Insights Into the Mechanism of Action of 13-Cis Retinoic Acid

The Pennsylvania State University

The Graduate School

College of Medicine

INSIGHTS INTO THE MECHANISM OF ACTION OF 13-CIS RETINOIC ACID

IN SUPPRESSING SEBACEOUS GLAND FUNCTION

A Thesis in

Molecular Medicine

Amanda Marie Nelson

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

May 2007

The thesis of Amanda Marie Nelson was reviewed and approved* by the following:

Diane M. Thiboutot Professor of Dermatology Thesis Advisor Chair of Committee

Gary A. Clawson Professor of Pathology, Biochemistry and Molecular Biology

Mark Kester Distinguished Professor of Pharmacology

Jeffrey M. Peters Associate Professor of Molecular Toxicology

Jong K. Yun Assistant Professor of Pharmacology

Craig Meyers Professor of Microbiology and Immunology Co-chair, Molecular Medicine Graduate Degree Program

*Signatures are on file in the Graduate School

iii ABSTRACT

Nearly 40-50 million people of all races and ages in the United States have acne, making it the most common skin disease. Although acne is not a serious health threat, severe acne can lead to disfiguring, permanent scarring; increased anxiety; and depression. Isotretinoin (13-cis Retinoic Acid) is the most potent agent that affects each of the pathogenic features of acne: 1) follicular hyperkeratinization, 2) the activity of Propionibacterium acnes, 3) inflammation and 4) increased sebum production. Isotretinoin has been on the market since 1982 and even though it has been prescribed for 25 years, extensive studies into its molecular mechanism of action in human skin and sebaceous glands have not been done. Since isotretinoin is a teratogen, there is a clear need for safe and effective alternative therapeutic agents. The studies undertaken in this thesis were designed to increase our understanding of the effects of 13-cis RA on the sebaceous gland and its mechanism of action in sebum suppression. It is well established that isotretinoin drastically reduces the size and lipid secretion of sebaceous glands. We hypothesized that isotretinoin decreases the size of the sebaceous gland by inducing cell cycle arrest and/or apoptosis and that sebum suppression is most likely an indirect result of the reduced size of the sebaceous gland. Our studies show that 13-cis RA, unlike 9-cis RA or ATRA, induces cell cycle arrest and apoptosis in SEB-1 sebocytes. Its ability to induce apoptosis is not inhibited in the presence of functional retinoic acid receptor (RAR) pan antagonist AGN 193198, suggesting an RAR- independent mechanism of apoptosis. Gene expression analysis was performed in cultured SEB-1 sebocytes that were treated with 13-cis RA and in biopsies of skin taken from patients that were treated for one week with isotretinoin. These data indicate that 13-cis RA increases expression of neutrophil gelatinase associated lipocalin (NGAL) and Tumor Necrosis Factor related apoptosis inducing ligand (TRAIL). In turn, we report that both NGAL and TRAIL induce apoptosis within SEB-1 sebocytes and, as such, are potential mediators of 13-cis RA induced apoptosis in human sebocytes. These studies into the mechanism of action of 13-cis RA in sebaceous glands suggest that 13-cis RA mediates its sebosuppressive effect through preferential induction of apoptosis in sebaceous glands. Furthermore, these data provide a rationale for drug discovery of alternative agents that are capable of selectively inducing apoptosis in sebaceous glands as a treatment for severe acne. iv TABLE OF CONTENTS

LIST OF FIGURES ...... ix

LIST OF TABLES...... xvi

LIST OF ABBREVIATIONS ...... xvii

ACKNOWLEDGEMENTS ...... xx

Chapter 1 Literature Review ...... 1

1.1 Introduction ...... 1 1.2 Sebaceous gland anatomy and physiology ...... 1 1.2.1 Skin...... 2 1.2.2 Anatomy of the Sebaceous Gland ...... 3 1.2.2.1 Histology ...... 3 1.2.2.2 Location ...... 4 1.2.2.3 Embryogenesis and Morphogenesis ...... 5 1.2.2.4 Physiology of the Sebaceous Gland: Holocrine Secretion ...... 6 1.2.2.5 Lipid Composition of Sebum...... 7 1.2.2.6 Function of Sebum...... 7 1.2.3 Regulation of sebaceous gland size and sebum production...... 8 1.3 Acne...... 10 1.3.1 Epidemiology ...... 10 1.3.2 Pathophysiology ...... 11 1.3.3 Classifications of acne lesions...... 13 1.3.4 Model systems for acne research-animal ...... 14 1.3.4.1 Rat preputial gland...... 14 1.3.4.2 Hamster flank organ and ear ...... 15 1.3.4.3 Rhino mouse...... 16 1.3.5 Models for acne research: isolated human sebaceous gland organ culture...... 17 1.3.6 Model systems for acne research: sebocyte cell culture ...... 18 1.3.7 Current Treatments for Acne ...... 19 1.3.7.1 Cleansers: follicular hyperkeratinization ...... 19 1.3.7.2 Antibiotics: P. acnes and inflammation ...... 20 1.3.7.3 Hormonal therapy: sebum suppression ...... 20 1.3.7.4 Topical Retinoids: inflammation, follicular hyperkeratinization..22 1.3.7.5 Oral Retinoid: Isotretinoin (Accutane®, 13-cis Retinoic Acid)....22 1.4 Retinoids...... 24 1.4.1 Retinoid Biology...... 25 1.4.1.1 What are retinoids?...... 25 1.4.1.2 Retinoid receptors...... 26 1.4.1.3 Retinoid binding proteins and retinoid metabolizing enzymes..28 1.4.1.4 Retinoid Function...... 30 1.4.1.4.1 Proliferation...... 30 1.4.1.4.2 Differentiation...... 32 1.4.1.4.3 Apoptosis...... 33 1.5 Retinoids in dermatology ...... 36 v 1.5.1 Significance of research project...... 36

Chapter 2 13-cis Retinoic Acid Induces Apoptosis and Cell Cycle Arrest in Human SEB- 1 Sebocytes ...... 37

2.1 Chapter Abstract...... 37 2.2 Introduction ...... 38 2.3 Results...... 39 2.3.1 13-cis RA exhibits a more rapid onset of growth inhibition of SEB-1 sebocytes compared to 9-cis RA and ATRA...... 39 2.3.2 13-cis RA significantly inhibits DNA synthesis in SEB-1 sebocytes....41 2.3.3 13-cis RA, but not 9-cis RA or ATRA, increases p21 levels in SEB-1 sebocytes...... 43 2.3.4 13-cis RA, but not 9-cis RA or ATRA, decreases cyclin D1 protein in SEB- 1 sebocytes...... 45 2.3.5 13-cis RA induces apoptosis in SEB-1 sebocytes but not in HaCaT keratinocytes or NHEK...... 45 2.3.6 13-cis RA specifically increases levels of cleaved caspase 3 in SEB-1 sebocytes...... 47 2.3.7 13-cis RA, but not 9-cis RA or ATRA, increases TUNEL staining in SEB-1 sebocytes...... 49 2.3.8 Apoptosis induction by 13-cis RA in SEB-1 sebocytes is not blocked by RAR antagonist AGN 193109...... 52 2.3.9 13-cis RA is isomerized to ATRA over time in SEB-1 sebocytes...... 52 2.4 Discussion...... 55 2.5 Materials and Methods...... 59 2.5.1 Cell Culture ...... 59 2.5.2 Effects of retinoids on SEB-1 proliferation ...... 60 2.5.3 3H thymidine incorporation assay ...... 60 2.5.4 Western blot analysis for p21, cyclin D1 and cleaved caspase 3 ...... 61 2.5.5 Annexin V-FITC/Propidium Iodide FACS Apoptosis Assay ...... 62 2.5.6 TdT-Mediated dUTP Nick End Labeling (TUNEL) Staining ...... 63 2.5.7 Quantitative Polymerase Chain Reaction (QPCR) ...... 64 2.5.8 HPLC ...... 64

Chapter 3 Array profiling of skin from patients on isotretinoin provides insights into potential mediators of its apoptotic effect on sebaceous glands...... 66

3.1 Chapter Abstract...... 66 3.2 Introduction ...... 67 3.3 Results...... 68 3.3.1 Patient selection and procedures...... 68 3.3.2 Histology reveals statistically significant decrease in sebaceous gland size after 8 weeks of treatment...... 69 3.3.3 Significant decreases in genes that regulate lipid metabolism were noted in the gene array expression analysis of skin biopsies taken from patients at 8 weeks into isotretinoin therapy...... 71 3.3.4 Gene expression analysis of skin from patients treated with 13-cis RA for one-week revealed significant increases in genes encoding calcium binding vi proteins, retinoid signaling molecules, solute carriers and serine proteases……………………………………………………………………..73 3.3.5 Gene expression analysis in SEB-1 sebocytes and HaCaT keratinocytes with 72 hour 13-cis RA treatment...... 75 3.3.6 QPCR verification of select genes from array analyses...... 78 3.3.7 Cluster Analysis ...... 80 3.3.8 Functional categorization of significantly changed genes...... 82 3.3.9 Promoter analysis of genes ...... 84 3.3.10 Comparison of gene changes at one-week and 8-week revealed only 3 common genes...... 85 3.3.11 Comparisons between one-week, SEB-1 sebocytes and HaCaT keratinocytes array data revealed only one gene in common between all three arrays...... 85 3.3.12 Immunohistochemistry and western analysis showed increased NGAL expression after 13-cis RA treatment in patient skin and our cell lines, respectively...... 87 3.3.13 Isotretinoin increased apoptosis in one-week patient sections...... 89 3.3.14 Purified NGAL protein induces apoptosis in SEB-1 sebocytes but not in HaCaT or NHEK keratinocytes...... 90 3.3.15 Apoptosis induced by NGAL is mediated by specific NGAL receptor isoforms...... 91 3.4 Discussion...... 92 3.5 Materials and Methods...... 98 3.5.1 Patient selection and tissue biopsies...... 98 3.5.2 Image analysis of sebaceous gland size ...... 99 3.5.3 Cell Culture ...... 99 3.5.4 Gene expression microarray analysis...... 100 3.5.5 Quantitative real-time polymerase chain reaction (QPCR)...... 100 3.5.6 Cluster Analysis ...... 101 3.5.7 Database promoter analysis of genes whose expression was significantly changed by 13-cis RA...... 101 3.5.8 Comparisons of gene expression arrays ...... 102 3.5.9 NGAL immunohistochemistry ...... 102 3.5.10 Western blotting...... 103 3.5.11 TdT-mediated dUTP Nick End Labeling (TUNEL) staining...... 103

Chapter 4 Mechanisms involved in induction of apoptosis in SEB-1 sebocytes:

Activation of the extrinsic death receptor pathway by Tumor Necrosis Factor related apoptosis inducing ligand (TRAIL) ...... 105

4.1 Chapter Abstract...... 105 4.2 Introduction ...... 106 4.3 Results...... 107 4.3.1 13-cis RA up-regulates genes involved in apoptosis in SEB-1 sebocytes. …………………………………………………………………………………107 4.3.2 13-cis RA increases cleaved caspase 8 to a greater extent than 9-cis RA or ATRA...... 108 4.3.3 13-cis RA increases TRAIL expression in SEB-1 sebocytes...... 109 4.3.4 TRAIL increases levels of cleaved caspase 3 in SEB-1 sebocytes ....111 vii 4.3.5 siRNA knockdown of TRAIL inhibits activation of caspase 3 by 13-cis RA ……………………………………………………………………………...…112 4.4 Discussion...... 116 4.5 Materials and Methods...... 120 4.5.1 Reagents ...... 120 4.5.2 Quantitative Polymerase Chain Reaction (QPCR) ...... 120 4.5.3 siRNA knockdown of TRAIL by nucleofection...... 121 4.5.4 Western blot analysis for TRAIL and cleaved caspase 3...... 122

Chapter 5 Development and Characterization of a Temperature Sensitive Sebocyte Cell Line (TSS-1)...... 123

5.1 Chapter Abstract...... 123 5.2 Introduction ...... 124 5.3 Results...... 124 5.3.1 Temperature Sensitive Sebocytes (TSS) persist in culture ...... 124 5.3.2 TSS sebocytes display a differentiated phenotype: Increased intracellular lipid and expression of the androgen receptor and the melanocortin 5 receptor...... 127 5.3.3 TSS sebocytes enter senescence following prolonged incubation at ‘restrictive’ temperatures...... 129 5.3.4 Total lipogenesis is increased at elevated temperatures in TSS sebocytes...... 131 5.3.5 Synthetic androgen R1881 increases and 13-cis RA decreases lipids in TSS-1 sebocytes ...... 134 5.3.6 13-cis RA decreases TSS-1 proliferation...... 135 5.3.7 13-cis RA induces apoptosis in TSS-1 sebocytes at the restrictive temperature...... 136 5.4 Discussion...... 138 5.5 Materials and Methods...... 142 5.5.1 Sebocyte Culture ...... 142 5.5.2 Establishment of TSS Sebocytes and Individual Clonal Cell Lines ....142 5.5.3 Cell Growth and Viability...... 143 5.5.4 Immunohistochemistry and Oil Red O Staining ...... 144 5.5.5 Western Analysis ...... 144 5.5.6 β-galactosidase Senescence Assay ...... 145 5.5.7 Lipogenesis Assay: 14C-actetate incorporation into neutral lipids ...... 145 5.5.8 TdT-Mediated dUTP Nick End Labeling (TUNEL) Staining ...... 146

Chapter 6 Discussion and Future Directions ...... 147

6.1 Introduction ...... 147 6.2 Rationale, hypothesis, and results of this work...... 147 6.3 Explanation of model ...... 148 6.4 Future Directions of this project ...... 150 6.4.1 Why is 13-cis RA superior to 9-cis RA or ATRA in the treatment of acne?...... 150 6.4.2 What RAR-independent events can lead to 13-cis RA-induced apoptosis?...... ……152 viii 6.4.3 Lipogenesis vs. Apoptosis: Does 13-cis RA preferentially affect one of these processes?...... 153 6.4.4 Does TRAIL mediate isotretinoin-induced apoptosis within the sebaceous gland? ...... 156 6.4.5 Why are the apoptotic effects of 13-cis RA limited to sebocytes and do not occur within keratinocytes?...... 158 6.5 Conclusion ...... 160

Appendix A Supplemental gene expression array tables ...... 161

A.1 All significantly changed genes after 8 weeks isotretinoin therapy...... 161 A.2 All significantly changed gene in SEB-1 sebocytes after 72 hours 13-cis RA treatment...... 178 A.3 All significantly changed genes in HaCaT keratinocytes after 72 hour 13-cis RA treatment...... 180

References…………………………………………………………………………………..182

ix LIST OF FIGURES

Figure 2: Hematoxylin and eosin stained longitudinal section of human sebaceous gland showing its multi-lobular structure...... 4

Figure 3: Signaling pathways and transcription factors involved in cell lineage determinations. As daughter cells migrate from the bulge region, changes in the expression patterns of numerous transcription factors determine their final cell lineage. Data is far from complete in this area; it is very likely that other pathways and transcription factors play a significant role in determining each cell lineage...... 6

Figure 4: Development of acne lesions Follicular hyperkeratinization, increased sebum production, active P. acnes bacteria, and inflammation all contribute to the development of acne. Diagram taken from “Fast Facts-Acne” 2004 Health Press Limited...... 12

Figure 5: Examples of the chemical structure of retinoids. Figure modified from Roos, T.C., Jugert, F. K., Merk, H. F., Bickers, D. R. Retinoid Metabolism in Skin. (1998) Pharmacolgical Reveiws 50(2): 315-333...... 26

Figure 6: Functions involving retinoids Figure taken from Napoli, JL. Biochemical pathways of retinoid transport, metabolism, and signal transduction. (1996) Clinical Immunology and Immunopathology. 80(3):S52-62...... 30

Figure 7: Simplified view of cell cycle control with focus on G1 and S phase of the cell cycle. Diagram from http://www.scielo.br/img/fbpe/rimtsp/v44n1/a07fig03.gif...... 31

Figure 8: Diagram of extrinsic and intrinsic apoptosis pathways. Type 1: extrinsic (death receptor) pathway. Type II: intrinsic (mitochondrial) pathway. Figure modified from I. Petak, J.A. Houghton. Pathology Oncology Research, Vol 7(2), 95-106, 2001...... 34

Figure 9: 13-cis RA, 9-cis RA and ATRA differentially inhibit SEB-1 sebocyte proliferation. (a-c) Time-dependent inhibition of SEB-1 sebocyte proliferation with individual retinoid compounds. SEB-1 cells were cultured in the presence of ethanol vehicle alone (0.01% or less; control), 0.1 µM, 0.5 µM or 1 µM concentrations of 13- cis RA, 9-cis RA, ATRA for 24, 48 or 72 hours. Attached cells were collected, stained with trypan blue, and counted manually by hemacytometer. Data represent mean ± SEM, n = 12. Statistical analysis was performed by ANOVA Two-Factor with Replication. * p < 0.05, **p < 0.01, *** p < 0.0001...... 40

Figure 10: 13-cis RA inhibits DNA synthesis to a greater extent than 9-cis RA or ATRA. (a-c) SEB-1 sebocytes were treated with ethanol vehicle (0.01% or less; control) or 0.1, 0.5, 1 µM concentrations of 13-cis RA, 9-cis RA or ATRA for 24, 48 x or 72 hours. 1µCi 3H thymidine was added to each sample 8 hours prior to collection. Cells were washed and collected for liquid scintillation counting. Data represent mean ± SEM, n ≥12. Statistical analysis was performed by ANOVA Two- Factor with Replication. *p < 0.005 and **p < 0.01...... 42

Figure 11: 13-cis RA increases p21 and decreases cyclin D1 proteins. (a) SEB-1 cells were treated with 0.1 µM, 1µM, 10 µM 13-cis RA or vehicle. (b-c) Parallel experiments were performed with 0.1 µM 0.5 µM, or 1 µM concentrations of 9-cis RA and ATRA. Blots were incubated with primary antibodies to p21 and β-actin for loading control normalization and analyzed by densitometry. (d) SEB-1 cells were treated with 0.1 µM, 1µM, 10 µM 13-cis RA or vehicle and blots were incubated with primary antibodies to cyclin D1 and β-actin. Magic Mark XP (MM) indicates band size. Blots are representative of a minimum of three western blots. Graphs represent normalized values relative to vehicle (control) expression of a minimum of three independent western blots. Mean ± SEM. * p < 0.05 ** p = 0.01...... 44

Figure 13: 13-cis RA induces cleaved caspase 3 expression in SEB-1 sebocytes. (a) SEB-1 sebocytes were treated with vehicle, 0.1 µM, 1 µM, or 10 µM 13-cis RA. (b-c) Parallel experiments were performed with 0.1 µM, 0.5 µM, or 1 µM concentrations of 9-cis RA or ATRA. Blots were incubated with primary antibodies to cleaved caspase 3 (1:1000) and actin (1:1000) for loading control normalization and analyzed by densitometry. p17 and p19 are cleaved caspase 3 active fragments. Blots are representative of a minimum of 4 independent experiments. Graph represents normalized values relative to vehicle (control) expression for 4 independent western blots. Data represent mean ± SEM * p < 0.01...... 48

Figure 14: Cleaved caspase 3 is not detected in NHEK treated with 13-cis RA. NHEK were treated with vehicle, 0.1 µM, 0.5 µM, or 1 µM concentrations of 13-cis RA or 1 µM staurosporine (S; positive control). Blots were incubated with primary antibodies to cleaved caspase 3 (1:1000) and β-actin (1:1000) for loading control normalization and analyzed by densitometry. Representative blot is shown...... 49

Figure 15: The increase in TUNEL staining with 13-cis RA is not inhibited in the presence of RAR pan antagonist AGN 193109. (a) Representative images of control, 0.1 µM, 1 µM, and 10 µM 13-cis RA, 9-cis RA, ATRA, and fenretinide treatments at 72 hours. (48 hour data not shown) (b) Quantification of the percentage of TUNEL positive stained cells per treatment at 48 and 72 hours. (9-cis RA not shown) Data represent mean + SEM, n = 6-12. Statistical analyses were performed with ANOVA Two Factor with Replication. * p < 0.01 ** p < 0.001 (c) xi Representative images of negative control, 1 µM 13-cis RA, AGN 193109, and 13- cis RA combined with 10 µM AGN 193109 at 72 hours. (48 hour data not shown) (d) Quantification of the percentage of TUNEL positive cells at 72 hours. Data represent mean + SEM, n = 12. Statistical analyses were performed with ANOVA Two Factor with Replication. * p < 0.05 when compared to control; + not statistically different. (e) QPCR verification of RAR antagonist AGN 193109 activity in SEB-1 sebocytes. Bars represent the efficiency corrected normalized average fold change of TIG1 under the experimental conditions as determined by REST-XL software. n = 4 ...... 51

Figure 16: 13-cis RA is isomerized to ATRA within SEB-1 sebocytes. HPLC analysis of (a) SEB-1 medium alone, (b) medium removed from SEB-1 sebocyte-containing plates and (c) SEB-1 sebocytes after 5 µM 13-cis RA treatment for the indicated times. Points are the average of duplicate samples...... 54

Figure 17: 13-cis RA decreases sebaceous gland volume. (a) Hematoxylin and eosin sections of back skin from patients before and after 8 weeks of treatment reveals a significant decrease in sebaceous gland volume. (b) Variable changes in sebaceous gland size were noted after one week of treatment compared to baseline biopsies (c) Area of sebaceous glands. Statistical significance was determined by paired t- test. Representative images are shown at a total magnification of 100X. Magnification bars = 250 µm...... 70

Figure 18: QPCR verification of gene array gene changes. (a) One-week (b) 8-week. Data represent the mean ± SEM of the fold change in gene expression as determined by REST-XL (QPCR) in 4-5 subjects compared to 6-8 subjects as determined by gene array. (c) SEB-1 (d) HaCaT (e) NHEK. Data represent the mean ± SEM of the fold change in gene expression as determined by REST-XL (QPCR) in 3 samples compared to 3 samples as determined by gene array.....79

Figure 19: Hierarchical clustering diagram of one-week isotretinoin patient samples. Hierarchical clustering was used to compute a dendrogram that assembled all genes and samples into a single tree. Patient samples included skin biopsies taken prior to treatment and at one-week of treatment. Normalized array data was imported into dChip software version 1.3. The information files for the Affymetrix human genome HG-U133A 2.0 array was obtained from www.dChip.org (8-week data not shown). Each row represents a single gene and each column represents a patient sample. (B=baseline and A=after treatment). The color reflects the level of expression when compared to the mean level of expression for the entire biopsy set. Red indicates expression higher than the mean and blue indicates lower expression than the mean...... 81

Figure 20: NGAL increased after one-week isotretinoin treatment. Immunohistochemistry for NGAL on sections on back skin taken before and at one- week treatment reveals notable increase in NGAL expression in sebaceous gland and hair follicle. Sections were incubated overnight with a 1:50 dilution of mouse monoclonal lipocalin 2/NGAL antibody. Negative control sections omitted primary antibody. All sections were counterstained with hematoxylin. NC=negative control; pre=before treatment; post=after treatment; SG=sebaceous gland; and Fol=follicle. Representative images are shown. Total magnification: 400X ...... 87 xii Figure 21: 13-cis RA increases NGAL protein expression. SEB-1 sebocytes, HaCaT keratinocytes and NHEK keratinocytes were treated with vehicle control or 13-cis RA (0.1 or 1 µM). Protein expression was verified by western blot. Blots were incubated with primary antibody to lipocalin 2 and β-actin for loading control normalization followed by densitometry. Graph represents normalized fold-change values relative to control expression for a minimum of six independent blots. Mean ± SEM...... 88

Figure 22: TUNEL staining in sebaceous glands increased in patient skin after one- week isotretinoin treatment. Two representative “after isotretinoin” images are shown. Skin sections were obtained from paraffin blocks of patients 9-15 and were subjected to TUNEL-peroxidase assay, according to manufacturer’s instructions. Assay controls included DNase I treated positive and negative controls with primary antibody omitted in negative control. At least 2 sections from every patient (before and after) were analyzed. Sections were counter-stained with hematoxylin. Data represent mean ± SD, n=6 patients; paired t-test was used for statistical analysis. Total magnification 400X...... 89

Figure 23: NGAL increases TUNEL staining in SEB-1 sebocytes SEB-1 sebocytes, HaCaT and NHEK keratinocytes were treated in duplicate with vehicle control, 1pg/mL, 10pg/mL, 1ng/mL and 10ng/mL purified recombinant human NGAL protein (R&D Systems) for 24 hours. (a) Representative images of SEB-1 sebocytes are shown. Total magnification 200X. (b) Quantification of the percentage of TUNEL positive stained cells per treatment at 24 hours. Data represent mean + SEM, n = 4- 6. Statistical analyses were performed with ANOVA Two Factor with Replication. * p< 0.05, ** p < 0.01, *** p < 0.0001...... 90

Figure 24: Cell-specific expression of 24p3R/NGAL-R isoforms is influenced by 13- cis RA. (a) Protein lysates (vehicle and 1µM 13-cis RA 48 hours) were immunoblotted with affinity purified 24p3 receptor antibody. Positive (+) and negative (-) control protein lysates obtained from Dr. Michael Green. Variable expression of receptor isoforms (short, long, and high molecular weight forms) are noted across cell lines and in response to 13-cis RA. (b) Relative quantification of isoforms. Blots were incubated with β-actin for loading control normalization followed by densitometry. Graph represents normalized fold-change values relative to control expression for three independent blots. Mean ± SD...... 92

Figure 25: Increased active caspase 8 with 13-cis RA treatment. Protein lysates from SEB-1 sebocytes treated with 1 µM concentrations of 13-cis RA, 9-cis RA and ATRA or vehicle control (0.01% ethanol) and 1 µM) for 48 hours were immunoblotted with mouse Caspase 8 antibody (Cell Signaling Technology). A representative blot is shown. Graph represents mean ± SD fold-change values normalized to control for 2 independent blots (13-cis RA) and 4 independent blots (9-cis RA and ATRA) ...... 109

Figure 26: TRAIL expression is increased by 13-cis RA treatment in SEB-1 sebocytes. (a) QPCR was performed for TRAIL at 24, 48 and 72 hours. Bars represent mean fold change of 3 independent samples as determined by REST-XL software. (b) TRAIL protein expression at 72 hours. Preliminary western blot xiii shown. Graph shows relative level of TRAIL protein for one experiment to date...... 111

Figure 27: TRAIL increases expression of cleaved caspase 3 protein. SEB-1 sebocytes were treated with increasing concentrations of purified recombinant human TRAIL (rhTRAIL) protein for 48 hours. (a) Representative blot is shown. (b) Graph represents normalized values relative to control expression of three independent western blots. Mean ± SD. * p < 0.05...... 112

Figure 28: QPCR shows siRNA knockdown of TRAIL mRNA siCONTROL and TRAIL siRNA (2 concentrations) were nucleofected into SEB-1 sebocytes using Amaxa Nucleofection Kit T and nucleofector device. Twenty-four hours later, 0.1 µM 13-cis RA was added to induce TRAIL expression. Total RNA was isolated at 24, 48 and 72 hours of 13-cis RA treatment. Graph represents fold-change in level of TRAIL mRNA for one sample...... 113

Figure 29: siRNA knockdown of TRAIL inhibits active caspase 3 protein expression. siCONTROL and TRAIL siRNA (2 concentrations) were nucleofected into SEB-1 sebocytes using Amaxa Nucleofection Kit T and nucleofector device. Twenty-four hours later, 0.1 µM 13-cis RA was added to induce TRAIL expression. Protein was isolated at 24, 48 and 72 hours of 13-cis RA treatment and subjected to immunoblotting with TRAIL (1:500) and cleaved caspase 3 (1:800) antibodies. Graphs represent normalized relative levels (compared to control) of TRAIL or cleaved caspase 3 protein of one sample...... 115

Figure 30: TSS sebocytes express SV40 large T antigen. TSS sebocytes (33ºC), SEB-1 sebocytes (37ºC) and HaCaT Keratinocytes (37ºC) were incubated with primary antibody to large T antigen and detected by anti-mouse, FITC secondary antibody. Cells were analyzed by fluorescence microscopy. Representative images are shown...... 125

Figure 31: SV40 large T antigen expression declines with increasing temperature within TSS sebocytes. TSS sebocytes were grown at 33ºC. Sebocytes then remained at 33ºC or were shifted to 37ºC or 41ºC for 72 hours to “shut-off” large T antigen protein. As a control, SEB-1 sebocytes were also grown under all three temperatures. Cells were incubated with primary antibody to large T antigen and detected by anti-mouse, FITC secondary antibody. Representative images are shown...... 126

Figure 32: TSS sebocyte growth declines with increasing temperature. TSS-1 sebocytes were plated and placed at 33ºC, 37ºC, 39ºC or 41ºC. Manual cells counts were performed every three days. Data points represent average of three independent samples at each time point...... 127

Figure 33: Oil Red O staining in TSS sebocytes increased with elevating temperatures. TSS and SEB-1 sebocytes were cultured at 33ºC, 37ºC and 41ºC temperatures f6r 6 days followed by O Red O staining to detect intracellular neutral lipids. Cells were counterstained with hematoxylin. Representative images are shown. Magnification: 400X...... 128 xiv Figure 34: Androgen receptor and melanocortin 5 receptor were expressed in differentiated TSS sebocyte cell lines. a) Total protein lysates from TSS-1, TSS- 2, TSS-3 and TSS-4 cell lines cultured at 33ºC, 37ºC and 39ºC for three days were analyzed for expression of androgen receptor (110kD) via western blotting. Representative blots are shown. b) Relative levels of androgen receptor expression. Mean ± SE, n= 3. c) Total protein lysates from TSS-1, TSS-2, TSS-3 and TSS-4 cell lines at cultured 33ºC, 37ºC and 39ºC were analyzed for expression of the melanocortin 5 receptor (32kD) via western blotting. Controls included lysates from SEB-1, rat preputial cells (Rat P.C.) and human placenta (Plac.) Magic Markers XP (M.M.) were used as size indicators...... 129

Figure 35: TSS-1 sebocytes entered senescence after prolonged incubation at 39C. TSS-1 sebocytes were cultured at 33ºC, 37ºC and 39ºC for 3, 5 or 7 days followed by β-galactosidase assay procedures. Representative images are shown. Magnification: 400X ...... 130

Figure 36: TSS-1 sebocyte culture and treatment model. This diagram outlines the timing and temperatures involved in using TSS-1 sebocytes. Experiments analyzing basal conditions are conducted on Day 5. Depending on the treatment length, assays are conducted on Days 6-9. In most cases, parallel plates at 33ºC and 39ºC are examined...... 131

Figure 37: TSS sebocyte lipogenesis was greatest at 37ºC incubation with little to no difference between 33 and 39ºC temperatures. TSS sebocytes were cultured and total lipogenesis was performed at 3 days (one-week data not shown) after temperature switch. Mean ± SEM; n = 6 samples. Statistical significance was determined by ANOVA Two Factor with Replication and considered significant if *p < 0.05...... 132

Figure 38: Incorporation of 14C acetate into lipids was greatest at 37ºC in TSS-1. TSS-1 sebocytes were cultured and lipogenesis assays were performed at 3 days (a) or one week (b) after temperature switch. Mean ± SD; n = 4 samples. C=cholesterol, FOH=fatty alcohol, OA=oleic acid, TAG=triglycerides, WE=wax esters, CO=cholesterol oleate, and SQ=squalene. Statistical significance was determined by paired t-test and considered significant if *p < 0.05. : All temperatures statistically different from each other...... 133

Figure 39: Synthetic androgen R1881 increased total lipogenesis in TSS-1 sebocytes. TSS-1 sebocytes were cultured as illustrated in Figure 36 and treated with R1881 (1 X 10-8 M) or vehicle alone (control) for 24 hours prior to lipogenesis assay. Assay was repeated three independent times. Mean ± SEM; n = 9. Statistical significance was determined by ANOVA Two Factor with Replication and considered significant if p < 0.05...... 134

Figure 40: 13-cis RA decreased total lipogenesis in TSS-1 sebocytes. TSS-1 sebocytes were cultured as illustrated in Figure 36 and treated with 13-cis RA (0.1 µM) or vehicle alone for 24 hours prior to lipogenesis assay. Assay was repeated three independent times. Mean ± SEM; n = 9. Statistical significance was determined by ANOVA Two Factor with Replication and considered significant if p < 0.05...... 135 xv Figure 41: 13-cis RA causes growth inhibition in TSS-1 sebocytes. (a) TSS-1 (33ºC) (b) TSS-1 (39ºC). Time-dependent inhibition of TSS-1 sebocyte proliferation. TSS-1 cells were cultured in the presence of ethanol vehicle alone (0.01% or less; control), 0.1 µM, 1 µM or 10 µM concentrations of 13-cis RA 24, 48 or 72 hours. Attached cells were collected, stained with trypan blue, and counted manually. Data represent mean ± SEM, n = 9. Statistical analysis was performed by ANOVA Two-Factor with Replication. * p < 0.05, **p < 0.01...... 136

Figure 42: TSS-1 sebocytes undergo apoptosis with 13-cis RA treatment: TUNEL Staining. Representative images of control, 0.1 µM, 1µM, and 10 µM 13-cis RA treatment at 48 and 72 hours at 39ºC. (b) Quantification of the percentage of TUNEL positive stained cells per treatment at 48 and 72 hours. Data represent mean + SEM, n = 6. Statistical analyses were performed with ANOVA Two Factor with Replication. * p < 0.05 ** p < 0.001 ...... 137

Figure 43: Model of 13-cis RA induces apoptosis and cell cycle arrest in SEB-1 sebocytes and human sebaceous glands...... 149

xvi LIST OF TABLES

Table 1: Comparison of animal models for studying sebaceous glands and acne.…17

Table 2: Comparisons of sebocyte cell culture models ...... 19

Table 3: Retinoid binding proteins with ligand identification. Table taken from Napoli, JL. Biochemical pathways of retinoid transport, metabolism, and signal transduction. (1996) Clinical Immunology and Immunopathology. 80(3):S52- 62………………………………………………………………………………………..29

Table 4: Retinoid induced apoptotic mechanisms...... 35

Table 5: Isotretinoin patient demographics ...... 69

Table 6: Selected significantly changed genes after 8 weeks isotretinoin therapy…72

Table 7: Significantly changed genes after 1 week isotretinoin therapy ...... 74

Table 8: Selected significantly changed gene in SEB-1 sebocytes after 13-cis RA treatment...... 76

Table 9: Selected significantly changed genes in HaCaT keratinocytes after 13-cis RA treatment ...... 77

Table 10: Functional categorization of significantly changed one-week genes.….82

Table 11: Protein domains enriched within genes significantly changed at 8 weeks...... 83

Table 12: Down-regulated pathways enriched within genes significantly changed at 8 weeks...... 84

Table 13: Common significantly changed genes within one-week isotretinoin, SEB- 1 sebocyte and HaCaT keratinocyte gene arrays...... 86

Table 14: Genes involved in apoptosis whose expression is significantly changed by 13-cis RA in SEB-1...... 107

xvii LIST OF ABBREVIATIONS

13-cis RA 13-cis retinoic acid 17β-HSD 17β hydroxysteroid dehydrogenase 3T3 albino Swiss mouse embryo fibroblast cell line 9-cis RA 9-cis retinoic acid ABC avidin:biotinylated enzyme complex ACTH adrenocorticotropic hormone AEC 3-amino-9-ethylcarbazole AGN 193109 RAR antagonist ANOVA analysis of variance APAF apoptotic protease activating factor AR androgen receptor Asebia sebaceous gland deficient mouse ATRA all-trans retinoic acid Bcl B cell lymphoma CD 3254 RXR pan-agonist CDK cyclin-dependent kinase CDKI cyclin-dependent kinase inhibitor cpm counts per minute CRABP cellular retinoic acid binding protein CYP cytochrome P450 DD death domain DED death effector domain DHEAS dehydroepiandrosterone sulfate DHT dihydrotestosterone DMEM Dulbecco's Modified Eagle Medium DMSO dimethyl sulfoxide DR direct repeat EDTA ethylenediamine tetraacetic acid EGF epidermal growth factor ER estrogen receptor ERK extracellular signal-related-kinase FACS fluorescent activated cell sorter FADD Fas-associated death domain FasL Fas Ligand FBS fetal bovine serum FDA Food and Drug Administration FDR false discovery rate FGFR fibroblast growth factor receptor FITC fluorescein isothiocyanate Gap1 growth phase 1 HaCaT human keratinocyte cell line H&E hematoxylin & eosin HPLC high performance liquid chromatography JNK c-Jun N terminal kinase xviii KGM keratinocyte growth medium MAPK mitogen-activated protein kinase mm millimeter MM or M.M. Magic Mark XP MSH melanocyte stimulating hormone NGAL neutrophil gelatinase associated lipocalin NHEK normal human epidermal keratinocytes P. acnes Propionibacterium acnes PBS phosphate buffered saline P.C. preputial cells PI propidium iodide PKC protein kinase C Plac. placenta PPAR peroxisome proliferator activated receptor QPCR quantitative real time polymerase chain reaction RA retinoic acid RALDH retinaldehyde dehydrogenase RAR retinoic acid receptor RARE retinoic acid receptor response element Rat P.C. rat preputial cells Rb retinoblastoma tumor suppressor RBP retinol binding protein REST Relative Expression Software Tool recombinant human tumor necrosis factor related apoptosis rhTRAIL inducing ligand RNA ribonucleic acid RNAi RNA interference ROR RAR-related orphan receptor RSV Rous Sarcoma Virus RT-PCR reverse transcription-polymerase chain reaction RXR rexinoid receptor, retinoid X receptor RXRE rexinoid receptor response element S staurosporine SAPK stress activated protein kinase SEB-1 SV-40 large T antigen immortalized sebocyte cell line SEM standard error of mean SERPINS serine protease inhibitors SG sebaceous gland Shh Sonic Hedgehog siRNA small interfering RNA SV40 simian virus 40 SZ95 SV-40 large T antigen immortalized sebocyte cell line TBP TATA binding protein Tcf3 transcription factor 3 TESS Transcription Element Search System TIG1 tazarotene-induced gene 1 TLR Toll-like receptor TNF tumor necrosis factor TNF-R tumor necrosis factor receptor xix TR thyroid receptor TRADD TNF-R1 associated death domain protein TRAIL tumor necrosis factor related apoptosis inducing ligand TSS temperature sensitive sebocyte cell line TUNEL TdT-Mediated dUTP Nick End Labeling VDR vitamin D receptor xx ACKNOWLEDGEMENTS

Words cannot express my gratitude for my thesis advisor and mentor, Dr. Diane Thiboutot M.D. I will forever by grateful to Diane for allowing me the freedom to explore and develop this project from the ground up. Throughout the entire thesis process, Diane was always supportive, encouraging, patient and positive. With her guidance, I was able to effectively learn from my mistakes, becoming more confident and trusting in my abilities as a budding scientist with each passing year. In addition to helping me succeed in science, she has shown me that balancing family and career is possible and just takes organization and flexibility. For all your support, I say ‘THANK YOU FOR EVERYTHING’. I look forward to our interactions in the years to come. The work presented in this thesis could not have been accomplished without the daily support of my co-workers and dear friends: Kathryn Gilliland, Zhaoyuan Cong, Kimberly Smith Dr. Heidi Devlin, and past members, Chelsea Billingsley and Dr. Terry Smith. I was warmly welcomed into the Thiboutot Lab by Kathy and Zhaoyuan, each of whom with smiles and lots of patience helped me get my research project started on the right foot. My daily conversations and interactions with all these wonderful people have positively influenced and challenged me to become the best scientist and person that I can be. For excellent scientific and technical advice, I must thank my thesis committee: Dr. Gary Clawson, Dr. Mark Kester, Dr. Jeffrey Peters and Dr. Jong Yun. These individuals always provided good advice and challenged me to think critically at every step of this project. I have truly enjoyed my time in graduate school at Penn State University College of Medicine. My graduate school classmates provided a listening ear and also the much needed, humor, to survive graduate school. In addition to classmates, I must thank Kathy Simon and Kathy Shuey, two of the best administrative professionals in the Graduate School. Both Kathy Simon and Kathy Shuey helped me tremendously with all the necessary paperwork from thesis committee appointments to graduation deadlines in addition to being available to help with any problem that arose during graduate school. Without them, I would still be lost. Finally, I would like to acknowledge the continual support from my family. To say thank you to my parents, Ken and Theresa Nelson, does not even begin to cover all they have done for me. Without a doubt, I would not be the person I am today without them. Brad, my loving husband, deserves a medal for his support throughout my graduate school career. He was xxi always there to encourage me and give me strength throughout these last 5 ½ years. I am so grateful to have him in my life.

Chapter 1

Literature Review

1.1 Introduction

Acne is one of the most common skin conditions encountered by dermatologists. Most acne occurs during adolescence, an already socially and psychologically challenging period in an individual’s life; although, it can persist into adulthood. There is no cure for acne, but successful treatments are available. The most effective drug for severe acne is isotretinoin. However, isotretinoin is a potent teratogen and as such, use of this drug is closely monitored through an FDA mandated registry program: iPLEDGE. Although isotretinoin has been prescribed for over 20 years, extensive studies into its molecular mechanism of action(s) in human skin and especially the sebaceous glands have not been done. By understanding the sebaceous gland and, specifically, the sebocyte response to isotretinoin in terms of changes in gene expression or cellular pathways, it may be possible for the development of alternative therapies for acne without the teratogenic side-effects. The first section of this chapter will review sebaceous gland anatomy and physiology. The second section will review epidemiological data on acne, the causes of acne, sebaceous gland model systems and current treatments of acne. Retinoid biology and functions are discussed within the final section of this chapter.

1.2 Sebaceous gland anatomy and physiology

Acne is a disease of the sebaceous gland. In order to understand the pathophysiology of this disease, it is important to understand the “normal” condition of the sebaceous gland. 2 1.2.1 Skin

The skin is the largest organ in the human body with an average area of 2 meters2 in adults. Its thickness ranges from 0.5mm to 4mm depending on body location with the thickest skin on the soles of the feet. Skin provides a physical and physiological barrier between the external environment and internal environment of the body. Skin has numerous functions including physical protection, wound healing, immune defense, sensory awareness, thermoregulation, secretion and permeability (Chuong et al, 2002). The skin is divided into two major components: epidermis and dermis. The epidermis is the most superficial layer of the skin and is composed of stratified squamous epithelium. Epidermal cells are called keratinocytes. The epidermis can be subdivided into 4-5 distinct layers. The deepest layer, stratum germinativium or stratum basale, is a single layer of proliferative cells that divide when necessary and also keeps the epidermis strongly attached to the underlying dermis. As cells migrate upwards into the stratum spinosum, stratum granulosum, and stratum lucidum layers, the cells are undergoing the differentiation process and acquiring new cytoskeletal framework, cell-cell connections, lipids and keratin proteins. Finally, fully differentiated cells reach the outermost layer of the skin, stratum corneum. This layer consists of numerous layers of flattened keratinocytes joined tightly together to form an impermeable layer. This top layer is constantly shed and under normal conditions, keratinocytes can make the journey from cell division to desquamation within a month (Blumenberg and Tomic-Canic, 1997). In addition to keratinocytes, the epidermis is composed of melanocytes, melanin producing cells responsible for our skin color; Langerhans cells, resident immune cells; and Merkel cells, which help with sensory perception (Kerr, 1999). The dermis is the deepest layer of our skin. It is composed of fibroblasts, abundant collagen and lesser amounts of elastic and reticular fibers. The dermis is strongly connected to the epidermis through hemidesmosomes and is the support system of the skin. The dermis contains blood and lymph vessels, glands (sweat and sebaceous), nerves (free and encapsulated nerve endings) and hair follicles (Figure 1). Blood vessels provide nutritional support as well as thermoregulation. Although located in the dermis, both sweat glands and sebaceous glands are derivatives of the epidermis and extend through the epidermis to the skin’s surface. Nerve endings receive and transmit information regarding temperature, pain, pressure and vibration (Kerr, 1999). 3

Beneath the dermis is a layer of subcutaneous tissue that is comprised of a thick layer of connective tissue and adipose tissue.

1.2.2 Anatomy of the Sebaceous Gland

1.2.2.1 Histology

Sebaceous glands, located in the dermis, are uni-lobular or multi-lobular structures that consist of acini connected to a common excretory duct composed of stratified squamous epithelium and are usually associated with a hair follicle. The glands are composed of lipid- producing sebocytes and keratinocytes that line the sebaceous ducts. Just inside the basement membrane at the periphery of the sebaceous gland is a basal cell layer composed of small, cuboidal, nucleated, highly mitotic sebocytes. Cells progress toward the middle of the gland and accumulate lipid droplets as they terminally differentiate. These fully differentiated sebocytes are 4 filled with lipid and lack all other cellular organelles. Surrounding the glands are connective tissue capsules composed of collagen fibers that provide physical support (Downie et al, 2004) (Figure 2).

Figure 2: Hematoxylin and eosin stained longitudinal section of human sebaceous gland showing its multi-lobular structure.

1.2.2.2 Location

Sebaceous glands are associated with hair follicles all over the body. A sebaceous gland associated with a hair follicle is termed a pilosebaceous unit. The glands may also be found in certain non-hairy sites including the eyelids (Meibomian glands), the nipples (Montgomery glands) and around the genitals (Tyson glands). Only the palms and soles, which have no hair follicles, are totally devoid of sebaceous glands. Sebaceous glands vary considerably in size, even in the same individual and in the same anatomic area. Most sebaceous glands are only a fraction of a millimeter in size. The largest glands and the greatest density of glands (up to 400- 900 glands per square centimeter) are found on the face and scalp (Downie, et al., 2004; Montagna and Parakkal, 1974; Strauss and Pochi, 1963). In the oral epithelium, sebaceous glands known as Fordyce spots are sometimes present. Fordyce spots are visible to the unaided eye because of their large size (up to 2 to 3 mm) and the transparency of the oral epithelium (Dreher and Grevers, 1995). In this location, the sebaceous ducts open directly to the surface. 5 1.2.2.3 Embryogenesis and Morphogenesis

In the human fetus, sebaceous glands develop in the 13th to 16th week of gestation from bulges (epithelial placodes) on the developing hair follicles (Muller et al, 1991; Williams et al, 1988). The bulge region of the follicle contains the epidermal stem cells that generate multiple cell lineages including epidermal and follicular keratinocytes as well as sebaceous glands. As daughter cells migrate from the bulge region, changes in the expression patterns of numerous transcription factors determine their final cell lineage. Multipotent progenitors of the bulge region express Tcf3 (transcription factor 3) and this expression represses terminally differentiated states including hair follicles and sebaceous glands (Nguyen et al, 2006). Wnt/wingless (Wnt) and Sonic Hedgehog (Shh) signaling pathways are intricately involved in embryonic patterning and cell fate decisions. Cells destined to become sebocytes have increased Shh and Myc signaling and decreased Wnt signaling (Figure 3). In transgenic mouse models, intact Wnt signaling promotes hair follicle differentiation, whereas inhibition of Wnt signaling by preventing Lef1/β-catenin interaction leads to sebocyte differentiation (Merrill et al, 2001). Loss-of-function and gain-of-function transgenic mouse models demonstrated that blocking Shh signaling inhibited normal sebocyte differentiation and constitutively activating Shh signaling resulted in increased number and size of sebaceous glands in skin (Allen et al, 2003). When fully formed, the glands remain attached to the hair follicles by a duct through which sebum flows into the follicular canal and eventually to the skin surface.

1.2.2.4 Physiology of the Sebaceous Gland: Holocrine Secretion

The sebaceous glands release lipids by disintegration of entire cells, a process known as holocrine secretion. The life span of a sebocyte from cell division to holocrine secretion is approximately 21-25 days (Plewig and Christophers, 1974; Plewig et al, 1971). Because of the constant state of renewal and secretion of the sebaceous gland, individual cells within the same gland are engaged in different metabolic activities dependent upon their differentiation state (Potter et al, 1979). The stages of this process are evident in the histology of the gland (Ito, 1984). The outermost cells, basal cell layer membrane, are small, nucleated, and devoid of lipid droplets. This layer contains the dividing cells that replenish the gland as cells are lost in the process of lipid excretion. As cells are displaced into the center of the gland, they begin to produce lipid, which accumulates in droplets. Eventually the cells become greatly distended with 7 lipid droplets and the nuclei and other subcellular structures disappear. As the cells approach the sebaceous duct, they disintegrate and release their contents. Less polar and more neutral lipids reach the skin surface (Nicolaides et al, 1970). Proteins, nucleic acids, and the membrane phospholipids are digested and apparently recycled during the disintegration of the cells.

1.2.2.5 Lipid Composition of Sebum

Human sebum contains cholesterol, cholesterol esters, squalene, wax esters, di- and triglycerides and fatty acids (Stewart and Downing, 1991). Squalene and wax esters are presumed to be unique to human sebum and distinguish sebum from the lipids of human internal organs, which contain minimal squalene and no wax esters. The particular fatty acid un- saturation patterns of the fatty acids in the triglycerides, wax esters, and cholesterol esters also distinguish human sebum from the lipids of other organs. The general desaturation pathway involves inserting a double bond between the ninth and tenth carbon of stearic acid (18:0) to produce oleic acid (18:1∆9). However, in human sebaceous glands, the predominant pattern is the insertion of a ∆6 double bond into palmitic acid (16:0) by delta-6 desaturase (fatty acid desaturase 2), which is the dominate fatty acid desaturase in sebaceous glands (Ge et al, 2003). The resulting product, sapienic acid (16:1∆6) is the major fatty acid of adult human sebum (Perisho et al, 1988). Elongation of the chain by two carbons and insertion of another double bond gives sebaleic acid (18:2∆5,8), which is unique to human sebum (Nicolaides, 1974).

1.2.2.6 Function of Sebum

The precise function of sebum in humans is unknown. It has been proposed that its solitary role is to cause acne (Cunliffe and Shuster, 1969). It has also been suggested that sebum reduces water loss from the skin’s surface and functions to keep skin soft and smooth, although evidence for these claims in humans is minimal. As demonstrated in the sebaceous gland deficient mouse (Asebia) model, glycerol derived from triglyceride hydrolysis in sebum is 8 critical for maintaining stratum corneum hydration (Flurh et al, 2003). Sebum has been shown to have mild antibacterial action, protecting the skin from infection by bacteria and fungi, because it contains immunoglobulin A, which is secreted from most exocrine glands (Gebhart et al, 1989). Vitamin E delivery to the upper layers of the skin protects the skin and its surface lipids from oxidation. Thus, sebum flow to the surface of the skin may provide the transit mechanism necessary for vitamin E to function (Thiele et al, 1999).

1.2.3 Regulation of sebaceous gland size and sebum production

Production of sebum is continuous and is not controlled by neural mechanisms, although substance P (neuropeptide) has been shown to increase proliferation and differentiation of sebaceous glands (Saint-Leger and Cohen, 1985; Thiboutot, 2004; Toyoda et al, 2002). The exact mechanisms underlying the regulation of human sebum production have not been defined. Clearly, sebaceous glands are regulated by androgens and retinoids, but recently, other factors, such as melanocortins, peroxisome proliferator-activated receptors (PPARs), and fibroblast growth factor receptors (FGFRs) have been postulated to play a role as well. It has long been recognized that sebaceous glands require androgenic stimulation to produce significant quantities of sebum. Androgen receptors have been localized to both the keratinocytes of the outer root sheath of hair follicles as well as the basal layer of the sebaceous gland (Kariya et al, 2005). Individuals with a genetic deficiency of androgen receptors (complete androgen insensitivity) have no detectable sebum secretion and do not develop acne (Imperato-McGinley et al, 1993). Conversely, addition of testosterone and dihydroepiandrosterone increases the size and secretion of sebaceous glands (Pochi and Strauss, 1969). There is still a question as to which androgen is physiologically significant. The most potent androgens are testosterone and dihydrotestosterone (DHT); however, levels of testosterone do not parallel the patterns of sebaceous gland activity. Sebum secretion starts to increase in children (5-6 years of age) during adrenarche although the levels of androgens are very low at this time (Pochi et al, 1977). It is possible that the sebaceous gland is responsive to these very low levels of androgens. In addition, testosterone levels are significantly higher in males than in females, with no overlap between the sexes, while average rates of sebum secretion are only slightly higher in males than in females, with considerable overlap between 9 the sexes (Harris, 1983; Thiboutot et al, 1999). The majority of females with acne have serum androgen levels that, although higher, are within normal limits and it has been hypothesized that locally-produced androgens within the sebaceous gland may contribute to acne (Levell, 1989; Lookingbill et al, 1985). The weak adrenal androgen, dehydroepiandrosterone sulfate (DHEAS), may regulate sebaceous gland activity through its conversion to testosterone and dihydrotestosterone within the sebaceous gland. The enzymes required to convert DHEAS to more potent androgens are present within sebaceous glands (Chen et al, 2002). The predominant isozymes in the sebaceous gland include the type 1 3β-hydroxysteroid dehydrogenase, the type 2 17β-hydroxysteroid dehydrogenase (17β-HSD) and the type 1 5α- reductase (Fritsch et al, 2001; Thiboutot et al, 1995; Thiboutot et al, 1998). Investigations into the influence of locally-produced androgens indicated that the activities of 5α-reductase and 17β-HSD enzymes within the sebaceous gland are not higher in male or female patients with acne compared to patient controls with no acne. Due to the small sample size, the influence of local androgen synthesis can not be ruled out (Thiboutot, et al., 1999). Clearly, androgens influence sebaceous glands and sebum production, although which androgens are important and the mechanism of their influence is not known. Melanocortins include melanocyte-stimulating hormone (MSH) and adrenocorticotropic hormone (ACTH). In rodents, melanocortins increase sebum production. Human primary sebocyte cultures treated with MSH have increased numbers of cytoplasmic lipid droplets (Zhang et al, 2003). Transgenic mice deficient in the melanocortin-5 receptor have hypoplastic sebaceous glands and reduced sebum production (Chen et al, 1997). The melanocortin-5 receptor has been identified in human sebaceous glands where it may play a role in the modulation of sebum production (Thiboutot et al, 2000). Further experimentation is required to test this hypothesis. Peroxisome proliferator activated receptors (PPARs) are orphan nuclear receptors that are similar to retinoid receptors in many ways. Each of these receptors form heterodimers with retinoid X receptors in order to regulate the transcription of genes involved in a variety of processes, including lipid metabolism and cellular proliferation and differentiation (Kim et al, 2001; Rosen et al, 1999; Schoonjans et al, 1996; Spiegelman et al, 1997). Rat preputial cells serve as a model for human sebocytes in the laboratory (Laurent et al, 1992). In rat preputial cells, agonists of the PPARα and PPARγ receptors induced lipid droplet formation in preputial sebocytes but not in epidermal cells while lineolic acid (PPARβ/δ agonist) induced lipid formation in both preputial sebocytes and epidermal cells (Rosenfield et al, 1999). Based on 10 the results from their studies, Rosenfield et. al. propose that PPARα activation plays a role in the beginning stages of lipogenesis, PPARβ/δ activation enhances the lipogenesis and PPARγ activation controls the transition to a more differentiated state complete with more lipid droplets within the cells, clearly identifying PPARs as a key player in sebocyte differentiation (Rosenfield, et al., 1999). Within human sebocytes, PPAR-α, -β/δ, and -γ receptor subtypes are expressed in basal sebocytes. PPAR-γ is also present in differentiated sebocytes (Chen et al, 2003; Downie et al, 2004; Trivedi et al, 2006). In patients receiving fibrates (PPAR-α agonists) for hyperliperdemia or thiazolidinediones (PPAR-γ agonists) for diabetes, sebum secretion rates are increased (Trivedi, et al., 2006) indicating that PPARs do play a role in sebocyte differentiation and maturation in humans. Fibroblast growth factor receptors 1 and 2 (FGFR1, 2) are expressed in the epidermis and skin appendages. Expression of FGFR3 and FGFR4 are localized to dermal vessels and microvessels and are notably absent in epidermis and appendages (Hughes, 1997). FGFR2 plays an important role during embroygenesis in skin formation (Li et al, 2001). Germline mutations in FGFR2 lead to Apert’s syndrome, which is commonly associated with acne. In addition, somatic mutations in the same location can lead to acne, but how this receptor is involved in sebaceous gland development and how its mutation leads to acne is unknown (Gilaberte et al, 2003; Munro and Wilkie, 1998). The regulation of human sebum production is complex. Advances are being made in this area which may lead to alternative therapies for the reduction of sebum and improvement in acne.

1.3 Acne

1.3.1 Epidemiology

Acne is one of the most prevalent skin conditions encountered by dermatologists, affecting nearly 85% of the people between the ages of 12 and 24 years including 40-50 million people in the United States each year (Cordrain et al, 2002; White, 1998). Although most 11 prevalent during adolescence, acne also affects infants, prepubescent children, and mature adults (Layton et al, 2004). Acne is not life-threatening; however, it does have a significant psychosocial impact. Embarrassment, low self-esteem, anxiety, anger, frustration, feelings of depression and social withdrawal may be associated with acne (Baldwin, 2002; Cunliffe, 1986; Dermatology, 2006). A meta-analysis performed by researchers on behalf of The American Academy of Dermatology and The Society of Investigative Dermatology demonstrated that the direct cost of acne (cystic and vulgaris) was $2.5 billion dollars in 2004. Furthermore, adding in the loss of productivity, unemployment and social impact of acne, the total cost of acne is estimated at $12 billion dollars per year in the United States (Bickers et al, 2006).

