Molecular Mechanism of LIGHT and SARM Focusedon the Transcriptional Activity-mediatedSignal Transduction

著者 引地 由紀子year 2017その他のタイトル 転写因子活性を介したシグナル伝達に注目した

LIGHTとSARMの分子メカニズム学位授与大学 筑波大学 (University of Tsukuba)学位授与年度 2016報告番号 12102甲第8125号URL http://hdl.handle.net/2241/00147663

Molecular Mechanism of LIGHT and SARM Focused on the Transcriptional Activity-mediated

Signal Transduction

January 2017

Yukiko HIKICHI

Molecular Mechanism of LIGHT and SARM Focused on the Transcriptional Activity-mediated

Signal Transduction

A Dissertation Submitted to

the Graduate School of Life and Environmental Sciences,

the University of Tsukuba

in Partial Fulfillment of the Requirements

for the Degree of Doctor of Philosophy in Biological Science

(Doctoral Program in Biological Sciences)

Yukiko HIKICHI

i

Table of Contents

Abbreviations............................................................................................................................... 1

Abstract ....................................................................................................................................... 5

General Introduction ................................................................................................................... 8

Figures .................................................................................................................................... 13

Part I ......................................................................................................................................... 18

Abstract .................................................................................................................................. 19

Introduction ........................................................................................................................... 20

Material and Methods ............................................................................................................ 22

Results .................................................................................................................................... 25

Discussion ............................................................................................................................... 28

Figures .................................................................................................................................... 32

Part II ........................................................................................................................................ 44

Abstract .................................................................................................................................. 45

Introduction ........................................................................................................................... 46

Material and Methods ............................................................................................................ 48

Results .................................................................................................................................... 54

Discussions ............................................................................................................................. 58

Tables ..................................................................................................................................... 63

Figures .................................................................................................................................... 71

General Discussion ..................................................................................................................... 92

Figures .................................................................................................................................. 101

Acknowledgements .................................................................................................................. 105

References ................................................................................................................................ 107

1

Abbreviations

2

AR androgen receptor

Ara-C 1--darabino-furanosyl cytosine

bp base pair

BPE bovine pituitary extract

BPH benign prostatic hyperplasia

BrdU 5-bromo-29-deoxyuridine

cDNA complementary deoxyribonucleic acid

COX5B cytochrome c oxidase subunit Vb

DAXX death-domain associated protein

DHT dihydrotestosterone

ELISA enzyme-linked immunosorbent assay

ER estrogen receptor

FBS fetal bovine serum

FHL2 four and a half LIM domains 2

FI fluorescence intensity

FL full-length

FLNA filamin A, alpha

GRIP1 GR-interacting protein 1

h hour

HVEM herpesvirus entry mediator

IGF insulin-like growth factor

JNK c-Jun N-terminal kinase

LIGHT homologous to lymphotoxins, shows inducible expression, and competes with

herpes simplex virus glycoprotein D for HVEM, a receptor expressed by T

lymphocytes

LOH loss of heterozygosity

3

LT lymphotoxin

LTR lymphotoxin-beta receptor

mAb monoclonal antibody

min minute

MMTV mouse mammary tumor virus

mRNA messenger ribonucleic acid

M2H mammalian two- hybrid

NCOA1 nuclear receptor coactivator 1

NFB nuclear factor kappa B

PAK6 p21protein-activated kinase 6

PBS phosphate-buffered saline

PIAS1 protein inhibitor of activated STAT, 1

PIAS3 protein inhibitor of activated STAT, 3

PIAS4 protein inhibitor of activated STAT, 4

PNRC proline-rich nuclear receptor coactivator 2

PrECs prostate epithelial cells

PSA prostate-specific antigen

RLU relative light unit

RMS rhabdomyosarcoma

rRNA ribosomal ribonucleic acid

RT-PCR reverse transcription polymerase chain reaction

SARM(s） selective androgen receptor modulator(s）

SEAP secreted alkaline phosphatase

S.E.M standard error of mean

SENP1 SUMO1/sentrin specific peptidase 1

SERM elective estrogen receptor modulator

4

SkMCs skeletal muscle cells

SM smooth muscle

TAF1 TAF1 RNA polymerase II, TATA box binding protein associated factor, 250kDa

TEF-1 transcriptional enhancer factor -1

TNF tumor necrosis factor

TNFR TNF receptor

TPA 12-o-tetracanoylphorbol-13-acetate

Ts testosterone

rRNA ribosomal ribonucleic acid

5

Abstract

6

Cell signaling is a kind of molecular mechanisms to transmit extracellular stimuli to

intracellular machineries. It is also known as signal transduction and includes cellular responses to

hormones, cytokines, neuro transmitters, and autacoids via corresponding receptors and they elicit a

variety of biological responses. Transcriptional activity-mediated signaling is one of major

components of cell signaling. It alters or modifies cellular functions through induction or

suppression of gene expression and its disruption or dysfunction causes various diseases including

cancers. Therefore, the understanding of cell signaling has attracted the interest of many researchers

in biomedical field. In this study, I have studied on the signal transduction mechanisms of tumor

necrosis factor (TNF) receptor and androgen receptor (AR); and explored novel target molecules for

anticancer agents.

In the first part, I have studied on the cell signaling which relates with the effect of LIGHT on

human rhabdomyosarcoma cell line RD. LIGHT is a member of the TNF superfamily, which plays

multiple roles in the development of immune system, and its receptors have been identified to be

lymphotoxin (LT) receptor (LTR) and the herpesvirus entry mediator (HVEM)/TR2. TNF is

known to induce two distinct signaling pathways, apoptosis via caspase activation and nuclear

factor kappa B (NFB) activation. LIGHT decreased the number of viable RD cells, modulated cell

cycle distribution, induced clear morphological change with concomitant expression of smooth

muscle (SM) -actin mRNA, and stimulated NFB-dependent transcriptional activity and

chemokine production. LT12, another TNF family ligand for LTR, had similar effects on RD

cells but TNF and LT did not. These results indicate that LIGHT suppresses growth of RD cells

and induces transformation into smooth muscle through LTR.

In the second part, I have studied on the cell signaling which underlie tissue selectivity of

TSAA-291, a steroidal antiandrogen which was previously used for benign prostatic hyperplasia

(BPH) treatment. AR is a member of the nuclear hormone receptor superfamily that is activated by

binding with either of endogenous androgens, testosterone (Ts), or dihydrotestosterone (DHT). It

elicits diverse range of biological changes and its signaling is also involved in the development of

7

tumors in prostate and some other tissues. Selective androgen receptor modulator (SARM) is a class

of AR ligands that specifically bind to AR and display either agonistic or antagonistic effects

depending on target organs. I have identified that TSAA-291 exhibited partial AR agonist activity

in in vitro reporter assay and in vivo pharmacological studies. Daily administration of TSAA-291

increased the weight of levator ani muscle without increasing the weight of the prostate and seminal

vesicle in castrated mice. Comprehensive cofactor recruitment analysis of AR using TSAA-291 and

endogenous AR ligands clarified that 12 of 112 cofactors including the protein inhibitor of activated

STAT, one (PIAS1) were differently recruited to AR in the presence of TSAA-291 or endogenous

androgens. These results indicate that TSAA-291 has a tissue-specific effect, as characteristic of

SARM. Moreover, my comprehensive analysis suggested that binding with ligand induces specific

conformation changes of AR, which modulates its surface topology and protein-protein interaction

with cofactors, such as PIAS1, and leads to tissue-specific transcriptional gene regulation.

Based on two studies, I propose that LIGHT and SARM(TSAA-291）can lead to a novel and

feasible therapeutic approach for rhabdomyosarcoma or hormone-related disease including cancer.

