UC Riverside Electronic Theses and Dissertations

UC Riverside UC Riverside Electronic Theses and Dissertations

Title The Embryotoxic Effects of Harm Reduction Tobacco Products on Osteoblasts Developing from Human Embryonic Stem Cells

Permalink https://escholarship.org/uc/item/32z2s123

Author Sparks, Nicole Renee

Publication Date 2018

License https://creativecommons.org/licenses/by/4.0/ 4.0

Peer reviewed|Thesis/dissertation

eScholarship.org Powered by the California Digital Library University of California

UNIVERSITY OF CALIFORNIA RIVERSIDE

The Embryotoxic Effects of Harm Reduction Tobacco Products on Osteoblasts Developing from Human Embryonic Stem Cells

A Dissertation submitted in partial satisfaction of the requirements for the degree of

Doctor of Philosophy

Environmental Toxicology

Nicole Renee Sparks

June 2018

Dissertation Committee: Dr. Nicole zur Nieden, Chairperson Dr. David Eastmond Dr. Yinsheng Wang

The Dissertation of Nicole Renee Sparks is approved:

Committee Chairperson

University of California, Riverside

ACKNOWLEDGEMENTS

I am beyond appreciative to have many people in my life that have made this dissertation possible.

I am grateful for the advice and the guidance I received from my principal investigator Nicole I. zur Nieden and the many cups of coffee we enjoyed together. Her knowledge and dedication to stem cell research and toxicology has been inspiring. She has been instrumental in my doctoral career and beyond.

For some of the work presented herein I had technical assistance from Lauren M.

Walker (mouse work), Ivann Martinez, Joseph Madrid (in vivo injections), Riley Bottom, and Cristina Soto. Thank you, Lauren, for lending your ear and pastries. Thank you to all my lab mates from the zur Nieden lab and friends made in the Riccomagno lab. We are like family and I have enjoyed my time shared with you all.

Thank you to the Environmental Toxicology doctoral program and my dissertation committee members Dr. David Eastmond and Dr. Yinsheng Wang. A special, thank you to

Dawn Loyola who has helped me every quarter of my doctoral career.

I would also like to acknowledge my funding sources: The Dean’s fellowship, fellowships from the graduate research mentorship program, a Cornelius Hopper Diversity and pre‐doctoral fellowship as well as a doctoral fellowship from the Tobacco Related

Disease Research Program and a fellowship from the International Foundation for Ethical

Research. Research funding came from the Tobacco Related Disease Research Program, the John Hopkins Center for Alternatives to Animal Testing as well as the National

Institutes for Dental and Craniofacial Research.

I would like to acknowledge Dr. Feodor Price (Harvard University), Dr. Derrick

Rancourt (University of Calgary), and Dr. Amander Clark (University of California, Los

Angeles) for kindly sharing genomic DNA of hESC and hiPSC lines analyzed in Appendix

Fig 1.2.2. I would like to thank Dr. Martín García‐Castro for kindly donating SOX10 and

PAX7 antibodies. I would like to thank Dr. Prue Talbot (University of California,

Riverside) for generously providing our lab with the mainstream and sidestream tobacco smoke solutions used for some of my work. I would also like to thank to Dr. Martin

Riccomagno (University of California, Riverside) for kindly providing the pBK-Flex shRNA plasmid and Eric Wong for technical assistance.

Finally, I am forever thankful to my husband Jason. His understanding, encouragement, and amazing patience has been my foundation. Without his support I could have not completed my dissertation. My amazing children, Gavin and Austen Sparks, who have shown patience and understanding and have kept me full of joy and laughter.

ABSTRACT OF THE DISSERTATION

The Embryotoxic Effects of Harm Reduction Tobacco Products on Osteoblasts Developing from Human Embryonic Stem Cells

Nicole Renee Sparks

Doctor of Philosophy, Graduate Program in Environmental Toxicology University of California, Riverside, June 2018 Dr. Nicole zur Nieden, Chairperson

Cigarette smoke is a mixture of over 7000 toxic components with at least 69 smoke chemicals producing cancers, suggesting that smoking is harmful to smokers. More importantly, the World Health Organization has estimated that 10% of the world’s population is involuntarily exposed to environmental tobacco smoke, a number that does not include developing embryos. However, tobacco exposure during embryonic development is one of the risk factors for developing a congenital birth defect. An understudied, yet important abnormality caused by maternal smoking is the improper development of the embryonic skeleton, which can result in long-term burdens on the affected child and their family, creating a need for investigation that is addressed in this thesis.

Due to the difficulty in quitting smoking while pregnant, women often look to use harm reduction tobacco products (HRTPs) because they are perceived as a safer alternative in comparison to conventional cigarettes. These products include ultra-filtered cigarettes with reduced carcinogen content, and chewing tobacco, which omits exposure to harmful

vi combustion products. However, there is a lack of investigation to reveal the molecular mechanisms involved in skeletal embryotoxicity induced by conventional and harm- reduction tobacco, the unraveling of which is another objective of this thesis.

In the process of this investigation, an in vitro model based on human embryonic stem cells was developed. These cells are the only currently available cell culture system, which mimics the development of the human embryo. As this thesis shows, their differentiation into osteoblasts in vitro can be used to study chemical toxicity and the molecular mechanisms thereof.

vii

TABLE OF CONTENTS

Acknowledgements ...... iv

Abstract ...... vi

List of figures ...... ix

List of tables ...... xii

Chapter 1: Introduction to stem cells ...... 1

Chapter 2: Low osteogenic yield in human pluripotent stem cells associates with differential neural crest promoter methylation ...... 14

Chapter 3: Pluripotent stem cells as tools to assess developmental toxicity: diversity instead of consolidation ...... 50

Chapter 4: Differential predictivity of human pluripotent stem cell lines in skeletogenic developmental toxicity assays correlates with their innate differentiation bias towards neural crest or mesodermal osteogenesis ...... 73

Chapter 5: Tobacco-induced oxidative stress disrupts osteogenic differentiation in human embryonic stem cells ...... 119

Chapter 6: Harm-reduction tobacco alters bone development through tobacco sensitive transcriptional regulators ...... 155

Conclusion ...... 191

APPENDIX 1: Supplemental information ...... 199

APPENDIX 2: List of differentially expressed genes ...... 208

viii

LIST OF FIGURES

Figure 1.1. Human blastocyst ...... 1

Figure 1.2. Pluripotent stem cell differentiation ...... 2

Figure 1.3. Stem cell division and differentiation ...... 3

Figure 1.4. Adult stem cell ...... 4

Figure 1.5. Generation and application of induced pluripotent stem cells (iPSCs) ...... 6

Figure 1.6. Contributions of neural crest, paraxial mesoderm, and lateral plate mesoderm to bone development...... 7

Figure 2.1. Cytogenetic, pluripotent, and genetic characterization of human induced pluripotent stem cell (hiPSC) lines ...... 24

Figure 2.2. Osteogenic differentiation of human embryonic stem cells ...... 25

Figure 2.3. Intermediate differentiation in H9 cells is through the neural crest lineage ...... 27

Figure 2.4. Osteogenic differentiation in RIV9 and RIV4 human induced pluripotent stem cells (hiPSCs) ...... 30

Figure 2.5. Passage‐specific analysis of bone parameters and gene expression ...... 32

Figure 2.6. Osteoblast origin correlates with methylation of the TWIST1 and PAX7 promoters ...... 34

Figure 2.7. Toxicity of 5‐fluorouracil (5FU) tested with H9 and RIV9 hPSCs ...... 36

Figure 3.1. Embryonic stem cell derivation and differentiation ...... 52

Figure 3.2. Embryonic stem cell test (EST) ...... 54

Figure 3.3. Human ESC osteogenic induction ...... 60

Figure 4.1. Human ESCs are more sensitive in predicting the developmental toxicity of 13cisRA, a known human teratogen, than mouse ESCs ...... 85

Figure 4.2. Differentiation and cell viability endpoints measured with RIV9 hiPSCs ...... 87

Figure 4.3. Human H9 ESCs preferentially undergo neural crest osteogenesis ...... 89

Figure 4.4. Concentration-response curves for representative neural crest-associated inhibitory compounds ...... 91

Figure 5.1. The embryotoxicity of tobacco exposure ...... 124

Figure 5.2. Tobacco dysregulates the expression of osteogenic genes ...... 135

Figure 5.3. Camel sidestream smoke induces apoptosis ...... 136

Figure 5.4. Camel sidestream smoke exacerbates apoptosis ...... 139

Figure 5.5. Tobacco exposure decreases antioxidant activity ...... 140

Figure 5.6. Tobacco increases production of ROS in differentiating human embryonic stem cells ...... 141

Figure 5.7. Tobacco exposure decreases antioxidant activity ...... 143

Figure 6.1. pBK-Flex-shRNA plasmid map ...... 162

Figure 6.2. FOXO facilitates osteogenic differentiation ...... 167

Figure 6.3. Harm reduction tobacco products diminish the expression of FOXO ...... 169

Figure 6.4. HRTP-mediated nuclear exclusion of FOXO ...... 171

Figure 6.5. FOXO shRNA clones reduce FOXO protein levels ...... 172

Figure 6.6. FOXO is essential for osteogenic differentiation ...... 173

Figure 6.7. Differential expression of genes treated with Camel Blue and Camel Snus tobacco extract ...... 174

Figure 6.8. Highly differentially expression genes sensitive to Camel Blue exposure during osteogenic differentiation ...... 176

Figure 6.9. STE-responsive differentially expressed genes ...... 177

Figure 6.10. Genes downregulated unique to Camel Snus tobacco extract treatment ...... 178

Figure 6.11. Camel Blue and Camel Snus tobacco extract regulate different biological processes in differentiating cells ...... 180

Figure 6.12. Camel Blue and Camel Snus tobacco extract negatively affect bone development through separate biological processes ...... 181

Figure 7.1. Harm-reduction tobacco products transcriptional regulation during osteogenesis ...... 197

Appendix Figure 1.2.1. Pilot differentiation studies for protocol development ...... 199

Appendix Figure 1.2.2. Kinetic and passage-specific analysis of additional gene expression associated with osteogenesis and early lineage determination ...... 200

Appendix Figure 1.2.3. Inter-and intra-line variability between different passages of hPSCs used for osteogenic differentiation ...... 201

Appendix Figure 1.2.4. Early lineage determining genes are differentially expressed in the three hPSC lines ...... 202

Appendix Figure 1.2.5. Correlation between PAX7 and TWIST1 methylation status and osteogenic differentiation yield across hPSC lines ...... 203

Appendix Figure 1.4.1. Testing of three control chemicals in the hESC-ESTo ...... 205

Appendix Figure 1.4.2. The effects of 13cisRA on somatic cells ...... 206

Appendix Figure 1.4.3. TWIST1 mRNA expression as endpoint in the hESC-ESTo ...... 207

LIST OF TABLES

Table 4.1. Half-maximal inhibitory concentrations for all tested endpoints and prediction of embryotoxicity classes ...... 94

Table 5.1. Apoptosis-related genes examined with the RT2 ProfilerTM qPCR array ...... 132

Table 5.2. Primer sequences used in this study. Ta, annealing temperature ...... 134

Table 6.1. Oligonucleotide sequences ...... 164

Appendix Table 1.2.1. Oligonucleotide sequences used in this study. Ta, annealing temperature ...... 204

xii

CHAPTER 1

Introduction to stem cells

Stem cells are unique cells, regardless of their source of origin, that are defined by the ability to self-renew and unspecialized with the capacity to differentiate into different cell lineages. The self-renewal and differentiation properties of stem cells are what make them different from the majority of somatic cells that are committed to a specific function, such as a calcifying osteoblast and a beating cardiomyocyte. As such, stem cells are a powerful developmental tool and can aid in the understanding of tissue development, cell- based therapies, drug discovery, screening of toxicants, and understanding of diseases and birth defects (Tandon and Jyoti, 2012).

One such stem cell type, the embryonic stem cell (ESC), is derived from the inner cell mass (ICM) of a pre-implantation blastocyst, a 3 to 5-day old embryo. The blastocyst is formed early in development after the first lineage commitment. Here, a subset of cells specifies towards the trophectoderm, the extraembryonic tissue like the placenta, and the remaining cells becoming the ICM, which develops into the future embryo (Fig 1.1).

ESCs do not have the ability to differentiate

Figure 1.1. Human Blastocyst. Blastocyst into the extraembryonic tissue, but instead are showing the inner cell mass (ICM) and trophectoderm. Zhang et al., 2009.

Figure 1.2. Pluripotent stem cell differentiation. Embryonic stem cells (ESCs) are derived from the inner cell mass of the blastocyst. ESCs have the ability to be propagated in culture and can differentiate to all cell types from the three germ layers.https://ltsa.sheridancollege.ca/apps/human-anatomy/quiz_labeling_femur.html?bgImage=; https://www.sciencelearn.org.nz/resources/1917-muscle-structure-muscle-under-the-microscope; http://jonlieffmd.com/blog/how-many-different-kinds-of-neurons-are-there; http://trollpasta.wikia.com/wiki/File:Red-blood-cells-on-white.jpg; https://noticieros.televisa.com/ultimas- noticias/cientificos-unam-crean-metodo-mas-preciso-diagnosticar-higado-graso/

2 pluripotent: they can give rise to cell types from all three germ layers (i.e., ectoderm, mesoderm, and endoderm) (Fig 1.2; Itskovitz-Eldor et al., 2000) and can undergo extensive self-renewal and proliferation in vitro. Due to these properties, ESCs can be utilized for prenatal and postnatal studies in the pluripotent and differentiated states, respectively.

In 1981 came the first reports of the derivation and methods of growing mouse

ESCs (mESCs) (Martin, 1981; Evans and Kaufman, 1981). Early mESC studies provided a foundational understanding of embryonic stem cell biology and advanced appropriate culture techniques to facilitate eventual derivation of human ESCs (hESCs) from cryopreserved human blastocysts (Thomson et al., 1998). Prolonged pluripotency and proliferation of ESCs is attributed to the expression of stage-specific pluripotency markers, high telomerase activity, and the absence of lineage markers. Human ESCs express

POU5F1, NANOG, SOX2, SSEA-3, and SSEA-4 for pluripotency and downregulate the respective genes to initiate differentiation (Hadjimichael et al., 2015). The unlimited self- renewal of hESCs is associated with maintenance of long telomeres, which somatic cells

Figure 1.3. Stem cell division and differentiation. Through asymmetrical division, the stem cell can self-renew maintaining stemness or choose to differentiate and lose stemness. During differentiation the cell will go through progenitor states before becoming a somatic cell (terminal differentiation).

do not possess (Huang et al., 2011). In somatic cells,

telomere shortening eventually leads to cellular

senescence and apoptosis.

Mouse and human ESCs can remain

undifferentiated and self-renew under the correct

culture condition. When ESC colonies are able to

touch or clump together, this can lead to

spontaneous differentiation. Their differentiation

can be directed into specific cell types (i.e.,

osteoblasts, cardiomyocytes, chondrocytes and

Figure 1.4. Adult stem cells. Tissues adult neurons) by changing their culture medium to stem cells reside in. https://www.livescience.com/37009- human-body.html suppress pluripotency and upregulate cell specific genes (zur Nieden et al., 2003; zur Nieden et al.,

2005; Okabe et al., 1996). When ESCs progress towards a particular lineage, they undergo an asymmetrical division process that creates a progenitor cell. Progenitor cells are distinct from ESCs in that they possess limited proliferation capacity and reduced potency.

Eventually, these cells will terminally differentiate into a mature somatic cell (Fig 1.3).

In contrast to pluripotent stem cells, adult stem cells (ASCs) are unspecialized cells derived from specific somatic tissues (Fig 1.4). ASCs have been identified in many tissues from all germ layers and can self-renew and differentiate into cell types they reside in, having multipotent or unipotent potential. For instance, a hematopoietic stem cell (HSC) resides in the bone marrow. These cells can only form blood cells. Thus, the role of ASCs

4 in the adult body is to maintain homeostasis and repair damaged tissue. ASCs possess limited self-renewal capacity and prefer to remain quiescent until needed. Though restricted into what cells types they can become, ASCs do have clinical relevance. In particular, HSCs and mesenchymal stem cells (MSCs) have been used for blood and bone regeneration (Reya et al., 2001). Due to their harvest not destroying an embryo, ASCs do not have the same ethical concerns as ESCs.

It has been the ethical concerns associated with the use of ESCs that has prohibited some countries from allowing ESC research (Dhar and Hsi-En, 2009) and alternative ways to create ethically unchallenged pluripotent cells have been sought. In a technique called somatic cell nuclear transfer (SCNT), a somatic cell nucleus is transferred into an enucleated oocyte, activated to proliferate and ESC-like cells derived from the emerging blastocyst (Wilmut et al., 1997; Gurdon and Uehlinger, 1966). SCNT offered the potential to generate various cells types for research and therapeutic usage, however proved to be labor intensive, inefficient, and raised concerns of human cloning. This changed with the generation of induced pluripotent stem cells (iPSCs) (Takahashi et al., 2006). Induced

PSCs are generated through the viral transduction of mouse somatic cells, often fibroblasts, encoding for four transcription factors (Pou5f1, Sox2, Klf4, and c-Myc; Takahashi et al.,

2006). Using the same technique, human induced pluripotent stem cells (hiPSCs) were generated a year later (Takahashi et al., 2007). Similar to hESCs, hiPSCs can self-renew and can differentiate into cell types from all three germ layers. Human iPSCs are hypothesized to be a powerful tool to study development and patient-specific diseases that

Figure 1.5. Generation and application of induced pluripotent stem cells (iPSCs). Somatic cells are reprogramed into iPSCs. iPSCs can be used for cell therapy and modelling of diseases in a dish. Diseases can be corrected through gene targeting to correct for mutation, or novel drugs can be identified to inhibit cell death. Stadtfeld and Hochedlinger, 2010. can be globally adapted (Fig 1.5). However, evidence has shown differences in methylation patterns and iPSCs retaining signatures of their tissue of origin (Lister et al., 2011) as also found in this thesis (chapter 2).

Osteogenesis from ESCs

Osteogenesis is the complex process of bone formation that occurs during embryonic development, bone homeostasis in adults, and bone remodeling after an injury. During embryonic development, bone develops in two ways: intramembranous ossification and endochondral ossification. Neural crest derived osteogenic precursor cells undergo intramembranous ossification to give rise to the flat bones (parts of the skull and pelvis).

The long bones, like the axial and appendicular skeleton, are derived from a mesoderm

origin that often, but not

exclusively, uses endochondral

ossification (Fig 1.6, reviewed in

Keller and zur Nieden, 2011).

Endochondral ossification has a

cartilaginous precursor

framework that is eventually

replaced with bone (Provot et al.,

2013; Gilbert, 2000). The

formation and maintenance of Figure 1.6. Contributions of neural crest, paraxial mesoderm, and lateral plate mesoderm to bone development. Neural crest cells give bone is tightly regulated between rise to bones of the craniofacial skeleton. Paraxial mesoderm and lateral plate mesoderm comprise the axial and appendicular skeleton. https://www.the- osteoblasts, the cells that blueprints.com/blueprints/humans/anatomy/26684/view/rib_cage/;htt p://www.clipartpanda.com/clipart_images/the-lower-jaw-14587528; http://etc.usf.edu/clipart/15400/15417/femur_15417.htm synthesize a calcified and mineralized extracellular matrix (ECM), and osteoclasts, the bone absorbing cells.

Regardless of mesoderm or neural crest developmental origin, bone progenitor cells condense at locations of skeletal formation. These future sites of bone formation are packed with MSCs which will differentiate into the osteoblasts and later the terminally mature osteoblasts, known as osteocytes. Osteoblast differentiation is controlled by stage

(proliferation, differentiation, mineralization, and maturation) specific expression of regulatory proteins.

Factors that control osteoblastogenesis from ESCs include ascorbic acid, β- glycerophosphate and 1,25-(OH)2 vitamin D3 (zur Nieden et al., 2003). Pioneering studies

7 from our laboratory showed that during mESC osteogenesis osteogenic genes are regulated in a temporal manner equivalent to embryonic and MSC osteogenesis: alkaline phosphatase

(ALPL) expression is followed by CBFA1, an osteogenic master regulatory transcription factor (Ducy et al., 1997) coding for runt-related transcription factor 2 (RUNX2), as well as osteopontin (OPN). Later during the differentiation, bone sialoprotein (BSP) and osteocalcin (OCN) are expressed marking mature osteoblasts, then followed by CAPG and

DESTRIN, which are associated with osteocytes (Ehnes et al., 2015). Mis-regulation at any step during osteogenesis can lead to a wide array of skeletal disorders and diseases (zur

Nieden et al., 2005, 2007).

Aims and Objectives

Epidemiological studies suggest cigarette smoking as a probable environmental factor for a variety of congenital anomalies (Surgeon General, 2010), including low bone mass

(Godfrey et al., 2001) and increased fracture risk (Parviainen et al., 2017), both of which affect skeletal health. However, there is a lack of investigation to reveal the molecular mechanisms involved in tobacco-induced skeletal toxicity. This thesis aims to provide knowledge on how tobacco exposure can perturb osteoblast differentiation by addressing four specific aims:

Specific Aim 1: Develop a protocol to differentiate human ESCs and iPSCs into osteoblasts in vitro (addressed in chapter 2)

Specific Aim 2: Develop an in vitro developmental osteotoxicity screening model using hESCs and hiPSCs to predict and identify developmentally toxic compounds (addressed in chapters 3 and 4)

Specific Aim 3: Determine the impact of tobacco-induced oxidative stress and downstream transcriptional changes during osteogenic differentiation (addressed in chapters 5 and 6)

References 1. Dhar D, Hsi-En Ho J. Stem cell research policies around the world. Yale J Biol Med 2009;82(3):113-5.

2. Evans MJ, Kaufman MH. Establishment in culture of pluripotential cells from mouse embryos. Nature 1981;292(5819):154-6.

3. Gilbert SF. Osteogenesis: The Development of Bones. In: Developmental Biology. 6th edition. Sunderland (MA): Sinauer Associates; 2000. Available from: https://www.ncbi.nlm.nih.gov/books/NBK10056/

4. Godfrey K, Walker-Bone K, Robinson S, Taylor P, Shore S, Wheeler T, Cooper C (2001) Neonatal bone mass: influence of parental birthweight, maternal smoking, body composition, and activity during pregnancy. J Bone Miner Res 16:1694–1703.

5. Gurdon JB, Uehlinger V. "Fertile" intestine nuclei. Nature 1966;210(5042):1240-1.

6. Hadjimichael C, Chanoumidou K, Papadopoulou N, Arampatzi P, Papamatheakis J, Kretsovali A. Common stemness regulators of embryonic and cancer stem cells. World J Stem Cells. 2015;7(9):1150-84.

7. How tobacco smoke causes disease: the biology and behavioral basis for smoking- attributable disease: a report of the Surgeon General. – Rockville, MD: Dept. of Health and Human Services, Public Health Service, Office of Surgeon General, 2010.

8. Huang J, Wang F, Okuka M, Liu N, Ji G, Ye X, Zuo B, Li M, Liang P, Ge WW, Tsibris JC, Keefe DL, Liu L. Association of telomere length with authentic pluripotency of ES/iPS cells. Cell Res. 2011;21(5):779-92.

9. Itskovitz-Eldor J, Schuldiner M, Karsenti D, Eden A, Yanuka O, Amit M, Soreq H, Benvenisty N. Differentiation of human embryonic stem cells into embryoid bodies compromising the three embryonic germ layers. Molecular Medicine 2000;6(2):88- 95.

10. Keller K, zur Nieden NI. 2011. Osteogenesis from pluripotent stem cells: Neural Crest or Mesodermal Origin? Embryonic Stem Cells Michael S. Kallos, IntechOpen, DOI:10.5772/24432.Availablefrom: https://www.intechopen.com/books/embryonic- stem-cells-differentiation-and-pluripotent-alternatives/osteogenesis-from- pluripotent-stem-cells-neural-crest-or-mesodermal-origin-

11. Lister R, Pelizzola M, Kida YS, Hawkins RD, Nery JR, Hon G, Antosiewicz-Bourget J, O'Malley R, Castanon R, Klugman S, Downes M, Yu R, Stewart R, Ren B, Thomson JA, Evans RM, Ecker JR. Hotspots of aberrant epigenomic reprogramming in human induced pluripotent stem cells. Nature 2011;471(7336):68-73.

12. Martin GR. Isolation of a pluripotent cell line from early mouse embryos cultured in medium conditioned by teratocarcinoma stem cells. Proc Natl Acad Sci U S A 1981;78(12):7634-8.

13. Okabe S, Forsberg-Nilsson K, Spiro AC, Segal M, McKay RD. Development of neuronal precursor cells and functional postmitotic neurons from embryonic stem cells in vitro. Mech Dev. 1996;59(1):89-102.

14. Parviainen R, Auvinen J, Pokka T, Serlo W, Sinikumpu JJ. Maternal smoking during pregnancy is associated with childhood bone fractures in offspring – A birth-cohort study of 6718 children. Bone. 2017;101:202-205.

15. Provot S, Schipani E, Wu J, Kronenberg H. Development of the Skeleton. In: Osteoporosis (Fourth Edition). Elsevier, 2013:97-126.

16. Reya T, Morrison SJ, Clarke MF, Weissman IL. Stem cells, cancer, and cancer stem cells. Nature 2001;414(6859):105-11.

17. Stadtfeld M, Hochedlinger K. Induced pluripotency: history, mechanisms, and applications. Genes Dev 2010;24(20):2239-63.

18. Stadtfeld M, Hochedlinger K. Induced pluripotency: history, mechanisms, and applications. Genes Dev. 2010 Oct 15;24(20):2239-63.

19. Takahashi K, Tanabe K, Ohnuki M, Narita M, Ichisaka T, Tomoda K, Yamanaka S. Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell 2007;131(5):861-72.

20. Takahashi K, Yamanaka S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 2006;126(4):663-76.

21. Tandon S, Jyoti S. Embryonic stem cells: An alternative approach to developmental toxicity testing. J Pharm Bioallied Sci. 2012;4(2):96-100.

22. Thomson JA, Itskovitz-Eldor J, Shapiro SS, Waknitz MA, Swiergiel JJ, Marshall VS Jones JM. Embryonic stem cell lines derived from human blastocysts. Science 1998;5391(282):1145-1147.

23. Wilmut I, Schnieke AE, McWhir J, Kind AJ, Campbell KH. Viable offspring derived from fetal and adult mammalian cells. Nature 1997;385(6619):810-3.

24. Zhang P, Zucchelli M, Bruce S, Hambiliki F, Stavreus-Evers A, Levkov L, Skottman H, Kerkelä E, Kere J, Hovatta O. Transcriptome profiling of human pre-implantation development. PLoS One. 2009;4(11):e7844.

25. Zhang P, Zucchelli M, Bruce S, Hambiliki F, Stavreus-Evers A, Levkov L, Skottman H, Kerkelä E, Kere J, Hovatta O. Transcriptome profiling of human pre-implantation development. PLoS One 2009;4(11):e7844.

26. zur Nieden NI, Kempka G, Ahr HJ. In vitro differentiation of embryonic stem cells into mineralized osteoblasts. Differentiation. 2003;71(1):18-27.

27. zur Nieden NI, Kempka G, Rancourt DE, Ahr HJ. Induction of chondro-, osteo- and adipogenesis in embryonic stem cells by bone morphogenetic protein-2: effect of cofactors on differentiating lineages. BMC Developmental Biology 2005;5:1.

28. zur Nieden NI, Price FD, Davis LA, Everitt RE, Rancourt DE. Gene profiling on mixed embryonic stem cell populations reveals a biphasic role for beta-catenin in osteogenic differentiation. Molecular Endocrinology 2007;21(3):674-85.

CHAPTER 2

Low Osteogenic Yield in Human Pluripotent Stem Cells Associates with Differential

Neural Crest Promoter Methylation

Nicole Renee Lee Sparks, Ivann Kenneth Carvajal Martinez, Cristina Helen Soto, Nicole

Isolde zur Nieden

Abstract

Human pluripotent stem cell‐derived osteoblasts possess great potential for utilization in bone disorder elucidation and repair; yet, while the general ability of human pluripotent stem cells to differentiate into osteoblasts and lay down bone‐specific matrix has been shown, previous studies lack the complete characterization of the process whereby such osteoblasts are derived as well as a comparison between the osteogenic efficiency of multiple cell lines. Here, we compared the osteogenic potential of two human induced pluripotent stem cell lines (RIV9 and RIV4) to human H9 embryonic stem cells. Generally capable of osteogenic differentiation, the overall osteogenic yield was lower in the RIV9 and RIV4 lines and correlated with differential expression of osteocalcin (OCN) in mature cultures and PAX7 and TWIST1 during early differentiation. In the undifferentiated cells the promoters of the latter two genes were differentially methylated potentially explaining the variation in differentiation efficiency. Furthermore, the expression signatures of selected neural crest and mesodermal genes and proteins suggested that H9 cells

14 preferentially gave rise to neural crest‐derived osteoblasts, whereas the osteoblasts in the

RIV9 cultures were generated both through a mesodermal and a neural crest route albeit each at a lower rate. These data suggest that epigenetic dissimilarities between multiple

PSC lines may lead to differences in lineage derivation and mineralization. Since osteoblast progenitors from one origin inadequately repair a defect in the other, these data underscore the importance of screening human pluripotent stem cells lines for the identity of the osteoprogenitors they lay down.

Introduction

The in vitro differentiation of human embryonic stem cells (hESCs) offers an invaluable source to study cellular development, the etiology of degenerative diseases, and the regeneration of dysfunctional tissues [1‐4]. Human ESCs are exceptionally appropriate for the in vitro generation of specialized cells due to their pluripotency. Researchers have developed in vitro protocols that differentiate ESCs from mouse [5‐7] and human [79] into functional osteoblasts that can mineralize their extracellular matrix (ECM), a hallmark of bone formation. Directed differentiation of ESCs towards an osteoblast lineage has been achieved using dexamethasone (DEX), which triggers the expression of the osteogenic genes runt‐related transcription factor 2 (RUNX2) and osteocalcin (OCN) [10, 11]. Our group routinely uses the active exogenous factor 1α,25 dihydroxy vitamin D3 (VD3) to enhance commitment of mouse or primate ESCs towards the osteoblast lineage [6, 7, 12].

However, the ethical concerns surrounding the use of hESCs have held back exploring the great expectations otherwise associated with hESCs. Human induced

pluripotent stem cells (hiPSCs), generated from the reprogramming of somatic cells, are

also truly pluripotent cells and provide an alternative mean to obtain ES‐like cells [13,14].

While believed to be broadly equivalent to hESCs based on morphology and gene

expression [1517], several reports have identified hiPSC lines to be different from hESCs

in DNA methylation and gene expression signatures [18‐21].

Increasingly, the capacity of iPSCs to differentiate into osteoblasts is at the center

of studies, as these cells may be used in cell‐based replacement therapies or embryotoxicity

screening assays [22, 23]. To direct osteogenic differentiation in hiPSCs prior studies

employed DEX addition to the culture medium or seeding these cells on three‐dimensional

(3D) scaffolds [24‐32]. While these studies have provided evidence that hiPSCs are

generally capable of generating osteoblasts they have not characterized whether they do so

with a similar efficiency as hESCs and whether they do so through the same process. This

is important, because osteoblasts may originate from the neural crest or the mesoderm, the

former having superior grafting abilities than the latter [33, 34]. Similarly, a culture process

which derives osteoblasts from the neural crest may not be suitable to uncover

embryotoxicants that cause limb malformations in vivo.

As it is thus clearly necessary to elucidate the efficacy of hiPSC technology prior to routine experimental utilization of these cells [35], here we compared the osteogenic differentiation potential of two hiPSC lines, RIV9 and RIV4, with hESCs of the H9 line.

Although RIV9 and RIV4 cells were generally capable of differentiating into functional matrix‐calcifying osteoblasts when induced with VD3, both of the hiPSC lines tested exhibited a lower potential to yield osteoblasts than the H9 cells as determined by bone‐

16 specific mRNA expression and quantification of calcium accumulation. Despite their lower differentiation potential, osteogenically induced RIV9 cells were equally sensitive to the strong embryotoxicant 5‐Fluorouracil indicating their potential usefulness for the screening and identification of putative skeletal toxicants.

However, in H9 hESCs, VD3 induction seemed to cause osteoblasts to form from a

source of neural crest cells, while RIV9 osteoblasts seemed to be specified primarily from

mesoderm and to a lesser degree the neural crest, whereas RIV4 cells seemed to prefer a

neuroepithelial path. Differentiation efficiency and route was not inherently associated

with the type of pluripotent stem cell, but rather seemingly correlated with the methylation

state of neural crest promoters that was found before differentiation was initiated. In

summary, our results indicate that before individual hPSC lines can come to use in

orthopedic regenerative medicine or for the study of molecular mechanisms accompanying

osteogenic differentiation, they need to be thoroughly characterized for the lineage origin

they derive osteoblasts from.

Materials and Methods

Human pluripotent stem cell lines

Cells of the hESC H9 line were acquired from WiCell (WiCell Research Institute).

RIV4 and RIV9 hiPSCs were generated from foreskin fibroblasts (HFF‐1, ATCC) by retroviral integration of OCT4, KLF4 and SOX2 at the University of California Riverside’s

Stem Cell Core Facility [36, 37]. To disrupt expression of T/Brachyury, human pluripotent stem cells (4 x 105 cells) were transfected with 0.3 µg brachyury CRISPR/Cas9 KO plasmid

(h; sc‐416539) and 0.3 µg brachyury HDR plasmid (h; sc416539‐HDR) from Santa Cruz using Effectene Transfection Reagent (Qiagen 301427). After 72 hours, transfectants were selected with 150 µg/ml puromycin (Sigma‐Aldrich P8833) for 3‐5 days. Resistant colonies that were double positive for green and red fluorescent protein were picked and expanded for analysis. Osteogenic differentiation was conducted from those clones, in which a reduction in BRACHYURY protein expression during differentiation was confirmed. Control clones were generated using 0.3 µg Control CRISPR/Cas 9 plasmid.

All cells were maintained on Matrigel (BD Biosciences) treated dishes in mTeSR® (Stem

Cell Technologies) in 5% CO2 and at 37°C. Pluripotent colonies were passaged every 5 days upon reaching 70% confluency by dissociating cells with accutase and a cell scraper.

Karyotyping

Standard G‐banding type chromosome and cytogenetic analysis were carried out at

the WiCell Research Institute.

Osteogenic Differentiation of hPSCs

Confluent pluripotent colonies (designated day 0) were switched to control

differentiation medium composed of Dulbecco’s Modified Eagle’s Medium (DMEM,

Gibco) 15% FBS (Atlanta), 1% non‐essential amino acids (NEAA; Gibco), 1:200

penicillin/streptomycin (Gibco), and 0.1 mM β‐mercaptoethanol (Sigma). On day 5,

control medium was supplemented with 10 mM β‐glycerophosphate (Sigma), 50 µg/ml

ascorbic acid (Sigma), and 50 nM 1,25(OH)2 Vitamin D3 (Calbiochem) [38].

Immunocytochemistry

Cells were fixed with 4% paraformaldehyde for 30 min at 4°C. After washing and permeabilizing (intracellular markers only) with 0.1% Triton‐X 100 (Sigma) for 15 min, cells were incubated in primary antibody diluted in PBS/10% FBS overnight at 4°C.

Primary antibodies used were POU5F1 (EMD Millipore, ab3209), NANOG (Cell

Signaling 4903), SOX2 (Cell Signaling, 3579), SSEA4 (EMD Millipore, MAB4304), OCN

(Abcam, ab1857), p75NTR (Abcam, ab8874), SOX10 (Santa Cruz, SC17342), TWIST1

(Abcam, ab50887), PDGRFα (Abcam, ab61216), and PAX7 [39] (mIgG1, Developmental

Studies Hybridoma Bank). Cells were then washed and incubated in the appropriate secondary antibody (AlexaFluor conjugated) for 2h. Cells were observed on a Nikon fluorescent microscope.

Histochemical analysis

Cultures were washed with PBS and fixed in 4% paraformaldehyde at 4°C for 30 min. For von Kossa stain, fixed samples were overlaid with silver nitrate solution (Ricca

Chemical Company) and illuminated under an intense light source for 1 h. Cultures were fixed with 5% sodium thiosulfate (Red Bird Service) for 2 minutes. Cells were also washed with PBS and overlaid with 2% w/v Alizarin Red S solution (Sigma‐Aldrich) for 5 minutes to visualize calcium‐rich areas. Prior to imaging, cultures were washed 3X with water followed by ascending alcohol washes (70%, 80%, 90%, 100% ethanol).

Characterization of bone: Alkaline phosphatase (ALP), inorganic phosphate, and calcium

assays

Cells were harvested in radioimmunoprecipitation (RIPA) buffer. Activity of the

ALP enzyme and calcium deposition were determined and calculated as previously

described [40]. Free inorganic phosphate (Pi) was measured with the PiPer phosphate assay

(Molecular Probes, Invitrogen) per the protocol of the manufacturer and as used by

Majumdar et al. [41]. The total free inorganic phosphate in the sample was taken from a

potassium phosphate standard curve. All measurements were normalized to the total

protein content of the sample measured with the Lowry method.

RNA isolation and quantitative PCR

Total RNA was extracted with the NucleoSpin RNA II protocol (Macherey Nagel).

RNA concentration was determined with a NanoDrop® 1000 spectrophotometer (Thermo

Scientific) at 260 nm. 25 ng of total RNA was used for cDNA synthesis with a mastermix including 1X reaction buffer, 0.5 mM dNTPs, 20 U/µL RNase inhibitor, 0.8 U/µL reverse transcriptase, and 1 μM random primer. 25ng cDNA transcripts were used for quantitative polymerase chain reaction (qPCR) with SYBR green on the MyiQ cycler (BIO‐RAD).

Reactions were setup for 10 minutes of denaturing at 94°C, followed by 40 cycles of denaturing at 94°C, and annealing/elongation at 60°C for 45 seconds. The n‐fold expression in target samples was calculated with the ΔΔCT method by standardizing CT values to GAPDH expression [42]. Primer sequences used in this study can be found in

Appendix Table 1.2.1. Pre‐designed primer pairs for GOOSECOID (GSC;

Hs00906630_g1), PAX6 (Hs00240871_m1), and SOX17 (Hs00751752_s1) were purchased from Life Technologies. All other primers were from Leung et al., 2016 [43].

Methylation‐specific (MS) PCR Analysis

To analyze the methylation status of the TWIST1 and PAX7 promoters, cells were

collected in lysis buffer (100 mM NaCl, 0.5% SDS, 20 mM Tris pH7.5, 50 mM EDTA)

with 0.2 mg of proteinase K (Sigma) and incubated at 60°C overnight. DNA was isolated

by standard phenolchloroform and isopropanol precipitation methods and quantified with

a NanoDrop® 1000 spectrophotometer. Bisulfite conversion was performed from 200 ng

of genomic DNA with the EZ DNA Methylation‐Lightning Kit (Zymo). The methylation

status of the PAX7 and TWIST1 promoters was then determined from the eluted DNA

using the MS‐nested PCR method [44]. Primer sequences for first and second rounds of

amplification for both methylated and unmethylated DNA are listed in Appendix Table

1.2.1 and are as previously published [44]. Cycling conditions for PAX7 were 10 minutes

at 95°C, followed by 40 cycles at 94°C for 25 seconds, 60°C for 25 seconds, and 72°C for

20 seconds. A final elongation of 2 minutes at 72°C followed. The first and second

reactions for TWIST1 were setup for 10 minutes at 95°C, followed by 40 cycles for round

one and 32 cycles in round 2 at 94°C for 25 seconds, 63.9°C for 25 seconds, 72°C for 20

seconds, and a final 2 minute elongation at 72°C. Amplified products were observed in a

2% agarose gel using ethidium bromide for detection.

MTT assay

Cell viability was determined with 3‐[4,5dimethylthiazol‐2‐yl]‐2,5‐

diphenylterazolium bromide (MTT) as described previously [45] and absorbance was read

at 595 nm in an iMark™ microplate reader (BIORAD).

Video bioinformatics

Phase contrast microscopic images of differentiating hPSCs were automatically

acquired at a magnification of 4X every 12h utilizing the BioStation CT (Nikon

Corporation, Tokyo, Japan). For each 24‐well dish well ten reference points of interest

were randomly selected as representative view fields. Raw images were processed,

enhanced and segmented using MATLAB software generated recipes (Mathworks, MA,

U.S.A) using a manual threshold of 33. Pixels below the threshold were removed to create

segmented images of calcified regions. Remaining pixels were counted to quantitatively

represent the degree of calcification from each image. The calcification rate was

determined from the number of calcified pixels in each image at tx subtracted from the

number of calcified pixels at t0 and divided by the hours of elapsed time.

Toxicity Testing

5‐Fluorouracil (5FU) was obtained from Sigma‐Aldrich and a stock solution prepared in dimethyl sulfoxide. Differentiation of hPSCs was initiated as described above and cells were treated for 20 days of osteogenesis with five concentrations of 5FU and a solvent only control. 5FU was replenished with each media change.

Statistical Analysis

All experiments were performed from 3 independent passages, each consisting of either three or five biological replicates (different wells). Video bioinformatics assessment was performed from a total of 30 areas within three independent culture wells per replicate.

For single and multiple comparisons, an unpaired student’s T‐test and a One‐Way ANOVA with Holm‐Sidak post‐hoc analysis was performed, respectively. A p-value of 0.05 or smaller was considered statistically significant (SigmaPlot, Systat Software, San Jose, CA).

