular cells, including nuclear morphology, DNA synthesis, DNA damage, and cytoskeleton structure using the CellProfiler Analyst. Our lab has developed a phenotype recognition pipeline, comprised of image pre-processing, object detection, feature extraction, training data, and classification. We found that BPA and its Analogs induced phenotypic 3D structural changes and M phase arrest in a dose-dependent manner. BPAF induced an accumulation of multinucleated cells, which was associated with an increase of DNA damage response, and impairment of cellular actin filaments. Overall, through the implementation of machine learning paired with image-based HCA, we have demonstrated a new and effective means of classifying multiple toxic endpoints in the co-culture model treated with BPA and its Analogs. Therefore, this ML-based HCA approach provided an in-depth analysis of high-content datasets, powered up image-based multivariate data, and allowed rapid and objective high-throughput screening for future environmental toxicity testing.
Dulbecco's Modified Eagle Medium (DMEM), Modified Eagle's Medium/Nutrient Mixture F-12 (DME/F12), horse serum, and penicillin-streptomycin were purchased from GE Healthcare Life Sciences (Logan, Utah.). Fetal bovine serum (FBS), 4,4′-(propane-2,2-diyl) diphenol (BPA, ≥99%), 4,4′-sulfonyldiphenol (BPS, 98%), 2,2′,6,6′-Tetrabromo-4,4′-isopropylidenediphenol (TBBPA, 97%), and neutral red (NR) were purchased from Sigma-Aldrich (St Louis, Mo.). Nu-Serum was purchased from BD BioScience (Redford, Mass.). 4-[1,1,1,3,3,3-Hexafluoro-2-(4-hydroxyphenyl)propan-2-yl]phenol (BPAF, 98%) was purchased from Alfa Aesar (Ward Hill, Mass.). 5-Bromo-2′-deoxyuridine (BrdU, 99%) was purchased from Thermo Scientific (Waltham, Mass.). 4% Paraformaldehyde was purchased from Boston Bioproducts (Ashland, Mass.).
The testicular cell co-culture model was established as reported in Example 1 Briefly, mouse spermatogonial cell line C18-4 was established via germ cells isolated from the testes of 6 day old Balb/c mice. This cell line was selected as it showed morphological features of type A spermatogonia, expressed testicular germ cell-specific genes, such as GFRA1, Dazl, Ret and stem cell-specific genes such as psiwi12 and prame11 (Hofmann et al., 2005). Mouse Leydig (TM3) and Sertoli (TM4) cell lines were purchased from ATCC, and these cells were isolated from prepubertal mouse gonads. C18-4 cells were maintained in DMEM composed of 5% FBS, and 100 U/ml streptomycin and penicillin in a 33° C., 5% CO2 humidified environment in a sub-confluent condition, and were passaged every 3-4 days TM3 and TM4 cells were cultured in DME/F12 composed of 1.25% FBS, 2.5% horse serum, and 100 U/ml streptomycin and penicillin at 37° C., 5% CO2 in a sub-confluent condition, and were passaged every 2-3 days. When the cells reached 70-80% confluence, a total of 1.5×104 cells per well were inoculated into a 96-well plate. The percentages of spermatogonial, Sertoli, and Leydig cells in the co-culture model were 80%, 15%, and 5%, respectively. The co-cultures were maintained in DMEM composed of 2.5% Nu-serum, and 100 U/ml streptomycin and penicillin in a 33° C., 5% CO2 humidified environment. After, an overnight incubation period, the co-cultures reached 100% confluence and were then treated with various doses of BPA, BPS, BPAF, and TBBPA for the indicated doses and time periods.
Cell viability was determined through the Neutral Red uptake assay, which is based on the ability of viable cells to incorporate NR dye into their lysosomes, while the dye is not retained if the cell dies (Repetto et al., 2008). The co-cultures were treated with various doses of BPA, or BPS, (25, 50, 100, 200 and 400 μM), and BPAF or TBBPA (2.5, 5, 10, 25 and 50 μM) for 24, 48, and 72 h. The vehicle controls were set as the co-cultures treated with the vehicle (0.05% DMSO) After the treatments, the culture medium was replaced by fresh medium containing NR (50 μg/ml). After a 3 h incubation period, the co-cultures were washed with phosphate buffered saline (PBS), and the NR dye was eluted with 100 μl of a 0.5% acetic acid/50% ethanol solution. The plate was then gently shaken, and absorbance values were measured at 540 nm with a Synergy HT microplate reader (BioTek, VT). Cell viability was presented as a percentage of the mean of vehicle controls after subtracting background readings.
To examine DNA synthesis, and conduct cell cycle analysis, the co-cultures were treated with various doses of BPA, or BPS (5, 10, 25, 50 and 100 μM), and BPAF or TBBPA (1, 2.5, 5, 10 and 15 μM) for 24, 48 and 72 h. BrdU incorporation, cell fixation, BrdU, and DNA staining followed the previously published protocol (Liang, et al., 2017b). the determination of DNA damage responses and cytoskeleton analysis was conducted after cell fixation. The cells were then permeabilized by a 0.1% Triton X-100 in PBS, and then incubated with a mouse-anti phosphor-histone H2AX (Ser139) (γ-H2AX) overnight at 4° C. After washing the cells with a BPS/BSA the cells were then incubated with goat anti-mouse Dylight 650, and Hoechst 33342 in a BPS/BSA solution for 90 minutes. Finally, prior to acquiring the images, the cells were stained with Alexa Fluor 488 Phallodin for 30 minutes to stain the F-actin (Liang, et al., 2017)
Multi-channel images were then automatically acquired using an Arrayscan™ VTI HCS reader (Thermo Scientific, MA) with the HCS Studio 2.0 Target Activation BioApplication module. Forty-nine fields per well were acquired at 20× and 40× magnification using a Hamamatsu ROCA-ER digital camera in combination with a 0.63× coupler and Carl Zeiss microscope optics in auto-focus mode. Image smoothing was conducted in order to reduce object fragmentation prior to the primary object identification that takes place in the first channel. Channel one (Ch1) applied the BGRFR 386_23 for Hoechst 33342 that was used for auto-focus, object identification, and segmentation. The border objects were then excluded. Channel two (Ch2) applied applied the BGRFR 549_15, which identified the BrdU staining and BGRFR 485_20 to identify the F-actin staining. BGRFR 650_13 was applied in Channel three (Ch3) and identified the γ-H2AX staining.