These would be typical example for the application of basic studies on the transcriptional

activity-mediated signaling to drug discovery.

8

General Introduction

9

As a part of mechanisms to maintain homeostasis of life, cells receive and respond to inputs

from their environment, such as nutrients, osmotic pressure, humoral factors, extracellular matrices

and cell-cell contacts. They are integrated or coordinated to regulate a variety of cellular

phenomena, such as gene expression, protein synthesis, metabolism, proliferation, morphogenesis,

differentiation and programmed cell death (Martin, 2003; Dangsheng, 2012). Cell signaling is a

cellular mechanism to receive, amplify and transmit external input at cell membrane, cytoplasm, or

nucleus. In that sense, cell signaling is also referred to as signal transduction. Molecular component

of this process include various type of proteins, such as receptors, enzymes, transcriptional factors,

transporters and structural proteins. Regardless of the nature of initiating signal, the selectivity of

cellular responses is primarily determined by the receptors which recognize corresponding signaling

molecules so called ligands. On the other hand, molecular events downstream of receptors overlap

each other. Therefore, it is still incompletely understood how the cells direct the traffic of highly

complicated signaling to exert precisely controlled vital phenomena. It is well established that

dysregulation of cell signaling causes many diseases such as cancer (Martin, 2003). In that sense,

understanding of cell signaling mechanism under various disease states would offer important

information for the development of novel therapeutic strategies. Among numerous signaling

pathways of cells, I focused attention to tumor necrosis factor (TNF) receptor and androgen

receptor (AR) signaling which are involved in tumorigenesis or severe cancer-related symptoms.

The direct relationship or crosstalk between TNF and AR signaling is still controversial, however, it

is suggested that androgen actions on TNF-induced signaling pathways. TNF binding to trimeric

TNFR1 leads to assembly of complex I, which then results in activation of nuclear factor kappa B

(NF-κB), c-Jun N-terminal kinase (JNK), and/or the caspase cascade, depending on cellular context.

All of these three TNF downstream signaling pathways can be regulated by androgen (Bosscher

KD et al., 2006).

Members of the TNF superfamily, including original TNF ligands, play important roles in cell

activation, proliferation, differentiation, apoptosis, immunoglobulin class switch, immune evasion,

10

and immune suppression (Smith et al., 1994; Aggarwal and Natarajan, 1996; Baker and Reddy,

1996; Locksley et al., 2001). They exert their biological functions via interaction with

transmembrane receptors, comprising the TNF receptor family. Ligand-receptor binding causes a

conformational change in the receptor, which triggers subsequent signaling cascade (Purves et al.,

2001). LIGHT is a member of the TNF ligand superfamily and it was first identified from database

based on structural similarity (Hikichi et al., Patent No. US6235878). It contains 240 amino acids

with an N-terminal cytosolic domain of 37 residues that precedes a stretch of 22 hydrophobic

residues characteristic of a type II transmembrane protein (Mauri et al., 1998; Figure 1A). LIGHT

showed 20%-30% structural identity to the other TNF ligands including FasL and TNF (Mauri et

al., 1998; Figure 1B). LIGHT binds two known as TNF receptor family, lymphotoxin (LT)-

receptor (LTR) and the herpesvirus entry mediator (HVEM)/TR2 (Mauri et al., 1998; Crowe et al.,

1994). As previous various studies have shown, LIGHT played a role in the immune modulation

(Zhai et al., 1998; Celik et al., 2009) and had a potential value in breast and colorectal cancer

therapy (Zhai et al., 1998; Harrop et al., 1998). Furthermore, it is known that LIGHT stimulated

NF-B transcriptional activity in the human primary hepatocytes and inhibits the apoptotic caspase

cascade induced by TNFMatsui et al., 2002; Figure 1C. Rhabdomyosarcoma (RMS) is the most

common soft tissue sarcoma in children originating from immature cells and it has a long-term

survival rate in children of only 50%-70% (Parham 1994; Pappo et al., 1995), and no standard

treatment for adults with recurrent/refractory RMS has yet been established (Sasada et al., 2016).

As shown in part I of this thesis, I have identified that LIGHT, a member of the TNF superfamily,

induces morphological changes and delays proliferation in the human RMS cell line RD (Hikichi et

al., 2001). Since LIGHT is a pleiotropic molecule initiating diverse biological functions depending

on the receptor expression profiles of the target cells, understanding mechanism underlying

LIGHT-mediated anti-tumor activity is valuable for discovering a novel therapeutic strategy for

RMS.

Another key signal pathway of my interest is AR signaling. Androgen is all-inclusive term for

11

group of hormones that play a role in male traits and reproductive activity. Testosterone (Ts) and

dihydrotestosterone (DHT), the major endogenous ligands for AR, are involved in a variety of

physiologic functions in human such as maintaining sexual function, germ cell development, fat

distribution, maintenance of muscle mass/strength, and central nervous systems. It has been known

that androgen deficiency in elder men is the most prevalent disorder of AR signaling (Bhasin et al.,

2006). Age-associated loss of muscle mass/strength and physical functional limitation increases a

risk of many chronic illnesses, such as chronic obstructive lung disease, congestive heart failure,

and cancer cachexia (Bassey et al,, 1992; Baumgartner et al., 1999; Baumgartner et al., 2000;

Melton et al., 2000; Roy et al,, 2002; Srinath and Dobs, 2014). These symptoms may improve with

Ts treatment, but various adverse effects such as prostate or latent prostate cancer would associate it

(Gooren, 2003). To overcome these problems of Ts, alternate and promising candidates for anabolic

therapies have long been explored. Recently, it was revealed that existing therapeutic agents such as

selective androgen receptor modulator (SARM) have a function to promote anabolic therapies

(Bhasin et al., 2006; Narayanan et al., 2008; Negro-Vilar 1999). SARM is a class of AR ligands that

bind AR and is known to display tissue-selective activation of androgenic signaling (Bhasin et al.,

2006; Narayanan et al., 2008). SARM can exert anabolic effects on the skeletal muscle and bone

without dose-limiting and Ts-like adverse effects. Furthermore, SARM has an anti-tumor activity,

especially on hormone-related cancer (Narayanan et al., 2014; Chisamore et al., 2016). Taken

together, there is an enormous promise for anabolic therapies with SARM that can improve physical

function and achieve anticancer effect without any adverse events.

While the physiological action and anticancer effect of SARM have been clarified, an

underlying mechanism for its tissue selectivity is still controversial. Narayanan et al. (2008)

reported that SARM and endogenous androgens activate several distinct signal pathways. But the

reason why only SARM is able to show tissue-specific transcriptional regulation and selective

biological effect remains unclear. I hypothesized that the tissue selectivity of SARM might be

achieved by SARM-specific patterns of cofactor recruitment to AR. To confirm my hypothesis, I

12

have constructed a comprehensive mammalian two-hybrid (M2H) system with full-length (FL)

human AR and a total of 112 cofactors in 293T cells. With this system, I have assessed the effects

of SARM and endogenous androgens on the pattern of cofactor recruitment to AR (Figure 2). As

shown in part II of this thesis, it was revealed for the first time that PIAS1 is key cofactor to explain

the tissue selectivity of SARM (Hikichi et al., 2015).

Here I summarize my findings on the cell signaling of extracellular and intracellular

receptor-ligand mediated systems. In part I, I have discovered that LIGHT suppresses growth and

induces transformation of RD cancer cells into smooth muscle cells. In part II, I have clarified that

TSAA-291, used for BPH treatment, is SARM and suggest that TSAA-291 may induce different

conformational changes of the AR. Its profile in cofactor recruitment, such as PIAS1, supports

tissue-specific activity of TSAA-291. Further studies targeting LIGHT and SARM are expected to

lead discovery of effective medicines for cancer or hormone-related disease.