Results

Pluripotency assessment of cell lines used

To apply hiPSC technology in the cell‐based replacement of diseased or injured bone tissue or the screening for skeletal toxicity and teratogenicity, the osteogenic differentiation potential of human induced pluripotent stem cells was first determined. To rule out that eventual differences in osteogenic differentiation capacity stemmed from differences in their pluripotency status, H9 hESCs and RIV4 and RIV9 hiPSCs were first compared for expression of pluripotency associated markers and absence of those associated with germ layer differentiation.

All lines displayed similar morphology and both hiPSC lines showed a normal karyotype (Fig. 2.1A, B). In the undifferentiated state, RIV9 and RIV4 lines expressed

POU5F1, SOX2, and NANOG transcription factors, which localized to the nucleus, and the membrane associated stage‐specific embryonic antigen 4 (SSEA4; Fig. 2.1C). A quantitative comparison of pluripotency marker mRNA levels showed no statistical

23 difference between the RIV and the H9 lines (Fig. 2.1D). Upon withdrawal of pluripotency conditions, all lines up‐regulated the expression of tri‐lineage genes (Fig. 2.1E‐G), which were found at very low levels in the undifferentiated cells, concomitant with a reduction in pluripotency mRNAs with increased time in differentiation conditions (DMEM+FBS).

These data indicated that all lines used in this study showed general tri‐lineage potential facilitating the assessment of their osteogenic differentiation potential.

Figure 2.1. Cytogenetic, pluripotent, and genetic characterization of human induced pluripotent stem cell (hiPSC) lines. (A): Morphology of H9 human embryonic stem cells (hESCs) and RIV9 and RIV4 hiPSCs. (B): Cytogenetic G‐banding analysis for RIV9 and RIV4 hiPSCs (46, XY). (C): Immunofluorescence staining for pluripotent markers. Insets: Nuclear counterstaining with 4′,6‐diamidino‐2‐phenylindole. (D): Quantitative polymerase chain reaction (qPCR) for mRNA expression of pluripotency and lineage markers on undifferentiated hPSCs. Expression is normalized to H9 hESCs and standardized to GAPDH, n = 3 independent samples ± SD. (E–G): qPCR mRNA expression of the same set of mRNAs on differentiating hPSCs, normalized to GAPDH, n = 3 independent samples ± SD. *p <0.05, Student's t test compared with expression in undifferentiated cells (mTeSR). Abbreviations: DAPI, 4′,6‐diamidino‐2‐phenylindole; GAPDH, glyceraldehyde 3‐phosphate dehydrogenase; T/BRA, T‐brachyury.

Characterization of osteoblast cells derived from hESCs

For osteogenic induction, we adopted a monolayer overgrowth culture approach previously described by Karp et al. [10] (for schematic see Fig. 2.2A) after attempts to execute an embryoid body‐based protocol failed due to excessive cell death after replating

Appendix Fig. 1.2.1A). We selected VD3 as an osteogenic supplement as it led to high levels of alkaline phosphatase (ALP) enzyme activity and superior calcification over DEX induction in pilot experiments (Appendix Fig. 1.2.1B).

Upon VD3 induction, differentiating ESCs showed black appearing deposits, which represent the calcification of the extracellular matrix, a hallmark of bone development from stem cells [46] (Fig.

2.2B). Cultures were positive for the major non‐collagenous bone protein OCN and Figure 2.2. Osteogenic differentiation of human embryonic stem cells. (A) Schematic representation of the osteogenic induction protocol. (B) Immunostaining with an osteocalcin antibody reveals presence of positive deposition of bone matrix in the vicinity of black deposits. Alizarin red S staining detected calcium ions in the same areas. Scale bars = 100 µm. (C) Quantitative measurement of deposited calcium over time, n = 5 independent replicates ± SD, calcium ions was *p <0.05 Student's t test compared with d0. (D) Quantitative mRNA analysis of RUNX2, SATB2, and OCN, normalized to GAPDH (n = 3 independent depicted by Alizarin samples ± SD), *p<0.05 one‐way analysis of variance over day something missing here. Triplicates were pooled for reverse transcriptase polymerase chain reaction analyses of the osteocyte genes CAPG and DESTRIN (inset). Abbreviations: Aliz Red S and Von Kossa S, alizarin red S; DAPI, 4′,6‐diamidino‐2‐phenylindole; ESC, embryonic stem cell; GAPDH, glyceraldehyde 3‐phosphate dehydrogenase; OCN, osteocalcin.

25 stains (Fig. 2.2B). Quantitatively, the levels of deposited calcium rose over non‐induced cultures by 10 days of the in vitro differentiation and had reached levels that were 78‐fold higher than the non‐osteogenic cells by day 30 (Fig. 2.2C). At late stages of osteogenesis, osteoblast‐specific expression of SATB2, RUNX2 and OCN transitioned into osteocyte specific expression of CAPG and DESTRIN [48] (Fig. 2.2D).

During early differentiation, the co‐expression of SOX10/PAX7 and

SOX10/TWIST suggested the presence of neural crest cells [39] (Fig. 2.3A), while these were absent in uninduced control cultures. Further, p75NTRpositive cells, often associated with migrating cranial neural crest cells [39, 49], were more abundant than in spontaneously differentiating cultures, and a few days later cells became positive for

PDGFRα, which associates with mesenchymal cells, a subsequent step in maturation (Fig.

2.3B). In line with this observation the expression of PAX7 and TFAP2α, two neural plate border specifier genes [48‐51], were over 470‐ and 3160‐fold up‐regulated over undifferentiated cells (Fig. 2.3C). Mimicking the timely expression during neural crest development, SOX10 and ZIC3 were expressed slightly later. Subsequently, DLX6, which marks osteoprogenitors from the neural crest was significantly up‐regulated over undifferentiated cells (Fig. 2.3C). In contrast, T/Brachyury (T/BRA, a pan‐mesodermal gene) [52], peaked simultaneously with PAX7 on day 5 of differentiation (Fig. 2.3D), but remained ~140‐fold lower than that of PAX7. PAX7 expression was also 98‐ and 49‐fold greater than the levels of the neuroepithelial SOX1 and PAX6, respectively (Fig. 2.3E), suggesting that hESCs primarily differentiated through a neural crest route.

Figure 2.3. Intermediate differentiation in H9 cells is through the neural crest lineage. (A, B) Double staining with antibodies against SOX10/PAX7, SOX10/TWIST1, and single stains with p75NTR, and platelet‐derived growth factor receptor α at indicated time points of differentiation. Scale bar = 100 µm. (C–E) Quantitative mRNA analysis of early lineage genes, normalized to GAPDH, n = 3 independent samples ± SD. Abbreviations: DAPI, 4′,6‐diamidino‐2‐phenylindole; GAPDH, glyceraldehyde 3‐phosphate dehydrogenase; PDGFR α, platelet‐derived growth factor receptor α; T/BRA, T‐brachyury.

Osteogenic Analysis of hiPSCs

We next tested whether the RIV9 and RIV4 cells also responded to VD3 with osteogenic differentiation. On day 30 of the culture process, both the RIV9 and RIV4 osteogenic cultures indeed displayed the characteristic black deposits in brightfield microscopy indicative of the mineralized extracellular matrix produced by osteoblasts (Fig.

2.4A). This was further supported by positive immunostaining for OCN, Alizarin Red S, and von Kossa stains. Both RIV lines also showed measurable levels of inorganic phosphate and calcium (Fig. 2.4B, C). While these data suggested that all tested lines were generally capable of differentiating into osteoblasts, we noted an inconsistency in the mineralization process between H9, RIV9 and RIV4 during our studies. For example, inorganic phosphate was significantly decreased in the RIV4 cell line compared to the H9 cells (Fig. 2.4C, P<0.05), while both RIV4 and RIV9 lines showed less calcium content on day 20 (P<0.05) by multiple decades (Fig. 2.4B). Consequently, the Ca/P ratio, an important parameter in the determination of bone quality [53], was lower in both tested hiPSC lines, not even reaching values of 0.58 associated with fetal bone [53] (Fig. 2.4D).

Only the Ca/P ratio for the H9 cells was determined at a close to physiological 2.6:1 [53‐

56].

We next examined when this difference first manifested itself. Morphological assessment suggested that both RIV4 and RIV9 cells calcified less than the H9 cells (Fig.

2.4D), and it appeared as if black deposit was visible earlier than in the H9 cells.

Quantitative calcium measurements revealed significantly more calcification in RIV9 hiPSCs than in H9 hESCs on day 5, a pattern that reverted by day 15 (Fig. 2.4E). Instead, the inorganic phosphate content did not show any discernable pattern over a 20‐day period

(Fig. 2.4E).

To quantitatively determine this apparent difference in kinetics, we applied an innovative time‐lapse morphometric assessment taking advantage of the black appearance of the calcium deposits in images [47]. Images of differentiating cultures were acquired

28 every 12 hours over the course of 20 days and image segmentation applied to separate black pixels from the background. The remaining black pixels were used as a measure for calcification. Both RIV lines exhibited fewer dark pixels confirming a lower degree of calcification in the hiPSC lines (Fig. 2.4F). Calculating the calcification rate by charting the addition of black pixels as a function of elapsed time revealed that RIV4 and RIV9 appeared to calcify earlier than H9 cells (Fig. 2.4G). Together these data imply that hiPSC differentiations varied in the rate that calcification occurred at, with considerably lower osteogenic potential in the RIV4 and RIV9 lines.

Passage‐to‐passage analysis of osteogenic hPSC cultures

We next aimed to elucidate whether the noted variability in differentiation yield was due to differences in differentiation potential between passages or truly inherent in the specific line tested. To examine this variability more closely, we next compared three experiments inoculated from subsequent passages of cells (denoted by Roman numerals I‐

III) for osteogenic yield and levels of expression of genes associated with different levels of culture/osteoblast maturation.

Figure 2.4. Osteogenic differentiation in RIV9 and RIV4 human induced pluripotent stem cells (hiPSCs). (A) Photomicrographs of unstained and stained osteogenic cultures from H9 human embryonic stem cells and RIv9 and RIV4 hiPSCs on day 30 of differentiation. Matrix mineralization detected with OCN, Aliz S, and Von Kossa stains. (B) The 20‐day‐old osteogenic hPSCs deposit significant levels of calcium ions over spontaneously differentiating cultures (n = 5 independent samples ± SD). *p<0.05 one‐way analysis of variance (ANOVA) over spontaneous differentiations. (C) Levels of inorganic phosphate levels measured with the PiPer assay on day 20 content, n = 5 independent samples ± SD, *p<0.05 one‐way ANOVA over H9. Bottom graph depicts calcium:phosphate ratios determined from Ca2+ and PiPer assays, day 20, n = 3 independent replicates ± SD. *p<0.05 one‐way ANOVA versus H9. (D) Morphological time course of hPSC osteogenic differentiation reveals differences in calcification yield between different hPSC lines. (E) Calcification time course analysis shows heterogeneity between osteogenesis in different hPSC lines. Levels of inorganic phosphate also vary depending on differentiation day; n = 3 independent replicates ± SD. #p<0.05 RIV9‐H9; *p<0.05 RIV4‐ H9, one‐way ANOVA. (F, G) RIV9 and RIV4 cultures begin their calcification process earlier than H9 cells as determined by video bioinformatics analysis, n = 10 technical replicates from three biological replicates ± SEM. Abbreviations: Aliz S, alizarin red S; DAPI, 4′,6‐diamidino‐2‐phenylindole; OCN, osteocalcin.

The overall yield in calcium deposit followed a similar pattern of continuous up‐

regulation in all three cell lines but was significantly lower in both RIV lines

(H9>RIV9>RIV4, Fig. 2.5A). At the gene expression level, consistent patterns of

expression that correlated between the three cell lines and with the calcification data were

found for OCN, PAX7 and TWIST1 with the latter inversely correlating to calcification

(Fig. 2.5B‐D).

In contrast, no other examined genes exhibited any clear time‐ or cell‐line dependent pattern of gene expression, although RUNX2 and OPN indicated continuous maturation over time in the H9 cultures despite one out of the three passages under‐ expressing when compared to the other two (Appendix Fig. 1.2.2A, B). Similarly, no time‐ dependent pattern could be observed for ALP enzyme activity or inorganic phosphate content (Appendix Fig. 1.2.2E, F).

To test whether the observed differences between the three cell lines were correlative to passage variability, we next calculated the n‐fold differences in gene expression, mineral content and enzyme activity across all time points and graphed them as box plots Appendix Fig. 1.2.3). These indicated that although some of the endpoints showed higher variability in the RIV9 cells, most of the endpoints exhibited equal variations in all cell lines, and none correlated with overall osteogenic yield.

Figure 2.5. Passage‐specific analysis of bone parameters and gene expression. (A) Calcium content in hPSC cultures inoculated from three different passages (I–III). Data points represent five technical replicates for each biological passage shown ± SD. *p<0.05 passage II over I; #p<0.05 passage III over I; Δp<0.05 passage III over II. (B–D) The mRNA expression of osteogenic genes was determined with quantitative polymerase chain reaction and normalized to GAPDH, n = 3 ± SD. RIV9 and RIV4 human induced pluripotent stem cells failed to show definitive osteogenic patterns. *p <0.05 passage II over I; #p <0.05 passage III over I; Δp<0.05 passage III over II, one‐way analysis of variance. Abbreviations: GAPDH, glyceraldehyde 3‐phosphate dehydrogenase; OCN, osteocalcin.

Calcification ability of hPSCs correlates with PAX7 and TWIST1 expression and

methylation status

Based on the association of PAX7 and TWIST1 expression measured in H9 ESCs

(Fig. 2.2G) with neural crest lineage determination, and the correlative calcification kinetics found for H9 ESCs, we next comparatively measured the difference in mRNA expression patterns of multiple genes associated with the neural crest as well as mesodermal lineage across the three cell lines. PAX7, TFAP2a, ZIC3 and MSX2 mRNA levels were significantly lower in both RIV lines than in H9 cells (Fig. 2.6A). In contrast, the mesodermal genes T/BRA, TBX6, MIXL1 and MIER1 were expressed in the RIV9 line at 2‐, 37‐, 2‐ and 18‐fold the levels found in H9 cells, respectively (Fig. 2.6A).

Contrastingly, RIV4 cultures had lower levels than RIV9 of most of the neural crest and mesodermal genes, but highly expressed SOX1 by 14‐fold over the H9 line (Appendix Fig.

1.2.4) potentially suggesting that their differentiation was halted at a neuroepithelial state

[57]. H9 cultures with a CRISPR/Cas9‐mediated disruption of T/BRA did not exhibit a reduction in calcification, while disruption of the same gene in RIV9 cells lessened the degree of calcification (Fig. 2.6B). This result potentially suggested that osteogenesis occurred independently of T/BRA in H9 cells but was dependent on a mesodermal transition state in RIV9 cells.

Given that iPSCs have been reported to bear epigenetic marks from their cell type

of origin [19] and given the high correlation between PAX7 and TWIST1 expression with

the calcification yield in the H9 line over the RIV9 line we next examined the methylation

status of both promoters. Methylation specific PCR supported the notion that the PAX7

Figure 2.6. Osteoblast origin correlates with methylation of the TWIST1 and PAX7promoters. (A) mRNA expression for genes associated with mesoderm or the neural crest was quantitatively analyzed with quantitative polymerase chain reaction (qPCR), n = 3 independent samples ± SD. *p<0.05 one‐ way analysis of variance (ANOVA) over H9. (B) Quantification of calcium content in control cultures of H9 and RIV9 cells and two clones each with a CRISPR–Cas9‐mediated disruption of T/BRACHYURY (T/BRACRISPR). Values represent means of three independent experiments, n = 5 technical replicates each ± SD. *p<0.05 one‐way ANOVA versus the respective controlCRISPR cell line. (C) Methylation‐specific PCR showed that TWIST1 was unmethylated (U) in the RIV9 line. The PAX7 promoter region was methylated (M) in the RIV9 line. (D) Quantitative analyses of time course mRNA expression in the RIV9 and H9 cells, normalized to GAPDH, pooled from n = 3 independent replicates ± SD. #p<0.05 RIV9‐H9; *p<0.05 RIV4‐H9, one‐way ANOVA. Abbreviations: GAPDH, glyceraldehyde 3‐phosphate dehydrogenase; GSC, goosecoid; M, methylated; T/BRA, T‐ brachyury; U, unmethylated. promoter region was hypomethylated in the H9 cells, while it was methylated in the RIV9 hiPSCs (Fig. 2.6C). This methylation pattern suggested that the activation of this locus was hindered in the RIV9 cells, which was reflected in the PAX7 mRNA levels. In turn, the evaluation of the TWIST1 promoter methylation status, a gene expressed by neural crest

cells once they have committed to a mesenchymal fate, revealed high methylation of the

TWIST1 promoter in the H9 line, whereas TWIST1 was less methylated in the RIV9 cell

line (Fig. 2.6C). In vivo, TWIST1 mediates cranial bone progenitor specification [58], and

is highly expressed in mesenchymal cells and then down‐regulated for further

differentiation to occur [59, 60]. The detected methylation status on the TWIST1 promoter

suggested that in the RIV9 cells TWIST1 was turned on prematurely, thwarting its proper

activation pattern during differentiation. Indeed, already in undifferentiated RIV9 cells

TWIST1 was expressed 2.2–fold higher over the H9 hESCs (Fig. 2.6D) and increased with

faster kinetics.

Next, we additionally analyzed the PAX7 and TWIST1 methylation status over a selection of different hESC and hiPSC lines and found differential methylation states that did not correlate with hESC or hiPSC identity (Appendix Fig. 1.2.5). However, the methylation status of these two promoters in the undifferentiated state did seem to be predictive of the osteogenic differentiation capacity of the respective lines. Moreover, the methylation status also predicted the ability of two karyotypically abnormal RIV lines,

RIV1 and RIV7, independently of their karyotypic abnormalities (Appendix Fig. 1.2.5B,

E). Together, these findings suggested that both PAX7 and TWIST1 methylation status may have a determining role in osteoblast development in PSCs.

RIV9 hiPSCs are able to predict skeletal teratogenesis despite their low osteogenic yield

While a low osteogenic differentiation potential of a pluripotent stem cell line could be of concern for its clinical application in cell‐based regenerative medicine, it may still

35 prove useful for the in vitro screening of putative skeletal toxicants despite its diminished osteogenic yield. To test this, we next examined the potential of the RIV9 line to accurately predict the toxicity of the strongly embryotoxic agent 5‐Fluorouracil (5FU) [62] by assessing its inhibition of differentiation compared to its cytotoxicity [45]. The half‐ maximal inhibitory concentrations (ID50) in the endpoint calcium content were not statistically different between the RIV9 and the H9 cells (P=0.090), neither were the half‐

Figure 2.7. Toxicity of 5‐fluorouracil (5FU) tested with H9 and RIV9 hPSCs. (A) Table of half‐maximal inhibitory doses for cytotoxicity (IC50) as determined from concentration‐response curves with 3‐[4,5‐dimethylthiazol‐2‐yl]‐ 2,5‐diphenylterazolium bromide assay and differentiation (ID50) with calcium assay. (B) Linear regression of RIV9 half‐maximal inhibition of differentiation graphed against H9 half‐maximal inhibition of differentiation. Both cell lines showed similar classification in predicting skeletal toxicity of 5FU. Abbreviations: 5FU, 5‐fluorouracil; hESC, human embryonic stem cell; hiPSC, human induced pluripotent stem cell; MTT, 3‐[4,5‐dimethylthiazol‐2‐yl]‐2,5‐ diphenylterazolium bromide.

36 maximal inhibitory concentrations of cytotoxicity (IC50) (P=0.172) (Fig. 2.7A). Plotting the measured reduction in calcium content for every tested concentration in 5FU‐treated osteogenic RIV9 cultures against the respective result found with the H9 cells (Fig. 2.7B) resulted in correlation coefficients close to 1, further demonstrating that RIV9 cells exhibit the same sensitivity to 5FU as H9 cells. These results potentially proved the usefulness of

PSC osteogenic cultures in the prediction of skeletal toxicants, despite different

differentiation kinetics or lower efficiencies.

Discussion

In line with previous studies, we show here that hPSC cell lines are generally

capable of osteogenic differentiation. However, to our knowledge this is the first report

that VD3, an osteogenic supplement that acts to trigger osteogenic gene expression and

calcification in mouse and marmoset ESCs [6, 12], can also enhance human ESC

osteogenic differentiation as well as induce osteogenic differentiation of human iPSCs.

Notably, calcification levels were significantly lower in the two tested hiPSC lines than in

the H9 hESC line. Where other studies have shown that osteoblasts can be derived from

iPSCs when grown on scaffolds or with culture additives in general, most of these relied

on qualitative von Kossa and Alizarin Red stains [8, 11]. Because only minimal

quantitative analyses such as ALP and OCN were used in past studies [10], a thorough

evaluation of the differences between hESCs and hiPSCs had not been performed [24, 63].

Using a broad range of techniques, including quantitative PCR, immunocytochemistry, and

innovative video bioinformatics we show here systematically that although osteogenically

differentiating hiPSCs can form a mineralized matrix, the two hiPSC lines tested exhibited

lower osteogenic efficiency than H9 hESCs by multiple decades. While we initially

hypothesized that this difference had to do with the type of cell studied we soon found

evidence that these differences in osteogenic yield were cell type independent.

In contrast to the results presented here, a recently published study on mouse ESCs and two mouse iPSC lines did not identify any differences in osteogenic yield based on global gene expression analysis [73]. However, some data in another previous study hints at a lower osteogenic yield in both mouse and human iPSC lines [64]. Potentially, these discrepancies could be either based on species‐specific differences of how differentiation proceeds or, alternatively, on the fact that some of the tested lines were grown on feeder cells while others were not, which may affect their pluripotency states. Indeed, future studies should examine whether returning all cell lines to naïve pluripotency would equalize the methylation states of the key promoters analyzed here and thus level the osteogenic yield in all lines and potentially across species.

For now, the notably low differentiation yield in some hPSC lines may prove problematic in future applications for regenerative purposes. This is because large numbers of transplantable cells will be needed for any type of tissue repair [65, 66]. Hence, a low differentiation yield would render the already problematic scaleup even more labor‐ intensive. Of highest concern, however, is the fact there have so far been no investigations into the tissue origin of the osteoblasts that hPSCs generate. This is especially puzzling since it is well known that neural crest‐derived frontal osteoblasts display a superior capacity to undergo osseous healing compared with calvarial osteoblasts of paraxial

38 mesoderm origin [67]. In a grafting study, Leucht and colleagues showed that neural crest‐ derived bone would graft and give rise to new bone regardless of whether they were placed into a mandibular or tibial defect [33]. In contrast, tibial bone placed in the mandibular defect, the mesodermal tibial cells made cartilage in the mandible instead of bone. Jheon and Schneider [34] have further suggested that bone of neural crest origin will graft better when placed with its bone tissue of origin.

We suggest here for the first time that in some hPSCs osteoblasts may arise through a different embryonic lineage than in others. Specifically, differential regulation of TWIST1 and PAX7 mRNA expression caused by differential promoter methylation may be the apparent cause for defective neural crest osteogenesis in the RIV9 and RIV4 cell lines, which is otherwise found in H9 cells.

The low levels of T/BRA and PAX7 expression in RIV4 hiPSCs potentially explains the decrease in RIV4 osteogenic yield, while positivity for SOX1 could indicate that most of the RIV4 cells specify to neuroepithelium [57], but then fail to differentiate further. In contrast, PAX7 and SOX10 expression and stage‐appropriate downregulation of TWIST1 mRNA correlate with a neural crest origin of H9‐derived osteoblasts. Interestingly, the

RIV9 cells highly expressed T/BRA, TBX6, MIXL and MIER1 mRNA with decreased

PAX7, ZIC1 and SOX1, indicating a preference for mesodermal osteogenesis in those cells.

We suggest this because mesendodermal differentiation is typically associated with a ~90‐ fold increase in the expression of T/BRA over undifferentiated cells [68], while in our cultures it is up‐regulated only 3.5‐fold. Emerging evidence suggests that presence PAX7 may function to push cells towards a neural crest lineage [49, 50]. Our findings of PAX7

39 gene expression and methylation status between the different human pluripotent stem cell lines coupled with high expression of TFAP2α, ZIC1 and ZIC3 appears to support this notion. We show that H9 cells express higher levels of the PAX7 gene coupled with an unmethylated promoter region. Thus, the RIV9 line could exhibit silencing of this locus through DNA methylation regulation.

We also concentrated on TWIST1 as a candidate for assessing osteogenic potential because TWIST1 seems to regulate the commitment of mesenchymally fated cells to the osteogenic lineage. While high TWIST1 levels maintain mesenchymal stem cells in their unspecified state [60], silencing of TWIST1 enhanced matrix mineralization and osteoblast gene expression in C3H10T1/2 cells [59]. In cranial cells, loss of TWIST1 leads to the maintenance of an epithelial character and reversal of EMT [69, 70]. Here we have revealed a potential role of gene‐specific hypomethylation in limiting osteogenic differentiation of the RIV9 line, which may have caused the early overexpression of TWIST1.

One further indication for a possible alternative osteoblast origin between the H9 and the RIV9 cells was the kinetic calcification analysis that we performed. The video bioinformatic calcification data indicated that calcification arose quicker in the RIV9 line than in the H9 cells, validating findings by Nishikawa and colleagues, who had suggested a timely discrepancy between the appearance of neural crest and mesoderm derived mesenchymal cells, whereby the mesoderm derived mesenchymal progenitors appeared five days earlier than those derived from the neuroepithelium [57].

While the lower osteogenic output of some hPSC lines could mean that the development of an appropriate therapy for skeletal regeneration may require further

40 investigations, this lower yield does not seem to preclude their use in the identification and prediction of the embryotoxicity associated with exposure to a chemical compound.

Researchers are utilizing hPSCs increasingly for developmental toxicity testing, however there is a lack of different tissue endpoints such as bone [71]. Furthermore, the capacity of hiPSCs as a screening tool has not been explored [72]. We show here that hiPSCs may be potentially useful in the in vitro prediction of embryotoxicity even though a specific line may have a lower osteogenic yield.

Conclusion

In summary, our study demonstrates that different hPSC lines, including individual

hiPSC clones, could have inherently different abilities to produce functional osteoblasts

potentially based on the methylation marks placed on promotor regions important for

lineage decisions. This could have an impact on the potential application domain of these

cells. While low differentiation yields may be ameliorated by simple changes to the

culturing protocol, for example EB‐mediated differentiations or alternative supplements,

future studies need to take into account the different capability of individual hPSC lines

and characterize the process by which they lay down osteoblasts. Also, it will be important

to investigate whether mesodermal‐derived osteoprogenitors and hPSC‐derived neural

crest cells repair calvarial and long bone defects equally well.

References

1. Dimos JT, Rodolfa KT, Niakan KK et al. Induced pluripotent stem cells generated from patients with ALS can be differentiated into motor neurons. Science 2008;321:1218– 1221.

2. Mellough CB, Sernagor E, Moreno‐Gimeno I et al. Efficient stage‐specific differentiation of human induced pluripotent stem cells toward retinal photoreceptor cells. Stem Cells 2012;30:673‐686.

3. Davis RP, Casini S, van den Berg CW et al. Cardiomyocytes derived from pluripotent stem cells recapitulate electrophysiological characteristics of an overlap syndrome of cardiac sodium channel disease. Circulation 2012;125:3079‐3091.

4. McClelland‐Descalzo DL, Ehnes DD, zur Nieden NI. Stem cells for osteodegenerative diseases: current studies and future outlook. Regen Med 2014;9:219‐230.

5. Buttery LD, Bourne S, Xynos JD et al. Differentiation of osteoblasts and in vitro bone formation from murine embryonic stem cells. Tissue Eng 2001;7:89–99.

6. zur Nieden NI, Kempka G, Ahr HJ. In vitro differentiation of embryonic stem cells into mineralized osteoblasts. Differentiation 2003;71:18–27.

7. Ding H, Keller KC, Martinez IKC et al. NOβ‐catenin crosstalk modulates primitive streak formation prior to embryonic stem cell osteogenic differentiation. J Cell Sci 2012;125:5564‐5577.

8. Sottile V, Thomson A, McWhir J. In vitro osteogenic differentiation of human ES cells. Cloning Stem Cells 2003;5:149–155.

9. Bielby RC, Boccaccini AR, Polak JM et al. In vitro differentiation and in vivo mineralization of osteogenic cells derived from human embryonic stem cells. Tissue Eng 2004;10:1518‐1525.

10. Karp JM, Ferreira LS, Khademhosseini A et al. Cultivation of human embryonic stem cells without the embryoid body step enhances osteogenesis in vitro. Stem Cells 2006;24:835–843.

11. Bielby RC, Pryce RS, Hench LL et al. Enhanced derivation of osteogenic cells from murine embryonic stem cells after treatment with ionic dissolution products of 58S bioactive solgel glass. Tissue Eng 2005;11:479‐488.

12. Trettner S, Findeisen A, Taube S et al. Osteogenic induction from marmoset embryonic stem cells cultured in feeder‐dependent and feeder‐independent conditions. Osteoporosis Int 2014;25:1255‐1266.

13. Yu J, Hu K, Smuga‐Otto K et al. Human induced pluripotent stem cells free of vector and transgene sequences. Science 2009;324:797–801.

14. Kim D, Kim CH, Moon JI et al. Generation of human induced pluripotent stem cells by direct delivery of reprogramming proteins. Cell Stem Cell 2009;4:472–476.

15. Yu J, Vodyanik MA, Smuga‐Otto K et al. Induced pluripotent stem cell lines derived from human somatic cells. Science 2007;318:1917–1920.

16. Takahashi K, Tanabe K, Ohnuki M et al. Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell 2007;131:861–872.

17. Park IH, Zhao R, West JA et al. Reprogramming of human somatic cells to pluripotency with defined factors. Nature 2008;451:141–146.

18. Chin MH, Mason MJ, Xie W et al. Induced pluripotent stem cells and embryonic stem cells are distinguished by gene expression signatures. Cell Stem Cell 2009;5:111‐123.

19. Kim K, Doi A, Wen B et al. Epigenetic memory in induced pluripotent stem cells. Nature 2010;467:285‐290.

20. Ohi Y, Qin H, Hong C et al. Incomplete DNA methylation underlies a transcriptional memory of somatic cells in human iPS cells. Nat Cell Biol 2011;13:541–549.

21. Dolgin E. Gene flaw found in induced stem cells. Nature 2010;464:663.

22. Nasu A, Ikeya M, Yamamoto T et al. Genetically matched human iPS cells reveal that propensity for cartilage and bone differentiation differs with clones, not cell type of origin. PLoS One 2013;8:e53771.

23. Sparks NR., zur Nieden NI. Pluripotent stem cells as tools to assess developmental toxicity: diversity instead of consolidation. In: Sahu SC and Casciano DA, eds. Handbook of Nanotoxicology, Nanomedicine and Stem Cell Use in Toxicology. Chickester, UK: John Wiley & Sons, 2014:299‐307.

24. Guzzo RM, Gibson J, Xu RH et al. Efficient differentiation of human iPSC‐derived mesenchymal stem cells to chondroprogenitor cells. J Cell Biochem 2013;114:480‐ 490.

25. Deyle DR, Khan I.F, Ren G et al. Normal collagen and bone production by genetargeted human osteogenesis imperfecta iPSCs. Mol Ther 2011;20:204‐213.

26. Quarto N, Leonard B, Li S et al. Skeletogenic phenotype of human Marfan embryonic stem cells faithfully phenocopied by patient‐specific induced‐pluripotent stem cells. PNAS 2011;109:215‐220.

27. Lee TJ, Jang J, Kang S et al. Mesenchymal stem cell‐conditioned medium enhances osteogenic and chondrogenic differentiation of human embryonic stem cells and human induced pluripotent stem cells by mesodermal lineage induction. Tissue Eng Part A 2014;20:1306‐1313.

28. de Peppo GM, Marcos‐Campos I, Kahler DJ et al. Engineering bone tissue substitutes from human induced pluripotent stem cells. PNAS 2013;110:8680‐8685.

29. Tang M, Chen W, Liu J et al. Human induced pluripotent stem cell‐derived mesenchymal stem cell seeding on calcium phosphate scaffold for bone regeneration. Tissue Eng Part A 2014;20:1295‐1305.

30. Zou L, Luo Y, Chen M et al. A simple method for deriving functional MSCs and applied for osteogenesis in 3D scaffolds. Sci Rep 2013;3:2243.

31. Ardeshirylajimi A, Hosseinkhani S, Parivar K et al. Nanofiber‐based polyethersulfone scaffold and efficient differentiation of human induced pluripotent stem cells into osteoblastic lineage. Mol Biol Rep 2013;40:42874294.

32. Kim MJ, Son MJ, Son MY et al. Generation of human induced pluripotent stem cells from osteoarthritis patient‐derived synovial cells. Arthritis Rheum 2011;63:3010‐3021.

33. Leucht P, Kim JB, Amasha R et al. Embryonic origin and Hox status determine progenitor cell fate during adult bone regeneration. Development 2008;135(17):2845‐ 54.

34. Jheon AH, Schneider RA. The cells that fill the bill: neural crest and the evolution of craniofacial development. J Dent Res 2009;88(1):12‐21.

35. Park S, Im GI. Embryonic stem cells and induced pluripotent stem cells for skeletal regeneration. Tissue Eng Part B Rev 2014;20:381‐391.

36. Chatterjee P, Cheung Y, Liew CG. Transfecting and nucleofecting human induced pluripotent stem cells. J Vis Exp 2011;56:e3110.

37. Torrez LB, Perez Y, Yang J et al. Derivation of neural progenitors and retinal pigment epithelium from common marmoset and human pluripotent stem cells. Stem Cells Int 2012;201:417865.

38. Kuske B, Savkovic V, zur Nieden NI. Improved media compositions for the differentiation of embryonic stem cells into osteoblasts and chondrocytes. Methods Mol Biol. 2011;690:195‐215.

39. Betters E, Liu Y, Kjaeldgaard A et al. Analysis of early human neural crest development. Dev Bio 2010;344:578‐592.

40. Davis LA, Dienelt A, zur Nieden NI. Absorption‐based assays for the analysis of osteogenic and chondrogenic yield. Methods Mol Biol 2011;690:255‐272.

41. Majumdar S, Ramachandran S, Cerione RA. New insights into the GTP binding/GTP hydrolytic cycle of large G proteins. J Biol Chem 2006;281:9219‐9226.

42. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real‐time quantitative PCR and the 2(‐Delta C(T)) Method Methods 2001;25:402‐408.

43. Leung AW, Murdoch B, Salem AF et al. WNT/β‐catenin signaling mediates human neural crest induction via a pre‐neural border intermediate. Development. 2016;143(3):398‐410.

44. Sung CO, Lee KW, Han S et al. Twist1 is up‐regulated in gastric cancer‐associated fibroblasts with poor clinical outcomes. Am J Pathol 2011;179:1827‐1838.

45. Kuske B, Pulyanina PY, zur Nieden NI. Embryonic Stem Cell Test: Stem cell use in predicting developmental cardiotoxicity and osteotoxicity. In: Harris C and Hansen JM, eds. Developmental Toxicology: Methods in Molecular Biology 2012;889:147‐179.

46. zur Nieden NI, Price FD, Davis LA et al. Gene profiling on mixed embryonic stem cell populations reveals a biphasic role for betacatenin in osteogenic differentiation. Mol Endocrinol 2007;21:674‐685.

47. zur Nieden NI, Davis LA, Rancout DE. Monolayer cultivation of osteoprogenitors shortens duration of the embryonic stem cell test while reliably predicting developmental osteotoxicity. Toxicology 2010;277:66‐73.

48. Guo D, Keightley A, Guthrie J et al. Identification of osteocyte‐selective proteins. Proteomics 2010;10:3688‐3698.

49. Basch ML, Bronner‐Fraser M, GarcíaCastro MI. Specification of the neural crest occurs during gastrulation and requires Pax7. Nature 2006;441:218‐222.

50. Murdoch B, DelConte C, García‐Castro MI. Pax7 lineage contributions to the mammalian neural crest. PLoS One 2012;7:e41089.

51. Li W, Cornell RA. Redundant activities of Tfap2a and Tfap2c are required for neural crest induction and development of other non‐neural ectoderm derivatives in zebrafish embryos. Dev Biol. 2007;304(1):338‐354.

52. Beddington RS, Rashblass P, Wilson V. Brachyury‐‐a gene affecting mouse gastrulation and early organogenesis. Dev Suppl 1992;157‐165.

53. Sornay‐Rendu E, Munoz F, Garnero P et al. Identification of osteopenic women at high risk of fracture: the OFELY study. J Bone Miner Res 2005;20:1813‐1819.

54. Sotiropoulou P, Fountos G, Martini N et al. Bone calcium/phosphorus ratio determination using dual energy X‐ray method. Phys Med 2015;31:307‐313.

55. Tzaphlidou M, Zaichick V. Calcium, phosphorus, calcium‐phosphorus ratio in rib bone of healthy humans. Biol Trace Elem Res. 2003;93:63‐74.

56. Quelch KJ, Melick RA, Bingham PJ et al. Chemical composition of human bone. Arch Oral Biol 1983;28:665‐674.

57. Takashima Y, Era T, Nakao K et al. Neuroepithelial cells supply an initial transient wave of MSC differentiation. Cell 2007;129:1377‐1388.

58. Yen HY, Ting MC, Maxson RE. Jagged1 functions downstream of Twist1 in the specification of the coronal suture and the formation of a boundary between osteogenic and non‐osteogenic cells. Dev Biol 2010;347:258‐270.

59. Miraoui H, Severe N, Vaudin P et al. Molecular silencing of Twist1 enhances osteogenic differentiation of murine mesenchymal stem cells: implication of FGFR2 signaling. J Cell Biochem 2010;110:1147‐1154.

60. Goodnough LH, Chang AT, Treloar C et al. Twist1 mediates repression of chondrogenesis by β‐catenin to promote cranial bone progenitor specification. Development 2012;139:4428‐4438.

61. Draper JS, Smith K, Gokhale P, Moore HD, Maltby E, Johnson J, Meisner L, Zwaka TP, Thomson JA, Andrews PW. Recurrent gain of chromosomes 17q and 12 in cultured human embryonic stem cells. Nat Biotechnol. 2004;22(1):53‐54.

62. Romero A, Vilanova E, Sogorb MA. Shortening and improving the embryonic stem cell test through the use of gene biomarkers of differentiation. J Toxicol 2011;286034.

63. Liu Y, Goldberg AJ, Dennis JE et al. Onestep derivation of mesenchymal stem cell (MSC)‐like cells from human pluripotent stem cells on a fibrillar collagen coating. PLoS One 2012;7:e33225.

64. Kanke K, Masaki H, Saito T et al. Stepwise differentiation of pluripotent stem cells into osteoblasts using four small molecules under serum‐free and feeder‐free conditions. Stem Cell Rep. 2014;2:751‐760.

65. Diekman BO, Christoforou N, Willard VP et al. Cartilage tissue engineering using differentiated and purified induced pluripotent stem cells. PNAS 2012;109:19172‐ 19177.

66. Yoshida Y, Yamanaka S. Recent stem cell advances: induced pluripotent stem cells for disease modeling and stem cell‐based regeneration. Circulation 2010;122:80‐87.

67. Quarto N, Wan DC, Kwan MD et al. Origin matters: differences in embryonic tissue origin and Wnt signaling determine the osteogenic potential and healing capacity of frontal and parietal calvarial bones. J Bone Miner Res 2010;25:1680‐1694.

68. Singh AM, Reynolds D, Cliff T et al. Signaling network crosstalk in human pluripotent cells: a Smad2/3‐regulated switch that controls the balance between self‐renewal and differentiation. Cell Stem Cell 2012;10:312‐26.

69. Bildsoe H, Loebel DA, Jones VJ et al. Requirement for Twist1 in frontonasal and skull vault development in the mouse embryo. Dev Biol 2009;331:176‐188.

70. Bildsoe H, Loebel DA, Jones VJ et al. The mesenchymal architecture of the cranial mesoderm of mouse embryos is disrupted by the loss of Twist1 function. Dev Biol 2013;374:295‐307.

71. Krug AK, Kolde R, Gaspar JA et al. Human embryonic stem cell‐derived test systems for developmental neurotoxicity: a transcriptomics approach. Arch Toxicol 2013;87:123‐143.

72. Scott CW, Peterts MF, Dragan YP. Human induced pluripotent stem cells and their use in drug discovery for toxicity testing. Toxicol Lett 2013;219:49‐58.

CHAPTER 3

Pluripotent Stem Cells as Tools to Assess Developmental Toxicity:

Diversity Instead of Consolidation

Nicole R.L. Sparks and Nicole I. zur Nieden

Introduction

The embryonic development of an organism represents a relatively short, yet vulnerable, span in its entire life cycle. Therefore, any teratogenic compound acting upon the developing organism during pregnancy may severely perturb development. In the early

1960s, the thalidomide disaster demonstrated the urgent need to identify and predict the developmental toxicity of putative toxicants before being released into the environment or for human use for the first time.

Nowadays, human populations are experiencing a rapid increase in exposure to

potential environmental toxicants in the form of pharmaceutical drugs, commercial

chemicals, industrial by-products, and wastes that are untested for safety. Most of these

chemicals need to be assessed for their health risks to determine acceptable exposure levels.

In vivo studies according to the Organization for Economic Co-operation and Development

(OECD) guidelines are currently performed to provide such data. It is estimated that

100,000 existing chemicals circulate the market in the United States and that 2000 new

substances enter it each year (Carpenter and Bushkin-Bedient, 2013; Locke and Bruce

Myers, 2010). Apart from the tremendous associated costs, in vivo testing of such a magnitude of chemicals would require enormous amounts of animals and a great deal of time. Clearly, in vivo testing methods would not be feasible to adequately test all current and putative substances. Hence, it is crucial to develop alternative in vitro methods to assess toxicity, especially teratogenicity, to test current and putative chemicals for their impact on prenatal development while at the same time reduce the use of animals and cut expenses.