Multi-Channel images were analyzed using HCS Studio™2.0 TargetActivation BioApplication. Multiple parameters of nuclei were characterized in HCA, including nuclei number, nuclear area, shape, and total DNA intensity. Nuclear shape measurements included P2A, a ratio of the nuclear perimeter squared to 4it*nucleus area (perimeter2/4π*enuclear area) to evaluate nuclear smoothness, and LWR, a ratio of the nuclear length to width, to measure nuclear roundness. For a fairly round and smooth object, the values for P2A and LWR are around 1.0. Total intensity was defined as total pixel intensity within a cell in the respective channel. The total intensities of BrdU, γ-H2AX, and F-actin of the individual cells were also quantified. With forty-nine 20× images acquired from each well, at least 1,000 cells were analyzed per well, and single-cell based data were extracted for further analysis. The experiments were performed with at least four biological replicates and repeated twice.
For ML-based HCA as illustrated in
In the current study, we quantified three key phenotypic changes in the co-culture model. The first change we noticed was the induction of multinucleated cells, a unique toxicity marker for BPAF observed in spermatogonia cells in the previous study, was once again examined in the current co-culture testicular cell model. As depicted in
The data obtained from the HCS Studio·8 2.0 TargetActivation BioApplication, CellProfiler and CellProfiler Analyst were exported and further analyzed using the JMP statistical analysis package (SAS Institute, NC). To remove cell clumps, nuclei with areas larger than 1000 μM2 were excluded. For each plate, the vehicle control showed consistent measurements for all endpoints tested. For intra-plate normalization, data were normalized to the overall scaling factors, which was the mean of medians of the vehicle controls in each plate. The single cell-based data were averaged for the well-based condition. BrdU-positive cells were set by the total intensity of BrdU in the control at over 25,000 pixels. γ-H2AX positive cells were set by the total intensity of γ-H2AX in the control at over 120,000 pixels. The median lethal concentrations (LC50) were calculated using a nonlinear regression curve fit on GraphPad Prism 5 (San Diego, Calif.). To examine the correlation between the cytoskeleton and DNA damage responses in multinucleated cells, the Spearman correlation analysis was conducted between the total intensity of F-actin and γ-H2AX for 24, 48, and 72 h using cell-based data. Data were presented as mean±standard deviation (SD). Statistical significance was determined using one-way ANOVA followed by Tukey-Kramer all pairs comparison. The p-value less than 0.05 denoted a significant difference compared to the vehicle control (*).
Both image processing workflows were tested on a desktop Windows 7 workstation (16 GB RAM) and CellProfiler pipelines were submitted for processing on a Linux cluster. Software versions used were CellProfiler version 2.1.2 v2015_08_05, CellProfiler Analyst 2.0 v2014_04_01, and ilastik version 0.5.12; links to Windows binary versions of these are available on the world wide web at cellprofile.org/published_pipelines.shtml.
To determine the appropriate concentrations of BPA and its Analogs to be used in the HCA experiments, cell viability was measured by the NR uptake assay.
Nuclear morphology is considered to be a sensitive endpoint for detecting chemical toxicity in HCA assays (Martin et al., 2014; O'Brien et al., 2006). In our HCA, we measured multiple morphological parameters of nuclei, including nuclear area, roundness (LWR) and smoothness (P2A).
In
To quantify the multinucleated cells in the co-cultures treated with the various compounds, we applied a supervised ML-based HCA of imaging data. In
Cell cycle progression is essential for germ cell renewal and progeny cell production (De Rooij and Russell, 2000). We previously developed a HCA assay to measure DNA synthesis and cell cycle progression.
In addition, we observed that BPA and its Analogs preturbed cell cycle progression in the co-culture modelin a dose-and time-dependent manner (
F-actin structures are involved in various cellular processes including germ cell nuclei remodeling, cytoplasm reduction, and cell movement during spermatogenesis (Niedenberger et al., 2013). In Sertoli cells, parallel actin bundles form ectoplasmic specialization (ES) to provide an immunological barrier for germ cells and regulate elongated spermatid orientation and spermatozoa release (Cheng and Mruk, 2002; Setchell, 2008; Wong et al., 2008). As shown in
In order to quantify phenotypic 3D structure in the co-cultures and the cells' response to chemical perturbations, the cells with F-actin bundles stretching across the cytosol were quantified based on a supervised ML approach.
Although current HCA has emerged as a powerful tool for analyzing multiparametric data for toxicity profiling, data exploration lags behind. High-content analysis has focused on one or two image-related features and has generated population-averaged readouts that simply reflect alterations of morphological features in each sample (Singh et al., 2014). In addition, the aggregation of single cell-level data usually masks phenotypic heterogeneity within cells, especially when a phenotypic change only occurred in a small specific subpopulation (Altschuler and Wu, 2010). Some phenotypic changes could be quantified by specific fluorescent labeling; however, it is difficult to evaluate multiple parameters precisely due to the multiplex fluorescent labels used within a single assay (Heynen-Genel et al., 2012). Therefore, it is essential to employ an advanced computational approach to integrate multi-dimensional HCA data on the single-cell level to precisely quantify the complex phenotypes of interest. ML that selects and integrates multiple features for automated phenotypic classification has been used to score various phenotypes in an unbiased manner (Bakal et al., 2007; Jones et al., 2009; Loo et al., 2007; Neumann et al., 2006). The classification algorithm was generated from a training dataset by manual annotation of some representative images according to the predefined classes. After the training period, the ML algorithm could automatically discriminate among the classes in the full dataset. This supervised ML approach reduced the workload that comes with the manual adaption of parameter sets, and increased objectivity, consistency, and accuracy in large-scale data sets (Sommer and Gerlich, 2013; Tarca et al., 2007). In recent years, ML combined with HCA has been used to generate phenotypic profiling in various research fields (Conrad and Gerlich, 2010; Fuchs et al., 2010; Fuller et al., 2016; Leonard et al., 2015; Mata et al., 2016; Schmitz et al., 2010).
Sertoli and Leydig cells play critical roles in maintaining spermatogenesis and reproductive functions by providing physiological and nutritional support for germ cell mitosis, meiosis and movement. These two cell types have been employed in various co-culture systems to improve the physical relevance of in vitro models, and examine the reproductive toxicities of various chemicals. Recently, our lab combined the C18-4 spermatogonial cell line, the TM3 Leydig cell line, and the TM4 Sertoli cell line to construct an animal-free testicular cell co-culture model that mimics the in vivo testicular structure. Leydig cells, TM3, and Sertoli cells, TM4, were previously established from the testis of immature BALB/c mice and exhibited distinct morphology and growth response to hormones (Mather, 1980). The incorporation of these testicular somatic cells in this co-culture model showed a distinct cord-like structure, high specificity, and high sensitivity in the classification of reproductive toxicants (Example 1).