1.3.2 Pathophysiology

Acne is a disease of the pilosebaceous unit. Acne is the culmination of the interaction of 4 distinct factors: 1) increased sebum production 2) increased follicular hyperkeratinization 3) the activity of Propionibacterium acnes and 4) inflammation (Figure 4). The precise sequence of events leading to acne is unknown, that is, which factor begins the process is a mystery.

Very early studies demonstrated that sebum excretion correlates with the severity of acne, in both males and females (Cotterill et al, 1971). Sebum production begins slowly with only a few follicles exuding sebum onto the skin’s surface but, sebum production increases significantly with age and pubertal stage. Furthermore, children who developed acne had higher sebum productivity than those who did not develop acne (Mourelatos et al, 2007). Increased sebaceous lipogenesis and sebum excretion does correlate with severity of acne in male patients (Cooper et al, 1976). Sebum production is stimulated by androgens (Pochi and Strauss, 1969) and inhibited by 13-cis retinoic acid, corresponding to improvement in acne (Jones et al, 1980). Follicular hyperkeratinization plays a role in the pathophysiology of acne. How hyperkeratinization occurs and contributes to acne is not known, although studies have investigated proliferative states of keratinocytes and adhesion of keratinocytes. Using the Ki-67 antibody as an indicator of cell proliferation, Knaggs et. al. demonstrated increased Ki-67 staining in normal follicles from acne-affected skin when compared to non-acne-affected skin. Furthermore, cell proliferation in acne lesions (comedones) was increased compared to normal follicles (Knaggs et al, 1994). To assess whether hyperkeratinization results from abnormal 13 cohesion between keratinocytes, the distribution of desmosomal components in normal and acne patients was compared. No differences between these components was detected between acne lesions (non-inflammatory vs. inflammatory) and between normal epithelium from control or acne subjects (Knaggs et al, 1994). This study does not rule out the possibility that other factors may play a role in increased cohesiveness of keratinocytes. The predominant organism in the follicular flora is the gram positive, anaerobic, pleomorphic diphtheroid Propionibacterium acnes (P. acnes) although aerobic Staphylococcus epidermidis may also be present (Marples et al, 1974; Puhvel et al, 1975). The microenvironment within the sebaceous gland is anaerobic; therefore, favoring the survival of P. acnes bacteria over others. The P. acnes bacterium relies on sebaceous lipids as a nutrient source and breaks down triglycerides into free fatty acids (Gribbon et al, 1993). Free fatty acids within sebum can be irritating and contribute to the inflammatory response (Ro and Dawson, 2005). P. acnes plays an important role in the production of inflammatory acne by stimulating the classical and alternative complement pathways (Webster, 1979). P. acnes releases a variety of lytic enzymes and cytokines that are chemotactic for inflammatory cells (Webster, 1982; Webster, 1979). In addition, P. acnes lysates are capable of stimulating the production of both proinflammatory cytokines/chemokines and anti-microbial peptides from keratinocytes and cultured sebocytes (SZ95), indicating that keratinocytes and sebocytes, themselves, may play a role in the pathogenesis of acne (Graham et al, 2004; Nagy et al, 2006). Toll-like receptor (TLR) activation plays a role in innate immune response (Aderem and Ulevitch, 2000). Studies have demonstrated increased TLR-2 and TLR-4 expression within acne lesions and have shown that P. acnes induces TLR-2 and TLR-4 expression in cultured keratinocytes, suggesting that these receptors are involved in acne-induced inflammation (Jugeau et al, 2005). Many myths surround the development of acne. Acne is not caused by poor diet, bad hygiene, or greasy cosmetics. Acne may flare around the time of menses or in times of increased stress, and is probably the result of hormonal influences (Dermatology, 2006).

1.3.3 Classifications of acne lesions

Acne vulgaris is the most common type of acne and is characterized by both non- inflamed and inflamed lesions as well as scaring. Non-inflamed lesions, also known as 14 comedones, are either closed comedones (whiteheads) or open comedones (blackheads). Closed comedones are classified according to size: microcomedones (1-mm in diameter) or macrocomedones (>2-mm in diameter). Open comedones are of similar size and are black in color due to the oxidation of the skin pigment, melanin. Inflamed lesions can be papules, pustules, or nodules and are red to yellow in color, lasting from 1-3 weeks. Most patients have a mixture of non-inflamed and inflamed lesions (Layton, et al., 2004). Other forms of acne include acne conglobata, which is characterized by comedones, cysts, and abscesses, and acne fulmians, which is characterized by inflamed nodules, cysts and plaques.

1.3.4 Model systems for acne research-animal

Acne research is limited by the lack of animal models. The only mammal that exhibits acne is the human. Currently, no one animal model completely mimics all aspects of the human situation; however, there are a few models that have been used to study the different aspects of acne.

1.3.4.1 Rat preputial gland

The rat preputial gland is useful to study sebaceous gland biochemistry and regulation of sebum production. The preputial glands are exocrine glands and pairs of glands are located near the genitals in both male and female rats. The secretions of the preputial glands are believed to play a role in both mating behavior (secreting pheromones) and territory marking (Pietras, 1981; Thody and Dijkstra, 1978). The preputial gland is structured identically to human sebaceous glands in that acini are connected to a branching duct system and more lipid-laden cells are present in the center of the gland (Laurent, et al., 1992). Gland size and secretions are influenced by systemic administration of hormones (androgens, progesterones, estrogens), which is similar to human sebaceous glands (Alves et al, 1986; Ebling et al, 1971; Ebling et al, 1969; Thody and Dijkstra, 1978). 15 Single cell suspensions can be made by disruption of the rat preputial gland. These individual cells grow as a monolayer when co-cultured with 3T3 fibroblast cells and are like sebaceous cells in that they are slow growing and express K4, a sebaceous-specific keratin. However, under these conditions, preputial cells did not exhibit the characteristic decrease in proliferation when exposed to retinoic acid (Laurent, et al., 1992). Like the whole preputial gland, hormone (estrogen) stimulation increases lipogenesis in preputial cells (Alves, et al., 1986). The preputial cell model system has two clear disadvantages: cells are not responsive to androgens and the percentage of each individual lipid varies dramatically from those in human sebum (Alves et al, 1986; Rosenfield, et al., 1999; Thiboutot, 2004).

1.3.4.2 Hamster flank organ and ear

The hamster flank organ (costovertebral gland) is another model to study sebaceous gland biochemistry and regulation of sebum production. The paired flank organs are located on the back of golden Syrian hamsters. The organ contains sebaceous glands, hair follicles and melanocytes which is similar to human skin (Franz et al, 1989). Because the glands are large (roughly 6-8mm in diameter) and paired, they are extremely useful for investigations of topical compounds, with the second gland being used as a non-treatment control. Androgen stimulation increases the size of the flank organs while retinoid stimulation significantly decreases the size of the flank organs; a response identical to that seen in human sebaceous glands (Ferrari et al, 1978; Gomez, 1981). Not all compounds which inhibit the flank organ have the same effect in the human sebaceous gland and therefore the usefulness of this model may be limited (Franz, et al., 1989). The hamster ear model has been extensively used as a model for human sebaceous glands. The ventral surface of the hamster ear contains multiple sebaceous glands (size of gland varies by location within ear) (Matias and Orentreich, 1983). These sebaceous glands are very similar to human sebaceous glands in that they have similar turn-over rates and size (Plewig and Luderschmidt, 1977). In addition, hamster ear sebaceous glands are sensitive to androgens and retinoids (Geiger, 1995; Matias and Orentreich, 1983). However, not all retinoids which inhibit hamster sebaceous glands are effective in humans. For example, 16 isotretinoin and etritinate are both effective in the hamster ear but only isotretinoin is effective in humans (Geiger, 1995).

1.3.4.3 Rhino mouse

The rhino mouse model is useful for screening anti-keratinizing agents as well as comedolytic agents. This mouse model has a recessive mutation on chromosome 14 that results in a mouse devoid of body hair and having wrinkled skin. The hair follicles become detached from the underlying dermis and are no longer active, instead becoming filled with sloughing keratinocytes to create huge numbers of hornfilled utriculi which resemble comedones (Seiberg et al, 1997). Topical application of salicylic acid, lactic acid, and benzoyl peroxide had partial de-scaling effects. Application of retinoic acid completely reversed the excessive scaling and “normalized” the wrinkled phenotype (Kligman and Kligman, 1979).

17 Table 1: Comparison of animal models for studying sebaceous glands and acne. Animal Model System Useful to Study Advantages Disadvantages GLANDS: androgen sebum composition responsive, testing of is significantly Rat Preputial sebaceous gland, systemic compounds different from Glands/Cells sebum CELLS: grown as human sebaceous monolayer, express glands sebocyte markers similar morphology to human sebaceous glands, not all effective sebaceous gland, other gland can serve as compounds are Hamster Flank Organ sebum control, testing of topical effective in human compounds, androgen sebaceous glands and retinoid responsive multiple sebaceous glands per ear, other ear can not all effective sebaceous gland, serve as control, testing of compounds are Hamster Ear sebum topical compounds, effective in human androgen and retinoid sebaceous glands responsive skin contains huge numbers of horn-filled utriculi that resemble comedones, able to test Rhino Mouse follicular keratinization ? topical agents that affect differentiation and lose of cohesion between keratinocytes

1.3.5 Models for acne research: isolated human sebaceous gland organ culture

Human sebaceous glands can be isolated and maintained in culture for 7 days. Guy et. al. have demonstrated that whole sebaceous glands in culture maintain cell division and lipogenesis rates as in vivo for up to 7 days. In addition, 13-cis retinoic acid inhibits lipogenesis as it does in vivo; however, in this system, testosterone did not increase lipogenesis as one would predict based on the sebaceous glands’ in vivo response in previous studies (Guy et al, 1996; Pochi and Strauss, 1969). The difficulty with this model system is obtaining skin samples, dissection of sebaceous glands, and the limited amount of time for experimentation.

18 1.3.6 Model systems for acne research: sebocyte cell culture

Primary sebocytes are very difficult to grow and maintain in cell culture. In one method, the epidermal and dermal layers of the skin are separated and the dermis is scraped to obtain sebocytes which are co-cultured with 3T3 fibroblasts (Doran et al, 1991). In a second method, whole sebaceous glands are isolated from the surrounding tissue and placed in culture medium. Primary sebocytes are detected as out-growths from the sebaceous lobules (Abdel-Naser, 2004; Xia et al, 1989). These cells are not highly proliferative, grow in colonies, have limited sub-culturing capabilities, and express proteins characteristic of differentiated sebocytes even though they do not completely differentiate as they would in vivo (Xia, et al., 1989; Zouboulis et al, 1994; Zouboulis et al, 1991). The sebocytes undergo holocrine rupture long before there are sufficient numbers of cells for experimentation, which limits their usefulness. Despite the difficulties in obtaining ample numbers of primary sebocytes, they respond to retinoids with decreased proliferation and decreased lipogenesis (Zouboulis et al, 1991). Primary sebocyte proliferation is stimulated by androgens and is blocked by spironolactone, an androgen receptor blocker (Zouboulis et al, 1998). In order to circumvent the problem of “low yield” of primary sebocytes, immortalized human sebaceous cell lines have been created by Simian Virus 40 (SV40) large T antigen immortalization of primary sebocytes: SZ95 and SEB-1 (Thiboutot et al, 2003; Zouboulis et al, 1999). Both of these cell lines 1) express characteristics of differentiated sebocytes; 2) produce sebocyte specific lipids, wax esters and squalene; 3) are androgen responsive; and 4) exhibit inhibited proliferation in the presence of retinoids (Nelson et al, 2006; Thiboutot, et al., 2003; Zouboulis, et al., 1999). The major benefits of these cell lines are 1) enough cells can be obtained for repeated experimentation and 2) investigators can examine sebocyte-specific responses. 19 Table 2: Comparisons of sebocyte cell culture models Useful to Cell Culture Model System Advantages Disadvantages Study sebaceous gland maintains characteristics of low sebocyte Primary Sebocytes regulation, sebocytes in vivo, retinoid numbers, difficult to sebum and androgen responsive obtain production sufficient numbers of cells unable to completely sebaceous for experiments; terminally gland responsive to androgens, differentiate; SV40 Immortalized Sebocyte Cell regulation, retinoids; produce large T antigen Lines (SZ95, SEB-1) sebum triglycerides, wax esters interferes with production and squalene; express "normal" growth and sebocyte specific markers differentiation

1.3.7 Current Treatments for Acne

There are numerous over-the-counter soaps, washes and preparations as well as dermatologist-prescribed drugs available to treat mild to severe acne. For the purposes of this thesis, I will discuss the mechanism of action of each major category in the treatment of acne.

1.3.7.1 Cleansers: follicular hyperkeratinization

Body washes and facial cleansers containing hydroxy acids and benzoyl peroxide have anti-acne properties. These products are available as soaps, creams, gels, washes, lotions, scrubs and peels. α-hydroxy acids, including lactic acid and glycolic acid, are water soluble and penetrate the dermis. β-hydroxy acids, like salicylic acid, are lipid soluble and penetrate the upper epidermis and pilosebaceous unit. Both acid forms decrease cohesion amongst keratinocytes and cause exfoliation, helping remove the “keratinocyte plug” (Davies and Marks, 1976; Van Scott and Yu, 1984). Benzoyl peroxide works by an anti-bacterial effect in the treatment of acne, thereby decreasing the numbers of P. acnes. 20 1.3.7.2 Antibiotics: P. acnes and inflammation

Oral antibiotics function to reduce acne by inhibiting protein synthesis and proliferation of P. acnes. Trimethoprim inhibits the enzyme dihydrofolate reductase, resulting in the inhibition of tetrahydrofolic acid synthesis, a key precursor in DNA purine and pyrimidine synthesis. This inhibition interferes with P. acnes proliferation. Antibiotics of the tetracycline family inhibit protein synthesis by binding near the A-site on the 30S bacterial ribosomal subunit thus interfering with the binding of the amino-acid charged tRNA required for mRNA translation (Spahn and Prescott, 1996; Wirmer and Westhof, 2006). Additionally, tetracycline family members, doxycycline and minocycline, may exhibit some anti-inflammatory properties by decreasing each of the following: lipase production by P. acnes, cytokine production, white blood cell chemotaxis and activity of matrix metalloproteinases (Higaki et al, 2004; Leyden, 2001; Li et al, 2006). Topical antibiotics such as erythromycin and clindamycin are effective in reducing the number of inflammatory lesions and inhibiting P. acnes proliferation They are available in many formulations. Antibiotics within the macrolide family work similarly to tetracyline family members, except that they irreversibly bind the 50S ribosomal subunit inhibiting translocation of peptidyl tRNA and interfering with protein synthesis. Erythromycin and clindamycin may also be given orally to treat acne.

1.3.7.3 Hormonal therapy: sebum suppression

Hormonal therapy, which functions to decrease sebum production, is available for female patients and is particularly helpful in those with “acne flares” around the time of menses. Anti-androgens, oral contraceptives and glucocorticoids are all types of hormonal therapy (Thiboutot and Chen, 2003). Anti-androgens function as androgen receptor blockers and include spironolactone, flutamide and cyproterone acetate. Spironolactone (aldosterone antagonist) is a steroidal androgen receptor blocker which has been used for over 20 years for the treatment of acne and hirsutism (Shaw and White, 2002; Yemisci et al, 2005). This drug decreases sebum production and inhibits type 2 17β-hydroxysteroid dehydrogenase, thereby inhibiting the conversion of 21 androstenedione to testosterone (Tremblay et al, 1999; Zouboulis et al, 1994). Additional indirect mechanisms include the inhibition of 5α−reductase and elevation of steroid hormone binding globulin (Archer and Chang, 2004). Cyproterone acetate, a progestin, acts as an androgen receptor blocker and inhibits ovulation. Its effectiveness in the treatment of acne results from its ability to reduce sebum production and perhaps comedogenesis (Stewart et al, 1986). Although not approved in the U.S., it is approved in Canada, Europe, and Asia for the treatment of severe acne that is resistant to traditional therapy. Flutamide is a non-steroidal androgen receptor blocker effective in the treatment of prostate cancer (Harper, 2006). It is converted to a potent metabolite, 2-hydroxyflutamide, which inhibits the binding of dihydrotestosterone (DHT) to androgen receptors (Brogden and Clissold, 1989). Oral contraceptives are combinations of two agents: an estrogen (most commonly ethinyl estradiol) and a progestin. Oral contraceptives also suppress ovulation, which results in a decrease in the ovarian production of androgens and subsequent decrease sebum production. Within the sebaceous gland, estrogen receptors (ERα and ERβ) are expressed in both basal and differentiating sebocytes (Pelletier and Ren, 2004; Thornton et al, 2003). The most active estrogen is estradiol, which is produced from testosterone by the action of the enzyme aromatase. Aromatase is active in the ovary, adipose tissue and other peripheral tissues. Estradiol can be converted to the less potent estrogen, estrone, by the action of 17β hydroxysteroid dehydrogenase (17βHSD) enzyme. Both aromatase and 17βHSD are present in the skin (Hay et al, 1982; Sawaya and Price, 1997). Estrogens regulate cell activities by binding to and activating estrogen receptors. Like estrogens, progestins act through nuclear hormone receptors triggering activation or suppression of progesterone target genes leading to a cell-specific response. In the sebaceous gland, progesterone receptor B is expressed in both basal and differentiating sebocytes (Kariya, et al., 2005). First and second generation progestins can react with the androgen receptor, thus aggravating acne, hirsutism, and androgenic alopecia (Thiboutot, 2000). Newer third generation progestins are more selective for the progesterone receptor than the androgen receptor. In addition, progesterone has been shown to modulate pro-inflammatory and anti-inflammatory cytokines (Davies et al, 2004). Whether progesterone receptor activation can lead to decreased inflammation in sebaceous glands is not yet known. Currently in the United States, there are only two oral contraceptives approved for use in the treatment of acne: 35 µg ethinyl estradiol/norgestimate (Ortho Tri-Cyclen® Ortho, Raritan, NJ) or 20-35 µg ethinyl estradiol/norethindrone acetate (Estrostep®, Parke Davis, Detroit, MI). 22 Glucocorticoids block androgen production by the adrenal gland and prednisone is the preferred agent due to the increased risk of adrenal suppression with dexamethasone (Thiboutot and Chen, 2003).

1.3.7.4 Topical Retinoids: inflammation, follicular hyperkeratinization

Topical retinoids are key in the treatment of acne due to their ability to inhibit the formation of microcomedones. Tretinoin, tazarotene and adapalene are topical retinoids that are currently available for the treatment of acne and are useful for inflammatory and non- inflammatory acne. Tretinoin, the acid form of vitamin A (all-trans retinoic acid, (ATRA) Retin A®), is the original topical retinoid discovered by Albert Kligman in 1967 (Kligman et al, 1969) and is available in multiple formulations. It functions by binding retinoid nuclear receptors: retinoic acid receptors (RAR α,β,γ) and rexinoid receptors (RXR α,β, γ) and modulating gene expression. The mechanism of action of retinoids will be discussed in great detail in section 1.4. Tretinoin use is associated with skin irritation including redness, burning, and excessive dryness and peeling, which decreases with long-term use. Interestingly, tretinoin is also effective against wrinkles. Tazarotene (synthetic retinoid) and adapalene (naphthoic acid derivative), RARβ and RARγ agonists, are effective against inflammatory and non-inflammatory acne similar to tretinoin; although, the typical side effects are diminished.

1.3.7.5 Oral Retinoid: Isotretinoin (Accutane®, 13-cis Retinoic Acid)

Isotretinoin was originally prescribed for and effective in the treatment of lamellar ichthyosis, a hereditary disorder of keratinization causing excessive scaling of the skin (Peck and Yoder, 1976). Dr. Peck noticed a very positive side effect of this drug. In the subset of his patients with acne who were receiving isotretinoin for ichthyosis, their acne showed dramatic improvement (Peck, 1979). On this observation, isotretinoin was pushed into patient clinical trials for acne treatment. Isotretinoin (13-cis RA), marketed under the trade name of Accutane® by Hoffman-La Roche, was approved by the U.S. Food and Drug Administration (FDA) in 1982 for the 23 treatment of recalcitrant nodular cystic acne. 13-cis RA is prescribed in cases where acne is severe, non-responsive to other treatments, or when physical scarring is present. Patients receive doses of 0.5mg/kg/day-2mg/kg/day; with the dose depending on age, gender and body weight. The course of treatment lasts approximately 16-20 weeks. In rare cases, a second course of treatment is needed. Most patients report minor side effects including chapped lips, dry eyes, and generally dry skin. Isotretinoin is a potent teratogen and, as such, two forms of contraception and monthly negative pregnancy tests are required of females of child-bearing age. Children exposed to 13- cis RA in utero have severe hindbrain and forebrain malformations, limb malformations, craniofacial and cardiovascular defects. If the fetus is not spontaneously aborted, children who survive are afflicted with serious motor and sensory delays and mental retardation (Lammer et al, 1985; Lammer et al, 1985). In April 2002, Hoffman-La Roche implemented the “System to Manage Accutane Related Teratogenicity (S.M.A.R.T)” program, aimed at preventing pregnant women from receiving isotretinoin. (Roche-Laboratories, 2001) As of March 1, 2006, isotretinoin use is restricted by iPLEDGE, an FDA-mandated risk management program that seeks to limit fetal exposure to isotretinoin. This program requires registration of patients, prescribing dermatologists and pharmacists. For more information, please visit https://www.ipledge.com. Isotretinoin belongs to the class of drugs known as retinoids, which includes all naturally occurring and synthetic derivatives of vitamin A. In regards to acne, experimental evidence supports the fact that 13-cis RA can influence each of the factors involved in the pathogenesis of acne: 1) follicular hyperkeratinization, 2) bacterial colonization of the follicle, 3) inflammation and 4) sebum production. Isotretinoin decreases follicular keratinization by approximately 50%. (Plewig et al, 2004) How this is accomplished is unknown; although, it is known that isotretinoin does not affect the metabolic activity of the keratinocytes in the follicular duct epithelium or interfollicular epidermis (Dalziel et al, 1987). Isotretinoin therapy causes a significant reduction in the gram positive, anaerobic Propionibacterium acnes bacteria, including antibiotic resistant strains, with levels of bacteria slowly returning to baseline after discontinuing treatment (Coates et al, 2005; Leyden et al, 1986). It is not known how this reduction is achieved, however it may have a(n) 1) direct killing effect on P. acnes, 2) indirect effect by decreasing sebum production, thereby removing the food supply, or 3) increasing the host’s defense mechanisms. In support of the latter, 24 retinoic acid supplementation has been shown to “prime” the immune system to protect against bacterial lipopolysaccharide (LPS) challenges in rats (Seguin-Devaux et al, 2005). 13-cis RA has been shown to competitively inhibit the 3α-hydroxysteriod activity of retinol dehydrogenase, leading to decreased androgen synthesis (Karlsson et al, 2003). In addition, it inhibits the migration of polymorphonuclear leukocytes and monocytes into the skin, supporting its role in reducing the inflammation that is associated with acne (Norris et al, 1987; Wozel et al, 1991). The classical and alternative complement activation pathways are stimulated by P. acnes, possibly contributing to the inflammatory response (Webster, 1979). P. acnes releases a variety of lytic enzymes and pro-inflammatory substances that are chemotactic for inflammatory cells (Webster, 1979). With the reduction of P. acnes from isotretinoin treatment, inflammation is likely to diminish. The majority of studies have examined the sebosuppressive effect of isotretinoin. Yet, how this sebosuppression is achieved is poorly understood. It is well established that isotretinoin drastically reduces the size and lipid secretion of sebaceous glands in human and animal models, in in vitro cell cultures of human sebocytes and in immortalized sebocyte cell lines, SZ95 and SEB-1 (Goldstein et al, 1982; Gomez and Moskowitz, 1980; Landthaler et al, 1980; Nelson, et al., 2006; Strauss et al, 1980; Zouboulis, et al., 1991; Zouboulis, et al., 1999; Zouboulis et al, 1991). Processes such as cell cycle arrest or apoptosis may explain the histological data in human skin biopsies that demonstrate a drastic decrease in the size, shape, and lipid content of sebaceous glands after 16 weeks of treatment with isotretinoin (Goldstein, et al., 1982). It seems likely that the decrease in sebum production may be the net result of sebaceous gland involution in response to isotretinoin treatment rather than its direct target.

1.4 Retinoids

There are no less than 4000 currently published review articles on retinoid biology and retinoid therapy for various diseases. The field of retinoid research is so vast, that I will give an overview of retinoid biology followed by a description of retinoid effects. 25 1.4.1 Retinoid Biology

1.4.1.1 What are retinoids?

In the simplest definition: retinoids are vitamin A (retinol) or natural or synthetic vitamin A-like derivatives. Structurally, retinoid molecules consist of a cyclic end group, either cyclohexenyl ring or aromatic ring, attached to a polyene chain, and ending in a polar group. Examples of retinoid structures are shown in Figure 5. Derivatives of the oxidized form of retinol, retinoic acid, have been extensively synthesized and to date, over 5000 retinoid compounds have been produced (Dawson and Hobbs, 1994; Roos et al, 1998). Vitamin A and its retinoid derivatives are lipid soluble molecules that can readily transverse the lipid bilayer of the plasma membrane. The classic view of retinoid functionality is that retinoids affect cellular functions by binding nuclear hormone receptors and affecting gene expression (Giguere et al, 1987; Mangelsdorf et al, 1995). However, within the cancer field, more attention has been focused on retinoid receptor-independent effects within the cell, such as modulation of signal transduction kinase cascades including mitogen-activated protein kinase (MAPK) pathways: extracellular signal-related-kinases (ERKs), p38MAPK, and stress activated protein kinase/c-Jun N terminal kinase (SAPK/JNK) (Nakagawa et al, 2003; Olson and Hallahan, 2004; Pettersson et al, 2004). 26

1.4.1.2 Retinoid receptors

Retinoid receptors belong to the superfamily of nuclear hormone receptors. This superfamily is subdivided into two groups: steroid nuclear hormone receptors, which include androgen and estrogen receptors; or the non-steroid receptors, which include the thyroid, vitamin D and retinoid receptors (Mangelsdorf, et al., 1995). Retinoids exert their specific cellular effects through activation of retinoic acid or retinoid X nuclear receptors (RARs, RXRs). There are three isoforms of the RAR receptor designated RARα, RARβ and RARγ, encoded by three separate genes located on chromosomes 17p21.1, 3p24 and 12q17, respectively (Brand et al, 1988; Mattei et al, 1988). Like RARs, three isoforms of RXRs (RXRα, RXRβ and RXRγ) have been identified and each is encoded by separate a gene located on chromosomes 9q34.3, 27 6q21.3 and 1q22, respectively (Hoopes et al, 1992; Mangelsdorf et al, 1990; Yu et al, 1991). Regardless of the type of retinoid receptor, all have two distinct functional domains: a DNA binding domain and a ligand binding domain (Mangelsdorf, 1994; Reichel and Jacob, 1993). Retinoid receptors form heterodimers or homodimers and bind to cis-acting DNA elements in the genome. The retinoic acid receptor consensus sequence (RARE) is a direct repeat (DR) of the half-site AGTTCA separated by 2 or 5 nucleotides (DR-2, DR-5) and is recognized by RAR-RXR heterodimers. For RXR homodimers, the retinoid X receptor consensus sequence (RXRE) is a DR-1 of AGGTCA (Mangelsdorf, et al., 1995). RARs form heterodimers with RXR isoforms while RXRs are promiscuous forming heterodimers with other nuclear receptor family members including vitamin D receptor (VDR), peroxisome proliferator activated receptor (PPAR), and thyroid receptor (TR) in addition to forming homodimers with itself (Mangelsdorf, et al., 1995). Currently, experimental evidence supports the view that, in the absence of a retinoid ligand, dimerized receptors repress gene transcription by interacting with co-repressor molecules. Upon ligand binding, this repression is released and gene transcription occurs (Minucci and Pelicci, 1999). Retinoid responsive genes with no obvious RARE or RXRE elements in their promoters, usually require de novo protein synthesis to induce their expression and, as such, show slower kinetics of expression. These genes are referred to as ‘delayed retinoid response genes’ (Arany et al, 2002; Chen et al, 2001). The retinoid receptors have different affinities for retinoid ligands. For the purpose of this thesis, I will focus on all-trans retinoic acid (ATRA), 9-cis retinoic acid (9-cis RA) and 13-cis retinoic acid (13-cis RA). The RARs recognize both ATRA and 9-cis RA with similar affinity (Kd ~ 1nM). The RXRs exclusively bind to 9-cis RA with affinities ranging from 14-18 nM depending on the receptor isoform (Allenby et al, 1993). 9-cis RA has a higher affinity for RAR than RXR receptors. Competitive binding assays with radio-labeled 9-cis RA and 50 µM 13-cis RA indicate that 13-cis RA does not bind RXRs; however, some experiments show that 13-cis RA can bind to RARs, specifically RARγ, with low affinity (Allenby, et al., 1993; Idres et al, 2002). In addition to the RARs and RXRs that have defined ligands and interactions, there are also RAR-related orphan receptors (RORa,b,c, etc.) that, as of yet, have no identified biological ligands. The vast majority of tissues express at least one RAR and RXR receptor. Expression of the multiple RAR and RXR isotypes varies by tissue and stage of development suggesting that each receptor performs some unique functions; although, some functional redundancy has also been identified (for Review see (Chambon, 1994; Chiba et al, 1997; Krezel et al, 1998). Skin expresses all 6 isoforms of the RAR and RXR receptors, with RARγ and RXRα the predominant 28 isoforms (Boehm et al, 2004; Chakravarti et al, 2006; Elder et al, 1992; Roos, et al., 1998). Within the sebaceous gland, expression of RARβ and RXRα has been detected (Boehm, et al., 2004; Downie, et al., 2004). .

1.4.1.3 Retinoid binding proteins and retinoid metabolizing enzymes

In addition to binding to retinoid receptors, the functions of retinoic acids are also controlled by retinoid binding proteins (RBP). Numerous retinoid binding proteins have been identified and are classified as extracellular binding proteins (lipocalins) or intracellular, cytosolic binding proteins. Each RBP demonstrates ligand-binding specificity (Table 3). RBPs are involved in retinoid absorption, transport within the blood and regulating levels of “free” retinoids (Napoli, 1996). Cellular Retinoic Acid Binding Protein I and II (CRABP I, II) have been identified in human skin and bind to both 9-cis RA and ATRA but neither binds to 13-cis RA (Napoli, 1996; Roos, et al., 1998; Rosdahl et al, 1997). 29 Table 3: Retinoid binding proteins with ligand identification. Table taken from Napoli, JL. Biochemical pathways of retinoid transport, metabolism, and signal transduction. (1996) Clinical Immunology and Immunopathology. 80(3):S52-62

MW Class/Protein Primary ligands Loci Prospective function (kDa) Extracellular lipid-binding proteins (lipocalins) RBP 21 Retinol Serum Retinol transporter β-lactoglobulin 18.3 Retinol? Milk Retinol transporter? E-RABP 18.5 RA = 9cRA Epididymis RA/9cRA transporter Intracellular lipid-binding proteins holo: substrate for LRAT and Many (e.g., liver, RoDH CRBP 14.6 Retinol » retinal kidney, testis) apo: stimulates REH; inhibits LRAT

holo: substrates for LRAT and CRBP(II) 14.6 Retinol = retinal Intestine retinal reductase Many (e.g., holo: substrate for RA RA » 9cRA > CRABP 15 testis, lung, metabolism; sequesters RA 13cRA » 9,13cRA kidney) and possibly RA metabolites RA » 9cRA > 9cRA Adult Skin, Same as for CRABP but with CRABP(II) 15.7 » 9,13cRA Embryo different affinities for RAs? Others RPE (retinal 11-cis-retinal, 11- Protects retinoids from CRALBP 33 pigment cis-retinol isomerization epithelium) Retinol, many IRBP 145 Lipid transporter others

Retinaldehyde dehydrogenases (RALDH) and cytochrome p450 (CYP)-dependent 4- hydroylases also play a role in regulating the activity of retinoids within cells. Retinaldehyde dehydrogenase converts retinaldehyde to the active retinoic acid form. Within the central nervous system RALDH2 expression is critical to generating the gradient of retinoic acid expression required for hindbrain patterning (Glover et al, 2006). To date, characterization of RALDH isoforms in human skin has not been done. CYP-dependent 4-hydroxylases convert retinoic acid forms to their 4-oxo-retinoic acid and 4-hydroxy-retinoic forms, which are more likely to be secreted from cells, shutting down the retinoid activity (Ramp et al, 1994; Williams and Napoli, 1985). However, it is known that these 4-oxo-RA products are themselves, capable of gene activation in human skin (Baron et al, 2005). 30 1.4.1.4 Retinoid Function

Retinoids have been called the master regulators of differentiation and play critical roles in development (McCaffery and Drager, 2000). Retinoic acid is essential for both embryonic and adult growth. Retinoids control patterning of the central nervous system influencing development of the hindbrain and spinal cord; the development of the cardiovascular system; development of the kidney, eye, ear; and the olfactory pathway, among others (Figure 6) (Glover, et al., 2006; Hyatt and Dowling, 1997; LaMantia et al, 2000; Maden, 2006; Mendelsohn et al, 1999; Mollard et al, 2000; Romand et al, 2006; Vermot et al, 2003). In short, retinoids control the processes of proliferation, differentiation and apoptosis throughout an organism’s life. Several studies indicate that the effects of retinoids on cell proliferation, differentiation and apoptosis are retinoid-specific and cell-type specific.

1.4.1.4.1 Proliferation

Progression through the cell cycle, from the Gap1 (or growth phase 1) to mitosis (M), is tightly regulated by the levels and activity of specialized groups of proteins known as cyclins, 31 cyclin-dependent kinases (CDK) and cyclin-dependent kinase inhibitors (CDKI) (Figure 7) (Golias et al, 2004). Retinoids influence each one of these groups of proteins. For example, retinoic acid induces growth arrest in myeloid cell lines by up-regulation of p21/CIP1 and p27/KIP1 (CDKIs); down-regulation of cyclin E and cyclin D1/D3, cyclin A and cyclin B; decreased CDK activity; and de-phosphorylation of pRb (Dimberg and Oberg, 2003). In EBV-immortalized B lymphocytes, ATRA-, 9-cis RA- and 13-cis RA- triggered growth arrest is associated with multiple changes in G1 regulatory proteins including decreased activity of CDK2, CDK4 and CDK6; decreased levels of cyclin D3 and cyclin A; as well as increased expression of p27/KIP1 (Zancai et al, 1998).

Figure 7: Simplified view of cell cycle control with focus on G1 and S phase of the cell cycle. Diagram from http://www.scielo.br/img/fbpe/rimtsp/v44n1/a07fig03.gif

Within the sebaceous gland, 13-cis RA has been shown to decrease proliferation of sebocytes as evidenced by decreased 3H-thymidine labeling after 12-weeks of treatment when compared to baseline (Landthaler, et al., 1980). No studies to date have examined the nature of this decrease in proliferation. 32 1.4.1.4.2 Differentiation

In the simplest definition, differentiation is a series of biochemical and structural changes by which cells become specialized in form and function. The most studied example of retinoids controlling differentiation is the patterning and formation of the hindbrain during embryonic development (for Review, (Glover, et al., 2006)). In this case, the timing of and the gradient expression of retinoic acid determines the antero-posterior as well as the dorso-ventral axes of the hindbrain (Avantaggiato et al, 1996; Glover, et al., 2006). Excess retinoic acid in the anterior region results in a “posteriorizing” of the hindbrain whereas decreased retinoic acid in the posterior region results in an “anteriorizing” of the hindbrain (Glover, et al., 2006; Thompson et al, 1969), which upsets the normal developmental patterns leading to deleterious defects. In addition, the gradient of retinoic acid expression influences the expression of the Hox genes, which are critical in rhombomere segment determination (Simeone et al, 1990; Wilkinson et al, 1989; Wood et al, 1994). Furthermore, the pattern formation of the hindbrain is specific to the type of retinoid (i.e.: retinoic acid (ATRA)). Exogenous administration of 9-cis RA in zebrafish has a more pronounced effect than ATRA in “posteriorizing” the hindbrain (Zhang et al, 1996). This central nervous system example illustrates the fact that the amount of retinoic acid and the specific isoform of retinoic acid present can have profound effects on differentiation. Retinoids also influence differentiation in skin. The earliest studies noted that deficiency of vitamin A in the diet of laboratory animals led to changes in the normal epithelium, with specific loss of the mucous secretory epithelium, while replenishing the vitamin A restored the normal phenotype (Wolbach and Howe, 1925). Retinoids inhibit the differentiation of keratinocytes as evidenced by decreased keratin 1, keratin 10, transglutaminase, loricrin, and filaggrin expression (Fisher and Voorhees, 1996). De Luca et al. have studied the effects of retinoids in mouse endocervical epithelia undergoing squamous metaplasia as a result of retinoid deficiency. Their studies show that vitamin A deficiency causes a simple-columnar epithelium to gradually become squamous metaplasia and that vitamin A concentration is a factor in maintaining a simple or more stratified epithelial morphology (De Luca et al, 1995). In addition to retinoid effects on epithelium, it can also influence the dermis. For example, it modulates the expression of the genes for hyaluronate and collagen, two major constituents of the dermis. It increases their expression, synthesis and concentration in the skin, helping to reduce wrinkle formation (Sorg et al, 2005). Within the sebaceous gland, retinoids inhibit 33 differentiation as determined by decreases in sebum secretion with 13-cis RA treatment (Strauss, et al., 1980).

1.4.1.4.3 Apoptosis

The definition of apoptosis is based on morphological characteristics. Apoptosis is a highly-regulated, well-orchestrated series of events culminating in nuclear condensation, DNA fragmentation, membrane-blebbing, cell shrinkage, and eventually phagocytosis of the dying cell (Wyllie et al, 1980). Two pathways leading to apoptosis have been well characterized: death receptor (extrinsic) and mitochondrial (intrinsic) apoptotic pathways. These two pathways converge with activation of caspase 3 and, in some cells, by activation of the protein Bid, a Bcl-2 family member. Death receptors on the cell’s surface detect extracellular stimuli and upon binding of their respective ligands, rapidly activate an intracellular caspase signaling cascade that results in apoptosis. The ligands for death receptors include tumor necrosis factor related apoptosis- inducing ligand (TRAIL) and Fas ligand (FasL, CD95L). All death receptors including Fas (CD95) and TRAIL-R1/R2 have an intracellular region termed the “death domain” (DD). It is this specific 80 amino acid sequence that allows the transmission of the apoptotic signal. DDs in the receptor recruit intracellular adaptor molecules (also containing DDs) that have “death effector domains” (DEDs). DEDs recruit and activate the “initiator” caspases 8 and 10 by cleavage. Initiator caspases proceed to activate “effector” caspases 3, 6, and 7, which by cleavage of their specific substrates (ie: PARP, α-fodrin) result in apoptosis of the cell. For example, TRAIL- mediated apoptosis is initiated by binding of TRAIL to a cell surface receptor, TRAILR1 (DR4) or TRAILR2 (DR5), which then recruits caspase 8 via the adaptor molecules, TNF-R1 associated death domain protein (TRADD) and Fas-associated death domain (FADD). Activated caspase 8 directly activates caspase 3, caspase 6, or caspase 7 or activates the intrinsic apoptosis pathway via cleavage and activation of Bid (Figure 8) (Slee et al, 1999; Smith et al, 2003; Wehrli et al, 2000). The mitochondrial apoptotic pathway is activated by intracellular damage sensed by the mitochondria itself resulting in permeability of the outer mitochondrial membrane and release of cytochrome c. When cytochrome c is released into the cytoplasm, it binds to apoptotic protease activating factor 1 (APAF-1) and through a conformational change, caspase 9 is recruited to this 34 complex and is activated. Caspase 9 (initiator caspase) is able to cleave caspase 3 causing further activation of the signaling cascade. The Bcl-2 (B cell lymphoma) family of proteins is intricately involved in intrinsic apoptosis by controlling the mitochondria permeability transition. This family contains both pro-apoptotic and pro-survival proteins. The delicate balance between these two opposing forces is important to whether or not a cell undergoes apoptosis (Figure 8) (Green and Kroemer, 2005; Lucken-Ardjomande and Martinou, 2005).

It is well established that retinoids induce apoptosis in numerous cell types, both normal cells and tumor cell lines. For example, 13-cis RA reduces the survival and genesis of murine hippocampal neurons in vivo (Crandall et al, 2004; Sakai et al, 2004). ATRA has been shown to 35 induce apoptosis in primary and metastatic melanoma cells as well as inducing growth arrest followed by apoptosis in orbital fibroblasts isolated from Graves’ disease patients (Pasquali et al, 2003; Zhang and Rosdahl, 2004). In leukemia cells, 9-cis RA inhibited cell growth and induced apoptosis to a greater extent than 13-cis RA or ATRA; however, in adult T cell leukemia cells, all three retinoids were equally effective (Fujimura et al, 2003; Koistinen et al, 2002). These studies are only a few of many in the literature that demonstrate that the actions of retinoids are unique and specific to the model used. Natural and synthetic retinoids induce apoptosis by activation of the extrinsic or intrinsic pathways or simultaneous activation of both pathways in a retinoid- and cell-type specific manner. Examples of retinoids and their mechanisms of apoptosis induction are listed in Table 4. The cellular targets of retinoids leading to activation of the cell death pathways are just beginning to be revealed.

Table 4: Retinoid induced apoptotic mechanisms Receptor Retinoid Mechanism Agonist

↓ Bcl-2, Bax, survivin, mitochondrial (Fujimura, et al., 2003) ATRA RAR membrane potential (Pratt et al, 2006)

↓ Bcl-2; Nur77/RXR nucleo-mitochondiral (Fujimura, et al., 2003) 9-cis RA RAR, RXR translocation (Lee et al, 2005) (Boehm et al, Bexarotene RXR ↑ TRAIL, transglutaminase 1995),(Altucci et al, 2004) Fenretinide non-classical ↑reactive oxygen species, ceramide induction (Wu et al, 2001) AGN193198 non-classical ↑ caspase activity (Keedwell et al, 2004)

↑caspase; Nur77/RXR nucleo-mitochondiral CD 437 non-classical (Zhang, 2007) translocation ↑ caspase activity ↓ mitochondrial membrane MX3550-1 non-classical (Chun et al, 2005) potential (Fujimura, et al., 2003) ↓ Bcl-xL, Bcl-2, mitochondrial membrane (Arce et al, 2005; 13-cis RA ? potential; ↑ Bax, cleaved caspase 8 Rigobello et al, 1999; Tosi et al, 1999)

1.5 Retinoids in dermatology

The use of retinoids in dermatology dates back to 1925 when abnormal keratinization was noticed in vitamin A-deficient animals by Wolbach and Howe (Wolbach and Howe, 1925). Because of their lipophilic properties as well as their abilities to affect proliferation and differentiation, retinoids have been prescribed for numerous skin conditions including ichthyosis, psoriasis, age spots, some skin cancers and acne.

1.5.1 Significance of research project

Isotretinoin, a known teratogen, is the second drug in the United States (after thalidomide) to have its use, as of March 1, 2006, become restricted within an FDA-mandated registry system, iPLEDGE, involving patients, physicians, pharmacies and wholesalers. With these new restrictions and the drug’s potent teratogenicity, it is extremely important to develop additional treatments for acne. Progress in this area has been hampered by a lack of understanding of the mechanism of action of 13-cis RA in the sebaceous gland. Elucidating the cellular processes, and possible cellular pathways, that are affected by 13-cis RA in sebocytes is a step toward understanding the overall molecular mechanism of action of this drug, which may lead to the identification of alternative strategies for the treatment of acne.

Chapter 2

13-cis Retinoic Acid Induces Apoptosis and Cell Cycle Arrest in Human SEB-1 Sebocytes

AM Nelson, KL Gilliland, Z Cong, DM Thiboutot. Journal of Investigative Dermatology (2006) 126: 2178-2189

2.1 Chapter Abstract

Isotretinoin (13-cis Retinoic Acid) is the most potent inhibitor of sebum production, a key component in the pathophysiology of acne, yet its mechanism of action remains largely unknown. The effects of 13-cis retinoic acid, 9-cis retinoic acid, and all-trans retinoic acid on cell proliferation, apoptosis, and cell cycle proteins were examined in SEB-1 sebocytes and keratinocytes. 13-cis retinoic acid causes significant dose-dependent and time-dependent decreases in viable SEB-1 sebocytes. A portion of this decrease can be attributed to cell cycle arrest as evidenced by decreased DNA synthesis, increased p21 protein expression, and decreased cyclin D1. Although not previously demonstrated in sebocytes, we report that 13-cis RA induces apoptosis in SEB-1 sebocytes as shown by increased Annexin V- FITC staining, increased TUNEL staining, and increased cleaved caspase 3 protein. Furthermore, the ability of 13-cis retinoic acid to induce apoptosis cannot be recapitulated by 9-cis retinoic acid or all- trans retinoic acid, and it is not inhibited by the presence of a retinoid acid receptor (RAR) pan- antagonist AGN 193109. Taken together these data indicate that 13-cis RA causes cell cycle arrest and induces apoptosis in SEB-1 sebocytes by a retinoid acid receptor (RAR) independent mechanism, which contributes to its sebosuppressive effect and the resolution of acne. 38 2.2 Introduction

Isotretinoin (13-cis Retinoic Acid, (13-cis RA)) is the most potent inhibitor of sebum production, a key component in the pathophysiology of acne. It is the only retinoid that dramatically reduces the size and secretion of sebaceous glands (Goldstein et al, 1982; Landthaler, et al., 1980; Strauss, et al., 1980). Despite the fact that isotretinoin is extremely effective against acne, surprisingly little is known regarding its molecular mechanism of action; although, advances are being made in this area. This unique retinoid has been shown to competitively inhibit the 3α-hydroxysteroid activity of retinol dehydrogenase leading to decreased androgen synthesis in vitro as well as inhibit the migration of polymorphonuclear leukocytes into the skin supporting its role in the reduction of inflammation that is associated with acne (Karlsson, et al., 2003; Wozel, et al., 1991). Numerous studies indicate that 13-cis RA and other retinoids affect cell cycle progression, differentiation, apoptosis and cell survival in a variety of cell types including human breast cancers, oral squamous cell carcinomas, lymphocytes and murine neurons (Cariati et al, 2000; Crandall, et al., 2004; Giannini et al, 1997; Pomponi et al, 1996; Sakai, et al., 2004; Toma et al, 1997). Like previous studies in other cell types, 13-cis RA has been shown to decrease sebocyte proliferation and inhibit sebocyte differentiation as indicated in histology specimens, primary sebocytes and SZ95 immortalized human sebocytes (Doran et al, 1980; Jones, et al., 1980; Landthaler, et al., 1980; Ridden et al, 1990; Strauss, et al., 1980; Zouboulis, et al., 1999; Zouboulis, et al., 1991; Zouboulis, et al., 1991). Although increased levels of caspase 3 were noted in SZ95 sebocytes 24 hours following treatment with 13-cis RA and inhibition of cell growth was evident at 7 days, other markers failed to indicate that SZ95 sebocytes were undergoing apoptosis (Wrobel et al, 2003; Zouboulis et al, 1993). We hypothesized that 13-cis RA reduces sebocyte counts by cell cycle arrest and/or apoptosis and that these effects might not be apparent within a 24-hour treatment period. In this chapter, we report that after 48 and 72 hours of treatment with 13-cis RA, but not 9-cis retinoic acid (9-cis RA) and all-trans retinoic acid (ATRA), inhibits growth and induces apoptosis in immortalized human SEB-1 sebocytes but not in HaCaT keratinocytes or normal human epidermal keratinocytes (NHEK). Furthermore, the retinoid acid receptor (RAR) pan antagonist, AGN 193109, does not block the apoptosis induced by 13-cis RA suggesting an RAR-independent apoptotic mechanism. We hypothesize that the ability of 13-cis RA to induce 39 cell cycle arrest and apoptosis in sebocytes contributes to the overall effect on suppression of sebum production and improvement in acne.

2.3 Results

2.3.1 13-cis RA exhibits a more rapid onset of growth inhibition of SEB-1 sebocytes compared to 9-cis RA and ATRA.

There is a significant dose-dependent decrease in cell count after 48 and 72 hours of treatment with 13-cis RA. At 48 hours, 13-cis RA concentrations of 0.1, 0.5, and 1 µM decreased cell count by 19, 22, 30%, respectively, when compared to vehicle (p < 0.05). After 72 hours, cell numbers were decreased by 19, 43, and 39% with 13-cis RA concentrations of 0.1 µM (p < 0.01), 0.5 µM (p < 0.0001), and 1 µM (p < 0.05), respectively (Figure 9a). No significant differences in cell number were noted at 24 hours of treatment. The effects of 9-cis RA and ATRA on SEB-1 sebocytes were noted beginning at 72 hours. Decreases of 39 and 43% were noted with 9-cis RA (0.5 and 1 µM, respectively) (p < 0.05). ATRA treatment (0.1 and 1 µM) significantly decreased cell number by 14 and 37%, respectively (p < 0.05) (Figure 9b,c). Overall, each of these three retinoids decreased SEB-1 sebocyte cell numbers at 72 hours, albeit to varying degrees, but effects were noted beginning at 48 hours with 13-cis RA. 40

Figure 9: 13-cis RA, 9-cis RA and ATRA differentially inhibit SEB-1 sebocyte proliferation. (a-c) Time-dependent inhibition of SEB-1 sebocyte proliferation with individual retinoid compounds. SEB-1 cells were cultured in the presence of ethanol vehicle alone (0.01% or less; control), 0.1 µM, 0.5 µM or 1 µM concentrations of 13-cis RA, 9-cis RA, ATRA for 24, 48 or 72 hours. Attached cells were collected, stained with trypan blue, and counted manually by hemacytometer. Data represent mean ± SEM, n = 12. Statistical analysis was performed by ANOVA Two-Factor with Replication. * p < 0.05, **p < 0.01, *** p < 0.0001. 41 2.3.2 13-cis RA significantly inhibits DNA synthesis in SEB-1 sebocytes.

13-cis RA (0.1, 0.5, and 1 µM) significantly decreased thymidine incorporation by approximately 3-fold at 72 hours (p < 0.01). No significant changes were noted at 24 or 48 hours. A 1.85-fold increase in 3H thymidine incorporation was noted when cells were treated with 1 µM 9-cis RA for 48 hours. ATRA concentrations of 0.5 and 1 µM decreased thymidine incorporation by approximately 1.8-fold at 24 and 72 hours, respectively (Figure 10a-c). 42

Figure 10: 13-cis RA inhibits DNA synthesis to a greater extent than 9-cis RA or ATRA. (a-c) SEB-1 sebocytes were treated with ethanol vehicle (0.01% or less; control) or 0.1, 0.5, 1 µM concentrations of 13-cis RA, 9-cis RA or ATRA for 24, 48 or 72 hours. 1µCi 3H thymidine was added to each sample 8 hours prior to collection. Cells were washed and collected for liquid scintillation counting. Data represent mean ± SEM, n ≥12. Statistical analysis was performed by ANOVA Two-Factor with Replication. *p < 0.005 and **p < 0.01. 43 2.3.3 13-cis RA, but not 9-cis RA or ATRA, increases p21 levels in SEB-1 sebocytes.

To further test the hypothesis that 13-cis RA changes cell cycle progression, expression of p21, a cell cycle inhibitor, was examined by western blot. p21 is a general cyclin dependent kinase inhibitor that blocks progression through the G1/S phase of the cell cycle. 13-cis RA significantly increased p21 protein expression after 48 and 72 hours (Figure 11a). Specifically, p21 levels increased on average 2.64-fold and 3.13-fold when cells were treated with 0.1 and 1 µM 13-cis RA, respectively, for 48 hours ( p = 0.008 and 0.05). After 72 hours of treatment, all concentrations tested increased p21 protein expression. Increases in p21 of 1.47-, 2.27-, and 3.01-fold were noted with 0.1, 1, and 10 µM concentrations of 13-cis RA, respectively. No significant differences in p21 expression were noted at 24 hours (data not shown). When SEB-1 sebocytes were treated with 9-cis RA or ATRA in concentrations of 0.1, 0.5 and 1 µM, no significant increases in p21 protein were noted at 48 or 72 hours (Figure 11b,c). 44

Figure 11: 13-cis RA increases p21 and decreases cyclin D1 proteins. (a) SEB-1 cells were treated with 0.1 µM, 1µM, 10 µM 13-cis RA or vehicle. (b-c) Parallel experiments were performed with 0.1 µM 0.5 µM, or 1 µM concentrations of 9-cis RA and ATRA. Blots were incubated with primary antibodies to p21 and β-actin for loading control normalization and analyzed by densitometry. (d) SEB-1 cells were treated with 0.1 µM, 1µM, 10 µM 13- cis RA or vehicle and blots were incubated with primary antibodies to cyclin D1 and β- actin. Magic Mark XP (MM) indicates band size. Blots are representative of a minimum of three western blots. Graphs represent normalized values relative to vehicle (control) expression of a minimum of three independent western blots. Mean ± SEM. * p < 0.05 ** p = 0.01 45 2.3.4 13-cis RA, but not 9-cis RA or ATRA, decreases cyclin D1 protein in SEB-1 sebocytes.

To further explore the possibility that 13-cis RA induces G1 arrest in SEB-1 sebocytes, cyclin D1 protein was examined by western blot. Cyclin D family members are expressed and function in controlling the progression from G1 to S phase in the cell cycle (Baldin et al, 1993). Overexpression of cyclin D1 shortens the duration of the G1 phase and is rate limiting in phase progression (Quelle et al, 1993). Therefore, cyclin D1 is a likely candidate to confirm the actions of 13-cis RA in inhibiting cell cycle progression by influencing the G1 to S phase transition. In SEB-1 sebocytes, 13-cis RA in concentrations of 0.1, 1 and 10 µM significantly decreased cyclin D1 protein at 72 hours. No significant effects of 13-cis RA were noted at 24 or 48 hours (24-hour data not shown). 9-cis RA or ATRA concentrations of 0.1, 0.5 and 1 µM did not reduce cyclin D1 protein levels at 72 hours (data not shown) (Figure 11d).

2.3.5 13-cis RA induces apoptosis in SEB-1 sebocytes but not in HaCaT keratinocytes or NHEK.

To determine if the effect of 13-cis RA on apoptosis is cell-type specific, time course experiments were conducted in SEB-1 sebocytes, HaCaT keratinocytes and normal human epidermal keratinocytes (NHEK). In SEB-1 sebocytes, no significant differences in apoptosis were noted in cells treated with 13-cis RA for 2, 4, 6 or 24 hours. A marginal, yet significant, increase in the percentage of cells in early apoptosis was noted in SEB-1 cells treated with 0.1 µM 13-cis RA: 2.03% to 2.49% at 48 hours and 2.19% to 2.84% at 72 hours (p < 0.01 at both time points). Significant increases in the percentage of cells in late apoptosis were noted at 48 and 72 hours with increasing concentrations of 13-cis RA (Figure 12a, late apoptosis shown). 46

Figure 12: 13-cis RA induces late apoptosis in SEB-1 sebocytes but not in HaCaT keratinocytes or NHEK. (a) SEB-1 cells were treated with vehicle (negative control), 13- cis RA (0.1 µM or 1 µM), or staurosporine (S) (positive control) for indicated times. (b) HaCaT cells were treated with vehicle, 13-cis RA (0.1 µM or 1 µM), or staurosporine (S) for the indicated times. (c) NHEK cells were treated with vehicle, 13-cis RA (0.1 µM or 1 µM), or staurosporine (S) for indicated times. In all experiments, cells were prepared according to manufacturer’s protocol for Annexin V-FITC / PI staining. (BD ApoAlert, BD Biosciences) Data was analyzed with Cell Quest Software and represent mean ± SEM, n ≥ 12. Statistical analysis was performed with ANOVA Two Factor with Replication. *p<0.01, **p<0.00001. 47 Specifically, 0.1 µM 13-cis RA increased the percentage of late apoptosis: 4.06% to 5.22% at 48 hours and 5.31% to 8.11% at 72 hours. 13-cis RA at 1 µM concentration caused increases from 3.64% to 5.08% and from 7.57% to 12.18% at 48 and 72 hours, respectively (Figure 12a). Nanomolar concentrations of 13-cis RA did not induce apoptosis at any of the time points examined (data not shown). In HaCaT keratinocytes, no significant differences in the percentage of cells in early or late stage apoptosis or necrosis were noted in cells treated with 0.1 µM 13-cis RA at all time points examined. 13-cis RA (1 µM) significantly increased the percentage of cell in early and late stage apoptosis at 24 and 48 hours. Yet, these increases were very minor, with the total percentage of cells in apoptosis with 13-cis RA treatment being less than 2% of the cells (Figure 12b). In experiments with NHEK, no significant differences were noted in cells treated with 13-cis RA, with the exception of an increase from 5.25% to 6.2% in late stage apoptosis at 2 hours with 1 µM 13-cis RA (Figure 12c). Apoptosis was significantly induced by staurosporine in SEB-1 sebocytes, HaCaT keratinocytes and NHEK showing that all three cell types are capable of undergoing apoptosis. No significant differences were noted between standard culture medium and ethanol vehicle controls in any cell type at any time point during these studies indicating that the concentrations of ethanol used in these experiments did not induce apoptosis.