Over the past 30 years, a number of in vitro approaches have been developed to screen chemicals for their teratogenic effects and circumvent traditional in vivo methods.

In 2002, three of these in vitro embryotoxicity assays were validated according to standards of the European Centre for the Validation of Alternative Methods (ECVAM): the micromass assay (Umansky, 1966), the whole embryo culture assay (Cuthbertson and

Beck, 1990), and the embryonic stem cell test (EST) (Genschow et al., 2000, 2002).

However, micromass and whole embryo culture assay are labor intensive and still require the killing of pregnant animals and/or embryos. The only in vitro mammalian-based developmental toxicity test that does not sacrifice pregnant animals is the EST, as it uses permanent cell lines, embryonic stem cells (ESCs), and 3T3 fibroblasts.

Embryonic Stem Cells

In recent years, scientific understanding of stem cells has greatly expanded the field of toxicology.

ESCs are derived from the inner cell mass of the blastocyst-stage embryo

(Figure 3.1). ESCs can maintain pluripotency for prolonged periods in vitro, have the ability to undergo sustained proliferation, and are able to differentiate and give rise to all three germ layers and thus all cells of the body (Vallier, Reynolds, and Figure 3.1. Embryonic stem cell derivation and Pedersen, 2004). ESCs express OCT- differentiation. Totipotent cells (zygote) can form all types of cells including the organism and the trophoblast. Pluripotent embryonic stem cells are derived from the inner cell mass of 3/4, NANOG, and SOX-2 (Chen and the blastocyst-stage embryo and can form all three germ layers, but not the extra embryonic structures. Daley, 2008) to maintain pluripotency, http://www.musclesused.com/types-muscles/ and the downregulation of these genes is associated with differentiation initiation (Rao and

Orkin, 2006). The prolonged lifespan of ESCs is attributed to high telomerase activity, which is typically low in somatic cells (Thomson et al., 1998). Thus, ESCs have an unlimited proliferation capacity in vitro, which does not hold true for adult stem cells

(Hiyama and Hiyama, 2007).

For ESCs to differentiate toward a specific lineage, expression of pluripotency markers must be suppressed. In addition, chemical cues can be added to further push cells toward a particular lineage. ESCs progressing toward a particular lineage first produce a daughter cell and a progenitor cell, in an asymmetrical division (Sommer and Rao, 2002).

Progenitor cells undergo a limited number of divisions (a distinction from ESCs) before terminally differentiating into a mature cell (Seaberg and van der Kooy, 2003). In a nonadherent environment, ESCs aggregate to form embryoid bodies (EBs), which comprise progenitor cells of various lineages. Owing to this ability to propagate themselves indefinitely and to mimic embryogenesis by producing a multitude of cell types, ESCs are a developmentally powerful tool for toxicological assays. For instance, cultures of spontaneously differentiated ESCs will contain contracting cardiomyocytes (Christoforou et al., 2008), a feature that is exploited in the classic EST.

Embryonic Stem Cell Test

The EST is the most promising mammalian in vitro assay to predict the embryotoxic potential of test chemicals. Many congenital abnormalities occur because of the adverse effects of deleterious agents on development processes, and as the EST predicts teratogenic potential during this growth period, it is the ideal test. It classifies substances into three classes of in vivo embryotoxicity (strong, weak, and not embryotoxic) (Scholz et al., 1999a; Genschow et al., 2002). Murine embryonic stem cells (mESCs) are utilized in the EST to assess the embryotoxic potential of the test substance as their in vitro differentiation recapitulates developmental processes and is characterized by gene

expression patterns indicative

of in vivo cardiogenesis.

3T3 fibroblasts are

also evaluated as they model the

effect of the chemical on mature

somatic cells (Figure 3.2). For

the classification of the

embryotoxic potential of the

test compound into the above-

named categories, three

Figure 3.2. Embryonic stem cell test (EST). Overview of the approach endpoints are determined: (i) and endpoints to assess the embryotoxic potential of a test chemical using mESCs and mouse fibroblasts. cell differentiation: the

concentration of the test compound that causes a 50% inhibition of the differentiation of ESCs into cardiomyocytes

(ID50); (ii) viability of immature cells: the concentration of the test compound resulting in a 50% decrease in the viability of ESCs [assessed by MTT (3(4,5-dimethylthiazol-2-yl)-

2,5-diphenoltetrazolium bromide) assay] (IC50); and (iii) viability of mature somatic cells: the concentration of the test compound resulting in a 50% decrease in the viability of 3T3 fibroblasts (IC50) (Scholz et al., 1999b). As previously mentioned, spontaneously differentiating EBs contain cell types of all three germ layers, amongst them beating cardiomyocytes. Indeed, the heart is the first organ to form in the developing embryo and

contracting clusters of cardiomyocytes can be microscopically identified in vitro, which

suggested that cardiomyocytes are a suitable endpoint for the EST.

In addition, the EST is the only in vitro test with a biostatistical prediction model that circumvents the use of animals to evaluate the embryotoxic potential of test compounds

(Scholz et al., 1999a,b; Genschow et al., 2000, 2002, 2004). Using this prediction model, the National Centre for Documentation and Evaluation of Alternative Methods to Animal

Experiments (ZEBET) at the German Federal Institute for Risk Assessment coordinated

ECVAM’s prevalidation and validation studies of 6 and 14 compounds, previously characterized by in vivo data, respectively. The prevalidation study evaluated the prediction model designed for the EST, and based on those results, the prediction model was reevaluated with the remaining 14 test substances in the validation study (Spielmann et al.,

2006). The validation study resulted in a 78% correct classification of the 20 test compounds (Genschow et al., 2000, 2002, 2004). In addition, the test displayed 100% predictivity for strong embryotoxic chemicals (Spielmann et al., 2006). The validation study comprises an intra- and interlaboratory exploration using six strong embryotoxic substances, seven weak substances, and seven nonembryotoxic substances under blind conditions in four laboratories (Spielmann et al., 2006). All four laboratories showed 100% predictivity and 85% precision of strong embryotoxic chemicals. Moreover, weak and nonembryotoxic correct classification of chemicals exhibited 83 and 75%, respectively.

Furthermore, there was a correlation between in vivo and in vitro classification in the four laboratories of 85, 83, 81, and 68%. These numbers show that the test could not decipher between weak and nonembryotoxic compounds, despite being successful in identifying

55 strong embryotoxic chemicals. Moreover, the test could not detect compounds whose metabolites act as embryotoxicants; however, the overall positive results gave the EST the green light to be considered for regulatory uses (Genschow et al., 2002).

With the validation from ECVAM, some pharmaceutical and chemical companies employed the EST and found it useful for early stages of product development. When originally invented, the EST was designed for access to all laboratories, including pharmaceutical and chemical companies, with a standard, validated set of procedures. A decade later, improvements were clearly needed for the EST. Pfizer utilized ECVAM’s validated EST with compounds used to validate the EST and in-house compounds

(Paquette et al., 2008). Their study showed that the test was robust and able to be used in other laboratories, however needed improvements. In pursuit of developing a standardized

EST for all laboratories, ECVAM held a workshop in 2003 to address the future of the test.

Key toxicologists determined that testing 20 compounds was not sufficient, and the choice of chemicals needed to be expanded for validating the test (Spielmann et al., 2006).

Additionally, molecular endpoints and additional differentiation pathways, such as neural and osteogenic, needed to be included. Furthermore, it was suggested that the successful use of human embryonic stem cells (hESCs) in embryotoxicity screens would abrogate interspecies differences and was more reliable to predict human effects, while still reducing the number of laboratory animals used. Thus, instead of further validation studies for standard use, more in vitro models for prediction were developed.

These modifications to the EST and new in vitro models were done in-house, which created diversity amongst screening methods instead of further consolidating the method.

Developmental toxicology must assess the efficacy of the models to produce a high throughput coherent set of procedures for all to adapt to. However, owing to these in-house developments, the new methods, some of which are reviewed later, were not tested for inter- and intra-laboratory capability and not validated although promising.

3.1 Improvements to the Embryonic Stem Cell Test

Although it was established early on that ESCs are a powerful and valuable

tool to examine the embryotoxic potential of a chemical, the originally tested cardiogenic

endpoint is problematic. The readout of the classic EST is based on the presence or absence

of beating cardiomyocytes and requires skilled personnel to microscopically detect such

cells. In addition to being a technically challenging endpoint, cardiogenesis may also not

be a valid endpoint for all teratogens. For example, a variety of developmental toxicants

result in limb, skeletal, or CNS (central nervous system) malformations (zur Nieden et al.,

2004). It was therefore deemed to be beneficial to differentiate ESCs into other lineages

that can be easily tested with the EST to identify developmental toxicity in other endpoint

organs (zur Nieden et al., 2004). For instance, the EST is one of few tests available to study

the effects of agents on skeletal development in vitro (zur Nieden et al., 2010a, 2010b; zur

Nieden and Baumgartner, 2010). Our group has established a method that abrogates the

need for expensive molecular assays and utilizes morphometric image analysis to assess

developmental osteotoxicity (zur Nieden et al., 2010a). We have used the mouse-based

assay to determine developmental osteotoxicity in pharmaceutical compounds and have

also investigated developmental osteotoxicity of chloride derivatives (zur Nieden and

Baumgartner, 2010).

In addition to the osteogenic endpoint, neural, chondrogenic, and endothelial endpoints were also developed to assess the embryotoxic potential on such cells (zur

Nieden et al., 2004; Festag et al., 2007a). This ability to test additional differentiation pathways is due to advances in stem cell biology that allow for the directed differentiation of the cells into specific lineages with high yield (zur Nieden et al., 2003, 2005; Buttery et al., 2001; Festag et al., 2007b). Interestingly, these studies highlight concentration- and tissue-specific effects of compounds. For example, at a specific concentration, retinoic acid, a classified strong teratogen, caused a 400% increase in osteoblast marker gene expression, a 50% decrease in a chondrocyte specific mRNA, and no disruption of neural or cardiomyocyte gene expression. These intriguing results underscore the importance of testing multiple differentiation pathways.

In addition to developing screens for multiple diverse differentiation pathways, laboratories made improvements to the EST by establishing molecular endpoint analyses that could enhance the predictivity of the test. Bremer et al. (1999) were the first to use transgenic mESCs, in which green fluorescent protein (GFP) linked to a cardiac specific promoter was used as a toxicological endpoint by quantifying the number of GFP positive cells in cultures treated with chemicals known to affect in vivo development with flow cytometry. At the same time, flow cytometric methods provide the ability of high- throughput screening and are a more robust and reproducible screening system. Seiler et al. (2002) also developed a quicker and more reproducible screen by discovering that the

58 flow-cytometry-based detection of tissue-specific markers in compound-treated mESCs could accurately predict teratogenicity. In addition to protein expression, we reported that changes in marker gene expression in treated cells as quantified by RT-PCR (reverse transcription polymerase chain reaction) are predictive of the teratogenic potential of the compound. Therefore, the analysis of gene expression is a promising tool to study developmental toxicology because embryogenesis is aided by the changes in the regulation of genes. An advantage of determining which genes are affected by specific compounds is that it can lead to further elucidation of the underlying mechanisms of the teratogenic effect.

Aside from tissue-specific protein and gene expression, the spontaneous extracellular field potential of differentiating cardiomyocytes measured with a microelectrode array system was found to be inhibited by embryotoxic chemicals (Koseki et al., 2010). Such promising studies underline the fact that it is worthwhile to include additional endpoints that enhance the predictivity of the EST.

Other improvements to the EST were attempts to reduce test duration. Although still quicker than an in vivo experiment, the first beating cardiomyocytes emerge at 9–

10days after in vitro culture, which does not facilitate the high throughput demands of testing the increased number of toxicants in the environment. It is necessary to develop an in vitro screening assay that reduces the time it takes to have the results in hand. Cardiac protein and gene expression endpoints mentioned above not only served the purpose of increasing predictivity but also allowed earlier detection of teratogenic effects on cardiogenesis. As such, quantitative PCR analysis of cardiac-specific myosin heavy chain

(MHC) expression reduced the test to 8days and still accurately assessed chemicals in

59 accordance with their in vivo potential (zur Nieden et al., 2001). In addition, the FACS-

EST (fluorescence-activated cell sorting embryonic stem cell test) produced the same sensitivity as the validated EST but reduced the length of time of the assay to 7days (Buesen et al., 2009; Seiler et al., 2004). While one would expect the appearance of mRNAs before the appearance of protein, the discrepancy in timing between these two studies may potentially be explained by the different serum lots used for the differentiation of the ESCs.

Similar attempts were made by our group to shorten the duration of skeletal endpoint differentiations (zur Nieden et al., 2010b). While we were successful in assessing osteotoxicity at the level of mRNA expression of bone marker genes, these endpoints could not be analyzed until 30days of osteogenic differentiation. To shorten the assay, we have made significant improvements to the culturing method and can now assess the teratogenic

Figure 3.3. Human ESC osteogenic induction. Osteogenic differentiation is initiated after the cultures become confluent (designated d0) with Dulbecco’s Modified Eagle Medium (DMEM) and fetal bovine serum (FBS). On day 5, 1,25α(OH)2 vitamin D3 (VD3), β-glycerophosphate (GP), and ascorbic acid (AA) are added to the medium to direct cells toward the osteogenic lineage. The presence of calcification is measurable on day 14, and by day 30, the culture is predominantly osteoblasts.

60 potential of a chemical in 14days (zur Nieden et al., 2010b). We currently use a variation of this method for human pluripotent stem cell induction to osteoblasts and assessment of toxicants (Figure 3.3).

3.2 Beyond the Embryonic Stem Cell Test

With lack of consensus on which in vitro tests or endpoints are suitable for

determining the embryotoxic potential of a compound, more developmental toxicity

screens are being developed. For instance, the adherent cell differentiation and cytotoxicity

(ACDC) assay is a murine-ESC-based model designed to (i) evaluate the effects of

chemical exposure on stem cell number and differentiation using a single culture technique,

(ii) quantify markers of cell number and cardiomyocyte differentiation, and (iii) provide an

assessment of multiple differentiation markers (Barrier et al., 2011). The model abrogates

the use of the EB-based differentiation system for an adherent method, which simplifies

the culturing technique and is completed in 9days. An advantage of this assay is the way

cell number and cardiomyocyte differentiation are assessed; both are done after fixation on

the same cells in the same culture wells. Cell number and MHC protein were both measured

using the in-cell western protocol develop ed to use with the Li-Cor Odyssey infrared

imaging system. This system measures two endpoints in each well with two infrared

channels at 700 and 800nm. In this study, Barrier at al. (2011) used the 700-nm channel to

quantify the relative cell number and the 800-nm channel to detect the MHC signal. This

assay is beneficial in that it simplifies culture techniques and can assess two endpoints in

one vessel—creating a high throughput assay that has the potential of full automation

(Barrier et al., 2011). Although promising, this assay does not include a prediction model

such as the EST.

More than 90% of chemicals that enter the body are metabolized in the liver into other forms, metabolites, which are a potential culprit that skews development (Wilkinson,

2005; Kleinstruer et al., 2011). Primary rat hepatocytes have been utilized to metabolize the compound in vitro before analysis. However, these cells are difficult to maintain and are plagued by interspecies variation (Greenhough, Medine, and Hay, 2010). Furthermore, cells are typically harvested from rats, which would not reduce animal use. Other currently available systems for mimicking liver metabolism include the use of S9 liver extracts, transgenic cell lines, hepatocyte-like cell monolayers, and 3D organotypic cultures (Adler et al., 2011; Esch, King, and Shuler, 2011; Giri et al., 2011; Landsiedel et al., 2011).

Pluripotent stem cell research has enabled the ability to derive hepatocyte-like cells from hESCs; however, the yield and efficiency are low and gene expression profiles vary

(Greenhough, Medine, and Hay, 2010). With a clear need for metabolism models, North

America and the European Union have created virtual liver initiatives. The US EPA

(Environmental Protection Agency) Virtual Liver Project (Environmental Protection

Agency (EPA), 2009) uses mathematical models to predict liver toxicity based on molecular targets, signaling pathways, and toxicity at the organ and organismal levels.

Similar to interspecies differences in the ability to metabolize substances, individuals too differ in their ability to metabolize chemicals (Xie et al., 2001), making it difficult to reliably predict toxicity using cells derived from a single donor (Greenhough, Medine, and

Hay, 2010). Here, the development of induced pluripotent stem cells (iPSCs) has

62 revolutionized stem-cell-based toxicity screening. iPSCs are generated by the reprogramming of somatic cells through the introduction of four key transcription factors:

Sox-2, Klf-4, Oct4, and c-Myc (Takahashi and Yamanaka, 2006; Takahashi et al., 2007).

These cells possess a number of advantages over hESCs, including patient-specific cell- based drug testing and are free from the ethical concerns associated with the use of blastocysts. The creation of liver toxicity screening technologies utilizing iPSC-derived hepatocyte-like cells would allow investigation into the effects of single nucleotide polymorphisms on drug metabolism and toxicity (Greenhough, Medine, and Hay, 2010).

Model Organism

Murine ESCs are able to recapitulate hallmarks of embryogenesis and murine-ESC- based embryotoxicity assays are thus a great tool to elucidate mechanisms of developmental toxicity. However, mouse models are not able to entirely mimic human development. Moreover, culturing methods and expression of certain genes differ between murine and human ESCs. For instance, human ESCs form flat colonies, do not survive as single cells, and require basic fibroblast growth factor (bFGF) in the culture medium to maintain an undifferentiated state, whereas murine ESCs have three-dimensional colonies, survive as single cells, and leukemia inhibitory factor is added to culture medium. These differences limit the transferability of predictions for human use made based on mouse cells. Indeed, in vivo developmental toxicity studies have shown high interspecies differences, despite the usefulness of the murine-based EST (Hurtt, Cappon, and Browning,

2003). Specifically, thalidomide has shown no teratogenic effects in mice, but causes

63 profound human congenital malformations (Gilbert, 2003; Fratta, Sigg, and Maiorana,

1965; Brent and Holmes, 1988).

Therefore, integrating human stem cells into developmental toxicity assays is necessary to accurately account for the human response to putative toxicants. In addition, these cells are highly attractive to the pharmaceutical industry in facilitating drug discovery development.

As a first step, Buzanska et al. (2009) utilized a human neural stem cell line derived from umbilical cord blood (HUCB-NSC) as a model to test developmental neurotoxicity.

The study showed that differently staged HUCB-NSCs were able to discriminate between neurotoxic and non-neurotoxic chemicals. The endpoints (cell viability, proliferation, apoptosis, and expression of cell-specific markers) were assessed 48h after exposure to test chemicals of early and late stages of HUCB-NSC. Cell viability was measured by the live/dead viability/cytotoxicity (LDVC) assay, MTT assay, or the resazurin reduction assay. Nuclear chromatin staining and immunocytochemical labeling for stage-specific markers were used to determine cell proliferation, apoptosis, and neural differentiation.

The LDVC assay allowed for the rapid discrimination of live and dead cells microscopically or with flow cytometric methods. The MTT and resazurin reduction assays are both colorimetric assays that rely on the ability of mitochondria to reduce MTT or resazurin, respectively, to measure viable cells. The downside of such assays is that they are not true viability tests because they measure mitochondrial potential, which could also be altered due to changes in cellular metabolism. Furthermore, apoptotic cell identification

64 could benefit from additional endpoints, such as caspase-3 quantification, to support chromatin condensation data.

Although some of the endpoint assays in the HUCB-NSC study could clearly be improved, an advantage of their approach was the usage of human cells. Advantageous from the standpoint of choice of test species, their approach differentiates neural progenitors as opposed to less committed cells and therefore evaluates particular stages of development only. In 2011, Colleoni et al. (2011) developed an alternative approach to assess neural teratogenicity using pluripotent human ESCs. The group differentiated human ESCs into neural rosettes, which represent neural plate and neural tube development, and exposed the cells to retinoic acid.

This treatment would induce neural tube defects, which is one of the most common birth defects, affecting nearly 0.5–2 newborns per 1000 (Colleoni et al., 2011). Their microarray analysis and qPCR data suggested dose-dependent misregulation of genes involved in neural tube formation and closure, such as the homeobox family genes and

Otx2. Furthermore, this in vitro modulation of neurulation markers in response to retinoic acid correlated with gene expression responses associated with in vivo retinoic acid exposure. Despite being a first promising step toward a predictive human-based in vitro model, this approach requires further validation of other teratogenic chemicals and lacks a prediction model to classify substances such as the one used in the classic EST.

In a similar approach, West et al. (2010) developed a hESC-based model that used metabolomics to detect small molecule changes in response to teratogen exposure, a concept that was first introduced by Cezar et al. (2007). In drug treated and untreated cells,

65 the group reported significant differences in metabolite abundances, with asymmetric dimethylarginine (ADMA) being the most predictive of teratogenicity, particularly neural tube closure. The discovered alteration of ADMA upon teratogen exposure suggests that it is a potential biomarker for neural tube defects when modified. However, further exploration is needed to detect biomarkers in toxicity of differentiation pathways. Novel approaches such as this will potentially elucidate the underlying mechanisms of developmental toxicity.

Conclusion

There are many demands in the field of developmental and predictive toxicology to create an in vitro system that meets the standards of academic and industrial toxicity screening. Although in vitro screening methods in developmental toxicology have tremendously advanced, there is still a lack of agreement on what constitutes a developmental toxicant and how these characteristics such as physiological endpoints, maternal toxicity, dose magnitude, and duration can be translated into, or derived from the in vitro model (Castoldi et al., 2008). The EST is the most promising validated in vitro developmental toxicity model, though its endpoints as well as the use of murine cells limit the utility of the test. These downfalls have created a frenzy of improvements and have led to new in vitro approaches (some including hESC models). New in vitro efforts would benefit the EST if toxicologists focused on standardizing a set of assays to assess developmental toxicity instead of creating diversity amongst the field.

References

1. Adler S, Basketter D, Creton S, Pelkonen O, van Benthem J, Zuang V, Andersen KE, Angers-Loustau A, Aptula A, Bal-Price A, Benfenati E, Bernauer U, Bessems J, Bois FY, Boobis A, Brandon E, Bremer S, Broschard T, Casati S, Coecke S, Corvi R, Cronin M, Daston G, Dekant W, Felter S, Grignard E, Gundert-Remy U, Heinonen T, Kimber I, Kleinjans J, Komulainen H, Kreiling R, Kreysa J, Leite SB, Loizou G, Maxwell G, Mazzatorta P, Munn S, Pfuhler S, Phrakonkham P, Piersma A, Poth A, Prieto P, Repetto G, Rogiers V, Schoeters G, Schwarz M, Serafimova R, Tähti H, Testai E, van Delft J, van Loveren H, Vinken M, Worth A, Zaldivar JM. Alternative (non-animal) methods for cosmetics testing: current status and future prospects-2010. Arch Toxicol .2011;85(5): 367–485.

2. Barrier M, Jeffay S, Nichols H, Chandler K, Hoopes M, Slentz-Kesler K, Hunter E. Mouse embryonic stem cell adherent cell differentiation and cytotoxicity (ACDC) assay. Reprod Toxicol 2011;31(4): 389–391.

3. Bremer S, van Dooren M, Paparella M, Kossolov E, Fleischmann BK, Hescheler J. Effects of embryotoxic chemicals on the in vitro differentiation of genetically engineered embryonic stem cells into cardiac cells. Toxicol In Vitro 1999;13(4–5): 645–650.

4. Brent R, Holmes L. Clinical and basic science lessons from the thalidomide tragedy: what have we learned about the cause of limb defects? Teratology 1988;38: 241–251. 5. Buesen R, Genschow E, Slawik B, Visan A, Spielmann H, Luch A, Seiler A. Embryonic stem cell test remastered: comparison between the validated EST and the new molecular FACS-EST for assessing developmental toxicity in vitro. Toxicologic Sci 2009;108(2): 389–400.

6. Buttery LDK, Bourne S, Xynos JD, Wood H, Hughes FJ, Hughes SPF, Episkopou V, Polak JM. Differentiation of osteoblasts and in vitro bone formation from murine embryonic stem cells. Tissue Eng 2001;7(1): 89–99.

7. Buzanska L, Sypecka J, Nerini-Molteni S, Compagnoni A, Hogberg HT, Del Torchio R, Domanska-Janik K, Zimmer J, Coeecke S. A human stem cell-based model for identifying adverse effects of organic and inorganic chemicals on the developing nervous system. Stem Cells 2009;27: 2591–2601.

8. Carpenter DO, Bushkin-Bedient S. Exposure to chemicals and radiation during childhood and risk for cancer later in life. J Adolesc Health. 2013;52(5 Suppl):S21-9.

9. Castoldi AF, Johansson C, Onishchenko N, Coccini T, Roda E, Vahter M, Ceccatelli S, Manzo L. Human developmental neurotoxicity of methylmercury: impact of variables and risk modifiers. Regul Toxicol Pharmacol 2008;51: 201–214.

10. Cezar GG, Quam JA, Smith AM, Rosa GJ, Piekarczyk MS, Brown JF, Gage FH, Muotri AR. Identification of small molecules from human embryonic stem cells using metabolomics. Stem Cell Dev 2007;16(6): 869–882.

11. Chen L, Daley GQ. Molecular basis of pluripotency. Hum Mol Genet 2008;17(R1): R23–R27.

12. Christoforou N, Miller RA, Hill CM, Jie CC, McCallion AS, Gearhart JD. Mouse ES cell-derived cardiac precursor cells are multipotent and facilitate identification of novel cardiac genes. J Clin Invest 2008;118(3): 894–903.

13. Colleoni S, Galli C, Gaspar JA, Meganathan K, Jagtap S, Hescheler J, Sachinidis A, Lazzari G. Development of a neural teratogenicity test based on human embryonic stem cells: response to retinoic acid exposure. Toxicol Sci 2011;124(2): 370–377.

14. Cuthbertson RA, Beck F. Postimplantation whole embryo culture: a new method for studying ocular development. Invest Ophthalmol Vis Sci 1990;31(8): 1653–1656.

15. Environmental Protection Agency (EPA). The U.S. Environmental Protection Agency’s Strategic Plan for Evaluating the Toxicity of Chemicals. 2009 http://www.epa.gov/spc/ toxicitytesting/docs/toxtest_strategy_032309.pdf.

16. Esch MB, King TL, Shuler ML. The role of body-on-a-chip devices in drug and toxicity studies. Annu Rev Biomed Eng 2011;13: 55–72.

17. Festag M, Viertel B, Steinberg P, Sehner C. An in vitro embryotoxicity assay based on the disturbance of the differentiation of murine embryonic stem cells into endothelial cells. II. Testing of compounds. Toxicol In Vitro 2007a;21(8): 1631–1640.

18. Festag M, Viertel B, Steinberg P, Sehner C. An in vitro embryotoxicity assay based on the disturbance of the differentiation of murine embryonic stem cells into endothelial cells. I. Testing of compounds. Toxicol In Vitro 2007b;21(8): 1619–1630.

19. Fratta ID, Sigg EB, Maiorana K. Teratogenic effects of thalidomide in rabbits, rats, hamsters, and mice. Toxicol Appl Pharmacol 1965;7: 268–286.

20. Genschow E, Scholz G, Brown N, Piersma A, Brady M, Clemann N, Huuskonen H, Paillard F, Bremer S, Becker K, Spielmann H. Development of prediction models for three in vitro embryotoxicity tests in an ECVAM validation study. In Vitro Mol Toxicol 2000;13(1): 51–66.

21. Genschow E, Spielmann H, Scholz G, Seiler A, Brown N, Piersman A, Brady M, Clemann N, Huuskonen H, Pailard F, Bremer S, Becker K. The ECVAM international validation study on in vitro embryotoxicity tests: results of the definitive phase and evaluation of prediction models. European Centre for the Validation of Alternative Methods. Altern Lab Anim 2002;30(2): 151–176.

22. Genschow E, Spielmann H, Scholz G, Pohl I, Seiler A, Clemann N, Bremer S, Becker K. Validation of the embryonic stem cell test in the international ECVAM validation study on three in vitro embryotoxicity tests. Altern Lab Anim 2004;32(3): 209–244.

23. Gilbert SF. Developmental Biology. 2003. Sinauer Associates, Inc.: Sunderland, MA.

24. Giri S, Acikgöz A, Pathak P, Gutschker S, Kürsten A, Nieber K, Bader A. Three dimensional cultures of rat liver cells using a natural self-assembling nanoscaffold in a clinically relevant bioreactor for bioartificial liver construction. J Cell Physiol 2011;227(1): 313–327.

25. Greenhough S, Medine C, Hay DC. Pluripotent stem cell derived hepatocyte like cells and their potential in toxicity screening. Toxicology 2010;278(3): 250–255.

26. Hiyama E, Hiyama K. Telomere and telomerase in stem cells. Br J Cancer 2007;96(7): 1020–1024.

27. Hurtt ME, Cappon GD, Browning A. Proposal for a tiered approach to developmental toxicity testing for veterinary pharmaceutical products for food-producing animals. Food Chem Toxicol 2003;41(5): 611–619.

28. Kleinstruer NC, Smith AM, West PR, Conard KR, Fontaine BR, Weir-Hauptman AM, Palmer JA, Knudsen TB, Dix DJ, Donley EL, Cezar GG. Identifying developmental toxicity pathways for a subset of ToxCast chemicals using human embryonic stem cells and metabolomics. Toxicol Appl Pharmacol 2011;257(1): 111–121.

29. Koseki N, Deguchi J, Yamada T, Funabashi H, Seki T. Improvement of the embryonic stem cell test endpoint analysis by use of field potential detection. J Toxicol Sci 2010;35(5): 619–629.

30. Landsiedel R, Fabian E, Tralau T, Luch A. Chemical toxicity testing in vitro using P450 expressing cell lines such as human CYP1B1. Nat Protoc 2011;6: 677–688.

31. Locke PA, Bruce Myers D Jr. Implementing the National Academy's Vision and Strategy for Toxicity Testing: opportunities and challenges under the U.S. Toxic Substances Control Act. J Toxicol Environ Health B Crit Rev. 2010;13(2-4):376-84.

32. Paquette JA, Kumpf SW, Streck RD, Thomson JJ, Chapin RE, Stedman DB. Assessment of the embryonic stem cell test and application and use in the pharmaceutical industry. Birth Defects Res B Dev Reprod Toxicol 2008;83(2): 104– 111.

33. Rao S, Orkin SH. Unraveling the transcriptional network controlling ES cell pluripotency. Genome Biol 2006;7(8): 230.

34. Scholz G, Pohl I, Genschow E, Klemm M, Spielmann H. Embryotoxicity screening using embryonic stem cells in vitro: correlation to in vivo teratogenicity. Cells Tissues Organs 1999a;165(3–4): 203–211.

35. Scholz G, Genschow E, Pohl I, Bremer S, Paparella M, Southee J, Spielmann H. Prevalidation of the embryonic stem cell test (EST)—a new in vitro embryotoxicity test. Toxicol In Vitro 1999b;13(4–5): 675–681.

36. Seaberg RM, van der Kooy D. Stem and progenitor cells: the premature desertion of rigorous definitions. Trends Neurosci 2003;26(3): 125–131.

37. Seiler A, Visan A, Pohl I, Genschow E, Buesen R, Spielmann H. Improving the embryonic stem cell test (EST) by establishing molecular endpoints of tissue specific

development using murine embryonic stem cells (D3 cells). ALTEX 2002;19(Suppl 1): 55–63.

38. Seiler A, Visan A, Buesen R, Genschow E, Spielmann H. Improvement of an in vitro stem cell assay for developmental toxicity: the use of molecular endpoints in the embryonic stem cell test. Reprod Toxicol 2004;18(2): 231–240.

39. Sommer L, Rao M. Neural stem cells and regulation of cell number. Prog Neurobiol 2002;66: 1–18.

40. Spielmann H, Seiler A, Bremer S, Hareng L, Hartung T, Ahr H, Faustman E, Haas U, Moffat GJ, Nau H, Vanparys P, Piersma A, Sintes JR, Stuart J. The practical application of three validated in vitro embryotoxicity tests. The report and recommendations of an ECVAM/ZEBET workshop (ECVAM workshop 57). Altern. Lab. Anim. 2006;34(5): 527–538.

41. Takahashi K, Yamanaka S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 2006;126(4): 663–676.

42. Takahashi K, Tanabe K, Ohnuki M, Narita M, Ichisaka T, Tomoda K, Yamanaka S. Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell 2007;131(5): 861–872.

43. Thomson JA, Itskovitz-Eldor J, Shapiro SS, Waknitz MA, Swiergiel JJ, Marshall VS, Jones JM. Embryonic stem cell lines derived from human blastocysts. Science 1998;5391(282): 1145–1147.

44. Umansky R. The effect of cell population density on the developmental fate of reaggregating mouse limb bud mesenchyme. Dev Biol 1966;13(1): 31–56.

45. Vallier L, Reynolds D, Pedersen RA. Nodal inhibits differentiation of human embryonic stem cells along the neuroectodermal default pathway. Dev Biol 2004;275(2): 403–421.

46. West PR, Weir AM, Smith AM, Donley EL, Cezar GG. Predicting human developmental toxicity of pharmaceuticals using human embryonic stem cells and metabolomics. Toxicol Appl Pharmacol 2010;247(1): 18–27.

47. Wilkinson GR. Drug metabolism and variability among patients in drug response. N Engl J Med 2005;352: 2211–2221.

48. Xie HG, Kim RB, Wood AJJ, Stein CM. Molecular basis of ethnic differences in drug disposition and response. Annu Rev Pharmacol Toxicol 2001;41: 815–850.

49. zur Nieden NI, Baumgartner L. Assessing developmental osteotoxicity of chlorides in the embryonic stem cell test. Reprod Toxicol 2010;30(2): 277–283.

50. zur Nieden NI, Kempka G, Ahr HJ. In vitro differentiation of embryonic stem cells into mineralized osteoblasts. Differentiation 2003;71(1): 18–27.

51. Zur Nieden NI, Kempka G, Ahr HJ. Molecular multiple endpoint embryonic stem cell test—a possible approach to test for the teratogenic potential of compounds. Toxicol Appl Pharmacol 2004;194(3): 257–269.

52. zur Nieden NI, Davis LA, Rancourt DE. Comparing three novel endpoints for developmental osteotoxicity in the embryonic stem cell test. Toxicol Appl Pharmacol 2010a;247(2): 91–97.

53. zur Nieden NI, Davis LA, Rancourt DE. Monolayer cultivation of osteoprogenitors shortens duration of the embryonic stem cell test while reliably predicting developmental osteotoxicity. Toxicology 2010b;277(1–3): 66–73.

54. zur Nieden NI, Kempka G, Rancourt DE, Ahr HJ. Induction of chondro-, osteo- and adipogenesis in embryonic stem cells by bone morphogenetic protein-2: effect of cofactors on differentiating lineages. BMC Dev Biol 2005;5: 1.

55. zur Nieden NI, Ruf LJ, Kempka G, Hildebrand H, Ahr HJ. Molecular markers in embryonic stem cells. Toxicol In Vitro 2001;15(4–5): 455–461.

CHAPTER 4

Differential predictivity of human pluripotent stem cell lines in skeletogenic

developmental toxicity assays correlates with their innate differentiation bias towards

neural crest or mesodermal osteogenesis

Nicole Renee Lee Sparks, Joseph Madrid, Riley Bottom, Nicole Isolde zur Nieden

Abstract

87.5% of registered human birth defects entail malformations of the skeleton, either in the craniofacial region or the limbs. A major preventable risk for the development of such deformities is chemical or drug exposure during pregnancy. However, unwanted chemical exposure may be prevented only if sufficient information on the developmental toxicity of such chemicals were available. The embryonic stem cell test (EST) comprehensively determines the potential of chemical entities to qualify as developmental toxicants based on multiple in vitro endpoints. These include the assessment of differentiation inhibition and the cytotoxic effect of the chemical in murine pluripotent mouse embryonic stem cells (ESCs) contrasted to the cytotoxic effect that this chemical elicits in fibroblasts. Due to the pluripotent nature of the ESCs used in this assay, the EST has been validated for multiple tissue endpoints including bone. However, potential interspecies variations in responses to chemicals raise concern about the efficacy of the rodent model to predict safe levels of compound exposure for humans. Thus, improvements

73 of the EST have focused on human-based modeling. Our present study demonstrates the promise of human pluripotent stem cells for in vitro developmental toxicity assays to uncover differentiation defects in the osteogenic lineage. First, our findings suggest that an assay based on human H9 ESCs can classify test compounds correctly, even for such compounds that classify incorrectly in the murine EST. Secondly, osteogenically differentiated H9 ESCs were more sensitive to chemicals known to negatively affect the developing neural crest, one of the two origins of bone tissue, than RIV9 human induced pluripotent stem cells (hiPSCs). This differential sensitivity correlated well with cell-line specific propensities to differentiate into neural crest or mesodermal osteoblasts.

Accordingly, analysis of TWIST1 mRNA, a gene primarily expressed in neural crest- derived osteogenic precursors, showed higher predictivity in the H9 cells than the RIV9 cells. We conclude from our results that the described test system using human pluripotent stem cells provides an accurate assay to identify and predict the potency of chemicals to elicit skeletogenic developmental toxicity that is biologically significant for humans.

Introduction

The etiology of major musculoskeletal birth defects remains largely unknown but plays a large role in public health concerns. Musculoskeletal birth defects that affect skeletal tissues are frequent (Parker et al., 2010) and impose life-long health consequences to the affected individuals and their families. Musculoskeletal anomalies include cleft lip and/or palate, oculo-auriculo-vertebral syndrome, limb reduction, and skeletal dysplasias.

Cleft lip/cleft palate, a congenital abnormality where the tissues of the upper lip or palate

74 do not fuse, occurs in 2-10/10,000 births (Burton, 2008). Limb reductions has an incidence of 2.5-7.06 per 10,000 births (Burton, 2008). Skeletal dysplasias, disorders that affect the development of the bone and cartilage, consists of hundreds of different classifications with an incidence of 2.1-2.3/10,000 births (Burton, 2008; Krakow and Rimoin, 2010). It is also plausible that these incidences could be higher because many anomalies do not manifest until later in childhood.

While genetic factors can be an underlying cause for some of these birth defects, others stem from involuntary environmental and chemical exposure of the fetus while developing in utero. To reduce the incidence rate of preventable birth defects, it is thus of utmost importance to identify chemicals that contribute to musculoskeletal maldevelopment. Stem-cell-based assays, which are also equipped to eliminate the use of animals, are gaining attention for this purpose. Among them, the Embryonic Stem Cell Test

(EST) is the most advanced (Genschow et al., 2000; Pulyanina and zur Nieden, 2012;

Kameoka et al., 2014).

It has been almost 25 years since Spielmann and colleagues introduced the mouse cell-based EST to assess the embryotoxic potential of chemical entities or blends in vitro

(Heuer et al., 1993; Scholz et al., 1999a; reviewed in Seiler and Spielmann, 2011).

Historically, this test utilized three endpoints. Mouse fibroblasts are exposed to a compound to assess maternal effects of the tested chemical. Any noted cytotoxic effect is then contrasted to the effect of the chemical on pluripotent mouse embryonic stem cell

(ESC) viability and differentiation propensity (Genschow et al., 2000; Scholz et al., 1999b).

While classically evaluating the inhibition of cardiac differentiation, a repertoire of

75 osteogenic, chondrogenic, endothelial, and neural differentiation was later added to test for cell lineage specific effects (de Jong et al., 2014; Festag et al., 2007a; 2007b; Hayess et al.,

2013; Stummann et al., 2009; zur Nieden et al., 2004; 2010). This addition of tissue endpoints in combination with molecular endpoints seemed necessary to improve the predictivity of the EST (Koseki et al., 2010; Schenk et al., 2010; Seiler et al., 2004; van

Dartel and Piersma, 2011). For example, in comparing the effects of compounds in the osteogenic and cardiogenic mouse EST, de Jong et al. (2014) showed the toxic effects of

2,3,7,8-tetrachlorodibenzo-p-dioxin on osteoblast, but not on cardiomyocyte differentiation.

Although highly predictive in comparison to other routinely used in vitro embryotoxicity assays (Scholz et al., 1999b; Genschow et al., 2002), the mouse-based EST still bares the risk of providing false negative or positive results based on potentially varied adverse outcome pathways in human versus the mouse. Changing the test species to non- human primate cells yielded differential responses between cells from rhesus monkey, marmoset monkey, and mouse, with no obvious dependency on chemical classes (Walker et al., 2014), supporting this hypothesis. Consequently, determining a safe exposure level for human use may be difficult for certain chemicals due to variations in response between different species, implying that human ESCs may be the most promising model to be used in such assays.

Indeed, commercially available mesenchymal stem cells derived from a human

ESC line have been evaluated for predicting skeletogenic developmental toxicity (Sittner et al., 2016). The utilization of mineralization defects and their comparison with

76 cytotoxicity of chemicals to assess exposure, represented a step in the right direction. The study did not include the comparison to cytotoxicity on somatic cells. More importantly, a disadvantage of the use of mesenchymal stem cells is that, while they represent a source for osteoblasts, they themselves may arise from different lineages during embryogenesis: neural crest or mesodermal cells. The commitment of neural crest and mesodermal cells may be affected differently by chemicals, which may be missed when assaying osteogenesis only from the vantage point of a mesenchymal cell.

Here, we evaluated the ability of a comprehensive human-based EST that analyzes all three known mouse EST endpoints to predict skeletal embryotoxicity based on full osteogenic lineage differentiation from undifferentiated hESCs. In addition, we expanded our assessment to a line of human induced pluripotent stem cells (hiPSCs), a source of pluripotent stem cells artificially created from somatic cells (Takahashi et al., 2007).