Nuclear morphological features have been suggested as useful indicators in various adverse cellular events (Eidet et al., 2014; Ikeguchi et al., 1999). In HCA assays, the quantitative assessment of multiple nuclear parameters have been demonstrated as sensitive markers for detecting early cytotoxic effects. In the present study, the quantification of nuclear morphology revealed that BPA treatment significantly altered nuclear area at a non-lethal dose, which is consistent with previous in vivo studies in which exposure to BPA induced abnormal nuclear morphology in rat mammary glands and mice testes (Ibrahim et al., 2016; Takao et al., 1999). In addition, the significant changes of P2A and LWR in the cells treated with BPA and BPS could be detected at lower doses in the co-cultures, when compared to the single cell cultures. These data suggested that the alteration of nuclear morphology in the co-cultures could be a sensitive endpoint in the detection of chemical toxicity.
In this study, we developed a ML-based phenotypic classification based on multiple subcellular features extracted from nuclear morphology, the texture and intensity of targeted proteins. These results revealed that BPAF induced dose- and time-dependent multinucleated cells in the co-cultures. The induction of multinucleated gonocytes has been reported as a reproductive toxicity marker in animal models and in humans in response to environmental chemicals, including di-(n-butyl) phthalate (DBP), BPA, andrographolide, and aflatoxin (Akbarsha and Murugaian, 2000; Barlow et al., 2004; Faridha et al., 2007; Gallegos-Avila et al., 2010; Mylchreest et al., 2002; Takao, et al., 1999). Although the underlying mechanism of multinucleated cell formation is still unclear, Faridha et al. reported that the generation of multinucleated spermatids was through the opening of the cytoplasmic bridge and merging of multiple cells (Faridha, et al., 2007). In addition, the overexpression of p190RhoGAP, an actin stress fiber regulator, has been associated with multinucleated phenotypes (Su et al., 2003). Furthermore, the perturbation of the cytoskeleton texture co-occurred with the formation of multinucleated cells, which could be explained by the alteration of compressive forces driven by perinuclear actin networks (Chen et al., 2015). Compared to non-multinucleated cells in the same treatment conditions, multinucleated cells exhibited higher correlations between cytoskeleton perturbation and DNA damage responses on the single-cell level, showing the unique biological characteristics of these cells. It has been reported that DNA damage induced dramatic alterations to nuclear and cytoplasmic actin, and F-actin polymerization served as a negative modulator in DNA damage responses (Belin et al., 2015; Chang et al., 2015; Wang et al., 2013; Zuchero et al., 2012). In future studies, the underlying mechanism of BPAF-induced multinucleation of testicular cells should be examined.
M phase, one of the most important events for successful cell reproduction, in which replicated chromosomes were segregated into two daughter cells (Nurse, 1990). The identification and quantification of cell populations in M phase usually requires additional staining using mitotic specific markers, such as anti-phosphorylated (ser10) H3 (Lyman et al., 2011). In one of our lab's previous studies, we developed an HCA approach to generate a cell cycle profile with discrete SubG1, G0/1, S and G2/M phases in the spermatogonial cell line (Liang, et al., 2017b). However, this profiler could not provide a quantification of cell population in M phase. Blasi et. al recently reported a label-free quantification of mitotic cell cycle phase by applying supervised ML to multi-dimensional features of single cells (Blasi et al., 2016). Our lab has since established a ML analysis pipeline in the co-cultures, which was able to recognize and quantify the cells in M phase based on morphological, texture and intensity features extracted from the multi-channel fluorescence staining. We have shown M phase arrest in the co-cultures treated with BPA and its Analogs in a dose- and time-dependent manner, illustrating the chemical specific effect on cell cycle progression. The results presented here are consistent with previous findings showing BPA exposure significantly perturbed spermatogenesis in animal models and inhibited cell proliferation in Sertoli cell TM4 and Leydig cell TM3 (Ali et al., 2014; Chen et al., 2016b; Liu et al., 2013; Pereira et al., 2014).
Actin, one of major component of the cytoskeleton, has been shown to play essential roles in cell movement, cargo transportation, acrosome reaction, and nuclear modification during spermatogenesis (Kierszenbaum and Tres, 2004; Sun et al., 2011). Alteration of F-actin intensity is a sensitive indicator for monitoring the adverse effects of environmental exposure. However, the quantification of F-actin total intensity might not reflect spatial alterations of F-actin filaments in the cytoplasm of cells. In the co-culture model, we observed two types of F-actin fibers, which included the dense cortex F-actin on the cell edge and the F-actin filaments stretching across the cytosol that assembled forming the 3D structure. When an ML approach was applied to recognize and quantify the cells with stretching F-actin bundles across cytosol, it demonstrated significant decreases of this specific phenotype in the co-cultures treated with different chemicals. In the previous study, a combination of cell-type specific markers and HCA cytoskeleton analysis revealed that Sertoli cells exhibited stretching F-actin bundles in the co-cultures (Example 1). Within the Sertoli cells, parallel actin bundles formed ectoplasmic specialization, which participates in spermatid head formation, cell movement, elongated spermatid orientation, and spermatozoa release. Damage to Sertoli cells has often resulted in germ cell degeneration and loss (Vidal and Whitney, 2014). Thus, the alteration of cells with stretching F-actin filaments suggested a potential loss of Sertoli cells and perturbation of 3D structure in the co-cultures. Findings supported the concept that if Sertoli cells are damaged then the blood-testis barrier which is crucial in the maintenance and initiation of spermatogenesis after toxicant-induced spermatogenesis, during the stages of spermatogonial differentiation. If the exposure to the toxicant is long enough this can permanently affect the morphology of F-actin in the Sertoli cells leading to infertility. These observations were consistent with previous in vivo studies that BPA or its Analogs treatment altered seminiferous tubule morphology in animal models.
γ-H2AX has been considered a highly specific and sensitive cellular marker for monitoring initiation of DNA damage (Ando et al., 2014; Fu et al., 2012; Garcia-Canton et al., 2013). Studies have shown that exposure to BPA or BPS induced DNA damage response in germ cells (Chen et al., 2016a; Liang et al., 2017a; Liu et al., 2014). In addition, the genotoxicity of BPA and its selected Analogs have been detected in multiple cell lines and followed a cell-type-specific and chemical-specific manner. In the co-cultures, we observed that BPA and BPS treatment did not induce γ-H2AX expression. One of the possible explanations of inconsistency between the co-culture and previous single cell type cultures could be due to differential biotransformation of BPA and its Analogs, and differing DNA damage repair capacity among the cell models. In human breast cancer cells, BPA was able to induce γ-H2AX at a low dose of 10 nM, but in HepG2 cell line, BPA was not able to induce γ-H2AX even at a dose of 100 μM, but bisphenol F was (Audebert et al., 2011; Pfeifer et al., 2015). For the other two compounds tested in the co-culture model, BPAF treatment significantly induced DNA damage response starting at 5 μM, and TBBPA induced a higher degree of DNA damage response at a dose of 15 μM, suggesting higher genotoxicity compared to BPA.