2.3.6 13-cis RA specifically increases levels of cleaved caspase 3 in SEB-1 sebocytes.

SEB-1 sebocytes were treated with 13-cis RA and four independent western blots were run to detect cleaved caspase 3. No cleaved caspase 3 was noted at 24 hours in negative control lanes or in cells treated with 13-cis RA. 13-cis RA significantly increased cleaved caspase 3 levels at 48 and 72 hours in SEB-1 sebocytes (Figure 13a). Specifically, 0.1 µM 13- cis RA and 1 µM 13-cis RA increased expression of cleaved caspase 3 an average of 3.58-fold and 3.33-fold (p < 0.01), respectively, at 48 hours. Small fold increases were noted at 72 hours that were not statistically significant. Although the magnitude of the increase in cleaved caspase 3 was greatest with 10 µM 13-cis RA at 48 hours, these results were not statistically significant. This is, most likely, due to the variability induced by the limited survival of the cells at this higher concentration. 48

To determine if the induction of apoptosis is specific to 13-cis RA, SEB-1 sebocytes were also treated with 0.1, 0.5 and 1 µM concentrations of 9-cis RA and ATRA. Again, no cleaved caspase 3 was detected at 24 hours post treatment in negative controls or with any 49 concentration of either retinoid. Furthermore, and unlike the case with 13-cis RA, no significant increases in cleaved caspase 3 were noted with either 9-cis RA or ATRA at 48 and 72 hours (Figure 13b,c). For additional confirmation that the apoptotic effect of 13-cis RA is specific to sebocytes, western blots for cleaved caspase 3 were performed on NHEK. No cleaved caspase 3 could be detected at any time point examined with NHEK cells treated with 13-cis RA. However, cleaved caspase 3 was detected with 1 µM staurosporine treatment further confirming that these cells are capable of undergoing apoptosis (Figure 14).

2.3.7 13-cis RA, but not 9-cis RA or ATRA, increases TUNEL staining in SEB-1 sebocytes.

To further test the hypothesis that 13-cis RA induces apoptosis in SEB-1 sebocytes and to confirm the results from the annexin V-FITC FACS experiments, we examined the effects of 13-cis RA on SEB-1 sebocytes by TUNEL assay. 13-cis RA (0.1 and 1 µM) increased the percentage of TUNEL-positive cells by 3.5- and 5.67-fold, respectively (p ≤ 0.01) at 48 hours, while each concentration increased the percentage of TUNEL-positive cells by approximately 13-fold at 72 hours (p ≤ 0.01) (Figure 15a,b). No differences were noted at 24 hours (data not shown). To compare the actions of 13-cis RA to its isomerization products, SEB-1 sebocytes were also treated with the same concentrations of 9-cis RA and ATRA and no significant increases in TUNEL-stained cells were noted at any time point examined (Figure 15a,b). Both 9-cis RA and ATRA had 1-3% TUNEL positive cells at all time points. Fenretinide is a synthetic retinoid known to induce apoptosis by ceramide or reactive oxygen species generation. 50 Fenretinide treatment in SEB-1 sebocytes significantly increased the percentage of TUNEL- positive cells in a dose-dependent manner at 48 and 72 hours (ranging from 15% to 85% positive cells) (Figure 15a,b). No significant increase in TUNEL staining was noted with retinoid X receptor (RXR) pan agonist, CD 3254, at 48 hours. However, 50 nM CD 3254 significantly increased TUNEL-positive cells from 3% to 48% at 72 hours (p < 0.01) (data not shown). 51

Figure 15: The increase in TUNEL staining with 13-cis RA is not inhibited in the presence of RAR pan antagonist AGN 193109. (a) Representative images of control, 0.1 µM, 1 µM, and 10 µM 13-cis RA, 9-cis RA, ATRA, and fenretinide treatments at 72 hours. (48 hour data not shown) (b) Quantification of the percentage of TUNEL positive stained cells per treatment at 48 and 72 hours. (9-cis RA not shown) Data represent mean + SEM, n = 6-12. Statistical analyses were performed with ANOVA Two Factor with Replication. * p < 0.01 ** p < 0.001 (c) Representative images of negative control, 1 µM 13-cis RA, AGN 193109, and 13-cis RA combined with 10 µM AGN 193109 at 72 hours. (48 hour data not shown) (d) Quantification of the percentage of TUNEL positive cells at 72 hours. Data represent mean + SEM, n = 12. Statistical analyses were performed with ANOVA Two Factor with Replication. * p < 0.05 when compared to control; + not statistically different. (e) QPCR verification of RAR antagonist AGN 193109 activity in SEB-1 sebocytes. Bars represent the efficiency corrected normalized average fold change of TIG1 under the experimental conditions as determined by REST-XL software. n = 4. 52

2.3.8 Apoptosis induction by 13-cis RA in SEB-1 sebocytes is not blocked by RAR antagonist AGN 193109

To determine if the effects of 13-cis RA on apoptosis are mediated by retinoic acid receptors (RARs), SEB-1 sebocytes were treated with 1 µM 13-cis RA in the presence of 10 µM AGN 193109, an RAR pan antagonist, and the TUNEL assay was performed. 13-cis RA alone significantly increased the percentage of TUNEL positive cells by approximately 5-fold at 48 and 72 hours. (p < 0.05). These increases were not inhibited in the presence of AGN 193109 at 48 and 72 hours (Figure 15d,e). To verify the activity of AGN 193109 within our cells at the time points examined in the TUNEL assay, we performed quantitative PCR for RAR responsive gene, tazarotene-induced gene 1 (TIG1). RAR activation induces the expression of TIG1 (Nagpal et al, 1996). In the presence of 1 µM 13-cis RA alone, TIG1 expression was approximately 13- and 17-fold higher than controls at 48 and 72 hours, respectively. With the addition of AGN 193109, TIG1 gene expression dramatically decreases at 48 and 72 hours and is lower than vehicle treated controls (Figure 15e).

2.3.9 13-cis RA is isomerized to ATRA over time in SEB-1 sebocytes.

To study the kinetics of 13-cis RA uptake in SEB-1 sebocytes and its possible isomerization to ATRA or 9-cis RA, SEB-1 sebocytes were treated with 13-cis RA and subjected to HPLC analysis. 13-cis RA remains relatively stable in standard culture medium for approximately 24 hours (Figure 16a). The concentration of 13-cis RA in standard culture medium alone is similar to the concentration in medium removed from SEB-1 sebocyte- containing plates (Figure 16a,b). The concentration within SEB-1 sebocytes increases to a maximum of 350 ng/mL at 12 hours, at which point the concentration declines for the duration of the experiment (Figure 16c). The concentration of ATRA in the medium alone and from plates containing SEB-1 sebocytes was much lower than 13-cis RA concentrations at the corresponding time points. The concentration of ATRA within SEB-1 sebocytes begins to rise at 12 hours and continues through the remaining time periods. 9-cis RA concentrations are 53 minimal at best, both in medium alone and in medium from SEB-1 containing plates during the time course. Within SEB-1 sebocytes, 9-cis RA concentrations range from 1.4 ng/mL at 0 hour to a maximum or 12ng/mL at 72 hours; these concentrations are magnitudes lower than either 13-cis RA or ATRA at the same time periods. 54

Determining the actions of isotretinoin on the sebaceous gland is essential in advancing our understanding of the molecular mechanism of action of this drug and in our search for safer therapeutic alternatives. Several studies indicate that the effects of retinoids on cell proliferation, cell cycle, and apoptosis are retinoid or cell-type specific. For example, growth inhibition with 13-cis RA has been reported in human breast cancer cell lines, primary glioblastoma cells, Epstein-Barr Virus-immortalized B lymphocytes, and oral squamous cell carcinoma cell lines (Bouterfa et al, 2000; Giannini, et al., 1997; Pomponi, et al., 1996; Toma, et al., 1997). In some cases the effects noted with 13-cis RA or 9-cis RA were not duplicated by ATRA (Bouterfa, et al., 2000). Most studies in other cell types suggest that retinoids cause a block in the G1/S phase of the cell cycle, triggering decreased S phase and an increased percentage of cells in the G0/G1 phase (Crandall, et al., 2004; Giannini, et al., 1997; Toma, et al., 1997). It is also well established that retinoids induce apoptosis in numerous cell types, both normal cells and tumor cell lines, although not previously demonstrated in sebocytes. For example, in doses comparable to those given for the treatment of acne in humans, 13-cis RA reduces the survival and genesis of murine hippocampal neurons in vivo (Crandall, et al., 2004; Sakai, et al., 2004). ATRA has been shown to induce apoptosis in primary and metastatic melanoma cells (Zhang and Rosdahl, 2004) as well as inducing growth arrest followed by apoptosis in orbital fibroblasts isolated from Graves’ disease patients (Pasquali, et al., 2003). In OCI/AML-2 retinoid-sensitive cell line subclones, derived from leukemia cells, 9-cis RA inhibited cell growth and induced apoptosis to a greater extent than 13-cis RA or ATRA (Koistinen, et al., 2002). These studies demonstrate that the actions of retinoids are unique and specific to the model used. The exact mechanism of action of 13-cis RA in the treatment of acne remains largely unknown. 13-cis RA has little to no ability to bind to cellular retinol-binding proteins or the RA nuclear receptors (RARs and RXRs) (Allenby, et al., 1993; Fogh K. et al, 1993; Levin et al, 1992). It has been suggested that 13-cis RA may, in fact, act as a pro-drug that is isomerized intracellularly to ATRA, which can bind to and activate RAR, leading to the overall inhibition of sebocyte proliferation (Tsukada et al, 2000). Our studies confirm that 13-cis RA is primarily isomerized to ATRA in SEB-1 sebocytes beginning at 24 hours. It is well established, however, that 13-cis RA is superior to either 9-cis RA or ATRA for sebosuppression (Geiger et al, 1996; Hommel et al, 1996; Ott et al, 1996). Alternatively, 13-cis RA may act in a receptor independent manner by influencing cellular signaling pathways through direct protein interactions as 56 demonstrated with other retinoids or by enzyme inhibition (Hoyos et al, 2000; Imam et al, 2001; Karlsson, et al., 2003; Zorn and Sauro, 1995). Previous studies have examined the actions of 13-cis RA, 9-cis RA and ATRA on cultured human sebocytes, SZ95 SV40-immortalized sebocytes, and rat preputial cells (Tsukada, et al., 2000; Wrobel et al, 2003; Zouboulis, et al., 1991; Zouboulis, et al., 1993). 13- cis RA at concentrations greater than 10-7 µM and ATRA (10-6 to 10-5 M) significantly decreased human sebocyte proliferation after 7 and 14 days (Zouboulis, et al., 1991; Zouboulis, et al., 1993). Studies of immortalized human sebocytes SZ95, showed that 13-cis RA, 9-cis RA and ATRA at concentrations of 10-7 M, all significantly reduced proliferation by approximately 50% after 9 days of treatment (Tsukada, et al., 2000). In primary rat preputial cells, ATRA and other RAR-selective agonists significantly decreased cell numbers after 9 days (Kim et al, 2000). Processes such as cell cycle arrest or apoptosis may explain the histological data in human skin biopsies that demonstrate a drastic decrease in size, shape, and lipid content of the sebaceous glands after 16 weeks of isotretinoin treatment (Goldstein, et al., 1982). Since proliferation studies in SZ95 sebocytes suggested that the effects of 13-cis RA and other retinoids may be noted after 7-9 days of treatment, we designed experiments to examine the early effects of 13-cis RA, 9-cis RA and ATRA on proliferation, cell cycle progression and apoptosis focusing on 24, 48 or 72 hours of treatment. Our proliferation studies show that 13-cis RA causes a dose-dependent decrease in cell count after 48 and 72 hours whereas 9-cis RA and ATRA show significant decreases beginning at 72 hours. We would expect that if our experiments were extended, the magnitude of this decrease would be greater as previously reported in SZ95 sebocytes after 9 days (Tsukada, et al., 2000). Overall, 13-cis RA at the concentrations tested in our studies act sooner in inhibiting proliferation than either 9-cis RA or ATRA. These data are supported by studies demonstrating an approximate 3-fold decrease in 3H-thymidine incorporation in SEB-1 sebocytes that were treated with 13-cis RA for 72 hours. This decrease is nearly 2-fold greater than the decreases produced by 9-cis RA or ATRA. This experiment suggests that 13-cis RA is more potent at growth inhibition than either 9-cis RA or ATRA in SEB-1 sebocytes. Further supporting the hypothesis that 13-cis RA causes a block in the G1/S phase of the cell cycle as demonstrated in other cell types, we show that 13-cis RA increases p21 protein and decreases cyclin D1 protein expression at 48 and 72 hours. Cyclin D1 protein expression decreases by approximately 50% by 72 hours which coincides with our 3H-thymidine studies where 13-cis RA had the most striking effect at 72 hours. Furthermore, cyclin D1 protein was 57 not decreased with 9-cis RA or ATRA at 72 hours, which is also consistent with our 3H- thymidine incorporation studies. No significant increases in p21 protein were noted with 9-cis RA or ATRA; although increasing trends were noted. Taken together these experiments show that in SEB-1 sebocytes, 13-cis RA is much more effective than 9-cis RA or ATRA in both decreasing the proportion of cells synthesizing DNA and inducing a G1/S phase block by increasing p21 and decreasing cyclin D1 protein expression. Studies in SZ95 sebocytes did not demonstrate apoptosis in cells treated for up to 24 hours with 13-cis RA (10-8 to 10-5 M) and assayed by DNA fragmentation and lactate dehydrogenase release (Wrobel, et al., 2003). At 24 hours, no changes in apoptosis were noted when SZ95 sebocytes were treated with 10-7 M 13-cis RA as assessed annexin V staining, cell death assays or FACS analysis and reverse transcription (RT)-PCR for the apoptotic proteins, Bcl-2 and Bax. Interestingly, in SZ95 sebocytes, 13-cis RA increased the levels of caspase 3 as detected by FACS analysis at 24 hours. Accordingly, in our studies of SEB-1 sebocytes, no increase in apoptosis was noted 24 hours after 13-cis RA as assayed by annexin V-FITC FACS. However, increases in early and late stage apoptosis were noted at 48 and 72 hours with concentrations of 13-cis RA similar to those used in SZ95 sebocytes, although the magnitude of the percentage of cells is small compared to the positive control, staurosporine. In contrast, the magnitude of the changes in apoptosis induced by 13-cis RA was much greater in the TUNEL assay. By extending our treatment times beyond 24 hours, we were able to detect the induction of apoptosis by 13-cis RA, which was verified by increased expression of cleaved caspase 3. Furthermore, the increase in apoptosis was limited to 13-cis RA as no significant increases in apoptosis were noted when SEB-1 sebocytes were treated with 9-cis RA or ATRA. The effects of 13-cis RA on apoptosis and growth inhibition may or may not be mediated by retinoid receptors. It is possible that the effects of 13-cis RA on apoptosis and growth inhibition may be mediated by other isomerization products such as 4-oxo-isotretinoin or 4- hydroxy-isotretinoin (Orfanos and Zouboulis, 1998). The 4-oxo metabolites of retinoids have been shown to be functionally active in human keratinocytes and fibroblasts by their ability to induce changes in gene expression (Baron, et al., 2005). Our data show that RAR pan- antagonist AGN 193109 sufficiently blocks RAR activation in the presence of 13-cis RA as measured by a significant decrease in TIG1 gene expression, yet does not block apoptosis induced by 13-cis RA in SEB-1 sebocytes, thus supporting the hypothesis that apoptosis induction via 13-cis RA is independent of RAR activation. Alternatively, apoptosis maybe 58 mediated through RXR nuclear receptor activation (Zhao et al, 2004). Using RXR pan-agonist, CD 3254, at a concentration of 50 µM, a significant increase in the percentage of TUNEL- positive SEB-1 was noted at 72 hours. Although our HPLC data indicate very low levels of 9-cis RA (a maximum of 12 ng/mL at 72 hours), RXR activation by 9-cis RA is possible (Allenby, et al., 1993) or 13-cis RA may be metabolized to another as yet unidentified metabolite that is capable of RXR activation. Alternatively, 13-cis RA may have effects that are independent of retinoid receptors. Interestingly, we showed that fenretinide; a synthetic retinoid known to induce apoptosis by primarily, RAR- and RXR-independent means is able to induce significant apoptosis in our SEB- 1 sebocytes. In fact, the degree of apoptosis induced by fenretinide at 48 hours is very similar to that observed with 13-cis RA treatment at 72 hours. Fenretinide induces apoptosis by elevating reactive oxygen species, and increases in activation of ceramide and caspases (Wu, et al., 2001). In addition, a retinoid-related molecule, AGN 193198 induces apoptosis without activation of the classical retinoid receptors (Balasubramanian et al, 2005; Keedwell, et al., 2004). It is possible that 13-cis RA acts similarly to fenretinide or AGN 193198 via receptor- independent mechanisms; although additional experiments are required to test this hypothesis. Since the actions of retinoids differ in various cell types and the effects of 13-cis RA are most profound on sebaceous glands in vivo, it is likely that the induction of apoptosis and cell cycle arrest is specific to sebocytes. 13-cis RA failed to induce apoptosis in HaCaT keratinocytes or NHEK. It is possible that with higher concentrations of 13-cis RA or longer treatment times that apoptosis may be induced in keratinocytes. Although there is no evidence in the literature of 13-cis RA-induced apoptosis in keratinocytes, ATRA and tazarotene (RAR β/γ selective agonist), have been shown to induce apoptosis in HaCaT keratinocytes (Louafi et al, 2003; Papoutsaki et al, 2004). Taken together, these experiments support the hypothesis that 13-cis RA specifically induces apoptosis in SEB-1 sebocytes and not in keratinocytes. In conclusion, our data indicate that 13-cis RA inhibits growth and induces apoptosis in SEB-1 sebocytes and not keratinocytes at concentrations that are therapeutically achievable in human plasma (Adamson, 1994; Almond-Roesler et al, 1998; Rollman and Vahlquist, 1986). Previous studies in human sebocytes and immortalized sebocytes have also documented growth inhibition with 13-cis RA, however, we have extended these studies to show that this growth inhibition is most likely due to influencing the G1/S phase of the cell cycle as evidenced by decreased DNA synthesis, increased p21 protein and decreased cyclin D1 protein. In addition, we report for the first time, that 13-cis RA also induces apoptosis in SEB-1 sebaceous 59 cells. The ability to induce apoptosis is specific to sebocytes, not keratinocytes, and is distinct from the effects observed with 9-cis RA or ATRA that may account, in part, for the superior efficacy of 13-cis RA in reducing sebum production. Furthermore, the induction of apoptosis by 13-cis RA does not appear to involve RAR nuclear receptors. Elucidating the cellular processes that are affected by 13-cis RA in sebocytes is a step toward understanding the overall molecular mechanism of action of this drug, which may lead to the identification of alternative strategies for the treatment of acne.

2.5 Materials and Methods

2.5.1 Cell Culture

The SEB-1 human sebocyte cell line was generated by transfection of secondary sebocytes with SV40 Large T antigen as previously described (Thiboutot, et al., 2003). SEB-1 sebocytes were cultured and maintained in standard culture medium containing: 5.5mM low glucose Dulbecco’s Modified Eagle Medium (DMEM) 3:1 Ham’s F12, 2.5% fetal bovine serum (FBS) , hydrocortisone (0.4µg/mL), adenine (1 X 10-8 M), insulin (10ng/mL), epidermal growth factor (3 ng/mL), cholera toxin (1.2 X 10-10 M) and antibiotics. HaCaT keratinocytes were cultured and maintained in 5.5 mM low glucose DMEM, 5% FBS and antibiotics. HaCaT keratinocytes served as a control cell line in annexin V-FITC FACS apoptosis assays. Normal human epidermal keratinocytes (NHEK)-neonatal, pooled (NHEK- neo, Clonetics Keratinocyte System, Cambrex Bioscience, Walkersville, MD) were cultured in keratinocyte growth medium-2 (KGM-2) (Cambrex Bioscience, Walkersville, MD). NHEK keratinocytes served as control cells in annexin V-FITC FACS apoptosis assays and western blots for cleaved caspase 3.

60 2.5.2 Effects of retinoids on SEB-1 proliferation

Retinoid compounds were purchased through SIGMA (St. Louis, MO): 13-cis RA (R 3255), 9-cis RA (R 4653) and ATRA (R 2625). Stock solutions of retinoids were handled under dimmed yellow light, dissolved in 100% ethanol to a concentration of 10 mM and stored under N2 gas at -20ºC until use. The RAR pan-antagonist AGN 193198 was obtained from Allergan (gift, Dr Rosh Chandraratna), dissolved in DMSO at a concentration of 10 mM and stored at - 70ºC until use. Treatments were made from retinoid stock solutions diluted to the appropriate concentration in standard culture mediums under dimmed yellow light. Staurosporine (S 5921, SIGMA, St Louis, MO) was solubilized in 100% ethanol at a concentration of 10 mM, stored at - 20ºC and diluted to desired final concentration in appropriate cell culture medium for a positive control in apoptosis assays. SEB-1 sebocytes (passage 20-23) were seeded at 4 X 104 cells per 35-mm plate and grown until approximately 40% confluent. Plates were treated with 0.1, 0.5 or 1 µM concentrations of 13-cis RA, 9-cis RA, ATRA or ethanol vehicle (0.01% or less; control) in triplicate for 24, 48 and 72 hours. Cells were detached using trypsin (0.05%), collected, and diluted in standard cell culture medium for manual cell counts using a hemacytometer. Cell viability was assessed using Trypan Blue dye exclusion. Each proliferation assay was performed three independent times. Analysis of variance (ANOVA) Two Factor with Replication was used for analysis. Results were considered significant if p < 0.05.

2.5.3 3H thymidine incorporation assay

SEB-1 sebocytes (passages 21-26) were seeded at 2.5 X 104 cells per well in 12-well plates and grown until 30-40% confluent. Wells were rinsed with phosphate-buffered saline (PBS) prior to the addition of 0.1, 0.5 or 1 µM concentrations of 13-cis RA, 9-cis RA, ATRA or ethanol vehicle (0.01% or less) alone in triplicate wells in standard culture medium. 3H thymidine (1µCi/well) was added a minimum of 8 hours prior to the end of the treatment period. At the end of the treatment period, medium was removed and cells were rinsed twice with PBS, detached using trypsin (0.05%) and collected for liquid scintillation counting. Each assay was performed a 61 minimum of three independent times. Statistical significance was determined with ANOVA Two Factor with Replication. Results were considered significant if p < 0.05

2.5.4 Western blot analysis for p21, cyclin D1 and cleaved caspase 3

To confirm the results from cell proliferation and apoptosis assays, protein levels of p21, cyclin D1 and cleaved caspase 3 were examined using western blot analysis in our various cell lines. p21, a cyclin dependent kinase inhibitor, blocks progression through the G1/S phase of the cell cycle. Cyclin D1 is specifically required for progression into S phase. Caspase 3, the key executioner caspase, is synthesized in the cell as a pro-caspase, which then becomes cleaved and activated when cells undergo apoptosis. Primary antibodies for p21 Waf/Cip1 (DCS60), cyclin D1 (DCS6), cleaved caspase 3 (Asp175) and β-actin as well as secondary anti-rabbit IgG horseradish peroxidase antibody were purchased from Cell Signaling Technology (Beverly, MA). Actin primary antibody and anti-mouse horseradish peroxide-linked secondary antibody were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). SEB-1 sebocytes (passage 20-26) were grown in 100-mm plates in standard culture medium until 50-75% confluent. Plates were rinsed with PBS and treated with 13-cis RA (0.1, 1 and 10 µM); 9-cis RA (0.1, 0.5 and 1 µM); ATRA (0.1, 0.5 and 1 µM); ethanol vehicle (0.01% or less) as a negative control; or 1 µM staurosporine as a positive control. Cells were treated for 24, 48 or 72 hours. NHEK cells (passage 3) were grown in 100-mm plates in standard culture medium until approximately 50-75% confluent. Plates were rinsed with PBS and treated with 13- cis RA (0.1, 0.5, and 1 µM); ethanol vehicle (0.01% or less); or 1 µM staurosporine for 2, 4, 6, 18, 24, 48 or 72 hours. Total cell protein lysates from adherent and floating cells of SEB-1 sebocytes and NHEK were collected, flash frozen in liquid nitrogen and stored at -80ºC until needed. Protein concentration of each sample was determined by the BCA Protein Assay (Pierce, Rockford, IL). Equal amounts of protein were run on NuPage 10% or 4-12% Bis-Tris Gels with MES Running Buffer (Invitrogen Life Technologies, Carlsbad, CA). Gels were transferred to polyvinylidene difluoride membrane, blocked for 1 hour at room temperature in 5% non-fat dry milk and incubated with 1:1,000 dilution of cleaved caspase 3 antibody (Asp175) rabbit monoclonal antibody, 1:1,000 dilution of cyclin D1 mouse monoclonal antibody or 1:8,000-15,000 dilution of 62 p21 mouse monoclonal antibody. Secondary anti-rabbit and –mouse horseradish peroxidase linked antibodies were used to detect primary antibodies. SuperSignal West Pico Chemiluminescent Substrate (Pierce, Rockford, IL) was used for protein detection. Blots were stripped with Restore Western Blot Stripping Buffer (Piece, Rockford, IL) and reprobed with β- actin or actin for a loading control. Films of blots were analyzed and quantified by densitometry with QuantityOne Software (Bio-Rad, Hercules, CA) after background subtraction. Western blots were repeated a minimum of three independent times. Data was analyzed with Student’s t-test and results were considered significant if p < 0.05.

2.5.5 Annexin V-FITC/Propidium Iodide FACS Apoptosis Assay

To determine if 13-cis RA induces apoptosis in SEB-1 sebocytes and the time course of this effect, the annexin V-FITC FACS assay was chosen (Martin et al, 1995). Apoptosis assays were performed in SEB-1 sebocytes, HaCaT keratinocytes and NHEK that were treated with 13- cis RA. SEB-1 sebocytes (passages 22-26) and HaCaT keratinocytes (passages 23-29) were seeded at 8 X 104 cells per 35-mm plate in their standard culture mediums and allowed to grow for 3 days, feeding once prior to treatment. Treatments consisted of standard culture medium or ethanol vehicle (0.01% or less) as negative controls; 1 µM staurosporine as a positive control; and 13-cis RA at final concentrations of 0.1 or 1 µM in SEB-1 sebocytes and HaCaT keratinocytes for the initial studies. For follow-up studies examining a possible 13-cis RA dose response, SEB-1 sebocytes were subjected to 0.1, 1, 10 nM 0.1, 1 and 10 µM as well as previously mentioned controls. All samples were run in triplicate and treatments were carried out for 2, 4, 6, 24, 48 and 72 hours. In parallel experiments, NHEK (passage 3) were grown in KGM-2 growth medium until 70% confluent. Treatments consisted of 0.1, 0.5, and 1 µM 13-cis RA; ethanol vehicle (0.01% or less) and 1 µM staurosporine. Samples were run in triplicate and assayed at 2, 4, 6, 18, 24, 48, and 72 hours. Each sample was prepared according to BD ApoAlert Annexin V Protocol (Cat no. K2025-1, BD Biosciences, Clontech, Palo Alto, CA). Ten thousand events (cells) were collected per sample using flow cytometry and debris was excluded by scatter gating. Single annexin V-FITC and propidium iodide-stained control samples determined quadrants for data analysis. Data analysis was by Cell Quest software (Becton Dickinson, Canada) and percentage of cells in early apoptosis, late apoptosis, necrosis, 63 and viable (unaffected) quadrants were calculated and compared by ANOVA Two Factor with Replication. Assay was performed three independent times. Results were considered significant if p < 0.05.

2.5.6 TdT-Mediated dUTP Nick End Labeling (TUNEL) Staining

SEB-1 sebocytes (passages 22-28) cultured in 12-well plates in standard medium until approximately 30-40% confluent. Wells were rinsed with PBS and were treated in triplicate with ethanol vehicle (0.01% or less) control, 13-cis RA, 9-cis RA or ATRA each in concentrations of 0.1, 1 or 10 µM. Retinoids were diluted in standard culture medium and treatments were carried out for 24, 48 and 72 hours. In parallel experiments, SEB-1 sebocytes (passages 22-24) were cultured as above and treated in triplicate with ethanol vehicle (0.01% or less), DMSO vehicle (0.01% or less), both vehicles together, 1 µM 13-cis RA alone, 10 µM AGN 193109 alone or 1 µM 13-cis RA + 10 µM AGN 193109 combination. Additional experiments were performed with fenretinide, a synthetic retinoid known to induce apoptosis via an RAR independent mechanism. (Wu, et al., 2001) Fenretinide (4-hydroxyphenyl-retinamide) was handled under dimmed yellow light and dissolved in 100% ethanol to create a 10 mM stock solution that was stored at -20ºC (H 7779, SIGMA, St Louis, MO). SEB-1 sebocytes were treated in triplicate with 0.1, 1, 10 µM concentrations. Furthermore, experiments were performed with RXR pan-agonist CD 3254 (Galderma R&D, Sophia Antipolis, France) CD 3254 was handled under normal light conditions and dissolved in DMSO to create a 10 mM stock solution that was stored at -20ºC until use. SEB-1 sebocytes were treated in triplicate with 1 and 50 nM concentrations. All compounds were diluted in standard culture mediums and applied for 48 and 72 hours. Each well was considered one sample. Samples were prepared by manufacturer’s instructions for In Situ Cell Death Detection Assay (Roche Applied Science, Indianapolis, IN). Additional assay controls included negative controls of labeling solution only and DNase I- treated wells as positive wells. Results were quantified by counting positive stained cells in 3 representative fields per well for each of the treatments carried out in triplicate. Each assay was performed three independent times; fenretinide and CD 3254 experiments were repeated twice. 64 Data analysis was performed using ANOVA Two Factor with Replication and considered significant if p < 0.05.

2.5.7 Quantitative Polymerase Chain Reaction (QPCR)

To verify RAR antagonist activity in TUNEL experiments, quantitative PCR was used to document down-regulation of the RAR target gene, tazarotene induced gene 1 (TIG1, retinoic acid receptor responder 1). SEB-1 sebocytes were handled, maintained and treated with 13-cis RA and RAR pan-antagonist AGN 193109 under conditions that were identical to those used in the TUNEL assays. Total RNA was isolated and QPCR performed as previously described (Trivedi et al, 2006). Primer-probe sets for TATA-binding protein (TBP; reference gene) and retinoic acid responder 1 (TIG1) were purchased from Applied Biosystems (Foster City, CA). Controls included “no template” and “no amplification” samples. The Relative Expression Software Tool (REST-XL) was used for data analysis.

2.5.8 HPLC

13-cis RA is reported to isomerize to ATRA in other cell types including SZ95 immortalized sebocytes (Tsukada, et al., 2000). To eliminate the possibility of an alternative pattern of isomerization and to study the kinetics of 13-cis RA uptake into SEB-1 sebocytes, we utilized liquid-liquid extraction, reverse phase HPLC with UV detection. SEB-1 sebocytes (passage 22) were grown to 80% confluence in 100-mm tissue culture plates. For “medium only” controls, SEB-1 medium alone was placed in empty 100-mm plates. 5 µM 13-cis RA was applied to SEB-1 sebocytes and “medium only” control plates in duplicate for 0, 2, 4, 6, 12, 18, 24, 48 and 72 hours. Experimental samples included medium collected from “medium only” control plates, medium from SEB-1 sebocyte plates and SEB-1 sebocyte cell pellet. Sample preparation was by liquid-liquid extraction with ethyl acetate. Ethyl acetate was evaporated and the residue was re-dissolved in a mixture of acetonitrile and purified water (80/20, vol/vol) before injection. Internal standard (acitretin), 13-cis RA, 9-cis RA and ATRA standards as well as quality control solutions were made and analyzed to generate the calibration curve. Samples 65 were injected into Agilent 1100 Series HPLC System (Agilent Technologies, Palo Alto, CA) using Nucleosil® 100-5 C18 (250 X 4mm2) HPLC columns (Macherey-Nagel Inc., Düren, Germany). Samples were eluted in a gradient solution composed of purified water and acetonitrile containing 0.2% acetic acid. Retinoid compounds were detected by UV detection at 350 nm.

We thank Drs. Johannes Voegel and Jean-Claude Caron of Galderma R&D for provision of compounds and HPLC analysis and Dr. Rosh Chandraratna of Allergan Inc. for provision of compounds. We also thank Nate Sheaffer of the Cell Science/Flow Cytometry Core Facility of the Section of Research Resources, Penn State College of Medicine, for excellent technical assistance with all FACS experiments and Anne Stanley of the Molecular Core Facility for assistance with densitometry analysis. Finally we thank Chelsea Billingsley for providing technical assistance.

This work is supported by NIH NIAMS R01 AR047820 to D.M.T. and the Jake Gittlen Cancer Research Foundation at the Pennsylvania State University College of Medicine.

Chapter 3

Array profiling of skin from patients on isotretinoin provides insights into potential mediators of its apoptotic effect on sebaceous glands.

3.1 Chapter Abstract

Although 13-cis retinoic acid (isotretinoin) is the most potent agent in the treatment of acne, its mechanism of action is still unknown. This is the first study to examine the effects of oral 13-cis retinoic acid on gene expression in the skin of acne patients. Gene expression analysis was performed on seven patients before and at one-week of treatment and also eight patients before and after 8-weeks of treatment as well as in our cell culture models: SEB-1 sebocytes and HaCaT keratinocytes for insights into cell specific effects of 13-cis RA. Histological estimation of sebaceous gland size was done and mRNA was isolated for gene array expression analysis using the Affymetrix system. Significant gene changes at 8-weeks are consistent with the known decreases in sebaceous gland lipid production induced by 13-cis RA. Significantly changed genes at one-week isotretinoin therapy can provide insight into the initial effects induced by this drug and these genes can be broadly categorized as tumor suppressors, protein processors or genes involved in transfer or binding of ions, amino acids, lipids or retinoids, including lipocalin 2. Lipocalins are small molecular weight proteins that regulate processes such as immune response, retinol transport, and prostaglandin synthesis. The lipocalin 2 gene product, neutrophil gelatinase associated lipocalin (NGAL), is known to function in innate immunity and can induce apoptosis. Immunohistochemistry on patient skin biopsies revealed increased NGAL after isotretinoin treatment with localization to the basal layer of the sebaceous gland and upper sebaceous duct. Our previous studies indicated that 13-cis RA can induce apoptosis in sebocytes. Purified NGAL protein induced significant apoptosis at 48 hours post-treatment in SEB-1 sebocytes, lending support to NGAL mediating the apoptosis actions of 13-cis RA. Together, these data provide rationale for further study of candidate genes, including lipocalin 2, that mediate retinoid response in the skin. 67 3.2 Introduction

13-cis retinoic acid (isotretinoin, 13-cis RA) is the most effective drug for the treatment of acne. It is a known teratogen whose use in the United States, as of March 1, 2006, is restricted within a registry system. The mechanism by which 13-cis RA reduces the size and secretion of sebaceous glands, normalizes follicular keratinization and improves acne is largely unknown. Its profound effect and improvement in acne was an unexpected finding during clinical trials for its use in ichthyosis (Peck, 1979). The lack of an animal model for acne has also deterred advances in the understanding of its mechanism of action. Because there are no safe alternatives to 13-cis RA that demonstrate comparable efficacy, insights into its mechanism of action are essential for alternative drug discovery. Transcriptional profiling represents a new powerful tool to examine changes in gene expression. When combined with advances in bioinformatics, transcriptional profiling can generate data that may target future hypothesis-driven investigation to specific genes or pathways. The goal of the present study is to gain broad insight into the potential pathways by which 13-cis RA exerts its clearing effect in acne. This is the first study to examine the effects of oral 13-cis RA on gene expression in the skin of acne patients. Array analysis was performed on skin samples which were taken from the backs of two cohorts of acne patients at baseline and after one-week or eight-weeks of isotretinoin therapy. The data generated distinct differences in the patterns of gene expression depending on the duration of therapy. In particular, marked decreases in the expression of genes involved in lipid metabolism were found at 8-weeks, which is in agreement with the marked histological decrease in sebaceous gland size. After one-week of treatment, a completely different profile emerged with significant changes in genes involved in differentiation, tumor suppression, serine proteases, serine protease inhibitors and solute transfer including lipocalin 2. In addition, gene array analysis was performed on our SEB-1 sebocytes and HaCaT keratinocytes, providing clues to possible cell-specific mechanisms of 13-cis RA. We explored in more detail the actions of lipocalin 2 in human skin and in our cell culture models, SEB-1 sebocytes and HaCaT and NHEK keratinocytes. Lipocalins are small molecular weight proteins that regulate processes such as immune response, retinol transport, and prostaglandin synthesis. The lipocalin 2 gene product, neutrophil gelatinase associated lipocalin (NGAL), is known to function in innate immunity and can induce apoptosis (Devireddy et al, 2005; Flo et al, 2004). Induction of lipocalin 2/NGAL may be one mechanism through which 13- 68 cis RA-induced apoptosis occurs in sebaceous glands. These data provide important clues to the effects of 13-cis RA that could advance our understanding of retinoid action not only in acne, but in other retinoid-responsive conditions such as psoriasis, leukemia and other cancers.

3.3 Results

3.3.1 Patient selection and procedures

A total of 15 patients that were prescribed isotretinoin by their dermatologist for their severe acne were enrolled in the study after signing informed consent forms. Early studies indicated that 13-cis RA drastically decreased sebaceous gland size after 16 weeks of treatment (Goldstein, et al., 1982); therefore, we chose an 8-week time point to examine the change in skin histology and gene expression in patients receiving the treatment for their severe acne. Eight patients had 5-mm punch biopsies of skin from their upper backs at baseline and at approximately 8 weeks into therapy. Sample analysis indicated a marked decrease in sebaceous gland size and decreased expression of numerous genes involved in lipid metabolism that are characteristic of sebaceous glands. In an effort to detect earlier gene changes, a second cohort of 7 patients was recruited to have biopsies preformed at baseline and one-week into treatment. Information regarding patient age, sex, the time of biopsy, and the dose of isotretinoin at the time of biopsy is presented in Table 5. The second biopsies from patients in Group 1 were performed during a regularly scheduled dermatology visit approximately 8 weeks into therapy. Due to scheduling issues, there was variation in the length of time at which the second biopsy was performed (9.12 ± 1.1 weeks). The doses of 13-cis RA depicted in Table 5 were those that the patients were receiving at the time of their second biopsy.

69 Table 5: Isotretinoin patient demographics

Group 1: 8 week study Group 2: 1 week study

Dose Biopsy Dose Biopsy Subject # Age Sex mg/kg/d (weeks) Subject # Age Sex mg/kg/d (days) 1 22 F 1 8 2 17 M 1 8 9 15 M 0.5 7 3 32 F 1 10 10 17 M 0.5 7 4 15 M 1 10 11 17 M 0.5 7 5 18 M 0.5 9 12 21 F 0.67 7 6 24 F 0.67 11 13 17 M 0.67 7 7 14 M 0.5 8 14 20 F 0.67 7 8 15 M 1 9 15 23 M 0.5 7 Mean ± SD 19.6 ± 6 0.83 ± 0.23 9.12 ± 1.1 Mean ± SD 18.5 ± 2.8 0.57 ± 0.09 7 ± 0

3.3.2 Histology reveals statistically significant decrease in sebaceous gland size after 8 weeks of treatment.

Hematoxylin and eosin staining was performed on sections of the baseline and treatment biopsies from both groups of patients (Figure 17a). In both cohorts, at baseline, sebaceous glands from the back were large and multi-lobulated. In Group 1 patients (8-week) treatment biopsies, sebaceous glands were markedly reduced in size by approximately 8.5-fold from baseline (Figure 17c). Glands lost their multi-lobular structure and sparse sebocytes were noted only in close proximity to hair follicles. This architecture and location closely mirrors the sebaceous glands in murine skin. For Group 2 patients (1-week), the changes in architecture were not as obvious, with only a decreasing trend in sebaceous gland size being noted (Figure 17b,c). 70

Figure 17: 13-cis RA decreases sebaceous gland volume. (a) Hematoxylin and eosin sections of back skin from patients before and after 8 weeks of treatment reveals a significant decrease in sebaceous gland volume. (b) Variable changes in sebaceous gland size were noted after one week of treatment compared to baseline biopsies (c) Area of sebaceous glands. Statistical significance was determined by paired t-test. Representative images are shown at a total magnification of 100X. Magnification bars = 250 µm.

71 3.3.3 Significant decreases in genes that regulate lipid metabolism were noted in the gene array expression analysis of skin biopsies taken from patients at 8 weeks into isotretinoin therapy.

In comparing the gene array data from the pre-treatment biopsies to the 8-week treatment biopsies, 197 genes were significantly up-regulated and 587 genes were significantly down regulated as determined by using a false discovery rate (FDR) of 0.05, that corresponds to a 5% chance of false positive genes among those genes considered significantly changed. Select genes that were up-regulated by approximately 2-fold or greater and genes that were down-regulated greater than 4-fold are listed in Table 6. For a complete listing of all significantly changed genes at 8-weeks, see A.1. Many of the down-regulated genes at 8-weeks are involved in the metabolism of steroids, cholesterol and fatty acids, which is consistent with the known decreases in sebaceous gland lipid production induced by 13-cis RA. 72 Table 6: Selected significantly changed genes after 8 weeks isotretinoin therapy

# Fold Change Gene Title* Symbol RARE 2.62 microseminoprotein, beta- MSMB RAR 2.03 collagen, type I, alpha 1 COL1A1 -6.31 hydroxyacid oxidase 2 (long chain) HAO2 -6.19 hydroxy-delta-5-steroid dehydrogenase, 3 beta HSD3B1 RAR -6.06 thioesterase domain containing 1 THEDC1 solute carrier organic anion transporter family, member -5.17 SLCO4C1 4C1 -5.1 male sterility domain containing 1 MLSTD1 RAR -5.09 phospholipase A2, group VII (PAF acetylhydrolase) PLA2G7 -4.9 fatty acid desaturase 1 FADS1 RAR,RXR -4.78 glycine dehydrogenase GLDC RAR, RXR -4.71 Galanin GAL RAR -4.57 PDZ domain containing 1 PDZK1 RAR, RXR -4.51 fatty acid binding protein 7, brain FABP7 -4.14 histone 1, H1c HIST1H1C -4.02 fatty acid binding protein 7, brain FABP7 -3.99 arachidonate 15-lipoxygenase, second type ALOX15B RAR -3.73 mucin 1, transmembrane MUC1 RXR -3.73 insulin induced gene 1 INSIG1 -3.68 3-hydroxy-3-methylglutaryl-Coenzyme A synthase 1 HMGCS1 RAR -3.6 Lipidosin BG1 RAR -3.59 sterol O-acyltransferase SOAT1 -3.37 fatty acid desaturase 2 FADS2 RAR, RXR -3.28 carnitine acetyltransferase CRAT RAR, RXR -3.25 hypothetical protein MAC30 MAC30 RAR, RXR -3.16 peroxisomal long-chain acyl-coA thioesterase ZAP128 RAR, RXR -3.08 chitinase 3-like 1 (cartilage glycoprotein-39) CHI3L1 RAR -3.03 apolipoprotein C-I APOC1 RAR,RXR -2.92 hydroxysteroid (11-beta) dehydrogenase 1 HSD11B1 -2.88 transmembrane protease, serine 11E TMPRSS11E -2.87 peroxisomal trans-2-enoyl-CoA reductase PECR -2.84 SA hypertension-associated homolog (rat) SAH -2.84 homogentisate 1,2-dioxygenase (homogentisate oxidase) HGD -2.83 dehydrogenase/reductase (SDR family) member 9 DHRS9 -2.83 steroid-5-alpha-reductase, alpha polypeptide 1 SRD5A1 -2.8 cell death-inducing DFFA-like effector a CIDEA RAR, RXR -2.78 SEC14-like 4 (S. cerevisiae) SEC14L4 RAR, RXR -2.77 Malic enzyme 1, NADP(+)-dependent, cytosolic ME1 RAR, RXR -2.73 phosphodiesterase 6A, cGMP-specific, rod, alpha PDE6A RAR -2.7 NAD(P) dependent steroid dehydrogenase-like NSDHL RAR, RXR -2.66 acetyl-Coenzyme A acetyltransferase 2 ACAT2 RAR -2.65 fatty acid 2-hydroxylase FA2H -2.61 farnesyl diphosphate synthase FDPS RAR -2.56 glycerol kinase GK RAR, RXR -2.52 3-hydroxy-3-methylglutaryl-Coenzyme A reductase HMGCR RAR *italics indicate genes involved in or linked to fatty acid or cholesterol metabolism # retinoic acid or rexinoid receptor response element located within first 1000 base pairs of gene

73 3.3.4 Gene expression analysis of skin from patients treated with 13-cis RA for one-week revealed significant increases in genes encoding calcium binding proteins, retinoid signaling molecules, solute carriers and serine proteases.

Array data was not generated for patient 11 due to insufficient quantity of RNA. In the array data from the remaining 6 patients, 38 genes were significantly up-regulated and 5 genes were significantly down-regulated when compared to baseline biopsies using a false discovery rate (FDR) of 0.05. Significantly changed genes are shown in Table 7. Many of the significantly up-regulated genes are known to be retinoid-responsive genes including: retinoic acid responder 1 [tazarotene induced gene 1 (TIG1)}, cellular retinol binding protein 1 and cellular retinoic acid binding protein 2. In addition, calcium-binding proteins (i.e. S100 proteins), serine proteases, serine protease inhibitors (serpins), lipocalin and solute carriers were significantly affected by 13-cis RA. 74 Table 7: Significantly changed genes after 1 week isotretinoin therapy

Fold Change Gene Title* Symbol RARE# 7.03 lipocalin 2 (oncogene 24p3) LCN2 RAR, RXR 6.2* S100 calcium binding protein A7 (psoriasin 1) S100A7 RAR, RXR 4.53 S100 calcium binding protein A9 (calgranulin B) S100A9 RAR 3.78 solute carrier family 12 (K/Cl transporters) SLC12A8 3.32 cytochrome P450, family 2, subfamily B CYP2B7P1 RAR 2.61 serine (or cysteine) proteinase inhibitor SERPINA3 RAR,RXR 2.61 retinoic acid receptor responder (TIG 1) RARRES1 2.35 transmembrane protease, serine 4 TMPRSS4 2.27 KIAA0125 KIAA0125 2.21 placenta-specific 8 PLAC8 2.08 Rhesus blood group, C glycoprotein RHCG RAR 2.04 pipecolic acid oxidase PIPOX RAR 1.99 S100 calcium binding protein P S100P 1.96 ERO1-like (S. cerevisiae) ERO1L RAR 1.92 ATPase, H+/K+ transporting, nongastric, alpha ATP12A 1.91 chemokine (C-C motif) ligand 2 CCL2 RAR, RXR 1.81 retinol binding protein 1, cellular RBP1 RAR, RXR 1.69 solute carrier family 6 (amino acid transporter) SLC6A14 1.67 E74-like factor 3 (ets domain transcription factor) ELF3 RAR 1.62 stimulated by retinoic acid gene 6 homolog STRA6 RAR, RXR 1.57 SEC14-like 2 (S. cerevisiae) SEC14L2 RAR, RAR 1.56 cellular retinoic acid binding protein 2 CRABP2 RAR, RXR 1.52 defensin, beta 1 DEFB1 1.51 calbindin 2, 29kDa (calretinin) CALB2 RAR, RXR 1.5 S100 calcium binding protein A2 S100A2 RAR 1.49 fucosyltransferase 3 FUT3 1.49 involucrin IVL RAR 1.49 interleukin 27 receptor, alpha IL27RA RAR, RXR 1.48 cytoplasmic polyadenylation element BP 1 CPEB1 1.43 CD24 antigen CD24 1.41 growth differentiation factor 15 GDF15 1.4 KIAA1462 KIAA1462 1.39 serine protease inhibitor, Kazal type 5 SPINK5 1.38 UDP-N-acetyl-alpha-D-galactosamine GALNT6 RAR 1.27 microtubule associated monoxygenase MICAL3 1.26 keratin 23 (histone deacetylase inducible) KRT23 -2.29 solute carrier family 26, member 3 SLC26A3 -2.27 phospholipase A2, group VII (PAF acetylhydrolase) PLA2G7 -2.13 phosphodiesterase 6A, cGMP-specific, rod, alpha PDE6A RAR -1.55 carboxypeptidase M CPM RXR -1.5 cysteine and glycine-rich protein 2 CSRP2 *italics indicated genes within chromosomal region 1q21: epidermal differentiation complex # retinoic acid or rexinoid receptor response element located within first 1000 base pairs of gene 75 3.3.5 Gene expression analysis in SEB-1 sebocytes and HaCaT keratinocytes with 72 hour 13-cis RA treatment.

Gene expression array analysis was performed on patient skin biopsies which contained mixed populations of cells. Our in vitro cell culture model, SEB-1, was used to examine sebocyte-specific gene changes induced by 13-cis RA. Three control samples and three samples treated with 0.1 µM 13-cis RA were analyzed. A total of 85 genes (78 different genes) were significantly influenced by 13-cis RA: 58 up-regulated and 27 down-regulated genes. Selected significantly changed genes are listed in Table 8. For a complete listing of all significantly changed genes, see A.2. The tumor suppressor, TIG1 and lipocalin 2 demonstrated the greatest changes in gene expression. In addition, there were changes in several genes involved in apoptosis and innate immunity such as TNFα-induced protein 2, TRAIL, interferon regulatory factor 1 (IRF1), interferon-induced proteins, NFκB, the death receptor, Fas and TIG3 (a.k.a. retinoic acid inducible gene 1 (RIG1)). TIG3 encodes an RNA helicase and represents a key intracellular protein that like the TLR3, can recognize viral double stranded RNA (dsRNA). (Sen and Sarkar, 2005; Yoneyama et al, 2004)

76 Table 8: Selected significantly changed gene in SEB-1 sebocytes after 13-cis RA treatment

Fold # Gene Title Gene Symbol RARE Change 12.25 retinoic acid receptor responder (tazarotene induced) 1 RARRES1 7.04 lipocalin 2 (oncogene 24p3) LCN2 RAR, RXR 5.95 tumor necrosis factor, alpha-induced protein 2 TNFAIP2 RAR, RXR 5.91 hydroxyprostaglandin dehydrogenase 15-(NAD) HPGD 4.64 cytochrome P450, family 1, subfamily B, polypeptide 1 CYP1B1 RAR 4.25 hydroxyprostaglandin dehydrogenase 15-(NAD) HPGD 4.18 tumor necrosis factor (ligand) superfamily, member 10 (TRAIL) TNFSF10 RAR 3.70 serine (or cysteine) proteinase inhibitor, clade B (ovalbumin), SERPINB3 RAR 3.43 insulin-like growth factor binding protein 3 IGFBP3 N/A 3.29 aldehyde dehydrogenase 1 family, member A3 ALDH1A3 3.22 retinoic acid receptor responder (tazarotene induced) 3 RARRES3 RXR 3.08 oxidised low density lipoprotein (lectin-like) receptor 1 OLR1 N/A 3.06 solute carrier family 1 (glial high affinity glutamate transporter) SLC1A3 3.00 growth differentiation factor 15 GDF15 RAR, RXR 2.60 cyclin-dependent kinase inhibitor 1A (p21, Cip1) CDKN1A RAR 2.42 interferon regulatory factor 1 IRF1 N/A 2.20 BTG family, member 2 BTG2 2.07 2',5'-oligoadenylate synthetase 1, 40/46kDa OAS1 1.85 GATA binding protein 3 GATA3 1.79 protein kinase C, alpha PRKCA RAR, RXR 1.70 nuclear factor of kappa light polypeptide gene enhancer in B-cells 2 NFKB2 RAR 1.69 Fas (TNF receptor superfamily, member 6) FAS RAR 1.66 dual specificity phosphatase 8 DUSP8 RAR, RXR 1.60 glutathione peroxidase 2 (gastrointestinal) GPX2 RAR 1.47 phosphoinositide-3-kinase, regulatory subunit 3 (p55, gamma) PIK3R3 RXR -3.12 dihydrofolate reductase DHFR N/A -2.48 glutamate dehydrogenase 1 GLUD1 N/A -2.27 ribonucleotide reductase M2 polypeptide RRM2 RXR -2.12 CD86 antigen (CD28 antigen ligand 2, B7-2 antigen) CD86 RAR, RXR -2.05 DNA replication complex GINS protein PSF1 PSF1 RXR -1.81 S100 calcium binding protein A10 S100A10 -1.71 phospholipase A2, group IVA (cytosolic, calcium-dependent) PLA2G4A -1.59 BUB3 budding uninhibited by benzimidazoles 3 homolog (yeast) BUB3

# retinoic acid or rexinoid receptor response element located within first 1000 base pairs of gene

In addition, we performed gene expression analysis on triplicate samples of HaCaT keratinocytes treated with 0.1 µM 13-cis RA or vehicle control for 72 hours. 13-cis RA induced a total of 54 significantly changed genes (47 different genes) using a false discovery rate (FDR) of 0.05: 53 up-regulated genes and only one down-regulated gene. Selected significantly changed 77 genes are listed in Table 9. For a complete listing of all significantly changed genes, see A.3. Results show unique gene changes that suggest HaCaT keratinocytes have powerful mechanisms in place to protect against retinoid induced apoptosis including TRIM31 that encodes an E3 ubiquitin ligase, and P450RAI2 (CYP26A), a potent retinoic acid 4-hydroxylase.

Table 9: Selected significantly changed genes in HaCaT keratinocytes after 13-cis RA treatment

Fold # Gene Title Gene Symbol RARE Change

3.56 lipocalin 2 (oncogene 24p3) LCN2 RAR, RXR 3.22 carcinoembryonic antigen-related cell adhesion molecule 5 CEACAM5 RXR 3.21 amiloride binding protein 1 (amine oxidase (copper-containing)) ABP1 3.21 cytochrome P450 retinoid metabolizing protein P450RAI-2 2.77 carcinoembryonic antigen-related cell adhesion molecule 6 CEACAM6 RXR 2.69 phospholipase A2, group X PLA2G10 RAR 2.33 fibulin 1 FBLN1 2.28 latexin protein LXN 2.11 kallikrein 6 (neurosin, zyme) KLK6 RAR 2.08 3'-phosphoadenosine 5'-phosphosulfate synthase 2 PAPSS2 RAR,RXR 1.96 plasminogen activator, tissue PLAT RAR 1.82 2',5'-oligoadenylate synthetase 1, 40/46kDa OAS1 1.81 GATA binding protein 3 GATA3 1.79 insulin-like growth factor binding protein 3 IGFBP3 RAR,RXR 1.76 nebulette NEBL 1.70 sarcospan (Kras oncogene-associated gene) SSPN 1.70 G protein-coupled receptor, family C, group 5, member B GPRC5B RAR 1.69 S100 calcium binding protein P S100P 1.67 prostaglandin I2 (prostacyclin) synthase PTGIS RAR, RXR 1.64 midkine (neurite growth-promoting factor 2) MDK RXR 1.64 annexin A9 ANXA9 1.63 involucrin IVL RAR 1.62 insulin-like growth factor binding protein 6 IGFBP6 RAR, RXR 1.54 phosphatidic acid phosphatase type 2A PPAP2A 1.48 lysophosphatidic acid phosphatase ACP6 1.34 retinoic acid induced 3 RAI3 1.30 2'-5'-oligoadenylate synthetase-like OASL RAR, RXR -2.10 Microfibril-associated glycoprotein-2 MAGP2

#retinoic acid or rexinoid receptor response element located within first 1000 base pairs of gene

78 3.3.6 QPCR verification of select genes from array analyses.