Together, our results illustrate the limitations of the mouse-based EST and emphasize the effectiveness of human-based in vitro models provided appropriate cell source/marker analysis combinations are used.

Materials and methods

Fibroblast culture

Human foreskin fibroblasts (HFFs) were graciously provided by Dr. Derrick

Rancourt (University of Calgary, Canada) at passage 1. The cells were maintained in high glucose L-glutamine Dulbecco’s modified Eagle’s medium (DMEM, Corning) with 10% fetal bovine serum (FBS, Atlanta), 1% non-essential amino acids (NEAA, Gibco), and

0.5% penicillin/streptomycin (10,000 units/10,000 units, Gibco). Murine 3T3 fibroblasts

(American Type Culture Collection, Rockville, MD, USA) were seeded on gelatin coated culture plates and cultured in high glucose L-glutamine DMEM supplemented with 15%

FBS and 1% penicillin/streptomycin (10,000 units/10,000 units).

Culture of embryonic and induced pluripotent stem cells

Cells of the human H9 ESC line were acquired from WiCell (WiCell Research

Institute). RIV9 hiPSC cells were previously generated from foreskin fibroblasts by retroviral integration of OCT4, KLF4 and SOX2 at the University of California Riverside’s

Stem Cell Core Facility (Chatterjee et al., 2011; Torrez et al., 2012; Sparks et al., 2018).

All human pluripotent cells were cultured on Matrigel (BD Biosciences) treated dishes and maintained as feeder-free cultures in mTeSR® media (Stem Cell Technologies) in 5% CO2 and at 37°C. Pluripotent colonies were passaged every 5 days upon reaching 70% confluency by dissociating cells with accutase and a cell scraper.

Murine D3 ESCs (American Type Culture Collection, Rockville, MD, USA) were expanded in high glucose DMEM containing L-glutamine, 15% FBS (selected batch), 1%

NEAA, and 1% penicillin/streptomycin (10,000 units/10,000 units), 0.1 mM 2- mercaptoethanol (Sigma), and 1000 U LIF/ml (Millipore). Cells were sub-cultured every

2–3 days with 0.25% Trypsin-EDTA (Life Technologies).

Osteogenic differentiation

Murine ESCs were induced to differentiate via aggregation into embryoid bodies and hanging drops at 750 cells/drop, in the presence of control differentiation medium

(CDM, mouse ESC maintenance medium without LIF) (zur Nieden et al., 2003).

Differentiating cells were replated on day 5 as a single cell suspension at a concentration of 50,000 cells/cm2 (zur Nieden et al., 2010a).

Confluent pluripotent human stem cell colonies previously maintained in mTeSR remained in their culture vessel (designated day 0), and the media was changed to CDM.

On day 5 of differentiation, cells from all species received osteogenic differentiation medium containing the induction factors β-glycerophosphate (10 mM), ascorbic acid (25

-8 μg/ml), and 1α,25-(OH)2 vitamin D3 (5 × 10 M) in CDM.

Histochemical analysis

Cells were washed with 1X PBS and fixed in 4% paraformaldehyde at 4°C for 1 hr on day 30 of differentiation. For von Kossa stain, fixed samples were overlaid with silver nitrate solution (Ricca Chemical Company) and illuminated under an intense light source for 1 hr. Un-reacted silver was removed with 5% sodium thiosulfate (Red Bird Service) for

2 minutes. Calcium-rich areas were also visualized with 2% w/v Alizarin Red S (Sigma-

Aldrich, pH 4.1-4.3). Cells were washed with PBS and overlaid with the Alizarin Red S solution for 5 minutes to form an Alizarin Red S-calcium complex. Cultures were washed

3X with water followed by ascending alcohol washes (70%, 80%, 90%, 100% ethanol).

Immunocytochemistry

Cells were washed with 1X PBS and fixed with 4% paraformaldehyde for 30 min at 4°C. Membranes were permeabilized for intracellular antigen staining with 0.1% Triton-

X 100 (Sigma) for 15 min and then incubated overnight at 4°C in PBS/10% FBS containing either OCN (Abcam, ab1857), SOX10 (Santa Cruz, SC17342), TWIST1 (Abcam, ab50887) and/or PAX7 (mIgG1, Developmental Studies Hybridoma Bank) antibody. Cells were then washed and incubated in the appropriate secondary antibody (AlexaFluor conjugated) for 2h and observed on a Nikon fluorescent microscope.

Test compounds

The chemicals 5-fluorouracil (5FU), all-trans-Retinoic acid (atRA), and penicillin

G sodium salt (PenG) were used in the original EST validation study (Genschow et al.,

2004) and were used here as positive and negative control model compounds. Exposure to

13-cis-Retinoic acid (13cisRA), a known human teratogen with lesser effects in the mouse

(Nau, 2001), was chosen for demonstrating potentially variable outcomes depending on cell lines examined. Cyclopamine, methoxyacetic acid (MAA), triadimefon, and triadimenol were selected for their known inhibitory effects on osteoblasts generated from the neural crest and/or the mesoderm (Jeong et al., 2004; Brown et al., 1984; Papis et al.,

2006; Zimmer et al., 2012). All test compounds were purchased from Sigma-Aldrich and either dissolved in DMSO or cell culture medium, based on solubility of the compound

(Buesen et al., 2009). Differentiation of ESCs was initiated as described above and cells were treated for 20 days of osteogenesis with seven or eight concentrations of test

80 compound. As per ECVAM standards, the maximum final concentration of any test chemical was 1000 µg/ml (Spielmann and Seiler, 2010). A solvent control containing the highest concentration of solvent was included to control for non-chemical specific effects.

All compounds were replenished with each media change.

Mitochondrial dehydrogenase activity assay

Cellular survival in response to test compounds was determined by 3-[4,5- dimethylthiazol-2-yl]-2,5-diphenylterazolium bromide (MTT, Spectrum Chemicals).

Briefly, cells were incubated with MTT (0.5 mg/ml) at 37 °C for 2 hr. After the supernatant was removed, cells were gently rocked in pre-warmed MTT desorb solution (0.7% sodium dodecyl sulfate (SDS) in 2-propanol) for 15 min, and the optical density of the solution was read at 570 nm in an iMark™ microplate reader (Bio-Rad). As the generation of the blue formazan product is proportional to the dehydrogenase activity, a decrease in the absorbance at 570 nm provided a direct measurement of the number of viable cells. Activity was normalized to solvent only controls and resulting percentages were graphed along the tested concentration range to construct a concentration–response curve. The half-maximal inhibitory effect (IC50) for each compound was subsequently established via linear interpolation of the curve.

Calcium assay

For quantification of extracellular matrix calcium, cells were harvested in radioimmunoprecipitation buffer (RIPA; 1XPBS, 1% NP-40, 0.5% sodium deoxycholate,

0.1% SDS, pH 7.4, and 1:100 protease inhibitors). Calcium deposition was determined based on calcium ions (Ca2+) reacting with Arsenazo III (Genzyme, Canada) to form a purple Ca-Arsenazo III complex, which was measured at 655 nm. The concentration of total calcium in the sample was calculated based on a CaCl2 standard and calcium content was normalized to the total protein content of the sample using the Lowry method (Bio-

Rad) (Davis et al., 2011; Kuske et al., 2012) as we have also done previously when screening with mouse cells (zur Nieden et al., 2004; zur Nieden et al., 2010a, b). This normalization is important to truly assess the differentiation level of the cultures without interference from cytotoxic or proliferative events. Calcium content was then standardized to solvent only controls and concentration-response curves charted. The half-maximal inhibitory dose (ID50Ca) for each compound was taken from linear interpolation of the curve.

RNA isolation and quantitative PCR

To analyze mRNA expression patterns upon differentiation or compound exposure cells were lysed in RA1 lysis buffer and total RNA was extracted according to the

NucleoSpin RNA II protocol (Macherey Nagel). RNA was quantified with a NanoDrop®

1000 spectrophotometer (Thermo Scientific) at 260 nm. 25 ng of total RNA was used as a template for cDNA synthesis with a mastermix including 1X reaction buffer, 0.5 mM dNTPs, 20 U/µL RNase inhibitor, 0.8 U/µL reverse transcriptase, and 1 μM random primer.

25 ng of cDNA transcripts were used for quantitative polymerase chain reaction (qPCR) with SYBR green on the MyiQ cycler (BIO-RAD). The primers for TWIST1 were 5’-

ACTGGCCTGCAAAACCATA-3’ (forward) and 5’-TGCATTTTACCATGGGTCCT-3’

(reverse). The primers for GAPDH were 5’-GAGTCAACGGATTTGGTCGT-3’ (forward) and 5’-TTGATTTTGGAGGGATCTCG-3’ (reverse). Reactions were setup for 10 minutes of denaturing at 94°C, followed by 40 cycles of denaturing at 94°C, and annealing/elongation at 60°C for 45 seconds. The n-fold expression in target samples was calculated with the ΔΔCT method by standardizing CT values to GAPDH expression (Livak and Schmittgen, 2001). Gene expression was normalized to solvent only controls and concentration-response curves charted. The half-maximal inhibitory dose (ID50TWIST1) for each compound was taken from linear interpolation of the curve.

Statistical analysis

Data presented are averages of three independent experiments (qPCR) or five technical replicates ± standard deviation. Significance was determined by using a web- based one-way ANOVA and Tukey HSD post hoc test

(http://faculty.vassar.edu/lowry/anova1u.html) or unpaired Student's t-test as appropriate.

Results

Comparison between predictivity for 13cisRA in mouse and human ESCs

To determine whether human cells would more accurately mimic phenotypes found after human exposure than mouse cells, we tested a selected set of four chemicals for which the outcome of in vivo exposures is adequately described. Penicillin G (PenG) was used as a non-embryotoxic chemical, 5-fluorouracil (5FU) as a strong embryotoxic chemical with

83 general cytotoxicity due to interference with DNA synthesis, and all-trans-Retinoic acid atRA because it has shown concentration-dependent toxicity on the osteogenic lineage

(Scholz et al., 1999; Genschow et al., 2000; zur Nieden et al., 2004; zur Nieden et al., 2010).

In addition, we included 13cisRA, which shows stronger exposure responses in humans than in rodents and the profound unwanted effects of retinoids on bone (Kraft et al., 1987;

Nau, 2001; DiGiovanna, 2001).

Towards this goal, mouse and human ESCs were differentiated according to Fig.

4.1A and concomitantly exposed to the test chemical. Murine and non-human primate

ESCs have been reported to differentiate into functional osteoblasts capable of mineralizing their extracellular matrix (ECM) (zur Nieden et al., 2003; Ding et al., 2012; Trettner et al.,

2014; Sparks et al., 2018). Solvent controls exhibited the expected mineralization of the matrix (Fig. 4.1B), which comprises the incorporation of calcium visible as black deposit in bright field microscopy (Fig. 4.1C). In the location of the mineralized areas, Alizarin

Red S and von Kossa stains positively confirmed deposition of calcium. The calcified dark areas of the ECM were also immuno-reactive to osteocalcin (Fig. 4.1C).

All chemicals were then tested in concentration response curves (Appendix Figs.

1.4.1 and 1.4.2) and half-maximal inhibitory concentrations for cell viability (IC50) and differentiation (ID50Ca) were established (Fig. 1D). In addition to the assessment of cell viability and differentiation inhibition on ESCs, the mouse EST compares the IC50 and

ID50 to the cell viability of 3T3 fibroblast cells of the same test compound (Scholz et al.,

1999), to represent the cytotoxic effect of the test chemical on the mother. For the inclusion

84 of this third endpoint in a human cell-based assay we measured cell viability on exposed human fibroblasts.

Figure 4.1. Human ESCs are more sensitive in predicting the developmental toxicity of 13cisRA, a known human teratogen, than mouse ESCs. (A) Schematic of protocols for osteogenic differentiation of mouse (D3) and human (H9) ESCs. (B) Calcium content in mouse and human ESC cultures differentiated per protocols in (A), either with the addition of the osteogenic factors Vitamin D3, ascorbic acid and β-glycerophosphate or without. (C) Photomicrographs of H9 hESC osteogenic cultures on day 30 of differentiation. Black deposits in brightfield images and positivity for von Kossa, Alizarin Red S and osteocalcin (OCN) indicate the presence of osteoblasts. Bar = 100 µm. (D) Mouse and human ESCs were exposed to listed compounds during differentiation. Calcium content was measured to determine the degree of osteogenic differentiation and mitochondrial dehydrogenase activity served as a measure of compound cytotoxicity. Human foreskin fibroblasts and mouse 3T3 fibroblasts were also exposed and assayed with the MTT assay. The listed half-maximal inhibitory concentrations (IC50 and ID50) were determined from concentration response curves that were generated for all cell lines and compounds (see also supplemental Figures 1 and 2) and are expressed in µg/mL. Mouse ESC and 3T3 fibroblast data for Pen G (penicillin G), 5FU (5- fluorouracil) and atRA (all-trans retinoic acid) taken from zur Nieden et al., 2004. (E) Correlation between half- maximal inhibitory doses found for cell viability (IC50) and differentiation (ID50Ca) with murine and human ESCs. (F) Classification of the test compounds per ECVAM’s biostatistical model. 5FU, 5-fluorouracil; atRA, all-trans Retinoic acid; hESC, human embryonic stem cell; hFF, human foreskin fibroblast; PenG, pnicillin G.

The three endpoints were then summarized in a biostatistical model that allowed for the classification of the test compound into three embryotoxicity classes: non-, weak, and strong embryotoxic potential (Genschow et al., 2000, 2002). This biostatistical model classified the three control compounds similarly between species (Fig. 4.1E), despite different sensitivity ranges (compare Appendix Fig. 1.4.1). For the strong embryotoxicant

5FU, human ESC cell viability and osteoblast differentiation endpoints were over 100-fold more sensitive compared to the mouse endpoints. Conversely, murine ESCs were more sensitive towards the strong embryotoxicant atRA than human ESCs, a pattern reflected when human fibroblasts were tested. Nonetheless, 5FU and atRA were predicted to be strong embryotoxicants and PenG to be non-embryotoxic matching their in vivo embryotoxicity classification.

For 13cisRA, half-maximal inhibitory doses were also lower when tested with human cells (Fig. 4.1D, Appendix Figure 1.4.2). The biostatistical model classified

13cisRA as non-embryotoxic (class I) based on the sensitivity of the mouse cells, and as moderately embryotoxic based on the human ESCs (class II; Fig. 4.1D). This suggests in vitro species-species variability in sensitivity to a compound that was similarly also found to be more embryotoxic in vivo. Half-maximal inhibitory doses for cell viability and differentiation obtained from both mouse and human experiments were next charted against each other in a linear regression model (Fig. 4.1F). Both charts showed a low correlation between the murine and human ESC endpoints producing R2 values of 0.52 and 0.57, respectively, again underscoring the differential sensitivity to toxicants between the species.

Figure 4.2. Differentiation and cell viability endpoints measured with RIV9 hiPSCs. (A) Mitochondrial dehydrogenase activity and (B) calcium content (mg Ca2+/mg protein) in response to chemical exposure. Both n=5±SD, *p<0.05 One-Way ANOVA. (C) Linear regression analysis for tested compounds between mitochondrial dehydrogenase activity and calcium content measured with H9 hESCs and RIV9 hiPSCs. 5FU, 5-fluorouracil; atRA, all-trans retinoic acid; 13cisRA, 13-cis-Retinoic Acid; hiPSC, human induced pluripotent stem cell; MTT, 3-[4,5- dimethylthiazol-2-yl]-2,5-diphenylterazolium bromide; PenG, penicillin G.

Evaluating the predictivity of RIV9 hiPSCs

Human iPSCs hold great promise in embryotoxicity screening assays, because they are cells generated from the reprogramming of somatic cells lacking the ethical challenges that hESCs face. To assess the suitability of an hiPSC cell line in the prediction of chemical embryotoxicity, RIV9 hiPSCs were osteogenically differentiated and simultaneously exposed to all four of the test chemicals. With the exception of PenG, all compounds inhibited cell viability and calcium content (Fig. 4.2A, B).

To summarize the data, we next charted the results obtained with the H9 cells to those obtained with the RIV9 cells. The correlation between the IC50 values was high with

R2 = 0.998 (Fig. 4.2C). A similar strong correlation (R2 = 0.97), was observed between the

ID50Ca of the two cell lines. When used in the biostatistical prediction model, the compounds were classified similarly, independently of the human PSC line that was used

(Table 4.1). Therefore, the similarity between the responses to the test compounds in the two cell lines suggested the usefulness of RIV9 hiPSCs for in vitro embryotoxic testing to identify and predict putative embryotoxic compounds based on this selected set of chemicals.

Assessment of chemicals with lineage specific action

Recent studies from our lab have suggested that the propensity of various human pluripotent stem cell lines for osteogenic differentiation correlated with the methylation pattern of promoters important for lineage specification while cells are in the undifferentiated state (Sparks et al., 2018). This seems to dictate how efficiently H9 and

Figure 4.3. Human H9 ESCs preferentially undergo neural crest osteogenesis. (A) Calcium content was measured on day 30 of differentiation and indicates the general capability of both human pluripotent stem cell lines to generate osteoblasts; n=5 ± SD, *p<0.05 Student’s t-test compared to untreated, spontaneously differentiating cultures. §P<0.05 Student’s t-test compared to H9 hESCs. (B) Antibody stains on 4 and 9-day –old hPSC cultures using PAX7, SOX10, and TWIST1 antibodies show sequential expression of border plate specifier proteins, neural crest specifier proteins and proteins associated with neural crest migration in the H9 cells, but only to a lesser degree in the RIV9 cells. Instead, the enhanced presence of TBRA in differentiating RIV9 cells may indicate a higher degree of mesodermal differentiation (Herrmann et al., 1990; Herrmann, 1991). (C) Quantitative mRNA expression of H9 hESCs and RIV9 hiPSCs for mesoderm (top row) or neural crest-associated genes (bottom row), n=3 ± SD. *p<0.05 Student’s t-test over H9 hESCs. Gbx2, gastrulation brain homeobox 2; GSC, goosecoid; Mier1, mesoderm induction early response 1; Mixl1, mix paired-like homeobox; Pax7, paired box 7; T-BRA, T-Brachyury; TBX6, T-box 6; TFAP2a, transcription factor AP-2 alpha; Sox10, SRY-box 10; Zic, Zic family member.

RIV9 cells calcify (Fig. 4.3A). It also relates to the kind of bone that is specified. Many cranial bones are formed through intramembranous ossification, in which mesenchymal precursors created from neural crest cells condense and directly differentiate into osteoblasts. In contrast, most bones, from the mesoderm lineage, undergo endochondral ossification, which requires a cartilage precursor to provide the framework for

89 vascularization and osteoblast calcification (Regard et al., 2012). In H9 cells, antibody stains on 4 and 9-day –old cultures using PAX7, SOX10, and TWIST1 antibodies show sequential expression of neural plate border specifier proteins, neural crest specifier proteins, and proteins associated with neural crest migration. In RIV9 cells these proteins were expressed to a lesser degree. Instead, the mesodermal protein, TBRA, (Herrmann et al., 1990; Herrmann, 1991) was enhanced. Quantitative mRNA expression for mesoderm and neural crest-associated genes also revealed that RIV9 cells significantly expressed mesoderm-associated genes over the H9 cells but expressed neural plate border specifier genes and neural crest specifiers at a significantly lower level. Additionally, H9 cells displayed a higher potential to generate neural crest cells when taken through a protocol that will directly induce this type of progenitor cell (Leung et al., 2016) (Fig. 4.3D). In line with our previous study, these data suggested that H9 cells display a higher propensity to generate neural crest osteoblasts (Sparks et al., 2018).

Thus, we hypothesized that the H9 cells could potentially be more sensitive to chemicals that specifically target the development of bone from the neural crest than the

RIV9 cells. To test this hypothesis, both cell lines were exposed to compounds with known in vivo toxicity to the neural crest: cyclopamine, methoxyacetic acid, triademefon and triademenol (Scott et al., 1989; Dunn et al., 1995; Menegola et al., 2005; Zimmer et al.,

2012). Indeed, the H9 cells showed inhibition of calcification at lower concentrations than the RIV9 cells (Fig. 4.4A). In addition, this differentiation defect occurred at concentrations lower than those that caused cytotoxicity (Fig.4.4B, C). Correlating cytotoxicity with differentiation inhibition of the respective compounds revealed lower R2 values in the H9

90 cells (Fig. 4.4D), again suggesting that H9 cells were more sensitive to neural crest inhibitory chemicals than RIV9 cells.

Figure 4.4. Concentration-response curves for representative neural crest-associated inhibitory compounds. (A-C) Human embryonic stem cells, induced pluripotent stem cells, and foreskin fibroblasts were analyzed on day 20 for concentration dependent effects on differentiation (Ca2+) and cell viability (mitochondrial dehydrogenase activity) after exposure to chemicals known to interfere with neural crest development in vivo, n=5±SD. *p<0.05 One-Way ANOVA exposed H9 hESCs versus solvent only, ΔP<0.05 One-Way ANOVA exposed RIV9 hiPSCs versus solvent only, §P<0.05 Student’s t-test H9 hESCs versus RIV9 hiPSCs. (D) Linear regression analysis between the endpoints of cell viability and differentiation (calcium) for neural crest inhibitory compounds. hESCs, human embryonic stem cells; hiPSCs, human induced pluripotent stem cells; MAA, methoxyacetic acid; MTT, 3-[4,5-dimethylthiazol-2- yl]-2,5-diphenylterazolium bromide.

Shortening assay duration by determining TWIST1 expression

Previous studies by our group and others have found positive associations with the measurement of lineage gene expression and prediction of embryotoxic potential (Schulpen et al., 2013, 2015b; Suzuki et al., 2011; Theunissen et al., 2012; zur Nieden et al., 2001,

2004). In addition to providing an objectively quantifiable endpoint, these means of differentiation analyses have also often led to the shortening of assay duration. Specifically, this is achieved by determining the expression of genes found in progenitors rather than fully differentiated cells. We thus hypothesized that the ability of the test compound to inhibit calcification may also correlate to a defect in specific mRNA expression patterns.

Specifically, we aimed to test whether the determination of TWIST1 levels, expression of which occurs early during differentiation prior to mineralization (Sparks et al., 2018), would shorten the chemical assessment and equally predict the chemical’s in vivo embryotoxicity classification. This hypothesis was based on prior literature that has described TWIST1 as an epithelial to mesenchymal transition marker with an active role in mediating osteogenic potential, specifically during the specification of cranial bone (Yen et al., 2010). Prior research has suggested that TWIST1 promoter methylation status and expression patterns can be indicative of osteogenic yield in pluripotent stem cell cultures

(Sparks et al., 2018).

The assessment of TWIST1 mRNA expression produced ID50 values lower than those seen for calcification. With H9 cells, ID50TWIST1 was lower for three of the eight tested chemicals (13cisRA, MAA, triademenol). RIV9 cells had six out of the eight chemicals tested with decreased TWIST1 (Table 4.1; Appendix Figure 1.4.3). Due to the

92 higher sensitivity of the TWIST1 mRNA endpoint over the calcium assay in most of the chemicals, the correlation between these two endpoints was lesser for the RIV9 cells than for the H9 cells (R2 0.34 versus 0.84) (Fig. 4.5A). Notably, the biostatistical model classified 13cisRA as strong embryotoxicant (class III) in the endpoint of TWIST1 gene expression for both cell lines (Table 4.1).

Figure 4.5. Half-maximal inhibitory doses for cytotoxicity (IC50) as determined from concentration-response curves with MTT assay and differentiation (ID50) with Arsenazo calcium assay and quantitative PCR for TWIST1. (A) Linear regression analysis showing degree of correlation between different endpoints. (B) Bar graphs charting the established half-maximal inhibitory doses for all PSC endpoints. *p<0.05 Student’s t-test ID50 versus IC50, §P<0.05 Student’s t-test H9 hESCs versus RIV9 hiPSCs. 5FU, 5-fluorouracil; atRA, all-trans retinoic acid; 13cisRA, 13-cis- Retinoic Acid; hESCs, human embryonic stem cells; hiPSCs, human induced pluripotent stem cells; IC50, half- maximal inhibitory concentration of cytotoxicity; ID50, half-maximal differentiation inhibiting concentration; MAA, methoxyacetic acid; MTT, 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenylterazolium bromide; PenG, penicillin G.

Integration of concentration-response curves and biostatistical models to predict skeletal developmental toxicity

Developmental toxicity of a compound can be assessed by its cytotoxicity or ability to inhibit differentiation. In our model, we specifically aim to uncover the effects of the toxicants on the developing skeleton specifically instead of widespread cytotoxicity.

Calcification and TWIST1 mRNA endpoints showed the same embryotoxicity classification of PenG, 5FU and atRA (Table 4.1) for both hPSC lines. However, the comparison between IC50 and ID50 suggested that some compounds were cytotoxic while others had the ability to inhibit differentiation prior to cytotoxic induction. (Fig 4.5B).

Determined by the ID50Ca2+, 5FU, atRA, and 13cisRA were cytotoxic to H9 cells while the NC chemicals only inhibited differentiation. 5FU, atRA, and cyclopamine were cytotoxic to H9 cells as assessed by ID50TWIST1. ID50Ca2+ of the RIV9 cells showed 5

Table 4.1. Half-maximal inhibitory concentrations for all tested endpoints and prediction of embryotoxicity classes.

94 out of the 8 tested compounds were cytotoxic. All compounds solely inhibited differentiation in the RIV9 ID50TWIST1 screen.

Discussion

Over the past decade multiple efforts have been launched to develop human-based in vitro assays to identify and predict in vivo toxicity to determine the safest chemicals.

Initial studies examined various techniques to quantify cell viability (Adler et al., 2008a, b; Pal et al., 2011). This was followed by different methods to analyze the differentiation behavior of the cells (Colleoni et al., 2011; Krug et al., 2013), often for compound exposure on fully differentiated cells (Jahnke et al., 2013; Mathur et al., 2015). While the latter studies have their own merit, compound exposure of differentiated cells represents toxicity only and not developmental toxicity, thus are inadequate to understand the potential of a compound to induce birth defects. In addition, most of these studies were lacking a comparison between mouse and human cells to evaluate the predictivity of the in vitro models. Indeed, comparisons of global gene expression changes in differentiating human and mouse ESC-derived neurons during valproic acid exposure, suggested a substantial difference in response between the mouse and the human cells (Schulpen et al., 2015a, b).

While only 988 genes were commonly regulated, a total of 4913 genes were differentially regulated between the two species, likely leading to a differential exposure outcome between the species.

Although representing key steps towards developing a human-based assay, prior studies with human-derived ESCs did not incorporate all three endpoints of the original

95 mouse EST (mEST). In addition to the assessment of cell viability and differentiation inhibition on ESCs, the mouse EST compares these two endpoints to the cell viability of

3T3 fibroblast cells of the same test compound (Scholz et al., 1999), signifying the cytotoxic effect of the test chemical on the mother. Together, the three endpoints are summarized in a biostatistical model that allows for the classification of the test compound into three embryotoxicity classes: non-, weak and strong embryotoxic (Genschow et al.,

2000, 2002). To this end, we tested a series of chemicals in our human pluripotent stem cell (hPSC)-based developmental toxicity assay to accurately determine compound potency using all three original EST endpoints. Furthermore, we sought to demonstrate species differences upon compound exposure. Lastly, we determined whether our test was sensitive enough to distinguish between compound inhibition of different embryonic lineages.

Human ESCs display multiple advantages in the field of toxicology testing due to their pluripotency, which includes unlimited proliferation capability and the ability to generate cells of all three germ layers from a human source. While mouse and human ESCs can both differentiate into the same cell types, there are differences between the ESCs that may inhibit the ability to extrapolate rodent data for human predictions. For example, undifferentiated human ESCs form flat colonies and are seeded as small cell aggregates, while mouse ESCs can be dissociated as single cells and form 3D colonies. Similarly, murine ESCs require the presence of leukemia inhibitory factory (LIF) for maintenance of a pluripotent state (Nichols et al., 1990), which is insufficient to maintain pluripotency in human ESCs (Humphrey et al., 2004). There is also variability of in vitro differentiation

96 into cardiomyocytes. Cardiac differentiation is induced via embryoid bodies, which are either formed using the hanging drop protocol (mouse ESCs) or by culturing cells in suspension (human ESCs). Also, the timing with which contractile activity is first detected and for how long it is sustained differs between cells of the two species (Kehat et al., 2001;

Metzger et al., 1994; Wobus et al., 1991). Thirdly, the efficiency with which contractile structures are formed varies substantially (Chow et al., 2013; Kehat et al., 2001).

Based on such differences, we hypothesized that including human cell lines can improve the ability of the EST to detect embryotoxicants with relevance for human exposure. Our results indeed demonstrated the benefit of including the developmental toxicity of hESC and hiPSC-derived osteoblasts in comparison to the osteogenic endpoint of the mESC based EST (mESTo). Contrasting cell viability and calcification of exposed mouse ESCs, hESCs, and hiPSCs with inhibition of cell viability in somatic fibroblasts equally predicted the classification of 5FU, atRA, and PenG according to their known in vivo teratogenicity classes. To our knowledge, this is the first time that all three of the mouse EST endpoints were tested together in a human model. However, despite predicting the same developmental toxicity classes as mouse cells, the IC50 and ID50Ca for 5FU was developmental toxic to human ESCs at a concentration 100-fold lower than mouse cells.

This may indicate that the human PSC-ESTo is a potentially more sensitive in vitro embryotoxicity screening assay to determine the adverse effects of a test compound on the developing skeleton than the mouse ESC-based assay. As such, the human-based test could yield more appropriate results to set human exposure limits during risk assessment and subsequent legal regulation of chemicals.

In vitro differentiation of hESCs into osteoblasts is a robust process and does not require specially trained personnel to detect the presence of black deposits in culture signifying osteoblast mineralization (zur Nieden et al., 2003). Indirectly, osteoblastogenesis can be quantitated through calcium content and a differentiation endpoint (zur Nieden et al., 2007). The differentiation of osteoblasts as determined by the level of calcium is a suitable endpoint for the EST to measure the osteotoxic potential of compounds (zur Nieden et al., 2010a,b; Walker et al., 2015). However, calcium content alone cannot detect the potency of a compound and could cause false negative results. This is possibly due to the number of proliferative cells in culture that cannot be controlled and will vary between all treatments, including the control. Previous studies have observed the robustness of calcium content as an endpoint for osteoblast differentiation in the mouse

EST (de Jong et al., 2012; Chen et al., 2015). In these studies, however, there are discrepancies in calcium levels for the tested toxic compounds. Specifically, Chen et al. reported increased levels of calcium in 5FU treated cells. Chen et al. (2015) and de Jong et al. (2012), both concluded that measuring the amount of calcium was not sensitive for the osteoblast EST. Instead, our prior work suggested the detection of calcium ions in the extracellular matrix to be a suitable endpoint for detecting a chemical’s potential to interfere with osteogenesis when normalized to the overall protein content (zur Nieden et al., 2010a, b; Kuske et al., 2012). Here, we apply this endpoint to the human ESTo since it is an indicator of the efficiency of the differentiation protocol. A potential reason for the differential conclusion between these studies may have been the failed accounting for the number of viable cells in culture in the other studies.

In human PSCs, 13cisRA potently inhibited osteogenesis in both cell lines. We specifically chose to test 13cisRA because it is highly teratogenic to humans resulting in cardiovascular, central nervous system, limb, and craniofacial abnormalities. It has been well documented that humans are more sensitive to the developmental toxic effects of

13cisRA than rabbits and rodents (Kochhar et al., 1984). To elicit similar developmental toxic effects in rodents, doses of 13cisRA can be up to 100-fold higher than what is harmful in humans (Kraft et al.,1987; Tzimas et al., 1994; Nau, 2001). Potentially, such species variability may be attributed to different metabolism and pharmacokinetics (Nau,

1986;1995;2001). Indeed, our human cell-based in vitro study supported the greater sensitivity of human cells to 13cisRA and classified 13cisRA as weakly toxic, while the murine embryonic stem cells classified it as nontoxic.

Overall, the development of an in vitro assay for skeletal toxicity is aggravated by the complexity of the bone formation process, including different embryonic origins of the mesenchymal cells that turn into osteoblasts and the process by which this occurs.

Depending on whether neural crest cells or mesodermal progenitors are impacted, craniofacial or limb defects will ensue. For instance, cyclopamine, methoxyacetic acid, and the triazoles triadimefon and triadimenol have known effects on neural crest cells and thus the craniofacial skeleton (Scott et al., 1989; Dunn et al., 1995; Menegola et al., 2005; Papis et al., 2006; Di Renzo et al., 2007; Zimmer et al., 2012). We tested the ability of our hPSC-

ESTo to identify these compounds as inhibitors of osteogenesis. Our focus was to select compounds that would induce craniofacial abnormalities to the developing fetus and determine if our two hPSC lines can yield differential sensitivity to the compounds. Our

99 previous study has uncovered that H9 hESCs predominantly generate osteoblasts of the NC lineage and our RIV9 hiPSCs make mesoderm-derived osteoblasts (Sparks et al., 2018). In early development, when the craniofacial region is derived from neural crest cells as they migrate from the neural tube, sonic Hedgehog (Shh) pathways are necessary for neural crest formation and proper facial development (Jeong et.al, 2004). Cyclopamine, as a Shh inhibitor has been known to induce limb and craniofacial teratogenesis in variety of species and targets neural crest cells specifically, while other cell types such as trunk neural crest cells are resistant to its effects (Dunn et.al, 1995). Methoxyacetic acid (MAA) is a general growth retardant with a relatively long half-life when injected into pregnant non-human primates (Scott, 1989). MAA treated rats have shown to elicit skeletal malformations

(Brown, 1984). Other studies demonstrate embryonic effects on the suture of the craniofacial region (Teixidó, 2012). Traidimefon and triadimenol are widely used as antifungal agents in agriculture and are derived from triazoles. Vertebrate systems from xenopus to zebrafish to rodents have resulted in craniofacial malformations due to toxicant exposure (Mengola et.al, 2005; Papis et.al, 2006; Grimes et al., 2008). Triadimenol and triadimefon have been shown to alter neural crest migration in human models and have well documented evidence that both chemicals alter neural cell migration in vivo and in vitro (Zimmer, 2012). Here, the H9-ESTo showed a dose-dependent inhibition of osteogenesis to these compounds, while they did not inhibit RIV9 calcification and showed cellular toxicity only in the highest dose of Triadimefon. Since our previous study presented in chapter 2 has uncovered that H9 hESCs predominantly generate osteoblasts of the neural crest lineage and that the RIV9 hiPSCs predominantly made mesoderm-derived

100 osteoblasts (Sparks et al., 2018), the higher potency of these chemicals in the H9-ESTo suggests the expected potential targeting of NC cells by these compounds. Thus, the concentration-response curves seem to indicate the potential of our test cell lines to distinguish osteotoxicity as it is chemical-specifically induced in the different osteogenic lineages.

We also evaluated the effects of the compounds at the molecular level based on the expression of an early mesenchymal gene, TWIST1, with the goal to reduce the test duration. When selecting a candidate gene for assessment in the EST, it is necessary to choose markers exclusive to the lineage under investigation and not such that are ubiquitously expressed in other tissues. This is due to the pluripotent nature of the cells, which, even when triggered into a certain lineage, will generate contaminating cell types unless some purification step is included. As such, we have previously assessed osteogenic differentiation inhibition with murine ESCs based on the expression of osteocalcin (OCN)

(zur Nieden et al., 2004). OCN mRNA and protein expression is exclusively associated with a maturation stage of osteoblasts, at which extracellular matrix is secreted that is then calcified (Aubin and Liu, 1996). Therefore, OCN expression can successfully be detected at later stages of ESC osteogenic differentiation (Sparks et al., 2018; zur Nieden et al.,

2003). In the current study, TWIST1 was selected over OCN as it is expressed during early stages of osteogenesis and is known to have an important role in proper bone development

(Miraoui et al., 2010). In mesenchymal stem cells, which are more specialized than ESCs, it is the down-regulation of this gene that directs osteoblast differentiation (Miraoui et al.,

2010), while continued upregulation of TWIST1 will result in the maintenance of a stem

101 cell state (Goodnough et al., 2012). Conversely, the absence of TWIST1 in vivo leads to

Saethre-Chotzen syndrome, which is characterized by untimely closure of the cranial sutures (Huang et al., 2014; Isenmann et al., 2009). Our results using TWIST1 mRNA analysis indicate that the compound’s innate toxicities can be assayed correctly when examining levels of this mRNA in hESCs. However, TWIST1 was not a suitable endpoint for the RIV9 hiPSCs. TWIST1 mRNA was dose dependently downregulated in the hiPSCs exposed to the NC chemicals. This expression data does not correlate to the calcification data and potentially resulting in a false-positive evaluation. While seemingly suitable for the hESCs, this finding warrants further studies but overall shows the need to include molecular endpoints in the EST that are cell-type specific. Previously assessed mRNA levels of Collagen type I or Sparc show the need to find early lineage markers that are tissue-specific (de Jong et al., 2012). While treatment of murine ESCs with methotrexate

(MTX), a childhood chemotherapy agent, resulted in the downregulation of Col1A1, this down-regulation may not solely reflect an inhibitory effect on osteogenic differentiation, since Col1A1 is expressed in virtually every tissue including lungs, liver, kidney and muscle (Lejard et al., 2011; Mia et al., 2014; Tsuchida et al., 2003; Uezumi et al., 2011).

Similarly, the potency of MTX to reduce Col1A1 expression was observed at a concentration analogous to its IC50, suggesting it was either inhibitory to Col1A1 expression due to the general cytotoxicity of the compound or that the assessed time point did not reflect the actual differentiation stage of the cells. Indeed, Col1A1 expression is typically found in pre-osteoblasts (Ducy, 2000) possibly explaining why other studies find a down-regulation of this gene beyond a certain differentiation stage (Bilousova et al.,

102

2011; Darcic et al., 2001; Kalajzic et al., 2002; zur Nieden et al., 2003). According to the in vivo teratogenicity of MTX, which is known for skeletal defects due to arrest of bone growth, reducing proliferation of chondrocytes, and inducing apoptosis in chondrocytes

(Fan et al., 2009; Nilsson et al., 1984; Xian et al., 2008, 2007), treatment of murine ESCs with this compound did also down-regulate Runx2, a master regulator of osteogenesis and exclusive to that lineage (Ducy et al, 1997). While seemingly a predictive indicator of skeletogenic teratogenicity, Runx2 is expressed later during osteogenesis than TWIST1

(Yousfi et al., 2002) and we therefore chose TWIST1 for this study. All data together suggests no one gene is a suitable catch all for assessing developmental toxicity across multiple cell lines and species, and a combination of markers would be the best indicator of embryotoxic potential.

Another alternative assessment in the field of embryotoxicity are high throughput screens (HTS), that may potentially meet the demand of testing the thousands of compounds entering the environment each year. Using a reporter gene assay, van de Burg et al. (2015) took a mechanistic approach in creating a predictive, 82%, high throughput assay to assess reproductive toxicity. Though the study tested for one tissue endpoint and lacked the recapitulation of developing cells that hESCs can provide, the system demonstrates the effectiveness and potential of such systems. Kameoka et al. (2014) performed a fast and predictive teratogenicity screen on hESCs differentiated into tissues of the three germ layers. This testing strategy demonstrates how a developmental toxicant can perturb an essential process, gastrulation, which could lead to a birth defect. While predictive, the study was limited in demonstrating inhibition of a functional tissue, resulting

103 in unknown anomalies. Additively, adverse outcome resulting from exposure during later development was missed. In the hESTo, however, cells are exposed throughout all phases of development and this assay thus shows the capability to uncover developmental osteogenic toxicity.

Conclusion

We have implemented the use of human pluripotent stem cells to improve the ability of the EST for human risk assessment of compounds. The assay has shown sensitivity of osteogenically differentiating cells going through different developmental pathways in test compounds. The human cells successfully predicted the embryotoxic potential of 5FU, atRA, and 13cisRA with contrasting sensitivity compared to the mouse

ESCs. This emphasizes interspecies differences in susceptibility to environmental toxicants. The hESC-derived osteoblasts showed sensitivity to chemicals known to affects

NC cells while the hiPSCs showed no inhibition of osteogenesis, noting that researchers should fully understand the differentiation process of the cells being used. In sum, the present study provides evidence that the use of human pluripotent stem cells may enhance the EST and proposes that the human EST is a reliable indicator for testing human hazards.

104

References

1. Adler S, Basketter D, Creton S, Pelkonen O, van Benthem J, Zuang V, Andersen KE, Angers-Loustau A, Aptula A, Bal-Price A, et al. Alternative (non-animal) methods for cosmetics testing: current status and future prospects—2010. Arch Toxicol. 2011;85(5):367-485.

2. Adler, S., Lindqvist, J., Uddenberg, K., Hyllner, J., Strehl, R. Testing potential developmental toxicants with a cytotoxicity assay based on human embryonic stem cells. Altern Lab Anim 2008a;36(2):129-40.

3. Adler, S., Pellizzer, C., Hareng, L., Hartung, T., Bremer, S. (2008b). First steps in establishing a developmental toxicity test method based on human embryonic stem cells. Toxicol In Vitro 2008b;22(1):200-11.

4. Aubin, J.E., and Liu, F. (1996). The osteoblast lineage. In principles of bone biology (ed. J. P. Bilezikian L. G. Raisz and G. A. Rodan), pp. 51-67. San Diego: Academic Press.

5. B.G. Herrmann, S. Labeit, A. Poustka, T.R. King, H. Lehrach. Cloning of the T gene required in mesoderm formation in the mouse. Nature, 343 (1990), pp. 617–622.

6. B.G. Herrmann. Expression pattern of the Brachyury gene in whole-mount TWis/TWis mutant embryos. Development, 113 (1991), pp. 913–917.