We observed a similar toxicity ranking of BPA and its selected Analogs in the co-culture model as compared to the single spermatogonial cell cultures. BPAF exerted the highest toxicity, followed by TBBPA, BPA, and BPS. This in vitro finding was further supported by an in vivo study that demonstrated that BPAF exposure uniquely impaired pregnancies and sexual development in rats at doses of ˜80 and ˜280 mg/kg, whereas BPA exposure did not alter these reproductive endpoints at similar levels (Sutherland. et al., 2017). Given the similarity of these two chemicals' in vitro estrogen and androgen receptor (ER, AR) activities, the differential reproductive toxicity both in vitro and in vivo potentially suggested that BPAF might partially exert its adverse effects on the reproductive system in an estrogen-independent manner. In addition, recent data has demonstrated that TBBPA did not interact with ER α/β or AR in a panel of in vitro bioassays, but it showed higher toxicity in both testicular cell models, when compared to BPA (Molina-Molina et al., 2013). Future studies will be critical to validate current findings and elucidate the mechanisms of action.
The ML-based HCA for the testicular cell co-culture model still has several limitations and several advantages. Although current supervised ML recognize certain phenotypes of interest based on the predefined knowledge, it still requires the creation of a training set spanning all replicates to improve accuracy. This is a time and labor extensive task, as it involves the manual sorting of images for each parameter examined and fluorescent label used. For the initial training batches, the individual sorting the images should be highly knowledgeable in the histopathology of the cells being examined. Another potential hang up on the ML-based HCA assay could be the computing power needed to efficiently operate this software. The amount of data that HCA produces can slow down most computers without the proper hardware. This ML-based HCA pipeline could take up to three hours per data set on a standard desktop computer. The better the computing power the less time that the extensive classification system will take to perform its tasks. This specific pipeline is set for the parameters examined in this study, if one wanted they would easily be able to manually adjust these parameters for a pipeline and if needed re-train their data set. The individual would then be able to examine the type of cells and parameters that were selected for their study. This in vitro model mimicked a functional in vivo physiology, through including three cell types, this model allowed for the creating of the BTB and also specialized niche's for spermatogonial cell differentiation, allowing our lab to establish more relevant data regarding toxicity to a human cell based model, rather than just one type of cell culture. This will allow for our lab to analyze the differential biotransformation of BPA and its Analogs and an understanding of how these metabolites can affect the different cell types. Another advantage portrayed by this HCA model is that future studies conducted regarding similar parameters examined, that the training set is already created and has shown with high sensitivity and specificity to be an accurate classifier of male reproductive toxicants. This pipeline may be able to help with a quicker and more efficient way of examining and classifying chemical toxicant's and may allow for regulatory decisions to be based off these findings.
In summary, we developed a ML-based HCA approach to characterize the testicular toxicity in a testicular cell co-culture model. Through the utilization of ML, our lab explored the added value of HCA to classify multiple cellular phenotypes and characterize the compound specific testicular toxicity of BPA and its selected Analogs. By integrating machine-learning based approaches with established HCA algorithms, it will be possible to uncover multi-dimensional data and quantify these phenotypic changes in large-scale exposure to environmental chemicals.
Akbarsha, M. A., and Murugaian, P. (2000). Aspects of the male reproductive toxicity/male antifertility property of andrographolide in albino rats: effect on the testis and the cauda epididymidal spermatozoa. Phytother Res 14(6), 432-5.
Ali, S., Steinmetz, G., Montillet, G., Perrard, M. H., Loundou, A., Durand, P., Guichaoua, M. R., and Prat, O. (2014). Exposure to low-dose bisphenol A impairs meiosis in the rat seminiferous tubule culture model: a physiotoxicogenomic approach. Plos One 9(9), e106245.
Altschuler, S. J., and Wu, L. F. (2010). Cellular Heterogeneity: Do Differences Make a Difference? Cell 141(4), 559-563.
Ando, M., Yoshikawa, K., Iwase, Y., and Ishiura, S. (2014). Usefulness of monitoring gamma-H2AX and cell cycle arrest in HepG2 cells for estimating genotoxicity using a high-content analysis system. J Biomol Screen 19(9), 1246-54.
Audebert, M., Dolo, L., Perdu, E., Cravedi, J. P., and Zalko, D. (2011). Use of the gamma H2AX assay for assessing the genotoxicity of bisphenol A and bisphenol F in human cell lines. Arch Toxicol 85(11), 1463-1473.
Bakal, C., Aach, J., Church, G., and Perrimon, N. (2007). Quantitative morphological signatures define local signaling networks regulating cell morphology. Science 316(5832), 1753-1756.
Barlow, N. J., McIntyre, B. S., and Foster, P. M. D. (2004). Male reproductive tract lesions at 6, 12, and 18 months of age following in utero exposure to di(n-butyl) phthalate. Toxicol Pathol 32(1), 79-90.
Belin, B. J., Lee, T., and Mullins, R. D. (2015). DNA damage induces nuclear actin filament assembly by Formin-2 and Spire-1/2 that promotes efficient DNA repair. Elife 4.
Blasi, T., Hennig, H., Summers, H. D., Theis, F. J., Cerveira, J., Patterson, J. O., Davies, D., Filby, A., Carpenter, A. E., and Rees, P. (2016). Label-free cell cycle analysis for high-throughput imaging flow cytometry. Nat Commun 7.
Breiman, L. (2001). Random forests. Mach Learn 45(1), 5-32.
Buchser, W., Collins, M., Garyantes, T., Guha, R., Haney, S., Lemmon, V., Li, Z., and Trask, O. J. (2004). Assay Development Guidelines for Image-Based High Content Screening, High Content Analysis and High Content Imaging. In Assay Guidance Manual (G. S. Sittampalam, N. P. Coussens, K. Brimacombe, A. Grossman, M. Arkin, D. Auld, C. Austin, J. Baell, B. Bejcek, T. D. Y. Chung, et al., Eds.), Bethesda (MD).