Quantitative real-time PCR for a select group of genes (based on fold changes) was performed on a subset of RNA samples (depending on the availability of remaining RNA from patient samples) in order to verify the direction and magnitude of the fold change in the 8-week and one-week array analysis. Additional QPCR experiments to verify select SEB-1 sebocytes and HaCaT keratinocytes gene changes were also performed. For verification of the changes noted in the one-array analysis, sufficient RNA was available from patients 9, 10, 13, 14, and 15 from among the 6 patients in this group. The genes verified included lipocalin 2 (LCN2), retinoic acid receptor responder 1 (RARRES1, TIG1), S100A7, serine protease inhibitor A3 (SERPINA3) and phospholipase A2 group 7 (platelet activating factor acetyl hydrolase) (Figure 18a). In the 8-week analysis, sufficient RNA was available from patients 1, 2, 4, and 7. Verification of changes in the level of expression of 3β- hydroxysteroid dehydrogenase (3βHSD1), HMG CoA reductase (HMGCR), phospholipase A2 group 7, insulin induced gene 1 (INSIG), carnitine acyltransferase (CRAT) and zinc finger binding protein 145 (ZBTB16) were performed by QPCR (Figure 18b). For the SEB-1 gene array, we verified LCN2, TIG1, insulin-like growth factor binding protein 3 (IGFBP3), GATA transcription factors 3 and 6 (GATA3, 6), ZBTB16 and solute carrier family member (SLC22A17) (Figure 18c). Four gene changes were verified for the HaCaT keratinocyte samples: LCN2, SLC22A17, RARRES1, and IGFBP3 (Figure 18d). Furthermore, QPCR analysis was performed on normal human epidermal keratinocytes (NHEK) treated with 0.1 µM 13-cis RA for 72 hours, even though gene expression analysis was not done. NHEK samples were used for comparison to HaCaT keratinocytes and as such, QPCR was performed for LCN2, SLC22A17, RARRES1, and IGFBP3 (Figure 18e). For all QPCR analyses, patient samples and cell line samples, the direction and magnitude of the change in expression for the selected genes is similar to that observed in the gene array analyses. 79

Using the computer software dChip (Li and Hung Wong, 2001), we performed hierarchical clustering of the entire set of genes that were significantly up- or down-regulated from our one-week and 8-week isotretinoin patient gene arrays. A two-way cluster analysis was done; the data was clustered by patient sample and also by genes exhibiting a similar expression profile. We found that the biopsy samples from the pre-treatment samples clustered into separate and distinct groups when compared with their respective after-treatment biopsies from both the one- and 8-week groups (Figure 19, one week data shown). Clustering of the patient samples (columns) is indicated at the top of the diagram. Genes with higher correlation coefficients among the standardized gene expression values across samples are clustered together by rows. Therefore, genes in the same cluster share similar expression patterns. A comparable analysis was performed with the SEB-1 gene array data; results showed control and 13-cis RA treated arrays clustering in separate and distinct groups (data not shown).

Gene expression analysis revealed numerous genes significantly affected by 13-cis RA treatment. In order to determine if 13-cis RA is preferentially influencing a particular subset of genes, we categorized them by the “dChip” computer software. Genes included on each array carry annotations that allow them to be grouped according to categories including “gene ontology”, “protein domains”, ”chromosomal location”, and “pathway”. It is important to note that annotations in each of these categories are not available for all genes on the arrays. Each category contains predefined terms for classifying genes. After hierarchical clustering, dChip assessed the significance of all functional categories within the cluster tree. The 42 genes changed in the one-week analysis and the 784 genes whose expression was significantly changed in the 8-week analysis were assessed for significant enrichment for the above listed categories. The p-value for this functional categorization is the probability of seeing x genes with a certain category occurring in a group of k genes at random, given n annotated genes on the array, of which m genes carry that specific annotation. Groups of genes mapping to a particular cluster with p < 0.001 were considered significant. From the 42 significantly changed genes at one-week, we identified 5 gene ontology terms, 3 protein domains and one chromosomal location that were enriched according to dChip (Table 10). The chromosomal location of 1q21 is the site of the epidermal differentiation complex. Genes that are located within this chromosomal region are indicated in italics in Table 7. “Pathway” analysis failed for this data; most likely due to the small number of significantly changed genes (only 42) and the fact that only three of these genes carried ”pathway” annotations.

Table 10: Functional categorization of significantly changed one-week genes. Gene Ontologies p-value Ectoderm development 0.0001 Epidermal development 0.00006 Response to pest, pathogen, or parasite 0.0007 Response to external biotic stimulus 0.0007 Vitamin binding 0.0006 Protein Domains Lipocalin related 0.00038 Calcium binding 0 Latexin 0 Chromosomal Location 1q21 0.00013

83 Within the 784 significantly changed genes at 8 weeks, we identified 98 gene ontology terms (data not shown), 21 protein domains, 0 chromosomal locations and 10 pathways that were significantly enriched (p < 0.001). Enriched protein domains are listed in Table 11; “anaphylotoxin/fibulin”, “fibronectin, type 1” and “collagen triple helix repeat” protein domains were significantly up-regulated in our list of genes, while all other protein domains were significantly down-regulated.

Table 11: Protein domains enriched within genes significantly changed at 8 weeks. # of genes in cluster/ Protein Domain Term total # of genes p-value with term on gene array Anaphylatoxin/fibulin 4 out of 9 0.00021 Fibronectin, type I 4 out of 10 0.00034 Collagen triple helix repeat 14 out of 104 3.2E-05 Polyprenyl synthetase 3 out of 5 0.000493 Carbohydrate kinase 6 out of 11 0.000001 AMP-dependent synthetase and ligase 11 out of 30 0 3-oxo-5-alpha-steroid 4-dehydrogenase 4 out of 6 0.000027 ERG4/ERG24 ergosterol biosynthesis protein 3 out of 4 0.000203 Cytochrome b5 11 out of 25 0 Peptidase T1, 20S proteasome 6 out of 25 0.000258 Enoyl-CoA hydratase/isomerase 5 out of 19 0.000543 Thiolase 4 out of 9 0.000211 Short-chain dehydrogenase/reductase 11 out of 51 0.000002 3-beta hydroxysteroid 4 out of 5 0.000009 dehydrogenase/isomerase H+-transporting two-sector ATPase, C subunit 5 out of 8 0.000121 CoA-binding domain 4 out of 4 0.000002 Transketolase, central region 5 out of 11 0.000028 Transketolase, C terminal 5 out of 8 0.000004 ATP-citrate lyase/succinyl-CoA ligase 4 out of 8 0.000121 Histone core 16 out of 80 0 Insulin-induced 3 out of 4 0.000203 * italics : protein domains that were significantly up-regulated

In the analysis of pathways, genes involved in each of the 10 pathways (Table 12) were down-regulated. Interestingly, each of these 10 down-regulated pathways is either directly involved in or linked to fatty acid and cholesterol metabolism, further supporting 13-cis RA’s role in sebum suppression. A subset of genes that mapped to these enriched pathways is indicated in italics in Table 6. 84 Table 12: Down-regulated pathways enriched within genes significantly changed at 8 weeks % of genes Pathway term p-value with RARE Electron transport chain 24 out of 135 0 76 Fatty acid degradation 12 out of 43 0.000003 87.5 Mitochondrial fatty acid β-oxidation 11 out of 31 0.000001 75 Krebs-TCA cycle 10 out of 54 0.000835 75

Reductive carboxylate cycle (CO2 fixation) 5 out of 14 0.000788 75 Fatty acid synthesis 8 out of 28 0.000121 75 Pentose phosphate//GenMAPP 5 out of 13 0.000531 75 Cholesterol biosynthesis 21 out of 30 0 75 Biosynthesis of steroids 15 out of 30 0 70 Terpenoid biosynthesis 6 out of 11 0.000012 66.7

3.3.9 Promoter analysis of genes

We analyzed the promoter regions of all significantly changed genes for one-week and 8-week patient array data as well as SEB-1 sebocyte and HaCaT keratinocyte arrays. The significantly changed genes containing retinoic acid receptor (RAR) or rexinoid receptor (RXR) response elements (RAREs) consensus sequences are indicated in their respective tables. Some genes did not have promoter sequences available from the Cold Spring Harbor Laboratory database and are indicated with N/A in each table. Of the selected 8-week changed genes listed in Table 2, 26 (60%) contained RAR or RXR consensus sequences. Interestingly, all genes located on chromosome 1q21 contain consensus sequences for retinoic acid receptors. A separate promoter analysis for the 8-week data indicated that of the 117 genes that mapped to the 10 enriched pathways, 75% of these genes contained RAR or RXR response elements in their promoters (Table 12). In the one-week analysis 21 of 42 (50%) significantly changed genes contained retinoid consensus sequences (Table 7). Of the 85 genes significantly changed in SEB-1 sebocytes, 39% (33) contained RAR or RXR consensus sequences in their promoters (Table 8) compared to 54% (29) of the 54 significantly changed genes in HaCaT keratinocytes (Table 9).

85 3.3.10 Comparison of gene changes at one-week and 8-week revealed only 3 common genes.

From among the 42 significantly changed genes at one-week and the 784 significantly changed genes at 8-weeks, only 3 common genes were found. These 3 genes were down- regulated 2- to 5-fold and include solute carrier family member 26, member 3 (SLC26A3); phospholipase A2, group VII, (PLA2G7); and phosphodiesterase 6A (PDE6A). In comparing the functional classification of the significantly changed genes, no commonalities were found between one-week and 8-week data with regard to “gene ontology”, “protein domain”, “pathway” or “chromosomal location”. The relative proportion of genes containing RAR or RXR consensus sequences is similar between one-week and 8-week data sets, 49% and 57%, respectively.

3.3.11 Comparisons between one-week, SEB-1 sebocytes and HaCaT keratinocytes array data revealed only one gene in common between all three arrays.

Pair-wise comparisons were made between significantly changed genes from the one- week isotretinoin, SEB-1 sebocytes, and HaCaT keratinocytes gene arrays (Table 13). Although not identical genes, family members of the S100 calcium binding proteins, and phospholipase A2 proteins were significantly changed in all three arrays. Surprisingly, 9 genes were found in common between SEB-1 sebocytes and HaCaT keratinocytes. 86 Table 13: Common significantly changed genes within one-week isotretinoin, SEB-1 sebocyte and HaCaT keratinocyte gene arrays.

One-week vs. SEB-1 sebocyte

Fold Fold change change SEB-1 Gene Name Symbol One-week sebocytes

1.41 3 Growth Differentiation Factor 15 GDF15 1.67 2.51 E74-like factor 3 ELF3 2.61 12.25 retinoic acid responder 1 RARRES1

One-week vs. HaCaT keratinocyte Fold Fold change change HaCaT Gene Name Symbol One-week keratinocytes 1.49 1.63 involucrin IVL 1.99 1.69 S100 calcium binding protein P S100P

SEB-1 sebocytes vs. HaCaT keratinocytes Fold Fold change change HaCaT Gene Name Symbol SEB-1 keratinocytes sebocytes 1.85 1.81 GATA binding protein 3 GATA3 2.07 1.82 2',5'-oligoadenylate synthetase 1, 40/46kDa OAS1 2.07 1.64 annexin A9 ANXA9 2.18 3.22 carcinoembryonic antigen-related cell adhesion molecule 5 CEACAM5 2.29 1.64 fucosidase, alpha-L-1, tissue FUCA1 2.74 3.68 tripartite motif-containing 31 TRIM31 3.43 1.79 insulin-like growth factor binding protein 3 IGFBP3 3.52 1.7 carcinoembryonic antigen-related cell adhesion molecule 1 CEACAM1 4.98 2.77 carcinoembryonic antigen-related cell adhesion molecule 6 CEACAM6

Within the 42, 85, and 54 significantly changed genes in the one-week, SEB-1 sebocyte and HaCaT keratinocyte gene array data, respectively, we identified only one gene in common among all three arrays: lipocalin 2. Lipocalin 2 is significantly up-regulated in all three arrays with approximately 7-fold increases within one-week and SEB-1 sebocytes, and 3.5-fold increase in HaCaT keratinocytes above their respective baseline or vehicle control samples. We chose to further characterize lipocalin 2 in human skin and in our cell lines because it was one of the most highly up-regulated genes by 13-cis RA in our entire study. Lipocalins are small molecular weight proteins that regulate processes such as immune response, retinol transport, and prostaglandin synthesis. The lipocalin 2 gene product, neutrophil gelatinase 87 associated lipocalin (NGAL), is known to function in innate immunity and can induce apoptosis (Devireddy, et al., 2005; Flo, et al., 2004; Tong et al, 2003).

3.3.12 Immunohistochemistry and western analysis showed increased NGAL expression after 13-cis RA treatment in patient skin and our cell lines, respectively.

When biopsies were taken from patients before and after isotretinoin treatment, a portion of the biopsy was formalin-fixed and paraffin-embedded for future studies, while the remainder was used for the aforementioned gene expression analysis. Based on availability, sections from one-week patients 9, 10, 13 and 14 were used. Immunohistochemistry on patient skin biopsies revealed increased NGAL after isotretinoin treatment with localization to the basal layer of the sebaceous gland and upper sebaceous duct (Figure 20).

Figure 20: NGAL increased after one-week isotretinoin treatment. Immunohistochemistry for NGAL on sections on back skin taken before and at one-week treatment reveals notable increase in NGAL expression in sebaceous gland and hair follicle. Sections were incubated overnight with a 1:50 dilution of mouse monoclonal lipocalin 2/NGAL antibody. Negative control sections omitted primary antibody. All sections were counterstained with hematoxylin. NC=negative control; pre=before treatment; post=after treatment; SG=sebaceous gland; and Fol=follicle. Representative images are shown. Total magnification: 400X 88 NGAL protein expression within our cell lines was verified by western blotting. SEB-1 sebocytes, HaCaT and NHEK keratinocytes were treated with 13-cis RA for 72 hours. NGAL protein expression increased approximately 10-fold in SEB-1 sebocytes with both concentrations of 13-cis RA tested when compared to control (p < 0.05). Non-significant fold increases of 1.17 and 1.40 with 13-cis RA concentrations of 0.1 µM and 1 µM, respectively, were noted in HaCaT keratinocytes with very similar results in NHEK (1.51- and 1.77-fold increases) (Figure 21). Of note is the higher level of NGAL expression within HaCaT and NHEK keratinocytes control samples when compared to SEB-1 control samples.

Figure 21: 13-cis RA increases NGAL protein expression. SEB-1 sebocytes, HaCaT keratinocytes and NHEK keratinocytes were treated with vehicle control or 13-cis RA (0.1 or 1 µM). Protein expression was verified by western blot. Blots were incubated with primary antibody to lipocalin 2 and β-actin for loading control normalization followed by densitometry. Graph represents normalized fold-change values relative to control expression for a minimum of six independent blots. Mean ± SEM.

89 3.3.13 Isotretinoin increased apoptosis in one-week patient sections.

Our previous work indicated that 13-cis RA induced apoptosis within SEB-1 sebocytes; most likely by an RAR independent mechanism (Nelson, et al., 2006). We were able to perform in vivo studies by using the baseline and one-week isotretinoin treated patient biopsy samples. The TUNEL-peroxidase assay was performed on sections from baseline and one-week treatment for patients 9 through 15. Patient 15 was omitted from analysis because sebaceous glands were not found in the sections (n = 6 pairs of samples). One week of isotretinoin treatment significantly increased the percentage of cells with TUNEL positive staining from 13.9% before treatment to 45.9% after treatment, an approximate 4-fold increase (p = 0.0001) (Figure 22). TUNEL staining is strongest in the basal nuclei and the early differentiated sebocytes, in a region similar to the distribution of NGAL.

Figure 22: TUNEL staining in sebaceous glands increased in patient skin after one-week isotretinoin treatment. Two representative “after isotretinoin” images are shown. Skin sections were obtained from paraffin blocks of patients 9-15 and were subjected to TUNEL-peroxidase assay, according to manufacturer’s instructions. Assay controls included DNase I treated positive and negative controls with primary antibody omitted in negative control. At least 2 sections from every patient (before and after) were analyzed. Sections were counter-stained with hematoxylin. Data represent mean ± SD, n=6 patients; paired t-test was used for statistical analysis. Total magnification 400X. 90 3.3.14 Purified NGAL protein induces apoptosis in SEB-1 sebocytes but not in HaCaT or NHEK keratinocytes.

The mouse form of NGAL, known as 24p3, has been shown to induce apoptosis in murine pro-B lymphocytic FL5.12 cells as well as other hematopoietic cells (Devireddy, et al., 2005). We assessed whether purified NGAL protein is capable of inducing apoptosis in our human cell lines using the TUNEL assay. SEB-1 sebocytes, HaCaT and NHEK keratinocytes were treated with increasing concentrations of purified recombinant human NGAL protein for 24 hours and cells were approximately 50% confluent at the time of assay. After 24 hours, NGAL significantly increased the percentage of TUNEL positive cells in SEB-1 sebocytes to a maximum of 35% with 1ng/mL NGAL treatment, but had no effect on HaCaT or NHEK keratinocytes (Figure 23). The percentage of cells in apoptosis with NGAL treatment in HaCaT keratinocytes never exceeded 1%, while approximately 5% of NHEK cells were TUNEL-positive in all treatments.

Figure 23: NGAL increases TUNEL staining in SEB-1 sebocytes SEB-1 sebocytes, HaCaT and NHEK keratinocytes were treated in duplicate with vehicle control, 1pg/mL, 10pg/mL, 1ng/mL and 10ng/mL purified recombinant human NGAL protein (R&D Systems) for 24 hours. (a) Representative images of SEB-1 sebocytes are shown. Total magnification 200X. (b) Quantification of the percentage of TUNEL positive stained cells per treatment at 24 hours. Data represent mean + SEM, n = 4-6. Statistical analyses were performed with ANOVA Two Factor with Replication. * p< 0.05, ** p < 0.01, *** p < 0.0001 91 3.3.15 Apoptosis induced by NGAL is mediated by specific NGAL receptor isoforms.

Recently, a cell surface receptor for 24p3/NGAL was identified in murine pro-B lymphocytic FL5.12 cells and the presence of this receptor is believed to be responsible for cell-specific susceptibility to apoptosis (Devireddy, et al., 2005). It is possible that NGAL functions similarly in our human cell lines. Based on sequence homology, the human homolog of 24p3R is predicted to be solute carrier member SLC22A17. We obtained the affinity purified antibody to the mouse 24p3 receptor and positive and negative control lysates from Dr. Michael Green (University of Massachusetts Medical School, Worcester, MA). SEB-1 sebocytes express the 24p3 R-Long (60kD) and the high molecular weight (HMW, 70kD) forms of the receptor. HaCaT keratinocytes express the 24p3 R-Short form (30kD) while NHEK do not express any form of the receptor (Figure 24). Interestingly, SEB-1, which expresses the 24p3R-L and HMW forms, undergoes apoptosis in response to 13-cis RA and NGAL treatment whereas HaCaT and NHEK do not. Treatment with 13-cis RA, which induces NGAL expression in all cell lines, halves the expression of 24p3R-L in SEB-1, decreases expression of HMW form by 3.5 fold in SEB-1 and increases 24p3R-S form in HaCaT keratinocytes. The functional significance of the various receptor forms is currently unknown; although, susceptibility to NGAL-induced apoptosis seems to correlate with expression of the high MW form (C. Gazin, Green lab, personal communication). 92

Figure 24: Cell-specific expression of 24p3R/NGAL-R isoforms is influenced by 13-cis RA. (a) Protein lysates (vehicle and 1µM 13-cis RA 48 hours) were immunoblotted with affinity purified 24p3 receptor antibody. Positive (+) and negative (-) control protein lysates obtained from Dr. Michael Green. Variable expression of receptor isoforms (short, long, and high molecular weight forms) are noted across cell lines and in response to 13-cis RA. (b) Relative quantification of isoforms. Blots were incubated with β-actin for loading control normalization followed by densitometry. Graph represents normalized fold-change values relative to control expression for three independent blots. Mean ± SD.

3.4 Discussion

The biological effects of 13-cis RA are complex and the pathway(s) by which sebum is decreased and acne is improved have yet to be fully elucidated. Early studies in the 1980’s demonstrated that 13-cis RA markedly diminishes the size and secretion of sebaceous glands after 16 weeks of isotretinoin treatment (Goldstein, et al., 1982). Our study demonstrates that 13-cis RA markedly decreases sebaceous gland volume by 8 weeks of treatment and a trend toward this reduction is apparent at one-week. This is consistent with the observations that sebum secretion can be markedly reduced by 13-cis RA as early as 2 weeks (Hughes and Cunliffe, 1994; Stewart et al, 1983). However, no study to date, has examined the global changes in gene expression that accompany these histological changes in the skin of patients treated with 13-cis RA. 93 Using gene expression analysis, we were able to demonstrate distinctly different patterns of gene expression at one-week and 8-weeks of treatment in patients receiving isotretinoin for severe acne. Hundreds of genes were significantly changed after 8-weeks of isotretinoin therapy. The preponderance of genes that were down-regulated involved lipid and sterol metabolism and are characteristically expressed, although not exclusively, within differentiated sebaceous glands (Thiboutot, et al., 2003). These changes at 8 weeks are consistent with the reduction in sebaceous gland size and volume that has been definitely demonstrated. The majority of genes that were up-regulated at this time point encode structural proteins of the extracellular matrix such as collagens, fibulin and fibronectin. These up-regulated genes are consistent with the known effects of retinoids on the extracellular matrix as reported in studies of photoaging (Weiss et al, 1988). It is clear that the down-regulation of genes involved in cholesterol and fatty acid metabolism and reduction in gland size are the net in vivo effects of 13-cis RA on sebaceous gland function. Gene changes at one-week are of particular interest because they may provide clues about the initial changes induced by this drug. Those early gene changes can be broadly categorized as tumor suppressors, protein processors, and genes involved in transfer or binding of ions, amino acids, lipids or retinoids. For example, tazarotene induced gene 1 (TIG1, retinoic acid responder 1 (RARRES1)) encodes a tumor suppressor belonging to the latexin family of proteins, whose promoter is methylated (CpG island) and therefore silenced in a variety of cancers (Mizuiri et al, 2005; Shutoh et al, 2005; Youssef et al, 2004; Zhang et al, 2004). The increase in expression of TIG1 induced by 13-cis RA may mediate the known effects of this drug in chemoprevention of skin cancer in addition to the known suppressive effects on sebocyte proliferation (Nelson, et al., 2006; Zouboulis, et al., 1991; Zouboulis, et al., 1993; Zouboulis, et al., 1999). Furthermore, genes encoding both serine proteases and serine protease inhibitors were up-regulated by 13-cis RA at one-week. Serine protease inhibitors (serpins) are involved in tissue remodeling and control of inflammation (Silverman et al, 2001). Increased expression of serpins have been reported in inflammatory processes such as psoriasis and inflammatory acne lesions (Takeda et al, 2002; Trivedi, et al., 2006). Since 13-cis RA is the most potent agent available to reduce severe inflammatory acne, it is possible that the up-regulation of serpins, which in turn, scavenge pro-inflammatory proteins, mediates the anti- inflammatory effect of 13-cis RA. In support of this hypothesis, our previous HPLC study demonstrated that 13-cis RA can be isomerized to ATRA within sebocytes and it has been 94 shown that SERPINA5 is capable of binding ATRA in vitro and may function in retinoid transport (Jerabek et al, 2001; Krebs et al, 1999). In addition, among the genes whose expression was significantly increased at one- week, there was an increased representation of genes within chromosomal location 1q21. CHR1q21 is the location of the epidermal differentiation complex and includes TIG1, fourteen of the S100 proteins, involucrin and cellular retinoid binding protein 2 among others. Retinoids are crucial to epidermal development and differentiation (Wolbach and Howe, 1925). S100 proteins modulate cellular differentiation, energy metabolism, cytoskeletal membrane interactions, and cell cycle progression (Eckert et al, 2004). S100 protein family members are up-regulated in our one-week patient array as well as our SEB-1 sebocytes and HaCaT keratinocytes gene arrays. Interestingly, the specific up-regulated S100 proteins are induced by oxidative or inflammatory stress and S100A7 (psoriasin) may function as a chemo-attractant agent for immune cells (Eckert, et al., 2004). One can speculate that initial up-regulation of S100 proteins by 13-cis RA is responsible for the “acne-flare” response observed in some patients receiving oral or topical retinoids for the treatment of their acne. Our array data confirm the increase in S100 protein noted with in vitro analysis of NHEK keratinocytes treated with 13-cis RA, 9-cis RA, ATRA and 4-oxo-13-cis RA (Baron, et al., 2005). The distinction between the patterns of gene expression induced by 13-cis RA at one- week and 8 weeks is substantiated by the finding that only 3 genes were commonly down- regulated at both time points. One of these genes, phosphodiesterase 6A (PDEA6) encodes the cyclic-GMP specific PDE6A alpha subunit, which is expressed in cells of the retinal rod outer segment. Mutations in PDE6A have been identified as one cause of autosomal recessive retinitis pigmentosa, associated with night blindness (Wang et al, 2001). Although night blindness is a known potential side effect of isotretinoin treatment, it has not been linked with changes in rod cell PDE6A. Gene expression analysis on patient samples is invaluable because it takes into account changes occurring in all compartments of the skin; changes that may not be reflected in isolated cell systems. To gain insight into any possible cell-type specific effects of 13-cis RA on gene expression we performed gene expression analysis on SEB-1 sebocytes and HaCaT keratinocytes. 13-cis RA actions are clearly specific to the cell type, as only 9 significantly changed genes were in common to both SEB-1 sebocytes and HaCaT keratinocytes. Within the one-week and SEB-1 sebocyte gene arrays, 13-cis RA significantly increased expression of multiple members of the solute carrier family of proteins. Recent data have 95 highlighted the importance of the solute carrier family of proteins in skin biology. For example, SLC12A8, which is up-regulated within our one-week analysis, encodes for a sodium/potassium/chloride transporter that has recently been identified as a candidate gene for psoriasis susceptibility contained within the PSORS5 locus of CHR3q (Hewett et al, 2002; Huffmeier et al, 2005). Although, no mutations in SLC12A8 have been linked to psoriasis, it is interesting to note that it is retinoid responsive. ATP12A, another significantly up-regulated gene at one-week, encodes sodium/potassium ATPase, an integral membrane protein involved in solute transport responsible for the hydrolysis of ATP coupled with the exchange of hydrogen and potassium ions across membranes. This particular protein regulates ion flux into melanosomes and affects tyrosinase activity (Watabe et al, 2004). Interestingly, another cation transporter (SLC24A5) has been identified as playing a prominent role in skin biology by regulating flux across the melanosome membrane in zebrafish and this gene links to skin color in humans (Lamason et al, 2005). No other studies to date have examined the other solute carrier molecules and their relationship to normal human skin or skin disorders. One gene was found in common between one-week, SEB-1 sebocyte and HaCaT keratinocyte gene expression arrays, lipocalin 2 (LCN2). LCN2 was one of the most highly up- regulated genes (approximately 7 fold in one-week and SEB-1 sebocytes and ~3.5 fold in HaCaT keratinocytes.) Lipocalins are small molecular weight proteins that regulate immune response, retinal transport, prostaglandin synthesis, renal tube morphogenesis, cell growth and metabolism (Flo, et al., 2004; Hanai et al, 2005; Newcomer and Ong, 2000; Yang et al, 2002). For these reasons, we studied LCN2 and its gene product neutrophil gelatinase associated lipocalin (NGAL) within human skin, SEB-1 sebocytes, HaCaT and NHEK keratinocytes. Lipocalin family members are found in all species including bacteria, plants and animals. Family members have very low amino acid similarity, but have three structurally conserved domains and are characterized by their ability to bind small hydrophobic, lipophilic molecules like retinol. Furthermore, most lipocalin family members are secreted proteins that bind to specific cell surface receptors. The prototypic lipocalin family member is retinol binding protein (RBP) (Akerstrom et al, 2000). Lipocalin 2 (NGAL, oncogene 24p3, uterocalin, α2-microglobulin related protein) is a 25kD secreted protein located on CHR9q34. Expression of LCN2/NGAL is ubiquitous with its expression being detected within bone marrow, uterus, prostate, salivary gland, stomach appendix, colon, trachea, and lung. NGAL expression has been documented within skin by Seo et. al., who showed its expression is restricted to inner root sheath and infundibulum of the hair follicle in normal skin (Seo et al, 2006). In line with previous studies, 96 immunohistochemistry on our one-week patient skin sections revealed increased NGAL expression within the basal layer of the sebaceous gland, upper sebaceous duct and hair follicle after 13-cis RA treatment. NGAL protein expression was increased in SEB-1 sebocytes, HaCaT and NHEK keratinocytes with 13-cis RA treatment. It is not altogether surprising that NGAL is increased in response to 13-cis RA as it has been shown to be increased with retinoid treatment in other model systems (Caramuta et al, 2006; Tong, et al., 2003). Documented ligands of NGAL include bacterial peptides, leukotriene B4, cholesterol oleate, retinol and retinoic acid (Akerstrom, et al., 2000; Kjeldsen et al, 2000). Recent studies into NGAL’s function focus on its anti-bacterial actions. Lipocalin knockout mice have an increased susceptibility to bacterial infections, but are otherwise healthy (Berger et al, 2006). NGAL can be secreted in response to Toll receptor activation by bacteria. NGAL binds bacterial siderophores and functions by sequestering iron from bacteria (Flo, et al., 2004). Since activation of Toll-like receptor 2 by P. acnes has been demonstrated in acne lesions it would be interesting to determine if NGAL released in response to 13-cis RA targets P. acnes. To this end, we performed preliminary Gram+ staining on skin sections for P. acnes bacterium before and after one-week isotretinoin therapy in patients 9, 10 and 13. Gram+ staining revealed P. acnes located at the base of the sebaceous gland and adjacent hair follicles (if present in the section) with a small amount extending into the middle of the sebaceous gland (data not shown). NGAL and P. acnes staining appear to localize to similar regions within the sebaceous gland and in 75% of sections with NGAL staining, P. acnes was also detected. Isotretinoin therapy causes a significant reduction in the gram positive, anaerobic P. acnes bacteria (including antibiotic resistant strains), with levels slowly returning to baseline after discontinuing treatment (Coates, et al., 2005; Leyden, et al., 1986). It remains to be determined if NGAL produced in response to 13-cis RA is capable of killing P. acnes, if P. acnes itself can stimulate NGAL production or a combination of both processes is occurring. Devireddy et. al. have demonstrated that secreted NGAL is capable of inducing apoptosis in F12.5 murine lymphocytes (Devireddy, et al., 2005). TUNEL staining on sections from the one-week patient biopsies demonstrated that 13-cis RA increased the percentage of cells undergoing apoptosis by 4-fold. This is the first in vivo evidence of 13-cis RA’s abilities to induce apoptosis within the human sebaceous gland. TUNEL staining is localized and strongest in the basal nuclei and the early differentiated sebocytes along the perimeter of the sebaceous glands. This distribution of TUNEL staining is remarkably similar to the localization of NGAL expression. Therefore it is possible that NGAL may be responsible for mediating the apoptotic 97 effect of 13-cis RA on sebaceous glands. Purified recombinant human NGAL protein induced apoptosis (TUNEL positive staining) within SEB-1 sebocytes. However, no apoptosis was detected within HaCaT or NHEK keratinocytes. Our in vitro cell culture models mirror the patient skin biopsies in that apoptosis is present within the sebaceous glands (sebocytes) and absent from the surrounding tissue or surface of the skin (keratinocytes). This data suggests it is possible for NGAL to be mediating the apoptotic effect of 13-cis RA on sebocytes. Lipocalins mediate their effects by binding to specific cell surface receptors. The recent identification of the mouse 24p3 receptor (24p3R) led to the identification of the highly conserved human homolog, SLC22A17, by GenBank searches (Devireddy, et al., 2005). The presence of this receptor correlated with apoptosis induction by NGAL. We obtained the purified 24p3R antibody from Dr. Michael Green and examined the expression of this receptor in our cell lines. We detected three different isoforms of this receptor. SEB-1 sebocytes express 24p3R-L and high MW (molecular weight) forms of this receptor. HaCaT keratinocytes express the 24p3R-S form of the receptor; this form of the receptor is an alternatively spliced variant lacking the first N-terminal 154 amino acid residues. NHEK keratinocytes do not express any form of the receptor. SEB-1 sebocytes undergo apoptosis in response to NGAL treatment and express the 24p3R-L and HMW forms of the receptor, while the keratinocytes which lack these forms of the receptor do not undergo apoptosis. Our findings suggest that susceptibility to NGAL-induced apoptosis correlates with expression of the 24p3R-L or HMW form of the receptor. Studies in Dr. Green’s lab also suggest that the HMW form is responsible for apoptosis sensitivity (C. Gazin, Green lab, personal communication). The identity/sequence of this high MW form remains unknown; although, phosphorylation, ubiquitination or glycosylation modifications of the 24p3R- L form are not suspected. Most intriguing is the retinoid-responsiveness of this receptor. Within SEB-1 sebocytes, 13-cis RA decreases the expression of 24p3R-L by approximately half and the HMW form all but disappears. To our knowledge, this is the first evidence of lipocalin cell surface receptors being regulated by retinoids. The functional significance of this decrease is unknown. It is exciting to speculate that the down-regulation of this receptor by 13-cis RA is a negative regulatory feed-back loop, turning off the apoptotic signal induced by the 13-cis RA up- regulation of NGAL. Future experiments are needed to test this hypothesis. The data presented in this study suggest that 13-cis RA initiates a temporal pattern of gene expression. Approximately half of the significantly changed genes in all gene arrays contain consensus sequences for RAR or RXR receptors. This study does not address the 98 question of whether the changes in gene expression are the direct result of 13-cis RA, one of its metabolites, or if retinoid receptors are involved. Instead, this study focused on the function of these significantly changed genes and how these functions may mediate the effect of 13-cis RA on sebaceous glands. It is clear that the down-regulation of genes involved in cholesterol and fatty acid metabolism and the reduction in size of the gland are the net in vivo effects of 13-cis RA on sebaceous gland function. Gene expression analysis at the one-week time-point provided insight into the initial changes induced by 13-cis RA, including up-regulation of NGAL. We further investigated NGAL localization and function and suggest that NGAL may, in part, mediate the apoptotic action of 13-cis RA on sebaceous glands. Our study provides rationale for further study of candidate genes, including lipocalin 2, that mediate retinoid response in the skin with the goal of discovering safer alternatives to oral retinoid use in the treatment of acne.

3.5 Materials and Methods

3.5.1 Patient selection and tissue biopsies

All protocols were approved by the Institutional Review Board of the Pennsylvania State University College of Medicine and were conducted according to the principles outlined in the Declaration of Helsinki. All subjects signed the informed consent form. Subjects included males and females ages 14 to 40 years who were scheduled by their dermatologist to receive treatment with 13-cis RA (isotretinoin, brand not noted) for severe acne. All aspects of the patients’ treatment with 13-cis RA apart from the skin biopsies were standard of care and were not part of this research. Exclusion criteria included patients on medications that could alter sebum excretion such as hormonal therapy (including oral contraceptives) or patients with underlying medical conditions requiring treatment with systemic medications that might interfere with the gene array analysis. In order to avoid multiple biopsies from each patient, two groups of patients were enrolled. Group 1 had a 5-mm punch biopsy of skin taken from their back before treatment and again after approximately 8 weeks of daily treatment with oral isotretinoin. Group 2 had biopsies before treatment and again at Day 7 of treatment (See Table 1 for treatment details). Biopsies were placed on ice and immediately transferred to the laboratory where they 99 were trimmed of fat and a small section of each biopsy was taken and paraffin-embedded for histology and immunohistochemistry. The remaining portion of the biopsy was flash frozen in liquid nitrogen and used for total RNA isolation.

3.5.2 Image analysis of sebaceous gland size

Following hematoxylin and eosin staining of sections from each of the biopsies, image analysis of sebaceous gland size was performed. Briefly, images were captured using a Spot digital camera (Diagnostic Instruments, Inc.) and measurements were obtained with Image Pro Plus Imaging Software Version 3.0 after spatial calibration with a micrometer slide under 10X magnification. All areas of sebaceous gland were circled using a free-hand measuring tool and the total area of the sebaceous gland was calculated in each section from the biopsy before treatment and from the biopsy during treatment for all subjects in the study. The mean area of sebaceous glands was calculated for each of the following groups: 8 week baseline, 8 week treatment, 1 week baseline and 1 week treatment. Paired t- tests (α =0.05) were performed to look for significant differences before and during treatment for the 8-week group and for the 1- week group.

3.5.3 Cell Culture

The SEB-1 human sebocyte cell line was generated by transfection of secondary sebocytes with SV40 Large T antigen as previously described (Thiboutot, et al., 2003). SEB-1 sebocytes were cultured and maintained in standard culture medium containing: 5.5mM Low Glucose DulBecco’s Modified Eagle Medium (DMEM) 3:1 Ham’s F12, 2.5% fetal bovine serum (FBS), hydrocortisone 0.4 µg/mL, adenine 1.8 X 10-4 M, insulin 10 ng/mL, epidermal growth factor (EGF) 3 ng/mL, cholera toxin 1.2 X 10-10 M, and antibiotics. HaCaT keratinocytes were cultured and maintained in 5.5 mM Low Glucose DMEM, 5% FBS and antibiotics. Normal Human Epidermal Keratinocytes-neonatal (pooled) [NHEK-neo, Clonetics Keratinocyte System, Cambrex Bioscience, Walkersville, MD] were cultured in Keratinocyte Growth Medium-2 (KGM- 2) (Cambrex Bioscience, Walkersville, MD). 100 13-cis RA (R 3255) was purchased through SIGMA (St. Louis, MO). Stock solutions of 13-cis RA were handled under dimmed yellow light, dissolved in 100% ethanol to a concentration of 10 mM and stored under N2 gas at -20ºC until use. Purified recombinant human NGAL protein (amino acids 21-198) was purchased from R&D Systems (Minneapolis, MN) ready to use and was stored at -20ºC until needed. Stock solutions were diluted to desired concentrations in standard sebocyte culture medium.

3.5.4 Gene expression microarray analysis

Total RNA was isolated from skin biopsies and DNase treated using the RNeasy Fibrous Tissue Kit (Qiagen Inc., Valencia, CA). Total RNA was isolated from SEB-1 sebocytes and HaCaT keratinocytes treated with 0.1 µM 13-cis RA or vehicle alone (0.001% ethanol) in three independent samples for 72 hours using a RNeasy kit (Qiagen Inc., Valencia, CA). Approximately 2µg of total RNA from each sample was used to generate double stranded cDNA using a T7-oligo (dT) primer. Biotinylated cRNA, produced through in vitro transcription, was fragmented and hybridized to an Affymetrix human U95Av2 microarray for SEB-1 sebocytes and U133A 2.0 microarray for all others. The arrays were processed on a GeneChip Fluidics Station 450 and scanned on an Affymetrix GeneChip Scanner (Santa Clara, CA). Expression signals were normalized as previously described (Irizarry et al, 2003; Irizarry et al, 2003; Trivedi, et al., 2006). Significant gene expression alterations were identified using Significance Analysis of Microarrays (SAM) computer software (Tusher et al, 2001). SAM controls the false positives resulting from multiple comparisons through controlling the false discovery rate (FDR) (Benjamini and Yekutieli, 2005). FDR is defined as the proportion of false positive genes among all genes that are considered significant.

3.5.5 Quantitative real-time polymerase chain reaction (QPCR)

Quantitative real-time PCR was performed to confirm the direction and magnitude of changes in the expression of select genes from the array data using Applied Biosystems’ Assays-on-Demand Taqman Universal PCR Master Mix and primer probe sets with ABI’s 7900HT Fast Real-Time PCR System with 384-well plate block module (Applied Biosystems, 101 Foster City, CA). Integrity of isolated RNA was verified by agarose gel electrophoresis. cDNA was generated from 1 µg of total RNA, primed with oligo-dT, using the Superscript First-Strand Synthesis System for reverse transcription-PCR (Invitrogen, Carlsbad, CA). Diluted cDNA samples were run for the reference gene TATA binding protein (TBP) as well as 6 genes of interest from the 8 week array analysis (3βHSD1, HMG CoA reductase 1, PLA2G7, INSIG1, Carnitine acyl transferase, ZFP145) and 5 genes of interest from the 1 week array analysis (lipocalin 2, [LCN2] RARRES1, S100A7, SERPIN A3, PLA2G7). For SEB-1, HaCaT and NHEK cell lines, 5-8 independent samples treated with 0.1 µM 13-cis RA or vehicle alone for 72 hours were analyzed by QPCR. Genes of interest included: LCN2, RARRES1, insulin-like growth factor binding protein 3 (IGFBP-3), GATA3, GATA6, ZBTB16, and LCN2 cell surface receptor (SLC22A17). Assay controls included samples omitting reverse transcriptase enzyme as well as samples without cDNA.

3.5.6 Cluster Analysis

Hierarchical clustering of patient samples and of significantly changed genes was performed using the normalized array data imported into dChip software version 1.3. The information files for the Affymetrix human genome HG-U133A 2.0 array was obtained from www.dChip.org. Separate cluster analyses were performed for one-week and 8-week patient gene arrays. Each row represents a single gene and each column represents a patient sample. (B=baseline and A=after treatment). The color reflects the level of expression when compared to the mean level of expression for the entire biopsy set. Red indicates expression higher than the mean and blue indicates lower expression than the mean.

3.5.7 Database promoter analysis of genes whose expression was significantly changed by 13-cis RA.

In an effort to understand which genes might be directly regulated by 13-cis RA or its metabolites 9-cis RA, which activates retinoic acid receptors (RARs) and rexinoid receptors (RXRs), or ATRA, which specifically activates RARs, the first 1000 basepairs (promoter regions included) of the top 136 genes with the greatest statistically significant fold change at 8-weeks were examined for retinoic acid response elements (RAREs). The first 1000 basepairs of all 102 genes significantly changed at one-week, within SEB-1 sebocytes and within HaCaT keratinocytes were also examined. The base pair sequences of each gene were obtained from the Cold Spring Harbor Laboratories Promoter Database (ftp://cshl.edu/pub/science/mzhanglab/PromoterSet) These sequences were scanned for RAREs using the predefined consensus sequences within the Transfac database through the Transcription Element Search System (TESS)

3.5.8 Comparisons of gene expression arrays

The one-week and 8-week array data was compared to identify common genes whose expression was influenced by 13-cis RA. The data from the functional categorization of significantly changed genes was similarly compared to identify common gene ontologies, protein domains, chromosomal locations or pathways that are affected by isotretinoin treatment. A comparison was also made between the percentage of significantly changed genes containing RAR or RXR consensus sequences. Pair-wise comparisons were also performed between one-week patient, SEB-1 sebocyte and HaCaT keratinocyte gene arrays to identify any significantly changed genes in common.

3.5.9 NGAL immunohistochemistry

Immunohistochemistry was performed on formalin-fixed paraffin-embedded human skin sections using the avidin-biotin complex method and AEC development (ABC kit and AEC Substrate Kit for Peroxidase, Vector Laboratories, Inc.; Burlingame, CA). Briefly, sections from patient (patients 9, 10, 13, 14) pre-treatment and after treatment biopsies were subjected to deparaffinization, rehydration and antigen retrieval prior to immunohistochemistry. Antigen retrieval was preformed using TRILOGY buffer (Cell Marque, Hot Springs, AR). Sections were incubated overnight with 1:50 dilution of mouse monoclonal Lipocalin 2 antibody (Abcam Inc, Cambridge MA). Negative control slides omitted primary antibody. Sections were counterstained with hematoxylin using standard procedures.

103 3.5.10 Western blotting

Mouse monoclonal antibody to lipocalin 2 (NGAL), used at 1:1000 dilution with overnight incubation (4ºC), was obtained from Abcam, Inc., (Cambridge, MA). Affinity purified 24p3 R (NGAL receptor) antibody, used at 1:2000 dilution overnight incubation (4ºC), was kindly provided by Dr. Michael Green (Howard Hughes Medical Institute, MA). Anti-rabbit horseradish peroxidase (HRP) linked secondary antibody was purchased from Cell Signaling Technology (Beverly, MA). Secondary Anti-mouse HRP antibody was purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Protein levels of NGAL and its receptor were examined using western blotting in our various cell lines as previously described (Nelson, et al., 2006). SEB-1 sebocytes (passages 22-26), HaCaT keratinocytes (passages 20-25) and NHEK (passage 3) were grown in 100-mm plates in standard culture medium until 50-75% confluent. Plates were rinsed with phosphate buffered saline and then treated with: 13-cis RA (0.1 µM, or 1 µM,) or ethanol vehicle (0.01% or less) as a negative control. Cells were treated for 48 or 72 hours. For NGAL-R blot, positive (Madin-Darby canine kidney; MDCK) and negative control lysates were obtained from Dr. Michael Green (University of Massachusetts Medical School, Worcester, MA). Blots were incubated with appropriate primary and secondary antibodies. SuperSignal West Pico Chemiluminescent Substrate (Pierce, Rockford, IL) was used for protein detection. β-actin was used as a loading control. Films of blots were analyzed and quantified by densitometry with QuantityOne Software (Bio-Rad, Hercules, CA.) after background subtraction. Western blots were repeated a minimum of 3-6 independent times. Data was analyzed using a Student’s t-test and results were considered significant if p < 0.05.

3.5.11 TdT-mediated dUTP Nick End Labeling (TUNEL) staining.

Sections from one-week before and after treatment biopsies from Patients 9-15 were used. Sections of skin were subjected to deparaffinization with xylenes, rehydration with graded ethanol series and permeabilization with 0.1% Triton-X, 0.1% sodium citrate in phosphate buffered saline according to manufacturer’s instructions. Sections were subjected to "In Situ Cell 104 Death Detection, Peroxidase" followed by counter staining with hematoxylin. Assay controls included DNase I treated positive and negative controls with primary antibody omitted in negative control. At least 2 sections from every patient (before and after) were analyzed. Results were analyzed and quantified by counting positive staining cells / total cells in sebaceous glands. Data represents mean ± SD, n = 6 patients; paired t-test was used for statistical analysis and considered significant if p < 0.05. SEB-1 sebocytes (passage 22-25); HaCaT keratinocytes (passage 24-25) and NHEK keratinocytes (passage 1) were cultured in 12-well plates in standard medium until approximately 30-40% confluent. Wells were rinsed with phosphate buffered saline (PBS) and were treated in duplicate with vehicle control, 1 pg/mL, 10 pg/mL, 1 ng/mL, or 10 ng/mL of recombinant human NGAL for 24 hours. Each well was considered one sample. Samples were prepared by manufacturer’s instructions for In Situ Cell Death Detection Kit, Fluorescein (Roche Applied Science, Indianapolis, IN). Assay controls included DNase I treated positive and negative controls with primary antibody omitted in negative control. Results were analyzed and quantified by counting positive staining cells / total cells in 3 representative fields per well for each of the treatments done in duplicate. Each assay was performed 2-3 independent times. Data analysis was performed using ANOVA Two Factor with Replication and considered significant if p < 0.05.

Chapter 4

Mechanisms involved in induction of apoptosis in SEB-1 sebocytes:

Activation of the extrinsic death receptor pathway by Tumor Necrosis Factor related apoptosis inducing ligand (TRAIL)

4.1 Chapter Abstract

Our previous work has shown that 13-cis RA induces apoptosis within SEB-1 sebocytes. Gene expression analysis indicates that 13-cis RA induces gene expression of mediators of apoptosis, particularly those involved with the extrinsic pathway including Tumor Necrosis Factor related apoptosis inducing ligand (TRAIL). Induction of TRAIL by 13-cis RA was confirmed by QPCR and western blotting. Furthermore, we verified that purified recombinant human TRAIL protein induces apoptosis within SEB-1 sebocytes confirming that TRAIL receptors are present on the surface of SEB-1 cells and the TRAIL apoptosis pathway is intact within our cell system. In addition, using RNA interference technology, we demonstrated that by decreasing TRAIL mRNA and protein expression in the presence of 13-cis RA leads to decreased cleaved caspase 3 protein, a marker for apoptosis induction. These data indicate that 13-cis RA preferentially activates the extrinsic apoptosis pathway through up-regulation of TRAIL and suggests that TRAIL plays a role in mediating apoptosis induced by 13-cis RA in SEB-1 sebocytes. Elucidating the cellular processes and pathways that are affected by 13-cis RA in sebocytes is a step toward understanding the overall molecular mechanism of action of this drug, which may lead to the identification of alternative strategies for the treatment of acne. 106 4.2 Introduction

The mechanism of sebosuppression by 13-cis RA is poorly understood. Although as early as the 1980’s, histological studies showed that isotretinoin dramatically reduces the size and secretion of the sebaceous gland in animal models. Previous work in our laboratory suggests that 13-cis RA decreases sebum production by influencing cell cycle and apoptotic pathways (Nelson, et al., 2006). Apoptosis, or programmed cell death, is absolutely essential during development when organisms need to eliminate unwanted cells in an orderly manner without invoking the inflammatory response. Apoptosis is a highly-regulated, well-orchestrated series of events culminating in nuclear condensation, DNA fragmentation, membrane-blebbing, cell shrinkage, and eventually phagocytosis of the dying cell (Wyllie, et al., 1980). Many different cellular stimuli can induce apoptosis including growth factor withdrawal, cellular stress, irreversible DNA damage, cytokine signaling and interference with pro-survival pathways. Two pathways of apoptosis have been well characterized: the death receptor (extrinsic) and the mitochondrial (intrinsic) apoptosis pathways. These two pathways converge with activation of caspase 3 and in some cells by activation of the protein Bid, a Bcl-2 family member. In this chapter, we investigated the apoptotic pathways activated by 13-cis RA. We were guided by gene expression analysis that indicated that 13-cis RA induces key genes involved in apoptosis. We report that 13-cis RA treatment up-regulates expression of Tumor necrosis factor Related Apoptosis Inducing Ligand (TRAIL) within SEB-1 sebocytes. Treatment with recombinant human purified TRAIL protein increased cleaved caspase 3 protein indicating that TRAIL induces apoptosis in SEB-1 sebocytes on its own. Utilizing siRNA knockdown, we have successfully inhibited TRAIL expression induced by 13-cis RA. In addition, knockdown of TRAIL protein decreased 13-cis RA induced-cleaved caspase 3 protein expression; thus indicating that TRAIL plays a role in mediating apoptosis induced by 13-cis RA in SEB-1 sebocytes. 107 4.3 Results

4.3.1 13-cis RA up-regulates genes involved in apoptosis in SEB-1 sebocytes.

Gene expression analysis of SEB-1 sebocytes treated with 0.1 µM 13-cis RA for 72 hours revealed significant up-regulation of genes involved in apoptosis (Table 14). Most interestingly, expression of tumor necrosis factor related apoptosis inducing ligand (TRAIL, TNFSF10) and TNF superfamily member 6, also known as Fas (CD95) receptor were increased. Both TRAIL and Fas exert their apoptotic effects through the extrinsic (death receptor) pathway of apoptosis, with Fas being a death receptor within the plasma membrane and TRAIL being a potent ligand (soluble or membrane bound) that binds to TRAIL-R1 and TRAIL-R2 receptors (Smith, et al., 2003). Up-regulation of both of these molecules by 13-cis RA suggests that 13-cis RA induces apoptosis via a death receptor pathway in sebocytes. In comparison, neither TRAIL nor Fas is significantly increased in HaCaT keratinocytes, nor do these cells undergo apoptosis in response to 13-cis RA (For significantly changed genes within HaCaT keratinocytes, see A.3).

Table 14: Genes involved in apoptosis whose expression is significantly changed by 13- cis RA in SEB-1 Fold Change Gene Symbol 12.25 tazarotene induced gene TIG1 7.04 lipocalin 2 (oncogene 24p3) LCN2 4.18 tumor necrosis factor apoptosis inducing ligand TNFSF10 3.43 insulin-like growth factor binding protein 3 IGFBP3 3.22 tazarotene induced gene TIG3 3.00 growth differentiation factor 15 GDF15 2.98 SRY (sex determining region Y)-box 4 SOX4 2.60 cyclin-dependent kinase inhibitor 1A (p21, Cip1) CDKN1A 2.42 interferon regulatory factor IRF1 1.7 Fas (TNF receptor superfamily, member 6) (CD95) Fas nuclear factor of kappa light polypeptide gene enhancer 2 1.70 (p49/p100) NFKB2

108

4.3.2 13-cis RA increases cleaved caspase 8 to a greater extent than 9-cis RA or ATRA.

To examine the possibility of death receptor pathway activation in SEB-1 sebocytes in response to 13-cis RA, caspase 8 protein expression was studied via western blot. Previous studies indicated that maximum apoptosis occurs at 72 hours post treatment with 13-cis RA. If activation of the death receptor pathway is responsible for this apoptosis, then caspase 8 cleavage and activation would likely occur prior to 72 hours. Therefore, we examined the expression of full length (p57), intermediate (p43/41) and active (p18) fragments of caspase 8 at 24 and 48 hours following 13-cis RA treatment. No cleaved caspase 8 was detected at 24 hours (data not shown). At 48 hours, 13-cis RA increases expression of active caspase 8 (p18) by approximately 7-fold compared to control (Figure 25). Experiments with 13-cis RA will be repeated at least one more time before statistical significance of fold change will be determined. Previous studies demonstrated that 13-cis RA induces apoptosis in SEB-1 sebocytes while 9-cis RA and ATRA do not. It is possible that 13-cis RA is more potent in activation of apoptotic pathways. Therefore, expression of full length (p57), intermediate (p43/41) and active (p18) fragments of caspase 8 at 48 hours following 9-cis RA or ATRA treatment was examined. In contrast to 13-cis RA, 9-cis RA and ATRA only increased expression of p18 by ~2-fold (Figure 25). Immunoblotting was performed with 1 µM 13-cis RA, 1 µM 9-cis RA and 1 µM ATRA protein lysates as these were available at the time of these experiments. Additional immunoblots will be performed with 0.1 µM concentrations in future experiments. 109

Figure 25: Increased active caspase 8 with 13-cis RA treatment. Protein lysates from SEB-1 sebocytes treated with 1 µM concentrations of 13-cis RA, 9-cis RA and ATRA or vehicle control (0.01% ethanol) and 1 µM) for 48 hours were immunoblotted with mouse Caspase 8 antibody (Cell Signaling Technology). A representative blot is shown. Graph represents mean ± SD fold- change values normalized to control for 2 independent blots (13-cis RA) and 4 independent blots (9-cis RA and ATRA)

4.3.3 13-cis RA increases TRAIL expression in SEB-1 sebocytes

Gene expression array analysis indicates that TRAIL is up-regulated 4.18-fold compared to control at 72 hours in SEB-1 sebocytes treated with 0.1 µM 13-cis RA. To confirm this finding and determine the time-line of TRAIL expression, we performed quantitative PCR (QPCR). Using QPCR, TRAIL mRNA is up-regulated by 3.792-, 3.259- and 1.883-fold after 24, 48 and 72 hours of treatment with 0.1 µM 13-cis RA, respectively. Fold-changes and statistical significance were determined with Relative Expression Software Tool (REST-XL) computer software program and as such, the algorithms used to compare data sets do not permit determination of standard error (Figure 26a). 110 TRAIL is a type II transmembrane protein expressed on the surface of cells. It also exists in a soluble form. Both the membrane form and the soluble form have been shown to induce apoptosis in a variety of tumor cell lines (Wiley et al, 1995). TRAIL protein expression at 72 hours was confirmed in SEB-1 by immunoblotting with available protein lysates from cells treated with 1µM 13-cis RA. Preliminary experiments indicate that the expression of the membrane and soluble forms of TRAIL protein is ~4.5-fold higher with 1µM 13-cis RA than control at 72 hours (Figure 26b). An additional experiment was performed at a later date in SEB-1 cells treated with 1 µM 9-cis RA and ATRA. TRAIL protein expression (membrane + soluble forms) is slightly increased with both treatments (approximately 1.5-fold increase) when compared to control (Figure 26b). Of note, the 9-cis RA and ATRA immunoblot was performed with slightly modified transfer conditions and a different lot number of TRAIL antibody which may explain the differences in control TRAIL protein levels between this blot and the 13-cis RA blot. Additional experiments with all three treatments together on the same blot need to be run to validate these perceived differences in TRAIL protein levels. Immunoblots will also be performed with 0.1 µM retinoid concentrations in future experiments.

111

Figure 26: TRAIL expression is increased by 13-cis RA treatment in SEB-1 sebocytes. (a) QPCR was performed for TRAIL at 24, 48 and 72 hours. Bars represent mean fold change of 3 independent samples as determined by REST-XL software. (b) TRAIL protein expression at 72 hours. Preliminary western blot shown. Graph shows relative level of TRAIL protein for one experiment to date.

4.3.4 TRAIL increases levels of cleaved caspase 3 in SEB-1 sebocytes

TRAIL-mediated apoptosis is initiated by binding of TRAIL to a cell surface receptor, TRAIL-R1 (DR4) or TRAIL-R2 (DR5). TRAIL-R1 or TRAIL-R2 genes were not included on our initial U95Av2 Affymetrix microarray chips. To determine if TRAIL can induce apoptosis in SEB- 1 sebocytes purified recombinant human TRAIL protein (R&D Systems, Minneapolis, MN) was added to our cells for 24 and 48 hours. Purified TRAIL protein induced apoptosis as evidenced by increased cleaved caspase 3 protein expression (fragment sizes p19 and p17) as early as 24 hours with statistically significant increases by 48 hours (Figure 27, 48 hour data shown). At 24 hours, cleaved caspase 3 protein expression was marginally increased from 1.1 to 2-fold compared to control. By 48 hours, TRAIL concentrations of 6ng/mL and higher, significantly 112 increased cleaved caspase 3 protein levels by at least 4-fold when compared to control. This result indicates that TRAIL receptors are present on the surface of SEB-1 cells and the TRAIL apoptosis pathway is intact within our cell system.

4.3.5 siRNA knockdown of TRAIL inhibits activation of caspase 3 by 13-cis RA.