13

Figures

Figure 1. Conceptual diagram and the sequence of LIGHT and alignment with TNF

Superfamily (A) Sequence of LIGHT. Amino-acid sequence deduced from the cDNA sequence.

The asterisk (*) indicates the position of the predicted N-glycosylation site. (B) Alignment of

LIGHT with TNF-related ligands. Sequences were aligned with ClustalW (Pam250 matrix)

(Macvector). This alignment excludes ligands for CD27, CD30, and Ox-40. The asterisks (*), the

contact residues in LT for TNFR60; bars, the -strands that form the scaffold in LT as

designated by convention. Homology regions are boxed with identical residues shaded.

(C) Schematic illustration for relationship between TNFligands and their receptors showed the

known interactions between members of TNF ligand family and TNF receptor family, and adaptor

proteins and signaling pathways associated with those receptors.

14

Figure 1A, B

Acytoplasmic tail

MEESVVRPSVFVVDGQTDIPFTRLGRSHRRQSCSVARVGLGLLLLLMGAGLAVQGWFLLQ

transmembrane

60

LHWRLGEMVTRLPDGPAGSWEQLIQERRSHEVNPAAHLTGANSSLTGSGGPLLWETQLGL 120

AFLRGLSYHDGALVVTKAGYYYIYSKVQLGGVGCPLGLASTITHGLYKRTPRYPEELELL 180

VSQQSPCGRATSSSRVWWDSSFLGGVVHLEAGEEVVVRVLDERLVRLRDGTRSYFGAFMV 240

receptor binding domain

*

B

LIGHT E V N P AA HL T G AH S S L - - - - - - - - - - - - T G SGG P L L W ET Q L - - - GL A FL RG - L S Y H D G A - - L V V T KAG Y Y YLTα T L K P AA HL I G DP S - - - - - - - - - - - - - - - - KQWS L L W RAWT - - - DRA FL Q DGF S L SWWS - - L L V P T SG I Y FLTβ P G L P AA HL I G AP L K - - - - - - - - - - - - - - - - GQ G L GW ET T K - - - EQA FL T SGT Q F S D A EG - L A L PQ DG L Y YTNF S D K P VA HVV A HP Q - - - - - - - - - - - - - - - - A E G Q L QW L NRR - - - A NAL L A NGY E L R DWQ - - L V V P S EG L Y LTRAIL P Q R V AA H I T G T RG R S W T L SSPNSK W EKA L G R K I W S W ES S R - - SGH S FL S N - L HL R N G E - - L V I HE KG F Y YFASL E L R K VA HL T G KS N S - - - - - - - - - - - - - - - RS M P L EW EDT Y - - - G I V L L SG - V KY K K GG - - L V I NE T G L Y FCD40L W P Q I AA HV I S EA S S - - - - - - - - - - - - - - K T T S V L QW A E KG - - - Y Y TMS NNL V T L E N G KQ - L T Y KRQG L Y Y41BBL R Q GM FA QLV A QNV L - - - - - - - - - - - - - - L I DG P L S W Y S DPGL - A GV SL T GGL S Y K E DTK EL Y YA KAG Y Y Y

LIGHT I Y S KV Q L GGV GC P - - - - - L GL AS T I T HG L Y K R T P R Y P E EL EL L V S QQS P - - - C Q RA T SS S R - - - - V WWD S

LTα V Y S QV V F S G KA Y S P - - KAT S S P L Y L AH EV Q L F SSQ Y P F HV P L L S S QK - - - - - - HV Y P GL Q E - - - - PWL N S

LTβ L YC L V G T RG RA P PGGGDPQGRS V T L RS S L Y RA GGA Y G P GT P EL L L EGA E T V T PV L D P A RRQG YG P LWY T S

TNF I Y S QV L F KGQ GC P - - - - - - S T HV L L T H T I S R I A V S Y Q T KV NL L S A I KS PC QR E T P E G A EAK - - - - PWT E P

TRAIL I Y S Q T Y F R F Q E E I K EN - - T KNDKQHVQ Y I Y K Y T S - Y P D F I L L H KS A RW - - - - - SC WS KDA E - - - - YG L Y S

FASL V Y S KV Y F RGQ SC N - - - - - - - - WL P L S H KV YM RWS K Y P Q DL VMME G KM - - - - - - MS Y C T T GQ - - - - MWA R S

CD40L I YA QV T F C S N R E A S - - - - - SQAP F I AS L C L K S PGR F E R I L L RAA N T HS - - - - - SA K P CGQQ - - - - - - - S -

41BBL V F F Q L E L R RV V A G E - - - - GSGS V S L AL HL Q P L RS A AGA AAL AL T V DL P - - - - - PA S S EARN - - - - - - SA F

LIGHT S F L G G V V H L E A G E E V V V R V L D E R - - - - L V R L R D G T R S Y F G A F M VLTα M Y H G A A F Q L T Q G D Q L S T H T D G I P - - - - H L V L S P S - T V F F G A F A LLTβ V G F G G L V Q L R R G E R V Y V N I S H P D - - - - M V D F A R G - K T F F G A V M V GTNF I Y L G G V F Q L E K G B R L S A E I N R P D - - - - Y L D F A E S G Q V Y F G I I A LTRAIL I Y Q G G I F E L K E N D R I F V S V T N E H - - - - L I D M D H E - A S F F G A F L V GFASL S Y L G A V F N L T S A D H L Y V N V S E L S - - - - L V N F E E S - Q T F F G L Y K LCD40L I H L G G V F E L Q P G A S V F V N V T D P S - - - - Q V S H G T G - F T S F G L L K L41BBL G F Q G R L L H L S A G Q R L G V H L E T E A R A R H A W Q L T Q G - A T V L G L F R V T P E I P A G L P S P R S E

F G H* * *

C * D* E***** * * * * *

A A’ A” B’ B** * * * **

15

Figure 1C

TNF LT

p55TNFR

p75TNFR

LT12 LIGHT

HVEM/TR2

LTR

Ligands

Receptors

FasL

Fas

TRAIL

DR4,5

Deathdomain

NF-B activation

Cell death

FADD

TRADD

TRAF

16

Figure 2. Conceptual diagram to identify key cofactors using M2H systems The tissue

selectivity of SARM is the result of the different patterns of cofactor recruitment to AR induced by

ligands. The cofactors that show significant different profiles between SARM and endogenous

androgens were selected by comprehensive M2H systems.

17

Endogenous androgens（DHT）

SARM（TSAA-291）

Figure 2

GAL4 binding site Luciferase

GAL4

AR cofactor

VP16

DHT

+++++

recruit

VP16

cofactor

GAL4 binding site Luciferase

GAL4

SARM（TSAA-291）

+

VP16

cofactorrecruit

18

Part I

LIGHT, a Member of the TNF Superfamily, Induces Morphological Changes and Delays

Proliferation in the Human Rhabdomyosarcoma Cell Line RD

19

Abstract

LIGHT is a member of the TNF superfamily, which binds two known receptors, LTR and

HVEM/TR2. I investigated the effects of LIGHT on the RMS cell line RD. LIGHT delayed cell

proliferation and induced morphological changes of the cells. These effects were not shown by

other TNF family ligands such as TNF and LT, which induced the transcriptional activity of

NF-B and NF-B-responsible chemokine productions in the same manner as did LIGHT. LT12,

another TNF family ligand for LTR, was shown to have similar activities in RD cells as LIGHT.

Both LIGHT and LT12 induced the expression of muscle-specific genes such as smooth muscle

(SM) -actin, while TNF and LT did not. These findings indicate that LIGHT may be a novel

inducer of RD cell differentiation associated with SM -actin expression through the LTR.