Calafat, A. M., Ye, X. Y., Wong, L. Y., Reidy, J. A., and Needham, L. L. (2008). Exposure of the US population to bisphenol A and 4-tertiary-octylphenol: 2003-2004. Environ Health Persp 116(1), 39-44.
Carpenter, A. E., Jones, T. R., Lamprecht, M. R., Clarke, C., Kang, I. H., Friman, O., Guertin, D. A., Chang, J. H., Lindquist, R. A., Moffat, J., et al. (2006). CellProfiler: image analysis software for identifying and quantifying cell phenotypes. Genome Biology 7(10).
Chang, C. Y., Leu, J. D., and Lee, Y. J. (2015). The Actin Depolymerizing Factor (ADF)/Cofilin Signaling Pathway and DNA Damage Responses in Cancer. Int J Mol Sci 16(2), 4095-4120.
Chen, B., Co, C., and Ho, C. C. (2015). Cell shape dependent regulation of nuclear morphology. Biomaterials 67, 129-136.
Chen, Y. C., Shu, L., Qiu, Z. Q., Lee, D. Y., Settle, S. J., Hee, S. Q., Telesca, D., Yang, X., and Allard, P. (2016a). Exposure to the BPA-Substitute Bisphenol S Causes Unique Alterations of Germline Function. Plos Genet 12(7).
Chen, Z. J., Zhang, K. S., Ge, L. C., Liu, H., Chen, L. K., Du, J., and Wang, H. S. (2016b). Signals involved in the effects of bisphenol A (BPA) on proliferation and motility of Leydig cells: a comparative proteomic analysis. Toxicol Res-Uk 5(6), 1573-1584.
Cheng, C. Y., and Mruk, D. D. (2002). Cell junction dynamics in the testis: Sertoli-germ cell interactions and male contraceptive development. Physiol Rev 82(4), 825-874.
Chong, Y. T., Koh, J. L. Y., Friesen, H., Duffy, S. K., Cox, M. J., Moses, A., Moffat, J., Boone, C., and Andrews, B. J. (2015). Yeast Proteome Dynamics from Single Cell Imaging and Automated Analysis (vol 161, pg 1413, 2015). Cell 162(1), 221-221.
Conrad, C., and Gerlich, D. W. (2010). Automated microscopy for high-content RNAi screening. Journal of Cell Biology 188(4), 453-461.
De Rooij, D. G., and Russell, L. D. (2000). All you wanted to know about spermatogonia but were afraid to ask. Journal of Andrology 21(6), 776-798.
Driffield, M., Harmer, N., Bradley, E., Fernandes, A. R., Rose, M., Mortimer, D., and Dicks, P. (2008). Determination of brominated flame retardants in food by LC-MS/MS: diastereoisomer-specific hexabromocyclododecane and tetrabromobisphenol A. Food Addit Contam A 25(7), 895-903.
Eidet, J. R., Pasovic, L., Maria, R., Jackson, C. J., and Utheim, T. P. (2014). Objective assessment of changes in nuclear morphology and cell distribution following induction of apoptosis. Diagn Pathol 9.
Elmore, S. A., Ryan, A. M., Wood, C. E., Crabbs, T. A., and Sills, R. C. (2014). FutureTox II: Contemporary Concepts in Toxicology: “Pathways to Prediction: In Vitro and In Silico Models for Predictive Toxicology”. Toxicol Pathol 42(5), 940-942.
EU. (2016). amending Annex XVII to Regulation (EC) No 1907/2006 of the European Parliament and of the Council concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) as regards bisphenol A. In (Vol. 2016/2235.
Faridha, A., Faisal, K., and Akbarsha, M. A. (2007). Aflatoxin treatment brings about generation of multinucleate giant spermatids (symplasts) through opening of cytoplasmic bridges: Light and transmission electron microscopic study in Swiss mouse. Reproductive Toxicology 24(3-4), 403-408.
FDA. (2013). Indirect Food Additives: Adhesives and Components of Coatings. In (Vol. 21 CFR 175, pp. 41840 -41843.
FDA. (2012). Indirect Food Additives: Polymers. In (Vol. 21 CFR 177, pp. 41899 -41902.
Fu, S. B., Yang, Y., Tirtha, D., Yen, Y., Zhou, B. S., Zhou, M. M., Ohlmeyer, M., Ko, E. C., Cagan, R., Rosenstein, B. S., et al. (2012). gamma-H2AX Kinetics as a Novel Approach to High Content Screening for Small Molecule Radiosensitizers. Plos One 7(6).
Fuchs, F., Pau, G., Kranz, D., Sklyar, O., Budjan, C., Steinbrink, S., Horn, T., Pedal, A., Huber, W., and Boutros, M. (2010). Clustering phenotype populations by genome-wide RNAi and multiparametric imaging. Mol Syst Biol 6, 370.
Fuller, J. A., Berlinicke, C. A., Inglese, J., and Zack, D. J. (2016). Use of a Machine Learning-Based High Content Analysis Approach to Identify Photoreceptor Neurite Promoting Molecules. Adv Exp Med Biol 854, 597-603.
Gallegos-Avila, G., Ancer-Rodriguez, J., Niderhauser-Garcia, A., Ortega-Martinez, M., and Jaramillo-Rangel, G. (2010). Multinucleation of spermatozoa and spermatids in infertile men chronically exposed to carbofuran. Reprod Toxicol 29(4), 458-60.
Garcia-Canton, C., Anadon, A., and Meredith, C. (2013). Assessment of the in vitro γH2AX assay by High Content Screening as a novel genotoxicity test. Mutation Research/Genetic Toxicology and Environmental Mutagenesis 757(2), 158-166.
Harris, S., Hermsen, S. A. B., Yu, X. Z., Hong, S. W., and Faustman, E. M. (2015). Comparison of toxicogenomic responses to phthalate ester exposure in an organotypic testis co-culture model and responses observed in vivo. Reproductive Toxicology 58, 149-159.
Harris, S., Shubin, S. P., Wegner, S., Van Ness, K., Green, F., Hong, S. W., and Faustman, E. M. (2016a). The presence of macrophages and inflammatory responses in an in vitro testicular co-culture model of male reproductive development enhance relevance to in vivo conditions. Toxicol in Vitro 36, 210-215.
Harris, S., Wegner, S., Hong, S. W., and Faustman, E. M. (2016b). Phthalate metabolism and kinetics in an in vitro model of testis development. Toxicol in Vitro 32, 123-131.
Haywood, M., Spaliviero, J., Jimemez, M., King, N. J. C., Handelsman, D. J., and Allan, C. M. (2003). Sertoli and germ cell development in hypogonadal (hpg) mice expressing transgenic follicle-stimulating hormone alone or in combination with testosterone. Endocrinology 144(2), 509-517.