RNA interference (RNAi) technology allows investigators to take advantage of normal cellular processes that recognize and degrade double-stranded RNA in a potent anti-viral response. By introducing specific small interfering RNA (siRNA) molecules, knockdown of target gene expression can be achieved (Dillon et al, 2005). We utilized ON-TARGETplus SMARTpool TRAIL and siCONTROL siRNA reagents (Dharmacon, Lafayette, CO) to examine the role of TRAIL in 13-cis RA induced apoptosis in SEB-1 sebocytes. siRNA experiments were conducted with two concentrations of siRNA, either 0.7 µg or 1.5 µg, to determine the optimal concentration necessary for specific “knock-down”. Twenty-four hours after nucleofection with TRAIL or siCONTROL (non-targeting) siRNA, cells were treated with 0.1 µM 13-cis RA to induce TRAIL expression. 113 Preliminary QPCR experiments with the 0.7 µg dose of siRNA to TRAIL indicated that TRAIL mRNA expression is decreased by 2.8-fold 24 hours after treatment with 0.1µM 13-cis RA compared to siCONTROL samples. At 48 and 72 hours of 13-cis RA treatment, TRAIL expression is decreased by approximately ~1.4 fold. Using 1.5 µg of siRNA to TRAIL in the same experiment where cells were treated with 0.1 µM 13-cis RA, TRAIL mRNA levels were decreased by 1.5 fold at 24 hours and a maximum decrease of 1.8 fold was found at 72 hours of 13-cis RA treatment (Figure 28).

24 hr 48 hr 72 hour 0.7 ug 1.5 ug 0.7 ug 1.5 ug 0.7 ug 1.5 ug 0

-0.5

-1

-1.5

-2

-2.5 Fold-change in TRAIL Fold-change mRNA -3

Due to limited amount of protein obtained from siRNA nucleofected samples, only one immunoblot could be run, therefore, the blot was incubated with both TRAIL and cleaved caspase 3 antibodies simultaneously, to prevent any loss of protein that may occur with “stripping” the blot. Both antibodies have been previously used successfully in separate experiments and no extraneous bands were noted that would interfere with interpretations of the results. TRAIL protein levels (membrane and soluble) were decreased with 1.5 µg of siCONTROL when compared to 0.7 µg siCONTROL (Figure 29, lanes 1 and 2, bands indicated by), which indicates that 1.5 µg dose is too high, leading to non-specific knockdown effects. 114 Furthermore, TRAIL protein levels with 1.5 µg siRNA to TRAIL were increased (lane 4), suggesting that this high dose of siRNA is toxic to the cells. Combined, these results suggest that the lower dose of siRNA is more effective for specific “knock-down” of TRAIL. Levels of membrane and soluble TRAIL were examined by western blot in cells that were nucleofected with two concentrations of siRNA to TRAIL and then treated for 24 and 48 hours with 0.1 µM 13-cis RA. The membrane form of TRAIL has a molecular weight of 32 kD whereas the molecular weight of the soluble form is 21kD. At 24 hours, no difference in the level of TRAIL protein (membrane and soluble) was noted between siCONTROL and siRNA to TRAIL samples (data not shown). At 48 hours, however, TRAIL protein levels were decreased by ~2-fold in cells nucleofected with 0.7 µg of siRNA to TRAIL when compared to the corresponding siCONTROL sample (Figure 29, lanes 1 and 3 bands indicated by). These results are from one experiment only and additional experiments to verify the decrease of TRAIL mRNA and protein levels with the 0.7 µg siRNA concentration will be performed. To determine if TRAIL “knockdown” affects apoptosis in SEB-1 sebocytes, total protein lysates from siRNA samples were subjected to immunoblotting with cleaved caspase 3 antibody which detects the cleaved/active forms (p19, p17) of caspase 3. No active caspase 3 protein was detected with 24 hours of 13-cis RA treatment (data not shown). This result is similar to previous studies in SEB-1 sebocytes. However, after 48 hours of 0.1 µM 13-cis RA treatment, active caspase 3 protein was decreased by ~2-fold in 0.7 µg siRNA TRAIL samples when compared to siCONTROL samples of the same dose (Figure 29, lanes 1 and 3, bands indicated by ). Active caspase 3 was decreased by ~1.5 fold with siRNA TRAIL concentration of 1.5 µg when compared with control of the same dose (Figure 29, lanes 2 and 4, bands indicated by ). Results are from one experiment only and additional experiments will be performed. 115

siCONTROL 2 1.2 siCONTROL siTRAIL 1.8 TRAIL 1 1.6

1.4 0.8 1.2 1 0.6 0.8 0.4 0.6 0.4 Relative level TRAIL protein TRAIL level Relative 0.2

0.2 3 caspase (active) cl. level Relative 0 0 0.7 ug 1.5 ug 0.7 ug 1.5 ug siRNA concentration siRNA concentration Figure 29: siRNA knockdown of TRAIL inhibits active caspase 3 protein expression. siCONTROL and TRAIL siRNA (2 concentrations) were nucleofected into SEB-1 sebocytes using Amaxa Nucleofection Kit T and nucleofector device. Twenty-four hours later, 0.1 µM 13- cis RA was added to induce TRAIL expression. Protein was isolated at 24, 48 and 72 hours of 13-cis RA treatment and subjected to immunoblotting with TRAIL (1:500) and cleaved caspase 3 (1:800) antibodies. Graphs represent normalized relative levels (compared to control) of TRAIL or cleaved caspase 3 protein of one sample.

In summary, siRNA to TRAIL decreases both TRAIL mRNA and protein levels. When TRAIL levels are decreased, cleaved caspase 3 protein levels are also decreased. These results suggest that TRAIL may be an important mediator of13-cis RA-induced apoptosis in SEB-1 sebocytes; however, as results are from one experiment only, additional experiments are needed to confirm this finding. 116 4.4 Discussion

The mechanism of 13-cis RA- induced apoptosis in SEB-1 sebocytes and human sebaceous glands is poorly understood. Elucidating the possible cellular pathways that are affected by 13-cis RA in sebocytes is a step toward understanding the overall molecular mechanism of action of this drug, which may lead to the identification of alternative strategies for the treatment of acne. To this end, studies in this chapter focused on determining which apoptosis pathway is activated by 13-cis RA in SEB-1 sebocytes. Gene expression analysis in SEB-1 sebocytes revealed significant up-regulation of a number of genes involved in apoptosis. The up-regulation of TRAIL and Fas suggests that 13-cis RA may be activating the extrinsic apoptosis pathway. Death receptors on the cell’s surface detect extracellular stimuli and, upon binding of their respective ligands, rapidly activate an intracellular caspase signaling cascade that results in apoptosis. The ligands for death receptors include Tumor necrosis factor Related Apoptosis- Inducing Ligand (TRAIL) and Fas ligand (FasL, CD95L). All death receptors including Fas (CD95) and TRAIL-R1/R2 have an intracellular region termed the “death domain (DD)”. It is this specific 80 amino acid sequence that allows the transmission of the apoptotic signal. DDs in the receptor recruit intracellular adaptor molecules (also containing DDs) that have “death effector domains (DED)”. DEDs recruit and activate the “initiator” caspases 8 and 10 by cleavage. Initiator caspases proceed to activate “effector” caspases 3, 6, and 7, which by cleavage of their specific substrates (i.e. PARP, α-fodrin) result in apoptosis of the cell. For example, TRAIL- mediated apoptosis is initiated by binding of TRAIL to a cell surface receptor, TRAILR1 (DR4) or TRAILR2 (DR5), which then recruits caspase 8 via the adaptor molecules, TNF-R1 associated death domain protein (TRADD) and Fas-associated death domain (FADD). Activated caspase 8 directly activates caspase 3, caspase 6, or caspase 7 or activates the intrinsic apoptosis pathway via cleavage and activation of Bid (Slee, et al., 1999; Smith, et al., 2003; Wehrli, et al., 2000). Our initial studies focused on caspase 8. Caspase 8 is the primary caspase recruited upon Fas or TRAIL-R1/-R2 death receptor activation by their respective ligands (Muzio et al, 1996; Srinivasula et al, 1996). 13-cis RA increased active caspase 8 (p18) by approximately 7- fold while a 2-fold increase was observed with 9-cis RA and ATRA treatment in SEB-1 sebocytes. Taken together, these findings suggest that the death receptor pathway is activated by 13-cis RA. Since TRAIL expression was significantly increased by 13-cis RA as determined 117 by gene expression analysis (4.2 fold over control), it seemed that TRAIL was a logical candidate as a mediator of activation of the extrinsic apoptotic pathway. We confirmed the increase in TRAIL expression and protein by QPCR and western blotting. TRAIL is a 281 amino acid, type II transmembrane protein expressed on the surface of cells. This tumor necrosis factor superfamily member also exists in a soluble form. Both the membrane form and soluble form have been shown to induce apoptosis in a variety of tumor cell lines. (Wiley, et al., 1995) TRAIL expression is detected in numerous normal tissues (spleen, prostate, ovary, lung, small intestine, colon, kidney and pancreas) including skin. (Stander and Schwarz, 2005; Wiley, et al., 1995) In normal human skin, TRAIL is expressed within the epidermis with stronger expression in the basal layers than superficial layers; infundibulum and outer root sheath of hair follicles; and in vitro studies have shown expression in keratinocytes and melanocytes (Stander and Schwarz, 2005). TRAIL-R1 and –R2 receptors and decoy TRAIL-R3 and –R4 are also expressed in the epidermis although expression is restricted to distinct layers: TRAIL-R1 (suprabasal layers), TRAIL-R2 (granular layer), and both TRAIL-R3 and TRAIL-R4 (basal layer) (Bachmann et al, 2001; Stander and Schwarz, 2005). Decoy receptors bind to TRAIL but lack or have truncated intracellular domains preventing transmission of the apoptotic signal (Marsters et al, 1997; Sheridan et al, 1997). TRAIL is unique in that it is supposedly non-reactive to normal human cells and specifically exerts its killing effect on tumor cells (Degli-Esposti et al, 1997). However, recent studies indicate that TRAIL induces apoptosis within 6-48 hours in primary human cells such as adult oligodendrocytes, prostate epithelial cells, primary hepatocytes and CD4+ T cells at concentrations of 10-100ng/mL (Armeanu et al, 2003; Herbeuval et al, 2005; Matysiak et al, 2002; Nesterov et al, 2002). To determine if TRAIL induces apoptosis in our SEB-1 sebocytes, we utilized recombinant human TRAIL protein, which maintains its bioactive function for a minimum of 24 hours in cell culture medium at 37ºC (personal communication, R&D Systems). The doses of TRAIL used in our studies (1-20ng) and the time frame of apoptosis induction noted in our cells (24-48 hours) are similar to other reports in the literature. The fact that TRAIL is capable of increasing cleaved caspase 3 protein expression within SEB-1 sebocytes suggests the presence of TRAIL-R1/-R2 receptors on the surface of SEB-1 sebocytes and that the TRAIL apoptosis pathway is intact within our cell system. RNA interference is a powerful tool for investigating protein function and target validation. Our siRNA knockdown experiments suggest that TRAIL may mediate 13-cis RA- induced apoptosis. In fact, our preliminary studies show that a 50% decrease in TRAIL protein 118 expression achieved with siRNA correlates with a 50% decrease in cleaved caspase 3 protein expression in the presence of 13-cis RA. Additional experiments are needed, however, to determine if TRAIL is the sole mediator of 13-cis RA- induced apoptosis in SEB-1 sebocytes. These studies do not directly address whether TRAIL induction is directly mediated by 13-cis RA or one of its isomers such as, 9-cis RA or ATRA, or a metabolite. Furthermore, additional studies are needed to assess whether retinoid receptor activation is required for this increase in TRAIL expression. The promoter region of the TRAIL gene does contain an RAR consensus sequence as identified by TESS computer software program. However, promoter mapping experiments conducted by Clarke et. al. demonstrated that there is no retinoic acid response element within 2kb of the transcription start site of TRAIL (Clarke et al, 2004). This is in agreement with our previous studies within SEB-1 sebocytes demonstrating that apoptosis was not blocked by a RAR pan-antagonist, suggesting an RAR independent mechanism to apoptosis induction. TRAIL up-regulation by 13-cis RA, therefore, most likely occurs by an indirect mechanism. Interferon regulatory factor 1 (IRF1) was identified as a critical factor in mediating TRAIL induction by retinoic acid in NB4 APL leukemia cells and SK-BR-3 breast cancer cells (Clarke, et al., 2004). Interestingly, 13-cis RA significantly up-regulates IRF1 gene expression (2.42 fold increase) in SEB-1 sebocytes (A.2). It may be possible that TRAIL up- regulation and the resulting apoptosis in SEB-1 sebocytes are due to IRF1. To definitively test this hypothesis, studies utilizing siRNA knockdown of IRF1 could be performed. To determine if 13-cis RA induced TRAIL mediated apoptosis is a relevant mechanism for 13-cis RA actions within the sebaceous gland, additional in vivo studies are needed. First, gene expression analysis of patient skin samples obtained at one week of isotretinoin (13-cis RA) therapy did not show up-regulation of TRAIL expression compared to baseline (Table 7); however, TRAIL expression is detected (“present”) within these patient biopsy samples. TRAIL up-regulation may be masked within the biopsies because of the extremely low levels of sebaceous gland mRNA relative to the total amount of mRNA from the skin biopsies. To validate TRAIL expression in vivo, immunohistochemistry for TRAIL on sections from baseline and one-week after isotretinoin can be performed. It would also be important to determine if TRAIL-R1 and –R2 receptors are expressed within the sebaceous gland. Our data (2.3.5) indicate that sebocytes, but not keratinocytes, undergo apoptosis in response to 13-cis RA. It would be interesting to determine if the cell-specific sensitivity to TRAIL accounts for the difference in the apoptotic response of sebocytes and keratinocytes to 13-cis RA. TRAIL-R1 and -R2 receptors are present within HaCaT keratinocytes and HaCaTs 119 undergo apoptosis in response to (500ng/mL) (Leverkus et al, 2003). Expression of TRAIL is “present” but is not significantly increased in HaCaT keratinocytes in response to 13-cis RA treatment according to our gene expression analysis. It would be important to confirm these findings via QPCR or western blotting. It is possible that 13-cis RA is not capable of inducing the level of TRAIL necessary for apoptosis induction in HaCaT keratinocytes. We can hypothesize that the cell-type specific sensitivity to TRAIL-induced apoptosis is one reason why 13-cis RA is incapable of inducing apoptosis within keratinocytes. Our studies are the first to demonstrate up-regulation of TRAIL by 13-cis RA. However, other retinoid compounds increase TRAIL expression. For example, ATRA and other RARα specific ligands up-regulate TRAIL expression, leading to apoptosis in acute promyelocytic leukemia (APL) cells (Altucci and Gronemeyer, 2002; Jimenez-Lara et al, 2004). In addition, retinoids including fenretinide, synthetic retinoid CD437, and ATRA, increase sensitivity to TRAIL-induced apoptosis by up-regulation of TRAIL-R1 and –R2 receptors in colon cancer, prostate cancer, and lung cancer cell lines (Kouhara et al, 2007; Sun et al, 2000; Sun et al, 2000). Recently, by increasing apoptosis, 13-cis RA alone or in combination with interferon, has been shown to be beneficial in the treatment of some forms of leukemia (Handa et al, 1997; Maeda et al, 1996); although the mechanism of apoptosis is unknown. Our data indicate that 13-cis RA up-regulates TRAIL in sebocytes and it may be possible that 13-cis RA up-regulates TRAIL expression in these cases, but additional experiments will be necessary to test this hypothesis. In summary, 13-cis RA-induced apoptosis is mediated, in part, by the up-regulation of TRAIL and activation of the extrinsic apoptosis pathway in SEB-1 sebocytes. How 13-cis RA up- regulates TRAIL expression in sebocytes and whether or not this drug alters TRAIL expression in vivo remains to be determined. Understanding the actions of 13-cis RA on a molecular level in terms of TRAIL regulation has implications in dermatology as well as in cancer therapeutics.

120 4.5 Materials and Methods

4.5.1 Reagents

N2 gas at -20ºC until use. Recombinant human TRAIL protein was purchased ready-to-use from R&D Systems (Minneapolis, MN) and stored at -20ºC. Stock solutions were diluted to desired concentrations in standard sebocyte culture medium containing: 5.5mM low glucose Dulbecco’s Modified Eagle Medium (DMEM) 3:1 Ham’s F12, 2.5% fetal bovine serum (FBS) , hydrocortisone (0.4µg/mL), adenine (1 X 10-8 M), insulin (10ng/mL), epidermal growth factor (3 ng/mL), cholera toxin (1.2 X 10-10 M) and antibiotics. TRAIL polyclonal rabbit antibody was purchased from Abcam, INC (Cambridge, UK). Cleaved caspase 3 (Asp175) rabbit monoclonal antibody and anti-rabbit HRP linked secondary antibody were purchased from Cell Signaling Technology (Beverly, MA).

4.5.2 Quantitative Polymerase Chain Reaction (QPCR)

Quantitative real-time PCR was performed to confirm the direction and magnitude of changes in the expression of select genes from the array data using Applied Biosystems’ Assays-on-Demand Taqman Universal PCR Master Mix and primer probe sets with ABI’s 7900HT Fast Real-Time PCR System with 384-well plate block module (Applied Biosystems, Foster City, CA). SEB-1 sebocytes (passage 22) were cultured in 60-mm plates until approximately 70% confluent. Sebocytes were treated with ethanol vehicle (0.01%) or 0.1 µM 13-cis RA in triplicate for 24, 48 and 72 hours. Total RNA was isolated using the RNeasy kit (Qiagen Inc., Valencia, CA) and its integrity was verified by agarose gel electrophoresis. cDNA was generated from 1 µg of total RNA, primed with oligo-dT, using the Superscript First-Strand Synthesis System for reverse transcription-PCR (Invitrogen, Carlsbad, CA). Diluted cDNA samples were run for the reference gene TATA binding protein (TBP) as well as TNFSF10 (TRAIL). Assay controls included samples omitting reverse transcriptase enzyme as well as 121 samples without cDNA. Data were analyzed using the REST-XL© program with efficiency corrected values and considered significant if p < 0.05.

4.5.3 siRNA knockdown of TRAIL by nucleofection

Nucleofection optimization and efficiency in SEB-1 sebocytes was determined by ‘Cell Line Optimization Nucleofector Kit’ according to manufacturer’s instructions (Amaxa Biosystems, INC. Gaithersburg, MD). Briefly, SEB-1 sebocytes in their logarithmic growth phase were nucleofected with 2 µg of pgmaxGFP DNA construct with each combination of nucleofector solution (V or L) and each nucleofector device program. (Amaxa Biosystems, INC., Gaithersburg, MD). The combination with the highest efficiency (GFP expression) and lowest mortality was chosen for future experiments: Nucleofector Kit T with program T-20. Efficiency of nucleofection was determined by GFP expression quantified with FACS. SEB-1 sebocytes (4 X 106 cells) were nucleofected with 2 µg of pgmaxGFP plating 1 X 106 cells per 100-mm culture plate. Control cells were “shocked” without GFP. Immediately after nucleofection, 500 µL RPMI 1640 medium was added to the cells and the cells were plated in normal sebocyte growth medium. Cells were trypsinized and collected 24, 48, 72 and 96 hours post-nucleofection and resuspended in phosphate buffered saline (PBS) for analysis by FACS. Mock-nucleofected SEB-1 sebocytes were used to determine negative, non-GFP containing cell populations and gates set accordingly. The percentages of cells expressing very high levels of GFP were 87%, 90%, 73%, and 57% at 24, 48, 72 and 96 hours post-nucleofection. ON-TARGET plus Human TNFSF10 (TRAIL) and siCONTROL siRNA duplex oligonucleotides were purchased from Dharmacon Research (Lafayette, CO). The transfection was performed as suggested by Dharmacon and Amaxa Biosystems with slight modifications. SEB-1 sebocytes, 2 X 106 cells per 100 µL reaction (solution + siRNA), were nucleofected with 54 pmols (0.7 µg) siCONTROL, 105 pmols (1.5 µg) siCONTROL, 54 pmols TRAIL and 105 pmols TRAIL to optimize siRNA amount needed for future experiments. 13-cis RA (0.1 µM) was added 24 hours post-nucleofection to induce TRAIL gene expression. Expression of TRAIL and extent of siRNA knockdown was verified by QPCR and western blotting for TRAIL at 48, 72 or 96 hours after nucleofection (24, 48 and 72 hour 13-cis RA treatment).

122 4.5.4 Western blot analysis for TRAIL and cleaved caspase 3

SEB-1 sebocytes (passage 22-26) were grown in 100-mm plates in standard culture medium until 50-75% confluent. Plates were rinsed with PBS and treated with 13-cis RA (1 µM); 9-cis RA (1 µM); ATRA (1 µM) or ethanol vehicle (0.01% or less) as a negative control. Cells were treated for 24, 48 or 72 hours. TRAIL and siCONTROL siRNA nucleofected cells (1 X106) were plated in 60-mm plates and treated with 0.1 µM 13-cis RA 24 hours post nucleofection. Total cell protein lysates from adherent and floating SEB-1 sebocytes were collected, flash frozen in liquid nitrogen and stored at -80ºC until needed. Protein concentration of each sample was determined by BCA Protein Assay (Pierce, Rockford, IL). Equal amounts of protein were run on NuPage 10% or 4-12% Bis-Tris Gels with MES Running Buffer (Invitrogen Life Technologies, Carlsbad, CA). Gels were transferred to polyvinylidene difluoride membrane, blocked for 1 hour at room temperature in 5% non-fat dry milk and incubated with 1:1,000 dilution of cleaved caspase 3 (Asp175) rabbit monoclonal antibody, or 1:500 dilution of TRAIL polyclonal rabbit antibody. Secondary anti-rabbit horseradish peroxidase linked antibodies were used to detect primary antibodies. SuperSignal West Pico Chemiluminescent Substrate (Pierce, Rockford, IL) was used for protein detection. Blots were stripped with 0.2M NaOH and reprobed with β-actin (1:4000). Films of blots were analyzed and quantified by densitometry with QuantityOne Software (Bio-Rad, Hercules, CA) after background subtraction. Data was analyzed with Student’s t-test and results were considered significant if p < 0.05.

Chapter 5

Development and Characterization of a Temperature Sensitive Sebocyte Cell Line (TSS-1)

5.1 Chapter Abstract

Current sebocyte model systems are able to mimic some important characteristics of sebaceous glands including androgen and retinoid sensitivity. The use of primary sebocytes for experiments is challenging due to their limited replication and incomplete terminal differentiation in culture. In order to circumvent this problem, investigators have developed Simian Virus 40 (SV40) immortalized sebocyte cell lines in their laboratories and currently, SZ95 and SEB-1, are the only available immortalized sebocyte cell lines. The major drawback to use of these particular cell models is the SV40 Large T antigen immortalization, which is necessary to achieve sufficient cell numbers, interferes with the normal differentiation program of the sebocytes. We have begun to develop a temperature sensitive SV40-immortalized sebocyte cell line in order to remove the influence of SV40 on our experiments. TSS-1 sebocytes are a suitable model system of human sebocytes because they 1) are able to be grown and maintained in culture, 2) express melanocortin 5 and androgen receptors (markers of differentiation), 3) demonstrate increased lipogenesis with androgen stimulation, 4) show inhibited lipogenesis with 13-cis RA treatment and 5) exhibit decreased cellular proliferation with 13-cis RA treatment, all characteristics of sebocytes in vivo. With additional characterization studies, we are hopeful that TSS-1 sebocytes will be proven to initiate a differentiation program that more closely resembles the in vivo program. 124 5.2 Introduction

Sebaceous gland research is limited by available model systems. Sebaceous glands are composed of sebocytes, which by their nature, undergo holocrine rupture upon maturing; thus, hindering the collection of sufficient cells to conduct experiments. Under normal conditions, sebocytes grow for a finite span of time; the life span of a sebocyte (in vivo) from cell division to holocrine secretion is approximately 21-25 days (Plewig and Christophers, 1974; Plewig, et al., 1971). In culture, isolated human sebaceous gland lobules give rise to primary sebocyte colonies that are a few centimeters in diameter and contain approximately 1 X 105 to 1 X 106 cells per colony. Cultured sebocytes accumulate lipids, which is evident by increases in Oil Red O staining and through immunohistochemical staining for sebocyte specific markers including EMA and OM-1 (Abdel-Naser, 2004; Zouboulis, et al., 1991). Experimentation with primary sebocytes in the laboratory, however, is not realistic due to the low numbers of attainable cells. In order to circumvent this problem, investigators have developed Simian Virus 40 (SV40) immortalized sebocyte cell lines in their laboratories (Thiboutot, et al., 2003; Zouboulis, et al., 1999). These immortalized cell lines allow for large-scale experiments and reproducibility of results. Currently, SZ95 and SEB-1 are the only available immortalized sebocyte cell lines. We have developed and characterized a temperature sensitive sebocyte cell line (TSS- 1). Using a temperature sensitive SV40 large T antigen DNA construct for immortalization, we are able to acquire the numbers of cells required for a given experiment and then “shut-down” the activity of the large T antigen by shifting cells to a restricted temperature, thus allowing the cells to ‘regain’ a more normal pattern of differentiation. It is our hope that this cell line will more accurately represent the sebocyte differentiation program that is noted in vivo. In the following chapter, we describe the development and current characterization of our newest immortalized sebocyte cell line.

5.3 Results

5.3.1 Temperature Sensitive Sebocytes (TSS) persist in culture

Primary sebocytes were stably transfected and transformed with DNA construct pRSV1609. The optimum temperature for this construct is 33ºC. Immunofluorescence staining 125 with an antibody to SV40 large T antigen indicates its presence in our new ‘TSS’ sebocytes and as a positive control, SEB-1 immortalized sebocytes. Large T antigen was absent from spontaneously immortalized HaCaT keratinocytes (negative control) (Figure 30).

The pRSV1609 construct encodes a Rous Sarcoma Virus (RSV) promoter to drive expression of a mutant SV40 large T antigen. While large T antigen is expressed at all temperatures, the protein is non-functional at the restrictive temperature. An amino acid change resulting from the mutation at base pair 1609 triggers protein instability at elevated temperatures leading to its unfolding and eventual degradation of the large T antigen protein. With this construct, large T antigen expression and functionality was optimal at 33ºC and hence, this was our permissive temperature. The stability and overall level of expression of this protein declined with increasing temperature (Figure 31). SEB-1 sebocytes, which contain wild-type large T antigen, were used for comparison.

126

Expression of SV40 large T antigen triggers uncontrolled cell growth. The most basic view of immortalization by large T antigen is that cell proliferation is induced by interference with Rb function and cell death is prohibited by blocking p53 normal functions, although other proteins and pathways may play some minor roles (for Review: (Ahuja et al, 2005)). As a check of the functionality of this protein, we performed manual cell counts every 3 days for 21 days at the permissive temperature and also at a range of elevated temperatures (37ºC, 39ºC and 41ºC) to determine which restrictive temperature was optimal for curbing cell growth but not triggering cell death. TSS sebocytes continually grew at 33ºC. The most rapid growth occured at 37ºC; however, this growth leveled off 9 days post plating. No increase in TSS sebocyte growth was noted under 39ºC conditions. With the 41ºC temperature, cell numbers did not increase (Figure 32). 127

33ºC 37ºC 700 39ºC 41ºC 600

500 4 400

300 Cells x 10 200

100

0 3 6 9 12151821 Days

After sufficient numbers of TSS sebocytes were obtained, we began generating clonal cell lines through serial dilution plating. Four distinct cell lines were produced: TSS-1 through TSS-4. During the process of clonal selection, some characterization studies were performed with “pooled” TSS sebocytes and then repeated where indicated with each of the clonal cell lines.

5.3.2 TSS sebocytes display a differentiated phenotype: Increased intracellular lipid and expression of the androgen receptor and the melanocortin 5 receptor.

Sebocyte differentiation is marked by an accumulation of intracellular lipids. The differentiation status of TSS sebocytes at 33ºC, 37ºC and 41ºC was monitored by Oil Red O 128 staining for intracellular neutral lipids. For comparison, Oil Red O staining was also done on SEB-1 sebocytes under the same conditions. The magnitude of difference in O Red O staining between SEB-1 and TSS sebocytes was not great. However, as the temperature increased, a corresponding increase in the amount of Oil Red O staining in both cell lines was noted with more staining in TSS-1 vs. SEB-1 at 41ºC (Figure 33).

In addition to lipid accumulation, differentiated sebocytes express androgen receptor and melanocortin 5 receptor (Miyake et al, 1994; Zhang et al, 2006). Both androgen receptor and melanocortin 5 receptor were expressed in the TSS-1, TSS-2, TSS-3 and TSS-4 sebocyte cell lines. Melanocortin 5 receptor expression was similar in all cells lines and was unchanged with respect to temperature. The level of androgen receptor expression varied considerably amongst the cell lines and was higher at the restrictive temperatures (Figure 34). Based on its higher level of expression of the androgen receptor, the TSS-1 cell line was chosen for all future studies. 129

Figure 34: Androgen receptor and melanocortin 5 receptor were expressed in differentiated TSS sebocyte cell lines. a) Total protein lysates from TSS-1, TSS-2, TSS-3 and TSS-4 cell lines cultured at 33ºC, 37ºC and 39ºC for three days were analyzed for expression of androgen receptor (110kD) via western blotting. Representative blots are shown. b) Relative levels of androgen receptor expression. Mean ± SE, n= 3. c) Total protein lysates from TSS-1, TSS-2, TSS-3 and TSS-4 cell lines at cultured 33ºC, 37ºC and 39ºC were analyzed for expression of the melanocortin 5 receptor (32kD) via western blotting. Controls included lysates from SEB-1, rat preputial cells (Rat P.C.) and human placenta (Plac.) Magic Markers XP (M.M.) were used as size indicators.

5.3.3 TSS sebocytes enter senescence following prolonged incubation at ‘restrictive’ temperatures.

By shifting the sebocytes to a higher temperature, the proliferative signal generated by expression of the large T antigen was “shut-down”. At this higher temperature (39ºC), it was also noted that the morphology of theTSS-1 sebocytes began to change. Our experiments depend on the cells’ ability to produce lipids and remain metabolically active, even if not in a proliferative state. To address the possibility that TSS-1 sebocytes enter a senescent state upon 130 shifting to the restrictive temperature, a standard β-galactosidase-associated senescence assay was performed (Dimri et al, 1995). With this assay, senescence is detected by an increase in b- galactosidase activity on provided substrate Xgal, translating to increased “blue” color within the cells. Shifting TSS-1 sebocytes from 33ºC to 39ºC resulted in sebocytes entering a senescent state after prolonged incubation at this higher temperature Figure 35. By 7 days, most TSS-1 sebocytes had entered senescence at 39ºC; some senescence was detected at 33ºC.

This assay was very important to the design of our future studies. Results from immunofluorescence studies and proliferation studies showed that sufficient SV40 large T antigen “shut-down” was achieved with 72-hour incubation at 39ºC. This assay illustrated that TSS-1 sebocytes will enter senescence after a prolonged period of time; therefore, limiting the amount of time a treatment can be applied. Based on the results from these three experiments, we designed our cell culture and treatment paradigm (Figure 36). 131

5.3.4 Total lipogenesis is increased at elevated temperatures in TSS sebocytes.

Sebocytes accumulate intracellular lipids upon differentiation. Some lipids produced are sebocyte-specific including wax esters and squalene. As a quantitative measure of total neutral lipid production, standard lipogenesis assays were performed with confluent TSS sebocytes maintained at 33ºC, 37ºC, and 41ºC for 3 days. Extremely low levels of lipids were detected at 41ºC (data not shown) therefore, the assay was repeated with temperatures of 33ºC, 37ºC and 39ºC. Unexpectedly, total lipogenesis was highest with a 37ºC incubation temperature, with little to no difference between 33ºC and 39ºC (Figure 37). The experiment was repeated using an incubation time of one week rather than 3 days. Again, at 37ºC, TSS sebocytes showed the greatest amount of lipogenesis, with little to no difference between 33ºC and 39ºC (data not shown). 132

80 *

cells/hour 70 6 60

40 Mean ± SEM 30

20 C-acetate incorporated/10 14 10

cpm cpm 0 37ºC 33ºC 39ºC

Incorporation of 14C-acetate into sebaceous lipids was assayed in the TSS-1 cell line after 3 days and at one week incubations at 33ºC, 37ºC and 39ºC. As expected from the previous studies in TSS sebocytes, most lipid classes were highest at 37ºC followed by 33ºC and 39ºC (Figure 38). Of note, is the lack of difference between the levels of acetate incorporation into sebaceous lipids at 33º and 39ºC at 3 days, possibly related to temperature requirements of the lipogenic enzymes. At the one-week time point, the decrease in lipogenesis at 39ºC is most likely due to cells entering senescence.

133

Although lipogenesis was maximal at 37ºC, subsequent lipogenesis assays were conducted at 39ºC; thus minimizing the effect of the SV40 large T antigen. 134 5.3.5 Synthetic androgen R1881 increases and 13-cis RA decreases lipids in TSS-1 sebocytes

Clinical observation and experimental evidence illustrate that androgens stimulate sebum production in vivo. To demonstrate that our cell culture model suitably mimics the in vivo situation, we treated TSS-1 (39ºC) with synthetic androgen R1881 (methyltrienolone) for 24 hours and measured total lipogenesis. TSS-1 sebocytes were androgen responsive. Androgen treatment increased total lipogenesis by 31% in TSS-1 sebocytes after 24 hours when compared to control (p = 0.003) (Figure 39).

13-cis RA is the most potent inhibitor of sebum production in vivo. To examine whether 13-cis RA exhibits a similar effect in our cell culture model, In experiments separate from the synthetic androgen treatment, TSS-1 (39ºC) sebocytes were treated with 0.1 µM 13-cis RA for 24 hours prior to total lipogenesis assay. 13-cis RA significantly decreased total lipogenesis by 22% as compared to vehicle control at 24 hours (p < 0.01) (Figure 40). 135

5.3.6 13-cis RA decreases TSS-1 proliferation

Growth inhibition by retinoids, including 13-cis RA, has been reported in numerous cell types; including the SEB-1 sebocyte model (Nelson, et al., 2006). To further validate the temperature sensitive sebocyte cell line, TSS-1 sebocytes were treated with 13-cis RA (0.1, 1 and 10 µM) for 24, 48 or 72 hours. All concentrations of 13-cis RA significantly decreased cell numbers at 48 hours (8-13% decrease from control) and 72 hours (7-31% decrease from control) in TSS-1 (33ºC), which is similar to the pattern observed in SEB-1 sebocytes (2.3.1). At 39ºC, cell numbers decreased at both the 48 and 72 hour time points with 13-cis RA concentrations (0.1, 1 or 10 µM) with a maximum decrease of 22% with 10 µM 13-cis RA at 72 hours. In general, a trend in decreased cell numbers is evident; although, not all decreases were significant (Figure 41). 136

Figure 41: 13-cis RA causes growth inhibition in TSS-1 sebocytes. (a) TSS-1 (33ºC) (b) TSS-1 (39ºC). Time-dependent inhibition of TSS-1 sebocyte proliferation. TSS-1 cells were cultured in the presence of ethanol vehicle alone (0.01% or less; control), 0.1 µM, 1 µM or 10 µM concentrations of 13-cis RA 24, 48 or 72 hours. Attached cells were collected, stained with trypan blue, and counted manually. Data represent mean ± SEM, n = 9. Statistical analysis was performed by ANOVA Two-Factor with Replication. * p < 0.05, **p < 0.01.

5.3.7 13-cis RA induces apoptosis in TSS-1 sebocytes at the restrictive temperature.

Sebaceous glands are retinoid responsive. Histologically, sebaceous glands from patients treated with 13-cis RA are markedly reduced in size and individual sebocytes appear undifferentiated, lacking the characteristic cytoplasmic accumulation of sebaceous lipids (Goldstein, et al., 1982; Landthaler, et al., 1980; Strauss, et al., 1980). Apoptosis may explain 137 the reduced size of the sebaceous gland and previous studies demonstrated that 13-cis RA induces apoptosis in SEB-1 sebocytes. Therefore, the TUNEL assay was performed in TSS-1 sebocytes (39ºC) treated with 13-cis RA (0.1, 1 and 10 µM) for 48 and 72 hours. At 48 hours, the percentage of TUNEL-positive cells significantly increased with 1 µM and 10 µM concentrations in a dose dependent manner (Figure 42). All concentrations of 13-cis RA at 72 hours increased the percentage of TUNEL-positive cells although much less than 48 hours. Most cells at the 72 hour time point were “floating” in medium and therefore clearly had undergone apoptosis but were not able to be assayed.

138 5.4 Discussion

Immortalization of primary cells by the large T antigen of DNA tumor virus, SV40, artificially disrupts the normal cell cycle and differentiation patterns. Large T antigen affects two major players in cell cycle control: Retinoblastoma (Rb) tumor suppressor and p53. Large T antigen binds and inactivates Rb, which acts as a negative regulator of normal cell proliferation. Under normal growth conditions, the Rb tumor suppressor and its subsequent phosphorylation, controls the transition from the G1 phase to the S phase of the cell cycle. Hypophosphorylated Rb binds E2F transcription factor members inhibiting the transcription of E2F regulated genes including cyclin family members and genes involved in DNA replication and repair (Ahuja, et al., 2005; Berthet et al, 2006; Ohtani et al, 1995). The association of Rb with E2F prevents progression through the cell cycle (Zhang et al, 1999). Phosphorylation of Rb by cyclin dependent kinases results in dissociation from E2F, allowing the cell to proceed into S phase (Berthet, et al., 2006; Nevins, 2001). DNA tumor viruses, including SV40 large T antigen, specifically recognize the hypophosphorylated forms of Rb thereby removing the ability of Rb to regulate E2F activities and allowing uncontrolled progression into S phase (Cooper, 1995). In addition, T antigen inhibits the function of p53 by blocking its DNA binding surface, thus inhibiting its ability to control gene expression (Bargonetti et al, 1992; Lilyestrom et al, 2006). The most basic view of immortalization by large T antigen is that cell proliferation is induced by interference with Rb function and cell death is prohibited by blocking p53 normal functions, although other proteins and pathways may play some minor roles (for Review: (Ahuja, et al., 2005)). Even though SV40 large T antigen immortalization of sebocytes allows for large numbers of cells to be acquired, it hinders the native sebocyte differentiation program. Differentiation of primary sebocytes involves slowing of cellular growth and accumulation of lipids within the cell (Rosenfield, 1989). This can be artificially induced in cell culture models with SV40 immortalization by allowing culture plates to reach a confluent state and the addition of adipogenic hormones including methylisobutylxanthine, dexamethasone and insulin (MDI) to cause accumulation of lipids. These compounds are known to trigger differentiation of 3T3-L1 pre-adipocytes into adipocytes (Student et al, 1980), a process that is analogous to the sebocyte differentiation program. Addition of these compounds adds to the complexity of understanding the sebocyte’s native differentiation program as the compounds themselves may affect other aspects of cellular development. 139 By using a temperature sensitive large T antigen construct, expression of the large T antigen can be controlled, allowing the sebocytes to “regain” a differentiation program more similar to sebocytes in vivo than other SV40 immortalized cell lines. The specific temperature sensitive construct (pRSV1609) obtained from Dr. Judy Tevethia (Penn State University College of Medicine, Hershey, PA) has an arginine to lysine amino acid change at position 357 and has been renamed ts357R-K (Loeber et al, 1989). In normal African green monkey kidney fibroblast CV-1 cells, the ts357R-K mutant was able to replicate at the permissive temperature (32ºC) and when shifted to 39ºC, replication was significantly reduced and completely blocked at 41ºC (Loeber, et al., 1989). These results precisely mirror the cell proliferation studies we performed with our TSS sebocyte cell line. Furthermore, characterization of the structural regions of large T antigen protein indicates that this amino acid substitution is in the middle of the p53 binding domain (Loeber, et al., 1989; Schmieg and Simmons, 1988). Previous studies have demonstrated that ts357R-K at the permissive temperature is able to bind to p53 preventing its function; however, at higher temperatures, p53 is not bound by ts357R-K and instead is able to regain its DNA binding abilities (Ray et al, 1996). A shift to the higher temperature, therefore, blocks SV40 inhibition of p53 function and allows the cell to regain control of p53 regulated gene expression. Inactivation of pRb alone is not sufficient for full transformation and cells will lose some features associated with transformation. Studies to verify that SV40 large T antigen does not bind p53 in our TSS-1 sebocyte cell line at 39ºC are currently in progress. In general, sebocytes in vitro undergo an incomplete terminal differentiation when compared to in vivo. Therefore, it is necessary to characterize the degree of differentiation within our new cell line. In rodents, melanocortins increase sebum production. Transgenic mice deficient in the melanocortin-5 receptor have hypoplastic sebaceous glands and reduced sebum production (Chen, et al., 1997). Melanocortin 5 receptor is expressed at the onset of differentiation and in lipid-containing, completely differentiated sebocytes and, as such, is a useful marker for accessing the differentiation status of our TSS sebocytes (Zhang, et al., 2006). Melanocortin 5 receptor is detected in each of our clonal cell lines at all temperatures, thus indicating the cell lines are capable of differentiation. The melanocortin-5 receptor may play a role in the modulation of sebum production. Further experimentation is required to test this hypothesis. Androgens regulate sebaceous gland development and differentiation in vivo. Individuals with a genetic deficiency of androgen receptors (complete androgen insensitivity) have no detectable sebum secretion and do not develop acne (Imperato-McGinley, et al., 1993). In 140 addition, testosterone or dihydroepiandrosterone (an adrenal precursor hormone) can increase the size and secretion of sebaceous glands (Pochi and Strauss, 1969). Immunohistochemistry studies have demonstrated the presence of androgen receptors within sebocytes (Blauer et al, 1991; Pelletier and Ren, 2004). Furthermore, using rat preputial cells, androgen receptor expression has been found to increase with sebocyte differentiation (Miyake, et al., 1994). Therefore, we examined androgen receptor expression and androgen-responsiveness within our newly developed cell line. Our studies indicate that TSS-1 sebocytes do not express androgen receptor at 33ºC. Androgen receptor expression does, however, increase with increasing temperature, suggesting that at 39ºC, TSS-1 sebocytes have undergone some degree of differentiation. In contrast, SEB-1 sebocytes do not express androgen receptor. Based on the high level of androgen receptor expression when compared to the other TSS cell lines, the TSS-1 cell line was chosen for additional characterization studies. As an assessment of androgen-responsiveness, we used synthetic androgen, R1881 (methyltrienolone), and measured total lipogenesis, which is analogous to sebum production in vivo. Synthetic androgen R1881 increased total lipogenesis within our TSS-1 sebocytes. Retinoids also regulate sebaceous glands by inhibiting differentiation in vivo. Histology revealed a marked decrease in the size and secretion of the sebaceous glands after 16 weeks isotretinoin treatment (Goldstein, et al., 1982). Retinoids (13-cis RA, ATRA and 9-cis RA) have been shown to inhibit proliferation on cultured human sebocytes, SZ95 and SEB-1 SV40- immortalized sebocyte cell lines, and rat preputial cells, although the magnitude of inhibition was retinoid- and length of treatment-dependent (Nelson, et al., 2006; Tsukada, et al., 2000; Wrobel, et al., 2003; Zouboulis, et al., 1991; Zouboulis, et al., 1993). TSS-1 sebocytes respond similarly to 13-cis RA treatment with the same characteristic decrease in cell proliferation that other human sebocyte models exhibit. Results of the proliferation studies in TSS-1 (33ºC) sebocytes parallel the results of previous studies in SEB-1 sebocytes; most likely due to the constitutive expression of SV40 large T antigen protein in both cell lines. Interestingly, the decrease in proliferation of the TSS-1 sebocytes at 39ºC is not as pronounced as at 33ºC. The percentage of TSS-1 (39ºC) sebocytes undergoing 13-cis RA-induced apoptosis is higher than the percentage seen in SEB-1 sebocytes in previous studies. 13-cis RA, the most potent inhibitor of sebum production in vivo, significantly decreased total lipogenesis in TSS-1 sebocytes. In summary, TSS-1 sebocytes are a suitable model system of human sebocytes because they 1) are able to be grown and maintained in culture, 2) express melanocortin 5 and androgen receptors (markers of differentiation), 3) demonstrate increased lipogenesis with 141 androgen stimulation, 4) show inhibited lipogenesis with 13-cis RA treatment and 5) exhibit decreased cellular proliferation with 13-cis RA treatment, all characteristics of sebocytes in vivo. Although this cell line has numerous positive features which are listed above, there are drawbacks to use of the model system. First, we are limited in the amount of time a treatment can be applied to the cells. Under our experiment paradigm, cells must be at 39ºC for 72 hours in order to “deactivate” the uncontrolled growth signal generated by functional large T antigen. Our initial characterization studies indicate that the TSS-1 sebocytes will enter a senescent state after a period of 7 days at the restricted temperature. Combined, this only allows for a 3-4 day treatment period. At the present time, this time period is sufficient for all our currently planned experiments; however, future experiments will have to be designed accordingly with this limitation in mind. A second potential drawback to our TSS-1 model system may be the induction of the heat shock response. By shifting TSS-1 sebocytes to 39ºC to “shut off” SV40 function, the sustained incubation at the higher temperature may induce the heat shock response which can interfere with the normal differentiation program. ‘Heat shock’ can rapidly induce transcription of heat shock proteins whose primary function is to resist thermal stress. In addition, heat shock can up-regulate signaling pathways directly involved in cell survival and cell death and, depending on the duration and intensity of the stress, may counteract or up-regulate the apoptotic response (Anckar and Sistonen, 2007; Nadeau and Landry, 2007; Voellmy and Boellmann, 2007). TSS-1 sebocytes were subjected to gene expression analysis at both temperatures (33 and 39; data not shown). Preliminary inspection of the results indicates that incubation at 39ºC induces expression of genes involved in the heat shock response, with approximately 10 heat shock proteins (HSP) and heat shock factor 2 binding protein (HSF2BP) significantly increased when compared to 33ºC control gene arrays. Interestingly, one of the most highly up-regulated genes, thioredoxin interacting protein (TXNIP, Vitamin D3 up-regulated protein 1, (VDUP1)) (up-regulated ~10-fold), has been shown to be induced under conditions of oxidative or heat shock stress. Overexpression of this protein triggered decreased proliferation and induced apoptosis within mouse fibroblasts (Junn et al, 2000). This may partially explain the observed increase in apoptosis sensitivity to 13-cis RA because up-regulation of this gene by increased temperature may ‘prime’ apoptotic pathways. Current sebocyte model systems are able to mimic some important characteristics of sebaceous glands, but sebocytes generally undergo incomplete terminal differentiation in culture. Using a temperature sensitive SV40 construct, we were able to develop the TSS-1 cell 142 line that exhibits the same important characteristics of previous models but is capable of more complete differentiation in vitro than previous models. With additional characterization studies, we are hopeful that TSS-1 sebocytes will be proven to initiate a differentiation program that more closely resembles the in vivo program.

5.5 Materials and Methods

5.5.1 Sebocyte Culture

Human skin was obtained from facial surgeries under a protocol approved by the Institutional Review Board of the Pennsylvania State University College of Medicine. Human sebaceous glands were dissected from facial skin and sebocyte cultures were established. Cells were co-cultured with mitomycin-C-inactivated 3T3 fibroblasts in medium containing: 5.5mM Low Glucose DulBecco’s Modified Eagle Medium (DMEM) 3:1 Ham’s F12, 2.5% fetal bovine serum (FBS), hydrocortisone 0.4 µg/mL, adenine 1.8 X 10-4 M, insulin 10 ng/mL, epidermal growth factor (EGF) 3 ng/mL, cholera toxin 1.2 X 10-10 M, and antibiotics. Primary sebocytes were detected as “outgrowths” from the sebaceous glands surrounding by the clearly distinct 3T3 fibroblasts. EDTA (0.02%) removes 3T3 fibroblasts from culture dishes and sebocytes were collected through trypsin digestion. Sebocytes were stored in freezing medium containing low glucose DMEM, 10% FBS, and 10%DMSO in liquid nitrogen cryo-storage until needed.

5.5.2 Establishment of TSS Sebocytes and Individual Clonal Cell Lines

Secondary sebocyte cultures were established as above from sebaceous glands dissected from the normal ear of an 80-y-old male and stored in liquid nitrogen. Primary sebocytes (p3) were recovered and plated on four 35-mm tissue culture plates with Bajor’s sebocyte medium containing: 10% FBS, high glucose, insulin and epidermal growth factor. Cells were transfected with DNA construct, pRSV1609, containing the sequence encoding the transforming SV40 large T protein with a RSV promoter (gift, Dr. Judy Tevethia) using Effectene Transfection Reagent kit (Qiagen Sciences, Maryland) with 0.5 µg DNA per cell culture plate 143 according to manufacturer instructions. The mutation at 1609 renders the protein non-functional at non-permissive temperatures due to its “unfolding”; thus targeting it for degradation. Successful transformation of sebocytes was confirmed in 5th passage sebocytes by localization of SV40 large T antigen in the nucleus of transformed cells using immunohistochemistry and continued passage in culture. We derived four cell lines (TSS-1 through TSS-4) using standard clonal dilution techniques on passage 15 sebocytes. Permissive temperature of 33ºC and the non-permissive temperature of 39ºC were empirically determined. All experiments conducted with TSS sebocytes or individual clonal lines were done in our standard sebocyte culture medium consisting of 5.5mM Low Glucose DulBecco’s Modified Eagle Medium (DMEM) 3:1 Ham’s F12, 2.5% fetal bovine serum (FBS), hydrocortisone 0.4 µg/mL, adenine 1.8 X 10-4 M, insulin 10 ng/mL, epidermal growth factor (EGF) 3 ng/mL, cholera toxin 1.2 X 10-10 M, and antibiotics.

5.5.3 Cell Growth and Viability

Cell growth and viability of TSS were calculated by plating equal numbers of cells per 35 mm plate followed by manual cell counts with a hemacytometer at days 3, 6, 9, 12, 15, 18 and 21. Cells were placed at 33ºC (permissive temperature) as well as 37ºC, 39ºC, and 41ºC to determine the optimal non-permissive temperature. Viability was assessed using Trypan Blue dye exclusion. Results are the average of two independent samples at each time point and temperature. To determine if 13-cis RA had any effect on TSS-1 sebocyte viability, cells were cultured at 33ºC and shifted to 39ºC for 72 hours prior to treatment with vehicle control, 0.1 µM, 1 µM or 10 µM for 24, 48 and 72 hours in triplicate. Parallel studies were conducted on TSS-1 sebocytes maintained at 33ºC. Cells were detached using trypsin (0.05%), collected, and diluted in standard cell culture medium for manual cell counts using a hemacytometer. Cell viability was assessed using Trypan Blue dye exclusion. Each assay was repeated three independent times. Statistical significance was determined by ANOVA Two Factor with Replication and considered significant if p < 0.05.

144 5.5.4 Immunohistochemistry and Oil Red O Staining

Transfection of sebocytes was confirmed by fluorescent immunohistochemistry with SV40 antibody. Polyclonal mouse SV40 large T antigen antibody (901/902) was obtained from Dr. Judy Tevethia. SEB-1 sebocytes were utilized as the positive control. Negative controls included HaCaT keratinocytes and TSS incubated without primary antibody. Cells were cultured and maintained under normal growing conditions. Antibody was used at 1:100 dilution with overnight incubation followed by 1:500 dilution of secondary anti-mouse, fluorescein antibody. Cells were examined under fluorescent microscopy and representative images are shown. Sebaceous phenotype of TSS cells was verified by confirmation of lipid droplets with Oil Red O staining. TSS sebocytes were cultured on slides placed at 33ºC. Six days later, cells were maintained at 33ºC or shifted to 37ºC, 39ºC or 41ºC for 72 hours prior to staining. Cells were fixed at room temperature with 10% formalin for 30 minutes followed by washings with phosphate buffered saline (PBS). Oil Red O staining solution consisting of Oil Red Oil (Sigma, St Louis, MO) and 99% isopropanol was added to slides for 15 minutes. Slides were rinsed with 50% isopropanol, then water and counterstained with hematoxylin. Slides were examined under light microscopy (400X total magnification) for presence of Oil Red O staining. Representative images were captured using a Spot digital camera (Diagnostic Instruments, Inc.).

5.5.5 Western Analysis

Total protein lysates were collected from TSS-1, TSS-2, TSS-3, TSS-4 and SEB-1 sebocytes. Each TSS-1 cell line was grown at 33ºC until approximately 70% confluent and then maintained at 33ºC or shifted to 37ºC or 39ºC for 3 days. Total protein lysates were collected, quantified and separated by electrophoresis as previously described in 2.5.4. Blots were incubated with 1:1000 dilution of polyclonal rabbit androgen receptor antibody (Cell Signaling Technology, Beverly, MA) or 1:500 dilution of polyclonal chicken melanocortin receptor 5 antibody (Cocalico Biologicals Inc, Reamstown, PA). Secondary anti-rabbit and –chicken horseradish peroxidase linked antibodies were used to detect primary antibodies. SuperSignal West Pico Chemiluminescent Substrate (Pierce, Rockford, IL) was used for protein detection. Blots were stripped with Restore Western Blot Stripping Buffer (Piece, Rockford, IL) and reprobed with β-actin (Cell Signaling Technology, Beverly, MA). Films of blots were analyzed and 145 quantified by densitometry with QuantityOne Software (Bio-Rad, Hercules, CA) after background subtraction. Androgen receptor western blots were repeated three independent times. Data was analyzed with Student’s t-test and results were considered significant if p < 0.05.

5.5.6 β-galactosidase Senescence Assay

Individual clonal cell lines TSS-1, TSS-2, TSS-3 and TSS-4 were cultured and placed at 33ºC, 37ºC, or 39ºC for 3, 5, or 7 days. After the appropriate time period, cells were washed with PBS and fixed with a PBS solution containing 2% formaldehyde, 0.2% glutaraldehyde.

Fixed cells were incubated overnight in a 37ºC “no-CO2” incubator with staining solution containing: 1mg/mL X-gal in dimethylformamide, 40mM 0.2M citric acid/sodium phosphate buffer, 5 mM ferrocyanide, 5 mM ferricyanide, 5M sodium chloride and 1M magnesium chloride. After incubation, plates were washed twice with PBS. Plates were examined under light microscopy for “degree of blue color” and representative images were captured at 400X total magnification.

5.5.7 Lipogenesis Assay: 14C-actetate incorporation into neutral lipids

TSS (passages 17-20) and individual TSS cell lines, TSS-1, TSS-2, TSS-3 and TSS-4 were subjected to our standard lipogenesis assay to measure the amount of neutral lipids within the cells at the different incubator temperatures (33ºC, 37ºC, 39ºC, 41ºC). SEB-1 sebocytes (constitutive SV40 expression) grown under the same conditions were used for comparisons. TSS and each TSS individual cell line (2 X 105 cells) were cultured in 35-mm plates and incubated at 33ºC until confluent. Once confluent, some plates remained at 33ºC while other plates were shifted to 37ºC, 39ºC or 41ºC for another 3 days prior to the assay to allow for SV40 large T antigen “shut-down”. Cells were fed every other day with standard sebocyte medium. Cells were collected by trypsinization and manually counted. Cells were resuspended in DMEM containing 1 µCi of 14C-acetate and incubated for 2 hours at 37ºC with shaking. Total lipids were extracted with ethyl ether and counted by liquid scintillation counting. Manual cell counts were performed to normalize data to cell number. Acetate incorporation into lipids was expressed as 146 “cpm 14C-acetate incorporated/106 cells/hour.” Each experiment was repeated 2-3 independent times. Additionally, TSS-1 sebocytes (passage 23) were subjected to a “neutral lipid pattern analysis”. Assay was performed as above, except after lipid extraction, lipids were separated by thin layer chromatography and individual lipids were counted by liquid scintillation counting according to published methods (Smith et al, 2006). Androgens are known to stimulate sebum production. To examine the androgen responsiveness of TSS-1, cells were treated with synthetic androgen R1881 (methyltrienolone; Perkin Elmer, Wellesley, MA) for 24 hours prior to total lipogenesis assay. In parallel experiments, the effect of 0.1 µM 13-cis RA (Sigma, St Louis, MO) was examined. TSS-1 cells were cultured in 35-mm culture plates and maintained at 33ºC until 80-90% confluent, feeding as necessary. To allow for large T antigen shut-down, all plates were shifted to 39ºC for 72 hours. TSS-1 sebocytes were treated in triplicate with vehicle controls, 1 X 10-8 M R1881, or 0.1 µM 13-cis RA for 24 hours. Total lipogenesis was performed as above and experiment was repeated 3 independent times; n = 9 samples. Statistical significance was determined with ANOVA Two Factor with Replication and considered significant if p < 0.05.

5.5.8 TdT-Mediated dUTP Nick End Labeling (TUNEL) Staining

TSS-1 sebocytes were cultured in 12-well plates and shifted to 39ºC for 72 hours prior to treatment. 13-cis RA (10 mM stock in ethanol) was diluted in standard culture medium and applied for 48 and 72 hours. Culture plates were approximately 60% confluent at time of treatment. Each well was considered one sample. Samples were prepared by manufacturer’s instructions for In Situ Cell Death Detection Assay (Roche Applied Science, Indianapolis, IN) Additional assay controls included DNase I-treated positive and negative controls, with negative controls receiving labeling solution only. Results were quantified by counting positively stained cells in 3 representative fields per well for each of the treatments carried out in triplicate. Each assay was performed two independent times; n = 6 samples. Data analysis was performed using ANOVA Two Factor with Replication and considered significant if p < 0.05.