20

Introduction

RMS is the most common soft tissue sarcoma in children originating from immature cells and

it has a long-term survival rate in children of only 50-70% (Pappo et al., 1995; Parham, 1994).

These tumors resemble primitive skeletal muscle-forming cells in appearance and are highly

aggressive, suggesting that RMS may arise from skeletal muscle cells that are arrested along the

normal myogenic pathway to maturation (Pappo et al., 1995; Parham, 1994). Several studies of the

molecular basis in RMS have shown RMS to be associated with loss of heterozygosity (LOH) at the

11p15 locus, which effects the expression of insulin-like growth factor (IGF), which is a growth

factor of RMS (Scrable et al., 1987; El-Badry et al., 1990). RMS is characterized by the expression

of several muscle-specific markers such as myogenic-promoting transcription factor MyoD.

Although the expression of such factors typically correlates with myogenic differentiation, RMS

fails to undergo terminal differentiation into skeletal muscle (Parham, 1994). Members of the TNF

family play important roles in cell activation, proliferation, differentiation, apoptosis,

immunoglobulin class switch, immune evasion, and immune suppression (Smith et al., 1994;

Aggarwal and Natarajan., 1996; Baker and Reddy., 1996; Locksley et al., 2001). Recently, a new

member of the TNF family, designated as LIGHT, was identified as a cellular ligand for both

HVEM, also designated as TR2, and LTR (Zhai et al., 1998; Mauri et al., 1998; Tamada et al.,

2000; Rooney et al., 2000). LIGHT mRNA is highly expressed in splenocytes, activated PBL, CD8+

tumor-infiltrating lymphocytes, granulocytes, and monocytes but not in the thymus or the tumor

cells examined to date (Zhai et al., 1998). LTR is prominent on epithelial cells but absent in T and

B lymphocytes (Crowe et al., 1994); it is also involved in the development of peripheral lymph

nodes and spleen architecture (Ettinger et al., 1996; Koni et al., 1997). LTR is the receptor for

LIGHT as well as membrane-bound LT12 trimers (Mauri et al., 1998; Crowe et al., 1994), and

HVEM/TR2 has been shown to be the receptor for LIGHT, LT, and herpes simplex virus envelope

glycoprotein D (Kwon et al., 1997; Marsters et al., 1997). Recently, LIGHT was reported to be able

21

to induce apoptosis in several tumor cells (Zhai et al., 1998), and has a CD28-independent

costimulatory activity leading to T-cell growth and differentiation (Tamada et al., 2000). Thus,

LIGHT is a pleiotropic molecule initiating diverse biological functions depending on the receptor

expression profiles of the target cells. I found that LIGHT delayed cell proliferation and induced

morphological changes in a human RMS cell line, RD, associated with the expression of smooth

muscle (SM) -actin mRNA. LT12, another TNF family ligand for LTR, had similar effects on

RD cells, but TNF and LT did not. Therefore, LIGHT may be a novel inducer of morphological

changes on RMS by expressing cytoskeletal protein(s) through LTR.

22

Material and Methods

Materials and cell culture

Human RMS cell line RD was purchased from the American Type Culture Collection

(Rockville, MD). The cells were maintained in DMEM containing 10% FBS and 1 mM sodium

pyruvate. Recombinant human TNF, LT, and LT12 were purchased from R&D Systems

(Abingdon, UK). TPA was purchased from Wako Pure Chemical Industries (Osaka, Japan).

RANTES and IL-8 productions were measured using Quantikine kits (R&D Systems). Soluble

LIGHT proteins were produced and purified as follows: a FL human LIGHT cDNA was obtained

from a SUPERSCRIPT human liver cDNA library (Gibco-BRL, MD) using the GENETRAPPER

cloning system (Gibco-BRL) with the following probes and colony-PCR primers:

5'-AGGTCAACCCAGCAGCGCATCTCA-3' and 5'-CACCATCACCCACGGCCTCTACAAG-3',

for cDNA cloning, 5'-AGGTCAACCCAGCAGCGCATCTCACAGG-3' and

5'-CAAATTAAACCGGGTACCATCACGCAGTCG-3', for colony-PCR. The extracellular region

(encoding Ile84 to Val240) of LIGHT was amplified from the cDNA by PCR using the following

primers: 5'-GAATTCGATACAAGAGCGAAGGTCTCACGAGGTC-3' and

5'-AAATCTAGATCCTTCCTTCACACCATGAAAGCCCC-3'. The PCR product was digested

with EcoRI and XbaI, and ligated into the EcoRI–XbaI site of pFLAG-CMV-1 expression vector

(Eastman Chemical Company, NY). The preprotrypsin-FLAG-LIGHT DNA region in the vector

was further digested with SacI and XbaI, and ligated into the SacI–XbaI site of pFAST-BAC1

vector (Gibco-BRL). The plasmid was infected into SF9 insect cells to generate the recombinant

LIGHT proteins according to the procedure of Bac-to-Bac Baculovirus Expression System

(Gibco-BRL). The FLAG-tagged soluble LIGHT proteins were purified with an anti-FLAG mAb

affinity column, ANTI-FLAG M2-Agarose (Sigma, MO).

23

Semiquantitative RT-PCR analysis

Semiquantitative PCR amplification using primers for either human SM -actin mRNA

(5'-GCTCACGGAGGCACCCCTGAA-3' and 5'-CTGATAGGACATTGTTAGCAT-3'), human

-actin mRNA (5'-TGACGGGGTCACCCACACTGTGCCCATCTA-3' and

5'-CTAGAAGCATTTGCGGTGGACGATGGAGGG-3'), human myogenin mRNA

(5'-CCGTGGGCGTGTAAGGTGTG-3' and 5'-ACGATGGAGGTGAGGGAGTGC-3'),

or human Id-1 mRNA (5'-CGAGGTGGTGCGCTGTCTGTCT-3' and

5'-TCGCCGTTGAGGGTGCTGAG-3') was performed using the Advantage 2 PCR Enzyme

Systems (Clontech, CA) to amplify the 591-bp fragment for SM -actin cDNA, 540-bp fragment

for -actin cDNA, 416-bp fragment for myogenin cDNA, or 315 bp fragment for Id-1 cDNA,

respectively. I used 18S rRNA cDNA detection as an internal control.

Cell proliferation assay

Cell proliferation assays were performed using the cell proliferation ELISA, BrdU kit (Roche,

NY). Briefly, after the RD cells (2500 cells/well) were cultured in 96-well plates with each ligand

for 4 days, the cells were labeled by adding BrdU solution for 1.5 h (final concentration of 10 mM),

before being washed and fixed. The cells were treated with an anti-BrdU-POD antibody for 1.5 h.

After washing, the cells were treated with POD substrate solution and the absorbance at 450 nm/690

nm was measured with a plate reader. In a separate procedure, the RD cells were plated in duplicate

in 25 cm2 flasks with each ligand for 6 days. The living cell number was determined using the

trypan blue exclusion method.

Flow cytometry

RD cells (1 × 106 cells) were cultured with or without LIGHT for 72 h, fixed in 70% ethanol

for 24 h at -20°C, and washed with PBS. The cells were incubated with 2 mg/ml RNaseA for 20

min at 37°C, and stained with propidium iodide for 30 min at room temperature. The DNA content

24

of the cells was determined using a flow cytometry FACScan (Becton–Dickinson, Germany).

NF-B transcriptional activity assay

The NF-B transcriptional activity was determined using a Mercury Pathway Profiling

System (Clontech). Briefly, RD cells (1 × 106 cells) were seeded in a 12-well plate for 1 day and

were then transfected with 0.5 mg pNF-B-SEAP vector using 1.5 ml FuGENE6 reagent (Roche)

for 20 h. After the cells were exposed with or without each ligand for the indicated times, the SEAP

activity in the culture media was determined using a Great EscAPe chemiluminescence detection kit

(Clontech).