Heynen-Genel, S., Pache, L., Chanda, S. K., and Rosen, J. (2012). Functional genomic and high-content screening for target discovery and deconvolution. Expert Opin Drug Discov 7(10), 955-68.
Hofmann, M. C., Braydich-Stolle, L., Dettin, L., Johnson, E., and Dym, M. (2005). Immortalization of mouse germ line stem cells. Stem cells (Dayton, Ohio) 23(2), 200-10.
Ibrahim, M. A. A., Elbakry, R. H., and Bayomy, N. A. (2016). Effect of bisphenol A on morphology, apoptosis and proliferation in the resting mammary gland of the adult albino rat. Int J Exp Pathol 97(1), 27-36.
Ikeguchi, M., Sakatani, T., Endo, K., Makino, M., and Kaibara, N. (1999). Computerized nuclear morphometry is a useful technique for evaluating the high metastatic potential of colorectal adenocarcinoma. Cancer 86(10), 1944-1951.
Jones, T. R., Carpenter, A. E., Lamprecht, M. R., Moffat, J., Silver, S. J., Grenier, J. K., Castoreno, A. B., Eggert, U. S., Root, D. E., Golland, P., et al. (2009). Scoring diverse cellular morphologies in image-based screens with iterative feedback and machine learning. P Natl Acad Sci USA 106(6), 1826-1831.
Jones, T. R., Kang, I. H., Wheeler, D. B., Lindquist, R. A., Papallo, A., Sabatini, D. M., Golland, P., and Carpenter, A. E. (2008). CellProfiler Analyst: data exploration and analysis software for complex image-based screens. Bmc Bioinformatics 9.
Kang, J. H., Kondo, F., and Katayama, Y. (2006). Human exposure to bisphenol A. Toxicology 226(2-3), 79-89.
Kierszenbaum, A. L., and Tres, L. L. (2004). The acrosome-acroplaxome-manchette complex and the shaping of the spermatid head. Arch Histol Cytol 67(4), 271-84.
Kitamura, S., Suzuki, T., Sanoh, S., Kohta, R., Jinno, N., Sugihara, K., Yoshihara, S., Fujimoto, N., Watanabe, H., and Ohta, S. (2005). Comparative study of the endocrine-disrupting activity of bisphenol A and 19 related compounds. Toxicological Sciences 84(2), 249-259.
Krewski, D., Acosta, D., Andersen, M., Anderson, H., Bailar, J. C., Boekelheide, K., Brent, R., Charnley, G., Cheung, V. G., Green, S., et al. (2010). Toxicity Testing in the 21st Century: A Vision and a Strategy. J Toxicol Env Heal B 13(2-4), 51-138.
Lakind, J. S., and Naiman, D. Q. (2011). Daily intake of bisphenol A and potential sources of exposure: 2005-2006 National Health and Nutrition Examination Survey. J Expo Sci Env Epid 21(3), 272-279.
Leonard, A. P., Cameron, R. B., Speiser, J. L., Wolf, B. J., Peterson, Y. K., Schnellmann, R. G., Beeson, C. C., and Rohrer, B. (2015). Quantitative analysis of mitochondrial morphology and membrane potential in living cells using high-content imaging, machine learning, and morphological binning. Bba-Mol Cell Res 1853(2), 348-360.
Liang, S., Yin, L., Shengyang Yu, K., Hofmann, M. C., and Yu, X. (2017a). High-Content Analysis Provides Mechanistic Insights into the Testicular Toxicity of Bisphenol A and Selected Analogs in Mouse Spermatogonial Cells. Toxicological sciences : an official journal of the Society of Toxicology 155(1), 43-60.
Liang, S., Yin, L., Shengyang Yu, K., Hofmann, M. C., and Yu, X. (2017b). High-Content Analysis Provides Mechanistic Insights into the Testicular Toxicity of Bisphenol A and Selected Analogs in Mouse Spermatogonial Cells. Toxicol Sci doi: 10.1093/toxsci/kfw178.
Liao, C. Y., and Kannan, K. (2013). Concentrations and Profiles of Bisphenol A and
Other Bisphenol Analogs in Foodstuffs from the United States and Their Implications for Human Exposure. J Agr Food Chem 61(19), 4655-4662.
Liao, C. Y., Liu, F., Guo, Y., Moon, H. B., Nakata, H., Wu, Q., and Kannan, K. (2012). Occurrence of Eight Bisphenol Analogs in Indoor Dust from the United States and Several Asian Countries: Implications for Human Exposure. Environ Sci Technol 46(16), 9138-9145.
Liu, C., Duan, W., Li, R., Xu, S., Zhang, L., Chen, C., He, M., Lu, Y., Wu, H., Pi, H., et al. (2013). Exposure to bisphenol A disrupts meiotic progression during spermatogenesis in adult rats through estrogen-like activity. Cell Death Dis 4, e676.
Liu, C., Duan, W. X., Zhang, L., Xu, S. C., Li, R. Y., Chen, C. H., He, M. D., Lu, Y. H., Wu, H. J., Yu, Z. P., et al. (2014). Bisphenol A exposure at an environmentally relevant dose induces meiotic abnormalities in adult male rats. Cell Tissue Res 355(1), 223-232.
Loo, L. H., Wu, L. F., and Altschuler, S. J. (2007). Image-based multivariate profiling of drug responses from single cells. Nat Methods 4(5), 445-453.
Lui, W. Y., Mruk, D., Lee, W. M., and Cheng, C. Y. (2003). Sertoli cell tight junction dynamics: Their regulation during spermatogenesis. Biology of reproduction 68(4), 1087-1097.
Lyman, S. K., Crawley, S. C., Gong, R., Adamkewicz, J. I., McGrath, G., Chew, J. Y., Choi, J., Holst, C. R., Goon, L. H., Detmer, S. A., et al. (2011). High-Content, High-Throughput Analysis of Cell Cycle Perturbations Induced by the HSP90 Inhibitor XL888. Plos One 6(3).
Martin, H. L., Adams, M., Higgins, J., Bond, J., Morrison, E. E., Bell, S. M., Warriner, S., Nelson, A., and Tomlinson, D. C. (2014). High-Content, High-Throughput Screening for the
Identification of Cytotoxic Compounds Based on Cell Morphology and Cell Proliferation Markers. Plos One 9(2).
Martin, M. T., Knudsen, T. B., Reif, D. M., Houck, K. A., Judson, R. S., Kavlock, R. J., and Dix, D. J. (2011). Predictive Model of Rat Reproductive Toxicity from ToxCast High Throughput Screening. Biol Reprod 85(2), 327-339.