Chapter 6

Discussion and Future Directions

6.1 Introduction

The initial observations of the effectiveness of isotretinoin in the treatment of acne came as an unexpected side effect of its use in the treatment of ichthyosis, a hereditary disorder of keratinization (Peck, 1979). Isotretinoin has been on the market since 1982 and even though it has been prescribed for 25 years, extensive studies into its molecular mechanism of action in human skin and sebaceous glands have not been done. The studies undertaken in this thesis are the beginning to understanding the effects of 13-cis RA on the sebaceous gland and its mechanism of action in sebum suppression.

6.2 Rationale, hypothesis, and results of this work

Isotretinoin is the most potent agent for sebosuppression, and yet, how this sebosuppressive action occurs is unknown. It is well established that isotretinoin drastically reduces the size and lipid secretion of sebaceous glands (Goldstein, et al., 1982). We hypothesized that isotretinoin decreases the size of the sebaceous gland by activation of cell cycle arrest and/or apoptosis pathways and that sebum suppression is most likely an indirect result of the reduced size of the sebaceous gland. Our hypothesis was based on the histology of the sebaceous gland after isotretinoin treatment.

In this thesis, we have shown that:

148 1) 13-cis RA, unlike 9-cis RA or ATRA, induces cell cycle arrest and apoptosis in SEB-1 sebocytes. Its ability to induce apoptosis is not inhibited in the presence of functional RAR pan antagonist AGN 193109, suggesting a non-RAR mediated mechanism.

2) Gene expression analysis of patient biopsies one-week into isotretinoin therapy provided insight into the initial changes induced by this drug, including NGAL up- regulation. Gene expression analysis on SEB-1 sebocytes and HaCaT keratinocytes provided clues to possible cell specific actions of 13-cis RA.

3) NGAL localization within the sebaceous gland and its ability to induce apoptosis in SEB-1 sebocytes suggest that NGAL may, in part, mediate the apoptotic action of 13-cis RA on sebaceous glands. We identified the possible human cell surface receptor for NGAL and cell-specific expression of its isoforms correlates with apoptosis sensitivity.

4) 13-cis RA, unlike 9-cis RA and ATRA, activates the extrinsic apoptosis pathway in SEB-1 sebocytes through up-regulation of TRAIL.

The results of this body of work have allowed us to generate a model by which we believe 13- cis RA is mediating apoptosis in sebocytes (Figure 43).

6.3 Explanation of model

Like other lipophilic retinoids, 13-cis RA can readily transverse the plasma membrane. 13-cis RA isomerizes to 9-cis RA and ATRA, which are known to activate their respective retinoid receptors and affect gene expression. 13-cis RA up-regulates, although probably indirectly through its isomers, metabolites or an as yet identified nuclear receptor, the expression of genes involved in cell-cycle arrest and apoptosis. Up-regulation of TRAIL leads to activation of the extrinsic apoptosis pathway and up-regulation of NGAL also leads to apoptosis, but the pathway (extrinsic or intrinsic) is not yet known.

149

Figure 43: Model of 13-cis RA induces apoptosis and cell cycle arrest in SEB-1 sebocytes and human sebaceous glands. 150 6.4 Future Directions of this project

There are still many unanswered questions. In the following section, I will discuss some of these important questions and how insights from our work can lead to advances in the field.

6.4.1 Why is 13-cis RA superior to 9-cis RA or ATRA in the treatment of acne?

13-cis RA is superior to either 9-cis RA or ATRA for sebosuppression and 13-cis RA has little to no ability to bind to cellular retinol-binding proteins or RAR and RXR receptors which are readily activated by both 9-cis RA and ATRA (Allenby, et al., 1993; Fogh K., et al., 1993; Geiger, et al., 1996; Hommel, et al., 1996; Levin, et al., 1992; Ott, et al., 1996). In terms of apoptosis, our work confirms that the actions of 13-cis RA are distinct from 9-cis RA or ATRA, with 13-cis RA activating the extrinsic apoptosis pathway more than either 9-cis RA or ATRA. Furthermore the apoptotic effect is not mediated by RAR activation, which suggests that 13-cis RA does not function as a pro-drug for ATRA, but instead is capable of influencing cellular processes on its own. Other investigators have proposed that the superior efficacy of 13-cis RA in acne treatment may be related to its: 1) unique pharmacokinetic properties within the cell, 2) alteration of cellular signaling pathways by covalent or non-covalent protein interactions, 3) ability to in-directly affect enzyme activity, 4) generation of alternative metabolites (other than 9- cis RA and ATRA) that have transcriptional activity or 5) binding to an as yet unidentified receptor (Baron, et al., 2005; Holmes et al, 2003; Hoyos, et al., 2000; Imam, et al., 2001; Karlsson, et al., 2003; Pettersson, et al., 2004; Zorn and Sauro, 1995).. Understanding the pharmacokinetic properties of 13-cis RA within our SEB-1 sebocytes is an important first step. In our SEB-1 sebocyte model, HPLC analysis showed that 13-cis RA concentrations were maximal within SEB-1 sebocytes twelve hours after treatment and concentrations of 9-cis RA and ATRA begin to increase after 24 hours. It is possible that the reason 13-cis RA is superior to 9-cis RA or ATRA is that it is more favorably absorbed by sebocytes or is more stable within sebocytes for a longer period of time. For example, ATRA may be rapidly hydroxylated and inactivated by 4-hydroxylases or it may bind to PKC and phosphorylate AKT initiating a prosurvival response as occurs in other cell systems (Bastien et al, 2006; Marikar et al, 1998; Ochoa et al, 2003). Initial HPLC experiments examining the metabolism of 13-cis RA, 9-cis RA and ATRA in sebocytes are necessary to differentiate between these possibilities. In addition, treatment with 13-cis RA may be more effective as an 151 acne therapy due to a combination of 13-cis RA, 9-cis RA and ATRA being present within the sebocytes simultaneously and acting synergistically to induce apoptosis. Our HPLC studies indicate that all three retinoids are present in measurable concentrations during the time period of apoptosis induction. It may be possible that all three retinoids synergistically contribute to the apoptotic effect and each may potentially work through a different apoptosis mechanism. To begin to address this possibility, combination retinoid treatments and HPLC analysis, along with our standard apoptosis assays would provide initial insight into this hypothesis. Finally, it may be possible that currently unidentified metabolites or isomerization products of 13-cis RA mediate its apoptotic actions in SEB-1 cells. Our HPLC studies did show two unknown “peaks” that could represent metabolites of 13-cis RA; additional characterization studies are needed to identify these unknown compounds. Beyond pharmacokinetic studies, understanding which cellular pathways are affected by 13-cis RA is crucial to understanding its mechanism of action. Experiments indicated that 13-cis RA induced apoptosis while ATRA or 9-cis RA did not; thus suggesting that apoptosis may be responsible for 13-cis RA’s effectiveness in the treatment of acne. Our gene array data indicated that 13-cis RA induced gene expression of mediators of both the intrinsic and extrinsic pathways of apoptosis. Synthetic retinoid MX3350-1 activates both the intrinsic and extrinsic apoptotic pathways independent of retinoid receptors (Chun, et al., 2005). It is possible that 13- cis RA, like MX3350-1, activates both the intrinsic and extrinsic pathways of apoptosis in sebocytes. In SEB-1 sebocytes, apoptosis occurs after 48 and 72 hours of 13-cis RA treatment and gene expression analysis reveals 13-cis RA up-regulates genes involved in apoptosis including potent apoptosis inducers TRAIL and FasL Due to the ‘delayed’ apoptotic response of SEB-1, it is possible that new gene transcription and protein synthesis is involved in mediating apoptosis. Apoptosis assays in the SEB-1 cell line using actinomycin D (DNA transcription inhibitor) and cyclohexamide (de novo protein synthesis inhibitor) in the presence of 13-cis RA could be performed to test this hypothesis. 13-cis RA- induced apoptosis may not require new gene products and instead directly activate the classical extrinsic or intrinsic apoptosis pathways driven by the activation of the multiple caspases within the cells (Figure 8). We have shown that 13-cis RA induced caspase 8 activation in sebocytes (most likely through TRAIL/TRAIL-R1/R2 activation) to a greater extent than 9-cis RA or ATRA. This would suggest preferential activation of the extrinsic apoptotic pathway by 13-cis RA; however, caspase 8 can activate Bid, leading to its truncation (tBid) and association with the mitochondria triggering intrinsic pathway activation. To fully explore the 152 possibility that 13-cis RA may activate both death receptor and mitochondrial mediated cell death, potential future experiments may include protein lysates subjected to western analysis for many key players involved this pathway such as Bid/tBid, caspase 9, and apoptosis activating factor 1 (APAF1). In order to design alternative therapies to exert potent effects on sebaceous gland function, it is necessary to understand how 13-cis RA transcriptionally regulates a pathway or induces a direct cellular effect that leads to apoptosis.

6.4.2 What RAR-independent events can lead to 13-cis RA-induced apoptosis?

Historically, retinoid compounds are known to exert their specific cellular effect by binding to and activating classical retinoid receptors, resulting in changes in gene expression. It is known that 13-cis RA is not capable of binding to these receptors (Allenby, et al., 1993). Recently, within the cancer field, more attention has been focused on retinoid receptor- independent effects within the cell, such as modulation of signal transduction kinase cascades including mitogen-activated protein kinase (MAPK) pathways: extracellular signal-related- kinases (ERKs), p38MAPK, and stress activated protein kinase/c-Jun N terminal kinase (SAPK/JNK) (Nakagawa, et al., 2003; Olson and Hallahan, 2004; Pettersson, et al., 2004). Activation of these pathways can result in cell survival, growth inhibition or cell death depending on the stimulus the cell receives. 13-cis RA or ATRA is able to induce apoptosis in medulloblastoma cells by phosphorylation of p38MAPK. Apoptosis was triggered by synthetic retinoid CD 437 in ovarian cancer cells by the same p38 phosphorylation; however, ATRA did not trigger apoptosis, indicating, once again, that there is cell-specificity in retinoid actions (Hallahan et al, 2003; Holmes, et al., 2003). ATRA induces sustained activation of ERK1/2 leading to apoptosis in MDA-MB-231 breast cancer cells while in SKBR-3 breast cancer cells protein kinase Cα (PKCα)-ERK pathway is disturbed leading to growth arrest (Nakagawa, et al., 2003; Pettersson, et al., 2004). Very few studies to date have evaluated 13-cis RA actions on these pathways and, in particular, no study has examined these pathways in sebocytes. Since 13-cis RA is known not to bind classical retinoid receptors, modulation of these important signaling cascades is an intriguing possibility that could be tested in future studies. Previous work in our laboratory has demonstrated that SEB-1 sebocytes do express p38, 153 ERK1/2 (p44/42), and SAPK/JNK (p54/46) proteins (Smith, in preparation). The potential activation of the p38/MAPK, ERK and SAPK/JNK pathways by 13-cis RA within SEB-1 sebocytes can be examined by western analysis for both total and phosphorylated forms of p38MAPK, ERK 1/2 and JNK to determine if/when a pathway first becomes activated and if this time course corresponds to apoptosis induction. If a particular pathway (i.e. p38/MAPK) is found to be activated in the presence of 13-cis RA, follow-up experiments utilizing specific pathway inhibitors SB203580, such as specific p38MAPKα/β inhibitor with no interference with ERK or JNK activity, could be used in combination with 13-cis RA to determine if apoptosis is blocked. Based on the scientific literature, we would anticipate that the p38/MAPK pathway or SAPK/JNK pathway will be activated in response to 13-cis RA treatment. These pathways are activated by inflammatory mediators, cytokines (TNF) and cellular stress; genes involved in these cellular processes are affected by 13-cis RA treatment as indicated by our gene expression analysis. These experiments are an initial step to clarifying receptor-independent mechanisms for the 13- cis RA. If 13-cis RA is found to influence a particular signal transduction kinase pathway, new avenues for drug development are open since many inhibitors of specific kinases are being developed. These results would have implications in cancer biology as well as treatment of acne since 13-cis RA is used in cancer chemoprevention and in the treatment of pediatric neuroblastomas, colorectal cancer and some forms of leukemia (Handa, et al., 1997; Maeda, et al., 1996; Recchia et al, 2007; Reynolds et al, 2003).

6.4.3 Lipogenesis vs. Apoptosis: Does 13-cis RA preferentially affect one of these processes?

The sebaceous gland undergoes constant turnover and renewal. Sebocytes divide from the basal layer, migrate into the center of the sebaceous gland and undergo terminal differentiation followed by holocrine rupture leading to the secretion of the lipid product, sebum. The molecular cues and pathways which control this natural process of elimination are still unknown. As previously mentioned, isotretinoin therapy significantly decreases sebum production and secretion. In vivo, 13-cis RA is known to decrease sebum secretion (Strauss, et al., 1980). In cultured human sebocytes, 13-cis RA effectively reduces the synthesis of triglycerides, wax esters, and free fatty acids; however, squalene synthesis remained unchanged while cholesterol synthesis was slightly increased (Zouboulis, et al., 1991). 13-cis RA (1 µM) has been shown to 154 significantly decrease the production of cholesterol, fatty alcohol, cholesterol esters and squalene in SEB-1 sebocytes after 72 hours of treatment (Trivedi, et al., 2006). Histological data also illustrates that isotretinoin drastically decreases the size of the sebaceous gland in vivo (Goldstein, et al., 1982). Initial studies indicated that isotretinoin decreases sebocyte proliferation (Landthaler, et al., 1980) within the sebaceous gland. This finding was confirmed in numerous in vitro cell culture experiments with primary human sebocytes, rat preputial cells and immortalized sebocyte cell lines (Kim et al, 2000; Nelson, et al., 2006; Tsukada, et al., 2000; Wrobel, et al., 2003; Zouboulis, et al., 1991; Zouboulis, et al., 1993). In addition to effects on cell proliferation our studies are the first to show that isotretinoin triggers significant apoptosis within sebaceous glands one-week into therapy as demonstrated by increases in the percentages of TUNEL positive nuclei when compared to baseline biopsies. Several lines of experimental evidence have shown that 13-cis RA affects lipogenesis as well as cell cycle and apoptosis of the sebaceous gland. However, it is unknown which process/pathway, lipogenesis or cell cycle arrest/apoptosis, is the primary target of 13-cis RA within the sebaceous gland. Our gene expression analysis on before and after 8-weeks or one- week isotretinoin treatment combined with parallel hematoxylin and eosin staining on these biopsies may provide some insight to which process is being preferentially affected. Initial gene expression analysis was performed before and after 8-weeks of isotretinoin therapy: 197 genes were significantly up-regulated and 587 genes were significantly down- regulated at 8-weeks when compared to baseline (A.1). Of the 197 genes that were significantly increased, the majority of genes encode structural proteins of the extracellular matrix such as collagens, fibulin and fibronectin. These up-regulated genes are consistent with the known effects of retinoids on the extracellular matrix as reported in studies of photoaging (Weiss, et al., 1988). Many of the down-regulated genes at 8-weeks are involved in the metabolism of steroids, cholesterol and fatty acids, which is consistent with the known decreases in sebaceous gland lipid production induced by 13-cis RA. It seems highly likely that the down-regulation of genes involved in cholesterol and fatty acid metabolism maybe related to the significant decrease in size of the sebaceous gland because hematoxylin and eosin staining revealed significant decreases in gland size at the 8-week time point. It became clear, based on these initial studies that the 8-week time point captures the overall net effect of isotretinoin therapy and that specific effects of isotretinoin on the sebaceous gland occurs prior to this time point. 155 Gene changes at one-week are of particular interest because they may provide clues about the initial changes induced by this drug. Those early gene changes can be broadly categorized as tumor suppressors, protein processors, and genes involved in transfer or binding of ions, amino acids, lipids or retinoids. Within the 42 significantly changed genes, there is increased representation of genes that are involved in or related to processes of ectoderm and epidermal development; pest, pathogen, parasite or external biotic response; and vitamin binding. The one-week gene expression analysis did not indicate any significant changes in genes involved in lipogenesis or key players in apoptosis pathways. Hematoxylin and eosin staining did indicate a decreasing trend in sebaceous gland volume at one-week of isotretinoin treatment. Together, gene expression analysis and hematoxylin and eosin staining suggest that the initial activity of isotretinoin triggers a decrease in sebaceous gland volume. As a result of the ‘shrinking’ sebaceous gland, decreases in genes involved in lipid metabolism are found at 8- weeks. To more definitively test this hypothesis, future gene expression analysis as well as hematoxylin and eosin staining could be performed before and at 4-weeks of isotretinoin treatment. This proposed experiment would allow for a more substantial correlation in the time frame of detectable changes in sebaceous gland volume compared to significant changes in genes involved in lipid metabolism within the sebaceous gland. Previous work in our laboratory demonstrated that insulin-like growth factor 1 (IFG-1) through the activation of its respective cell-surface receptor increases lipogenesis in SEB-1 sebocytes through Sterol Response Element Binding Protein 1 (SREBP-1) transcription factor dependent and independent pathways and has identified the phosphoinositol 3 kinase (PI3-K) pathway as the primary signaling cascade involved in lipogenesis (Smith, et al., 2006)(Smith, in preparation). If 13-cis RA has a direct effect on lipogenesis in SEB-1 sebocytes, it is possible that it influences the actions of SREBP-1 or the PI3-Kinase pathway leading to the documented decreases in lipogenesis (Trivedi, et al., 2006). The PI3-Kinase/AKT pathway is known to mediate cell proliferation, differentiation and cell survival. To date, no study has examined the effects of 13-cis RA or any other retinoid on these pathways in terms of lipogenesis. However, in mouse cell systems, retinoic acid (ATRA) has been shown to both activate and inhibit PI3- Kinase pathway resulting in differentiation and cell cycle arrest. (Bastien et al, 2005) Future studies designed to examine SREBP-1 expression and activity as well as activation or inhibition of the PI3-kinase pathway in the presence of 13-cis RA, may provide direct evidence of 13-cis RA effects on the process of lipogenesis within sebaceous glands. 156 6.4.4 Does TRAIL mediate isotretinoin-induced apoptosis within the sebaceous gland?

Our studies are the first to demonstrate up-regulation of TRAIL by 13-cis RA. We have identified TRAIL as a possible mediator of 13-cis RA-induced apoptosis in SEB-1 sebocytes. TRAIL expression is up-regulated in response to 13-cis RA and purified TRAIL protein induces apoptosis in SEB-1 sebocytes which suggests the TRAIL signaling pathway is intact within our cells. First and foremost, it is important to validate our in vitro results within the sebaceous gland. Immunohistochemistry using antibodies to TRAIL on sections of skin from patients that were treated with isotretinoin for one-week would be an important piece of the puzzle to determine if TRAIL is involved in mediating the effects of isotretinoin in vivo. Future studies could examine the expression of TRAIL receptor isoforms in sebaceous glands and SEB-1 sebocytes. Previous studies have examined the effects of retinoids on TRAIL receptor expression, although no study to date has examined expression within sebaceous glands. Retinoids including fenretinide, synthetic retinoid CD437, and ATRA, increase sensitivity to TRAIL-induced apoptosis by up-regulation of TRAIL-R1 and –R2 receptors, the receptors responsible for receiving and transmitting the TRAIL-mediated apoptotic signal (Kouhara, et al., 2007; Sun, et al., 2000; Sun, et al., 2000). TRAIL-R1 and –R2 receptors and decoy TRAIL-R3 and –R4 are expressed in the epidermis (Bachmann, et al., 2001; Stander and Schwarz, 2005). Experimental evidence suggests that, in normal and tumor cells, the sensitivity and/or resistance to TRAIL mediated apoptosis is associated with the relative levels of ‘active’ death receptors to decoy death receptors. (Daniels et al, 2005; Hesry Vincent, 2006; Kan Kondo, 2006; Qin et al, 2001; Sanlioglu et al, 2007). Immunohistochemical and western analysis of the TRAIL receptor isoforms and their changes with respect to 13-cis RA treatment could be performed in sebaceous glands and SEB-1 sebocytes, respectively, and would provide important validation for TRAIL as a mediator of isotretinoin induced apoptosis within the sebaceous gland. Understanding the mechanism of TRAIL induction by 13-cis RA (i.e. direct or in-direct up-regulation) can provide insight into 13-cis RA’s mechanism of action in acne therapy as well as when used as cancer therapy. Recently, by increasing apoptosis, 13-cis RA alone or in combination with interferon, has been shown to be beneficial in the treatment of some forms of leukemia (Handa, et al., 1997; Maeda, et al., 1996) although the mechanism of apoptosis is unknown. It is possible that TRAIL is responsible this increased apoptosis. The mechanism of 157 TRAIL induction by retinoids is not defined, although advances are being made in this area. Studies in our laboratory of the TRAIL gene promoter region indicates that it does contain an RAR consensus sequence as identified by TESS computer software program. However, promoter mapping experiments conducted by Clarke et. al. demonstrated that there is no retinoic acid response element within 2Kb of the transcription start site of TRAIL (Clarke, et al., 2004). The latter is in agreement with our previous studies within SEB-1 sebocytes demonstrating that apoptosis was not blocked by a RAR pan-antagonist, suggesting an RAR independent mechanism to apoptosis induction. TRAIL up-regulation by 13-cis RA, therefore, most likely occurs by an indirect mechanism. In support of this hypothesis, Interferon regulatory factor 1 (IRF1) was identified as a critical factor in mediating TRAIL induction by retinoic acid in NB4 APL leukemia cells and SK-BR-3 breast cancer cells (Clarke, et al., 2004). Interestingly, 13-cis RA significantly up-regulates IRF1 gene expression (2.42 fold increase) in SEB-1 sebocytes (A.2). It may be possible that 13-cis RA induced TRAIL up-regulation in SEB-1 sebocytes is due to increases in IRF1 expression. To definitively test this hypothesis, studies utilizing siRNA knockdown of IRF1 in the presence of 13-cis RA could be performed. Furthermore, additional studies are needed to assess whether retinoid receptor activation is required for this increase in TRAIL or IRF1 expression. Our studies have shown that 13-cis RA induced apoptosis is most likely not mediated by RAR activation as studies with RAR pan-antagonist AGN 193109 do not block apoptosis in the presence of 13-cis RA. If TRAIL mediates 13-cis RA induced apoptosis, we would hypothesize that TRAIL expression is not mediated by RAR activation. Studies proposed above would examine the role of IRF1 in mediating TRAIL expression in response to 13-cis RA and it is equally important to determine if RAR activation is required for increases in IRF-1 if IRF-1 mediates increases in TRAIL expression. The IRF1 promoter is not available in the TESS database and no previous study has directly accessed whether a RAR or RXR consensus sequence is located within the IRF-1 promoter; although previous studies have shown that IRF1 is induced by ATRA which would suggest that it does contain a retinoid receptor response element within its promoter. Studies utilizing RAR pan-antagonist AGN 193109 or siRNA to the RAR subtypes in the presence of 13- cis RA and examining IRF1 and TRAIL mRNA and protein expression via quantitative polymerase chain reaction (QPCR) and western blotting could be performed to address whether RAR receptor activation is required for increased expression of both genes. Understanding the mechanism by which 13-cis RA regulates TRAIL expression can lead to advances in acne treatments as well as a greater understanding of its use as cancer therapy. 158 If future experiments definitively demonstrate that TRAIL mediates apoptosis in response to 13- cis RA treatment in sebaceous glands, then drugs/compounds that increase TRAIL expression, its active receptors’ expression or sensitivity to TRAIL mediated apoptosis may be useful as an acne treatment. Histone deacetylase inhibitors and TRAIL receptor agonistic mono-antibodies have been shown increase the TRAIL sensitivity of prostate cancer cells with minimal effect on the normal prostate epithelium and are being used in conjunction with current prostate chemo- and gene- therapies (Kasman et al, 2006; Shimada et al, 2007; Vanoosten et al, 2007). Asoprisnil, a novel selective progesterone receptor modulator, up-regulates TRAIL and its active receptors (TRAIL-R1/R2) in leiomyoma cells resulting in increased apoptosis, however, no effect is noticed on the surrounding normal smooth muscle cells (Chen et al, 2006; Sasaki et al, 2007). Sebaceous glands are known to express the progesterone receptor (Kariya, et al., 2005) and it is not yet known what role the progesterone receptor and its activation may play in the development or resolution of acne. Additional research into these drugs and their effects on sebaceous gland physiology may illuminate an alternative to isotretinoin for acne treatment.

6.4.5 Why are the apoptotic effects of 13-cis RA limited to sebocytes and do not occur within keratinocytes?

Several studies indicate that the effects of retinoids on cell proliferation, differentiation, and apoptosis are retinoid- or cell-type specific. Histological evidence indicates the effects of 13- cis RA are most profound on sebaceous glands, with little to no alteration in the surrounding dermis or epidermis of the skin (Landthaler, et al., 1980). Our studies clearly demonstrated that 13-cis RA triggers apoptosis within sebocytes and does not affect keratinocytes (Nelson, et al., 2006) as assayed in both human patients and in vitro cell culture assays. To date, no study has detected apoptosis within keratinocytes in response to 13-cis RA treatment, although other retinoids including ATRA and tazarotene (RAR β/γ selective agonist) have been shown to induce apoptosis in HaCaT keratinocytes (Louafi, et al., 2003; Papoutsaki, et al., 2004). The reason for the cell-specificity of 13-cis RA-induced apoptosis is not yet known but initial studies done within our laboratory and others may provide some clues. Gene expression analysis revealed that 13-cis RA significantly affects different genes in SEB-1 sebocytes than those affected in HaCaT keratinocytes, with only 9 significantly changed genes in common. Clearly, 13-cis RA has cell-specific effects as less than 10% of changed genes are in commonly changed between the two cell lines. Within significantly changed genes 159 from the HaCaT gene array, “cellular differentiation” and “morphogenesis” gene ontology classifications were significantly enhanced (data not shown). Significantly up-regulated genes that fall into these classifications include: involucrin, insulin-like growth factor binding proteins 3 and 6 (IGFBP3, 6), bone morphogenetic protein 3 (BMP3) and kallikrein 5 and 6 (KLK5, 6) and all these genes are known to regulate cellular differentiation and proliferation in a variety of different cell types including keratinocytes (Eckert et al, 2004; Eckert and Green, 1986; Edmondson et al, 2005; Faucheux et al, 1997; Kishibe et al, 2007). This supports the idea that 13-cis RA does not induce an apoptosis pathway (as in sebocytes) but instead influences other cellular pathways within keratinocytes. Although, the ontology term of ‘apoptosis’ was not indicated within the significantly changed genes within SEB-1 sebocytes, numerous genes involved in apoptosis were up-regulated including TRAIL, FasL, LCN2, IGFBP3, IRF1. The question then becomes “why is 13-cis RA affecting these pathways and not apoptosis?” A possible answer to this question is provided within the list of significantly changed genes. Gene expression analysis revealed increased expression of TRIM31 that encodes an E3 ubiquitin ligase, and P450RAI2 (CYP26A), a potent retinoic acid 4-hydroxylase after 13-cis RA treatment. Up-regulation of these genes suggests that within HaCaT keratinocytes, 13-cis RA is rapidly degraded/metabolized and implies that HaCaT keratinocytes have powerful mechanisms in place to protect against the actions of retinoids. Törmä et. al. demonstrated that HaCaT keratinocytes, when compared to normal epidermal keratinocytes, have lower levels of retinoid binding proteins, increased metabolism of retinol and retinoic acid and high levels of p450RAI; all of which suggest that HaCaT keratinocytes do not achieve or maintain high levels of retinoids intracellularly (Torma et al, 1999) and it is possible that these lower levels affect differentiation pathways and not activate apoptotic pathways. Since 1925, retinoids have been shown to influence and normalize differentiation in keratinocytes but the exact mechanism of how this is accomplished is unknown (Wolbach and Howe, 1925). Future investigations into these cellular pathways and its possible ‘retinoid protection’ enzymes would be of interest to completely understand the actions of 13-cis RA on skin and its beneficial effects on keratinizing disorders. 160 6.5 Conclusion

The studies in this thesis are the beginning to understanding the effects of 13-cis RA on the sebaceous gland and its mechanism of action in sebum suppression. Our studies suggest that 13-cis RA mediates its sebosuppressive effect through preferential induction of apoptosis in sebaceous glands. Work in this area is far from complete. A fundamental understanding of 13- cis RA actions in sebaceous glands is needed before meaningful investigations into safer drug alternatives for acne treatment can be pursued.

ºoº ºoº ºoº

Appendix A

Supplemental gene expression array tables

A.1 All significantly changed genes after 8 weeks isotretinoin therapy

Some genes may be listed twice; indicates separate probe sets on Affymetrix gene array chips. Fold Gene Title Gene Symbol Change 2.54 microseminoprotein, beta- MSMB 2.03 microseminoprotein, beta- MSMB 2.00 kraken-like dJ222E13.1 1.98 ------1.85 coagulation factor C homolog, cochlin (Limulus polyphemus) COCH lymphocyte cytosolic protein 2 (SH2 domain containing leukocyte 1.77 protein of 76kDa) LCP2 1.73 fibulin 1 FBLN1 1.71 hypothetical protein MGC27165 MGC27165 1.68 carbohydrate (N-acetylglucosamine 6-O) sulfotransferase 6 CHST6 1.67 KIAA0527 protein KIAA0527 1.67 fibulin 1 FBLN1 1.66 collagen, type VI, alpha 2 COL6A2 1.64 collagen, type V, alpha 1 COL5A1 1.63 SAM domain, SH3 domain and nuclear localisation signals, 1 SAMSN1 1.63 filamin A, alpha (actin binding protein 280) FLNA 1.62 butyrophilin, subfamily 3, member A3 BTN3A3 1.61 collagen, type V, alpha 1 COL5A1 1.61 protein phosphatase 1, regulatory (inhibitor) subunit 16B PPP1R16B 1.60 tryptase beta 2 TPSB2 1.60 insulin-like growth factor binding protein 4 IGFBP4 1.60 microfibrillar-associated protein 2 MFAP2 1.59 tryptase beta 2 TPSB2 1.59 tryptase beta 2 TPSB2 1.57 ------1.57 slit homolog 3 (Drosophila) SLIT3 1.56 Rho-related BTB domain containing 3 RHOBTB3 1.56 insulin-like growth factor binding protein 5 IGFBP5 1.56 microfibrillar-associated protein 4 MFAP4 1.56 tryptase beta 2 TPSB2 1.55 collagen, type VI, alpha 2 COL6A2 1.55 collagen, type IV, alpha 2 COL4A2 1.55 fibulin 1 FBLN1 1.54 natural killer cell transcript 4 NK4 1.54 hematopoietic cell-specific Lyn substrate 1 HCLS1 1.53 tenascin N TNN 162 1.53 fibronectin 1 /// fibronectin 1 FN1 1.53 collagen, type VI, alpha 1 COL6A1 1.53 protease, serine, 11 (IGF binding) PRSS11 1.52 collagen, type V, alpha 2 COL5A2 1.52 procollagen C-endopeptidase enhancer PCOLCE 1.51 protein tyrosine phosphatase, receptor type, C PTPRC integrin, beta 2 (antigen CD18 (p95), lymphocyte function-associated 1.51 antigen 1; macrophage antigen 1 (mac-1) beta subunit) ITGB2 1.51 collagen, type V, alpha 1 COL5A1 1.51 laminin, alpha 2 (merosin, congenital muscular dystrophy) LAMA2 1.51 tryptase beta 2 TPSB2 1.50 tryptase beta 2 TPSB2 matrix metalloproteinase 2 (gelatinase A, 72kDa gelatinase, 72kDa 1.50 type IV collagenase) MMP2 1.49 FLJ00133 protein FLJ00133 1.49 protein phosphatase 1, regulatory (inhibitor) subunit 16B PPP1R16B 1.49 WNT inhibitory factor 1 WIF1 1.49 ------1.48 prostaglandin D2 synthase 21kDa (brain) PTGDS 1.48 fibronectin 1 FN1 1.48 thrombospondin 4 THBS4 1.48 complement component 1, s subcomponent C1S 1.48 protein tyrosine phosphatase, receptor type, G PTPRG 1.47 dynamin 1 DNM1 1.47 complement component 3 C3 serine (or cysteine) proteinase inhibitor, clade G (C1 inhibitor), member 1.47 1, (angioedema, hereditary) SERPING1 1.47 DKFZP586K1520 protein DKFZP586K1520 lymphocyte cytosolic protein 2 (SH2 domain containing leukocyte 1.47 protein of 76kDa) LCP2 1.46 fibronectin 1 FN1 1.46 insulin-like growth factor binding protein 5 IGFBP5 1.46 forkhead box D1 FOXD1 1.46 platelet-derived growth factor receptor, beta polypeptide PDGFRB 1.46 DKFZP564J102 protein DKFZP564J102 1.46 KIAA1102 protein KIAA1102 1.46 collagen, type XIV, alpha 1 (undulin) COL14A1 1.45 CDW52 antigen (CAMPATH-1 antigen) CDW52 1.45 ribonuclease T2 RNASET2 colony stimulating factor 1 receptor, formerly McDonough feline 1.44 sarcoma viral (v-fms) oncogene homolog CSF1R 1.44 ankyrin 2, neuronal ANK2 1.44 major histocompatibility complex, class II, DP alpha 1 HLA-DPA1 1.44 huntingtin interacting protein 1 HIP1 1.43 collagen, type VI, alpha 3 COL6A3 1.43 CD3Z antigen, zeta polypeptide (TiT3 complex) CD3Z 1.43 v-yes-1 Yamaguchi sarcoma viral related oncogene homolog LYN 1.43 major histocompatibility complex, class II, DP beta 1 HLA-DPB1 1.43 matrix metalloproteinase 23B MMP23B 1.42 fibronectin 1 FN1 1.42 TBC1 (tre-2/USP6, BUB2, cdc16) domain family, member 1 TBC1D1 1.42 Ras association (RalGDS/AF-6) domain family 2 RASSF2 163 1.42 tenascin XB TNXB 1.42 lysyl oxidase-like 2 LOXL2 1.42 mannosidase, alpha, class 1A, member 1 MAN1A1 1.42 insulin-like growth factor binding protein 5 IGFBP5 1.42 LIM domain only 2 (rhombotin-like 1) LMO2 1.42 Src-like-adaptor SLA 1.41 G protein-coupled receptor 124 GPR124 tissue inhibitor of metalloproteinase 3 (Sorsby fundus dystrophy, 1.41 pseudoinflammatory) TIMP3 1.41 chondroitin sulfate proteoglycan 2 (versican) CSPG2 1.41 TRAF3-interacting Jun N-terminal kinase (JNK)-activating modulator T3JAM 1.41 neurotrophic tyrosine kinase, receptor, type 2 NTRK2 1.40 phosphatidic acid phosphatase type 2B PPAP2B tissue inhibitor of metalloproteinase 3 (Sorsby fundus dystrophy, 1.40 pseudoinflammatory) TIMP3 1.39 chemokine (C-X-C motif) ligand 12 (stromal cell-derived factor 1) CXCL12 1.39 KIAA1102 protein KIAA1102 1.39 RNA binding protein with multiple splicing RBPMS Homo sapiens cDNA FLJ36690 fis, clone UTERU2008707, highly similar to COMPLEMENT C1R COMPONENT PRECURSOR (EC 1.39 3.4.21.41). --- 1.38 CD3D antigen, delta polypeptide (TiT3 complex) CD3D 1.38 lymphocyte-specific protein tyrosine kinase LCK 1.38 BTB (POZ) domain containing 3 BTBD3 fibroblast growth factor receptor 3 (achondroplasia, thanatophoric 1.37 dwarfism) FGFR3 1.37 lipase, hepatic LIPC 1.37 slit homolog 3 (Drosophila) SLIT3 1.37 polymerase I and transcript release factor PTRF 1.37 nuclear receptor subfamily 2, group F, member 1 NR2F1 1.37 Rac/Cdc42 guanine nucleotide exchange factor (GEF) 6 ARHGEF6 1.37 reticulon 1 RTN1 1.37 ------1.37 fucosyltransferase 8 (alpha (1,6) fucosyltransferase) FUT8 1.37 cystatin C (amyloid angiopathy and cerebral hemorrhage) CST3 Homo sapiens T cell receptor beta chain BV20S1 BJ1-5 BC1 mRNA, 1.36 complete cds --- 1.36 up-regulated in liver cancer 1 UPLC1 1.36 protein S (alpha) PROS1 fibroblast growth factor receptor 1 (fms-related tyrosine kinase 2, 1.35 Pfeiffer syndrome) FGFR1 1.35 plakophilin 1 (ectodermal dysplasia/skin fragility syndrome) PKP1 1.35 CUG triplet repeat, RNA binding protein 2 CUGBP2 1.35 major histocompatibility complex, class I, E HLA-E 1.35 ------1.35 CD2 antigen (p50), sheep red blood cell receptor CD2 1.35 up-regulated in liver cancer 1 UPLC1 1.35 HLA-G histocompatibility antigen, class I, G HLA-G 1.35 adrenergic, alpha-2A-, receptor ADRA2A 1.34 Rho GDP dissociation inhibitor (GDI) beta ARHGDIB 1.34 discs, large homolog 5 (Drosophila) DLG5 1.34 major histocompatibility complex, class I, F HLA-F 164 1.34 CD37 antigen CD37 1.34 interferon-induced protein 35 IFI35 1.34 major histocompatibility complex, class II, DQ beta 2 HLA-DQB2 1.34 likely ortholog of mouse IRA1 protein IRA1 1.34 hypothetical protein FLJ10770 KIAA1579 1.33 laminin, gamma 1 (formerly LAMB2) LAMC1 1.33 cat eye syndrome chromosome region, candidate 1 CECR1 1.33 collagen, type IV, alpha 2 COL4A2 1.33 phospholipase C-like 2 PLCL2 1.33 immunoglobulin superfamily containing leucine-rich repeat ISLR 1.33 KIAA1102 protein KIAA1102 Homo sapiens T cell receptor beta chain BV20S1 BJ1-5 BC1 mRNA, 1.33 complete cds --- 1.33 FK506 binding protein 10, 65 kDa FKBP10 1.33 Lysosomal-associated multispanning membrane protein-5 LAPTM5 1.33 CD97 antigen CD97 1.33 drebrin 1 DBN1 1.32 biglycan BGN 1.32 latent transforming growth factor beta binding protein 1 LTBP1 1.32 sprouty-related, EVH1 domain containing 2 SPRED2 1.32 dihydropyrimidinase-like 2 DPYSL2 1.32 protein tyrosine phosphatase, receptor type, M PTPRM 1.32 SET and MYND domain containing 3 SMYD3 1.32 HLA-G histocompatibility antigen, class I, G HLA-G 1.32 zinc finger protein FLJ10697 1.32 CD81 antigen (target of antiproliferative antibody 1) CD81 1.32 EGF-containing fibulin-like extracellular matrix protein 2 EFEMP2 1.31 lipoma HMGIC fusion partner LHFP MADS box transcription enhancer factor 2, polypeptide C (myocyte 1.31 enhancer factor 2C) MEF2C 1.31 F-box and leucine-rich repeat protein 7 FBXL7 1.31 hypothetical gene BC008967 BC008967 1.31 CD34 antigen CD34 S100 calcium binding protein A4 (calcium protein, calvasculin, 1.31 metastasin, murine placental homolog) S100A4 ras-related C3 botulinum toxin substrate 2 (rho family, small GTP 1.31 binding protein Rac2) RAC2 1.31 alpha-2-macroglobulin A2M 1.30 major histocompatibility complex, class II, DQ beta 1 HLA-DQB1 1.30 receptor tyrosine kinase-like orphan receptor 1 ROR1 1.30 potassium channel tetramerisation domain containing 12 KCTD12 1.30 chondroitin sulfate proteoglycan 2 (versican) CSPG2 1.30 interferon induced transmembrane protein 1 (9-27) IFITM1 1.30 transforming growth factor, beta receptor II (70/80kDa) TGFBR2 1.29 adaptor-related protein complex 2, beta 1 subunit AP2B1 1.29 KIAA1518 protein KIAA1518 1.29 RAB6 interacting protein 1 RAB6IP1 1.29 hypothetical protein FLJ21868 FLJ21868 1.29 protein tyrosine phosphatase, receptor type, B PTPRB 1.28 hypothetical protein FLJ11588 FLJ11588 1.28 caldesmon 1 CALD1 1.28 protein tyrosine phosphatase, non-receptor type substrate 1 PTPNS1 165 1.28 selectin P ligand SELPLG 1.27 chromosome 1 open reading frame 21 C1orf21 1.27 phosphoglycerate dehydrogenase PHGDH 1.27 quaking homolog, KH domain RNA binding (mouse) QKI 1.27 phospholipid transfer protein PLTP 1.26 intercellular adhesion molecule 2 ICAM2 1.26 fms-related tyrosine kinase 3 ligand FLT3LG 1.26 hypothetical protein DKFZp564K0822 DKFZP564K0822 1.25 ATPase, Ca++ transporting, plasma membrane 4 ATP2B4 1.25 microtubule-associated protein 4 MAP4 1.25 caldesmon 1 CALD1 1.25 brain abundant, membrane attached signal protein 1 BASP1 1.24 DIX domain containing 1 DIXDC1 1.24 A kinase (PRKA) anchor protein 11 AKAP11 1.23 solute carrier family 39 (zinc transporter), member 14 SLC39A14 1.23 CDC14 cell division cycle 14 homolog B (S. cerevisiae) CDC14B integrin, alpha V (vitronectin receptor, alpha polypeptide, antigen 1.23 CD51) ITGAV 1.22 transducer of ERBB2, 2 TOB2 1.22 ring finger protein 38 RNF38 1.21 fasciculation and elongation protein zeta 2 (zygin II) FEZ2 1.21 myocardin-related transcription factor B MRTF-B 1.20 glutamate receptor, ionotropic, N-methyl D-aspartate-like 1A GRINL1A -7.92 hydroxyacid oxidase 2 (long chain) HAO2 hydroxy-delta-5-steroid dehydrogenase, 3 beta- and steroid delta- -7.88 isomerase 1 HSD3B1 -6.91 hypothetical protein FLJ11106 FLJ11106 -6.78 solute carrier organic anion transporter family, member 4C1 SLCO4C1 -6.58 fatty acid desaturase 1 FADS1 -6.39 fatty acid desaturase 1 FADS1 -6.25 PDZ domain containing 1 PDZK1 glycine dehydrogenase (decarboxylating; glycine decarboxylase, -5.91 glycine cleavage system protein P) GLDC -5.83 fatty acid binding protein 7, brain FABP7 -5.62 hypothetical protein FLJ10462 FLJ10462 phospholipase A2, group VII (platelet-activating factor acetylhydrolase, -5.52 plasma) PLA2G7 -5.47 fatty acid desaturase 2 FADS2 -4.90 galanin GAL -4.79 arachidonate 15-lipoxygenase, second type ALOX15B -4.79 fatty acid binding protein 7, brain FABP7 -4.77 fatty acid desaturase 1 FADS1 -4.23 histone 1, H1c /// histone 1, H1c HIST1H1C -4.21 mucin 1, transmembrane MUC1 -3.99 3-hydroxy-3-methylglutaryl-Coenzyme A synthase 1 (soluble) HMGCS1 -3.92 insulin induced gene 1 INSIG1 -3.91 lipidosin BG1 sterol O-acyltransferase (acyl-Coenzyme A: cholesterol -3.84 acyltransferase) 1 SOAT1 -3.80 lipidosin BG1 -3.73 dehydrogenase/reductase (SDR family) member 9 DHRS9 -3.72 variable charge, Y-linked, 2 VCY2 166 -3.71 apolipoprotein C-I APOC1 -3.71 carnitine acetyltransferase CRAT -3.67 SA hypertension-associated homolog (rat) SAH -3.65 UDP glycosyltransferase 2 family, polypeptide A1 UGT2A1 -3.59 hypothetical protein MAC30 MAC30 hyperpolarization activated cyclic nucleotide-gated potassium channel -3.54 3 HCN3 -3.49 phosphodiesterase 6A, cGMP-specific, rod, alpha PDE6A -3.48 ureidopropionase, beta UPB1 -3.48 hypothetical protein MAC30 MAC30 -3.40 mucin 1, transmembrane MUC1 -3.38 hypothetical protein MAC30 MAC30 -3.37 SA hypertension-associated homolog (rat) SAH -3.34 transitional epithelia response protein TERE1 -3.34 hydroxysteroid (11-beta) dehydrogenase 1 HSD11B1 -3.34 peroxisomal long-chain acyl-coA thioesterase ZAP128 Homo sapiens mRNA; cDNA DKFZp564P142 (from clone -3.24 DKFZp564P142) --- -3.24 insulin induced gene 1 INSIG1 -3.20 solute carrier family 26, member 3 SLC26A3 -3.17 homogentisate 1,2-dioxygenase (homogentisate oxidase) HGD -3.11 NAD(P) dependent steroid dehydrogenase-like H105E3 farnesyl diphosphate synthase (farnesyl pyrophosphate synthetase, -3.11 dimethylallyltranstransferase, geranyltranstransferase) FDPS -3.06 solute carrier family 27 (fatty acid transporter), member 2 SLC27A2 -3.05 peroxisomal trans 2-enoyl CoA reductase PECR -3.04 DESC1 protein DESC1 -3.02 cytochrome P450, family 4, subfamily F, polypeptide 8 CYP4F8 -3.02 cell death-inducing DFFA-like effector a CIDEA -2.97 fatty acid 2-hydroxylase FA2H -2.97 chitinase 3-like 1 (cartilage glycoprotein-39) CHI3L1 -2.95 chitinase 3-like 1 (cartilage glycoprotein-39) CHI3L1 steroid-5-alpha-reductase, alpha polypeptide 1 (3-oxo-5 alpha-steroid -2.91 delta 4-dehydrogenase alpha 1) SRD5A1 -2.83 glycerol kinase GK -2.83 malic enzyme 1, NADP(+)-dependent, cytosolic ME1 acetyl-Coenzyme A acetyltransferase 2 (acetoacetyl Coenzyme A -2.78 thiolase) ACAT2 -2.78 insulin induced gene 1 INSIG1 -2.78 7-dehydrocholesterol reductase DHCR7 -2.76 ------2.75 hypoxia-inducible protein 2 HIG2 -2.74 fructose-1,6-bisphosphatase 1 FBP1 -2.64 fatty acid synthase FASN -2.64 ------2.63 chromosome 6 open reading frame 105 C6orf105 -2.62 histone 1, H2bc HIST1H2BC -2.60 3-hydroxy-3-methylglutaryl-Coenzyme A reductase HMGCR -2.59 arginase, type II ARG2 -2.54 abhydrolase domain containing 5 ABHD5 steroid-5-alpha-reductase, alpha polypeptide 1 (3-oxo-5 alpha-steroid -2.53 delta 4-dehydrogenase alpha 1) SRD5A1 167 -2.51 abhydrolase domain containing 5 ABHD5 -2.50 sterol-C4-methyl oxidase-like SC4MOL -2.50 acyl-Coenzyme A dehydrogenase, C-4 to C-12 straight chain ACADM -2.50 glycerol kinase GK ELOVL family member 5, elongation of long chain fatty acids -2.49 (FEN1/Elo2, SUR4/Elo3-like, yeast) ELOVL5 -2.49 B-cell receptor-associated protein 29 BCAP29 -2.48 solute carrier family 27 (fatty acid transporter), member 2 SLC27A2 -2.48 glycerol kinase GK -2.48 chromosome 14 open reading frame 137 C14orf137 -2.45 myogenic factor 3 /// myogenic factor 3 MYOD1 -2.45 malic enzyme 1, NADP(+)-dependent, cytosolic ME1 -2.44 apolipoprotein C-I APOC1 -2.44 CGI-100 protein CGI-100 acetyl-Coenzyme A acyltransferase 2 (mitochondrial 3-oxoacyl- -2.43 Coenzyme A thiolase) ACAA2 -2.42 glycerol kinase GK -2.42 deoxyribonuclease I-like 2 DNASE1L2 -2.42 cut-like 2 (Drosophila) CUTL2 -2.42 NAD(P) dependent steroid dehydrogenase-like H105E3 -2.39 nuclear receptor binding factor 1 CGI-63 -2.38 cystathionase (cystathionine gamma-lyase) CTH -2.35 paraoxonase 3 PON3 -2.33 emopamil binding protein (sterol isomerase) EBP -2.33 melanocortin 5 receptor MC5R -2.32 7-dehydrocholesterol reductase DHCR7 -2.32 mevalonate (diphospho) decarboxylase MVD -2.30 dopa decarboxylase (aromatic L-amino acid decarboxylase) DDC -2.30 branched chain aminotransferase 2, mitochondrial BCAT2 -2.29 superoxide dismutase 2, mitochondrial SOD2 -2.29 fatty-acid-Coenzyme A ligase, long-chain 2 FACL2 -2.29 intraflagellar transport protein IFT20 LOC90410 -2.27 carnitine acetyltransferase CRAT -2.27 retinol dehydrogenase 11 (all-trans and 9-cis) RDH11 -2.27 G protein-coupled receptor 64 GPR64 -2.26 adipose differentiation-related protein ADFP -2.25 phosphatidylcholine transfer protein PCTP -2.25 steroid-5-alpha-reductase, alpha polypeptide 1 SRD5A1 -2.23 lysophosphatidic acid phosphatase ACP6 -2.22 acyl-Coenzyme A oxidase 2, branched chain ACOX2 ELOVL family member 5, elongation of long chain fatty acids -2.22 (FEN1/Elo2, SUR4/Elo3-like, yeast) ELOVL5 -2.21 Krueppel-related zinc finger protein H-plk -2.20 abhydrolase domain containing 5 ABHD5 -2.20 molybdenum cofactor sulfurase MOCOS -2.20 dual specificity phosphatase 4 DUSP4 -2.20 cytochrome b-5 CYB5 -2.20 retinol dehydrogenase 11 (all-trans and 9-cis) RDH11 -2.18 potassium inwardly-rectifying channel, subfamily J, member 15 KCNJ15 branched chain keto acid dehydrogenase E1, beta polypeptide (maple -2.17 syrup urine disease) BCKDHB -2.16 transmembrane 7 superfamily member 2 TM7SF2 168 -2.15 cytochrome b-5 CYB5 -2.15 cytochrome P450, family 4, subfamily F, polypeptide 2 CYP4F2 acetyl-Coenzyme A acetyltransferase 2 (acetoacetyl Coenzyme A -2.15 thiolase) ACAT2 -2.13 emopamil binding protein (sterol isomerase) EBP -2.12 zinc finger protein 43 (HTF6) ZNF43 -2.11 cytochrome b-5 CYB5 -2.11 propionyl Coenzyme A carboxylase, beta polypeptide PCCB -2.11 acyl-Coenzyme A dehydrogenase family, member 8 ACAD8 -2.11 phosphomevalonate kinase PMVK -2.10 serum/glucocorticoid regulated kinase 2 SGK2 -2.09 N-acetylneuraminate pyruvate lyase (dihydrodipicolinate synthase) NPL Homo sapiens cDNA FLJ16053 moderately similar to MITOGEN- -2.09 ACTIVATED PROTEIN KINASE KINASE KINASE 5 (EC 2.7.1.-) --- -2.08 peroxisomal membrane protein 2, 22kDa PXMP2 -2.08 histone 1, H2ae HIST1H2AE -2.06 selenoprotein X, 1 SEPX1 -2.06 histone 1, H2bg HIST1H2BG -2.06 sorbitol dehydrogenase SORD Homo sapiens transcribed sequence with moderate similarity to protein -2.06 ref:NP_084526.1 (M.musculus) h --- -2.05 chromosome 22 open reading frame 20 C22orf20 -2.05 aldehyde dehydrogenase 3 family, member B2 ALDH3B2 1-acylglycerol-3-phosphate O-acyltransferase 3 /// 1-acylglycerol-3- -2.04 phosphate O-acyltransferase 3 AGPAT3 solute carrier family 25 (mitochondrial carrier; peroxisomal membrane -2.03 protein, 34kDa), member 17 SLC25A17 -2.02 peroxisomal membrane protein 4, 24kDa PXMP4 -2.02 alpha-methylacyl-CoA racemase AMACR -2.01 isopentenyl-diphosphate delta isomerase IDI1 -2.01 glucose-6-phosphate dehydrogenase G6PD -1.99 pyruvate kinase, liver and RBC PKLR -1.98 isopentenyl-diphosphate delta isomerase IDI1 -1.97 transketolase (Wernicke-Korsakoff syndrome) TKT -1.96 calcium binding protein P22 CHP -1.95 KIAA0626 gene product KIAA0626 -1.95 aconitase 1, soluble ACO1 -1.94 mevalonate kinase (mevalonic aciduria) MVK -1.92 ras homolog gene family, member I ARHI -1.92 hypothetical protein MGC4172 MGC4172 -1.92 hypothetical protein FLJ22679 FLJ22679 acetyl-Coenzyme A acyltransferase 1 (peroxisomal 3-oxoacyl- -1.92 Coenzyme A thiolase) ACAA1 -1.92 acetoacetyl-CoA synthetase AACS -1.92 3-hydroxy-3-methylglutaryl-Coenzyme A reductase HMGCR -1.92 peroxisomal biogenesis factor 11A PEX11A -1.91 peroxisomal biogenesis factor 11A PEX11A -1.91 carnitine palmitoyltransferase II CPT2 -1.91 LAG1 longevity assurance homolog 4 (S. cerevisiae) LASS4 -1.91 hypothetical protein LOC283537 LOC283537 -1.90 glutathione peroxidase 3 (plasma) GPX3 -1.90 hypothetical protein DKFZp547M236 DKFZp547M236 169 -1.90 gamma-aminobutyric acid (GABA) A receptor, alpha 4 GABRA4 -1.90 ------elongation of very long chain fatty acids (FEN1/Elo2, SUR4/Elo3, -1.89 yeast)-like 4 ELOVL4 -1.89 pyruvate dehydrogenase (lipoamide) alpha 1 PDHA1 -1.89 farnesyl-diphosphate farnesyltransferase 1 FDFT1 -1.89 3-hydroxyisobutyryl-Coenzyme A hydrolase HIBCH -1.89 lipin 1 LPIN1 -1.88 peroxisomal biogenesis factor 3 PEX3 solute carrier family 25 (mitochondrial carrier; peroxisomal membrane -1.87 protein, 34kDa), member 17 SLC25A17 -1.86 hypothetical protein HSPC111 HSPC111 -1.86 glutathione peroxidase 3 (plasma) GPX3 -1.85 2',5'-oligoadenylate synthetase 1, 40/46kDa OAS1 -1.85 progesterone receptor membrane component 1 PGRMC1 -1.85 sterol-C5-desaturase (ERG3 delta-5-desaturase homolog, fungal)-like SC5DL -1.85 ------1.85 electron-transferring-flavoprotein dehydrogenase ETFDH -1.85 chitinase 3-like 2 CHI3L2 -1.85 zinc finger protein 254 ZNF254 -1.85 superoxide dismutase 2, mitochondrial SOD2 -1.84 farnesyl-diphosphate farnesyltransferase 1 FDFT1 -1.84 brain expressed, associated with Nedd4 BEAN -1.84 synapsin II SYN2 zinc finger protein 145 (Kruppel-like, expressed in promyelocytic -1.83 leukemia) ZNF145 -1.82 zinc finger protein 91 (HPF7, HTF10) ZNF91 -1.82 ------1.82 transketolase (Wernicke-Korsakoff syndrome) TKT -1.81 enoyl Coenzyme A hydratase domain containing 1 ECHDC1 phosphogluconate dehydrogenase /// phosphogluconate -1.81 dehydrogenase PGD -1.81 ATP citrate lyase ACLY -1.80 electron-transferring-flavoprotein dehydrogenase ETFDH -1.79 isocitrate dehydrogenase 1 (NADP+), soluble IDH1 -1.79 peroxisome biogenesis factor 13 PEX13 -1.79 zinc finger protein 43 (HTF6) ZNF43 -1.78 peroxisomal biogenesis factor 3 PEX3 -1.78 aminolevulinate, delta-, synthase 1 ALAS1 -1.78 glyceronephosphate O-acyltransferase GNPAT -1.77 aldehyde dehydrogenase 3 family, member B2 ALDH3B2 -1.77 sorbitol dehydrogenase SORD -1.77 ATP citrate lyase ACLY -1.77 solute carrier family 15 (oligopeptide transporter), member 1 SLC15A1 -1.77 pleckstrin homology-like domain, family A, member 2 PHLDA2 -1.76 NAD kinase FLJ13052 -1.76 histone 1, H2bk HIST1H2BK -1.75 FLJ23311 protein FLJ23311 -1.75 sorting nexin 13 SNX13 -1.75 aminoacylase 1 ACY1 -1.75 TRAF6-inhibitory zinc finger protein TIZ -1.75 thiosulfate sulfurtransferase (rhodanese) TST 170 -1.75 lipin 1 LPIN1 -1.74 fatty-acid-Coenzyme A ligase, long-chain 5 FACL5 -1.74 2',5'-oligoadenylate synthetase 1, 40/46kDa OAS1 -1.73 cytochrome P450, family 4, subfamily F, polypeptide 3 CYP4F3 -1.73 chromosome 21 open reading frame 5 C21orf5 -1.73 anterior gradient 2 homolog (Xenopus laevis) AGR2 -1.72 fatty-acid-Coenzyme A ligase, long-chain 2 FACL2 -1.72 creatine kinase, mitochondrial 1 (ubiquitous) CKMT1 -1.72 zinc finger protein 165 ZNF165 -1.72 YDD19 protein YDD19 -1.71 ------1.71 H2B histone family, member S H2BFS -1.71 acetyl-Coenzyme A carboxylase alpha ACACA -1.71 alpha-methylacyl-CoA racemase AMACR -1.71 ATP citrate lyase ACLY -1.71 dicarbonyl/L-xylulose reductase DCXR -1.70 calsyntenin 3 CLSTN3 -1.70 desmocollin 2 DSC2 solute carrier family 31 (copper transporters), member 1 /// solute -1.69 carrier family 31 (copper transporters), member 1 SLC31A1 -1.69 calmodulin-like 3 CALML3 solute carrier family 25 (mitochondrial carrier; Graves disease -1.69 autoantigen), member 16 SLC25A16 -1.69 SAR1a gene homolog 2 (S. cerevisiae) SARA2 -1.68 interleukin 1, beta IL1B -1.68 L-3-hydroxyacyl-Coenzyme A dehydrogenase, short chain HADHSC -1.68 KIAA0626 gene product KIAA0626 -1.68 peroxiredoxin 2 PRDX2 -1.67 cytochrome P450, family 51, subfamily A, polypeptide 1 CYP51A1 dihydrolipoamide branched chain transacylase (E2 component of branched chain keto acid dehydrogenase complex; maple syrup urine -1.67 disease) DBT -1.67 silver homolog (mouse) SILV -1.66 peroxiredoxin 2 /// peroxiredoxin 2 PRDX2 -1.66 transmembrane 4 superfamily member 6 TM4SF6 -1.66 nudix (nucleoside diphosphate linked moiety X)-type motif 4 NUDT4 -1.65 cyclin-dependent kinase 5 CDK5 -1.65 hypothetical protein FLJ20574 FLJ20574 -1.64 N-terminal Asn amidase LOC123803 -1.64 hypothetical protein FLJ13263 FLJ13263 -1.64 transferrin receptor (p90, CD71) TFRC -1.63 peroxisomal biogenesis factor 16 PEX16 -1.63 low density lipoprotein receptor (familial hypercholesterolemia) LDLR -1.63 N-terminal Asn amidase LOC123803 -1.63 neuronal protein NP25 -1.63 DnaJ (Hsp40) homolog, subfamily C, member 3 DNAJC3 -1.62 homogentisate 1,2-dioxygenase (homogentisate oxidase) HGD solute carrier family 7 (cationic amino acid transporter, y+ system), -1.62 member 5 SLC7A5 -1.62 hypothetical protein FLJ22649 similar to signal peptidase SPC22/23 FLJ22649 -1.62 immediate early response 3 IER3 -1.62 carcinoembryonic antigen-related cell adhesion molecule 1 (biliary CEACAM1 171 glycoprotein) -1.62 brain protein 44-like BRP44L -1.61 histone 1, H2ag HIST1H2AG -1.61 Wiskott-Aldrich syndrome-like WASL -1.60 phosphoinositide-3-kinase, class 2, gamma polypeptide PIK3C2G -1.60 chromosome 9 open reading frame 16 C9orf16 -1.60 RNA-binding protein FLJ20273 holocarboxylase synthetase (biotin-[proprionyl-Coenzyme A- -1.60 carboxylase (ATP-hydrolysing)] ligase) HLCS -1.60 24-dehydrocholesterol reductase DHCR24 ems1 sequence (mammary tumor and squamous cell carcinoma- -1.60 associated (p80/85 src substrate) EMS1 acetyl-Coenzyme A acyltransferase 1 (peroxisomal 3-oxoacyl- -1.59 Coenzyme A thiolase) ACAA1 -1.59 CGI-111 protein CGI-111 Homo sapiens transcribed sequence: basic leucine-zipper protein -1.58 BZAP45; --- -1.58 peptidylprolyl isomerase F (cyclophilin F) PPIF -1.58 FBJ murine osteosarcoma viral oncogene homolog B FOSB -1.58 cytidine deaminase CDA -1.58 aldo-keto reductase family 1, member A1 (aldehyde reductase) AKR1A1 -1.58 solute carrier family 16 (monocarboxylic acid transporters), member 7 SLC16A7 -1.57 transmembrane 4 superfamily member 6 TM4SF6 -1.56 progesterone receptor membrane component 1 PGRMC1 -1.56 inositol polyphosphate-4-phosphatase, type II, 105kDa INPP4B -1.56 3-hydroxyisobutyryl-Coenzyme A hydrolase HIBCH Homo sapiens RAB15, member RAS onocogene family, mRNA (cDNA -1.56 clone IMAGE:4866926), with apparent retained intron --- -1.56 peroxisomal biogenesis factor 16 PEX16 -1.55 L-3-hydroxyacyl-Coenzyme A dehydrogenase, short chain HADHSC carcinoembryonic antigen-related cell adhesion molecule 1 (biliary -1.54 glycoprotein) CEACAM1 -1.54 AAIR8193 UNQ8193 -1.54 geranylgeranyl diphosphate synthase 1 GGPS1 -1.54 hypothetical protein HSPC111 HSPC111 -1.54 stromal cell protein LOC55974 elongation of very long chain fatty acids (FEN1/Elo2, SUR4/Elo3, -1.54 yeast)-like 1 ELOVL1 -1.54 methylcrotonoyl-Coenzyme A carboxylase 1 (alpha) MCCC1 -1.53 DKFZP564B167 protein DKFZP564B167 -1.53 multiple coagulation factor deficiency 2 MCFD2 -1.52 transferrin receptor (p90, CD71) TFRC -1.52 etoposide induced 2.4 mRNA EI24 -1.52 chromosome 20 open reading frame 24 C20orf24 -1.52 guanine nucleotide binding protein (G protein), gamma 4 GNG4 protein phosphatase 1B (formerly 2C), magnesium-dependent, beta -1.52 isoform PPM1B -1.52 Pirin PIR -1.52 ------1.51 melan-A MLANA -1.51 peptidylprolyl isomerase F (cyclophilin F) PPIF -1.51 keratin 16 (focal non-epidermolytic palmoplantar keratoderma) KRT16 172 -1.51 G antigen 5 GAGE5 -1.51 serine hydroxymethyltransferase 1 (soluble) SHMT1 -1.51 hypothetical protein FLJ20152 FLJ20152 -1.50 CGI-51 protein CGI-51 -1.50 chromosome 21 open reading frame 33 C21orf33 -1.50 ribonuclease P1 RNASEP1 -1.50 progestin and adipoQ receptor family member III PAQR3 -1.50 translational inhibitor protein p14.5 UK114 -1.49 histone 2, H2aa HIST2H2AA -1.49 hypothetical protein BC016005 LOC129642 -1.49 putative protein similar to nessy (Drosophila) C3F -1.49 fatty-acid-Coenzyme A ligase, long-chain 3 FACL3 -1.49 histone 3, H2a HIST3H2A -1.49 RAB27A, member RAS oncogene family RAB27A -1.48 solute carrier organic anion transporter family, member 4C1 SLCO4C1 solute carrier family 25 (mitochondrial carrier; oxoglutarate carrier), -1.48 member 11 SLC25A11 -1.48 biotinidase BTD -1.48 acyl-Coenzyme A dehydrogenase, very long chain ACADVL -1.48 cystatin B (stefin B) CSTB -1.47 CGI-100 protein CGI-100 -1.47 G protein-coupled receptor 143 GPR143 -1.47 DNA segment, Chr 15, Wayne State University 75, expressed D15Wsu75e -1.47 interleukin 13 receptor, alpha 1 IL13RA1 mitochondrial carrier homolog 2 (C. elegans) /// mitochondrial carrier -1.47 homolog 2 (C. elegans) MTCH2 -1.47 uncharacterized hypothalamus protein HT009 HT009 -1.47 c-Mpl binding protein LOC113251 -1.47 dihydrofolate reductase DHFR -1.47 RAR-related orphan receptor A RORA NADH dehydrogenase (ubiquinone) Fe-S protein 1, 75kDa (NADH- -1.47 coenzyme Q reductase) NDUFS1 -1.47 platelet-activating factor acetylhydrolase 2, 40kDa PAFAH2 -1.47 pyridoxal (pyridoxine, vitamin B6) kinase PDXK -1.47 ubiquitin carrier protein E2-EPF -1.46 hydroxysteroid (17-beta) dehydrogenase 7 HSD17B7 -1.46 v-Ki-ras2 Kirsten rat sarcoma 2 viral oncogene homolog KRAS2 -1.46 low molecular mass ubiquinone-binding protein (9.5kD) QP-C -1.46 pre-B-cell colony-enhancing factor PBEF -1.46 DKFZP566O084 protein DKFZp566O084 -1.46 hypothetical protein FLJ10525 FLJ10525 -1.46 fumarate hydratase FH -1.46 ring finger protein 128 RNF128 -1.45 biotinidase BTD -1.45 histone 2, H2aa HIST2H2AA -1.45 histone 1, H2bg HIST1H2BG -1.45 sushi-repeat protein SRPUL -1.45 nudix (nucleoside diphosphate linked moiety X)-type motif 4 NUDT4 -1.45 mitochondrial ribosomal protein S16 MRPS16 -1.45 Nijmegen breakage syndrome 1 (nibrin) NBS1 -1.45 pyruvate dehydrogenase complex, component X PDHX -1.45 CGI-65 protein CIA30 173 -1.44 ethanolamine kinase EKI1 -1.44 histone 1, H2bh HIST1H2BH sialyltransferase 7D ((alpha-N-acetylneuraminyl-2,3-beta-galactosyl- -1.44 1,3)-N-acetyl galactosaminide alpha-2,6-sialyltransferase) SIAT7D -1.44 tumor protein D52-like 1 TPD52L1 -1.44 ------1.44 GrpE-like 1, mitochondrial (E. coli) GRPEL1 -1.44 RNA terminal phosphate cyclase domain 1 RTCD1 -1.44 protease, serine, 8 (prostasin) PRSS8 -1.44 CGI-90 protein CGI-90 3-hydroxybutyrate dehydrogenase (heart, mitochondrial) /// 3- -1.44 hydroxybutyrate dehydrogenase (heart, mitochondrial) BDH -1.44 family with sequence similarity 3, member C FAM3C -1.44 glycine cleavage system protein H (aminomethyl carrier) GCSH ATP synthase, H+ transporting, mitochondrial F0 complex, subunit c -1.44 (subunit 9) isoform 3 ATP5G3 -1.44 membrane-associated nucleic acid binding protein MNAB -1.44 platelet-activating factor acetylhydrolase 2, 40kDa PAFAH2 -1.43 hypothetical protein FLJ11011 FLJ11011 -1.43 solute carrier family 35 (UDP-galactose transporter), member A2 SLC35A2 elongation of very long chain fatty acids (FEN1/Elo2, SUR4/Elo3, -1.43 yeast)-like 1 ELOVL1 -1.43 ------1.43 peroxisomal biogenesis factor 7 PEX7 -1.43 Nit protein 2 NIT2 -1.43 gamma-glutamyl carboxylase GGCX -1.43 activated leukocyte cell adhesion molecule ALCAM -1.43 FLJ20202 protein FLJ20202 -1.43 NAD(P)H dehydrogenase, quinone 2 NQO2 -1.43 geranylgeranyl diphosphate synthase 1 GGPS1 -1.42 chloride channel 3 CLCN3 -1.42 lactamase, beta 2 CGI-83 -1.42 signal-transducing adaptor protein-2 STAP2 excision repair cross-complementing rodent repair deficiency, -1.42 complementation group 1 (includes overlapping antisense sequence) ERCC1 Sjogren syndrome antigen A2 (60kDa, ribonucleoprotein autoantigen -1.42 SS-A/Ro) SSA2 -1.41 KIAA0186 gene product KIAA0186 -1.41 uracil-DNA glycosylase UNG -1.41 histone 1, H2bd HIST1H2BD -1.41 cytosolic nonspecific dipeptidase (EC 3.4.13.18) CN2 -1.41 sortilin 1 SORT1 -1.41 MRS2-like, magnesium homeostasis factor (S. cerevisiae) MRS2L -1.41 pyruvate dehydrogenase (lipoamide) beta PDHB -1.41 pre-B-cell colony-enhancing factor PBEF -1.41 v-raf murine sarcoma viral oncogene homolog B1 BRAF -1.40 Nijmegen breakage syndrome 1 (nibrin) NBS1 -1.40 hypothetical protein FLJ10849 FLJ10849 -1.40 methylcrotonoyl-Coenzyme A carboxylase 2 (beta) MCCC2 -1.40 ubiquitin specific protease 2 USP2 -1.40 tumor rejection antigen (gp96) 1 TRA1 -1.40 RAB27A, member RAS oncogene family RAB27A 174 -1.40 L-3-hydroxyacyl-Coenzyme A dehydrogenase, short chain HADHSC -1.40 thioesterase superfamily member 2 THEM2 -1.40 chromosome 14 open reading frame 1 C14orf1 -1.40 KIAA0626 gene product KIAA0626 -1.40 mediator of RNA polymerase II transcription, subunit 8 homolog (yeast) MED8 -1.40 chromosome 1 open reading frame 27 C1orf27 -1.40 DnaJ (Hsp40) homolog, subfamily D, member 1 DNAJD1 -1.40 cleavage and polyadenylation specific factor 5, 25 kDa CPSF5 -1.40 glycoprotein, synaptic 2 GPSN2 -1.40 riboflavin kinase FLJ11149 -1.40 histone 1, H2be HIST1H2BE -1.39 dual specificity phosphatase 4 DUSP4 -1.39 stromal cell-derived factor 2-like 1 SDF2L1 -1.39 KIAA0033 protein KIAA0033 -1.39 enoyl Coenzyme A hydratase 1, peroxisomal ECH1 -1.39 programmed cell death 8 (apoptosis-inducing factor) PDCD8 -1.39 hypothetical protein FLJ22353 FLJ22353 -1.39 proteasome (prosome, macropain) subunit, alpha type, 5 PSMA5 -1.39 biliverdin reductase B (flavin reductase (NADPH)) BLVRB -1.39 chromosome 14 open reading frame 87 C14orf87 -1.39 phosphoglucomutase 1 PGM1 -1.38 secretory carrier membrane protein 1 SCAMP1 -1.38 CGI-04 protein CGI-04 -1.38 testis enhanced gene transcript (BAX inhibitor 1) TEGT -1.38 interferon-related developmental regulator 1 IFRD1 -1.38 uncharacterized hematopoietic stem/progenitor cells protein MDS031 MDS031 -1.38 ribokinase RBSK -1.38 tumor protein D52-like 1 TPD52L1 -1.37 dystrophin related protein 2 DRP2 -1.37 pyruvate dehydrogenase (lipoamide) beta PDHB -1.37 hypothetical protein dJ473B4 DJ473B4 -1.37 myosin, light polypeptide 4, alkali; atrial, embryonic MYL4 dihydrolipoamide dehydrogenase (E3 component of pyruvate dehydrogenase complex, 2-oxo-glutarate complex, branched chain -1.37 keto acid dehydrogenase complex) DLD -1.37 solute carrier family 39 (zinc transporter), member 8 SLC39A8 -1.37 basigin (OK blood group) BSG -1.37 interleukin 1 receptor, type II IL1R2 -1.36 nasopharyngeal epithelium specific protein 1 NESG1 -1.36 ------1.36 HUS1 checkpoint homolog (S. pombe) HUS1 -1.36 ------1.36 biotinidase BTD -1.36 PRKC, apoptosis, WT1, regulator PAWR -1.36 RAS protein activator like 1 (GAP1 like) RASAL1 -1.36 heat shock 70kDa protein 9B (mortalin-2) HSPA9B -1.36 ubiquinol-cytochrome c reductase core protein I UQCRC1 -1.36 chromosome 9 open reading frame 16 C9orf16 -1.35 hydroxysteroid (17-beta) dehydrogenase 12 HSD17B12 -1.35 solute carrier family 39 (zinc transporter), member 8 SLC39A8 UDP glycosyltransferase 2 family, polypeptide B28 /// UDP -1.35 glycosyltransferase 2 family, polypeptide B28 UGT2B28 175 -1.35 NADH dehydrogenase (ubiquinone) flavoprotein 2, 24kDa NDUFV2 -1.35 ethanolamine kinase EKI1 -1.35 cyclin-dependent kinase inhibitor 1A (p21, Cip1) CDKN1A -1.35 carboxypeptidase D CPD -1.35 transaldolase 1 TALDO1 -1.35 interleukin 1 receptor, type II IL1R2 -1.35 malonyl-CoA:acyl carrier protein transacylase (malonyltransferase) MT -1.35 v-maf musculoaponeurotic fibrosarcoma oncogene homolog F (avian) MAFF -1.35 cytochrome c oxidase subunit Vb COX5B -1.35 carboxypeptidase D CPD solute carrier family 25 (mitochondrial carrier; ornithine transporter) -1.35 member 15 SLC25A15 -1.35 putative L-type neutral amino acid transporter KIAA0436 ATPase, H+ transporting, lysosomal 21kDa, V0 subunit c'' /// ATPase, -1.35 H+ transporting, lysosomal 21kDa, V0 subunit c'' ATP6V0B solute carrier family 11 (proton-coupled divalent metal ion transporters), -1.35 member 2 SLC11A2 -1.34 hypothetical protein DKFZp434G0522 DKFZp434G0522 -1.34 CDK5 regulatory subunit associated protein 1 CDK5RAP1 -1.34 solute carrier family 35, member B1 SLC35B1 -1.34 peroxisomal farnesylated protein PXF -1.34 fumarate hydratase FH -1.34 proteasome (prosome, macropain) subunit, alpha type, 4 PSMA4 NADH dehydrogenase (ubiquinone) Fe-S protein 8, 23kDa (NADH- -1.34 coenzyme Q reductase) NDUFS8 -1.34 eukaryotic translation initiation factor 2B, subunit 3 gamma, 58kDa EIF2B3 -1.33 ------1.33 proteasome (prosome, macropain) subunit, alpha type, 7 PSMA7 -1.33 associated molecule with the SH3 domain of STAM AMSH -1.33 elongation factor, RNA polymerase II, 2 ELL2 -1.33 NADH dehydrogenase (ubiquinone) 1 beta subcomplex, 3, 12kDa NDUFB3 -1.33 transcription factor Dp-2 (E2F dimerization partner 2) TFDP2 excision repair cross-complementing rodent repair deficiency, -1.33 complementation group 1 (includes overlapping antisense sequence) ERCC1 -1.33 Niemann-Pick disease, type C1 NPC1 -1.33 hypothetical protein MGC2574 MGC2574 -1.33 proteasome (prosome, macropain) subunit, alpha type, 3 PSMA3 -1.33 ORM1-like 2 (S. cerevisiae) ORMDL2 -1.33 GrpE-like 1, mitochondrial (E. coli) GRPEL1 proteasome (prosome, macropain) subunit, beta type, 5 /// proteasome -1.33 (prosome, macropain) subunit, beta type, 5 PSMB5 solute carrier family 11 (proton-coupled divalent metal ion transporters), -1.33 member 2 SLC11A2 -1.33 solute carrier family 22 (organic cation transporter), member 1-like SLC22A1L sorbin and SH3 domain containing 1 /// sorbin and SH3 domain -1.33 containing 1 SORBS1 -1.33 fructose-1,6-bisphosphatase 2 FBP2 NADH dehydrogenase (ubiquinone) 1, subcomplex unknown, 2, -1.32 14.5kDa NDUFC2 -1.32 translocase of inner mitochondrial membrane 13 homolog (yeast) TIMM13 -1.32 cytochrome c oxidase subunit VIIb COX7B -1.32 ------176 likely ortholog of mouse exocyst component protein 70 kDa homolog (S. cerevisiae) Exo70: exocyst component protein 70 kDa homolog (S. -1.32 cerevisiae) EXO70 -1.32 dipeptidylpeptidase 3 DPP3 -1.32 RAB38, member RAS oncogene family RAB38 -1.32 popeye domain containing 2 POPDC2 -1.32 heat shock 70kDa protein 9B (mortalin-2) HSPA9B -1.32 secretory carrier membrane protein 1 SCAMP1 -1.32 glutamate-cysteine ligase, modifier subunit GCLM -1.32 erythroid associated factor ERAF -1.32 hypothetical protein FLJ10375 FLJ10375 -1.31 early growth response 3 EGR3 solute carrier family 5 (sodium-dependent vitamin transporter), member -1.31 6 SLC5A6 -1.31 malignant T cell amplified sequence 1 MCTS1 -1.31 interferon-related developmental regulator 1 IFRD1 -1.31 H2A histone family, member X H2AFX -1.31 DnaJ (Hsp40) homolog, subfamily A, member 3 DNAJA3 -1.31 recombination protein REC14 REC14 glucan (1,4-alpha-), branching enzyme 1 (glycogen branching enzyme, -1.31 Andersen disease, glycogen storage disease type IV) GBE1 -1.31 CGI-51 protein CGI-51 dodecenoyl-Coenzyme A delta isomerase (3,2 trans-enoyl-Coenzyme -1.31 A isomerase) DCI -1.31 glutathione transferase zeta 1 (maleylacetoacetate isomerase) GSTZ1 -1.31 ------1.30 eukaryotic translation initiation factor 4E EIF4E -1.30 growth hormone inducible transmembrane protein GHITM NADH dehydrogenase (ubiquinone) Fe-S protein 7, 20kDa (NADH- -1.30 coenzyme Q reductase) NDUFS7 -1.30 mitochondrial ribosomal protein S33 MRPS33 -1.30 B-cell receptor-associated protein 31 BCAP31 -1.30 chromosome 14 open reading frame 2 C14orf2 -1.30 HSPC171 protein HSPC171 -1.30 D-dopachrome tautomerase DDT -1.30 potassium channel, subfamily K, member 1 KCNK1 -1.29 ubiquitin-conjugating enzyme E2A (RAD6 homolog) UBE2A -1.29 D-aspartate oxidase DDO -1.29 histone 2, H2be HIST2H2BE -1.29 ------1.29 solute carrier family 30 (zinc transporter), member 1 SLC30A1 -1.29 ------1.29 Nijmegen breakage syndrome 1 (nibrin) NBS1 -1.28 Krueppel-related zinc finger protein H-plk -1.28 ClpX caseinolytic protease X homolog (E. coli) CLPX -1.28 FLJ20288 protein FLJ20288 -1.28 hypothetical protein MGC4276 similar to CG8198 MGC4276 -1.28 translocase of inner mitochondrial membrane 23 homolog (yeast) TIMM23 -1.28 c-myc binding protein MYCBP -1.28 trefoil factor 2 (spasmolytic protein 1) TFF2 -1.28 translocase of outer mitochondrial membrane 70 homolog A (yeast) TOMM70A -1.27 cytochrome c oxidase subunit Va COX5A 177 -1.27 mitochondrial ribosomal protein S18A MRPS18A -1.27 polyposis locus protein 1 DP1 Homo sapiens transcribed sequence with strong similarity to protein -1.27 pir:A32800 (H.sapiens) A32800 chaperonin GroEL precursor - human --- -1.27 pantothenate kinase 3 PANK3 -1.26 ------1.26 glutaredoxin 2 GLRX2 ATP synthase, H+ transporting, mitochondrial F0 complex, subunit c -1.26 (subunit 9) isoform 3 ATP5G3 -1.26 unc-5 homolog B (C. elegans) UNC5B -1.26 mitochondrial translation optimization 1 homolog (S. cerevisiae) MTO1 -1.26 ganglioside-induced differentiation-associated protein 1-like 1 GDAP1L1 -1.26 huntingtin interacting protein 2 HIP2 -1.26 integral type I protein P24B -1.26 AUT-like 2, cysteine endopeptidase (S. cerevisiae) AUTL2 -1.26 cytochrome c oxidase subunit Vb COX5B -1.26 T-cell leukemia translocation altered gene TCTA -1.26 mitochondrial ribosomal protein L33 MRPL33 diazepam binding inhibitor (GABA receptor modulator, acyl-Coenzyme -1.26 A binding protein) DBI -1.25 ATPase, H+ transporting, lysosomal 16kDa, V0 subunit c ATP6V0C likely ortholog of rat vacuole membrane protein 1 /// likely ortholog of rat -1.25 vacuole membrane protein 1 VMP1 -1.25 cytochrome c oxidase subunit VIIa polypeptide 2 (liver) COX7A2 -1.25 metaxin 1 MTX1 -1.25 fatty acid binding protein 6, ileal (gastrotropin) FABP6 -1.25 postsynaptic protein CRIPT CRIPT -1.25 aspartylglucosaminidase AGA diazepam binding inhibitor (GABA receptor modulator, acyl-Coenzyme -1.25 A binding protein) DBI -1.25 NADH dehydrogenase (ubiquinone) 1 alpha subcomplex, 3, 9kDa NDUFA3 succinate dehydrogenase complex, subunit C, integral membrane -1.25 protein, 15kDa SDHC Homo sapiens transcribed sequence with weak similarity to protein pir:PC4369 (H.sapiens) PC4369 olfactory receptor, HT2 - human -1.25 (fragment) --- -1.24 6-pyruvoyltetrahydropterin synthase PTS -1.24 succinate-CoA ligase, GDP-forming, alpha subunit SUCLG1 -1.24 adaptor-related protein complex 3, sigma 2 subunit AP3S2 -1.24 cytochrome c oxidase subunit Vb /// cytochrome c oxidase subunit Vb COX5B -1.24 zinc finger protein 430 ZNF430 -1.24 chromosome 6 open reading frame 79 C6orf79 -1.23 recombination protein REC14 REC14 -1.23 ras homolog gene family, member E ARHE -1.23 F-box only protein 9 FBXO9 -1.23 ATPase, H+ transporting, lysosomal 34kDa, V1 subunit D ATP6V1D -1.22 peptidase (mitochondrial processing) beta PMPCB -1.22 DNA segment on chromosome X (unique) 9879 expressed sequence DXS9879E -1.21 CGI-147 protein CGI-147 -1.21 malate dehydrogenase 1, NAD (soluble) MDH1 -1.21 synaptogyrin 1 SYNGR1 -1.21 proteasome (prosome, macropain) subunit, alpha type, 1 PSMA1 178 -1.21 chromosome 14 open reading frame 2 C14orf2 ATP synthase, H+ transporting, mitochondrial F0 complex, subunit f, -1.20 isoform 2 ATP5J2 -1.20 ubiquitin specific protease 3 USP3 -1.20 cytochrome c oxidase subunit VIa polypeptide 1 COX6A1 NADH dehydrogenase (ubiquinone) Fe-S protein 3, 30kDa (NADH- -1.20 coenzyme Q reductase) NDUFS3 -1.20 ubiquinol-cytochrome c reductase core protein II UQCRC2 ATP synthase, H+ transporting, mitochondrial F1 complex, gamma -1.18 polypeptide 1 ATP5C1