7. Bilousova, G., Jun, du H., King, K.B., De Langhe, S., Chick, W.S., Torchia, E.C., Chow, K.S., Klemm, D.J., Roop, D.R., Majka, S.M. (2011). Osteoblasts derived from induced pluripotent stem cells form calcified structures in scaffolds both in vitro and in vivo. Stem Cells. 29(2):206-16.

8. Brown NA, Holt D, Webb M. The teratogenicity of methoxyacetic acid in the rat. Toxicol Lett. 1984 Jul;22(1):93-100.

9. Brown, N.A., Holt, D., Webb, M. (1984). The teratogenicity of methoxyacetic acid in the rat. Toxicol Lett. 22(1):93-100.

105

10. Buesen R, Genschow E, Slawik B, Visan A, Spielmann H, Luch A, Seiler A. Embryonic stem cell test remastered: comparison between the validated EST and the new molecular FACS-EST for assessing developmental toxicity in vitro. Toxicol Sci. 2009 Apr;108(2):389-400.

11. Burton, B. K., P, K. (Ed.), & BK, B. (Ed.) (2008). Congenital malformations: evidence- based evaluation and management. New York: Mc-Graw- Hill.https://www.scholars.northwestern.edu/en/publications/congenital- malformations-evidence-based-evaluation-and-management-3

12. Carney, E.W., Pottenger, L.H., Johnson, K.A., Liberacki, A.B., Tornesi, B., Dryzga, M.D., Hansen, S.C., Breslin, W.J. (2003). Significance of 2-methoxypropionic acid formed from beta-propylene glycol monomethyl ether: integration of pharmacokinetic and developmental toxicity assessments in rabbits. Toxicol. Sci. 71(2):217-28.

13. Centers for Disease Control and National Birth Defects Prevention Network (NBDPN). Preventing birth defects. Available at http://www.nbdpn.org/current/2008pdf/PrevBDBroch.pdf.

14. Chen, X., Hansen, D.K., Merry, G., DeJarnette, C., Nolen, G., Sloper, D., Fisher, J.E., Harrouk, W., Tassinari, M.S., Inselman, A.L. (2015). Developing osteoblasts as an endpoint for the mouse embryonic stem cell test. Reprod. Toxicol. 53:131-40.

15. Chow, M.Z., Boheler, K.R., Li, R.A. (2013). Human pluripotent stem cell-derived cardiomyocytes for heart regeneration, drug discovery and disease modeling: from the genetic, epigenetic, and tissue modeling perspectives. Stem. Cell. Res. Ther. 4(4):97.

16. Dacic, S., Kalajzic, I., Visnjic, D., Lichtler, A.C., Rowe, D.W. (2001). Col1a1-driven transgenic markers of osteoblast lineage progression. J. Bone Miner. Res. 16(7):1228- 36.

17. Davis, L.A., Dienelt, A., zur Nieden, N.I. (2011). Absorption-based assays for the analysis of osteogenic and chondrogenic yield. Methods Mol Biol. 690:255-72.

18. de Jong, E., van Beek, L., Piersma, A.H. (2014). Comparison of osteoblast and cardiomyocyte differentiation in the embryonic stem cell test for predicting embryotoxicity in vivo. Reprod. Toxicol. 48:62-71.

106

19. Di Renzo F, Broccia ML, Giavini E, Menegola E. Antifungal triazole derivative triadimefon induces ectopic maxillary cartilage by altering the morphogenesis of the first branchial arch. Birth Defects Res B Dev Reprod Toxicol. 2007;80(1):2-11.

20. Dienelt, A., and zur Nieden, N.I. (2011). Hyperglycemia impairs skeletogenesis from embryonic stem cells by affecting osteoblast and osteoclast differentiation. Stem Cells Dev. 20(3):465-74.

21. DiGiovanna JJ. Isotretinoin effects on bone. J Am Acad Dermatol 2001;45(5):S176- 82.

22. Ding, H., Keller, K.C., Martinez, I.K., Geransar, R.M., zur Nieden, K.O., Nishikawa, S.G., Rancourt, D.E., zur Nieden, N.I. (2012). NO-β-catenin crosstalk modulates primitive streak formation prior to embryonic stem cell osteogenic differentiation. J. Cell Sci. 125(Pt 22):5564-77.

23. Ducy P. (2000). Cbfa1: a molecular switch in osteoblast biology. Dev. Dyn. 219(4):461-71.

24. Ducy, P., Zhang, R., Geoffroy, V., Ridall, A.L., Karsenty, G. (1997). Osf2/Cbfa1: a transcriptional activator of osteoblast differentiation. Cell. 89(5):747-54.

25. Dunn MK, Mercola M, Moore DD. Cyclopamine, a steroidal alkaloid, disrupts development of cranial neural crest cells in Xenopus. Dev Dyn. 1995 Mar;202(3):255- 70.

26. Eckhoff, C., and Willhite, C.C. (1997). Embryonic delivered dose of isotretinoin (13- cis-retinoic acid) and its metabolites in hamsters. Toxicol. Appl. Pharmacol. 146(1):79- 87.

27. Fan, C., Cool, J.C., Scherer, M.A., Foster, B.K., Shandala, T., Tapp, H., Xian, C.J. (2009). Damaging effects of chronic low-dose methotrexate usage on primary bone formation in young rats and potential protective effects of folinic acid supplementary treatment. Bone. 44(1):61-70.

28. Festag, M., Viertel, B., Steinberg, P., Sehner, C. (2007a). An in vitro embryotoxicity assay based on the disturbance of the differentiation of murine embryonic stem cells into endothelial cells. I. Testing of compounds. Toxicol. In Vitro. 21(8):1619-30.

107

29. Festag, M., Viertel, B., Steinberg, P., Sehner, C. (2007b). An in vitro embryotoxicity assay based on the disturbance of the differentiation of murine embryonic stem cells into endothelial cells. II. Testing of compounds. Toxicol. In Vitro. 21(8):1631-40.

30. Genschow, E., Scholz, G., Brown, N., Piersma, A., Brady, M., Clemann, N., Huuskonen, H., Paillard, F., Bremer, S., Becker, K., Spielmann, H. (2000). Development of prediction models for three in vitro embryotoxicity tests in an ECVAM validation study. In Vitr. Mol. Toxicol. 13(1):51-66.

31. Goodnough, L.H., Chang, A.T., Treloar, C., Yang, J., Scacheri, P.C., Atit, R.P. (2012). Twist1 mediates repression of chondrogenesis by β-catenin to promote cranial bone progenitor specification. Development. 139(23):4428-4438.

32. Grimes AC, Erwin KN, Stadt HA, Hunter GL, Gefroh HA, Tsai HJ, Kirby ML. PCB126 exposure disrupts zebrafish ventricular and branchial but not early neural crest development. Toxicol Sci. 2008;106(1):193-205.

33. Hayess, K., Riebeling, C., Pirow, R., Steinfath, M., Sittner, D., Slawik, B., Luch, A., Seiler, A.E. (2013). The DNT-EST: a predictive embryonic stem cell-based assay for developmental neurotoxicity testing in vitro. Toxicology. 314(1):135-47.

34. Heuer J, Bremer S, Pohl I, Spielmann H. Development of an in vitro embryotoxicity test using murine embryonic stem cell cultures. Toxicol In Vitro. 1993 Jul;7(4):551-6. PubMed PMID: 20732251.

35. Hou, Z., Zhang, J., Schwartz, M.P., Stewart, R., Page, C.D., Murphy, W.L., Thomson, J.A. (2013). A human pluripotent stem cell platform for assessing developmental neural toxicity screening. Stem Cell Res. Ther. 4 Suppl 1:S12.

36. Huang, Y., Meng, T., Wang, S., Zhang, H., Mues, G., Qin, C., Feng, J.Q., D'Souza, R.N., Lu, Y. (2014). Twist1- and Twist2-haploinsufficiency results in reduced bone formation. PLoS One. 9(6):e99331.

37. Humphrey, R.K., Beattie, G.M., Lopez, A.D., Bucay, N., King, C.C., Firpo, M.T., Rose-John, S., Hayek, A. (2004). Maintenance of pluripotency in human embryonic stem cells is STAT3 independent. Stem Cells. 22(4):522-530.

108

38. Isenmann, S., Arthur, A., Zannettino, A.C., Turner, J.L., Shi, S., Glackin, C.A., Gronthos, S. (2009). TWIST family of basic helix-loop-helix transcription factors mediate human mesenchymal stem cell growth and commitment. Stem Cells. 27(10):2457-68.

39. Jahnke, H.G., Steel, D., Fleischer, S., Seidel, D., Kurz, R., Vinz, S., Dahlenborg, K., Sartipy, P., Robitzki, A.A. (2013). A novel 3D label-free monitoring system of hES- derived cardiomyocyte clusters: a step forward to in vitro cardiotoxicity testing. PLoS One. 8(7):e68971.

40. Jeong J, Mao J, Tenzen T, Kottmann AH, McMahon AP. Hedgehog signaling in the neural crest cells regulates the patterning and growth of facial primordia. Genes Dev. 2004 Apr 15;18(8):937-51.

41. Kalajzic, I., Kalajzic, Z., Kaliterna, M., Gronowicz, G., Clark, S.H., Lichtler, A.C., Rowe, D. (2002>). Use of type I collagen green fluorescent protein transgenes to identify subpopulations of cells at different stages of the osteoblast lineage. J. Bone Miner. Res. 17(1):15-25.

42. Kameoka, S., Babiarz, J., Kolaja, K., Chiao, E. (2014). A high-throughput screen for teratogens using human pluripotent stem cells. Toxicol. Sci. 137(1):76-90.

43. Kehat, I., Kenyagin-Karsenti, D., Snir, M., Segev, H., Amit, M., Gepstein, A., Livne, E., Binah, O., Itskovitz-Eldor, J., Gepstein, L. (2001). Human embryonic stem cells can differentiate into myocytes with structural and functional properties of cardiomyocytes. J. Clin. Invest. 108(3):407-14.

44. Kochhar DM, Penner JD, Tellone CI. Comparative teratogenic activities of two retinoids: effects on palate and limb development. Teratog Carcinog Mutagen. 1984;4(4):377-87.

45. Koseki, N., Deguchi, J., Yamada, T., Funabashi, H., Seki, T. (2010). Improvement of the embryonic stem cell test endpoint analysis by use of field potential detection. J. Toxicol. Sci. 35(5):619-29.

46. Kraft, J.C., Kochhar, D.M., Scott, W.J., Nau, H. (1987). Low teratogenicity of 13-cis- retinoic acid (isotretinoin) in the mouse corresponds to low embryo concentrations

109

during organogenesis: comparison to the all-trans isomer. Toxicol. Appl. Pharmacol. 87(3):474-82.

47. Krakow D, Rimoin DL. The skeletal dysplasias. Genetics in Medicine 2010;12(6):327- 41.

48. Krug, A.K., Kolde, R., Gaspar, J.A., Rempel, E., Balmer, N.V., Meganathan, K., Vojnits, K., Baquié, M., Waldmann, T., Ensenat-Waser, R, et al. (2013). Human embryonic stem cell-derived test systems for developmental neurotoxicity: a transcriptomics approach. Arch. Toxicol. 87(1):123-43.

49. Kugler J, Kemler R, Luch A, Oelgeschläger M. Editor's Highlight: Identification and Characterization of Teratogenic Chemicals Using Embryonic Stem Cells Isolated From a Wnt/β-Catenin-Reporter Transgenic Mouse Line. Toxicol Sci. 2016;152(2):382-94.

50. Kuske, B., Pulyanina, P.Y., zur Nieden, N.I. (2012). Embryonic stem cell test: stem cell use in predicting developmental cardiotoxicity and osteotoxicity. Methods Mol. Biol. 889:147-79.

51. Lejard, V., Blais, F., Guerquin, M.J., Bonnet, A., Bonnin, M.A., Havis, E., Malbouyres, M., Bidaud, C.B., Maro, G., Gilardi-Hebenstreit, P., Rossert, J., Ruggiero, F., Duprez, D. (2011). EGR1 and EGR2 involvement in vertebrate tendon differentiation. J. Biol. Chem. 286(7):5855-67.

52. Leung AW, Murdoch B, Salem AF, Prasad MS, Gomez GA, García-Castro MI. WNT/β-catenin signaling mediates human neural crest induction via a pre-neural border intermediate. Development 143(3):398-410 (2016).

53. Liu, W., Toyosawa, S., Furuichi, T., Kanatani, N., Yoshida, C., Liu, Y., Himeno, M., Narai, S., Yamaguchi, A., Komori, T. (2001). Overexpression of Cbfa1 in osteoblasts inhibits osteoblast maturation and causes osteopenia with multiple fractures. J. Cell Biol. 155(1):157-66.

54. Livak, K.J., Schmittgen, T.D. (2001). Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta C(T)) Method. Methods. 25:402-408.

55. Makori, N., Peterson, P.E., Blankenship, T.N., Dillard-Telm, L., Hummler, H., Hendrickx, A.G. (1998). Effects of 13-cis-retinoic acid on hindbrain and craniofacial

110

morphogenesis in long-tailed macaques (Macaca fascicularis). J. Med. Primatol. 27(4):210-9.

56. March of Dimes. (2010). March of dimes data book for policy makers: Maternal, infant, and child health in the United States. http://www.marchofdimes.org/advocacy/data- book-for-policy-makers-maternal-and-infant-health-in-the-united-states.aspx.

57. Mathur, A., Loskill, P., Shao, K., Huebsch, N., Hong, S., Marcus, S.G., Marks, N., Mandegar, M., Conklin, B.R., Lee, L.P., Healy, K.E. (2015). Human iPSC-based cardiac microphysiological system for drug screening applications. Sci. Rep. 5:8883.

58. McCarthy N, Liu JS, Richarte AM, Eskiocak B, Lovely CB, Tallquist MD, Eberhart JK. Pdgfra and Pdgfrb genetically interact during craniofacial development. Dev Dyn. 245(6):641-652 (2016).

59. Menegola E, Broccia ML, Di Renzo F, Massa V, Giavini E. Craniofacial and axial skeletal defects induced by the fungicide triadimefon in the mouse. Birth Defects Res B Dev Reprod Toxicol. 2005 Apr;74(2):185-95.

60. Metzger, J.M., Lin, W.I., Samuelson, L.C. (1994). Transition in cardiac contractile sensitivity to calcium during the in vitro differentiation of mouse embryonic stem cells. J. Cell Biol. 126(3):701-11.

61. Mia, M.M., Boersema, M., Bank, R.A. (2014). Interleukin-1β attenuates myofibroblast formation and extracellular matrix production in dermal and lung fibroblasts exposed to transforming growth factor-β1. PLoS One. 9(3):e91559.

62. Miraoui, H., Severe, N., Vaudin, P., Pagès, J.C., Marie, P.J. (2010). Molecular silencing of Twist1 enhances osteogenic differentiation of murine mesenchymal stem cells: implication of FGFR2 signaling. J. Cell Biochem. 110(5):1147-1154.

63. Nakamura K, Kusakawa, S, Tanoue, A. (2011). Assessment of Embryotoxicity and Teratogenicity by the Embryonic Stem Cell Test. In: Craig Atwood ed. Methodological Advances in the Culture, Manipulation and Utilization of Embryonic Stem Cells for Basic and Practical Applications. 10.5772/14644.

64. Narsinh, K.H., Sun, N., Sanchez-Freire, V., Lee, A.S., Almeida, P., Hu, S., Jan, T., Wilson, K.D., Leong, D., Rosenberg, J., Yao, M., Robbins, R.C., Wu, J.C. (2011).

111

Single cell transcriptional profiling reveals heterogeneity of human induced pluripotent stem cells. J. Clin. Invest. 121(3):1217-21.

65. Nau H. (2001). Teratogenicity of isotretinoin revisited: species variation and the role of all-trans-retinoic acid. J. Am. Acad. Dermatol. 45(5):S183-7.

66. Nau H. Chemical structure-teratogenicity relationships, toxicokinetics and metabolism in risk assessment of retinoids. Toxicol Lett 82-83 (1995)975–979.

67. Nau H. Species differences in pharmacokinetics and drug teratogenesis, Env Heal. Perspect. 70 (1986) 113–129.

68. Nichols, J., Evans, E.P., Smith, A.G. (1990). Establishment of germ-line-competent embryonic stem (ES) cells using differentiation inhibiting activity. Development. 110(4):1341-1348.

69. Nilsson, O.S., Bauer, F.C., Broström, L.A., Nilsonne, U. (1984). Effect of the antineoplastic agent methotrexate on experimental heterotopic new bone formation in rats. Cancer Res. 44(4):1653-6.

70. Pal, R., Mamidi, M.K., Das, A.K., Bhonde, R. (2011). Human embryonic stem cell proliferation and differentiation as parameters to evaluate developmental toxicity. J. Cell. Physiol. 226(6):1583-95.

71. Papis E, Bernardini G, Gornati R, Prati M. Triadimefon causes branchial arch malformations in Xenopus laevis embryos. Environ Sci Pollut Res Int. 2006 Jul;13(4):251-5.

72. Parker, S.E., Mai, C.T., Canfield, M.A., Rickard, R., Wang, Y., Meyer, R.E., Anderson, P., Mason, C.A., Collins, J.S., Kirby, R.S., Correa, A., National Birth Defects Prevention Network. (2010). Updated National Birth Prevalence estimates for selected birth defects in the United States, 2004-2006. Birth Defects Res. A Clin. Mol. Teratol. 88(12):1008-16.

73. Planello, A.C., Ji, J., Sharma, V., Singhania, R., Mbabaali, F., Müller, F., Alfaro, J.A., Bock, C., De Carvalho, D.D., Batada, N.N. (2014). Aberrant DNA methylation reprogramming during induced pluripotent stem cell generation is dependent on the choice of reprogramming factors. Cell Regen. (Lond). 3(1):4.

112

74. Pulyanina PY, zur Nieden NI (2012). Pluripotent stem cells as tools to assess developmental toxicity. In: Berhardt LV. (ed). Advances in Medicine and Biology. Volume 33. Nova Science Publishers, Inc., ISBN 978-1-61324-784-6

75. Regard, J.B., Zhong, Z., Williams, B.O., Yang, Y. (2012). Wnt signaling in bone development and disease: making stronger bone with Wnts. Cold. Spring. Harb. Perspect. Biol. 4(12).

76. Schenk, B., Weimer, M., Bremer, S., van der Burg, B., Cortvrindt, R., Freyberger, A., Lazzari, G., Pellizzer, C., Piersma, A., Schäfer, W.R., Seiler, A., Witters, H., Schwarz, M. (2010). The ReProTect Feasibility Study, a novel comprehensive in vitro approach to detect reproductive toxicants. Reprod. Toxicol. 30(1):200-18.

77. Scholz G, Pohl I, Genschow E, Klemm M, Spielmann H. Embryotoxicity screening using embryonic stem cells in vitro: correlation to in vivo teratogenicity. Cells Tissues Organs. 1999a;165(3-4):203-11.

78. Scholz, G., Genschow, E., Pohl, I., Bremer, S., Paparella, M., Raabe, H., Southee, J., Spielmann, H. (1999b). Prevalidation of the embryonic stem cell test (EST)-A new in vitro embryotoxicity test. Toxicol. In Vitro. 13(4-5):675-81.

79. Schulpen, S.H., Pennings, J.L., Piersma, A.H. (2015a). Gene Expression Regulation and Pathway Analysis after Valproic Acid and Carbamazepine Exposure in a Human Embryonic Stem Cell-Based Neurodevelopmental Toxicity Assay. Toxicol. Sci. 146(2):311-20.

80. Schulpen, S.H., Robinson, J.F., Pennings, J.L., van Dartel, D.A., Piersma, A.H. (2013). Dose response analysis of monophthalates in the murine embryonic stem cell test assessed by cardiomyocyte differentiation and gene expression. Reprod. Toxicol. 35:81-8.

81. Schulpen, S.H., Theunissen, P.T., Pennings, J.L., Piersma, A.H. (2015b). Comparison of gene expression regulation in mouse- and human embryonic stem cell assays during neural differentiation and in response to valproic acid exposure. Reprod. Toxicol. 56:77-86.

113

82. Scott WJ, Fradkin R, Wittfoht W, Nau H. Teratologic potential of 2-methoxyethanol and transplacental distribution of its metabolite, 2-methoxyacetic acid, in non-human primates. Teratology. 1989 Apr;39(4):363-73.

83. Seiler, A.E., and Spielmann, H. (2011). The validated embryonic stem cell test to predict embryotoxicity in vitro. Nat. Protoc. 6(7):961-78.

84. Seiler, A.E., Visan, A., Buesen, R., Genschow, E., Spielmann, H. (2004). Improvement of an in vitro stem cell assay for developmental toxicity: the use of molecular endpoints in the embryonic stem cell test. Reprod. Toxicol. 18(2):231-40.

85. Sittner D, Huhse B, Steinfath M, Luch A, Seiler AE.M. Osteogenic Differentiation of Human Embryonic Stem Cell-Derived Mesenchymal Progenitor Cells as a Model for Assessing Developmental Bone Toxicity In Vitro. Applied In Vitro Toxicology. September 2016, 2(3): 127-142.

86. Smith CL, Tallquist MD. PDGF function in diverse neural crest cell populations. Cell Adh Migr, 4(4):561-566 (2010).

87. Sparks NRL, Martinez IKC, Soto CH, zur Nieden NI. Low Osteogenic Yield in Human Pluripotent Stem Cells Associates with Differential Neural Crest Promoter Methylation. Stem Cells 2018;36(3):349-362.

88. Spielmann, H., and Seiler, AE. (2010). ECVAM DB-ALM: Protocols. Embryonic stem cell test DB-ALM protocol n° 113. http://ecvam- dbalm.jrc.ec.europa.eu/public_view_doc2.cfm?id=DC5ABDF7AC30F1B7DF7EF27 E87D68AAC7180BB0BC12CB10496CDA74B54630A05A3291B895581F634.Acce ssed9/04/2015.

89. Stummann, T.C., Hareng, L., Bremer, S. (2009). Hazard assessment of methylmercury toxicity to neuronal induction in embryogenesis using human embryonic stem cells. Toxicology. 257(3):117-26.

90. Suzuki, N., Ando, S., Yamashita, N., Horie, N., Saito, K. (2011). Evaluation of novel high-throughput embryonic stem cell tests with new molecular markers for screening embryotoxic chemicals in vitro. Toxicol. Sci. 124(2):460-71.

114

91. Swanson, A.B., Swanson, G.D., Tada, K. (1983). A classification for congenital limb malformation. J Hand Surg Am. 8(5 Pt 2):693-702.

92. Takahashi, K., Tanabe, K., Ohnuki, M., Narita, M., Ichisaka, T., Tomoda, K., Yamanaka, S. (2007). Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell. 131(5):861–72.

93. Takakura N, Yoshida H, Ogura Y, Kataoka H, Nishikawa S, Nishikawa S. PDGFR alpha expression during mouse embryogenesis: immunolocalization analyzed by whole-mount immunohistostaining using the monoclonal anti-mouse PDGFR alpha antibody APA5.J Histochem Cytochem. 45(6):883-893 (1997).

94. Teixidó E, Piqué E, Gómez-Catalán J, Llobet JM. Assessment of developmental delay in the zebrafish embryo teratogenicity assay. Toxicol In Vitro. 2013 Feb;27(1):469-78. doi: 10.1016/j.tiv.2012.07.010. Epub 2012 Aug 8.

95. Theunissen, P.T., Robinson, J.F., Pennings, J.L., van Herwijnen, M.H., Kleinjans, J.C., Piersma, A.H. (2012). Compound-specific effects of diverse neurodevelopmental toxicants on global gene expression in the neural embryonic stem cell test (ESTn). Toxicol. Appl. Pharmacol. 262(3):330-40.

96. Thomson, J.A., Itskovitz-Eldor, J., Shapiro, S.S., Waknitz, M.A., Swiergiel, J.J., Marshall, V.S., Jones, J.M. (1998). Embryonic stem cell lines derived from human blastocysts. Science. 282(5391):1145–7.

97. Trainor, P. (2010). Craniofacial Birth Defects: The role of neural crest cells in the etiology and pathogenesis of treacher Collins syndrome and potential for prevention. Am. J. Med. Genet. A. 152A(12):2984-94.

98. Trettner, S., Findeisen, A., Taube, S., Horn, P.A., Sasaki, E., zur Nieden, N.I. (2014). Osteogenic induction from marmoset embryonic stem cells cultured in feeder- dependent and feeder-independent conditions. Osteoporos. Int. 25(4):1255-66.

99. Tsuchida, K., Zhu, Y., Siva, S., Dunn, S.R., Sharma, K. (2003). Role of Smad4 on TGF- beta-induced extracellular matrix stimulation in mesangial cells. Kidney Int. 63(6):2000-9.

115

100. Tzimas G, Bürgin H, Collins MD, Hummler H, Nau H. The high sensitivity of the rabbit to the teratogenic effects of 13-cis-retinoic acid (isotretinoin) is a consequence of prolonged exposure of the embryo to 13-cis-retinoic acid and 13-cis-4-oxo-retinoic acid, and not of isomerization to all-trans-retinoic acid. Arch Toxicol. 1994;68(2):119- 28.

101. Uezumi, A., Ito, T., Morikawa, D., Shimizu, N., Yoneda, T., Segawa, M., Yamaguchi, M., Ogawa, R., Matev, M.M., Miyagoe-Suzuki, Y., Takeda, S., Tsujikawa, K., Tsuchida, K., Yamamoto, H., Fukada, S. (2011). Fibrosis and adipogenesis originate from a common mesenchymal progenitor in skeletal muscle. J. Cell Sci. 124(Pt 21):3654-64.

102. United States Government Accountability Office (GAO). 2009. Chemical Regulation: Options for enhancing the effectiveness of the toxic substances control act. http://www.gao.gov/new.items/d09428t.pdf

103. van Dartel, D.A., and Piersma, A.H. (2011). The embryonic stem cell test combined with toxicogenomics as an alternative testing model for the assessment of developmental toxicity. Reprod. Toxicol. 32(2):235-44.

104. van der Burg, B., Pieterse, B., Buist, H., Lewin, G., van der Linden, S.C., Man, H.Y., Rorije, E., Piersma, A.H., Mangelsdorf, I., Wolterbeek, A.P., Kroese, E.D., van Vugt-Lussenburg, B. (2015). A high throughput screening system for predicting chemically-induced reproductive organ deformities. Reprod. Toxicol. 55:95-103.

105. Walker, L., Baumgartner, L., Keller, K.C., Ast, J., Trettner, S., zur Nieden, N.I. (2015). Non-human primate and rodent embryonic stem cells are differentially sensitive to embryotoxic compounds. Toxicol. Rep. 2:165-174.

106. Wobus, A.M., Wallukat, G., Hescheler, J. (1991). Pluripotent mouse embryonic stem cells are able to differentiate into cardiomyocytes expressing chronotropic responses to adrenergic and cholinergic agents and Ca2+ channel blockers. Differentiation. 48(3):173-82.

107. Xian, C.J., Cool, J.C., Scherer, M.A., Fan, C., Foster, B.K. (2008). Folinic acid attenuates methotrexate chemotherapy-induced damages on bone growth mechanisms and pools of bone marrow stromal cells. J. Cell Physiol. 214(3):777-85.

116

108. Xian, C.J., Cool, J.C., Scherer, M.A., Macsai, C.E., Fan, C., Covino, M., Foster, B.K. (2007). Cellular mechanisms for methotrexate chemotherapy-induced bone growth defects. Bone. 41(5):842-50.

109. Xu, J., Murphy, S.L., Kochanek, K., Bastian, B.A. (2016). Deaths: Final data for 2013. Centers for Disease Control and Prevention National Vital Statistics Reports. 64(2):1-118. http://www.cdc.gov/nchs/data/nvsr/nvsr64/nvsr64_02.pdf

110. Yen, H.Y., Ting, M.C., Maxson, R.E. (2010). Jagged1 functions downstream of Twist1 in the specification of the coronal suture and the formation of a boundary between osteogenic and non-osteogenic cells. Dev. Biol. 347(2):258-270.

111. Yousfi, M., Lasmoles, F., Marie, P.J. (2002). TWIST inactivation reduces CBFA1/RUNX2 expression and DNA binding to the osteocalcin promoter in osteoblasts. Biochem. Biophys. Res. Commun. 297(3):641-4.

112. Zimmer B, Lee G, Balmer NV, Meganathan K, Sachinidis A, Studer L, Leist M. Evaluation of developmental toxicants and signaling pathways in a functional test based on the migration of human neural crest cells. Environ Health Perspect. 2012 Aug;120(8):1116-22.

113. zur Nieden, N.I., and Baumgartner, L. (2010a). Assessing developmental osteotoxicity of chlorides in the embryonic stem cell test. Toxicol. Appl. Pharmacol. 30(2):277-83.

114. zur Nieden, N.I., Davis, L.A., Rancourt, D.E. (2010b). Comparing three novel endpoints for developmental osteotoxicity in the embryonic stem cell test. Toxicol. Appl. Pharmacol. 247(2):91-97.

115. zur Nieden, N.I., Davis, L.A., Rancout, D.E. (2010c). Monolayer cultivation of osteoprogenitors shortens duration of the embryonic stem cell test while reliably predicting developmental osteotoxicity. Toxicology. 277:66-73.

116. zur Nieden, N.I., Kempka, G., Ahr, H.J. (2003). In vitro differentiation of embryonic stem cells into mineralized osteoblasts. Differentiation. 71(1):18-27.

117

117. zur Nieden, N.I., Kempka, G., Ahr, H.J. (2004). Molecular multiple endpoint embryonic stem cell test—a possible approach to test for the teratogenic potential of compounds. Toxicol. Appl. Pharmacol. 194(3):257-69.

118. zur Nieden, N.I., Price, F.D., Davis, L.A., Everitt, R.E., Rancourt, D.E. (2007). Gene profiling on mixed embryonic stem cell populations reveals a biphasic role for beta-catenin in osteogenic differentiation. Mol. Endocrinol. 21(3):674-85.

118

CHAPTER 5

Tobacco-induced oxidative stress disrupts osteogenic differentiation in human embryonic

stem cells.

Nicole RL Sparks and Nicole I zur Nieden

Introduction

Every 4.5 minutes, a baby is born with a birth defect in the United States. According to the Center for Disease Control, birth defects are the major cause of infant deaths, accounting for 1 in 5 infant deaths. Of the babies born with an abnormality, approximately one-third have a defect of the head and face (Trainor, 2010) including cleft lip/palate, improperly fused skull, small or absent skull and facial bones, and malformed nose, eyes, ears, and teeth (Carmichael et al., 2008; MOD, 2010). Both functionally and socially such birth defects can have a significant impact on the patient's quality of life. For example, the

CDC estimates the lifetime therapy costs for children born each year with cleft lip and/or palate to be US$697 million (Trainor, 2010). Causes of birth defects can be attributed to the environment, pharmaceuticals, genetics, viruses, nutrition, and diet (NCBDDD, CDC).

Approximately 7% of all congenital defects are caused by exposure to known teratogens

(Chung, 2004), which is defined as any substance (i.e., chemical, infectious agent, or physical condition) that has the ability to cause abnormal development to the embryo or fetus while in the mother’s womb (Polifka and Friedman, 2002). It is specifically during

119 periods of rapid cell division and differentiation that the embryo is most susceptible to teratogenic agents (Chung, 2004). Despite the great strides made in developmental biology in the last century, the adverse mechanisms associated with embryonic exposure to teratogens is poorly understood and, as such, is of great interest.

Tobacco products are one of the most common environmental toxicants/teratogens

and tobacco exposure is the single most preventable cause of disease, disability, and death

in the United States (CDC). Through anti-smoking initiatives, the prevalence of smoking-

related diseases has declined since 1965, from 43% to 18% today (Hall and Doran, 2016).

However, present tobacco use is attributed to nearly 500,000 deaths per year in the United

states, including an estimated 42,000 deaths from secondhand smoke exposure, and 8.6

million people live with a serious illness caused by smoking (CDC; Hall and Doran, 2016).

The costs associated to smoking-related illnesses in the United States is estimated to be

more than $300 billion per year (Xu et al., 2014; U.S. Department of Health and Human

Services, 2014a). Adverse effects of tobacco smoke have been demonstrated by

epidemiological and in vivo studies that associate smoke with diseases affecting the heart,

lungs, and various cancers (WHO, 2011). In addition, disorders of the adult skeleton have

been linked to cigarette smoke. Tobacco use has been associated with increased risk of

developing osteoporosis, increased risk of fractures, delayed rate of bone healing, 4-5%

decreased bone mineralization, and obstruction to bone surgery results (Akhter et al., 2005;

Rothem et al., 2009; AAOS, 2016).

Maternal smoking during pregnancy exposes the unborn fetus to tobacco constituents. Of the 7000 constituents of cigarette smoke, many chemicals like nicotine and

120 carcinogenic compounds can cross the placenta (Wickström, 2007; Bruin et al., 2010). Of all the children today, it is estimated that almost 1 out of 13 will die prematurely due to smoking (AAOS, 2016). Furthermore, in infants and children, secondhand smoke is the number one cause of asthma attacks, respiratory and ear infections, and sudden infant death syndrome (U.S. Department of Health and Human Services, 2006; 2014b). Despite this knowledge, 10-20% of United States women smoke during pregnancy with the highest smoking rates attributed to women of lesser education and low socioeconomic background

(Tong et al., 2009; Curtin and Matthews, 2016). Exposure to tobacco smoke increases the mother’s risk of spontaneous abortion, premature delivery, infertility, and fetal abnormalities such as long-term effects on brain function and cognition and improper development of the skeleton (Shiverick and Salafia, 1999). Furthermore, maternal smoking has shown to increase the chance of cleft lip and/or palate and premature closure of coronal sutures (craniosynostosis) (Carmichael et al., 2008; MOD, 2010). While studies have demonstrated the detrimental effects of maternal smoking to the developing skeleton, there is a lack of investigation to understand the pathogenic mechanisms of tobacco-induced skeletal teratogenicity in the developing embryo.

The known adverse health outcomes associated with tobacco use has led to the public misconception that alternative forms of tobacco like ultra-filtered and low nicotine/tar cigarettes or non-combustible chewing tobacco are safer than traditional cigarettes (Byron et al., 2017). Pregnant women who experience difficulty with smoking cessation may use these perceived safer alternatives, termed harm-reduction tobacco products (HRTPs). Maternal use of low nicotine and tar cigarettes, like Camel Blue, is a

121 risk factor for low birth weight (Mitchell et al., 2002). In comparison to cigarette smoking, it has been suggested that Snus can be a less harmful option due to not forming the combustible constituents of smoke (Ashford, 2016). However, there is growing evidence out of Sweden that maternal Snus use may be associated to oral cleft birth defects

(Gunnerbeck et al., 2014). While tobacco use during pregnancy has been linked to cognitive detriments, low birth weight, and skeletal defects, there is a lack of studies focusing on how tobacco disrupts bone development. With the increased use of HRTPs, there is a gap in knowledge of the molecular mechanisms of HRTP-induced skeletal teratogenicity.

Prior data

I have previously established a human ESC in vitro model of osteogenesis (chapter

2) and adapted the protocol to be used in a human embryonic stem cell test to predict the osteotoxicity caused by chemical compounds (chapter 4). Our lab has used the differentiation protocol and human EST to test popular tobacco brands for their ability to inhibit osteogenesis in differentiating hESCs. Since tobacco smoke exposure can occur through mainstream (MS) smoke, which is actively inhaled by the smoker, and sidestream

(SS) smoke, which burns off the tip of the cigarettes, both forms of smoke were tested. In addition, exposure can also happen orally using non-combustible products such as chewing tobacco, also called Snus. Among the tested tobacco products were Camel (conventional)

MS and SS smoke, Camel Blue (previously known as Camel Lights) MS and SS smoke, and Camel Suns tobacco extract (STE). Conventional Camel and Camel Blue MS smoke

122 showed no cytotoxicity nor detrimental effects to osteogenesis (Fig 5.1A). In contrast, the

SS smoke of conventional Camel and Camel Blue and Camel STE treated cultures, dose- dependently inhibited osteogenesis. Of importance, the differentiating hESCs exposed to

Camel Blue SS and STE showed a differentiation defect at concentrations that did not elicit cytotoxicity, whereas the conventional Camel SS caused inhibition due to cytotoxicity

(orange arrows).

In zebrafish, developmental exposure to Camel, Camel Blue, and STE induces cartilaginous defects in the developing head and spine as well as a mineralization delay

(Fig. 5.1C). In mice that were exposed to Camel Blue and STE in utero, multiple skeletal defects can be found at E17.5 that resemble metabolic bone disorders and are characterized by low bone mineral density (Fig 5.1D). In sum, using in vitro human pluripotent stem cells (hPSCs) and in vivo mouse embryo models, we provide here the first evidence that

HRTPs are toxic to the developing skeleton. This chapter will investigate the molecular mechanisms behind the observed differentiation defect. The data revealed that effective doses of Camel, Camel Blue, and STE generated excessive levels of the reactive oxygen species (ROS), superoxide anion and hydrogen peroxide, during early stages of osteogenesis along with elevated NADPH oxidase activity. However, HRTPs generated

ROS at intermediate levels compared to conventional Camel SS, suggesting a mechanism for the difference between cytotoxicity caused by the conventional tobacco and the subtoxic differentiation inhibition noted in HRTP-treated cells. Together these data support the notion that HRTPs initiate oxidative stress and disrupt embryonic skeletal development as will be outlined below.

123

Figure 5.1. The embryotoxicity associated with tobacco exposure. A) Concentration-response curves of Camel, Camel Blue, and Camel Snus Tobacco Extract (STE). Concentration-responses were determined by MTT assay of H9 hESCs and HFF cells and calcium assay of H9 hESCs; n=5, *p<0.05 One-Way ANOVA and Tukey HSD exposed versus untreated control. Data courtesy of Ivann Martinez, zur Nieden lab. B) Compiled list of IC50 and ID50 values of all screened tobacco products. Non-effective and effective doses were determined by the curves. C) Zebrafish exposure to Camel or STE negatively regulates pharyngeal arch development. On day 5 post fertilization, a COL2- reporter, which is expressed early in development, distinguished variable craniofacial defects between exposure to the two tobacco products. White arrows point to shortened development of the head and absent arches. Exposure also caused curvature of the spine (black arrows). Data courtesy of Dr. Maria Bondesson-Boulin and Omran Karmach, zur Nieden lab. D) Mouse pups exposed in utero to Camel Blue or STE exhibited porous skulls (white arrow heads). hESCs, human embryonic stem cells; HFF, human foreskin fibroblasts; IC50, half-maximal inhibitory concentration of cytotoxicity; ID50, half-maximal differentiation inhibiting concentration; MTT, 3-[4,5- dimethylthiazol-2-yl]-2,5-diphenylterazolium bromide. Col2:mcherry labels chondrocytes expressing collagen 2.

124

Materials and Methods hESC culture

Human ESCs of the H9 line were maintained on Matrigel (BD Biosciences) coated plates in serum-free mTeSR® medium (Stem Cell Technologies). Cultures were passaged every 4 days with accutase.

Osteogenic differentiation

Human ESC confluent colonies were given control differentiation medium consisting of Dulbecco’s modified Eagle's medium (DMEM; Gibco, US) containing 15%

FBS (Atlanta Biologicals, US, selected batches), 1% non-essential amino acids (NEAA;

Gibco, US), 1:200 penicillin/streptomycin (Gibco, US), and 0.1 mM β-mercaptoethanol

(Sigma, US) for 5 days as described (Sparks et al., 2018; Chapter 2). Subsequently, control differentiation medium was supplemented with osteogenic factors: 0.1 mM β- glycerophosphate (βGP; Sigma, US), 50 µg/ml ascorbic acid (AA; Sigma, China), and

-7 1.2x10 M 1,25(OH)2 Vitamin D3 (VD3; Cayman Chemical, US).

Production of tobacco solutions

Commercially available conventional (Camel) and harm reduction (Camel Blue) brands of cigarettes were purchased from a local retail dealer and used to make mainstream

(MS) and sidestream (SS) smoke solutions with a method described previously in detail

(Knoll and Talbot, 1998). Smoke solutions were generated using a University of Kentucky smoking machine that took a 2.2 second puff of MS every minute. MS smoke solution was

125 generated by pulling 30 puffs of MS smoke through 10 ml of DMEM culture medium.

During MS smoke production, SS smoke solution was produced by collecting the smoke that burned off the end of the cigarette and pulling it through 10 ml of DMEM. SS smoke was collected continuously, while MS smoke was collected during each puff. Both MS and

SS solutions were made at concentrations of 3 puff equivalents (PE). One PE of MS smoke is the amount of smoke in one puff that dissolves in 1 ml of culture medium. For SS smoke, one PE is the amount of SS smoke that dissolves in 1 ml of culture medium during one minute of burning. Immediately after preparation, smoke solutions were filtered through a

0.2 µm Acrodisc® PSF Syringe Filter (Pall Corporation, US), aliquoted into sterile

Eppendorf tubes, and stored in a -80°C freezer until used. Desired PEs were acquired through serial dilutions, and experiments were done using either MS or SS at concentrations as indicated and an untreated control.

Tobacco extract was created at 10% (w/v) in DMEM. Resulting solution was centrifuged for 10 min at 450 × g, to remove large debris. Supernatant was collected and centrifuged at 13,000 × g for 1 h, to remove finer debris. pH was adjusted to 7.4 and extract sterilized through a 0.45µm filter. FBS was added to a final concentration of 15%.