Mata, G., Radojevic, M., Smal, I., Morales, M., Meijering, E., and Rubio, J. (2016). Automatic Detection of Neurons in High-Content Microscope Images Using Machine Learning Approaches. I S Biomed Imaging, 330-333.
Mather, J. P. (1980). Establishment and characterization of two distinct mouse testicular epithelial cell lines. Biology of reproduction 23(1), 243-52.
Mattiazzi Usaj, M., Styles, E. B., Verster, A. J., Friesen, H., Boone, C., and Andrews, B. J. (2016). High-Content Screening for Quantitative Cell Biology. Trends Cell Biol 26(8), 598-611.
Merrick, B. A., Paules, R. S., and Tice, R. R. (2015). Intersection of toxicogenomics and high throughput screening in the Tox21 program: an NIEHS perspective. Int J Biotechnol 14(1), 7-27.
Molina-Molina, J. M., Amaya, E., Grimaldi, M., Saenz, J. M., Real, M., Fernandez, M. F., Balaguer, P., and Olea, N. (2013). In vitro study on the agonistic and antagonistic activities of bisphenol-S and other bisphenol-A congeners and derivatives via nuclear receptors. Toxicol Appl Pharm 272(1), 127-136.
Mylchreest, E., Sar, M., Wallace, D. G., and Foster, P. M. D. (2002). Fetal testosterone insufficiency and abnormal proliferation of Leydig cells and gonocytes in rats exposed to di(n-butyl) phthalate. Reproductive Toxicology 16(1), 19-28.
Neumann, B., Held, M., Liebel, U., Erfle, H., Rogers, P., Pepperkok, R., and Ellenberg, J. (2006). High-throughput RNAi screening by time-lapse imaging of live human cells. Nat Methods 3(5), 385-390.
Neumann, B., Walter, T., Heriche, J. K., Bulkescher, J., Erfle, H., Conrad, C., Rogers, P., Poser, I., Held, M., Liebel, U., et al. (2010). Phenotypic profiling of the human genome by time-lapse microscopy reveals cell division genes. Nature 464(7289), 721-727.
Niedenberger, B. A., Chappell, V. K., Kaye, E. P., Renegar, R. H., and Geyer, C. B. (2013). Nuclear localization of the actin regulatory protein Palladin in sertoli cells. Mol Reprod Dev 80(5), 403-13.
Nurse, P. (1990). Universal Control Mechanism Regulating Onset of M-Phase. Nature 344(6266), 503-508.
O'Brien, P. J., Irwin, W., Diaz, D., Howard-Cofield, E., Krejsa, C. M., Slaughter, M. R., Gao, B., Kaludercic, N., Angeline, A., Bernardi, P., et al. (2006). High concordance of drug-induced human hepatotoxicity with in vitro cytotoxicity measured in a novel cell-based model using high content screening. Arch Toxicol 80(9), 580-604.
Paul Friedman, K., Watt, E. D., Hornung, M. W., Hedge, J. M., Judson, R. S., Crofton, K. M., Houck, K. A., and Simmons, S. O. (2016). Tiered High-Throughput Screening Approach to Identify Thyroperoxidase Inhibitors Within the ToxCast Phase I and II Chemical Libraries. Toxicological Sciences 151(1), 160-80.
Payne, A. H. (1990). Hormonal-Regulation of Cytochrome-P450 Enzymes, Cholesterol Side-Chain Cleavage and 17-Alpha-Hydroxylase C17-20 Lyase in Leydig-Cells. Biology of reproduction 42(3), 399-404.
Pereira, E. F., Aracava, Y., DeTolla, L. J., Jr., Beecham, E. J., Basinger, G. W., Jr., Wakayama, E. J., and Albuquerque, E. X. (2014). Animal models that best reproduce the clinical manifestations of human intoxication with organophosphorus compounds. J Pharmacol Exp Ther 350(2), 313-21.
Peretz, J., Vrooman, L., Ricke, W. A., Hunt, P. A., Ehrlich, S., Hauser, R., Padmanabhan, V., Taylor, H. S., Swan, S. H., VandeVoort, C. A., et al. (2014). Bisphenol A and Reproductive Health: Update of Experimental and Human Evidence, 2007-2013. Environ Health Persp 122(8), 775-786.
Pfeifer, D., Chung, Y. M., and Hu, M. C. T. (2015). Effects of Low-Dose Bisphenol A on DNA Damage and Proliferation of Breast Cells: The Role of c-Myc. Environ Health Persp 123(12), 1271-1279.
Repetto, G., del Peso, A., and Zurita, J. L. (2008). Neutral red uptake assay for the estimation of cell viability/cytotoxicity. Nat Protoc 3(7), 1125-1131.
Richter, C. A., Birnbaum, L. S., Farabollini, F., Newbold, R. R., Rubin, B. S., Talsness, C. E., Vandenbergh, J. G., Walser-Kuntz, D. R., and vom Saal, F. S. (2007). In vivo effects of bisphenol A in laboratory rodent studies. Reprod Toxicol 24(2), 199-224.
Rochester, J. R. (2013). Bisphenol A and human health: A review of the literature. Reproductive Toxicology 42, 132-155.
Schmitz, M. H., Held, M., Janssens, V., Hutchins, J. R., Hudecz, O., Ivanova, E., Goris, J., Trinkle-Mulcahy, L., Lamond, A. I., Poser, I., et al. (2010). Live-cell imaging RNAi screen identifies PP2A-B55alpha and importin-betal as key mitotic exit regulators in human cells. Nat Cell Biol 12(9), 886-93.
Setchell, B. P. (2008). Blood-Testis Barrier, Junctional and Transport Proteins and Spermatogenesis. Adv Exp Med Biol 636, 212-233.
Shukla, S. J., Huang, R., Austin, C. P., and Xia, M. (2010). The future of toxicity testing: a focus on in vitro methods using a quantitative high-throughput screening platform. Drug Discov Today 15(23-24), 997-1007.
Singh, S., Carpenter, A. E., and Genovesio, A. (2014). Increasing the Content of High-Content Screening: An Overview. Journal of Biomolecular Screening 19(5), 640-650.
Sommer, C., and Gerlich, D. W. (2013). Machine learning in cell biology—teaching computers to recognize phenotypes. Journal of Cell Science 126(24), 5529-5539.