A.2 All significantly changed gene in SEB-1 sebocytes after 72 hours 13-cis RA treatment

Some genes may be listed twice; indicates separate probe sets on Affymetrix gene array chips. Fold Gene Gene Title Change Symbol 12.25 retinoic acid receptor responder (tazarotene induced) 1 RARRES1 9.89 retinoic acid receptor responder (tazarotene induced) 1 RARRES1 7.04 lipocalin 2 (oncogene 24p3) LCN2 5.95 tumor necrosis factor, alpha-induced protein 2 TNFAIP2 5.91 hydroxyprostaglandin dehydrogenase 15-(NAD) HPGD carcinoembryonic antigen-related cell adhesion molecule 6 (non-specific 4.98 cross reacting antigen) CEACAM6 4.64 cytochrome P450, family 1, subfamily B, polypeptide 1 CYP1B1 4.30 cytochrome P450, family 1, subfamily B, polypeptide 1 CYP1B1 4.25 hydroxyprostaglandin dehydrogenase 15-(NAD) HPGD 4.18 tumor necrosis factor (ligand) superfamily, member 10 TNFSF10 3.70 serine (or cysteine) proteinase inhibitor, clade B (ovalbumin), member 3 SERPINB3 3.65 homeo box A5 HOXA5 carcinoembryonic antigen-related cell adhesion molecule 1 (biliary 3.52 glycoprotein) CEACAM1 3.43 insulin-like growth factor binding protein 3 IGFBP3 3.29 aldehyde dehydrogenase 1 family, member A3 ALDH1A3 3.22 retinoic acid receptor responder (tazarotene induced) 3 RARRES3 3.08 oxidised low density lipoprotein (lectin-like) receptor 1 OLR1 3.06 solute carrier family 1 (glial high affinity glutamate transporter), member 3 SLC1A3 3.00 growth differentiation factor 15 GDF15 2.98 SRY (sex determining region Y)-box 4 SOX4 2.74 tripartite motif-containing 31 TRIM31 2.60 cyclin-dependent kinase inhibitor 1A (p21, Cip1) CDKN1A 2.51 E74-like factor 3 (ets domain transcription factor, epithelial-specific ) ELF3 2.48 breast carcinoma amplified sequence 1 BCAS1 2.46 tetratricopeptide repeat domain 9 TTC9 2.42 interferon regulatory factor 1 IRF1 179 2.42 interferon-induced protein with tetratricopeptide repeats 3 IFIT3 2.35 tripartite motif-containing 31 TRIM31 2.29 fucosidase, alpha-L- 1, tissue FUCA1 2.23 interferon-induced protein with tetratricopeptide repeats 2 IFIT2 2.20 BTG family, member 2 BTG2 2.20 interferon-induced protein with tetratricopeptide repeats 2 IFIT2 2.18 carcinoembryonic antigen-related cell adhesion molecule 5 CEACAM5 2.09 proteasome (prosome, macropain) subunit, beta type, 10 PSMB10 2.08 vascular cell adhesion molecule 1 VCAM1 2.07 annexin A9 ANXA9 2.07 2',5'-oligoadenylate synthetase 1, 40/46kDa OAS1 2.06 ------1.94 PTK6 protein tyrosine kinase 6 PTK6 1.85 GATA binding protein 3 GATA3 1.82 2',5'-oligoadenylate synthetase 1, 40/46kDa OAS1 1.79 protein kinase C, alpha PRKCA 1.78 EPH receptor B2 EPHB2 1.78 RNA binding protein with multiple splicing RBPMS 1.77 protein kinase D2 PRKD2 1.76 B-cell linker BLNK major histocompatibility complex, class I, B /// major histocompatibility 1.71 complex, class I, C HLA-B 1.71 glycine receptor, beta GLRB 1.71 integrin, beta 6 ITGB6 nuclear factor of kappa light polypeptide gene enhancer in B-cells 2 1.70 (p49/p100) NFKB2 1.69 Fas (TNF receptor superfamily, member 6) FAS ATPase, aminophospholipid transporter (APLT), Class I, type 8A, member 1.66 1 ATP8A1 1.66 dual specificity phosphatase 8 DUSP8 1.62 solute carrier organic anion transporter family, member 3A1 SLCO3A1 1.60 glutathione peroxidase 2 (gastrointestinal) GPX2 1.59 KIBRA protein KIBRA 1.56 sequestosome 1 SQSTM1 1.47 phosphoinositide-3-kinase, regulatory subunit 3 (p55, gamma) PIK3R3 -4.68 keratin 6A /// keratin 6C KRT6A -3.94 FK506 binding protein 5 FKBP5 -3.78 keratin 6A /// keratin 6C KRT6A ELOVL family member 5, elongation of long chain fatty acids (FEN1/Elo2, -3.25 SUR4/Elo3-like, yeast) ELOVL5 -3.12 dihydrofolate reductase DHFR -2.79 pro-melanin-concentrating hormone PMCH -2.72 zinc finger and BTB domain containing 16 ZBTB16 -2.57 pyruvate dehydrogenase kinase, isoenzyme 4 PDK4 -2.48 glutamate dehydrogenase 1 GLUD1 -2.27 ribonucleotide reductase M2 polypeptide RRM2 -2.25 transcription elongation factor A (SII)-like 4 TCEAL4 -2.12 CD86 antigen (CD28 antigen ligand 2, B7-2 antigen) CD86 -2.05 DNA replication complex GINS protein PSF1 PSF1 180 -2.03 protein tyrosine phosphatase, non-receptor type 1 PTPN1 -2.02 ------1.96 coilin COIL -1.92 collagen, type IV, alpha 6 COL4A6 -1.91 gamma-aminobutyric acid (GABA) A receptor, alpha 2 GABRA2 -1.89 ------S100 calcium binding protein A10 (annexin II ligand, calpactin I, light -1.81 polypeptide (p11)) S100A10 -1.80 thrombospondin 1 THBS1 -1.79 ubiquinol-cytochrome c reductase hinge protein UQCRH -1.71 phospholipase A2, group IVA (cytosolic, calcium-dependent) PLA2G4A protein tyrosine phosphatase-like (proline instead of catalytic arginine), -1.71 member b PTPLB -1.64 CDC6 cell division cycle 6 homolog (S. cerevisiae) CDC6 -1.59 BUB3 budding uninhibited by benzimidazoles 3 homolog (yeast) BUB3 -1.49 hypothetical protein LOC130074 LOC130074

A.3 All significantly changed genes in HaCaT keratinocytes after 72 hour 13-cis RA treatment

Some genes may be listed twice; indicates separate probe sets on Affymetrix gene array chips. Fold Gene Gene Title Change Symbol 3.68 tripartite motif-containing 31 TRIM31 3.56 lipocalin 2 (oncogene 24p3) LCN2 3.22 carcinoembryonic antigen-related cell adhesion molecule 5 CEACAM5 3.21 amiloride binding protein 1 (amine oxidase (copper-containing)) ABP1 3.21 cytochrome P450 retinoid metabolizing protein P450RAI-2 2.77 carcinoembryonic antigen-related cell adhesion molecule 6 CEACAM6 2.69 phospholipase A2, group X PLA2G10 2.61 tripartite motif-containing 31 TRIM31 2.56 tripartite motif-containing 31 TRIM31 2.48 carcinoembryonic antigen-related cell adhesion molecule 6 CEACAM6 2.33 fibulin 1 FBLN1 2.28 latexin protein LXN 2.21 mucin 4, tracheobronchial MUC4 2.19 chromosome 11 open reading frame 8 C11orf8 2.11 kallikrein 6 (neurosin, zyme) KLK6 2.08 3'-phosphoadenosine 5'-phosphosulfate synthase 2 PAPSS2 1.96 plasminogen activator, tissue PLAT 1.91 chromosome 11 open reading frame 8 C11orf8 1.90 SLAM family member 7 SLAMF7 181 1.87 sushi-repeat protein SRPUL 1.82 2',5'-oligoadenylate synthetase 1, 40/46kDa OAS1 1.81 GATA binding protein 3 GATA3 1.80 neural precursor cell expressed, developmentally down-regulated 9 NEDD9 1.79 insulin-like growth factor binding protein 3 IGFBP3 1.79 fibulin 1 FBLN1 1.76 nebulette NEBL 1.70 sarcospan (Kras oncogene-associated gene) SSPN 1.70 G protein-coupled receptor, family C, group 5, member B GPRC5B carcinoembryonic antigen-related cell adhesion molecule 1 (biliary 1.70 glycoprotein) CEACAM1 1.69 S100 calcium binding protein P S100P prostaglandin I2 (prostacyclin) synthase /// prostaglandin I2 (prostacyclin) 1.67 synthase PTGIS carcinoembryonic antigen-related cell adhesion molecule 1 (biliary 1.66 glycoprotein) CEACAM1 1.65 Homo sapiens cDNA clone IMAGE:3865861, partial cds --- 1.64 SKI-like SKIL 1.64 fucosidase, alpha-L- 1, tissue FUCA1 1.64 midkine (neurite growth-promoting factor 2) MDK 1.64 annexin A9 ANXA9 1.63 involucrin IVL 1.62 insulin-like growth factor binding protein 6 IGFBP6 UDP-N-acetyl-alpha-D-galactosamine:polypeptide N- 1.60 acetylgalactosaminyltransferase 12 (GalNAc-T12) GALNT12 1.57 ------1.54 phosphatidic acid phosphatase type 2A PPAP2A 1.54 basic helix-loop-helix domain containing, class B, 3 BHLHB3 1.51 insulin-like growth factor binding protein 3 IGFBP3 1.49 bone morphogenetic protein 3 (osteogenic) BMP3 1.48 lysophosphatidic acid phosphatase ACP6 myxovirus (influenza virus) resistance 1, interferon-inducible protein p78 1.48 (mouse) MX1 1.47 integrin, beta-like 1 (with EGF-like repeat domains) ITGBL1 1.45 cathepsin H CTSH 1.39 vacuolar protein sorting 28 (yeast) VPS28 1.35 FLJ21963 protein FLJ21963 1.34 retinoic acid induced 3 RAI3 1.30 2'-5'-oligoadenylate synthetase-like /// 2'-5'-oligoadenylate synthetase-like OASL -2.10 Microfibril-associated glycoprotein-2 MAGP2

182 References

Abdel-Naser MB: Selective cultivation of normal human sebocytes in vitro; a simple modified technique for a better cell yield. Exp Dermatol 13: 562-6, 2004.

Adamson PC: Pharmacokinetics of all-trans-retinoic acid: clinical implications in acute promyelocytic leukemia. Semin Hematol 31: 14-7, 1994.

Aderem A, Ulevitch RJ: Toll-like receptors in the induction of the innate immune response. Nature 406: 782-7, 2000.

Ahuja D, Saenz-Robles MT, Pipas JM: SV40 large T antigen targets multiple cellular pathways to elicit cellular transformation. Oncogene 24: 7729-45, 2005.

Akerstrom B, Flower DR, Salier JP: Lipocalins: unity in diversity. Biochim Biophys Acta 1482: 1- 8, 2000.

Allen M, Grachtchouk M, Sheng H, Grachtchouk V, Wang A, Wei L, et al: Hedgehog signaling regulates sebaceous gland development. Am J Pathol 163: 2173-2178, 2003.

Allenby G, Bocquel MT, Saunders M, Kazmer S, Speck J, Rosenberger M, et al: Retinoic acid receptors and retinoid X receptors: interactions with endogenous retinoic acids. Proc Natl Acad Sci U S A 90: 30-4, 1993.

Almond-Roesler B, Blume-Peytavi U, Bisson S, Krahn M, Rohloff E, Orfanos CE: Monitoring of isotretinoin therapy by measuring the plasma levels of isotretinoin and 4-oxo-isotretinoin. A useful tool for management of severe acne. Dermatology 196: 176-81, 1998.

Altucci L, Gronemeyer H: Decryption of the retinoid death code in leukemia. J Clin Immunol 22: 117-23, 2002.

Altucci L, Wilhelm E, Gronemeyer H: Leukemia: beneficial actions of retinoids and rexinoids. Int J Biochem Cell Biol 36: 178-82, 2004.

Alves AM, Thody AJ, Fisher C, Shuster S: Measurement of lipogenesis in isolated preputial gland cells of the rat and the effect of oestrogen. Journal of Endocrinology 109: 1-7, 1986.

Anckar J, Sistonen L: Heat shock factor 1 as a coordinator of stress and developmental pathways. Adv Exp Med Biol 594: 78-88, 2007.

Arany I, Whitehead WE, Grattendick KJ, Ember IA, Tyring SK: Suppression of Growth by All- trans Retinoic Acid Requires Prolonged Induction of Interferon Regulatory Factor 1 in Cervical Squamous Carcinoma (SiHa) Cells. Clin. Diagn. Lab. Immunol. 9: 1102-1106, 2002.

Arce F, Gatjens-Boniche O, Vargas E, Valverde B, Diaz C: Apoptotic events induced by naturally occurring retinoids ATRA and 13-cis retinoic acid on human hepatoma cell lines Hep3B and HepG2. Cancer Lett 229: 271-81, 2005.

183 Archer JS, Chang RJ: Hirsutism and acne in polycystic ovary syndrome. Best Practice & Research Clinical Obstetrics & Gynaecology Polycystic Ovary Syndrome: Challenges for the Clinician 18: 737-754, 2004.

Armeanu S, Lauer UM, Smirnow I, Schenk M, Weiss TS, Gregor M, et al: Adenoviral gene transfer of tumor necrosis factor-related apoptosis-inducing ligand overcomes an impaired response of hepatoma cells but causes severe apoptosis in primary human hepatocytes. Cancer Res 63: 2369-72, 2003.

Avantaggiato V, Acampora D, Tuorto F, Simeone A: Retinoic acid induces stage-specific repatterning of the rostral central nervous system. Dev Biol 175: 347-57, 1996.

Bachmann F, Buechner SA, Wernli M, Strebel S, Erb P: Ultraviolet Light Downregulates CD95 Ligand and Trail Receptor Expression Facilitating Actinic Keratosis and Squamous Cell Carcinoma Formation. 117: 59-66, 2001.

Balasubramanian S, Chandraratna RA, Eckert RL: A novel retinoid-related molecule inhibits pancreatic cancer cell proliferation by a retinoid receptor independent mechanism via suppression of cell cycle regulatory protein function and induction of caspase-associated apoptosis. Oncogene 24: 4257-70, 2005.

Baldin V, Lukas J, Marcote MJ, Pagano M, Draetta G: Cyclin D1 is a nuclear protein required for cell cycle progression in G1. Genes Dev 7: 812-21, 1993.

Baldwin HE: The interaction between acne vulgaris and the psyche. Cutis 70: 133-9, 2002.

Bargonetti J, Reynisdottir I, Friedman PN, Prives C: Site-specific binding of wild-type p53 to cellular DNA is inhibited by SV40 T antigen and mutant p53. Genes Dev 6: 1886-98, 1992. Baron JM, Heise R, Blaner WS, Neis M, Joussen S, Dreuw A, et al: Retinoic Acid and its 4-Oxo Metabolites are Functionally Active in Human Skin Cells In Vitro. J Invest Dermatol 125: 143-53, 2005.

Bastien J, Plassat JL, Payrastre B, Rochette-Egly C: The phosphoinositide 3-kinase//Akt pathway is essential for the retinoic acid-induced differentiation of F9 cells. Oncogene 25: 2040- 2047, 2005.

Benjamini Y, Yekutieli D: Quantitative trait Loci analysis using the false discovery rate. Genetics 171: 783-90, 2005.

Berger T, Togawa A, Duncan GS, Elia AJ, You-Ten A, Wakeham A, et al: Lipocalin 2-deficient mice exhibit increased sensitivity to Escherichia coli infection but not to ischemia-reperfusion injury. Proc Natl Acad Sci U S A 103: 1834-9, 2006.

Berthet C, Klarmann KD, Hilton MB, Suh HC, Keller JR, Kiyokawa H, et al: Combined loss of Cdk2 and Cdk4 results in embryonic lethality and Rb hypophosphorylation. Dev Cell 10: 563-73, 2006.

Bickers DR, Lim HW, Margolis D, Weinstock MA, Goodman C, Faulkner E, et al: The burden of skin diseases: 2004: A joint project of the American Academy of Dermatology Association and 184 the Society for Investigative Dermatology. Journal of the American Academy of Dermatology 55: 490-500, 2006.

Blauer M, Vaalasti A, Pauli SL, Ylikomi T, Joensuu T, Tuohimaa P: Location of Androgen Receptor in Human Skin. J Invest Dermatol 97:264-268. 97: pp.264-268, 1991.

Blumenberg M, Tomic-Canic M: Human epidermal keratinocyte: keratinization processes. Exs 78: 1-29, 1997.

Boehm MF, Zhang L, Zhi L, McClurg MR, Berger E, Wagoner M, et al: Design and synthesis of potent retinoid X receptor selective ligands that induce apoptosis in leukemia cells. J Med Chem 38: 3146-55, 1995.

Boehm N, Samama B, Cribier B, Rochette-Egly C: Retinoic-acid receptor beta expression in melanocytes. Eur J Dermatol 14: 19-23, 2004.

Bouterfa H, Picht T, Kess D, Herbold C, Noll E, Black PM, et al: Retinoids inhibit human glioma cell proliferation and migration in primary cell cultures but not in established cell lines. Neurosurgery 46: 419-30, 2000.

Brand N, Petkovich M, Krust A, Chambon P, de The H, Marchio A, et al: Identification of a second human retinoic acid receptor. Nature 332: 850-3, 1988.

Brogden RN, Clissold SP: Flutamide. A preliminary review of its pharmacodynamic and pharmacokinetic properties, and therapeutic efficacy in advanced prostatic cancer. Drugs 38: 185-203, 1989.

Caramuta S, De Cecco L, Reid JF, Zannini L, Gariboldi M, Kjeldsen L, et al: Regulation of lipocalin-2 gene by the cancer chemopreventive retinoid 4-HPR. Int J Cancer 119: 1599-606, 2006.

Cariati R, Zancai P, Quaia M, Cutrona G, Giannini F, Rizzo S, et al: Retinoic acid induces persistent, RARalpha-mediated anti-proliferative responses in Epstein-Barr virus-immortalized b lymphoblasts carrying an activated C-MYC oncogene but not in Burkitt's lymphoma cell lines. Int J Cancer 86: 375-84, 2000.

Chakravarti N, El-Naggar AK, Lotan R, Anderson J, Diwan AH, Saadati HG, et al: Expression of retinoid receptors in sebaceous cell carcinoma. Journal of Cutaneous Pathology 33: 10-17, 2006.

Chambon P: The retinoid signaling pathway: molecular and genetic analyses. Semin Cell Biol 5: 115-25, 1994.

Chen J, Maltby KM, Miano JM: A Novel Retinoid-Response Gene Set in Vascular Smooth Muscle Cells. Biochemical and Biophysical Research Communications 281: 475-482, 2001.

Chen W, Kelly M, Opitz-Araya X, Thomas R, Low M, Cone R: Exocrine gland dysfunction in MC5-R deficient mice: evidence for coordinated regulation of exocrine gland function by melanocortin peptides. Cell 91: 789-798, 1997.

185 Chen W, Ohara N, Wang J, Xu Q, Liu J, Morikawa A, et al: A Novel Selective Progesterone Receptor Modulator Asoprisnil (J867) Inhibits Proliferation and Induces Apoptosis in Cultured Human Uterine Leiomyoma Cells in the Absence of Comparable Effects on Myometrial Cells. J Clin Endocrinol Metab 91: 1296-1304, 2006.

Chen W, Thiboutot D, Zouboulis C: Cutaneous androgen metabolism: basic research and clinical perspectives. J Invest Dermatol 119: 992-1007, 2002.

Chen W, Yang C, Sheu H, Seltmann H, Zouboulis C: Expression of peroxisome proliferator- activated receptor and CCAAT/enhancer binding protein transcription factors in cultured human sebocytes. J Invest Dermatol 121: 441-447, 2003.

Chiba H, Clifford J, Metzger D, Chambon P: Specific and Redundant Functions of Retinoid X Receptor/Retinoic Acid Receptor Heterodimers in Differentiation, Proliferation, and Apoptosis of F9 Embryonal Carcinoma Cells. J. Cell Biol. 139: 735-747, 1997.

Chun KH, Pfahl M, Lotan R: Induction of apoptosis by the synthetic retinoid MX3350-1 through extrinsic and intrinsic pathways in head and neck squamous carcinoma cells. Oncogene 24: 3669-77, 2005.

Chuong CM, Nickoloff BJ, Elias PM, Goldsmith LA, Macher E, Maderson PA, et al: What is the 'true' function of skin? Exp Dermatol 11: 159-87, 2002.

Clarke N, Jimenez-Lara AM, Voltz E, Gronemeyer H: Tumor suppressor IRF-1 mediates retinoid and interferon anticancer signaling to death ligand TRAIL. Embo J 23: 3051-60, 2004.

Coates P, Vyakrnam S, Ravenscroft JC, Stables GI, Cunliffe WJ, Leyden JJ, et al: Efficacy of oral isotretinoin in the control of skin and nasal colonization by antibiotic-resistant propionibacteria in patients with acne. Br J Dermatol 153: 1126-36, 2005.

Cooper GM. Oncogenes. Boston: Jones and Bartlett, 1995.

Cooper MF, McGrath H, Shuster S: Sebaceous lipogenesis in human skin. Variability with age and with severity of acne. Br J Dermatol 94: 165-72, 1976.

Cordrain L, Lindeberg S, Hurtado M, Hill K, Eaton S, Brand-Miller J: Acne vulgaris: a disease of western civilization. Arch Dermatol 138: 1584-1590, 2002.

Cotterill J, Cunliffe W, Williamson B: Severity of acne and sebum excretion rate. British Journal of Dermatology 85: 93-94 1971.

Crandall J, Sakai Y, Zhang J, Koul O, Mineur Y, Crusio WE, et al: 13-cis-retinoic acid suppresses hippocampal cell division and hippocampal-dependent learning in mice. Proc Natl Acad Sci U S A 101: 5111-6, 2004.

Cunliffe WJ: Acne and unemployment. Br J Dermatol 115: 386, 1986.

Cunliffe WJ, Shuster S: Pathogenesis of acne. Lancet 1: 685-7, 1969.

186 Dalziel K, Barton S, Marks R: The effects of isotretinoin on follicular and sebaceous gland differentiation. Br J Dermatol 117: 317-23, 1987.

Daniels RA, Turley H, Kimberley FC, Liu XS, Mongkolsapaya J, Ch'En P, et al: Expression of TRAIL and TRAIL receptors in normal and malignant tissues. Cell Res 15: 430-438, 2005.

Davies M, Marks R: Studies on the effect of salicylic acid on normal skin. Br J Dermatol 95: 187- 92, 1976.

Davies S, Dai D, Wolf DM, Leslie KK: Immunomodulatory and transcriptional effects of progesterone through progesterone A and B receptors in Hec50co poorly differentiated endometrial cancer cells. J Soc Gynecol Investig 11: 494-9, 2004.

Dawson MI, Hobbs PD. The Synthetic Chemistry of Retinoids. In: The Retinoids: Biology, Chemistry and Medicine (Second ed.), edited by M. B. Sporn, A. B. Roberts and D. S. Goodman. New York, New York: Raven Press, Ltd., 1994, p. 5-178.

De Luca LM, Darwiche N, JOnes CS, Scita G: Retinoids in differentiation and neoplasia. Scient Am Sci Med 2: 28-37 1995.

Degli-Esposti MA, Dougall WC, Smolak PJ, Waugh JY, Smith CA, Goodwin RG: The novel receptor TRAIL-R4 induces NF-kappaB and protects against TRAIL-mediated apoptosis, yet retains an incomplete death domain. Immunity 7: 813-20, 1997.

Dermatology AAo. Myths about Acne and FAQs. Skin Care Physicians, 2006.

Dermatology AAo. The Social Impact of Acne. In: AcneNet American Academy of Dermatology, 2006.

Devireddy LR, Gazin C, Zhu X, Green MR: A cell-surface receptor for lipocalin 24p3 selectively mediates apoptosis and iron uptake. Cell 123: 1293-305, 2005.

Dillon CP, Sandy P, Nencioni A, Kissler S, Rubinson DA, Van Parijs L: Rnai as an experimental and therapeutic tool to study and regulate physiological and disease processes. Annu Rev Physiol 67: 147-73, 2005.

Dimberg A, Oberg F: Retinoic acid-induced cell cycle arrest of human myeloid cell lines. Leuk Lymphoma 44: 1641-50, 2003.

Dimri GP, Lee X, Basile G, Acosta M, Scott G, Roskelley C, et al: A Biomarker that Identifies Senescent Human Cells in Culture and in Aging Skin in vivo. PNAS 92: 9363-9367, 1995.

Doran TI, Baff R, Jacobs P, Pacia E: Characterization of Human Sebaceous Cells in vitro. J Invest Dermatol 1991 96: pp.341-348, 1991.

Doran TI, Vidrich A, Sun TT: Intrinsic and extrinsic regulation of the differentiation of skin, corneal and esophageal epithelial cells. Cell 22: 17-25, 1980.

187 Downie MM, Sanders D, Maier L, Stock D, Kealey T: Peroxisome proliferator-activated receptor and farnesoid X receptor ligands differentially regulate sebaceous differentiation in human sebaeous gland organ cultures in vitro. Br J Dermatol 151: 766-775, 2004.

Downie MMT, Guy R, Kealey T: Advances in sebaceous gland research: potential new approaches to acne management. International Journal of Comestic Science 26: 291-311, 2004.

Dreher A, Grevers G: [Fordyce spots. A little regarded finding in the area of lip pigmentation and mouth mucosa]. Laryngorhinootologie 74: 390-2, 1995.

Ebling FJ, Ebling E, McCaffery V, Skinner J: The response of the sebaceous glands of the hypophysectomized-castrated male rat to 5 -dihydrotestosterone, androstenedione, dehydroepiandrosterone and androsterone. J Endocrinol 51: 181-90, 1971.

Ebling FJ, Ebling E, Skinner J: The influence of pituitary hormones on the response of the sebaceous glands of the male rat to testosterone. J Endocrinol 45: 245-56, 1969.

Eckert RL, Broome AM, Ruse M, Robinson N, Ryan D, Lee K: S100 proteins in the epidermis. J Invest Dermatol 123: 23-33, 2004.

Eckert RL, Crish JF, Efimova T, Dashti SR, Deucher A, Bone F, et al: Regulation of involucrin gene expression. J Invest Dermatol 123: 13-22, 2004.

Eckert RL, Green H: Structure and evolution of the human involucrin gene. Cell 46: 583-589, 1986.

Edmondson SR, Thumiger SP, Kaur P, Loh B, Koelmeyer R, Li A, et al: Insulin-like growth factor binding protein-3 (IGFBP-3) localizes to and modulates proliferative epidermal keratinocytes in vivo. British Journal of Dermatology 152: 225-230, 2005.

Elder JT, Astrom A, Pettersson U, Tavakkol A, Krust A, Kastner P, et al: Retinoic acid receptors and binding proteins in human skin. J Invest Dermatol 98: 36S-41S, 1992.

Faucheux C, Ulysse F, Bareille R, Reddi AH, Amedee J: Opposing Actions of BMP3 and TGF[beta]1 in Human Bone Marrow Stromal Cell Growth and Differentiation. Biochemical and Biophysical Research Communications 241: 787-793, 1997.

Ferrari RA, Chakrabarty K, Beyler AL, Wiland J: Suppression of sebaceous gland development in laboratory animals by 17alpha-propyltestosterone. J Invest Dermatol 71: 320-3, 1978.

Fisher GJ, Voorhees JJ: Molecular mechanisms of retinoid actions in skin. Faseb J 10: 1002- 13, 1996.

Flo TH, Smith KD, Sato S, Rodriguez DJ, Holmes MA, Strong RK, et al: Lipocalin 2 mediates an innate immune response to bacterial infection by sequestrating iron. Nature 432: 917-21, 2004.

Flurh JW, Mao-Qiang M, Brown BE, Wertz PW, Crumrine D, Sundberg JP, et al: Glycerol regulates stratum corneum hydration in sebaceous gland deficient (Asebia) mice. Journal of Investigative Dermatology 120: 728-737, 2003. 188 Fogh K., Voorhees J. J., Astrom A.: Expression, Purification, and Binding Properties of Human Cellular Retinoic Acid-Binding Protein Type I and Type II. Archives of Biochemistry and Biophysics 300: 751-755, 1993.

Franz TJ, Lehman PA, Pochi P, Odland GF, Olerud J: The hamster flank organ model: is it relevant to man? J Invest Dermatol 93: 475-9, 1989.

Fritsch M, Orfanos C, Zouboulis C: Sebocytes are the key regulators of androgen homeostasis in human skin. J Invest Dermatol 116: 793-800, 2001.

Fujimura S, Suzumiya J, Yamada Y, Kuroki M, Ono J: Downregulation of Bcl-xL and activation of caspases during retinoic acid-induced apoptosis in an adult T-cell leukemia cell line. Hematol J 4: 328-35, 2003.

Ge L, Gordon J, Hsuan C, Stenn K, Prouty S: Identification of the delta-6 desaturase of human sebaceous glands: expression and enzyme activity. J Invest Dermatol 120: 707-714, 2003.

Gebhart W, Metze D, Jurecka W: Identification of secretory immunoglobulin A in human sweat and sweat glands. J Invest Dermatol 92: 648, 1989.

Geiger J-M, Hommel L, Harms M, Suarat J-H: Oral 13-cis Retinoic Acid is superior to 9-cis retinoic acid in sebosuppression in human beings. Journal of the American Academy of Dermatology 34: 513-515, 1996.

Geiger J: Retinoids and sebaceous gland activity. Dermatol 1995, 1995.

Giannini F, Maestro R, Vukosavljevic T, Pomponi F, Boiocchi M: All-trans, 13-cis and 9-cis retinoic acids induce a fully reversible growth inhibition in HNSCC cell lines: implications for in vivo retinoic acid use. Int J Cancer 70: 194-200, 1997.

Giguere V, Ong ES, Segui P, Evans RM: Identification of a receptor for the morphogen retinoic acid. Nature 330: 624-9, 1987.

Gilaberte M, Puig L, Alomar A: Isotretinoin treatment of acne in a patient with Apert Syndrome. Pediatric Dermatology 20: 443-446, 2003.

Glover JC, Renaud JS, Rijli FM: Retinoic acid and hindbrain patterning. J Neurobiol 66: 705-25, 2006.