Western blot and immunocytochemistry

Western blot analysis was performed using a ProtoBlotII AP System (Promega, Germany).

Briefly, the RD cells (7 × 104 cells) were cultured with reagents for 6 days and lysed with a high

salt buffer (0.6 M KCl in 10 mM Tris, pH 7.5, and protease inhibitors). The soluble proteins

collected by centrifugation from the lysates were separated by 2-15% SDS-polyacrylamide gels,

and then transferred onto nitrocellulose membranes. The membranes were then incubated with a

mouse anti-skeletal myosin monoclonal antibody (mAb) MY-32 (Zymed, CA) at 1:100 dilutions for

1 h at room temperature. After washing, the blots were incubated for 1 h with an anti-mouse IgG (H

+ L) AP Conjugate (Promega) at 1:5000 dilutions before being exposed to the substrate solution.

For immunocytochemistry, the RD cells (6 × 103 cells) were cultured with various reagents for 6

days and then fixed in ethanol:acetone (1:1) for 30 min at 220°C before being blocked with PBS

containing 1% BSA. The cells were then incubated with MY-32 mAb for 1 h, following

HRP-anti-mouse IgG F(ab)'2 (ICN/Cappel, OH) as the 2nd antibody against MY-32 mAbs.

After washing, diaminobenzidine (Sigma, MO) was used as a peroxidase substrate to visualize the

stained cells.

25

Results

Growth Delay Induced by LIGHT

As shown in Figure 3A, I found that LIGHT at a concentration above 6 ng/ml had a

substantial growth inhibitory effect on a human embryonal RMS cell line RD. Though LIGHT

inhibited the total viable RD cell number by up to about one-third during the first 6 days compared

with the control growth (Figure 3B), re-exposure of the cells with an excess amount of LIGHT (up

to 50 ng/ml) did not result in complete suppression of growth. Furthermore, when the cell cycle

distribution of the LIGHT-treated cells was compared with that of control cells by flow cytometry,

the G0/G1 percentage of the cells increased only slightly, from 49 to 60%, even after 6 days of

treatment (Figure 3C). Therefore, the effect of LIGHT on the RD cells seemed to involve a delay of

cell proliferation rather than a frank arrest of the growth. A similar inhibitory effect was observed

with LT LT12, while the other TNF family ligands, such as TNF and LT, had only a weak

suppressive effect on RD cell proliferation (20% inhibition according to the trypan blue exclusion

method) (Figure 3B). Interestingly, LT synergistically stimulated the inhibitory effect of LT12,

as shown in Figure 3.

Morphological Changes Caused by LIGHT

Observations with phase-contrast microscopy showed that the first 2 days of treatment with

50 ng/ml of LIGHT induced no morphological changes in the RD cells compared with control cells

(Figure 4A). After the 5th day of culture, however, more than half of the cells treated with LIGHT

showed an evident increase in elongated cytoplasm hypertrophy and formed multinucleated

myotube-like cells (Figure 4B). Cells treated with LT12 showed similar morphological changes

to those treated with LIGHT (Figure 4C). LT synergistically promoted the changes induced by

LT12, whereas LT alone did not (Figure 4D, E). As shown in Figure 4F, TPA, a known inducer

of RD cell differentiation (Aguanno et al., 1990; Bouche et al., 1993), induced a different

26

morphology that seen in the control or LIGHT-treated cells. This suggests that LIGHT causes a

qualitatively different induction of changes in the morphology of RD cells than does TPA.

NF-B Activation and Chemokine Productions

To examine whether LIGHT regulates the transcriptional activity of NF-B in the manner of

other ligands, I monitored the capacity of each ligand that induced the expression of representative

NF-B-responsive proteins, such as IL-8 or RANTES in RD cells. As shown in Figure 5, all ligands

induced comparative biological responses including IL-8 and RANTES productions in the cells

cultured for 3 days. The productivity of each chemokine by LIGHT was similar to that by LT12.

Further experiments were performed to determine whether this responsiveness was related to the

transcriptional activity of NF-B. I used a SEAP reporter gene construct driven by four tandem

copies of Kappa () enhancer element (B4;8) as an NF-B responsive sequence in the Mercury

Pathway Profiling Systems (Clontech), as described in materials and methods, and found that all

ligands induced comparative levels of SEAP in RD cells. By contrast, the activity of NF-B

induced by LIGHT was weaker than that induced by other ligands (Figure 6). Thus, even though

engagement of TNF or LT induces normal transcriptional activity of NF-B in RD cells, the

activated NF-B was not able to transactivate the growth delay or morphological conversions of RD

cells.

Induction of Smooth Muscle -Actin Gene Expression

To investigate possible roles of LIGHT in regulating morphological changes of RD cells, I

examined apparent alterations of muscle regulatory genes and their protein products. By

semiquantitative RT-PCR, I analyzed the accumulation of muscle-specific gene transcripts such as

SM -actin, myogenin, and Id-1. LIGHT and LT12 induced expression of SM -actin mRNA,

whereas other ligands such as TNF or LT did not. The expression of -actin, myogenin, and Id-1

were not changed in LIGHT-treated RD cells (Figure 7). It has been reported that RD cells treated

27

with TPA increase their expression of cytoskeletal proteins such as skeletal muscle myosin and

skeletal muscle -actin, increase their binding of 125I--bungarotoxin, and increase the

phosphorylation of several proteins including -PKC (Aguanno et al., 1990; Bouche et al., 1993).

While TPA did not modify the expression of SM -actin or myogenin in the present study, it

induced skeletal myosin expression in RD cells according to a combined analysis using both

immunocytochemistry (Figure 8A) and Western blot analysis (Figure 8B) with an anti-skeletal

muscle myosin mAb MY-32. Conversely, no positive cells or skeletal myosin proteins were found

in RD cells treated with LIGHT according to this analysis (Figure 8A, B). Immunocytochemistry

and Western blot analysis using an anti-smooth muscle myosin heavy chain mAb F126.16D9

(Biocytex, France) showed that none of the ligands tested changed the expression level of smooth

muscle myosin proteins (data not shown).

28

Discussion

LIGHT is a member of the TNF superfamily, which binds two known receptors, LTR and

HVEM/TR2. My results indicated that LIGHT plays an important role in the differentiation process

of human RMS cell line RD. LIGHT caused a marked morphological change in the RD cells

characterized by growth delay with elongated cytoplasm hypertrophy (Figures 3 and 4) and

increased SM -actin expression (Figure 7). Furthermore, my results demonstrated that such

changes in the RD cells are not caused by TNF or LT, except for LT12, which is another TNF

family ligand specifically bound to LTR. Both TNF and LT stimulated the activation of NF-B

and the production of NF-B-responsive chemokine in the manner of both LIGHT and LT12,

indicating that LTR signaling may directly activate differentiation pathways in RD cells (Figures 5

and 6).

RD is an embryonal RMS resembling normal fetal skeletal muscle in morphology and it

expresses several muscle-specific genes such as the myogenic-promoting transcription factor MyoD.