Antioxidant treatment

The following were added at the start of osteogenic differentiation and tobacco treatment (previously described) at a final concentration of: ascorbic acid (AA; Sigma-

Aldrich, USA) 100 µM, 10 µM dl-α-tocopherol acetate (vitamin E; Supelco, Sigma-

126

Aldrich, USA), and glutathione reduced ethyl ester (GSH-EE; Sigma-Aldrich, USA) 500

μM. The antioxidant medium was replaced with each media change.

RNA isolation and quantitative PCR

Total RNA was extracted with the NucleoSpin RNA II protocol (Macherey Nagel).

RNA concentration was determined with a NanoDrop® 1000 spectrophotometer (Thermo

Reactions were setup for 10 minutes of denaturing at 94°C, followed by 40 cycles of denaturing at 94°C, and annealing/elongation at 60°C for 45 seconds. The n‐fold expression in target samples was calculated with the ΔΔCT method by standardizing CT values to GAPDH expression. Primer sequences used in this study can be found in Table

5.1.

·- Superoxide anion (O2 ) detection

Generation of superoxide anion was determined using a Lumimax Superoxide

Anion Detection Kit (Agilent Technologies, USA). H9 cultures were trypsinized, washed with PBS, and resuspended in 1mL of fresh medium to incubate for 30 minutes at 37°C. A total of 5 × 105 of the cells were resuspended in 190 μL of superoxide anion (SOA) assay medium and 5 μL of 4.0 mM luminol solution plus 5 μL of 5.0 mM enhancer medium were

127 added to the SOA medium and incubated at room temperature for 30 min. The chemiluminescent light emissions of superoxide anion were measured with a luminometer

(Lucetta™).

·- NADPH Oxidase O2 detection

NADPH oxidases (NOX) are superoxide anion generating enzymes. Superoxide

·- anion, O2 , was detected by the chemiluminescent probe lucigenin, where lucigenin is

·- reduced by O2 to a lucigenin cation radical (Wang et al., 2011; Shah et al., 2013). Briefly, cultured cells were washed with 1X PBS and lysed in 300 µL of lysis buffer (20 mM

KH2PO4, pH 7.0, 1 mM EGTA, protease inhibitors). 100 µL of lysed cells were added to

900 µL of assay buffer consisting of 50 mM phosphate buffer, pH 7.0, containing 1 mM

EGTA, 150 mM sucrose, 5 µM lucigenin (Sigma M8010), and 100 µM NADPH (Sigma

N5130). Chemiluminescent light emission in terms of relative light units was measured in a luminometer every 3 min for 30 min. There was no measurable activity in the absence of

NADPH. Protein content was measured with the Lowry method using the Bio-Rad DC protein assay reagent. All measurements were normalized to its respective protein amount.

Superoxide anion production was expressed as relative chemiluminescence (light) units

(RLU)/mg protein.

·- MitoSOX™ mitochondrial (O2 ) detection

Mitochondrial superoxide formation was detected by incubating cells in the dark with 2.5 μM MitoSOX™ Red dye (excitation/emission at λ = 510 nm/580 nm;

128

ThermoFisher M36008) in PBS for 10 min. The average fluorescence intensity per cell for each experimental group of cells was calculated using NIH ImageJ analysis software as described by Fitzpatrick (2014) and Jensen (2013).

Detection of hydrogen peroxide

Dihydrorhodamine 123 (Invitrogen D23806) is an uncharged and nonfluorescent

H2O2 indicator that can passively diffuse across membranes where it is oxidized to cationic rhodamine 123 which localizes in the mitochondria. Cells were incubated in 1µM of DHR-

123 in PBS for 30 min at 37°C. Cells were washed twice with PBS and harvested in 500µL cold PBS for flow analysis in Beckman Coulter Flow cytometer detected at excitation/emission at λ = 485/535 nm. Gating was set using unstained control samples by forward scatter and side scatter light. For each sample, 10,000 events were collected.

Increased green fluorescence intensity indicates increased ROS H2O2.

Mitochondrial Membrane Potential (ΔΨm)

JC‐1 (5,5′,6,6′‐tetrachloro‐1,1′,3,3′‐tetraethylbenzimidazoyl carbocyanine iodide;

Invitrogen, T3168) is a lipophilic cationic dye that enters the inner mitochondrial matrix in its monomeric form when the mitochondrial membrane is polarized. When the mitochondrion has a high ΔΨm, the dye crosses the membrane and forms J‐aggregates, which appear red under UV light. If ΔΨm is low, the dye remains in its monomeric form and fluoresces green. The membrane‐sensitive dye was prepared at 1 µg/mL working solution in PBS from stock solution (5mg/mL in DMSO) prior to use. Cells were harvested

129 in microcentrifuge tubes, centrifuged, and incubated with 500µL of working solution for

15 min in a humidified incubator at 37°C, 5% CO2. Following staining, the cells were washed twice in PBS and resuspended in cold PBS for flow analysis in Beckman Coulter

Flow cytometer. To exclude debris, samples were gated to unstained samples from the same differentiation day based on forward scatter and side scatter light. Using the

“medium” setting flow rate, 10,000 events were collected. Mitochondrial potential was detected at excitation/emission at λ = 590/610 nm for JC-1 aggregates and excitation/emission at λ = 485/535 nm for monomers.

Superoxide dismutase (SOD)

Cultures were washed with PBS, harvested in 300 µL of RIPA buffer, and stored at -20 °C. Once thawed, the SOD assay kit was utilized per manufacturer’s instructions

(Cayman Chemical 706002). 10 µL of samples were combined with 200 µL of radical detector. The addition of 20 µL xanthine oxidase initiated the reaction, followed by a 20- minute room temperature incubation. The absorbance was measured at 450 nm.

Absorbance values for each sample were compared to a 7-point SOD standard curve

(provided by the kit) to determine SOD activity. The SOD activity of each sample was normalized to its total protein content determined by the Lowry method.

Catalase (CAT)

Cultures were washed with PBS, harvested in 300µL of RIPA buffer, and stored at

-20 °C. Once thawed, the catalase assay kit was utilized per manufacturer’s instructions

130

(Cayman Chemical 707002). To sample wells, 30µL of methanol was added, followed by

20µL of hydrogen peroxide to initiate the reaction. After 20 minutes of incubation, 30µL of potassium hydroxide terminated the reaction. Next, purpald, a chromogen, was added to measure the colorimetric change due to formaldehyde formation. Subsequently, each reaction underwent a 5-minute incubation with 10µL catalase potassium periodate and the absorbance was read at 540nm. Catalase activity was determined from a 7-point catalase formaldehyde standard curve (provided by the kit). Catalase activity was standardized to its total protein content.

Caspase assay

Guava Caspase-3/7-FAM Kit (Millipore 4500-0540) was used to measure caspase-

3/7 activity. Cell viability was simultaneously evaluated using 7-AAD. Cells were washed with 1X PBS and incubated with 10 µl of caspase reagent for 1 h at 37 °C in a 5% CO2 incubator. At the end of incubation, cells were washed with provided wash buffer, and incubated in 1X apoptosis buffer with 7-AAD for 10 min at room temperature. After incubation, cells were washed and changed to 1X PBS and analyzed with a Nikon fluorescent microscope.

qRT-PCR Apoptosis array

The RT2 Profiler™ human Apoptosis qPCR Array (PAHS-012ZA, Qiagen) was used to measure the expression levels of 84 apoptosis-related genes (Table 5.1). Total RNA was isolated and quantified as previously described. A total of 0.5 µg of RNA was used for

131 cDNA synthesis and mixed with the kit supplied RT2 SYBR Green Fluor qPCR Master

Mix and distributed onto a 96-well array with a final concentration of 36.7 ng of cDNA per reaction. The qPCR was performed on a Bio-Rad iQ5 cycler according to the RT2 Profiler

PCR Array instructions under the following conditions: 95°C for 10 min, then 40 cycles at

95°C for 15 sec and 60°C for 1 min. Five separate housekeeping genes (ACTB, B2M,

GAPDH, HPRT1, RPLP0) within each array were used for normalization by calculating the ΔΔCT for each gene of interest. Fold changes of gene expression and were analyzed by using RT2 PCR array data analysis web portal (Qiagen, https://www.qiagen.com/us/shop/genes-and-pathways/data-analysis-center-overview- page/) and a heatmap generated with heatmapper online software (Babicki et al., 2016).

Induction of Apoptosis Death Domain Receptors: CRADD, FADD, TNF, TNFRSF10B (DR5). DNA Damage & Repair: ABL1, CIDEA, CIDEB, TP53 (p53), TP73. Extracellular Apoptotic Signals: CFLAR (Casper), DAPK1, TNFRSF25 (DR3). Other Pro-Apoptotic Genes: BAD, BAK1, BAX, BCL10, BCL2L11, BID, BIK, BNIP3, BNIP3L, CASP1, CASP10 (MCH4), CASP14, CASP2, CASP3, CASP4, CASP6, CASP8 (FLICE), CD27 (TNFRSF7), CD70(TNFSF7), CYCS, DFFA, DIABLO (SMAC), FAS (TNFRS F6), FASLG (TNFSF6), GADD45A, HRK, LTA(TNFB), NOD1 (CARD4), PYCARD (TMS1, ASC), TNFRSF10A (TRAIL-R), TNFRSF9, TNFSF10 (TRAIL), TNFSF8, TP53BP2, TRADD, TRAF3.

Anti-Apoptotic AKT1, BAG1, BAG3, BAX, BCL2, BCL2A1 (BFL1), BCL2L1 (BCLXL), BCL2L10, BCL2L2, BFAR, BIRC3 (c- IAP2), BIRC5, BIRC6, BNIP2, BNIP3, BNIP3L, BRAF, CD27 (TNFRSF7), CD40LG (TNFSF5), CFLAR(Casper), DAPK1, FAS (TNFR SF6), HRK, IGF1R, IL10, MCL1, NAIP (BIRC1), NFKB1, NOL3, RIPK2, TNF, XIAP (BIRC4).

Regulation of Apoptosis Negative Regulation of Apoptosis: BAG1, BAG3, BCL10, BCL2, BCL2A1 (BFL1), BCL2L1 (BCLXL), BCL2L10, BCL2L2, BFAR, BIRC2 (c-IAP1), BIRC3 (c-IAP2), BIRC6, BNIP2, BNIP3, BNIP3L, BRAF, CASP3, CD27 (TNFRSF7), CD40LG (TNFSF5), CFLAR (Casper), CIDEA, DAPK1, DFFA, FAS (TNFRSF6), IGF1R, MCL1, NAIP (BIRC1), NOL3, TP53 (p53), TP73, XIAP (BIRC4). Positive Regulation of Apoptosis: ABL1, AKT1, BAD, BAK1 BAX, BCL2L11, BID, BIK, BNIP3, BNIP3L, CASP1 (ICE), CASP10 (MCH4), CASP14, CASP2, CASP4, CASP6, CASP8 (FLICE), CD40 (TNFRSF5), CD70 (TNFSF7), CIDEB, CRADD, FADD, FASLG (TNFSF6), HRK, LTA (TNFB), LTBR, NOD1 (CARD4), PYCARD (TMS1, ASC), RIPK2, TNF, TNFRSF10A (TRAILR), TNFRSF10B (DR5), TNFRSF25 (DR3), TNFRSF9, TNFSF10 (TRAIL), TNFSF8, TP53 (p53), TP53BP2, TRADD, TRAF2, TRAF3.

Death Domain Receptors CRADD, DAPK1, FADD, TNFRSF10A (TRAILR), TNFRSF10B (DR5), TNFRSF11B (OPG), TNFRSF1A(TNFR1), TNFRSF1B, TNFRSF21, TNFRSF25 (DR3), TRADD.

Caspases & Regulators Caspases: CASP1 (ICE), CASP10 (MCH4), CASP14, CASP2, CASP3, CASP4, CASP5, CASP6, CASP7, CASP8 (FLICE), CASP9, CFL AR (Casper), CRADD, PYCARD (TMS1, ASC). Caspase Activation: AIFM1 (PDCD8), APAF1, BAX, BCL2L10, CASP1 (ICE), CASP9, NOD1 (CARD4), PYCARD (TMS1, ASC), TNFRSF10A (TRAIL-R), TNFRSF10B (DR5), TP53 (p53). Caspase Inhibition: CD27 (TNFRSF7), XIAP (BIRC4).

Table 5.1. Apoptosis-related genes examined with the RT2 ProfilerTM qPCR array.

132

Calcium assay

For quantification of extracellular matrix calcium, cells were harvested in radioimmunoprecipitation buffer (RIPA; 1XPBS, 1% NP-40, 0.5% sodium deoxycholate,

0.1% SDS, pH 7.4, and 1:100 protease inhibitors) with extensive scraping using the pipet tip. Calcium deposition was determined based on calcium ions (Ca2+) reacting with

Arsenazo III (Genzyme, Canada) to form a purple Ca-Arsenazo III complex, which was measured at 655 nm. The concentration of total calcium in the sample was calculated based on a CaCl2 standard and calcium content was normalized to the total protein content of the sample using the Lowry method (Bio-Rad) as we have also done previously when screening with mouse cells (zur Nieden et al., 2004; zur Nieden et al., 2010a, b). This normalization is important to truly assess the differentiation level of the cultures without interference from cytotoxic or proliferative events. Calcium content was then standardized to solvent only controls and concentration-response curves charted. The half-maximal inhibitory dose (ID50Ca) for each compound was taken from linear interpolation of the curve.

Statistical analysis

Data presented are averages of three independent experiments ± standard deviation.

Significance was determined by using a web-based one-way ANOVA and Tukey HSD post hoc test (http://faculty.vassar.edu/lowry/anova1u.html) or unpaired Student's t-test as appropriate.

133

Gene Forward Primer Reverse Primer Ta GAPDH 5'-GAGTCAACGGATTTGGTCGT-3' 5'-TTGATTTTGGAGGGATCTCG-3' 60 SOD2 5'-GGCCAAGGGAGATGTTACAA-3' 5'-GGCTGTCAGCTTCTCCTTAAAC-3' 60 Catalase 5'-CCCCAACTATTACCCCAACAG-3' 5'-ATGTTCTCACACAGGCGTTTC-3' 60 OCN 5'-GGCAGCGAGGTAGTGAAGAG-3' 5'-CTGGAGAGGAGCAGAACTGG-3' 60 TWIST1 5'-ACTGGCCTGCAAAACCATAG-3' 5'-TGCATTTTACCATGGGTCCT-3' 60

Table 5.2. Primer sequences used in this study. Ta, annealing temperature

Results

Sidestream smoke reduces OCN expression in osteogenically differentiating hESCs

Prior evidence suggested that Camel and Camel Blue SS smoke and Camel STE

inhibit in vitro differentiation of osteoblasts (Fig 5.1A). Here, we examined the expression

levels of OCN, a gene that codes for the non-collagenous extracellular matrix (ECM)

protein OCN, which is exclusively synthesized by osteoblasts and used as a marker for

bone mineralization (Ducy and Karsenty, 1995) in osteoblasts differentiated from hESCs.

The dose-response curve showed that Camel, Camel Blue, and STE gradually suppressed

OCN expression with increasing concentrations (Fig 5.2A) supporting our earlier findings.

The inhibitory effect on osteogenesis manifested early at day 7 of the differentiation

as osteoprogenitor genes were down-regulated at this early time point. For instance, the

master regulatory transcription factors TWIST1 and RUNX2 were differentially expressed

(Fig 5.2B). Specifically, effective doses of Camel Blue and STE resulted in the 1.79- and

1.62-fold upregulation of TWIST1 compared to the untreated control, respectively. RUNX2

was downregulated in Camel Blue and STE, to 0.58- and 0.68-fold respectively. We also

134 analyzed the expression of SATB2, a nuclear protein involved in osteoblast development.

Developing mouse embryos show expression of SATB2 in the first branchial arch (Zhang et al., 2011; Dobreva et al., 2006) and SATB2 null mice have craniofacial defects (Dobreva et al., 2006). When assessed at E15.5, Satb2-/- tibiae showed reduced osteoblast mineralization. Camel Blue showed no regulation over SATB2; however, STE treated cultures had a 1.99-fold upregulation (Fig 5.2B).

Figure 5.2. Tobacco dysregulates the expression of osteogenic genes. (A) Concentration-response curves for differentiation (ID50) of OCN quantitative PCR. Dose-dependently, OCN mRNA expression decreased. (B-D) Osteogenic promoting genes showed HRTP-misregulation. (B) TWIST1 was significantly upregulated in ED cell. (C) RUNX2 mRNA was reduced and (D) SATB2 showed a 1.99-fold increase in STE effective dose. UNT, untreated; NED, non-effective dose; ED, effective dose; STE, Camel Snus tobacco extract; OCN, osteocalcin; TWIST1, Twist Family BHLH Transcription Factor 1; RUNX2, Runt-related transcription factor 2; SATB2, Special AT-rich sequence-binding protein 2. n = 3 independent samples ± SD. *p<0.05 one‐way ANOVA versus untreated control.

135

Sidestream Camel smoke exposure causes apoptosis, while Sidestream Camel Blue and

STE do not

Because the MTT assay performed as part of the tobacco hEST, suggested Camel

to be cytotoxic (compare Fig. 5.1A), Caspases 3 and 7 activity levels were measured in an

effective dose (ED) determined from the concentration response curves. Effective doses

were defined as the doses at which the ECM calcium was reduced to 50%. A respective

non-effective dose (NED, at 100-fold below the effective) was run alongside the effective

dose in addition to an untreated control. On day 7 of differentiation, caspase 3/7 activation

was observed only in the Camel ED showing more apoptotic cells compared to the

untreated control, respective NED and all the HRTP-treated cultures (Fig 5.3).

We further aimed to

understand the cytotoxicity of Camel

SS compared to Camel Blue SS and

STE treated cultures on the

molecular level and hypothesized

that alterations in gene expression of

apoptotic mediators were causing

apoptosis to occur in Camel ED. On

day 7 of osteogenic differentiation, Figure 5.3. Camel sidestream smoke induces apoptosis. we employed a human apoptosis Conventional Camel showed increased apoptosis determined by Caspase 3/7 activation. Camel Blue and Camel STE did not show apoptotic cells in effective doses (ED). Co-administration of array, simultaneously analyzing 84 ascorbic acid (AA) reversed apoptotic activity. hESCs, human embryonic stem cells; STE, Snus tobacco extract; NED, Non- apoptosis-related genes. As shown effective dose; ED, effective dose; UT, untreated.

136 in the heatmap (Fig 5.4A), our qPCR array results showed significant differential regulation of apoptosis-related genes in the Camel ED. The majority of the genes were upregulated. Further, hierarchical clustering showed that the effective doses of Camel Blue and STE did not have commonly affected genes, neither had they a close association to the genes regulated in Camel ED (Fig 5.4A). These data indicate potentially different pathways of apoptosis regulation.

Camel ED mRNA analysis resulted in significant upregulation of the following of pro-apoptotic genes (Fig 5.4B-D). In particular, the following 12 genes were significantly upregulated: ABL1, GADD45, CIDEA, CIDEB, DFFA, AIF, HRK, AIFM, CYCS, CASP9,

BNIP3L, and BAX, which are pro-apoptotic genes associated with response to DNA damage and mitochondrial stress (Wang, 2000; Salvador et al., 2013; Fig 5.4B; Table 5.2).

In contrast, Camel Blue ED treated cells showed upregulation of only a subset of these genes (AIF, CIDEB, AIFM, CYTC, and CASP9). The most striking difference was the specific regulation of ABL1 in Camel ED. Camel STE upregulated 3 genes out of the 12:

CIDEA, AIF, and AIFM. In addition, both Camel Blue and STE upregulated CASP4 mRNA, while Camel SS did not (Fig 5.4E). While the activation of most caspases is associated with apoptosis, caspase 4 has been shown to respond specifically to endoplasmic reticulum (ER) stress (Hitomi et al., 2004), potentially indicating that osteogenic inhibition in HRTP-treated cultures could be partially due to ER stress.

To validate the finding that Camel SS exposure induced apoptosis by inducing mitochondrial stress, we measured the mitochondrial membrane potential, which is determined by the flux through the electron transport chain and generation of a proton

137

gradient across the membrane. This was done using JC-1, a cationic dye that enters and

accumulates in the mitochondrial matrix and forms j-aggregates. J-aggregates fluoresce red

and indicate an increase in membrane potential. Strong JC-1 aggregates, an indication of a

healthy cell, were observed in all non-effective doses, HRTP effective doses, and control

cultures (Fig 5.4C). In contrast, Camel effective dose-treated cells had diminished

aggregates and increased monomers, as JC-1 remained in its monomeric form as

determined by its green fluorescence. Suggesting that that the mitochondrial membrane

potential collapsed, and the cells were in a state of stress. This data indicated that exposure

to Camel SS smoke can result in mitochondrial dysfunction.

Inhibition of osteogenesis is linked to oxidative stress

Next, we investigated whether oxidative stress, a common link between toxicant exposure and cytotoxicity, was responsible for the observed adverse outcomes associated with tobacco exposure. To causally address this, caspase 3/7 activation was measured in cells treated with ascorbic acid (AA) concomitantly with exposure to an effective dose. As expected, if apoptosis were to occur via oxidative stress, the presence of AA diminished the caspase 3/7 signal in the Camel SS ED (Fig 5.3).

Co-administration of tobacco products and antioxidants (i.e., AA, vitamin E (VE),

and glutathione reduced ethyl ester (GSH-OEt) reestablished calcium deposition to control

levels (Fig 5.5A). These results suggest that excessive ROS may lead to cytotoxicity via

apoptosis due to the activation of caspase 3/7.

138

Figure 5.4. Camel tobacco products de-regulate apoptosis-related genes. A) Clustering of 84 apoptotic genes transcriptionally responsive between non-effective (NED) and effective (ED) doses of Camel, Camel Blue, and Camel Snus tobacco extract (STE) as measured with the RT2 ProfilerTM qPCR array. Camel clustered away from the other treatments with positive regulation of apoptosis. Each row represents the relative expression of a single gene. B) Graphs show the overview of significantly upregulated apoptotic genes categorized by apoptotic response. (C) Reduced mitochondrial membrane potential measured by JC-1 dye is a consequence of Camel toxicity. Mitochondrial permeability was not affected by HRTPs. n = 3 independent samples ± SD. *p < 0.05 one‐way ANOVA versus untreated control (UNT).

139

HRTPs generate subtoxic levels of ROS

Oxidative stress is the imbalance between reactive oxygen species (ROS) and antioxidants. Therefore, we next investigated which specific ROS species were responsible for observed oxidative stress. ROS are highly reactive molecules that consists of radical

·- − and non-radical oxygen species such as superoxide anion (O2 ), hydroxyl radical (OH ),

·- and hydrogen peroxide (H2O2). O2 is the primary ROS produced and quickly turned over

− to generate the more aggressive secondary ROS (H2O2 and OH ; Sharma et al., 2012; Circu

·- and Aw, 2010). Mainstream smoke solutions did not generate significant levels of O2 .

However, we confirmed that effective doses of Camel and Camel Blue SS smoke and STE

·- generated O2 in H9 cells differentiating into osteoblasts on day 7 (Fig 5.6A), which is

Figure 5.5. Tobacco exposure decreases antioxidant activity. Effective (ED) doses of Camel, Camel Blue, and Camel Snus tobacco extract (STE), determined by dose response curves (compare prior data), were compared against non- treated controls. Co-administration of antioxidants ascorbic acid (AA), vitamin E (VE), or glutathione reduced ethyl ester (GSH-OEt) with its respective ED of tobacco product (A) rescued osteoblast inhibition and (B) diminished O2·- production. hESCs, human embryonic stem cells. n = 5 technical replicates for calcium assay, n = 3 independent samples for O2·- determination ± SD. *p<0.05 one‐way ANOVA versus untreated control. UNT, untreated; O2·-, superoxide anion

140 when the osteoprogenitor genes were deregulated (compare Fig.

5.2B). The highest increase in superoxide anion content was registered in the Camel SS exposed cells, followed by the

Camel Blue SS exposed cells and then STE. Effective doses of differentiating cells treated with

·- antioxidants alleviated O2 Figure 5.6. Tobacco increases production of ROS in differentiating human embryonic stem cells. (A) Graph shows total superoxide anion levels (Fig 5.5B.) In addition, (O2·-) measured with Luminol, normalized to cell number and found to be highest in the effective doses. Camel and Camel Blue mainstream ·- DHR123 measured a significant (MS) smoke did not show a difference in O2 levels. (B) Measurement of H2O2 DHR fluorescence by flow cytometry. ED of tobacco products had significant increases in ROS levels compared to the untreated increase in H2O2 on day 7 in all control and NED cells. C) NOX activity was measured by lucigenin- based luminescence assay, effective doses generated increased levels treatments (Fig 5.6B). of ROS. D) Quantitative data expressing fluorescence of MitoSOX positive cells. Mitochondria were a major source of ROS production in Camel treated cells. n = 3 independent samples ± SD. *p<0.05 one‐way ROS are a natural ANOVA versus untreated control. MS, mainstream smoke; SS, sidestream smoke; H2O2, C. Blue, Camel Blue; Hydrogen peroxide; byproduct generated through the STE, Snus tobacco extract. activation of enzymes such as NADPH oxidase (NOX), a membrane enzyme, and mitochondrial oxidases (Circu and Aw, 2010). Hence, NOX activity was next assessed by the lucigenin chemiluminescence method. NOX activity was increased in effective doses over relevant non-effective doses on day 7 of differentiation with a 1.68, 1.8, and 2.04-fold increase by 3 min and 1.73, 2.32, and 1.71-fold increase at 30 min in the order of Camel

ED, Camel Blue ED and Camel STE, respectively (Fig 5.6C).

141

·- Another major source of O2 generation is in the mitochondria. The tobacco-

induced increase in mitochondrial superoxide anion was confirmed by MitoSOX for Camel

SS. However, mitochondrial ROS production was not affected by the HRTPs Camel Blue

and STE (Fig 5.6D), indicating mitochondrial dysfunction was occurring in the Camel

treated cultures only. These data suggest that the observed cytotoxicity of Camel SS may

·- be due to excessive levels of O2 formed by the mitochondria with a secondary role for

·- NOX. In contrast, NOX appear to be the major contributor of O2 in the HRTP-treated

cells. Therefore, the detrimental effect of the HRTPs on osteogenic differentiation can

·- potentially be attributed to subtoxic levels of O2 that appear to support cell survival, but

instead inhibit differentiation along the proper lineage.

Tobacco-induced decrease in antioxidant enzyme activity

When the equilibrium between ROS and antioxidant levels is disrupted, oxidative stress occurs. To assess the presence and/or lack of endogenous antioxidant contributions by antioxidant enzymes, enzyme activity levels of superoxide dismutase (SOD), the first

ROS detoxifying enzyme, and catalase (CAT) were measured. SOD is a free scavenging

·- enzyme that dismutates O2 into H2O2. In turn, H2O2 is degraded to water and molecular oxygen by the enzyme CAT. Interestingly, there was a significant reduction in SOD and catalase activity (Fig 5.7A), which was evident at the mRNA level for SOD2, but not CAT

(Fig 5.7B).

142

Figure 5.7. Tobacco exposure decreases antioxidant activity. (A) Superoxide dismutase (SOD) and catalase (CAT; free scavenging enzymes) activity levels were lowest in the effective doses. SOD dismutates superoxide anion to hydrogen peroxide and catalase breaks down hydrogen peroxide to water and oxygen. (B) C) SOD2 but not CAT had decreased mRNA compared against the non-treated control. n = 5 technical replicates ± SD for enzyme activity. n = 3 independent samples ± SD for mRNA quantification. *, p < 0.05 one‐way ANOVA) versus untreated control. NED, non-effective dose; ED, effective dose; Hydrogen peroxide; STE, Snus tobacco extract.

Discussion

Earlier work from our lab has shown that administration of tobacco products decreases osteoblast production during differentiation. However, the mechanisms responsible for osteogenic inhibition have not been fully elucidated. This present study shows that tobacco products increase ROS in H9 hESCs as they are differentiating into osteoblasts. Our findings further suggest that NOX and the mitochondrial electron transport chain are contributing sources of ROS production. Other enzymes, such as xanthine oxidase can also produce superoxide anion, but only a small amount of ROS are generated by the normal activity of xanthine oxidase. However, conditions like ischemia have been attributed to high ROS levels generated by xanthine oxidase (Ahmad et al., 2017). During oxidative stress, memebrane lipids are also susceptible to oxidation by ROS. The oxidation results in the formation of a lipid radical and initiating lipid peroxidation (Phaniendra et al., 2015). Lipid peroxidation disrupts memebrane integrity and plays a role in various

143 pathological conditions such as cardiovascular diseases and neurodegeneration (Nam,

2011). These other generators of ROS such as lipid peroxidation and xanthine oxidase need to be further explored to determine the main contributor(s) of ROS in the HRTP-mediated inhibition of skeletal development. Tobacco exposure attenuated the activity of SOD and

CAT. Importantly, osteogenic inhibition was reversed by supplementation of antioxidants during differentiation causally linking tobacco exposure, oxidative stress and inhibition of osteoblast development.

Oxidative stress can occur due to aging and exposure to radiation, drug therapies, and environmental toxicants (Valko et al., 2007; Tilg et al., 2008). Oxidative stress can lead to the damage of cell membranes, DNA, proteins, and enzymes (Birben et al., 2012), which can result in damage to organs. Bone diseases such as osteoporosis have been attributed to oxidative damage. For example, osteoporosis in postmenopausal women has been shown to associate with activation of NOX coupled with reduced function of antioxidant enzymes (Sendur et al., 2009). In skeletal disorders that cause low bone mineral density, studies indicate oxidative stress to have a role in the pathogenesis of the disease

(Tilg et al., 2008). It has also been suggested that oxidative stress supports osteoclast differentiation and activity through the upregulation of RANKL and induces osteoblast apoptosis (Tilg et al., 2008; Guo et al., 2017). Both osteoblast reduction and increased osteoclast activity would disrupt osteogenesis and perpetuate decreased bone mass making bones fragile and susceptible to fractures.

The combustion of a cigarette can produce excessive levels of ROS that would not

be present in the tobacco leaf or ash alone (Huang et al., 2005). Cigarette filters are not

144 equipped to remove the ROS. The gas and solid phase of SS smoke has high levels of free

·- radicals (Valavanidis et al., 2009). Further, water-soluble components can produce O2 and subsequently H2O2, showing that the involuntary exposure to cigarette smoke (passive smoking) is a health hazard to vulnerable populations such as the developing fetus, infants, and young children. Smokeless tobacco has also been demonstrated to induce oxidative stress and lead to extensive DNA damage and potentially cancers of the mouth (Bagchi et al., 2001; Katakwar et al., 2016). It is therefore plausible, that 1) chemicals other than combustion products only may be causes of health issues, and that 2) tobacco-induced oxidative stress can target the bone specifically to cause disorders. Yet, it remains to be uncovered if oxidative stress can contribute to tobacco-associated skeletal birth defects.

The degree of elevation of ROS can have different effects on the vitality of the cell.

At low levels, ROS generation could be beneficial and necessary for homeostasis. ROS produced by macrophages have been shown to suppress T cell-dependent arthritis

(Gelderman et al., 2007). Importantly, low levels of ROS can support hematopoietic stem cell (HSC) self-renewal and proliferation. With increased levels of ROS, cell cycle inhibitors such as p16INK4 are activated to stop HSC self-renewal without damage to the cell (Ito et al., 2006). In this study, the data suggest that excessive levels of ROS result in cytotoxicity as seen in Camel treated cultures. Conversely, Camel Blue and STE generate

ROS at lower levels that do not lead to cytotoxicity but results in a HRTP-mediated skeletal differentiation defect. This notion was further supported by the high apoptotic activity in

Camel treated cultures as uncovered with the apoptosis array, which showed upregulation of pro-apoptotic and DNA damage genes. In the HRTP cultures instead, only a small subset

145

of such genes was upregulated. In particular, the upregulation of CASP4 in the HRTP

exposed cells may point towards differentiation inhibition through an inflammation/ER-

stress response. Therefore, conventional and harm-reducing tobacco products might show

different modes of bone inhibition.

The hypothesis that oxidative stress is a major modulator of development is relatively new but has already been examined for a variety of chemicals. With recently increased interest in redox biology and disease in the past few years, teratogenic agents such as thalidomide, benzo[a]pyrene, and phenytoin have been linked to ROS-mediated teratogenesis. Specifically, the action of thalidomide, an agent known to cause limb defects in humans and rabbits, can be reversed with pre-treatment of the free-radical spin trapping agent phenylbutylnitrone (Lee et al., 2011). The attenuation of osteogenesis in this case can thus potentially be attributed to ROS at levels that did not elicit cytotoxicity and is reversible with the up-regulation of free-radical scavenging enzymes, as also shown here.

Other studies have reported a ROS-mediated reduction in adult osteogenesis

(Almeida et al., 2007; Ambrogini et al., 2010; Iyer et al., 2013). In contrast to our work, these previous studies look at the role of ROS on adult stem cells, such as mesenchymal stem cells (MSCs). Results obtained with adult stem cells can be indicative of age-related disorders, such as osteoporosis, however fail to address how the tissue originally develops.

In contrast, our study demonstrates skeletal defects during embryonic development in association with oxidative stress. Specifically, it is the use of human ESCs that allows for the assessment of differentiation defects that are directly relevant to the developing human embryo and fetus.

146

In summary, we have demonstrated that the HRTPs Camel Blue and STE are detrimental to the development of the embryonic skeleton. We suggest the underlying mechanism of HRTP osteoblast inhibition, in part, be due to the generation of sub-toxic

ROS levels. How this sub-toxic oxidative stress may potentially cause differentiation inhibition molecularly will be addressed in the following chapter.

147

References

1. AAOS. Tobacco Use and Orthopaedic Surgery. American Academy of Orthopaedic Surgeons. Information Statement 1047. February 2016 https://www.aaos.org/uploadedFiles/PreProduction/About/Opinion_Statements/advist mt/1047%20Tobacco%20Use%20and%20Orthopaedic%20Surgery%20(3).pdf [accessed April 7, 2018]

2. Ahmad G, Almasry M, Dhillon A, Abuyyash M, Kothandaraman N, Cakar Z. (2017) Overview and Sources of Reactive Oxygen Species (ROS) in the Reproductive System. In: Agarwal A. et al. (eds) Oxidative Stress in Human Reproduction. Springer, Cham.

3. Akhter MP, Lund AD, Gairola CG. 2005. Bone biomechanical property deterioration due to tobacco smoke exposure. Classified Tissue International. 77(5):319-326.

4. Almeida M, Han L, Martin-Millan M, O'Brien CA, Manolagas SC. Oxidative stress antagonizes Wnt signaling in osteoblast precursors by diverting beta-catenin from T cell factor- to forkhead box O-mediated transcription. J Biol Chem. 2007;282(37):27298-305.

5. Ambrogini E, Almeida M, Martin-Millan M, Paik JH, Depinho RA, Han L, Goellner J, Weinstein RS, Jilka RL, O'Brien CA, Manolagas SC. FoxO-mediated defense against oxidative stress in osteoblasts is indispensable for skeletal homeostasis in mice. Cell Metab. 2010;11(2):136-46.

6. Ashford K. What is Snus? WebMD 2016. https://www.webmd.com/smoking- cessation/features/Snus-tobacco-health-risks#1

7. Bagchi M, Kuszynski CA, Balmoori J, Joshi SS, Stohs SJ, Bagchi D. Protective effects of antioxidants against smokeless tobacco-induced oxidative stress and modulation of Bcl-2 and p53 genes in human oral keratinocytes. Free Radic Res. 2001;35(2):181-94.

8. Birben E, Sahiner UM, Sackesen C, Erzurum S, Kalayci O. Oxidative stress and antioxidant defense. World Allergy Organ J. 2012;5(1):9-19.

9. Bruin J, Gerstein H, Holloway A. Long-Term Consequences of Fetal and Neonatal Nicotine Exposure: A Critical Review. Toxicol Sci. 2010;116(2):364-74.

148

10. Byron MJ, Jeong M, Abrams DB, Brewer NT. Public misperception that very low nicotine cigarettes are less carcinogenic Tobacco Control Published Online First: 23 January 2018.

11. Carmichael SL, Ma C, Rasmussen SA, Honein MA, Lammer EJ, Shaw GM. Craniosynostosis and maternal smoking. Birth Defects Res A Clin Mol Teratol 2008;82(2):78-85.

12. CDC - Smoking & Tobacco Use. Centers for Disease Control and Prevention. http://www.cdc.gov/tobacco/. [accessed April 7, 2018]

13. Chung W. “Teratogens and their effects” in The New Public Health: An Introduction for the 21st Century. Chapter 23. New York: Columbia University Press 2004. http://www.columbia.edu/itc/hs/medical/humandev/2004/Chpt23-Teratogens.pdf

14. Circu ML, Aw TY. Reactive oxygen species, cellular redox systems, and apoptosis. Free Radic Biol Med. 2010;48(6):749-62.

15. Curtin SC, Matthews TJ. Smoking Prevalence and Cessation Before and During Pregnancy: Data From the Birth Certificate, 2014. Natl Vital Stat Rep. 201610;65(1):1- 14.

16. Dobreva G, Chahrour M, Dautzenberg M, Chirivella L, Kanzler B, Fariñas I, Karsenty G, Grosschedl R. SATB2 is a multifunctional determinant of craniofacial patterning and osteoblast differentiation. Cell. 2006;125(5):971-86.

17. Ducy P, Karsenty G. Two distinct osteoblast-specific cis-acting elements control expression of a mouse osteocalcin gene. Mol Cell Biol. 1995;15(4):1858-69.

18. Fitzpatrick M. Measuring cell fluorescence using ImageJ. Read the Docs. 2014. https://theolb.readthedocs.io/en/latest/imaging/measuring-cell-fluorescence-using- imagej.html [accessed November 18, 2015]

19. Gelderman KA, Hultqvist M, Pizzolla A, Zhao M, Nandakumar KS, Mattsson R, Holmdahl R. Macrophages suppress T cell responses and arthritis development in mice by producing reactive oxygen species. J Clin Invest. 2007;117(10):3020-8.

149

20. Gunnerbeck A, Edstedt Bonamy AK, Wikström AK, Granath F, Wickström R, Cnattingius S. Maternal snuff use and smoking and the risk of oral cleft malformations- -a population-based cohort study. PLoS One. 2014 Jan 15;9(1):e84715.

21. Guo S, Fei HD, Ji F, Chen FL, Xie Y, Wang SG. Activation of Nrf2 by MIND4-17 protects osteoblasts from hydrogen peroxide-induced oxidative stress. Oncotarget. 2017;8(62):105662-105672.

22. Hall W, Doran C. How Much Can the USA Reduce Health Care Costs by Reducing Smoking? PLoS Med. 2016; 13(5): e1002021.

23. Huang MF, Lin WL, Ma YC. A study of reactive oxygen species in mainstream of cigarette. Indoor Air. 2005; 15(2):135-40.

24. Iyer S, Ambrogini E, Bartell SM, Han L, Roberson PK, de Cabo R, Jilka RL, Weinstein RS, O'Brien CA, Manolagas SC, Almeida M. FOXOs attenuate bone formation by suppressing Wnt signaling. J Clin Invest. 2013;123(8):3409-19.

25. Jensen C. Quantitative Analysis of Histological Staining and Fluorescence Using ImageJ. Anat Rec (Hoboken). 2013;296(3):378-81.

26. Karatza AA, Varvarigou A, Beratis NG. 2003. Growth up to 2 years in relationship to maternal smoking during pregnancy. Clin Pediatr (Phila). 42:533–541.

27. Katakwar P, Metgud R, Naik S, Mittal R. Oxidative stress marker in oral cancer: A review. J Cancer Res Ther. 2016;12(2):438-46.

28. Kawakita A, Sato K, Makino H, Ikegami H, Takayama S, Toyama Y, Umezawa A. 2008. Nicotine acts on growth plate chondrocytes to delay skeletal growth through the alpha7 neuronal nicotinic ac-etylcholine receptor. PLoS One. 3(12):e3945.

29. Knoll M, Talbot P. Cigarette smoke inhibits oocyte cumulus complex pick-up by the oviduct in vitro independent of ciliary beat frequency. Reprod Toxicol 1998;12(1):57- 68.

150

30. Lee CJ, Gonçalves LL, Wells PG. Embryopathic effects of thalidomide and its hydrolysis products in rabbit embryo culture: evidence for a prostaglandin H synthase (PHS)-dependent, reactive oxygen species (ROS)-mediated mechanism. FASEB J. 2011;25(7):2468-83.

31. March of Dimes. 2010. March of dimes data book for policy makers: Maternal, infant, and child health in the United States. http://www.marchofdimes.org/advocacy/data- book-for-policy-makers-maternal-and-infant-health-in-the-united-states.aspx. [accessed April 7, 2018]

32. Mitchell EA, Thompson JM, Robinson E, Wild CJ, Becroft DM, Clark PM, Glavish N, Pattison NS, Pryor JE. Smoking, nicotine and tar and risk of small for gestational age babies. Acta Paediatr. 2002;91(3):323-8.

33. Nam TG. Lipid peroxidation and its toxicological implications. Toxicol Res. 2011;27(1):1-6.

34. NCBDDD (National Center on Birth Defects and Developmental Disabilities). Birth Defects: Data & Statistics. Centers for Disease Control and Prevention 2018. https://www.cdc.gov/ncbddd/birthdefects/data.html [accessed April 7, 2018]

35. Phaniendra A, Jestadi DB, Periyasamy L. Free radicals: properties, sources, targets, and their implication in various diseases. Indian J Clin Biochem. 2015;30(1):11-26.

36. Polifka JE and Friedman JM. 2002. Medical genetic: 1. Clinical teratology in the age of genomics. CMAJ. 167(3):265-73.

37. Rothem DE, Rothem L, Soudry M, Dahan A, Eliakim R. Nicotine modulates bone metabolism-associated gene expression in osteoblast cells. J Bone Miner Metab. 2009;27(5):555-61.