Goldstein JA, Comite H, Mescon H, Pochi PE: Isotretinoin in the treatment of Acne. Archives of Dermatology 118: 555-558, 1982.

Goldstein JA, Socha-Szott A, Thomsen RJ, Pochi PE, Shalita AR, Strauss JS: Comparative effect of isotretinoin and etretinate on acne and sebaceous gland secretion. J Am Acad Dermatol 6: 760-5, 1982.

Golias CH, Charalabopoulos A, Charalabopoulos K: Cell proliferation and cell cycle control: a mini review. International Journal of Clinical Practice 58: 1134-1141, 2004.

189 Gomez EC: Differential effect of 13-cis-retinoic acid and an aromatic retinoid (Ro 10-9359) on the sebaceous glands of the hamster flank organ. J Invest Dermatol 76: 68-9, 1981.

Gomez EC, Moskowitz RJ: Effect of 13-cis-retinoic acid on the hamster flank organ. J Invest Dermatol 74: 392-7, 1980.

Graham GM, Farrar MD, Cruse-Sawyer JE, Holland KT, Ingham E: Proinflammatory cytokine production by human keratinocytes stimulated with Propionibacterium acnes and P. acnes GroEL. Br J Dermatol 150: 421-8, 2004.

Green DR, Kroemer G: Pharmacological manipulation of cell death: clinical applications in sight? J Clin Invest 115: 2610-7, 2005.

Gribbon EM, Cunliffe WJ, Holland KT: Interaction of Propionibacterium acnes with skin lipids in vitro. J Gen Microbiol 139: 1745-51, 1993.

Guy R, Ridden C, Kealey T: The improved organ maintenance of the human sebaceous gland: modeling in vitro the effects of epidermal growth factor, androgens, estrogens, 13-cis retinoic acid, and phenol red. J Invest Dermatol 106: 454-60, 1996.

Hallahan AR, Pritchard JI, Chandraratna RA, Ellenbogen RG, Geyer JR, Overland RP, et al: BMP-2 mediates retinoid-induced apoptosis in medulloblastoma cells through a paracrine effect. Nat Med 9: 1033-8, 2003.

Hanai J, Mammoto T, Seth P, Mori K, Karumanchi SA, Barasch J, et al: Lipocalin 2 diminishes invasiveness and metastasis of Ras-transformed cells. J Biol Chem 280: 13641-7, 2005.

Handa H, Hegde UP, Kotelnikov VM, Mundle SD, Dong LM, Burke P, et al: The effects of 13-cis retinoic acid and interferon-alpha in chronic myelogenous leukemia cells in vivo in patients. Leuk Res 21: 1087-96, 1997.

Harper JC: Antiandrogen therapy for skin and hair disease. Dermatol Clin 24: 137-43, v, 2006.

Harris HH: Sustainable rates of sebum secretion in acne patients and matched normal control subjects. J Am Acad Dermatol 8: 200, 1983.

Hay JB, Hodgins MB, Donnelly JB: Human skin androgen metabolism and preliminary evidence for its control by two forms of 17 ß-hydroxysteroid oxidoreductase. J Endocrinology 93: 403-413, 1982.

Herbeuval JP, Hardy AW, Boasso A, Anderson SA, Dolan MJ, Dy M, et al: Regulation of TNF- related apoptosis-inducing ligand on primary CD4+ T cells by HIV-1: role of type I IFN-producing plasmacytoid dendritic cells. Proc Natl Acad Sci U S A 102: 13974-9, 2005.

Hesry Vincent P-: Sensitivity of prostate cells to TRAIL-induced apoptosis increases with tumor progression: DR5 and caspase 8 are key players. The Prostate 66: 987-995, 2006.

Hewett D, Samuelsson L, Polding J, Enlund F, Smart D, Cantone K, et al: Identification of a psoriasis susceptibility candidate gene by linkage disequilibrium mapping with a localized single nucleotide polymorphism map. Genomics 79: 305-14, 2002. 190

Higaki S, Nakamura M, Morohashi M, Yamagishi T: Propionibacterium acnes biotypes and susceptibility to minocycline and Keigai-rengyo-to. International Journal of Dermatology 43: 103- 107, 2004.

Holmes WF, Soprano DR, Soprano KJ: Early events in the induction of apoptosis in ovarian carcinoma cells by CD437: activation of the p38 MAP kinase signal pathway. Oncogene 22: 6377-86, 2003.

Hommel L, Geiger JM, Harms M, Saurat JH: Sebum excretion rate in subjects treated with oral all-trans-retinoic acid. Dermatology 193: 127-30, 1996.

Hoopes CW, Taketo M, Ozato K, Liu Q, Howard TA, Linney E, et al: Mapping of the mouse Rxr loci encoding nuclear retinoid X receptors RXR alpha, RXR beta, and RXR gamma. Genomics 14: 611-7, 1992.

Hoyos B, Imam A, Chua R, Swenson C, Tong GX, Levi E, et al: The cysteine-rich regions of the regulatory domains of Raf and protein kinase C as retinoid receptors. J Exp Med 192: 835-45, 2000.

Huffmeier U, Lascorz J, Traupe H, Bohm B, Schurmeier-Horst F, Stander M, et al: Systematic linkage disequilibrium analysis of SLC12A8 at PSORS5 confirms a role in susceptibility to psoriasis vulgaris. J Invest Dermatol 125: 906-12, 2005.

Hughes BR, Cunliffe WJ: A prospective study of the effect of isotretinoin on the follicular reservoir and sustainable sebum excretion rate in patients with acne. Arch Dermatol 130: 315-8, 1994.

Hughes SE: Differential expression of the fibroblast growth factor receptor (FGFR) multigene family in normal human adult tissues. The Journal of Histochemistry and Cytochemistry 45: 1005-1019, 1997.

Hyatt GA, Dowling JE: Retinoic acid. A key molecule for eye and photoreceptor development. Invest Ophthalmol Vis Sci 38: 1471-5, 1997.

Idres N, Marill J, Flexor MA, Chabot GG: Activation of Retinoic Acid Receptor-dependent Transcription by All-trans-retinoic Acid Metabolites and Isomers. J. Biol. Chem. 277: 31491- 31498, 2002.

Imam A, Hoyos B, Swenson C, Levi E, Chua R, Viriya E, et al: Retinoids as ligands and coactivators of protein kinase C alpha. Faseb J 15: 28-30, 2001.

Imperato-McGinley J, Gautier T, Cai LQ, Yee B, Epstein J, Pochi P: The androgen control of sebum production. Studies of subjects with dihydrotestosterone deficiency and complete androgen insensitivity. J Clin Endocrinol Metabol 76: 524-528, 1993.

Irizarry RA, Gautier L, Cope LM. The Analysis of Gene Expression Data: Methods and Software: Spriger Verlag, 2003.

191 Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U, et al: Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 4: 249-64, 2003.

Ito M: New findings on the proteins of sebaceous glands. J Invest Dermatol 82: 381, 1984.

Jerabek I, Zechmeister-Machhart M, Binder BR, Geiger M: Binding of retinoic acid by the inhibitory serpin protein C inhibitor. Eur J Biochem 268: 5989-96, 2001.

Jimenez-Lara AM, Clarke N, Altucci L, Gronemeyer H: Retinoic-acid-induced apoptosis in leukemia cells. Trends Mol Med 10: 508-15, 2004.

Jones H, Blanc D, Cunliffe WJ: 13-cis retinoic acid and acne. Lancet 2: 1048-9, 1980.

Jugeau S, Tenaud I, Knol AC, Jarrousse V, Quereux G, Khammari A, et al: Induction of toll-like receptors by Propionibacterium acnes. British Journal of Dermatology 153: 1105-1113, 2005.

Junn E, Han SH, Im JY, Yang Y, Cho EW, Um HD, et al: Vitamin D3 Up-Regulated Protein 1 Mediates Oxidative Stress Via Suppressing the Thioredoxin Function. J Immunol 164: 6287- 6295, 2000.

Kan Kondo SYTSNTTKMIYS: Cisplatin-dependent upregulation of death receptors 4 and 5 augments induction of apoptosis by TNF-related apoptosis-inducing ligand against esophageal squamous cell carcinoma. International Journal of Cancer 118: 230-242, 2006.

Kariya Y, Moriya T, Suzuki T, Chiba M, Ishida K, Takeyama J, et al: Sex steroid hormone receptors in human skin appendage and its neoplasms. Endocr J 52: 317-25, 2005.

Karlsson T, Vahlquist A, Kedishvili N, Torma H: 13-cis-retinoic acid competitively inhibits 3 alpha-hydroxysteroid oxidation by retinol dehydrogenase RoDH-4: a mechanism for its anti- androgenic effects in sebaceous glands? Biochem Biophys Res Commun 303: 273-8, 2003.

Kasman L, Lu P, Voelkel-Johnson C: The histone deacetylase inhibitors depsipeptide and MS- 275, enhance TRAIL gene therapy of LNCaP prostate cancer cells without adverse effects in normal prostate epithelial cells. Cancer Gene Ther 14: 327-334, 2006.

Keedwell RG, Zhao Y, Hammond LA, Qin S, Tsang KY, Reitmair A, et al: A retinoid-related molecule that does not bind to classical retinoid receptors potently induces apoptosis in human prostate cancer cells through rapid caspase activation. Cancer Res 64: 3302-12, 2004.

Kerr JB. Atlas of Functional Histology. Bacelona, Spain: Mosby International Limited, 1999.

Kim M, Ciletti N, Michel S, Reichert U, Rosenfield R: The role of specific retinoid receptors in sebocyte growth and differentiation in culture. J Invest Dermatol 114: 349-353, 2000.

Kim MJ, Ciletti N, Michel S, Reichert U, Rosenfield RL: The role of specific retinoid receptors in sebocyte growth and differentiation in culture. Journal of Investigative Dermatology 114: 349-53, 2000.

192 Kim MJ, Deplewski D, Ciletti N, Michel S, Reichert U, Rosenfield RL: Limited cooperation between peroxisome proliferator-activated receptors and retinoid X receptor agonists in sebocyte growth and development. Mol Genet Metab 74: 362-9, 2001.

Kishibe M, Bando Y, Terayama R, Namikawa K, Takahashi H, Hashimoto Y, et al: Kallikrein 8 Is Involved in Skin Desquamation in Cooperation with Other Kallikreins. J. Biol. Chem. 282: 5834- 5841, 2007.

Kjeldsen L, Cowland JB, Borregaard N: Human neutrophil gelatinase-associated lipocalin and homologous proteins in rat and mouse. Biochim Biophys Acta 1482: 272-83, 2000.

Kligman AM, Fulton JE, Jr., Plewig G: Topical vitamin A acid in acne vulgaris. Arch Dermatol 99: 469-76, 1969.

Kligman LH, Kligman AM: The effect on rhino mouse skin of agents which influence keratinization and exfoliation. J Invest Dermatol 73: 354-8, 1979.

Knaggs H, Holland K, Morris C, Wood E, Cunliffe W: Quantification of cellular proliferation in acne using the monoclonal antibody Ki-67. J Invest Dermatol 102: 89-92., 1994.

Knaggs H, Hughes B, Morris C, Wood E, Holland D, Cunliffe w: Immunohistochemical study of desmosomes in acne vulgaris. Br J Dermatol 130: 731-737., 1994.

Koistinen P, Zheng A, Saily M, Siitonen T, Mantymaa P, Savolainen ER: Superior effect of 9-cis retinoic acid (RA) compared with all-trans RA and 13-cis RA on the inhibition of clonogenic cell growth and the induction of apoptosis in OCI/AML-2 subclones: is the p53 pathway involved? Br J Haematol 118: 401-10, 2002.

Kouhara J, Yoshida T, Nakata S, Horinaka M, Wakada M, Ueda Y, et al: Fenretinide up- regulates DR5/TRAIL-R2 expression via the induction of the transcription factor CHOP and combined treatment with fenretinide and TRAIL induces synergistic apoptosis in colon cancer cell lines. Int J Oncol 30: 679-87, 2007.

Krebs M, Uhrin P, Vales A, Prendes-Garcia MJ, Wojta J, Geiger M, et al: Protein C inhibitor is expressed in keratinocytes of human skin. J Invest Dermatol 113: 32-7, 1999.

Krezel W, Ghyselinck N, Samad TA, Dup, eacute, Val, et al: Impaired Locomotion and Dopamine Signaling in Retinoid Receptor Mutant Mice. Science 279: 863-867, 1998.

LaMantia AS, Bhasin N, Rhodes K, Heemskerk J: Mesenchymal/epithelial induction mediates olfactory pathway formation. Neuron 28: 411-25, 2000.

Lamason RL, Mohideen MA, Mest JR, Wong AC, Norton HL, Aros MC, et al: SLC24A5, a putative cation exchanger, affects pigmentation in zebrafish and humans. Science 310: 1782-6, 2005.

Lammer EJ, Chen DT, Hoar RM, Agnish ND, Benke PJ, Braun JT, et al: Retinoic acid embryopathy. N Engl J Med 313: 837-41, 1985.

193 Lammer EJ, Flannery DB, Barr M: Does isotretinoin cause limb reduction defects? Lancet 2: 328, 1985.

Landthaler M, Kummermehr J, Wagner A, Plewig G: Inhibitory Effects of 13 cis-Reinoic Acid on Human Sebaceous Glands. Archives of Dermatological Research 269: 297-3069, 1980.

Laurent SJ, Mednieks MI, Rosenfield RL: Growth of sebaceous cells in monolayer culture. In vitro Cell Dev Biol 28A: 83-89, 1992.

Layton AM, Thiboutot DM, Bettoli V. Fast Facts-Acne. Oxford, UK: Health Press Limited, 2004

Lee K-W, Ma L, Yan X, Liu B, Zhang X-k, Cohen P: Rapid Apoptosis Induction by IGFBP-3 Involves an Insulin-like Growth Factor-independent Nucleomitochondrial Translocation of RXR{alpha}/Nur77. J. Biol. Chem. 280: 16942-16948, 2005.

Levell MJ: Acne is not associated with abnormal plasma androgens. Br J Dermatol 120: 649, 1989.

Leverkus M, Sprick MR, Wachter T, Denk A, Brocker E-B, Walczak H, et al: TRAIL-Induced Apoptosis and Gene Induction in HaCaT Keratinocytes: Differential Contribution of TRAIL Receptors 1 and 2. 121: 149-155, 2003.

Levin A, Bosakowski T, Kazmer S, Grippo JF: 13-cis Retinoic Acid does not bind to retinoic acid receptors alpha, beta and gamma. Toxicologist 12: 181, 1992.

Leyden JJ: Current issues in antimicrobial therapy for the treatment of acne. J Eur Acad Dermatol Venereol 15 Suppl 3: 51-5, 2001.

Leyden JJ, McGinley KJ, Foglia AN: Qualitative and quantitative changes in cutaneous bacteria associated with systemic isotretinoin therapy for acne conglobata. J Invest Dermatol 86: 390-3, 1986.

Li C, Guo H, Xu X, Weinberg W, Deng C-x: Fibroblast growth factor receptor 2 (Fgfr2) plays an important role in eyelid and skin formation and patterning. Developmental Dynamics 222: 471- 483, 2001.

Li C, Hung Wong W: Model-based analysis of oligonucleotide arrays: model validation, design issues and standard error application. Genome Biol 2: RESEARCH0032, 2001.

Li R, Luo X, Pan Q, Zineh I, Archer DF, Williams RS, et al: Doxycycline alters the expression of inflammatory and immune-related cytokines and chemokines in human endometrial cells: implication in irregular uterine bleeding. Hum. Reprod. 21: 2555-2563, 2006.

Lilyestrom W, Klein MG, Zhang R, Joachimiak A, Chen XS: Crystal structure of SV40 large T- antigen bound to p53: interplay between a viral oncoprotein and a cellular tumor suppressor. Genes Dev 20: 2373-82, 2006.

Loeber G, Tevethia MJ, Schwedes JF, Tegtmeyer P: Temperature-sensitive mutants identify crucial structural regions of simian virus 40 large T antigen. J Virol 63: 4426-30, 1989. 194

Lookingbill DP, Horton R, Demers LM, Egan N, Marks JG, Jr., Santen RJ: Tissue production of androgens in women with acne. J Am Acad Dermatol 12: 481-7, 1985.

Louafi F, Stewart CE, Perks CM, Thomas MG, Holly JM: Role of the IGF-II receptor in mediating acute, non-genomic effects of retinoids and IGF-II on keratinocyte cell death. Exp Dermatol 12: 426-34, 2003.

Lucken-Ardjomande S, Martinou JC: Regulation of Bcl-2 proteins and of the permeability of the outer mitochondrial membrane. C R Biol 328: 616-31, 2005.

Maden M: Retinoids and spinal cord development. J Neurobiol 66: 726-38, 2006.

Maeda Y, Miyatake J, Sono H, Matsuda M, Tatsumi Y, Horiuchi F, et al: 13-cis retinoic acid inhibits growth of adult T cell leukemia cells and causes apoptosis; possible new indication for retinoid therapy. Intern Med 35: 180-4, 1996.

Mangelsdorf DJ: Vitamin A receptors. Nutr Rev 52: S32-44, 1994.

Mangelsdorf DJ, Ong ES, Dyck JA, Evans RM: Nuclear receptor that identifies a novel retinoic acid response pathway. Nature 345: 224-9, 1990.

Mangelsdorf DJ, Thummel C, Beato M, Herrlich P, Schutz G, Umesono K, et al: The nuclear receptor superfamily: the second decade. Cell 83: 835-9, 1995.

Marikar Y, Wang Z, Duell EA, Petkovich M, Voorhees JJ, Fisher GJ: Retinoic acid receptors regulate expression of retinoic acid 4-hydroxylase that specifically inactivates all-trans retinoic acid in human keratinocyte HaCaT cells. J Invest Dermatol 111: 434-9, 1998.

Marples RR, Leyden JJ, Stewart RN, Mills OH, Jr., Kligman AM: The skin microflora in acne vulgaris. J Invest Dermatol 62: 37-41, 1974.

Marsters SA, Sheridan JP, Pitti RM, Huang A, Skubatch M, Baldwin D, et al: A novel receptor for Apo2L/TRAIL contains a truncated death domain. Curr Biol 7: 1003-6, 1997.

Martin SJ, Reutelingsperger CP, McGahon AJ, Rader JA, van Schie RC, LaFace DM, et al: Early redistribution of plasma membrane phosphatidylserine is a general feature of apoptosis regardless of the initiating stimulus: inhibition by overexpression of Bcl-2 and Abl. J Exp Med 182: 1545-56, 1995.

Matias JR, Orentreich N: The hamster ear sebaceous glands. I. Examination of the regional variation by stripped skin planimetry. J Invest Dermatol 81: 43-6, 1983.

Mattei MG, de The H, Mattei JF, Marchio A, Tiollais P, Dejean A: Assignment of the human hap retinoic acid receptor RAR beta gene to the p24 band of chromosome 3. Hum Genet 80: 189- 90, 1988.

Matysiak M, Jurewicz A, Jaskolski D, Selmaj K: TRAIL induces death of human oligodendrocytes isolated from adult brain. Brain 125: 2469-80, 2002.

195 McCaffery P, Drager UC: Regulation of retinoic acid signaling in the embryonic nervous system: a master differentiation factor. Cytokine Growth Factor Rev 11: 233-49, 2000.

Mendelsohn C, Batourina E, Fung S, Gilbert T, Dodd J: Stromal cells mediate retinoid- dependent functions essential for renal development. Development 126: 1139-48, 1999.

Merrill B, Gat U, DasGupta R, Fuchs E: Tcf3 and Lef1 regulate lineage differentiation of multipotent stem cells in skin. Genes & Development 15: 1688-1705, 2001.

Minucci S, Pelicci PG: Retinoid receptors in health and disease: co-regulators and the chromatin connection. Semin Cell Dev Biol 10: 215-25, 1999.

Miyake K, Ciletti N, Liao S, Rosenfield RL: Androgen receptor expression in the preputial gland and its sebocytes. Journal of Investigative Dermatology 103: 721-5, 1994.

Mizuiri H, Yoshida K, Toge T, Oue N, Aung PP, Noguchi T, et al: DNA methylation of genes linked to retinoid signaling in squamous cell carcinoma of the esophagus: DNA methylation of CRBP1 and TIG1 is associated with tumor stage. Cancer Science 96: 571-577, 2005.

Mollard R, Ghyselinck NB, Wendling O, Chambon P, Mark M: Stage-dependent responses of the developing lung to retinoic acid signaling. Int J Dev Biol 44: 457-62, 2000.

Montagna W, Parakkal PF. The Structure and Function of Skin. New York, NY, U.S.A: Academic Press, 1974.

Mourelatos K, Eady EA, Cunliffe WJ, Clark SM, Cove JH: Temporal changes in sebum excretion and propionibacterial colonization in preadolescent children with and without acne. Br J Dermatol 156: 22-31, 2007.

Muller M, Jasmin JR, Monteil RA, Loubiere R: Embryology of the hair follicle. Early Hum Dev 26: 159-66, 1991.

Munro CS, Wilkie AOM: Epidermal mosaicism producing localized acne: somatic mutation in FGFR2. The Lancet 352: 704-705, 1998.

Muzio M, Chinnaiyan AM, Kischkel FC, O'Rourke K, Shevchenko A, Ni J, et al: FLICE, a novel FADD-homologous ICE/CED-3-like protease, is recruited to the CD95 (Fas/APO-1) death-- inducing signaling complex. Cell 85: 817-27, 1996.

Nadeau SI, Landry J: Mechanisms of activation and regulation of the heat shock-sensitive signaling pathways. Adv Exp Med Biol 594: 100-13, 2007.

Nagpal S, Patel S, Asano A, Johnson A, Duvic M, Chandraratna R: TIG1 and TGI2 (tazarotene- induced genes 1 and 2) are novel retinoic acid receptor responsive genes in skin. J Invest Dermatol 106: 818, 1996.

Nagy I, Pivarcsi A, Kis K, Koreck A, Bodai L, McDowell A, et al: Propionibacterium acnes and lipopolysaccharide induce the expression of antimicrobial peptides and proinflammatory cytokines/chemokines in human sebocytes. Microbes Infect 8: 2195-205, 2006.

196 Nakagawa S, Fujii T, Yokoyama G, Kazanietz MG, Yamana H, Shirouzu K: Cell growth inhibition by all-trans retinoic acid in SKBR-3 breast cancer cells: involvement of protein kinase Calpha and extracellular signal-regulated kinase mitogen-activated protein kinase. Mol Carcinog 38: 106-16, 2003.

Napoli JL: Biochemical pathways of retinoid transport, metabolism, and signal transduction. Clin Immunol Immunopathol 80: S52-62, 1996.

Nelson AM, Gilliland KL, Cong Z, Thiboutot DM: 13-cis Retinoic acid induces apoptosis and cell cycle arrest in human SEB-1 sebocytes. J Invest Dermatol 126: 2178-89, 2006.

Nesterov A, Ivashchenko Y, Kraft AS: Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) triggers apoptosis in normal prostate epithelial cells. Oncogene 21: 1135-40, 2002.

Nevins JR: The Rb/E2F pathway and cancer. Hum Mol Genet 10: 699-703, 2001.

Newcomer ME, Ong DE: Plasma retinol binding protein: structure and function of the prototypic lipocalin. Biochim Biophys Acta 1482: 57-64, 2000.

Nguyen H, Rendl M, Fuchs E: Tcf3 Governs Stem Cell Features and Represses Cell Fate Determination in Skin. Cell 127: 171-183, 2006.

Nicolaides N: Skin lipids: their biochemical uniqueness. Science 186: 19-26, 1974.

Nicolaides N, Ansari MN, Fu HC, Lindsay DG: Lipid compsition on comedones compared with that of human skin surface in acne patients. J Invest Dermatol 54: 487-95, 1970.

Norris DA, Osborn R, Robinson W, Tonnesen MG: Isotretinoin produces significant inhibition of monocyte and neutrophil chemotaxis in vivo in patients with cystic acne. J Invest Dermatol 89: 38-43, 1987.

Ochoa WF, Torrecillas A, Fita I, Verdaguer N, Corbalan-Garcia S, Gomez-Fernandez JC: Retinoic acid binds to the C2-domain of protein kinase C(alpha). Biochemistry 42: 8774-9, 2003.

Ohtani K, DeGregori J, Nevins JR: Regulation of the cyclin E gene by transcription factor E2F1. Proc Natl Acad Sci U S A 92: 12146-50, 1995.

Olson JM, Hallahan AR: p38 MAP kinase: a convergence point in cancer therapy. Trends Mol Med 10: 125-9, 2004.

Orfanos CE, Zouboulis CC: Oral retinoids in the treatment of seborrhoea and acne. Dermatology 196: 140-7, 1998.

Ott F, Bollag W, Geiger JM: Oral 9-cis-retinoic acid versus 13-cis-retinoic acid in acne therapy. Dermatology 193: 124-6, 1996.

Papoutsaki M, Lanza M, Marinari B, Nistico S, Moretti F, Levrero M, et al: The p73 gene is an anti-tumoral target of the RARbeta/gamma-selective retinoid tazarotene. J Invest Dermatol 123: 1162-8, 2004.

197 Pasquali D, Bellastella A, Colantuoni V, Vassallo P, Bonavolonta G, Rossi V, et al: All-trans retinoic acid- and N-(4-hydroxyphenil)-retinamide-induced growth arrest and apoptosis in orbital fibroblasts in Graves' disease. Metabolism 52: 1387-92, 2003.

Peck GL: Prolonged remissions of cystic acne with 13-cis-retinoic acid. N Engl J Med 300: 329, 1979.

Peck GL, Yoder FW: Treatment of lamellar ichthyosis and other keratinising dermatoses with an oral synthetic retinoid. Lancet 2: 1172-4, 1976.

Pelletier G, Ren L: Localization of sex steroid receptors in human skin. Histol Histopathol 19: 629-36, 2004.

Perisho K, Wertz PW, Madison KC, Stewart ME, Downing DT: Fatty acids of acylceramides from comedones and from the skin surface of acne patients and control subjects. Journal of Investigative Dermatology 90: 350-3, 1988.

Pettersson F, Couture M-C, Hanna N, Miller Jr. W: Enhanced retinoid-induced apoptosis of MDA-MB-231 breast cancer cells by PKC inhibitors involved activation of ERK. Oncogene 23: 7053-7066, 2004.

Pietras RJ: Sex pheromone production by preputial gland: the regulatory role of estrogen. Chem. Senses 6: 391-408, 1981.

Plewig G, Christophers E: Renewal rate of human sebaceous glands. Acta Derm Venereol 54: 177-82, 1974.

Plewig G, Christophers E, Braun-Falco O: Cell transition in human sebaceous glands. Acta Derm Venereol 51: 423-8, 1971.

Plewig G, Dressel H, Pfleger M, Michelsen S, Kligman AM: Low dose isotretinoin combined with tretinoin is effective to correct abnormalities of acne. J Dtsch Dermatol Ges 2: 31-45, 2004.

Plewig G, Luderschmidt C: Hamster ear model for sebaceous glands. J Invest Dermatol 68: 171-6, 1977.

Pochi PE, Strauss JS: Sebaceous gland response in man to the administration of testosterone, ∆4- androstenedione, and dehydroisoandrosterone. J Invest Dermatol 52: 32-36, 1969.

Pochi PE, Strauss JS: Sebaceous gland response in man to the administration of testosterone, delta-4-androstenedione, and dehydroisoandrosterone. J Invest Dermatol 52: 32-6, 1969.

Pochi PE, Strauss JS, Downing DT: Skin surface lipid composition, acne, pubertal development, and urinary excretion of testosterone and 17-ketosteroids in children. J Invest Dermatol 69: 485- 9, 1977.

Pomponi F, Cariati R, Zancai P, De Paoli P, Rizzo S, Tedeschi RM, et al: Retinoids irreversibly inhibit in vitro growth of Epstein-Barr virus-immortalized B lymphocytes. Blood 88: 3147-59, 1996.

198 Potter JE, Prutkin L, Wheatley VR: Sebaceous gland differentiation. I. Separation, morphology and lipogenesis of isolated cells from the mouse preputial gland tumor. J Invest Dermatol 72: 120-7, 1979.

Pratt MA, Niu MY, Renart LI: Regulation of survivin by retinoic acid and its role in paclitaxel- mediated cytotoxicity in MCF-7 breast cancer cells. Apoptosis 11: 589-605, 2006.

Puhvel SM, Reisner RM, Amirian DA: Quantification of bacteria in isolated pilosebaceous follicles in normal skin. J Invest Dermatol 65: 525-31, 1975.

Qin JZ, Bacon P, Chaturvedi V, Nickoloff BJ: Role of NF-kappaB activity in apoptotic response of keratinocytes mediated by interferon-gamma, tumor necrosis factor-alpha, and tumor- necrosis-factor-related apoptosis-inducing ligand. J Invest Dermatol 117: 898-907, 2001.

Quelle DE, Ashmun RA, Shurtleff SA, Kato JY, Bar-Sagi D, Roussel MF, et al: Overexpression of mouse D-type cyclins accelerates G1 phase in rodent fibroblasts. Genes Dev 7: 1559-71, 1993.

Ramp U, Gerharz CD, Eifler E, Biesalski HK, Gabbert HE: Effects of retinoic acid metabolites on proliferation and differentiation of the clonal rhabdomyosarcoma cell line BA-HAN-1C. Biology of the Cell 81: 31-37, 1994.

Ray S, Anderson ME, Tegtmeyer P: Differential interaction of temperature-sensitive simian virus 40 T antigens with tumor suppressors pRb and p53. J Virol 70: 7224-7, 1996.

Recchia F, Saggio G, Cesta A, Candeloro G, Di Blasio A, Amiconi G, et al: Phase II study of interleukin-2 and 13-cis-retinoic acid as maintenance therapy in metastatic colorectal cancer. Cancer Immunol Immunother 56: 699-708, 2007.

Reichel RR, Jacob ST: Control of gene expression by lipophilic hormones. Faseb J 7: 427-36, 1993.

Reynolds CP, Matthay KK, Villablanca JG, Maurer BJ: Retinoid therapy of high-risk neuroblastoma. Cancer Lett 197: 185-92, 2003.

Ridden J, Ferguson D, Kealey T: Organ Maintenance of Human Sebaceous Glands: In Vitro Effects of 13-cis Retinoic Acid and Testosterone. J Cell Sci 95: pp.125-136., 1990.

Rigobello MP, Scutari G, Friso A, Barzon E, Artusi S, Bindoli A: Mitochondrial permeability transition and release of cytochrome c induced by retinoic acids. Biochem Pharmacol 58: 665- 70, 1999.

Ro BI, Dawson TL: The role of sebaceous gland activity and scalp microfloral metabolism in the etiology of seborrheic dermatitis and dandruff. J Investig Dermatol Symp Proc 10: 194-7, 2005.

Roche-Laboratories. System to Manage Accutane Related Teratogenicity -S.M.A.R.T Guide to Best Practices. 2001.

Rollman O, Vahlquist A: Oral isotretinoin (13-cis-retinoic acid) therapy in severe acne: drug and vitamin A concentrations in serum and skin. J Invest Dermatol 86: 384-9, 1986. 199 Romand R, Dolle P, Hashino E: Retinoid signaling in inner ear development. J Neurobiol 66: 687-704, 2006.

Roos TC, Jugert FK, Merk HF, Bickers DR: Retinoid metabolism in the skin. Pharmacol Rev 50: 315-33, 1998.

Rosdahl I, Andersson E, Kagedal B, Torma H: Vitamin A metabolism and mRNA expression of retinoid-binding protein and receptor genes in human epidermal melanocytes and melanoma cells. Melanoma Res 7: 267-74, 1997.

Rosen ED, Sarraf P, Troy AE, Bradwin G, Moore K, Milstone DS, et al: PPAR gamma is required for the differentiation of adipose tissue in vivo and in vitro. Molecular Cell 4: 611-7, 1999.

Rosenfield RL: Relationship of sebaceous cell stage to growth in culture. J Invest Dermatol 92: 751-4, 1989.

Rosenfield RL, Kentsis A, Deplewski D, Ciletti N: Rat preputial sebocyte differentiation involves peroxisome proliferator-activated receptors. Journal of Investigative Dermatology 112: 226-32, 1999.

Saint-Leger D, Cohen E: Practical study of qualitative and quantitative sebum excretion on the human forehead. Br J Dermatol 113: 551-7, 1985.

Sakai Y, Crandall JE, Brodsky J, McCaffery P: 13-cis Retinoic acid (accutane) suppresses hippocampal cell survival in mice. Ann N Y Acad Sci 1021: 436-40, 2004.

Sanlioglu, Koksal, Ciftcioglu, Baykara, Luleci, Sanlioglu: Differential Expression of TRAIL and its Receptors in Benign and Malignant Prostate Tissues. The Journal of Urology 177: 359-364, 2007.

Sasaki H, Ohara N, Xu Q, Wang J, DeManno DA, Chwalisz K, et al: A novel selective progesterone receptor modulator asoprisnil activates tumor necrosis factor-related apoptosis- inducing ligand (TRAIL)-mediated signaling pathway in cultured human uterine leiomyoma cells in the absence of comparable effects on myometrial cells. J Clin Endocrinol Metab 92: 616-23, 2007.

Sawaya ME, Price VH: Different levels of 5α-reductase type I and II, aromatase, and androgen receptor in hair follicles of women and men with androgenetic alopecia. J invest Dermatol 109: 296-300, 1997.

Schmieg FI, Simmons DT: Characterization of the in vitro interaction between SV40 T antigen and p53: mapping the p53 binding site. Virology 164: 132-40, 1988.

Schoonjans K, Staels B, Auwerx J: The peroxisome proliferator activated receptors (PPARs) and their effects on lipid metabolism and adipocyte differentiation. Biochim Biophys Acta 1302: 93-109, 1996.

200 Seguin-Devaux C, Hanriot D, Dailloux M, Latger-Cannard V, Zannad F, Mertes PM, et al: Retinoic acid amplifies the host immune response to LPS through increased T lymphocytes number and LPS binding protein expression. Mol Cell Endocrinol 245: 67-76, 2005.

Seiberg M, Siock P, Wisniewski S, Cauwenbergh G, Shapiro SS: The effects of trypsin on apoptosis, utriculi size, and skin elasticity in the Rhino mouse. J Invest Dermatol 109: 370-6, 1997.

Sen GC, Sarkar SN: Hitching RIG to action. Nat Immunol 6: 1074-6, 2005.

Seo SJ, Ahn JY, Hong CK, Seo EY, Kye KC, Lee WH, et al: Expression of neutrophil gelatinase-associated lipocalin in skin epidermis. J Invest Dermatol 126: 510-2, 2006.

Shaw JC, White LE: Long-term safety of spironolactone in acne: results of an 8-year followup study. J Cutan Med Surg 6: 541-5, 2002.

Sheridan JP, Marsters SA, Pitti RM, Gurney A, Skubatch M, Baldwin D, et al: Control of TRAIL- induced apoptosis by a family of signaling and decoy receptors. Science 277: 818-21, 1997.

Shimada O, Wu X, Jin X, Nouh MA, Fiscella M, Albert V, et al: Human agonistic antibody to tumor necrosis factor-related apoptosis-inducing ligand receptor 2 induces cytotoxicity and apoptosis in prostate cancer and bladder cancer cells. Urology 69: 395-401, 2007.

Shutoh M, Oue N, Aung PP, Noguchi T, Kuraoka K, Nakayama H, et al: DNA methylation of genes linked with retinoid signaling in gastric carcinoma: expression of the retinoid acid receptor beta, cellular retinol-binding protein 1, and tazarotene-induced gene 1 genes is associated with DNA methylation. Cancer 104: 1609-19, 2005.

Silverman GA, Bird PI, Carrell RW, Church FC, Coughlin PB, Gettins PG, et al: The serpins are an expanding superfamily of structurally similar but functionally diverse proteins. Evolution, mechanism of inhibition, novel functions, and a revised nomenclature. J Biol Chem 276: 33293- 6, 2001.

Simeone A, Acampora D, Arcioni L, Andrews PW, Boncinelli E, Mavilio F: Sequential activation of HOX2 homeobox genes by retinoic acid in human embryonal carcinoma cells. Nature 346: 763-6, 1990.

Slee EA, Adrain C, Martin SJ: Serial killers: ordering caspase activation events in apoptosis. Cell Death Differ 6: 1067-74, 1999.

Smith KJ, Diwan H, and Skelton H: Death receptors and their role in dermatology, with particular focus on tumor necrosis factor-related apoptosis-inducing ligand receptors. Int J Dermatol 42: 3- 17, 2003.

Smith TM, Cong Z, Gilliland KL, Clawson GA, Thiboutot DM: Insulin-like growth factor-1 induces lipid production in human SEB-1 sebocytes via sterol response element-binding protein-1. J Invest Dermatol 126: 1226-32, 2006.

Smith TM, Gilliland KL, Clawson GA, Thiboutot, DM. IGF-1 induces SREBP-1 expression and lipogenesis in SEB-1 sebocytes via PI3Kinase/AKT pathway activation. in preparation. 201

Sorg O, Kuenzli S, Kaya G, Saurat JH: Proposed mechanisms of action for retinoid derivatives in the treatment of skin aging. J Cosmet Dermatol 4: 237-44, 2005.

Spahn CM, Prescott CD: Throwing a spanner in the works: antibiotics and the translation apparatus. J Mol Med 74: 423-39, 1996.

Spiegelman BM, Hu E, Kim JB, Brun R: PPAR gamma and the control of adipogenesis. Biochimie 79: 111-2, 1997.

Srinivasula SM, Ahmad M, Fernandes-Alnemri T, Litwack G, Alnemri ES: Molecular ordering of the Fas-apoptotic pathway: the Fas/APO-1 protease Mch5 is a CrmA-inhibitable protease that activates multiple Ced-3/ICE-like cysteine proteases. Proc Natl Acad Sci U S A 93: 14486-91, 1996.

Stander S, Schwarz T: Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) is expressed in normal skin and cutaneous inflammatory diseases, but not in chronically UV- exposed skin and non-melanoma skin cancer. Am J Dermatopathol 27: 116-21, 2005.

Stewart ME, Benoit AM, Stranieri AM, Rapini RP, Strauss JS, Downing DT: Effect of oral 13-cis Retinoic Acid at three dose levels on sustainable rates of sebum secretion and on acne. Journal of the American Academy of Dermatology 8: 532-538, 1983.

Stewart ME, Downing DT. Chemistry and function of mammalian sebaceous lipids. In: Advances in Lipid Research, edited by P. M. Elias. San Diego: Academic Press, 1991, p. 263.

Stewart ME, Greenwood R, Cunliffe WJ, Strauss JS, Downing DT: Effect of cyproterone acetate-ethinyl estradiol treatment on the proportions of linoleic and sebaleic acids in various skin surface lipid classes. Archives of Dermatological Research 278: 481-5, 1986.

Strauss JS, Pochi PE. The hormonal control of human sebaceous gland. In: Advances in the Biology of Skin: The sebaceous glands, edited by W. Montagna, R. Ellis and A. Silver. Oxford: Pergamon Press, 1963, p. 220-254.

Strauss JS, Stranieri AM, Farrell LN, Downing DT: The effect of marked inhibition of sebum production with 13 cis-retinoic acid on skin surface lipid composition. J Invest Dermatol 74: 66-7, 1980.

Student AK, Hsu RY, Lane MD: Induction of fatty acid synthetase synthesis in differentiating 3T3-L1 preadipocytes. J Biol Chem 255: 4745-50, 1980.

Sun SY, Yue P, Hong WK, Lotan R: Augmentation of tumor necrosis factor-related apoptosis- inducing ligand (TRAIL)-induced apoptosis by the synthetic retinoid 6-[3-(1-adamantyl)-4- hydroxyphenyl]-2-naphthalene carboxylic acid (CD437) through up-regulation of TRAIL receptors in human lung cancer cells. Cancer Res 60: 7149-55, 2000.

Sun SY, Yue P, Lotan R: Implication of multiple mechanisms in apoptosis induced by the synthetic retinoid CD437 in human prostate carcinoma cells. Oncogene 19: 4513-22, 2000. Takeda A, Higuchi D, Takahashi T, Ogo M, Baciu P, Goetinck PF, et al: Overexpression of serpin squamous cell carcinoma antigens in psoriatic skin. J Invest Dermatol 118: 147-54, 2002. 202 Thiboutot D: Acne and Rosacea: New and Emerging Therapies. Dermatol Clinics 18: 63-71, 2000.

Thiboutot D: Regulation of human sebaceous glands. J Invest Dermatol 123: 1-12, 2004.

Thiboutot D, Chen W: Update and future of hormonal therapy in acne. Dermatology 206: 57-67, 2003.

Thiboutot D, Gilliland K, Light J, Lookingbill D: Androgen metabolism in sebaceous glands from subjects with and without acne [see comments]. Archives of Dermatology 135: 1041-5, 1999.

Thiboutot D, Harris G, Iles V, Cimis G, Gilliland K, Hagari S: Characterization of the 5a- reductase enzyme in isolated sebaceous glands and whole skin obtained from face, scalp and nonacne-prone regions. J Invest Dermatol 104: 607, 1995.

Thiboutot D, Jabara S, McAllister J, Sivarajah A, Gilliland K, Cong Z, et al: Human skin is a steroidogenic tissue: Steroidogenic enzymes and cofactors are expressed in epidermis, normal sebocytes, and an immortalized sebocyte cell line (SEB-1). J Invest Dermatol 120: 905-914, 2003.

Thiboutot D, Martin P, Volikos L, Gilliland K: Oxidative activity of the type 2 isozyme of 17β- hydroxysteroid dehydrogenase (17β-HSD) predominates in human sebaceous glands. J Invest Dermatol 111: 390-395, 1998.

Thiboutot D, Sivarajah A, Gilliland K, Cong Z, Clawson G: The melanocortin 5 receptor is expressed in human sebaceous glands and rat preputial cells. J Invest Dermatol 115: 614-619, 2000.

Thiele JJ, Weber SU, Packer L: Sebaceous gland secretion is a major physiologic route of vitamin E delivery to skin. J Invest Dermatol 113: 1006-10, 1999.

Thody AJ, Dijkstra H: Effect of ovarian steroids and alpha-melanocyte-stimulating hormone on preputial gland sex attractants in the female rat [proceedings]. J Endocrinol 77: 48P-49P, 1978.

Thompson JN, Howell JM, Pitt GA, McLaughlin CI: The biological activity of retinoic acid in the domestic fowl and the effects of vitamin A deficiency on the chick embryo. Br J Nutr 23: 471-90, 1969.

Thornton MJ, Taylor AH, Mulligan K, Al-Azzawi F, Lyon CC, O'Driscoll J, et al: The distribution of estrogen receptor beta is distinct to that of estrogen receptor alpha and the androgen receptor in human skin and the pilosebaceous unit. J Investig Dermatol Symp Proc 8: 100-3, 2003.

Toma S, Isnardi L, Raffo P, Dastoli G, De Francisci E, Riccardi L, et al: Effects of all-trans- retinoic acid and 13-cis-retinoic acid on breast-cancer cell lines: growth inhibition and apoptosis induction. Int J Cancer 70: 619-27, 1997.

Tong Z, Wu X, Kehrer JP: Increased expression of the lipocalin 24p3 as an apoptotic mechanism for MK886. Biochem J 372: 203-10, 2003.

203 Torma H, Rollman O, Vahlquist A: The vitamin A metabolism and expression of retinoid-binding proteins differ in HaCaT cells and normal human keratinocytes. Arch Dermatol Res 291: 339-45, 1999.

Tosi P, Pellacani A, Visani G, Ottaviani E, Ronconi S, Zamagni E, et al: In vitro treatment with retinoids decreases bcl-2 protein expression and enhances dexamethasone-induced cytotoxicity and apoptosis in multiple myeloma cells. Eur J Haematol 62: 143-8, 1999.

Toyoda M, Nakamura M, Morohashi M: Neuropeptides and sebaceous glands. Eur J Dermatol 12: 422-7, 2002.

Tremblay MR, Luu-The V, Leblanc G, Noel P, Breton E, Labrie F, et al: Spironolactone-related inhibitors of type II 17beta-hydroxysteroid dehydrogenase: chemical synthesis, receptor binding affinities, and proliferative/antiproliferative activities. Bioorg Med Chem 7: 1013-23, 1999.

Trivedi NR, Cong Z, Nelson AM, Albert AJ, Rosamilia LL, Sivarajah S, et al: Peroxisome proliferator-activated receptors increase human sebum production. J Invest Dermatol 126: 2002-9, 2006.

Trivedi NR, Gilliland KL, Zhao W, Liu W, Thiboutot DM: Gene array expression profiling in acne lesions reveals marked upregulation of genes involved in inflammation and matrix remodeling. J Invest Dermatol 126: 1071-9, 2006.

Tsukada M, Schroder M, Roos TC, Chandraratna RA, Reichert U, Merk HF, et al: 13-cis retinoic acid exerts its specific activity on human sebocytes through selective intracellular isomerization to all-trans retinoic acid and binding to retinoid acid receptors. J Invest Dermatol 115: 321-7, 2000.

Tusher VG, Tibshirani R, Chu G: Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci U S A 98: 5116-21, 2001.

Van Scott E, Yu R: Hyperkeratinization, corneocyte cohesion, and alpha hydroxy acids. J Am Acad Dermatol 11: 867-879. 1984.

Vanoosten RL, Earel JK, Jr., Griffith TS: Histone deacetylase inhibitors enhance Ad5-TRAIL killing of TRAIL-resistant prostate tumor cells through increased caspase-2 activity. Apoptosis 12: 561-71, 2007.

Vermot J, Niederreither K, Garnier JM, Chambon P, Dolle P: Decreased embryonic retinoic acid synthesis results in a DiGeorge syndrome phenotype in newborn mice. Proc Natl Acad Sci U S A 100: 1763-8, 2003.

Voellmy R, Boellmann F: Chaperone regulation of the heat shock protein response. Adv Exp Med Biol 594: 89-99, 2007.

Wang Q, Chen Q, Zhao K, Wang L, Wang L, Traboulsi EI: Update on the molecular genetics of retinitis pigmentosa. Ophthalmic Genet 22: 133-54, 2001. Watabe H, Valencia JC, Yasumoto K, Kushimoto T, Ando H, Muller J, et al: Regulation of tyrosinase processing and trafficking by organellar pH and by proteasome activity. J Biol Chem 279: 7971-81, 2004. 204 Webster G: Mechanisms of Propionibacterium acnes-mediated inflammation in acne vulgaris. Semin Dermatol 1: 299, 1982.

Webster GF: Complement activation in acne vulgaris. Consumption of complement by comedones. Infect Immuno 26: 183, 1979.

Webster GF: Neutrophil lysosomal release in response to Propionibacterium acnes. J Invest Dermatol 72: 209, 1979.

Wehrli P, Viard I, Bullani R, Tschopp J, French LE: Death receptors in cutaneous biology and disease. J Invest Dermatol 115: 141-8, 2000.

Weiss JS, Ellis CN, Headington JT, Tincoff T, Hamilton TA, Voorhees JJ: Topical tretinoin improves photoaged skin. A double-blind vehicle-controlled study. Jama 259: 527-32, 1988.

White GM: Recent findings in the epidemiologic evidence, classification, and subtypes of acne vulgaris. J Am Acad Dermatol 39: S34-7, 1998.

Wiley SR, Schooley K, Smolak PJ, Din WS, Huang CP, Nicholl JK, et al: Identification and characterization of a new member of the TNF family that induces apoptosis. Immunity 3: 673- 82, 1995.

Wilkinson DG, Bhatt S, Cook M, Boncinelli E, Krumlauf R: Segmental expression of Hox-2 homoeobox-containing genes in the developing mouse hindbrain. Nature 341: 405-9, 1989.

Williams JB, Napoli JL: Metabolism of retinoic acid and retinol during differentiation of F9 embryonal carcinoma cells. Proc Natl Acad Sci U S A 82: 4658-62, 1985.

Williams ML, Hincenbergs M, Holbrook KA: Skin lipid content during early fetal development. J Invest Dermatol 91: 263-8, 1988.

Wirmer J, Westhof E: Molecular contacts between antibiotics and the 30S ribosomal particle. Methods Enzymol 415: 180-202, 2006.

Wolbach SB, Howe PR: Tissue changes following deprivation of fat soluble vitamin A. J Exp Med 43: 753, 1925.

Wood H, Pall G, Morriss-Kay G: Exposure to retinoic acid before or after the onset of somitogenesis reveals separate effects on rhombomeric segmentation and 3' HoxB gene expression domains. Development 120: 2279-85, 1994.

Wozel G, Chang A, Zultak M, Czarnetzki BM, Happle R, Barth J, et al: The effect of topical retinoids on the leukotriene-B4-induced migration of polymorphonuclear leukocytes into human skin. Arch Dermatol Res 283: 158-61, 1991.

Wrobel A, Seltmann H, Fimmel S, Muller-Decker K, Tsukada M, Bogdanoff B, et al: Differentiation and apoptosis in human immortalized sebocytes. J Invest Dermatol 120: 175- 181, 2003.

205 Wu JM, DiPietrantonio AM, Hsieh TC: Mechanism of fenretinide (4-HPR)-induced cell death. Apoptosis 6: 377-88, 2001.

Wyllie AH, Kerr JF, Currie AR: Cell death: the significance of apoptosis. Int Rev Cytol 68: 251- 306, 1980.

Xia LQ, Zouboulis C, Detmar M, Mayer-da-Silva A, Stadler R, Orfanos CE: Isolation of human sebaceous glands and cultivation of sebaceous gland-derived cells as an in vitro model. Journal of Investigative Dermatology 93: 315-21, 1989.

Yang J, Goetz D, Li JY, Wang W, Mori K, Setlik D, et al: An iron delivery pathway mediated by a lipocalin. Mol Cell 10: 1045-56, 2002.

Yemisci A, Gorgulu A, Piskin S: Effects and side-effects of spironolactone therapy in women with acne. J Eur Acad Dermatol Venereol 19: 163-6, 2005.

Yoneyama M, Kikuchi M, Natsukawa T, Shinobu N, Imaizumi T, Miyagishi M, et al: The RNA helicase RIG-I has an essential function in double-stranded RNA-induced innate antiviral responses. Nat Immunol 5: 730-7, 2004.

Youssef EM, Chen XQ, Higuchi E, Kondo Y, Garcia-Manero G, Lotan R, et al: Hypermethylation and silencing of the putative tumor suppressor Tazarotene-induced gene 1 in human cancers. Cancer Res 64: 2411-7, 2004.

Yu VC, Delsert C, Andersen B, Holloway JM, Devary OV, Naar AM, et al: RXR beta: a coregulator that enhances binding of retinoic acid, thyroid hormone, and vitamin D receptors to their cognate response elements. Cell 67: 1251-66, 1991.

Zancai P, Cariati R, Rizzo S, Boiocchi M, Dolcetti R: Retinoic acid-mediated growth arrest of EBV-immortalized B lymphocytes is associated with multiple changes in G1 regulatory proteins: p27Kip1 up-regulation is a relevant early event. Oncogene 17: 1827-36, 1998.

Zhang H, Rosdahl I: Expression of p27 and MAPK proteins involved in all-trans retinoic acid- induced apoptosis and cell cycle arrest in matched primary and metastatic melanoma cells. Int J Oncol 25: 1241-8, 2004.

Zhang HS, Postigo AA, Dean DC: Active transcriptional repression by the Rb-E2F complex mediates G1 arrest triggered by p16INK4a, TGFbeta, and contact inhibition. Cell 97: 53-61, 1999.

Zhang J, Liu L, Pfeifer GP: Methylation of the retinoid response gene TIG1 in prostate cancer correlates with methylation of the retinoic acid receptor beta gene. Oncogene 23: 2241-9, 2004. Zhang L, Li WH, Anthonavage M, Eisinger M: Melanocortin-5 receptor: a marker of human sebocyte differentiation. Peptides 27: 413-20, 2006.

Zhang LI, Anthonavage M, Huang Q, Li W-H, Eisinger M: Proopiomelanocortin Peptides and Sebogenesis. Ann NY Acad Sci 994: 154-161, 2003.

Zhang X-k: Targeting Nur77 translocation. Expert Opinion on Therapeutic Targets 11: 69-79, 2007. 206 Zhang Z, Balmer JE, Lovlie A, Fromm SH, Blomhoff R: Specific teratogenic effects of different retinoic acid isomers and analogs in the developing anterior central nervous system of zebrafish. Dev Dyn 206: 73-86, 1996.

Zhao Y, Qin S, Atangan LI, Molina Y, Okawa Y, Arpawong HT, et al: Casein Kinase 1α Interacts with Retinoid X Receptor and Interferes with Agonist-induced Apoptosis. Journal of Biological Chemistry 279: 30844-30849, 2004.

Zorn NE, Sauro MD: Retinoic acid induces translocation of protein kinase C (PKC) and activation of nuclear PKC (nPKC) in rat splenocytes. Int J Immunopharmacol 17: 303-11, 1995.

Zouboulis CC, Akamatsu H, Stephanek K, Orfanos CE: Androgens affect the activity of human sebocytes in culture in a manner dependent on the localization of the sebaceous glands and their effect is antagonized by spironolactone. Skin Pharmacology 7: 33-40, 1994.

Zouboulis CC, Korge B, Akamatsu H, Xia LQ, Schiller S, Gollnick H, et al: Effects of 13-cis- retinoic acid, all-trans-retinoic acid, and acitretin on the proliferation, lipid synthesis and keratin expression of cultured human sebocytes in vitro. J Invest Dermatol 96: 792-7, 1991.

Zouboulis CC, Korge BP, Mischke D, Orfanos CE: Altered Proliferation, Synthetic Activity, and Differentiation of Cultured Human Sebocytes in the Absence of Vitamin A and Their Modulation by Synthetic Retinoids. J Invest Dermatol 101: pp.628-633., 1993.

Zouboulis CC, Krieter A, Gollnick H, Mischke D, Orfanos CE: Progressive differentiation of human sebocytes in vitro is characterized by increasing cell size and altering antigen expression and is regulated by culture duration and retinoids. Experimental Dermatology 3: 151-60, 1994.

Zouboulis CC, Seltmann H, Neitzel H, Orfanos CE: Establishment and characterization of an immortalized human sebaceous gland cell line (SZ95). Journal of Investigative Dermatology 113: 1011-20, 1999.

Zouboulis CC, Xia L, Akamatsu H, Seltmann H, Fritsch M, Hornemann S, et al: The human sebocyte culture model provides new insights into development and management of seborrhoea and acne. Dermatology 196: 21-31, 1998.

Zouboulis CC, Xia LQ, Detmar M, Bogdanoff B, Giannakopoulos G, Gollnick H, et al: Culture of human sebocytes and markers of sebocytic differentiation in vitro. Skin Pharmacology 4: 74-83, 1991.

Zouboulis CC, Zia L, Korge B, Gollnick H, Orfanos CE. Cultivation of Human Sebocytes in vitro: Cell Characterization and Influence of Synthetic Retinoids. In: Retinoids: 10 Years On, edited by J.-H. Saurat Basel, Karger, 1991, p. pp.254-273.

VITA

Amanda Marie Nelson

Education Ph.D – Molecular Medicine Pennsylvania State University College of Medicine, Hershey PA 2001-2007

Bachelor of Science 1996-2000 Purdue University, West Lafayette, IN

Teaching Experience

Pennsylvania State University Graduate School Teaching Certificate, TWT Certificate 2007

Human Anatomy and Physiology Laboratory Teaching Assistant 2006 Messiah College, Pennsylvania

Molecular Genetics and Cell Biology Tutor 2004-2005 Pennsylvania State University College of Medicine, Hershey PA

Honors/Awards

La Roche-Posay Laboratoire Pharmaceutique The North American Foundation Research Award 2007-2008

Induction into Phi Beta Kappa 1999

Invited Lectures

Advances in Acne Research Symposium 2007 Society of Investigative Dermatology and Galderma R&D, Los Angeles, CA.

“Insights into Retinoid Actions in Sebaceous Glands.” Galderma R&D, France. 2006

Publications

AM Nelson, KL Gilliland, W Zhao, A Zaenglein, W Lui, DM Thiboutot. Lipocalin 2/NGAL up-regulation by 13-cis retinoic acid induces apoptosis in human skin and SEB-1 sebocytes. In preparation.

SB Clarke, AM Nelson, RE George and DM Thiboutot. Pharmacologic Modulation of Sebaceous Gland Activity: Mechanisms and Clinical Applications. Dermatologic Clinics. April 2007 Vol 25, Issue 2

AM Nelson and DM Thiboutot. Biology of Sebaceous Glands. Fitzpatrick’s Dermatology in General Medicine: Seventh Edition. In press. (October 2007).

AM Nelson, KL Gilliland, Z cong, and DM Thiboutot. 13-cis retinoic acid induces apoptosis and cell cycle arrest in human SEB-1 sebocytes. J Invest Dermatol 126(10):2178-89, 2006

NR Trivedi, Z Cong, AM Nelson, AJ Albert, LL Rosamilia, S Savarajah, KL Gilliland, W Liu, DT Mauger, RA Gabbay, and DM Thiboutot. Peroxisome Proliferator-Activated Receptors Increase Human Sebum Production. J Invest Dermatol 126(9):2002-9, 2006.