RD cells are capable of only a limited and abortive spontaneous myogenic differentiation, probably

because they lack functional p53 (Germani et al., 1994) and have homozygous gene deletion of

p16ink4 gene (Urashima et al., 1999). Recent evidence suggests that transfection with a

temperature-sensitive p16 mutant (E119G) gene in RD cells, under a permissive culture condition,

reduced CDK6-associated kinase activity, induced G1 growth arrest, and induced morphological

change coupled with the expression of myogenin and myosin light chain genes (Urashima et al.,

1999). However, Knudsen et al. (1998) reported that ectopic expression of p21cip1, p16ink4, or

p27kip1 in RD cells caused the cell growth arrest, but not detectable expression of myogenic

markers such as myosin heavy chain, indicating that these activities alone are not sufficient for RD

cells to differentiate. In any case, the failure of RMS to undergo terminal differentiation into

skeletal muscle may be one mechanism by which these cells gain the growth advantage necessary

for tumor formation. While the growth inhibitory effect of LIGHT on RD cells was obvious at a

29

concentration of 6 ng/ml (Figure 3A), at 50 ng/ml LIGHT, which is an excessive amount, the

inhibition increased only marginally, suppressing the cell number by one-third compared with the

control (Figure 3B). When I compared the cell cycle distribution in LIGHT-treated cells with that

for control cells, the G0/G1 phase percentage was only slightly increased, from 49% to 60% (Figure

3C). The flow cytometry analysis was performed after excluding the differentiated cells with

elongated cytoplasm hypertrophy and multinucleated myotube-like cells. Thus, there may be the

possibility that the differentiated cells preferentially arrest cell growth in the culture system.

Therefore, the reduction in cell number by LIGHT may be due to morphological conversion of part

of the RD cells rather than a typical growth arrest or cytotoxicity. Since I did not investigate the

growth characteristics of the morphologically changed cells, more experiments are needed to better

understand these phenomena.

TPA is known to be a differentiation reagent of RD cells accompanied with inducible

expression of muscle-specific genes such as skeletal muscle -actin and myosin light chain genes.

Both LIGHT- and LT12-treated RD cells markedly expressed SM -actin gene (Figure 7), but

TPA did not. By contrast, TPA slightly reduced the expression of Id-1 mRNA, a negative regulator

of MyoD, but LIGHT and LT12 did not (Figure 7). Conversely, skeletal myosin expression was

increased in TPA-treated RD cells, but not in those treated by LIGHT or LT12 (Figure 8). These

results suggest that the differentiation state induced by LIGHT and LT12 is completely different

from that induced by TPA. The SM -actin gene is well-known to be activated during the early

stage of embryonic cardiovascular development, switched off in late stage heart tissue, and replaced

by cardiac and skeletal -actins. It also appears during vascular development, and becomes the

most abundant protein in adult vascular smooth muscle cells. Tissue-specific expression of SM

-actin is required for the principal force-generating capacity of the vascular smooth muscle cells.

Therefore, LIGHT might activate the transcriptional machinery necessary for transdifferentiation

from a skeletal- to a smooth-muscle lineage through LTR.

Signaling by TNF family members is initiated by an aggregation of specific cell surface

30

receptors. TNF, LT, LT12, and LIGHT exhibit distinct but overlapping patterns of binding to

four cognate receptors: TNF receptor type 1 (TNFR1), TNF receptor type 2 (TNFR2), LTR, and

HVEM/TR2, which together define a core group within the larger TNF superfamily. TNF binds

two receptors, TNFR1 and TNFR2, and LT binds TNFR1, TNFR2, and HVEM/TR2. LT12,

predominantly expresses in activated T cells and specifically binds LTR. Although LIGHT binds

both LTR and HVEM/TR2, the cross-utilization of the receptors suggests functional redundancy

of the ligand. When I examined the expression levels of these four receptors via RT-PCR, I

observed that they all expressed in the cells. However, the expression level of each receptor mRNA

was quite different; the lowest being HVEM/TR2 (40 copies/ng total RNA), with the expression

levels of LTR, TNFR1, and TNFR2 being 125-, 215-, and 45-fold higher than that of HVEM/TR2,

respectively (data not shown). Rooney et al. suggested that LTR is necessary and sufficient for

LIGHT-mediated apoptosis in a human adenocarcinoma cell line HT29 (Rooney et al., 2000).

However, Zhai et al. reported that both LTR and HVEM/TR2 are involved cooperatively in the

LIGHT-mediated killing of tumor cells, including HT29 cells (Zhai et al., 1998). Furthermore, it

has been reported that only the activation of LTR by cross-linking with an anti- LTR mAb could

induce growth arrest and chemokine production in A375 melanoma cells (Degli-Esposti et al.,

1997), although I did not observe these effects in the cells by LIGHT or LT12. This discrepancy

might reflect different ligand sensitivity for the cell line I used. Since I did not determine whether

the function of HVEM/TR2 is sufficient for ligand-mediated signal activation on RD cells, the

possibility of the phenotype conversion through the HVEM/TR2 needs to be studied in more detail.

RMS is the most frequent soft tissue malignancy in pediatric patients. It is known that several

distinct histological subtypes of RMS have been described: alveolar, embryonal, botryoid, and

undifferentiated (Pappo et al., 1995, Parham, 1994). Although I examined additional RMS cell lines,

Hs729, A673, and A-204, to confirm whether LIGHT can also induce differentiation in these cell

lines, I did not observe any effects in them (data not shown). Therefore, there might be certain RMS

cell types which have the sensitivity to LIGHT. Several studies using RD cells have been attempted

31

to regress the phenotype by evaluating agents that alter cellular growth as well as differentiation

both in in vitro and in vivo. Pyrimidine analogues such as GR-891 (Marchal et al., 1999) and Ara-C

(Crouch et al., 1993) are reported to induce growth inhibition and to stimulate the differentiation

processes and terminal myogenic differentiation of RD cells correlating with several differentiation

markers.

The present study may be the first to report morphological changes in RD cells induced by

the proteinaceous reagents LIGHT and LT12. It is hoped that further investigation of these

ligands and their potential role could lead to a novel and useful therapeutic approach for RMS.

32

Figures

Figure 3. Growth inhibitory activity of LIGHT on RD cells. (A) RD cells were treated with

varying concentrations of LIGHT (◆), TNF (■), LT (▲), LT12 (●), and LT plus LT1

2 (×) for 4 days, and OD 450 nm/690 nm was measured by a plate reader to determine the amount

of BrdU incorporation. (B) RD cells (105 cells/well) were cultured for 6 days with 50 ng/ml of each

ligand, and the viable cells in each well were counted after staining with trypan blue. (C) RD cells

cultured with or without 50 ng/ml LIGHT for 3 days were stained with propidium iodide for 30 min

at room temperature, and after establishing fractionated mononuclear cell populations, the DNA

content of the cells was determined by flow cytometry. FI, fluorescence intensity.

33

Figure 3

Via

ble

ce

ll n

um

be

r（x1

06 c

ells）

0

1

2

3

4

5

6

none LIGHT TNFα LTα LTα1β2 LTα plus

LTα1β2

0 200 400 600 800 1000

NoneG0/G1 : 49%

S : 25%G2/M : 26%

Ce

ll n

um

be

r

FI

0 200 400 600 800 1000

LIGHTG0/G1 : 60%

S : 18%G2/M : 22%

Ce

ll n

um

be

r

A4

50

/69

2

【ligand】 （ng/ml）

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

0 3.125 6.25 12.5 25 50

A B

C

FI

34

Figure 4. Morphology of each of the samples of ligand-treated RD cells. Each phase-contrast

image represents control RD cells (A), RD cells treated with 50 ng/ml LIGHT (B), LT12 (C),

LT (D), LT plus LT12 (E), or 100 mg/ml TPA (F) for 5 days. All images were acquired

at the same magnification using a × 10 objective lens.

35

Figure 4

36

Figure 5. Chemokine production of RD cells by LIGHT. RD cells were cultured with the

indicated concentrations of each ligand for 2 and 3 days. IL-8 and RANTES were measured by each

specific ELISA kit.