38. Sendur OF, Turan Y, Tastaban E, Serter M. Antioxidant status in patients with osteoporosis: a controlled study. Joint Bone Spine. 2009;76(5):514-8.

39. Shah A, Xia L, Goldberg H, Lee KW, Quaggin SE, Fantus IG. Thioredoxin-interacting protein mediates high glucose-induced reactive oxygen species generation by

151

mitochondria and the NADPH oxidase, Nox4, in mesangial cells. J Biol Chem. 2013;288(10):6835-48.

40. Sharma P, Jha AB, Dubey RS, Pessarakli M. Reactive Oxygen Species, Oxidative Damage, and Antioxidative Defense Mechanism in Plants under Stressful Conditions. Journal of Botany, vol. 2012, Article ID 217037.

41. Shiverick KT and Salafia C. 1999. Cigarette smoking and pregnancy I: ovarian, uterine and placental effects. Placenta. 20:265-272.

42. Tilg H, Moschen AR, Kaser A, Pines A, Dotan I. Gut, inflammation and osteoporosis: basic and clinical concepts. Gut. 2008;57(5):684-94.

43. Tong VT, Jones JR, Dietz PM, D’Angelo D, Bombard JM, CDC. 2009. Trends in smoking before, during, and after pregnancy - Pregnancy Risk Assessment Monitoring System (PRAMS), United States, 31 sites, 2000-2005. MMWR Surceill Summ. 58(4):1-29.

44. Trainor P. 2010. Craniofacial Birth Defects: The role of neural crest cells in the etiology and pathogenesis of treacher Collins syndrome and potential for prevention. Am J Med Genet A. 152A(12):2984-94.

45. U.S. Department of Health and Human Services. The Health Consequences of Smoking—50 Years of Progress: A Report of the Surgeon General. Atlanta: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 2014a.

46. U.S. Department of Health and Human Services: Let’s Make the Next Generation Tobacco Free: Your Guide to the 50th Anniversary Surgeon General’s Report on Smoking and Health. Atlanta: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 2014b.

47. U.S. Department of Health and Human Services: The Health Consequences of Involuntary Exposure to Tobacco Smoke: A Report of the Surgeon General. Atlanta: U.S. Department of Health and Human Services, Centers for Disease Control and

152

Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 2006.

48. Valavanidis A, Vlachogianni T, Fiotakis K. Tobacco smoke: involvement of reactive oxygen species and stable free radicals in mechanisms of oxidative damage, carcinogenesis and synergistic effects with other respirable particles. Int J Environ Res Public Health. 2009;6(2):445-62.

49. Valko M, Leibfritz D, Moncol J, Cronin MT, Mazur M, Telser J. Free radicals and antioxidants in normal physiological functions and human disease. Int J Biochem Cell Biol. 2007;39(1):44-84.

50. Wang JY. Regulation of cell death by the Abl tyrosine kinase. Oncogene. 2000;19(49):5643-50.

51. Wang X, Son YO, Chang Q, Sun L, Hitron JA, Budhraja A, Zhang Z, Ke Z, Chen F, Luo J, Shi X. NADPH oxidase activation is required in reactive oxygen species generation and cell transformation induced by hexavalent chromium. Toxicol Sci. 2011;123(2):399-410.

52. WHO report on the global tobacco epidemic, 2011: warning about the dangers of tobacco. World Health Organization.

53. Wickström R. Effects of nicotine during pregnancy: human and experimental evidence. Curr Neuropharmacol. 2007;5(3):213-22.

54. Xu X, Bishop E, Kennedy S, Simpson S, Pechacek T. Annual Healthcare Spending Attributable to Cigarette Smoking: An Update. American Journal of Preventive Medicine 2014;48(3):326–33.

55. zur Nieden NI, Davis LA, Rancourt DE. Comparing three novel endpoints for developmental osteotoxicity in the embryonic stem cell test. Toxicol Appl Pharmacol. 2010a;247(2):91-7.

153

56. zur Nieden NI, Davis LA, Rancourt DE. Monolayer cultivation of osteoprogenitors shortens duration of the embryonic stem cell test while reliably predicting developmental osteotoxicity. Toxicology. 2010b;277(1-3):66-73.

57. zur Nieden NI, Kempka G, Ahr HJ. Molecular multiple endpoint embryonic stem cell test--a possible approach to test for the teratogenic potential of compounds. Toxicol Appl Pharmacol. 2004;194(3):257-69.

58. zur Nieden NI1, Price FD, Davis LA, Everitt RE, Rancourt DE. Gene profiling on mixed embryonic stem cell populations reveals a biphasic role for beta-catenin in osteogenic differentiation. Mol Endocrinol. 2007;21(3):674-85.

154

CHAPTER 6

Harm-reduction tobacco alters bone development through tobacco sensitive

transcriptional regulators

Nicole Renee Lee Sparks and Nicole Isolde zur Nieden

Introduction

Transcriptional regulation of skeletal development is an important step for the correct expression of genes to determine the fate of a developing cell. Even though osteoblasts arise from mesenchymal cells of neural crest or mesoderm origin, each of the pathways to generate that mesenchymal cell has their own regulatory expression profile.

Misregulation of these transcription factors has been identified to play a role in hypoplastic clavicles, various dental defects, delayed ossification of the skull bones, premature fusion of the skull sutures, and a number of other skeletal defects (Jensen et al., 2010). Because of their importance, alteration of such regulatory genes due to toxicant exposure can be detrimental to skeletal development. As previously discussed (chapter 4), toxicants can inhibit osteogenic differentiation of human embryonic stem cells (hESCs) and neural crest

(NC)- and mesoderm-derived osteoblasts can respond differently to chemical agents during differentiation. Our prior work has also shown that when osteogenically differentiating hESCs are exposed to popular brands of tobacco products, their differentiation potential is decreased. Maternal smoking has been shown to be detrimental to bone development of

155 newborns (Carmichael et al, 2008; Kawakita et al., 2008; Karatza et al., 2003). However, it remains to be uncovered if regulatory genes of osteoblast differentiation are tobacco- sensitive.

While many signaling cascades influence transcription factors, oxidative stress can also disrupt transcriptional activity of genes essential for regulating the transcriptional profile of osteogenesis. As outlined in the last chapter, cells are equipped to counteract

ROS and prevent oxidative stress through various mechanisms including the actions of free-scavenging enzymes SOD and CAT. This response requires the activation of the redox-sensitive forkhead foxO (FOXO) family of transcription factors. Therefore, we hypothesized that FOXOs may be misregulated upon tobacco exposure ultimately diminishing the ability to remove ROS.

The FOXO family is evolutionarily conserved with a vital role in cell cycle control, longevity, and regulation of redox mechanisms. Members of the FOXO family include

FOXO1, FOXO3a, FOXO4, and FOXO6 in humans (Huang and Tindall, 2007). FOXO1 and 3 have similar expression patterns in adult tissue, while FOXO6 is specific to the brain

(van der Vos and Coffer, 2008). Drosophila melanogaster and Caenorhabditis elegans studies first identified FOXO and named it dFOXO and DAF-16, respectively (Huang and

Tindall, 2007). Drosophila melanogaster and Caenorhabditis elegans have shown that the transcription factors are controlled by the evolutionary conserved insulin- phosphotidylinositol 3 kinase-protein kinase B (insulin-PI3K-PKB) signal transduction cascade, which promotes the nuclear exclusion of FOXO (Essaghir et al., 2009). To be active, FOXO is post-translationally modified through phosphorylation/

156 dephosphorylation, methylation, ubquitination, and acetylation/deacetylation (Brown and

Webb, 2017). The same mechanisms can regulate FOXO negatively with the addition of miRNAs (Brown and Webb, 2017).

Reports have linked FOXO regulation in oxidative stress-induced osteoblast inhibition. In the absence of FOXO3, bone mass in cortical and cancellous bone was decreased with increased oxidative stress in 3-month-old mice. In a mouse triple knockout of FOXO 1/3/4, enhanced apoptosis and bone loss was observed (Ambrogini et al., 2010).

Further, overexpression of FOXO3 in adult osteoblasts supported bone formation, osteoblast yield, and reduced oxidative stress. While these cases demonstrate the role of

FOXO in adult bone homeostasis, it is unknown if such mechanisms are at play during embryonic skeletal development. It this study, we investigate the role of FOXO during osteoblast differentiation using human embryonic stem cells (hESCs) investigating a link between tobacco exposure, FOXO regulation and downstream gene activation.

Materials and Methods

Human ESC maintenance and differentiation.

Human ESCs were maintained in mTeSR medium as previously described. Briefly, osteogenesis was initiated on confluent colonies with DMEM and 15% FBS for the first 5 days of differentiation. At that time, β‐glycerophosphate, ascorbic acid, and 1,25(OH)2

Vitamin D3 were added to the culture medium for the remainder of the differentiation

(previously described). Throughout differentiation, cells were exposed to either non- effective doses (NED) or effective doses (ED) of Camel SS, Camel Blue SS, or Camel

157

STE. The FOXO concentration-response curve was established from 7 concentrations with an untreated control. Smoke solutions were replenished with each media change.

Neural crest cell induction

H9 hESCs colonies maintained in mTeSR were dissociated with accutase and

plated as singles cells in differentiation medium (DMEM/F12, 2% B27 supplement

[ThermoFisher], 1x Glutamax [Gibco], 0.5% bovine serum albumin). ROCK inhibitor

Y27632 (10 µM Tocris Bioscience) was added for the first 24 hrs. CHIR 99021 (3 µM

Tocris Bioscience) was added to the medium for the first two days.

Immunocytochemistry

Cells were fixed with 4% paraformaldehyde for 30 min at 4°C.and permeabilized with 0.1% Triton‐X 100 for 15 min, cells were incubated in primary antibody diluted in

PBS/10% FBS overnight at 4°C. Primary antibodies used were FOXO1 (ab39670),

FOXO3a (ab23683), SOX10 (Santa Cruz, SC17342), and TFAP2α (C83E10 Cell

Signaling). Cells were then washed and incubated in the appropriate secondary antibody

(AlexaFluor conjugated) and DAPI (4',6-Diamidino-2-Phenylindole, Dihydrochloride) for

2h. Cells were observed on a Nikon Eclipse Ti-S fluorescent microscope.

Histochemical analysis

Cultures were washed with PBS and fixed in 4% paraformaldehyde at 4°C for 30 min. For Von Kossa stain, fixed samples were overlaid with silver nitrate solution (Ricca

158

Chemical Company) and illuminated under an intense light source for 1 h. Cultures were then incubated with 5% sodium thiosulfate (Red Bird Service) for 2 minutes to visualize the von Kossa stain. Cells were also washed with PBS and overlaid with 2% w/v Alizarin

Red S solution (Sigma‐Aldrich) for 5 minutes to visualize calcium‐rich areas. Prior to imaging, cultures were washed 3X with water followed by ascending alcohol washes (70%,

80%, 90%, 100% ethanol).

RNA isolation and quantitative PCR

Total RNA was extracted with the NucleoSpin RNA II protocol (Macherey Nagel).

RNA concentration was determined with a NanoDrop® 1000 spectrophotometer (Thermo

GAPDH expression (Table 5.1).

Protein extraction

On day 7 of differentiation, whole cell, nuclear, and cytoplasmic protein were collected. Prior to collection, cell cultures were pretreated with 1mM sodium orthovanadate

159

(stable for 30 minutes) for 30 minutes to inhibit protein tyrosine phosphatases (PTPs), which regulates the phosphorylation state of multiple signaling molecules. This pre- treatment preserves the phosphorylation of proteins. Whole cell collections were carried out with RIPA buffer (RIPA: 1XPBS, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, pH 7.4, 1:100 phosphatase inhibitors, and 1:100 protease inhibitors)) with Halt protease and phosphatase inhibitor cocktails (Thermo Scientific) and stored at -80°C. Nuclear and cytoplasmic fractions were extracted with the NE-PER Nuclear and Cytoplasmic

Extraction kit as per the manufacturer’s instructions. Harvested cultures were lysed in CER

I buffer provided with the kit (including protease inhibitors, sodium orthovanadate, sodium fluoride, and phenylmethylsulfonyl fluoride. Samples were centrifuged for 5 minutes at maximum setting (~16,000g) and the supernatant extracted and stored at -80°C

(cytoplasmic portion). The pellet was resuspended in NER buffer provided with the kit

(including protease inhibitors, sodium orthovanadate, sodium fluoride, and PMSF). For 40 minutes samples were intermittently vortexed, then centrifuged for 10 minutes to collect the supernatant (nuclear content) and stored at -80°C. Protein concentrations were determined by the Lowry method described in chapter 2.

Western Blot analysis

Equal amounts of protein were loaded into a 10% SDS/polyacrylamide gel and ran at 100V for 1 hour. Proteins were transferred from gels onto PVDF membranes at 30V in a transfer chamber overnight at 4°C, apparatus contained 1X transfer buffer (100 mL 10X transfer buffer [30.2 g Tris-Base, 188 g Glycine, and 1 L ddH2O], 200mL methanol, and

160

700ml ddH2O). Membranes were blocked in 5% BSA/TBST-T for 60 min. The membranes were then at 4°C overnight in one of the following primary antibodies: rabbit anti-FOXO1

(ab39670 Abcam), rabbit anti-FOXO3a (ab23683 Abcam), mouse anti-β-catenin (Thermo

Scientific), mouse anti-TATA binding protein (Abcam), and mouse anti-Tubulin (Abcam).

Horseradish peroxidase-conjugated anti-rabbit IgG or anti-mouse IgG (Santa Cruz

Biotechnology) was used as a secondary antibody diluted in 10% FBS/PBS. Enhanced chemiluminescence (SuperSignal West Pico Chemiluminescence) was used to detect immunoreactive proteins on the Bio-Rad Gel Doc system.

Immunoprecipitation analysis

Three hundred micrograms of nuclear lysates were subjected to immunoprecipitation with 2 µg of β-catenin primary antibody. Proteins bound to protein

A-sepharose CL-4B beads (GE Healthcare) were washed and equal amounts of protein, determined by beta-catenin Western blot, were loaded into a 10% SDS/polyacrylamide gel.

Gels were blotted onto PVDF membranes as described in Western blot protocol (above), incubated with FOXO1 or FOXO3 primary antibodies and subsequently with secondary antibodies for chemiluminescence visualization.

shRNA expression vector construct

FOXO shRNA 21 bp sequences were determined from MISSION® shRNA

Plasmid DNA (Sigma Aldrich). For vector construction, BamHI (5') and HindIII (3')

161 restriction sites were included at the end of the oligo. A loop sequence with an XhoI restriction site was inserted between the sense and antisense sequences. Table 6.1 contains the oligonucleotide sequences, where green indicates the sequence for insertion into the

BamHI or HindIII restriction enzyme sites, and red indicates the short hairpin loop sequence. The sense shRNA sequence is in bold lettering and the antisense strand follows after the loop sequence. Two shRNAs were designed for each FOXO of interest. Forward

(45 µM) and reverse (45 µM) oligos were annealed in 1X annealing buffer (10x annealing buffer: 1M K-acetate, 0.3 M HEPES-KOH pH7.4, and 20 mM Mg-acetate). Annealing was carried out in a thermocycler set to 95°C and programmed to reduce by 1°C every minute until the reaction reached 25°C. 1 µg of the cloning vector, pBK-Flex-shRNA (Fig 6.1), was double digested with 1.0 unit of BamHI and 1.0 unit of HindIII in 50 µL reaction volume (1X

NEB Buffer 2.1 New England

BioLabs) at 37 °C for 1.5 hours.

The digested cloning vector was resolved on a 1% agarose gel and purified using the GeneJet PCR

Purification kit (ThermoFisher).

The annealed oligos were ligated overnight at room temperature to Figure 6.1. pBK-Flex-shRNA plasmid map. shRNA of FoxO1 or FoxO3 was inversely inserted at BamHI and HindIII restriction the digested cloning vector in a 10 sites. shRNA, short hairpin RNA.

162

µL reaction volume (2 µL cloning vector, 2 µL annealed oligonucleotides, 1X ligation buffer, and 1 .0 unit of ligase).

Bacterial transformation

The ligation mixture (5 µL) was transformed into Stbl2™ competent cells (ThermoFisher).

50 µL aliquots of Stbl2 stored at -80 °C were thawed on ice. The ligated DNA was added to the competent bacterial cells and incubated on ice for 5 minutes. The DNA-bacterial cell mixture was heat showed for 30 seconds at 42 °C and then placed back on ice for 2 minutes.

500 µL of Luria-Bertani (LB) broth was added to the mixture and then incubated for 30 minutes at 37 °C. The mixture (100 µL) was spread onto ampicillin (100 µg/mL) LB agar plates and incubated overnight (12-18 hours) at the 37 °C. Individual colonies were selected and grown in 5 mL of LB Broth with ampicillin overnight in a shaking incubator at 37 °C set to 175 RPM. After incubation, the plasmid DNA was purified using the

E.Z.N.A.®Plasmid Mini Extraction Kit (Omega Bio-tek). Purified plasmids were restriction enzyme digested with 1.0 unit of XhoI in 1X CutSmart® Buffer (New England

BioLabs) and resolved on a 1% agarose gel. Plasmids that yield two products at 6.6 kb and

1.6 kb had the correct insertion of the FOXO shRNA into the vector.

Generation of cre-inducible FOXO shRNA stable cell lines

1 µg of pBK-Flex-FOXO1-1 (sequence 1), pBK-Flex-FOXO1-2 (sequence 2), pBK-Flex-FOXO3-1 (sequence 1), and pBK-Flex-FOXO3-2 (sequence 2) plasmids were linearized with 1.0 unit NotI restriction enzyme in 1X NEBuffer 3.1 buffer. The digested

163 plasmids were purified using the GeneJet PCR Purification kit. Single cell human ESCs in mTeSR were transfected with 200 nanograms of linearized pBK-Flex-FOXO1-1, pBK-

Flex-FOXO1-2, pBK-Flex-FOXO3-1, or pBK-Flex-FOXO3-2 using the Effectene

Transfection Reagent (Qiagen). 72-96 h after transfection, individual colonies were selected and plated into a 96-well format. The individual colonies are now referred to as clones with a denoted number. Once confluent, the clones were expanded to a 48-well format and half of the culture was assessed for the presence of GFP by PCR analysis. On day 5 of differentiation, FOXO shRNA clones were transfected with 500 ng of Cre recombinase, generously donated by Dr. Martin Riccomagno, University of California

Riverside. In this Cre-dependent system, the absence of Cre keeps the shRNA in an antisense orientation—the system off. In the presence of Cre, the cassette flips and results in a sense “on” orientation.

FOXO1-1 Forward GATCCatctacgagtggatggtcaaACTCGAGAttgaccatccactcgtagatTTTTTTGGA FOXO1-1 Reverse AGCTTCCAAAAAAatctacgagtggatggtcaaTCTCGAGTttgaccatccactcgtagatG FOXO1-2 GATCCgccggagtttagccagtccaaACTCGAGAttggactggctaaactccggcTTTTTTGGA Forward FOXO1-2 Reverse AGCTTCCAAAAAAgccggagtttagccagtccaaTCTCGAGTttggactggctaaactccggcG FOXO3-1 Forward GATCCggacaatagcaacaagtatacACTCGAGAgtatacttgttgctattgtccTTTTTTGGA FOXO3-1 AGCTTCCAAAAAAggacaatagcaacaagtatacTCTCGAGTgtatacttgttgctattgtccG Reverse FOXO3-2 Forward GATCCcatgttcaatgggagcttggaACTCGAGAtccaagctcccattgaacatgTTTTTTGGA FOXO3-2 AGCTTCCAAAAAAcatgttcaatgggagcttggaTCTCGAGTtccaagctcccattgaacatgG Reverse

Table 6.1. Oligonucleotide sequences.

164

Dot Blot analysis

FOXO shRNA clones grown in a 48-well format were lysed in RIPA buffer (RIPA;

1XPBS, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, pH 7.4, and 1:100 protease inhibitors) and 2 µL of protein lysates were spotted on strips of PVDF membrane grids.

The membranes were dried and blocked in 5% BSA in TBS-T buffer for 30 min. Blots were incubated in FOXO1, FOXO3, or Tubulin primary antibodies for 30 min. HRP-linked secondary antibodies were incubated for 30 min. SuperSignal West Pico

Chemiluminescence was used to detect immunoreactive proteins on the Bio-Rad Gel Doc system

RNA isolation, library preparation, and sequencing

Total RNA of HRTP treated or control cultures were extracted with the NucleoSpin

RNA II protocol. 100 ng of each RNA sample was used as input for RNA sample preparations. All samples passed initial quality testing using Bioanalyzer 2100 (Agilent), assuring RIN values > 9. Ribosomal RNA was depleted using the NEBNext® Poly(A) mRNA Magnetic Isolation Module. The libraries were constructed using NEBNext®

Ultra™ II Directional RNA Library Prep Kit for Illumina. Libraries were submitted for 75- bp single-end sequencing by Illumina HiSEQ 2000 at the Genomics Core in the UCR

Institute of Integrated Genome Biology (IIGB). A total of three samples for each time course (each sample in triplicate) were multiplexed and sequenced in one lane, each of which yielded ∼240 million reads (∼27 million reads per sample).

165

Single-end sequencing reads were uploaded to the Galaxy web platform

(usegalaxy.org) (Afgan., et al., 2016) for analysis. The human reference genome

(GRCh38/hg38) was aligned to the reads with the Tophat v2.1.0 algorithm (Trapnell et al.,

2009). The data aligned by Tophat were processed by Cufflinks (Trapnell et al., 2010) to assemble transcripts and to measure their relative abundance in fragments per kilobase of exon per million fragments mapped (FPKM). Assembled transcripts from experimental samples were compared with the annotated transcriptome (GRCh38.89) downloaded from

UCSC genome browser and examined for differential expression using Cuffdiff. The following criteria were used to select differentially expressed genes: (i) a fold change (FC) of at least 1 or more and (ii) a false discovery rate (FDR) threshold of 5%. Principle component analysis was generated in R on rlog (Love et al., 2014) transformation on the normalized count data of the differentially expressed genes. Heatmaps were generated also in R with the rlog transformation and clustered based on Pearson correlation. Gene ontology (GO) analysis was performed using DAVID Bioinformatics Resources 6.8

(Huang et al., 2009a, 2009b). GO terms were submitted to REVIGO to make a graph-based summary of the differentially regulated genes (Supek et al., 2011). Cytoscape software v3.6.1 was used to construct GO term networks from the REVIGO summary (Shannon et al., 2003). VENNY 2.1 (Oliveros, 2007-2015) was used to compare lists of genes in the different treatment cultures.

166

Statistical analysis

Data presented are averages of three independent experiments ± standard deviation.

Significance was determined by using a web-based one-way ANOVA and Tukey HSD post hoc test (http://faculty.vassar.edu/lowry/anova1u.html) or unpaired Student's t-test as appropriate.

Results

FOXO1 and FOXO3 activity increases during osteogenesis

Published studies have shown that FOXO is activated and necessary for osteoblast maintenance (Iyer et al., 2013). However, it is not clear if FOXOs are needed during the

Figure 6.2. FoxOs are expressed during osteogenic differentiation. A) Time-course of FoxO1 and FoxO3 expression during osteogenic differentiation assessed by immunocytochemistry. FoxO presence correlates to neural crest cell- mediated osteogenic differentiation. FoxO1, FoxO3, and SOX10 positive cells appeared on day 7 of differentiation. B) FoxO (green), SOX10 (orange), and TFAP2α (red) with DAPI (blue) immunostaining on day 5 neural crest cells differentiated from hESCs (Leung et al., 2016). hESCs, human embryonic stem cells; TFAP2a, transcription factor AP-2 alpha; Sox10, SRY-box 10. Scale bar = 100 µm

167 entire differentiation process beginning with a pluripotent cell. To determine the presence of FOXOs during early osteogenic differentiation, we performed immunocytochemistry on hESC osteogenic cultures using time points corresponding to germ layer specification and lineage determination (Fig 6.2). Both FOXO1 and FOXO3a appeared to be transiently expressed during these early stages of cellular commitment. They were present in the nucleus as cells exited from pluripotency (day 1), but then down-regulated after.

Previous work has demonstrated that our differentiation protocol predominantly gives rise to osteoblasts of the neural crest lineage and to a lesser degree to mesoderm

(chapter 2). After the addition of osteogenic factors to the medium, FOXOs were concomitantly expressed with SOX10 on day 7, indicating that both FOXO isoforms are found in neural crest cells. When using a 5-day neural crest induction protocol (Leung et al., 2016, described in chapter 2), we found cells that were triple-positive for FOXO1 or

FOXO3 and the neural crest markers SOX10 and TFAP2α. These data suggest that FOXOs may play a role during osteogenic differentiation and, at least in part, during neural crest- associated osteogenic differentiation.

HRTP represses the expression of FOXO1 and FOXO3

As it became evident that FOXOs are expressed during in vitro osteogenesis, we next asked whether FOXO transcription was altered by HRTPs (Fig 6.3A). The osteogenically differentiating cultures were treated with different concentration of HRTPs and harvested on day 7, the time point that was previously found to show altered antioxidant

168

enzyme activity as well

as ROS accumulation

(see chapter 5). A dose-

response curve shows a

concentration-

dependent decrease in

FOXO mRNA levels,

similar to that found for

RUNX2 and TWIST1

(Fig 5.1A). Indeed, the

Figure 6.3. Harm reduction tobacco products diminish FOXO expression. A) effective doses of Quantitative PCR concentration-response curve of FOXO1 and FOXO3. FOXO mRNAs were downregulated with increasing doses of Camel Blue and Camel Snus HRTPs chosen tobacco extract exposure. B) Selected effective doses (ED) used in studies decreased FOXO1 and FOXO3 expression on day 7 of differentiation. n = 3 independent samples ± SD, *p<0.05 FOXO1 expression, one‐way ANOVA versus throughout this study untreated control and Δp<0.05 FOXO3 expression, one-way ANOVA versus untreated control. NED, non-effective dose; STE, Camel Snus tobacco extract. decreased FOXO mRNA levels (Fig 6.3B). However, the cytotoxic conventional Camel SS did not have a regulatory effect on the transcription of FOXO.

HRTP excludes FOXO from the nucleus.

We next determined whether the HRTP-mediated reduction in FOXO mRNAs manifested at the protein level. While cytoplasmic and whole protein levels did not show a misregulation (Fig 6.4 A, B), FOXO1 and FOX3 protein levels were markedly reduced in the effective doses of Camel Blue and Camel STE (Fig 6.4B), suggesting that FOXOs were

169 inhibited from transcriptional activation upon HRTP exposure. Overall, this finding may suggest why the SOD2 and CAT mRNAs were down-regulated in HRTP-treated cells.

A co-factor for FOXO1 and 3 to activate the oxidative stress defense targets SOD2 and CAT is β-catenin (CTNNB1) (Iyer et al., 2013; Almeida et al., 2007). In addition, the

Wnt/CTNNB1 signaling pathway plays an essential role in differentiation during skeletal development (Ding et al., 2012; zur Nieden et al., 2007). Specifically, in the absence of

CTNNB1, craniofacial development is inhibited, and long bones are under-mineralized and shortened (Brault et al., 2001; Nishimoto et al., 2015).

To determine whether FOXO absence affected CTNNB1 binding in differentiating osteoblast, we examined if the two proteins interact in the nucleus. FOXO1 and FOXO3 both associated with CTNNB1 in untreated and NED cultures of Camel Blue and STE. The degree of this interaction was decreased in ED of Camel Blue and STE (Fig 6.4A).

However, CTNNB1 protein levels were not altered due to HRTP exposure, thus suggesting that the diminished interaction between FOXO and CTNNB1 is driven by the diminished nuclear availability of FOXO. Generally, this finding may also point to an inability of additional proteins to shuttle into the nucleus.

170

Figure 6.4. HRTP-mediated nuclear exclusion of FOXO. A-C) Western blot analysis of cytoplasmic (A), whole cell (B) and nuclear (C) protein fractions. Cytoplasmic and whole cell lysates do not show reduced FOXO protein in Camel Blue and STE effective doses (ED), while nuclear FOXO localization was decreased in Camel Blue and STE treated cells on day 7. D) Camel Blue and STE decrease CTNNB1/FOXO interaction. Nuclear protein extracts were immunoprecipitated with CTNNB1 and Western blots performed. Densitometry suggests decreased presence of CTNNB1/FOXO complex in ED. CTNNB1, beta-catenin; STE, Camel Snus tobacco extract.

171 shRNA-mediated knockdown of FOXO mimics exposure-associated inhibition of osteogenesis

To causally determine the link between FOXO availability and early commitment of cells that may continue through differentiation to give rise to osteoblasts, we examined the effect of an shRNA-mediated FOXO knockdown on osteogenic differentiation. FOXO hairpin sequences were cloned into the pBK-Flex shRNA vector and human ESCs were then transfected with the resulting shFOXO1 or shFOXO3 plasmids to create stable cell

Figure 6.5. Activation of FOXO shRNA reduces FOXO protein levels. (A) Dot blots were performed on whole cell lysates after the addition of Cre recombinase to H9 hESCs stably transfected with a hairpin against the respective FOXO as indicated. 2 µL dots of lysates were probed with primary antibodies. shRNA-mediated knockdown of FoxO (orange boxes) was determined to select clones for future experiments. (B) shFOXO clones were osteogenically differentiated. Cre recombinase was administered on day 5 and cells were fixed on day 7. Clones showed a reduction in FoxO nuclear localization and overall fluorescence intensity compared to the Mock control. Scale bar = 100 µm.

172 lines. The stable clones were then taken through an osteogenic differentiation and on day

5 of differentiation, Cre was added to induce knockdown. Knockdown was confirmed with a dot blot analysis (Fig. 6.5A). On day 7, immunocytochemistry analysis also confirmed nuclear absence of the FOXOs (Fig 6.5B). On day 20, a calcification analysis showed reduced calcium content in all but one of the shRNA-FOXO clones (Fig 6.6 A,B). In this, the FOXO knockdown mimicked HRTP-mediated inhibition of osteogenesis. Together, these data suggest that HRTP exposure negatively effects FOXO expression, which in turn can lead to decreased osteoblast differentiation.

Figure 6.6. FOXOs are required for osteogenic differentiation. (A) Calcium quantification in cellular lysates measured by Arsenazo III showed reduced calcium levels compared to mock transfected cells, n=5± SD, *p<0.05 one-way ANOVA vs mock. (B) FOXO1 and FOXO3 knockdown cells exhibited attenuated osteogenic differentiation. The appearance of the black deposits of a calcified extracellular matrix were reduced (brightfield microscopy). Alizarin and von Kossa staining also showed reduced mineralization. Scale bar = 500 µm.

HRTPs induce unique expression profiles in differentiating osteoblasts While FOXO target promoters are well known in association with oxidative stress removal, it is unclear which down-stream transcriptional changes occur together with a reduction in nuclear FOXO during osteogenesis and/or early osteoprogenitor commitment.

Hence, HRTP-associated changes in transcriptional regulation were determined globally

173 using RNA-seq on day 7 differentiated cells. Principal component analysis (PCA) on normalized data showed that triplicates clustered together within their treatment (Fig 6.7A).

Differential expression analysis comparing untreated cells to Camel Blue and STE uncovered that 606 and 181 genes were differentially expressed, respectively (Fig 6.7B).

Interestingly, this analysis showed that there were more differentially regulated genes in

Camel Blue-treated cells versus STE-treated cells. In determining overlapping differentially expressed genes (DEGs) between the two HRTPs, a VENN analysis was conducted. The result showed that Camel Blue had 590 genes differentially expressed STE had 163 unique genes expressed (Appendix 2). There were no common DEGs in the

Figure 6.7. Differential expression of genes treated with Camel Blue and Camel Snus tobacco extract as identified with RNA-seq. (A) Principle component analysis (PCA) demonstrates clustering of Camel Blue and STE treated replicates on day 7 of differentiation. (B) Total number of genes dysregulated in Camel Blue and STE cells compared to untreated control or Camel Blue compared to STE. (C) Venn diagram of Camel Blue and STE compared to untreated genes differentially regulated from panel B. (D) Venn diagram shows the total number of genes differentially regulated in comparison between untreated to STE and Camel Blue to STE. UNT, untreated; STE, Camel Snus tobacco extract; CB, Camel Blue; NED, non-effective dose; ED, effective dose.

174 combined up- and downregulated genes between Camel Blue and STE (Fig 6.7C). The

HRTPs shared 11 genes downregulated and 1 gene upregulated. In comparison, analysis of

Camel Blue to STE revealed 832 genes were differentially expressed between the two treatment groups (Fig 6.7D). PCA and heatmap findings suggested that Camel Blue and

STE had their own HRTP-sensitive transcriptome. The top DEGs were determined by

Log2FC of ≥1.0 (Fig 6.8; Fig 6.9). In the genes that were differentially expressed between untreated to STE and Camel Blue to STE, there were 61 genes in common (Fig 6.10).

175

Figure 6.8. Highly differentially expressed genes sensitive to Camel Blue exposure during osteogenic differentiation. Graph of dysregulated genes with ≥1.0-Log2 fold change (FC).

176

Figure 6.9. STE-responsive differentially expressed genes. Graph of dysregulated genes with ≥1.0-Log2 fold change (FC). STE, Camel Snus tobacco extract.

177

Figure 6.10. Genes downregulated unique to Camel Snus tobacco extract treatment. Graph shows the Log2 fold change (FC) of 61 commonly downregulated genes by STE. Unt, untreated; CB, Camel Blue; STE, Camel Snus tobacco extract.

178

Gene ontology (GO) analysis of Camel Blue and STE DEGs displayed biological processes with unique networks that are each known to have a role in the pathogenesis of bone defects (Fig 6.11). Significant GO terms of p-values less than 0.05 for Camel Blue were found for downregulated genes associated with cell cycle regulation, interaction of

CTNNB1/TCF, transcriptional regulation, and nucleosome assembly (Fig 6.12A). The upregulated genes in Camel Blue were involved with mitochondrial assembly, drug, and response to ROS (Fig 6.12B). STE downregulated genes were specific for developmental pathways such as somitogenesis, mesoderm formation, and bone morphogenesis (Fig

6.12A). GO analysis revealed 3 distinct categories for STE upregulated genes for ECM organization, cell shape regulation, and phagocytosis (Fig 6.12B). Notably, Camel Blue downregulated IGF1R, EP300, EGFR, and PTCH2 genes (Fig 6.8). Downregulation of any of these genes results in a reduction of bone (Linder et al., 2018; Klein et al., 2016; Hu et al., 2015; Xian et al., 2010). Interestingly, IGF1R, EP300, and EGFR are FOXO targets

(Chiu et al., 2016; Guntur and Rosen, 2013; van der Heide and Smidt, 2005). STE had the following downregulated genes: T, EOMES, CDX1, CDX2, MSX2, CER1, dysregulation of any of which can result in axial, appendicular, or craniofacial malformations. (Fig 6.9;

Pennimpede et al., 2012; Savory et al., 2009; Hart et al., 2002, Zhu et al., 2016;

Papaioannou, 2014, Papaioannou, 2014; Satokata et al., 2000). Together the RNA-seq results showed that genes that are regulated early in osteoblast differentiation are indeed sensitive to Camel Blue and STE exposure.

179

Figure 6.11. Camel Blue and Camel Snus tobacco extract regulate different biological processes in differentiating cells. A) Gene ontology networks of Camel Blue. B) STE gene ontology networks. More biological processes were dysregulated in the Camel Blue treated cells. STE altered pathways involved in early development. STE, Camel Snus tobacco extract; GO, gene ontology.

180

Figure 6.12. Camel Blue and Camel Snus tobacco extract negatively affect bone development through separate biological processes. A) Gene ontology (GO) analysis of downregulated genes. B) Geno ontology (GO) of upregulated genes. C) Gene ontology of diseases of downregulated genes. STE, Camel Snus tobacco extract; GO, gene ontology.

Discussion Tobacco use is associated with several adverse health outcomes, including defects in bone development. It is important to understand the specific genes and pathways of bone development that are dysregulated due to tobacco exposure. Other groups have focused on the role of osteoclast activity to lead to a skeletal defect. However, there is a gap in understanding of how altered gene expression during osteoblast differentiation can describe osteogenic inhibition. Here we provide here the first report of the effects of Camel Blue

181 and STE on human ESC osteoblast differentiation, identified a potential role of FOXO1/3 including a global gene expression analysis.

FOXOs are crucial in mediating longevity, yet their exact role during embryonic development is not well understood. Other studies have shown that the decrease of FOXO in the adult mouse skeleton will inhibit bone homeostasis (Ambrogini et al., 2010). We expand here on this knowledge of FOXO in bone development. Specifically, we show that

FOXOs are present on the first day of hESC differentiation and downregulated further into differentiation. After osteogenic differentiating factors are added to the culture medium, both FOXO1 and FOXO3 are increasingly expressed on day 7 of differentiation, showing that FOXO may have a role in early osteogenesis in a time dependent manner and potentially suggesting that expression of FOXOs may be controlled by the vitamin D3 receptor (as vitamin D3 is one of the three osteogenic supplements).

Further, it appears FOXOs play a specific part in the differentiation of neural-crest osteoblasts. Previously we had shown that our osteogenic induction protocol gives rise osteoblasts from the NC-lineage, with positive identification of NC cells on day 7. When we put the H9 hESCs through a NC-specific differentiation protocol, the cells were positive for NC markers as well as FOXO.

HRTP exposure to osteogenically differentiating cells dose-dependently decreased

FOXO1 and FOXO3 mRNA expression. The decrease in FOXO was in agreement with previous findings that osteogenesis decreased with increasing concentrations of Camel

Blue and STE (chapter 5). In addition, FOXO transcriptional activity, determined by SOD2 expression, was decreased in our HRTP-treated cells. However, we did not see

182 misregulation of CAT. This is in line with previous findings from our laboratory, in which chromatin immunoprecipitation studies in murine ESCs could not confidently determine

FOXO1 and FOXO3 contribution to CAT expression either (McClelland-Descalzo et al.,

2016). While CAT is efficient at reducing hydrogen peroxide, CAT is confined to the peroxisomes and may not be the major contributor of hydrogen peroxide detoxification.

RNA-seq analysis uncovered the major hydrogen peroxide detoxifying enzyme, glutathione peroxidase (GPX) was downregulated in HRTP-treated cells. Though not concluded to be regulated by FOXO, the lack of GPX further demonstrates the breakdown in cellular defense mechanisms.

In response to environmental stimuli, FOXO interacts with CTNNB1 to upregulate mechanisms to thwart oxidative stress. However, the complex of these two can prevent

CTNNB1 from complexing with LEF/TCF, a crucial interaction for osteogenesis (Iyer et al., 2013). Immunoprecipitation analysis showed a reduced interaction between FOXO and

CTNNB1. Further, we found lower nuclear protein levels of FOXO. FOXO regulation is controlled through protein-protein interactions and post-translation modifications (Brown and Webb, 2018). For example, acetylation of FOXO leads to its activation and remain in the nucleus to turn on its target genes. Therefore, it is plausible that HRTPs can either prevent FOXO from shuttling into the nucleus or make it prone to exiting through post- translational modifications (Daitoku, et al., 2011).

Others have shown that the silencing of FOXO decreases osteoblast yield (Iyer,

2013; Ambrogini et al., 2010; Teixeira et al., 2010; Almeida et al., 2007). In these studies,

FOXO was deleted in mature osteoblasts or at the adult MSC stage. These data represent

183 defects in skeletal homeostasis relevant to an adult. In contrast to those prior studies, we removed FOXO at an early stage of development and the results we gathered pertain to the developing fetus. During osteogenic differentiation of the shFOXO1 and shFOXO3 clones, the cultures phenocopied the reduction of osteoblast production as seen with the HRTP treatment. These data support the notion that FOXOs reduces bone formation, however add an earlier level of regulation that was previously unknown. Consequently, the hypothesis that HRTPs might suppress FOXOs yielding osteoblast inhibition upon exposure seems feasible.

RNA-seq results revealed that each Camel Blue and Camel STE had their own distinct effects on osteogenesis. Camel Blue had DEGs that broadly targeted cellular functions such as DNA regulation (nucleosome assembly and histone modification), mitochondrial function, and growth receptors (EGFR and IGF1R). Whereas STE had an overwhelmingly amount of DEGs associated to paraxial mesoderm. Nevertheless, each of these mechanistic pathways is linked to bone development. In the data set, we uncovered

FOXO targets or genes associated with FOXO such as, C/EBPα, EP300, EGFR, IGF1R,

EOMEs, and NODAL. Specifically, all the downregulated genes resulted in diseases associated with bone development and mineralization determined by GO analysis disease pathway (Fig 6.12C). However, further investigation is required for the confirmation of these RNA-seq results and whether these genes are truly regulated by FOXOs.

In sum, this study found that exposure to harm-reduction tobacco products are detrimental to the developing skeleton. Our results suggest depletion of FOXO1 and

FOXO3 mRNA and protein in Camel Blue and STE treated cells. RNA-seq analysis

184 identified HRTP-specific DEGs that are linked to adverse skeletal development.

Ultimately, this study is closing the gap on understanding the underlying molecular mechanisms involved in the pathology of HRTP-induced skeletal defects.

185

References 1. Afgan E, Baker D, van den Beek M, Blankenberg D, Bouvier D, Čech M, Chilton J, Clements D, Coraor N, Eberhard C, Grüning B, Guerler A, Hillman-Jackson J, Von Kuster G, Rasche E, Soranzo N, Turaga N, Taylor J, Nekrutenko A, Goecks J. The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2016 update. Nucleic Acids Res. 2016;44(W1):W3-W10.