Stossi, F., Bolt, M. J., Ashcroft, F. J., Lamerdin, J. E., Melnick, J. S., Powell, R. T., Dandekar, R. D., Mancini, M. G., Walker, C. L., Westwick, J. K., et al. (2014). Defining Estrogenic Mechanisms of Bisphenol A Analogs through High Throughput Microscopy-Based Contextual Assays. Chem Biol 21(6), 743-753.
Su, L., Agati, J. M., and Parsons, S. J. (2003). p190RhoGAP is cell cycle regulated and affects cytokinesis. Journal of Cell Biology 163(3), 571-582.
Sun, X., Kovacs, T., Hu, Y. J., and Yang, W. X. (2011). The role of actin and myosin during spermatogenesis. Mol Biol Rep 38(6), 3993-4001.
Sutherland., V., McIntyre., B., Pelch., K., Waidyanatha., S., Conley., J. M., Gray., L. E., and Foster, P. M. A Comparison of In Vivo Reproductive and Developmental Toxicity (DART) Endpoints for Bisphenol AF and Bisphenol A2017, Baltimore, Md.
Takao, T., Nanamiya, W., Nagano, I., Asaba, K., Kawabata, K., and Hashimoto, K. (1999). Exposure with the environmental estrogen bisphenol A disrupts the male reproductive tract in young mice. Life Sci 65(22), 2351-2357.
Tarca, A. L., Carey, V. J., Chen, X. W., Romero, R., and Draghici, S. (2007). Machine learning and its applications to biology. Plos Comput Biol 3(6), 953-963.
Vandenberg, L. N., Hauser, R., Marcus, M., Olea, N., and Welshons, W. V. (2007). Human exposure to bisphenol A (BPA). Reproductive Toxicology 24(2), 139-177.
Vidal, J. D., and Whitney, K. M. (2014). Morphologic manifestations of testicular and epididymal toxicity. Spermatogenesis 4(2), e979099.
vom Saal, F. S., Akingbemi, B. T., Belcher, S. M., Birnbaum, L. S., Crain, D. A., Eriksen, M., Farabollini, F., Guillette, L. J., Hauser, R., Heindel, J. J., et al. (2007). Chapel Hill bisphenol A expert panel consensus statement: Integration of mechanisms, effects in animals and potential to impact human health at current levels of exposure. Reproductive Toxicology 24(2), 131-138.
Wang, L., Wang, M., Wang, S. Y., Qi, T. Y., Guo, L. J., Li, J. J., Qi, W. J., Ampah, K. K., Ba, X. Q., and Zeng, X. L. (2013). Actin Polymerization Negatively Regulates p53 Function by Impairing Its Nuclear Import in Response to DNA Damage. Plos One 8(4).
Wong, E. W. P., Mruk, D. D., and Cheng, C. Y. (2008). Biology and regulation of ectoplasmic specialization, an atypical adherens junction type, in the testis. Bba-Biomembranes 1778(3), 692-708.
Yin, L., Wei, H., Liang, S., and Yu, X. (2017). An animal-free in vitro three-dimensional testicular cell co-culture model for evaluating male reproductive toxicants. Toxicological sciences : an official journal of the Society of Toxicology, 43-60.
Yu, X. Z., Hong, S., Moreira, E. G., and Faustman, E. M. (2009). Improving in vitro Sertoli cell/gonocyte co-culture model for assessing male reproductive toxicity: Lessons learned from comparisons of cytotoxicity versus genomic responses to phthalates. Toxicol Appl Pharm 239(3), 325-336.
Yu, X. Z., Sidhu, J. S., Hong, S., and Faustman, E. M. (2005). Essential role of extracellular matrix (ECM) overlay in establishing the functional integrity of primary neonatal rat sertoli cell/gonocyte co-cultures: An improved In vitro model for assessment of male reproductive toxicity. Toxicological Sciences 84(2), 378-393.
Zanella, F., Lorens, J. B., and Link, W. (2010). High content screening: seeing is believing. Trends Biotechnol 28(5), 237-245.
Zuchero, J. B., Belin, B., and Mullins, R. D. (2012). Actin binding to WH2 domains regulates nuclear import of the multifunctional actin regulator JMY. Mol Biol Cell 23(5), 853-863.
The complete disclosure of all patents, patent applications, and publications, and electronically available material (including, for instance, nucleotide sequence submissions in, e.g., GenBank and RefSeq, and amino acid sequence submissions in, e.g., SwissProt, PIR, PRF, PDB, and translations from annotated coding regions in GenBank and RefSeq) cited herein are incorporated by reference in their entirety. Supplementary materials referenced in publications (such as supplementary tables, supplementary figures, supplementary materials and methods, and/or supplementary experimental data) are likewise incorporated by reference in their entirety. In the event that any inconsistency exists between the disclosure of the present application and the disclosure(s) of any document incorporated herein by reference, the disclosure of the present application shall govern. The foregoing detailed description and examples have been given for clarity of understanding only. No unnecessary limitations are to be understood therefrom. The disclosure is not limited to the exact details shown and described, for variations obvious to one skilled in the art will be included within the disclosure defined by the claims.
Unless otherwise indicated, all numbers expressing quantities of components, molecular weights, and so forth used in the specification and claims are to be understood as being modified in all instances by the term “about.” Accordingly, unless otherwise indicated to the contrary, the numerical parameters set forth in the specification and claims are approximations that may vary depending upon the desired properties sought to be obtained by the present disclosure. At the very least, and not as an attempt to limit the doctrine of equivalents to the scope of the claims, each numerical parameter should at least be construed in light of the number of reported significant digits and by applying ordinary rounding techniques.
As used herein, the term “about,” when referring to a value or to an amount of mass, weight, time, volume, concentration or percentage is meant to encompass variations of in some embodiments±20%, in some embodiments±10%, in some embodiments±5%, in some embodiments±1%, in some embodiments±0.5%, and in some embodiments±0.1% from the specified amount, as such variations are appropriate to perform the disclosed method.
Notwithstanding that the numerical ranges and parameters setting forth the broad scope of the disclosure are approximations, the numerical values set forth in the specific examples are reported as precisely as possible. All numerical values, however, inherently contain a range necessarily resulting from the standard deviation found in their respective testing measurements.
All headings are for the convenience of the reader and should not be used to limit the meaning of the text that follows the heading, unless so specified.
This application claims the benefit of U.S. Provisional Application Ser. No. 62/523,342, filed Jun. 22, 2017, which is incorporated by reference herein.
This invention was made with government support under R21 OH 010473 awarded by the CDC, and R43 ES 027374 awarded by the National Institutes of Health. The government has certain rights in the invention. (37 CFR 401.14(f) (4))
| Number | Date | Country | |
|---|---|---|---|
| 62523342 | Jun 2017 | US |