37

Figure 5

None, Light, TNFα, LTα plus LTα1β2, LTα, LTα1β2

8

7

6

5

4

3

2

1

0

RA

NT

ES

（ng

/ml）

0 10 50 50100

2day 3day【ligand】 （ng/ml）

9

8

7

6

5

4

3

2

1

0

I L

-8 （

ng/m

l）

0 10 50 501002day 3day

【ligand】 （ng/ml）

38

Figure 6. NF-B transcriptional activity of LIGHT on RD cells. RD cells transfected with 0.5

mg of pNF-B-SEAP vector (Clontech) were cultured with 50 ng/ml LIGHT (■), LT12 (●),

TNF (×), LT (-), LT plus LT12 (▲), or the control (◆) for the indicated times. After the

treatment, the SEAP activity in culture media was determined using a Great EscAPe

chemiluminescence detection kit.

39

Figure 6

20

18

16

14

12

10

8

6

4

2

03 4 5 6 7

Treatment time （h）

Flu

ore

sc

en

ce

ra

tio

40

Figure 7. Expression of muscle-specific genes in RD cells by LIGHT. After the RD cells were

cultured with 50 ng/ml of each ligand or 100 mg/ml of TPA for 6 days, RT-PCR was performed on

each sample of cells using primers specific for SM -actin, Id-1, myogenin, -actin, and rRNA. The

PCR products were resolved on 1% agarose gel containing ethidium bromide to stain the DNA.

41

Figure 7

42

Figure 8. Expression of skeletal muscle-specific myosin heavy chain proteins in TPA- and

LIGHT-treated RD cells. (A) RD cells plated onto coverslips were cultured with or without 50

ng/ml of each ligand or 100 mg/ml of TPA for 6 days. After being fixed, each treated sample of

cells was stained with an anti-skeletal muscle-specific myosin heavy chain mAb MY-32. All images

were acquired at the same magnification using a × 10 objective lens. (B) For Western blot

analysis, each ligand-treated sample of cells was cultured for 6 days and lysed, and for each a 50 mg

protein extract was used for Western blot analysis with MY-32 mAb.

43

Figure 8

44

Part II

Selective Androgen Receptor Modulator Activity of a Steroidal Antiandrogen TSAA-291

and its Cofactor Recruitment Profile

45

Abstract

SARM specifically binds to the AR and exert agonistic or antagonistic effects on target

organs. In this study, I investigated the SARM activity of TSAA-291, previously known as a

steroidal antiandrogen, in mice because TSAA-291 was found to possess partial AR agonist activity

in reporter assays. In addition, to clarify the mechanism underlying its tissue selectivity, I

performed comprehensive cofactor recruitment analysis of AR using TSAA-291 and DHT, an

endogenous androgen. The AR agonistic activity of TSAA-291 was more obvious in reporter assays

using skeletal muscle cells than in those using prostate cells. In castrated mice, TSAA-291

increased the weight of the levator ani muscle without increasing the weight of the prostate and

seminal vesicle. Comprehensive cofactor recruitment analysis via M2H methods using 293T cells

revealed that among a total of 112 cofactors, 12 cofactors including the PIAS1 were differently

recruited to androgen receptor in the presence of TSAA-291 and DHT. Further M2H analysis using

primary cells confirmed that TSAA-291 recruited PIAS1 to AR more potently in skeletal muscle

cells than in prostate cells. Prostate displayed higher PIAS1 expression than skeletal muscle. Forced

expression of the PIAS1 augmented the transcriptional activity of the AR, and silencing of PIAS1 by

siRNAs suppressed the secretion of PSA, an androgen responsive marker. My results demonstrate

that TSAA-291 has SARM activity and suggest that TSAA-291 may induce different

conformational changes of the AR and recruitment profiles of cofactors such as PIAS1 compared

with DHT to exert tissue-specific activity.

46

Introduction

Androgens, such as Ts and DHT, have diverse physiological actions in men such as the

maintenance of muscle mass/strength, sperm production, fat distribution, and sex drive (Mooradian

et al., 1987). Various symptoms such as loss of libido, decreases in muscle mass/strength, and mood

disorder in aging men with low blood Ts levels are improved by Ts treatment; however, Ts and

DHT are associated with various undesirable side effects (Gooren, 2003). To overcome the

limitations of Ts treatment, several strategies have been investigated. One major approach has been

the development of SARM. SARM is defined as compounds that specifically bind to the AR and

exert agonistic/antagonistic effects in a tissue-specific manner (Negro-Vilar, 1999). SARM

displaying tissue-selective anabolic effects in muscle and the bone without exerting adverse effects

on prostate or latent prostate cancer would be a useful therapeutic option for sarcopenia,

osteoporosis, muscle wasting associated with cancer cachexia, and aging-associated functional

limitations (Allan et al., 2007; Miner et al., 2007).

While the molecular mechanism underlying the tissue selectivity of SARM activity is unclear,

SARM activity may be understood in the context of the well-established concept of selective

estrogen receptor modulator (SERM) activity. Shang and Brown (2002) reported that cell type- and

promoter-specific differences in steroid receptor coactivator-1 recruitment play a critical role in

determining SERM function in the breast and uterus and offered a paradigm for understanding the

action of SERM in other important targets. The transcriptional activity of AR is modulated via its

interactions with cofactors, including coactivators that enhance AR activity and corepressors that

inhibit AR activity (Brady et al., 1999; Dotzlaw et al., 2002; Fu et al., 2000; Yang et al., 2002).

However, to the best of my knowledge, no comprehensive cofactor recruitment analysis of AR has

been performed with a M2H system using endogenous ligands and SARM.

TSAA-291 is a steroidal antiandrogen previously used for BPH in Japan. In the present study,

I investigated the SARM activity of TSAA-291 in mice because TSAA-291 was found to possess

47

partial AR agonist activity in reporter assays. In addition, to clarify the mechanism underlying its

tissue selectivity, I performed comprehensive cofactor recruitment analysis of AR using TSAA-291

and DHT.

48

Material and Methods

Reagents

TSAA-291 (16β-ethyl-17-hydroxyestr-4-en-3-one) and TSAA-272

(17-hydroxy-16-isopropyl-estr-4-en-3-one) were previously synthesized in my laboratory (Goto

et al., 1978).

Animals

Male Crj:ICR mice (5 weeks old) were purchased from Charles River Laboratories, Japan

(Yokohama, Japan). The animals were housed in a temperature-controlled room (23°C ± 2°C) with

a 12-h/12-h light/dark cycle and given ad libitum access to food (CE-2; CLEA Japan, Inc. Tokyo,

Japan) and water. All procedures were performed according to protocols approved by the

Institutional Animal Care and Use Committee of Pharmaceutical Research Division, Takeda

Pharmaceutical Company Limited.

In vivo pharmacokinetic study

Male Crj:ICR mice were randomly assigned to treatment groups and subsequently treated

with TSAA-291. Plasma was collected via heart puncture under anesthesia at the indicated time

points (each time point, n = 3), and the concentration of TSAA-291 was determined by

high-performance liquid chromatography.

In vivo effects of Selective Androgen Receptor Modulator compounds on androgen-dependent

organ weight

Male Crj:ICR mice were castrated at the beginning of the study. A group of sham-operated

male mice was also included as an intact control. The castrated animals were allocated into groups

of five animals according to body weight and subsequently treated with TSAA-291, TSAA-272, or

49

vehicle for the indicated period as described in Figure 2 legends. In addition, the intact animals

were treated with vehicle during the treatment period. The drugs were dissolved in 20% benzyl

benzoate:corn oil (20:80, vol/vol) and administered via daily subcutaneous injections. On the day

after the final treatment, the animals were weighed; the blood was then collected from the

abdominal aorta under anesthesia and the mice were euthanized by decapitation. The levator ani

muscle, prostate, and seminal vesicle were subsequently excised and weighed.