2. Almeida M, Han L, Martin-Millan M, O'Brien CA, Manolagas SC. Oxidative stress antagonizes Wnt signaling in osteoblast precursors by diverting beta-catenin from T cell factor- to forkhead box O-mediated transcription. J Biol Chem. 2007;282(37):27298-305.

3. Ambrogini E, Almeida M, Martin-Millan M, Paik JH, Depinho RA, Han L, Goellner J, Weinstein RS, Jilka RL, O'Brien CA, Manolagas SC. FoxO-mediated defense against oxidative stress in osteoblasts is indispensable for skeletal homeostasis in mice. Cell Metab. 2010;11(2):136-46.

4. Babicki S, Arndt D, Marcu A, Liang Y, Grant JR, Maciejewski A, Wishart DS. Heatmapper: web-enabled heat mapping for all. Nucleic Acids Res 2016;44(W1):W147-53.

5. Brault V, Moore R, Kutsch S, Ishibashi M, Rowitch DH, McMahon AP, Sommer L, Boussadia O, Kemler R. Inactivation of the beta-catenin gene by Wnt1-Cre-mediated deletion results in dramatic brain malformation and failure of craniofacial development. Development 2001;128(8):1253-64.

6. Brown AK, Webb AE. Regulation of FOXO Factors in Mammalian Cells. Curr Top Dev Biol. 2018;127:165-192. 7. Carmichael SL, Ma C, Rasmussen SA, Honein MA, Lammer EJ, Shaw GM. Craniosynostosis and maternal smoking. Birth Defects Res A Clin Mol Teratol 2008;82(2):78-85.

8. Chiu CF, Chang YW, Kuo KT, Shen YS, Liu CY, Yu YH, Cheng CC, Lee KY, Chen FC, Hsu MK, Kuo TC, Ma JT, Su JL. NF-κB-driven suppression of FOXO3a contributes to EGFR mutation-independent gefitinib resistance. Proc Natl Acad Sci U S A. 2016;113(18):E2526-35.

186

9. Daitoku H, Sakamaki J, Fukamizu A. Regulation of FoxO transcription factors by acetylation and protein-protein interactions. Biochim Biophys Acta. 2011;1813(11):1954-60.

10. Ding H, Keller KC, Martinez IK, Geransar RM, zur Nieden KO, Nishikawa SG, Rancourt DE, zur Nieden NI. NO-β-catenin crosstalk modulates primitive streak formation prior to embryonic stem cell osteogenic differentiation. J Cell Sci. 2012;125(Pt 22):5564-77.

11. Essaghir A, Dif N, Marbehant CY, Coffer PJ, Demoulin JB. The transcription of FOXO genes is stimulated by FOXO3 and repressed by growth factors. J Biol Chem. 2009;284(16):10334-42.

12. Guntur AR, Rosen CJ. IGF-1 regulation of key signaling pathways in bone. Bonekey Rep. 2013;2:437.

13. Hart AH, Hartley L, Sourris K, Stadler ES, Li R, Stanley EG, Tam PP, Elefanty AG, Robb L. Mixl1 is required for axial mesendoderm morphogenesis and patterning in the murine embryo. Development. 2002;129(15):3597-608.

14. Hedrick SM, Hess Michelini R, Doedens AL, Goldrath AW, Stone EL. FOXO transcription factors throughout T cell biology. Nat Rev Immunol. 2012 Sep;12(9):649- 61. doi: 10.1038/nri3278.

15. Hu Z, Wang Y, Sun Z, Wang H, Zhou H, Zhang L, Zhang S, Cao X. miRNA-132-3p inhibits osteoblast differentiation by targeting Ep300 in simulated microgravity. Sci Rep. 2015;5:18655.

16. Huang DW, Sherman BT, Lempicki RA. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res 2009a;37(1):1-13.

17. Huang DW, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID Bioinformatics Resources. Nature Protoc 2009b;4(1):44-57.

18. Huang W, Yang S, Shao J, Li YP. Signaling and transcriptional regulation in osteoblast commitment and differentiation. Front Biosci. 2007;12:3068-92.

187

19. Iyer S, Ambrogini E, Bartell SM, Han L, Roberson PK, de Cabo R, Jilka RL, Weinstein RS, O'Brien CA, Manolagas SC, Almeida M. FOXOs attenuate bone formation by suppressing Wnt signaling. J Clin Invest. 2013;123(8):3409-19.

20. Jensen ED, Gopalakrishnan R, Westendorf JJ. Regulation of gene expression in osteoblasts. Biofactors. 2010;36(1):25-32.

21. Karatza AA, Varvarigou A, Beratis NG. 2003. Growth up to 2 years in relationship to maternal smoking during pregnancy. Clin Pediatr (Phila). 42:533–541.

22. Kawakita A, Sato K, Makino H, Ikegami H, Takayama S, Toyama Y, Umezawa A. 2008. Nicotine acts on growth plate chondrocytes to delay skeletal growth through the alpha7 neuronal nicotinic ac-etylcholine receptor. PLoS One. 3(12):e3945.

23. Kim D, Langmead B, Salzberg SL. HISAT: a fast spliced aligner with low memory requirements. Nat Methods. 2015;12(4):357-60.

24. Klein C, Zwick A, Kissel S, Forster CU, Pfeifer D, Follo M, Illert AL, Decker S, Benkler T, Pahl H, Oostendorp RA, Aumann K, Duyster J, Dierks C. Ptch2 loss drives myeloproliferation and myeloproliferative neoplasm progression. J Exp Med. 2016;213(2):273-90.

25. Linder M, Hecking M, Glitzner E, Zwerina K, Holcmann M, Bakiri L, Ruocco MG, Tuckermann J, Schett G, Wagner EF, Sibilia M. EGFR controls bone development by negatively regulating mTOR-signaling during osteoblast differentiation. Cell Death Differ. 2018. [Epub ahead of print]

26. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 2014;15(12):550.

27. Nishimoto S, Wilde SM, Wood S, Logan MP. RA Acts in a Coherent Feed-Forward Mechanism with Tbx5 to Control Limb Bud Induction and Initiation. Cell Rep. 2015;12(5):879-91.

28. Oliveros, JC. (2007-2015) Venny. An interactive tool for comparing lists with Venn's diagrams. http://bioinfogp.cnb.csic.es/tools/venny/index.html

188

29. Papaioannou VE. The T-box gene family: emerging roles in development, stem cells and cancer. Development. 2014;141(20):3819-33.

30. Pennimpede T, Proske J, König A, Vidigal JA, Morkel M, Bramsen JB, Herrmann BG, Wittler L. In vivo knockdown of Brachyury results in skeletal defects and urorectal malformations resembling caudal regression syndrome. Dev Biol. 2012;372(1):55-67.

31. Salvador JM, Brown-Clay JD, Fornace AJ Jr. Gadd45 in stress signaling, cell cycle control, and apoptosis. Adv Exp Med Biol. 2013;793:1-19.

32. Satokata I, Ma L, Ohshima H, Bei M, Woo I, Nishizawa K, Maeda T, Takano Y, Uchiyama M, Heaney S, Peters H, Tang Z, Maxson R, Maas R. Msx2 deficiency in mice causes pleiotropic defects in bone growth and ectodermal organ formation. Nat Genet. 2000;24(4):391-5.

33. Savory JG, Pilon N, Grainger S, Sylvestre JR, Béland M, Houle M, Oh K, Lohnes D. Cdx1 and Cdx2 are functionally equivalent in vertebral patterning. Dev Biol. 2009;330(1):114-22.

34. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13(11):2498-504.

35. Supek F, Bošnjak M, Škunca N, Šmuc T. REVIGO summarizes and visualizes long lists of gene ontology terms. PLoS One. 2011;6(7):e21800.

36. Trapnell C, Pachter L, Salzberg SL. TopHat: discovering splice junctions with RNA- Seq. Bioinformatics. 2009 May 1;25(9):1105-11.

37. Trapnell C, Williams BA, Pertea G, Mortazavi A, Kwan G, van Baren MJ, Salzberg SL, Wold BJ, Pachter L. Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat Biotechnol. 2010;28(5):511-5.

38. van der Heide LP, Smidt MP. Regulation of FoxO activity by CBP/p300-mediated acetylation. Trends Biochem Sci. 2005;30(2):81-6.

189

39. van der Vos KE, Coffer PJ. FOXO-binding partners: it takes two to tango. Oncogene. 2008;27(16):2289-99.

40. Xian L, Wu X, Pang L, Lou M, Rosen CJ, Qiu T, Crane J, Frassica F, Zhang L, Rodriguez JP, Xiaofeng Jia, Shoshana Yakar, Shouhong Xuan, Argiris Efstratiadis, Mei Wan, Xu Cao. Matrix IGF-1 maintains bone mass by activation of mTOR in mesenchymal stem cells. Nat Med. 2012;18(7):1095-101.

41. Zhang J, Tu Q, Grosschedl R, Kim MS, Griffin T, Drissi H, Yang P, Chen J. Roles of SATB2 in osteogenic differentiation and bone regeneration. Tissue Eng Part A. 2011;17(13-14):1767-76.

42. Zhu J, Kwan KM, Mackem S. Putative oncogene Brachyury (T) is essential to specify cell fate but dispensable for notochord progenitor proliferation and EMT. Proc Natl Acad Sci U S A. 2016;113(14):3820-5.

43. zur Nieden NI1, Price FD, Davis LA, Everitt RE, Rancourt DE. Gene profiling on mixed embryonic stem cell populations reveals a biphasic role for beta-catenin in osteogenic differentiation. Mol Endocrinol. 2007;21(3):674-85.

190

CONCLUSION

Cigarette smoke, a major environmental toxicant, is a complex mixture of over

7,000 chemicals. Tobacco use is the most preventable cause of death in the United States and is a health risk for many diseases including congenital abnormalities. Pregnant mothers who experience difficulties in quitting smoking, they may turn to harm-reducing tobacco products like “Lights” cigarettes or non-combustible chewing tobacco. While, these harm- reduction tobacco products (HRTPs) may seem to be a safe alternative for the smoker, the unborn child may be susceptible to unwanted effects. Because it unknown how HRTPs can affect the developing skeleton, in this study, we used human embryonic stem cells to uncover the potential adverse effects of HRTP on the developing skeleton.

Human embryonic stem cells (hESCs) are a developmentally powerful tool to understand embryotoxicity. Human ESCs are derived from the inner cell mass of the blastocyst and have the capability to give rise to every cell type in the body, including the bone forming osteoblasts. Because hESCs can recapitulate embryogenesis, hESCs hold great promise in assessing the effects of toxicants on cell differentiation and viability. Our laboratory has uncovered a promising alternative approach to study developmental bone toxicity that can replace in vitro murine-based models and the reliance on in vivo animal models.

In this work, we provide the first evidence that HRTPs may be developmentally toxic to the developing skeleton. We show that H9 hESCs have the ability to be differentiated into osteoblasts. The use of 1,25-dihydroxyvitamin D3 with ascorbic acid and β-glycerophosphate enhances osteogenesis and upregulates bone-specific genes. These

191 osteoblasts exhibit the hallmark mineralization of the extracellular matrix. In addition, it appears that our differentiation protocol, primarily gives rise to osteoblasts from the neural crest lineage, determined by mRNA upregulation of PAX7 and SOX10 and decreased

TBRA.

Human ESCs are a great model for assessing the embryotoxic potential of a chemical. The zur Nieden lab has been at the forefront of in vitro embryotoxicity screening models for skeletal development. We employed our osteogenic differentiation model to assess the effects of HRTP-mediated differentiation inhibition. Further, our aim was to uncover the mechanisms that lead to osteoblast inhibition. Preliminary data from our laboratory demonstrated that when osteogenically differentiating hESCs are exposed to extracts of popular brands of tobacco their differentiation potential is decreased.

Specifically, we exposed hESCs to mainstream (MS) smoke, what is actively inhaled by the smoker, sidestream (SS) smoke, which burns off the tip of the cigarette, or and Snus, fine grain chewing tobacco, tobacco extract (STE). We chose the popularly publicized and commonly used brand of Camel tobacco products. Mainstream smoke Camel conventional and Camel ‘Lights” renamed to Camel Blue, termed for being ultra-filtered and having lower nicotine and lower tar levels, was not cytotoxic to the differentiating cultures. Camel

SS, Camel Blue SS, and Camel STE inhibited osteogenesis measured by the reduction in calcium levels. One method tobacco may cause osteoblast inhibition is through reactive oxygen species (ROS) production. Oxidative stress, the imbalance between ROS levels and free scavenging enzymes, are known to damage DNA and protein that can result in apoptosis, aging, and disease. We found that Camel, Camel Blue, and STE generated

192

2- excessive levels ROS, O and H2O2, during early stages of osteogenesis. It appeared that the higher levels of ROS generated by Camel SS smoke yielded a cytotoxic effect. In contrast, Camel Blue and STE generated sub-toxic levels of ROS that allowed for continued cell survival but inhibited osteoblast differentiation. With the high levels of ROS, we found diminished antioxidant activity in the tobacco treated cells, thus leaving the tobacco treated cells with an inability to counteract the ROS, therefore, suggesting that

HRTPs, Camel Blue and STE, may be more harmful to the developing skeleton by perturbing differentiation.

Enzyme antioxidant activity is regulated through the transcription factor FOXO.

The measured high levels of ROS combined with reduced antioxidant activity suggests transcriptional dysregulation of FOXO. The FOXO family of transcription factors counteracts ROS through upregulation of antioxidants (SOD2 and CATALASE) and promotes osteoblast differentiation and development. HRTP concentration-response curves showed that FOXOs were dose-dependently downregulated. Western blot analysis of nuclear protein of HRTP-treated cells showed diminished levels of FOXO compared to the untreated control. There were no detectable differences of FOXO protein in cytoplasmic fractions and whole cell protein. Beta-catenin, essential for bone formation, is a co- transcriptional activator and is known to interact with FOXOs to upregulate responses to oxidative stress. Immunoprecipitation of HRTP protein lysates showed reduced

FOXO/beta-catenin interaction in the nucleus. Further investigation is needed for the shuttling of FOXO, but it is plausible that HRTP exposure is dysregulating cellular shuttling machinery and preventing the translocation of FOXO into the nucleus and, hence,

193 inhibiting FOXO target genes. Through short hairpin RNA (shRNA) knockdown of FOXO, we found that the transcription factor in necessary during embryonic differentiation. The lack of FOXO reduced osteogenic differentiation and phenocopied HRTP-mediated differentiation defects. Future studies will determine if the knockdown of FOXO during hESC-osteogenic differentiation results in high levels of ROS. FOXOs may have a key role in toxicant-induced skeletal defects.

RNA-sequencing has allowed for the global profiling of all the RNA within a cell.

Global profiling of embryonic development can play and instrumentntal role in identifying abberant pathways driving embryotoxicity. In this study, RNA-sequencing globally identified HRTP-responsive genes. There were more downregulated genes than upregulated genes, and a number of the identified altered genes are known to play a role in osteogenesis. We uncovered that Camel Blue and STE have different pathways to elicit osteogenic inhibition. Of all the differentially regulated genes identified, only 12 genes were shared between the two HRTPs. Gene ontology further demonstrated that different biological processes affects between the two HRTPs. Interestingly, STE had downregulated genes associated to paraxial mesoderm, known for axial bone development.

Preliminary ongoing studies, in our lab have discovered rib malformations in mouse pups exposed to STE in utero. Camel Blue showed downregulation of growth factors receptors crucial to bone development and defects in DNA machinery such as nucleosome assembly.

Further studies are needed to confirm the RNA-seq findings but while the end result of osteoblast inhibition is the same between the two HRTPs, the pathway misregulated is not the same.

194

The overall obejective here was to determine transcriptional regulators that are responsive to HRTP exposure that can result in skeletal malformations. FOXOs play a part in multiple cell processes from ROS detoxification, cell proliferation, DNA repair, apoptosis, and having a supportive role in osteoblast differentiation through the expression of FOXO target genes. Upstream effectors of FOXO, Phosphatidylinositol 3-kinases

(PI3K)-AKT pathway blocks FOXO target gene expression. Growth factors trigger the

PI3K-AKT pathway and allowing for AKT to phosphorylate FOXO proteins at three conserved residues. The AKT-mediated phosphorylation of FOXO results in the nuclear exclusion of FOXO and, therefore, the suppression of FOXO target genes (Martins et al.,

2016). Consequently, any compound that can act upon the pathway has the potential to perturb cellular processes and create unwanted effects. Based on findings from FOXO protein, mRNA levels, and RNA-seq data, that upon HRTP exposure FOXO protein levels may have been reduced through AKT-mediated phosphorylation that resulted in the inability to counteract ROS and disrupted osteogenesis. Overall, HRTPs may have caused differential activation of pathways that controlled transcriptional regulators that inhibited skeletal differentiation (Fig 7.1).

We have successfully established an in vitro model to study the effects of tobacco exposure on the developing skeleton using human embryonic stem cells differentiating into osteoblasts. We found that FOXO is required for proper osteogenic development.

Importantly, small changes in gene expression over a collective of genes can result in defects in differentiation. Global gene analysis identified HRTP-sensitive genes that led to perturbation of skeletal development. All together, these findings provide much needed

195 valuable insights in understanding the pathology of skeletal defects that can be used by global initiatives such as the Center for Disease Control and Prevention and the Tobacco-

Related Disease Research Program to help lead in the prevention of women smoking while pregnant and smoking around infants and young children.

196

Figure 7.1. Harm-reduction tobacco products transcriptional regulation during osteogenesis. A schematic of the working hypothesis of HRTP-mediated FOXO regulation. Chemcals that target receptor tyrosine kinases have the potential to control cellular processes (i.e., proliferation, apoptosis, protein synthesis, and metabolism). Specifically, receptor tyrosine kinase stimulation occurs through the interaction with growth factors. The upregulation of FGF suggests the activation of the PI3K-AKT signaling pathway and, consequently, the exclusion of FOXO (dashed line) from the nucleus facilitated by AKT phosphorylation. Furthermore, decreased nuclear FOXO resulted in the cell’s inability to remove intracellular ROS and inhbit osteogeneis. Validation of RTKs targeted by HRTPs to regulate FOXO is necessary in further understanding of skeletal development dysregulation. PI3K, Phosphatidylinositol 3-kinases; RTKs, receptor tyrosine kinases. Types of RTKs: FGFR, fibroblast growth factor receptor; EGFR, epidermal growth factor receptor; IGF1R, insulin-like growth factor 1 receptor; PDGFR, platelet-derived growth factor receptor; and VEGFR, vascular endothelial growth factor receptor.

197

APPENDIX

198

APPENDIX 1

Chapter 2: Supplemental information

Appendix Figure 1.2.1. Pilot differentiation studies for protocol development. To determine whether VD3 can induce osteogenesis in human PSCs as it does in mouse ESCs [6], dissociated hESC clusters were first forced to aggregate and form embryoid bodies (EBs), an approach commonly used for mouse and non-human primate ESCs [5, 11, 12, 46]. Human ESCs formed EBs when in the maintenance medium, mTeSR (A). However, upon enzymatic dissociation for single cell replating, in DMEM +20% FBS, an essential step during mouse osteogenic induction [47], low cell attachment and survival was noted. (B) Assays for ALP activity and calcium deposition in H9 hESCs differentiated according to the schematic shown in Fig. 2A show that VD3 outperforms DEX, n=5 independent samples ± SD, *p<0.05 Student’s t test. DEX, dexamethasone; VD3, 1α,25 dihydroxy vitamin D3.

199

Appendix Figure 1.2.2. Kinetic and passage-specific analysis of additional gene expression associated with osteogenesis and early lineage determination. (A-D) mRNA expression of early and late osteogenic genes was determined with qPCR and normalized to GAPDH, n=3±SD. RIV9 and RIV4 hiPSCs failed to show definitive osteogenic patterns. *p<0.05 passage II over I; #p<0.05 passage III over I; Δp<0.05 passage III over II, One-way ANOVA. (E) ALP activity supported findings for ALP mRNA expression, (n=5±SD). *p<0.05 passage II over I; #p<0.05 passage III over I; Δp<0.05 passage III over II, One-way ANOVA. (F) Inorganic phosphate in hPSC cultures inoculated from three different passages (I-III). Data points represent 5 technical replicates for each biological passage shown ± SD. *p<0.05 passage II over I; #p<0.05 passage III over I; Δp<0.05 passage III over II. ALP, alkaline phosphatase; Runx2, Runt-related transcription factor 2; OPN, osteopontin; GAPDH, glyceraldehyde 3-phosphate dehydrogenase.

200

Appendix Figure 1.2.3. Inter-and intraline variability between different passages of hPSCs used for osteogenic differentiation. Data points were normalized between different passages of hPSC lines and expressed as n-fold over H9 per time point of analysis. N-fold regulations were then analyzed in box plots. This revealed that the passage-to-passage variability between the three tested lines is similar for most endpoints.

201

Appendix Figure 1.2.4. Early lineage determining genes are differentially expressed in the three hPSC lines. Messenger expression for genes associated with mesoderm, neural crest and neuroepithelium was quantitatively analyzed with qPCR, on n=3 independent samples ± SD from cells harvested during differentiation. *p<0.05 One-way ANOVA over H9. ZIC1 was low in RIV9 cells, but not statistically different than in H9 cells. GBX2 showed a trend similar to the other neural crest genes, but expression in RIV9 was not statistically different than in H9 cells. NTE5was highly expressed in RIV9 cells. Although selected as it encodes CD73, an antigen often used to isolate mesenchymal stem cells and therefore expected to associate with a cell that specifies into the mesenchymal lineage from the neural crest, its expression in the RIV9 cultures may simply determine the undergoing mesenchymal commitment of mesodermal cells. Similarly, DLX5 and DLX6 were investigated as they represent markers of neural crest osteoprogenitors. However, their high expression in RIV9 cells, which showed little expression of neural crest genes, may be interpreted as marking all osteoprogenitor cells independently of lineage heritage.

202

Appendix Figure 1.2.5. Correlation between PAX7 and TWIST1 methylation status and osteogenic differentiation yield across hPSC lines. (A) Methylation-specific PCR was performed across multiple different hESC and hiPSC lines. The result indicated varying degrees of methylation at the TWIST1 and PAX7 promoters independently of cell type. (B- E) The linkage between PAX7 and TWIST1 and methylation state was further explored in RIV1 and RIV7 cells, which represent two hiPSC lines with abnormal karyotypes. RIV1 hiPSCs exhibited trisomy 12, which is a common aberration acquired by PSCs in culture [61] associated with a competitive growth advantage. While chromosome 12 codes for multiple pro-neural crest and pro-osteogenic genes, such as SP7, WNT1, WNT10b, IGF1 and VDR, RIV1 cells had incorporated 71 µg of Ca2+ per mg protein by day 20. Although slightly more than the 44 µg in RIV9 cells this was barely similar to the 90 µg measured in H9 cells possibly due to the two promoters showing patterns more similar to RIV9 cells. RIV7 cells showed an abnormal karyotype with trisomy 5 in 35% of the cells and additional isolated occurrence of trisomy 12 and an isochromosome of the short arm of chromosome 3. The PAX7 and TWIST1 methylation patterns in this line were more like the neural crest conducive H9 methylation patterns and thus these cells displayed enhanced calcification. M, methylated; U, unmethylated.

203

Gene Forward Primer Reverse Primer Ta ALP 5’-CCACGTCTTCACATTTGGTG-3’ 5’-AGACTGCGCCTTGGTAGTTGT-3’ 60 GAPDH 5’-GAGTCAACGGATTTGGTCGT-3’ 5’-TTGATTTTGGAGGGATCTCG-3’ 60 NANOG 5’-GATTTGTGGGCCTGAAGAAA-3’ 5’-AAGTGGGTTGTTTGCCTTTG-3’ 60 OCN 5’-GGCAGCGAGGTAGTGAAGAG-3’ 5’-CTGGAGAGGAGCAGAACTGG-3’ 60 OPN 5’-GAAACGAGTCAGCTGGATG-3’ 5’-TGAAATTCATGGCTGTGGAA-3’ 60 PAX7 5’-CACTGTGACCGAAGCACTGT-3’ 5’-GTCAGGTTCCGACTCCACAT-3’ 60 POU5F1 5’-GTACTCCTCGGTCCCTTTCC-3’ 5’-CAAAAACCCTGGCACAAACT-3’ 60 SOX1 5’-TGTTGGCATCTAGGTCTTGGC-3’ 5’-TTTGAATTGGAGCGACACTGC-3’ 60 T/BRA 5’-AAGAAGGAAATGCAGCCTCA-3’ 5’-TACTGCAGGTGTGAGCAAGG-3’ 60 TWIST1 5’-ACTGGCCTGCAAAACCATA-3’ 5’-TGCATTTTACCATGGGTCCT-3’ 60 RUNX2 5’-TTTGCACTGGGTCATGTGTT-3’ 5’-TGGCTGCATTGAAAAGACTG-3’ 60 SATB2 5’-GCAGTTAAGCCAGAGCCAAC-3’ 5’-GTTGTCGGTGTCGAGGTTTT-3’ ZIC1 5’-GTCCTACACGCATCCCAGTT-3’ 5’-GCGATAAGGAGCTTGTGGTC-3’ 60 Round 1 of TWIST1 Methylation-Specific nested PCR amplification TWIST1 methylated 5’-CGTTAGGGTTCGGGGGGGC-3’ 5’-ACCTAACCCGAACGACCG-3’ 58 TWIST1 unmethylated 5’-GTTGTGTTAGGGTTTGGGGGT-3’ 5’-ACCTAACCCGAACGACCG-3’ 58 Round 2 of TWIST1 Methylation-Specific nested PCR amplification TWIST1 methylated 5’-GGGGCGTTGTTCGTATGTTTC-3’ 5’-CACCCGCCTCCTAACCG-3’ 58 TWIST1 unmethylated 5’-TTTGGGGGTGTTGTTTGTATGTTTT-3’ 5’-TCCACCCACCTCCTAACCA-3’ 58 PAX7 Methylation-Specific PCR 5’-TTTTATGAATAAAAGAAAAGTTCGA- 5’-CTAAACGAAAAAAACAAAAAATAACG- PAX7 methylated 60 3’ 3’ 5’-TTTTATGAATAAAAGAAAAGTTTGA- PAX7 unmethylated 5’-AACAAAAAAAACAAAAAATAACAAA-3’ 60 3’

Appendix Table 1.2.1. Oligonucleotide sequences used in this study. Ta, annealing temperature.

204

Chapter 4: Supplemental Information

Appendix Figure 1.4.1. Testing of three control chemicals in the hESC-ESTo. (A) Photomicrographs of unstained and stained H9 hESC osteogenic cultures on day 30 of differentiation. Bar = 100 µm. (B- D) Concentration response curves for PenG, 5FU and atRA. (B) mg Ca2+ per mg protein measured in hESC osteogenic cultures after chemical exposure normalized to solvent controls. (C) Cell viability on hESCs measured with MTT assay. (D) Cell viability of exposed human foreskin fibroblasts (HFF). Values are representative of five technical replicates ± SD for each biological replicate, *p<0.05 One- Way ANOVA. hESC, human embryonic stem cell; HFF, human foreskin fibroblast; PenG, penicillin G; 5FU, 5-fluorouracil; atRA, all-trans Retinoic acid.

205

Appendix Figure 1.4.2. The effects of 13cisRA on somatic cells. (A) Endpoints in the hESC-ESTo: Human fibroblast (HFF) viability determined by MTT. (B) Endpoints in the mESTo: cell viability tested on mouse 3T3 fibroblasts. n=5±SD for each experimental replicate, *p<0.05 One-Way ANOVA.

206

Appendix Figure 1.4.3. TWIST1 mRNA expression as endpoint in the hESC-ESTo. (A) Effects of tested compounds on the expression of the osteogenic progenitor gene TWIST1 on day 10 of differentiation determined with qPCR and normalized to GAPDH, n=3±SD. *p<0.05 One-Way ANOVA over untreated control. (B) Correlation between half-maximal inhibitory doses found for cell viability (IC50) and differentiation (ID50Ca) with human ESCs and iPSCs. H9 hESC, human embryonic stem cell; RIV9 hiPSC, human induced pluripotent stem cell; PenG, penicillin G; 5FU, 5-fluorouracil; 13cisRA, 13-cis Retinoic acid; atRA, all-trans Retinoic acid.

207

APPENDIX 2

Chapter 6: List of differentially expressed genes

Camel Blue UP genes

RPL7AP30 NOP10 PFDN4 MRPS21 POLR2H RPL34 EIF1B MGST3 MTX2 RPRM MRPL27 CD9 RNF181 TIMMDC1 PLGRKT RP11-452N17.1 LINC01405 PGLS ECI1 RPS20P10 RP11-734E19.1 ATP5A1P2 CTSL WFDC2 EDF1 MRPS33 RPL24P4 SRP9 FAUP1 MFSD3 RPL7AP6 DGUOK TCEAL9 UQCRFS1 PMVK SDF2L1 POLR2K INO80B,WBP1 DYNLT1 SF3B6 TPRKB PTGES2-AS1 SF3B5 MRPL12,SLC25A10 DUSP23 AGPAT2 BEX3 S100A6 RPS19P1 PFN1P1 HNRNPLP1 SPAG7 CLPTM1L SAT1 RP11-298C3.2 ZCCHC9 PRDX5 RPL35P5 CTB-33G10.1 ZNF593 RAB13 CETN3 RPS12 COX5A GAMT RBX1 UQCR10 CIR1 RPS4XP11 RGS16 ID1 C9orf142 RP11-641D5.1 CCDC71L COX7A2 PSMB6 HDHD5 MT2A TPM2 GAR1 SNRPG FDX1 TMEM101 RPL6P27 LINC00998 TMEM54 ARL6IP4,OGFOD2 BIK NUDT1 RP11-114H7.1 CKS2 RPP25L RWDD1 RPL5P17 RIDA RP1-159A19.3 ISOC2 C19orf43 TXN RP1-278E11.3 BLOC1S4 ATP5E BLVRB RPL34P18 PYCARD TACSTD2 FTH1P8 TRMT112 RPS7 RPS27AP16 ABRACL METTL26 NDUFS5 RPS15 HSD17B10 TMEM150C PGF NDUFV2P1 MDK CLDND2,ETFB GTF3C6 NDUFA12 RPL36AL PITHD1 CHCHD2 C12orf57,RNU7-1 MCRIP1 CIB1 RPS20,SNORD54 SOD1 C20orf27 CYBA FTLP3 RAC3 RP3-340B19.2 PSMA7 SDHAF1 PIR ATP5LP2 HMGB2 AC007969.5 NDUFA8 HSPB1 ATPIF1 SNORD50,VPS29 NMRK2 DRAP1 COX5B CCS ILF3-AS1 RBMX2 GPX1 SSBP1,TAS2R6P

208

MRPS34 RP11-92K2.2 SCGB3A2 LGALS1 NAT14 PDCD5 RP1-182O16.1 TIMM10 GAL RPS20P14 LINC01315 SCAND1 RARRES2 CKS1B FIS1 SAC3D1 DPY30,MEMO1 RPS7P10 NDUFAB1 MED10 COMTD1 AP2S1 NDUFA13,YJEFN3 SFT2D1 FAM96A RPL37P2 TNNT1 KCTD14,NDUFC2 C20orf96 UNC50 HEBP1 PLA2G2A DNPH1 MAMDC4,PHPT1 HMGN3 RPS21 RP11-864N7.2 RPL7AP50 ADAM15,DCST1 MRPL54 RPL27 DPP7 SNRPEP4 COA6 DPM3 LINC01356 PSMG3 NDUFB1 IFI6 ATP5I SEC61G AURKAIP1 ID3 RP11-220D10.1 EEF1B2P2 CHCHD1 RPL35AP21 SNHG5,SNORD50B RPL24P2 CCDC34 HPN DNAAF3,TNNI3 FOXD3-AS1 H2AFJ C1orf35 TCF15 RP11-234N17.1 RPL34P27 RPL18P11 SNRPFP1 RP11-365H23.1 RP11-112J1.1 COMMD6 FGF17 PLAC9 EBPL MIR1307,USMG5 MINOS1P2 MRPL52 NDUFB4P12 HSD17B14 RPL24P8 MIR4784,MZT2A SNHG19 GADD45GIP1 RABAC1 DAXX RPS27A SIVA1 LGALS16 CCDC167 UBL5 TMEM145 RPL23AP65 SEC61B MIB2 METRN UQCRQ RPL17P36 DBI DDRGK1 FBXL15 NDUFS7 PCSK1N DPPA5 RP11-466H18.1 CLEC11A TPGS1 SC22CB-1E7.1 NDUFAF8 DGCR6 COPS9 LSM5 GADD45G FKBP2 FUT5,NDUFA11 IL32,RNU1-125P RPL24 ALKBH7 HOPX UPF3B MRPL41 LUM RPL22L1 TMEM219 CGA RPS21P4 LAMTOR4 AC073043.1 LAMTOR2 PLSCR3,TMEM256 HOXB-AS1,AS3 NDUFS6 PCMTD2 NR2F2-AS1 TMEM208 TMA7 SNRPD2 UQCRH

209

Camel Blue DOWN genes

SNORD3A PRH1,PRR4 NOMO2 RMRP LINC00632 CLASP1 7SK AATBC ZFX RN7SL3 C10orf90 RP11-791G16.4 HIST1H1B CCT6P3,RP11-460N20.3 PRLR HIST1H3B LINC01719 HK2P1 PTCH2,RNU5E-6P RP11-520B13.8 NSD1 HIST1H1E MIRLET7D,MIRLET7F1 MAPK6PS4 HIST1H1A ZNF221 BRWD1 SCARNA2 NPIPB3,SMG1P3 OSBPL10 HIST1H2BE IL6STP1 MED13L HIST1H4B ZNF460,ZNF543 VEZF1P1 HIST1H2AI FMN1 RAPGEF5 HIST1H2AH IGF1R MAML2 HIST1H2BM MIAT,MIATNB RNF169 HIST1H4D USP37 EGFR HIST1H2BL KCNQ1OT1 BNC2 HIST1H2AD,HIST1H3D FAT3 AC092835.2 HIST1H3A SNORD88B,SNORD88C FBN2 HIST1H4H SERPINI2,WDR49 RP3-425P12.4 HIST2H2BF INO80D ZDHHC20P4 HIST1H2AL,HIST1H2BN LPCAT1 ZNF43 HIST1H2BC SNHG15,SNORA9 RP3-514P16.1 TERC HERC2 ATRNL1 RCC1,SNHG3,SNORA73 RP11-547D13.1 BMPR2 RNU2-2P,WDR74 LPP PPFIBP1 HIST1H4C PTPRD IPO8P1,KCNH1 HIST1H2AM,HIST1H3J LCOR OPRD1 RNU12 42802 PARP14 HIST4H4,RP11-174G6.5 RBL1 STXBP5 HIST1H3H SLC8A1 CDC42BPA HIST1H2AG TNKS ZNF780A,ZNF780B HIST1H2BD FGD4 TAF1L CH507-513H4.5 RBFOX2 ACADM,RABGGTB SNHG17,SNORA71 MGA MIR4523,TAOK1 COL15A1 AFF2 ALPK3 SNORD118,TMEM107 HELZ RUNX1T1 SNORD3B-2 PWRN1 HIPK3 HIST3H2BB TRPS1 PEAK1 CTD-2134A5.3 RP11-95M15.2 ASXL2 HIST1H2BK,HIST1H4I SNORA11,TRO RPL18A,SNORA68 HIST1H1C HELB ATP11A HIST1H2APS4,HIST1H3G RP11-815J4.6 NOTCH2 CDKL1 DCHS1 WDR82P2 HIST1H4J TMEM132B CDON HIST2H2BE ARNT2,MIR5572 TNPO1P2 MIR4687,STIM1 GNAQP1 ZNF124 AC007362.3 BRWD3 TJP1 TXLNB PTCH1 FGD6 RP11-1100L3.4 TBX10 TLK1P1 SLC5A9 ZNF10,ZNF268 AC073046.25,TET3 HIST1H2AC RP11-296E7.1 ITGB8 RP11-544A12.4 SHANK2 DOCK1 UCA1 PHKA1P1 EXPH5 USP3-AS1 EPPK1 KIAA1549 GRIA1 AFF4 GTF2I MIR6516,SCARNA16 PPARGC1A PRICKLE2 CACNA2D4 EP300,MIR1281 SPATA5 RP11-310E22.4 CNOT6LP1 PPP1R9A STARD13-AS AC009487.6 MSNP1 ZNF66 PPM1L ZNF469

210

CTD-2651B20.1,SHF CLIC4P1 LTN1 RP11-30P6.1 EIF4BP7 TEK WDFY3 AFAP1 PRKCA CECR2 GABPAP KIF21B WDR82P1 RFX7 DSP RP1-102E24.6 AGAP1 MAP3K9 CLIC4P3 RIF1 CYP51A1P2 SCDP1 EXOC5P1 FNDC3B TNRC6B HAUS6P1 TEX15 RP11-428P16.2 GAPVD1 ADCY9 PHC1P1 YWHAZP3 CHRDL1 MTR ARHGAP5 ZNF507 PPARGC1B BCLAF1P2 DIP2B BBC3,MIR3191 LMTK2 ZNF609 ZKSCAN8 CTB-52I2.4 AKAP11 NHLRC2 LSS TLK2P1 RP13-88F20.1 ZNF557 FAM60CP CIT,MIR1178 INSR EIF4HP1 KIAA0355 ZDBF2 CDCP1 AGO1 USP13 LRCH1 FREM2 FOXK1 CHST3 POLR2A PDE4D BCORL1 XYLT1 DNAJC6,RNU2-15P CDK12 RREB1 UBN2 ATP12A LIFR,MIR3650 TET2 EIF2S3L EIF3LP1 KIRREL EPG5 WNK3 MAP1A RP4-775C13.1 NRIP1 FLNC ARHGEF12 PTPN23 API5P1 DANT1 ARHGAP32 EIF4BP3 ATP9A NF2 HIST3H2A HSPA8P8 FYCO1 PARP4 FAM131B AC063961.1,KIAA1217 ANKRD50 GABRQ LINC00694,ZNF445 SETD1B GLTSCR1L SOGA1 TBC1D16 SPRY4 CDKL5 SYT14 BCL9 SNORA40,SSH1 IQGAP3 BEND3 INTS6P1 ZBTB40 YBX1P10 ZNF732 HIPK2 PSAT1P3 AC004893.11,TRRAP BZW1P1 NPTXR SLX4 CYP51A1P1 RLF PPATP1 KIAA1671 ZBED4 KHSRPP1 ROCK2 SCAF4 GRID2 AC006026.9

211

Camel Snus Tobacco Extract UP Genes

KDR RP11-472N13.2 HLTF LINC01374 HPN RNU6-1274P KLF9 EEF1A1P39 CHP2 RP11-483F11.7 LINC00698 RP1-305G21.1 ACTN3 RP11-313I2.6 RP11-324H9.1 RN7SL519P ADRA2A RP11-321F8.4 TXNIP RP11-124N19.4 ATRX RP11-594C13.2 PPFIA4 RNU1-47P MIR922,RUBCN RP11-457D20.1 MIGA1,RNA5SP21 CTD-2520I13.1 PAX3 LA16c-349E11.1 CCDC180,RP11-23J9.4 RN7SL525P ICAM5 Y_RNA ITGB2 SFTPB C15orf52,RNA5SP392 RN7SKP185 PLCZ1 RP11-674E16.1 PRDM7 RNU6-922P MEPE RP13-1016M1.2 FAM129C RN7SL218P RP11-553L6.5,ZBTB20 RN7SL802P DRAIC CTD-2201E18.5 RP11-88I18.2 RP3-380B8.4 RNU6-871P RP11-15J23.1 LINC00639 HIST1H2AA ARL11 RP5-874C20.7 LINC00271 DUTP5 RP5-1182A14.5 SNORA64 NDUFS5P3 KB-1090H4.2 U1 SUMO2P18 RP11-15I11.2 NOP56P2 RNVU1-17 RN7SL338P RP11-284L19.1 RP11-92C4.3

212

Camel Snus Tobacco Extract DOWN Genes

AC008277.1 MAP1A CH507-513H4.5 LINC01225,LINC01226 RP11-415J8.3 REEP1 RNU5E-1 STS BCRP3,KB-1995A5.4 FAM43A SP5 TBX3 MIR4687,STIM1 SMPDL3A DOCK3,MIR4787 CAPN6 MSRB3 MSX2 CDIP1 CDH5 RP1-225E12.2,TXLNB PDE4D SNHG15,SNORA9 ANKRD1 RNU4-2 ZFHX3 CER1 GATA6 CCKBR GATA2 RNU4-1 PGF FAM89A,MIR1182 RPL5P34 T GREM2 MRGPRX1 IFI6 RNY1 RHOU ZRANB2-AS2 TET2 HSD3B1 ZNF703 PRR9 FTH1P7 RNVU1-7 HSPB8 LINC02095,SOX9-AS1 CTSV FYB S100A16 EOMES CSRP1 CDX1 C1orf115 OLFML1 CLDN4 MIXL1 KRT19P1 MGLL RPL37AP8 RP11-384K6.6 H2AFY KRT80 KRT8P45 BBC3,MIR3191 TACSTD2 CDX2 EFNB2 GCM1 KRT18P27 HAND1 RP5-827C21.1 PLXNA2 RP2 KANK4 CDC42EP1 U1 S100A11 LRRN2 RHOB TIMP3 HMOX1 WLS MYL6P3 RNU6ATAC PRH1,PRR4,TAS2R12 KCNK13 RP11-794A8.1 SNORA11,TRO RNU5B-1 VIPR1-AS1 RP11-525G12.1 VGLL1 HIST1H2BA DIO3 RP11-89N17.2 VIT LYPD1 SP6 CEBPA CGA FAM26E ACADM,SNORD45 ADGRG1 MUC15 PLEKHA6 NODAL

213