The effect of TSAA-291 on the sexual behavior in castrated mice

The effect of continuous treatment with TSAA-291 on the induction of sexual behavior in

castrated male mice. The castrated male mice were treated with TSAA-291 (3, 10, or 30 mg/kg,

subcutaneous) or vehicle for 11-14 days and were mated for 2 consecutive days with female mice in

the proestruos stage on a one-on-one basis (n = 5). The induction rates of pseudopregnancy (black

bar) and menstrual disorder (white bar) were confirmed by the observation of vaginal smears every

morning.ΨP < 0.05 versus vehicle by Fisher’s exact test.

Cell culture

Human embryonic kidney cells (293T), human breast tumor cells (BT-474), monkey kidney

fibroblasts (COS-7), and human prostate normal peripheral zone cells (RWPE-1) were purchased

from the American Type Culture Collection (ATCC, Manassas, VA, USA). PrECs and SkMCs

were purchased from Lonza Japan (Tokyo, Japan). 293T and COS-7 cells were maintained in

DMEM (Life Technologies, Inc., Carlsbad, CA, USA), containing 10% FBS (Lot. C10031, Life

Technologies, Inc.). BT-474 cells were maintained in RPMI1640 (Life Technologies, Inc.)

containing HEPES and 10%FBS. RWPE-1 cells were maintained in keratinocyte serum-free

medium containing EGF and BPE (Life Technologies, Inc.). PrECs and SkMCs were cultured in

special media, PrEGM™ BulletKit (Lonza) and SkGM™ BulletKit (Lonza), respectively,

according to the product description of the cells. All cell lines were maintained at 37°C in a 5%CO2

50

atmosphere. For androgen response experiments in M2H and reporter assays, to exclude the

influence of the steroids contained in the medium, 293T, COS-7, and BT-474 cells were cultured in

phenol red-free medium supplemented with charcoal dextran-stripped FBS (DCC-FBS) (Life

Technologies, Inc.) before use in the assay. PrECs and SkMCs were cultured in a medium without

BPE and EGF for 2 days to exclude the influence of the steroids contained in BPE before use in the

assay.

AR binding assays

Radiolabeled mibolerone (3 nM) and test compounds were added to a solution containing the

wild-type AR, and the mixture was incubated at 4C for 3 h. The bound and free forms of

mibolerone were separated by the dextran/charcoal method. The radioactivity of bound mibolerone

was measured, and the inhibitory rate of the test compound was calculated. The Ki value was

calculated as follows: Ki = IC50/[1 + (Ligand)/Kd]; (Ligand) = 3 × 10−9 M, Kd = 1.82 × 10−9 M.

Reporter assays

COS-7 cells

COS-7 cells were transfected with 2.5 μg of the FL hAR -pcDNA3.1 (hAR-pcDNA3.1) and

2.5 μg of pGL4.12/Mouse Mammary Tumor Virus (MMTV)-LTR (pGL4.12/MMTV-LTR) using

Lipofectamine™ 2000 (Life Technologies, Inc.), according to the manufacturer's instructions. Four

hours after transfection, the cells were plated in each well of a 96-well plate at a density of 1 × 104

cells/well, followed by treatment with DHT or the test compound for 4 h. The luciferase activity in

the cells was measured using a Bright-Glo Luciferase assay kit (Promega, Madison, WI, USA) and

a Wallac 1420 ARVO MX multi-label counter (Perkin Elmer, Waltham, MA, USA).

51

PrECs and SkMCs (primary cells)

PrECs and SkMCs were transfected with 1.25 μg of human AR-pcDNA3.1 (hAR-pcDNA3.1)

and 1.25 μg of human PSA2-Luc-pGL3 (hPSA2-Luc-GL3) using Lipofectamine™ LTX (Life

Technologies, Inc.) and PLUS reagent (Life Technologies, Inc.) according to the manufacturer’s

instructions. Six hours after transfection, the cells were seeded on 96-well poly-L-lysine–coated

white plates at a density of 2.0 × 104 cells/well. After overnight culture, DHT or TSAA-291 was

added and cultured for an additional 24 h. Luciferase activity was measured as described above.

The effect of PIAS1 on AR transactivation

In the reporter assays studying the effect of PIAS1 on AR transactivation, 293T cells were

prepared as previously described. 293T cells were transfected with 1 g of hAR-pcDNA3.1 and 1

g of 6kPSA-pGL3 together with the human PIAS1-pcDNA3.1 plasmid at the indicated

concentration using LipofectamineTM2000. The total plasmid amount was set as 3.6 g. Four hours

after transfection, the cells were plated in each well of a 96-well plate at a density of 1 × 104 cells /

well and cultured for 4 h. The cells were treated with 100 nM DHT or 100 nM TSAA-291 and

further cultured for 24 h. Luciferase activity was measured as previously described.

Mammalian Two-Hybrid assays

AR-cofactor interactions were evaluated by the Checkmate™ Mammalian Two-Hybrid

System (Promega) in 293T cells. The Gal4-DNA binding domain was fused with FL human AR

(hAR-pBIND vector), and the VP16 activation domain was fused to each of the cofactor genes

(cofactor-pACT vectors). The luciferase reporter plasmid with five copies of the GAL4 binding site

(pG5luc vector) was purchased from Promega. 293T cells were seeded in phenol red-free

DMEM-5% DCC-FBS medium on 6-well poly-L-lysine-coated plates at a density of 8.6 × 105

cells/well for 24 h before transfection. 293T cells were transfected with 1666 ng of hAR-pBIND,

1666 ng of cofactor-pACT, and 1666 ng of pG5luc vector using Lipofectamine™ 2000. Six hours

52

after transfection, the cells were harvested, suspended in DMEM-5% DCC-FBS medium, and

seeded on 96-well poly-L-lysine-coated white plates. After a 24-h culture, Ts, DHT, or the test

compound was added to each well. After an additional 24 h of culture, luciferase activity was

measured as described above.

For the M2H assays with RWPE-1 cells and SkMCs, were suspended in steroid-reduced

medium and seeded on 10-cm dishes at a density of 80% confluency for 2 days before transfection.

RWPE-1 cells and SkMCs were transfected with 5000 ng of hAR-pBIND, 5000 ng of

cofactor-pACT, and 5000 ng of pG5luc vectors by the same method. Six hours after transfection,

RWPE-1 cells and SkMCs were harvested, and the cell suspensions of RWPE-1 cells or SkMCs

were seeded on 96-well poly-L-lysine-coated white plates at a density of 5 × 104 or 2 × 104

cells/well, respectively. After a 24-h culture, Ts, DHT, or the test compound was added to each well

at 10 pM to 3 M. After an additional 24-h culture, luciferase activity was measured as described

above.

BT-474 PSA assays

Silencing using 15 nM siRNA was performed using LipofectamineTM RNAiMAX (Life

Technologies, Inc.) via the reverse transfection method according to the manufacturer’s protocol.

The following three types of PIAS1 siRNAs were used:

5′-CCUUACGACUUACAAGGAUUAGAUU-3′(siRNA#1),

5′-UAUUCACUUUCACACAAAGAUUGGG-3′(siRNA#2),

5′-GGUCCAGUUAAGUUUGU-3′(siRNA#3) (Life Technologies, Inc). The Silencer® Select

Negative Control No. 2 siRNA is used as a non-targeting negative control (Life Technologies, Inc).

Twenty-four hours after siRNA transfection, DHT (0.1, 1, 10, or 100 nM) was added. The amount

of PSA in the conditioned medium was determined using an ELISA kit (Dainippon Pharmaceutical,

Osaka Lapan).

53

Quantitative real-time PCR

The gene expression levels of cofactors were quantified by quantitative real-time PCR. Total

RNA of human normal prostate (lot No. 7060206) and skeletal muscle tissue (lot No. 8062503A)

was purchased from Clontech Laboratories (Mountain View, CA, USA). Each cDNA for these total

RNAs was synthesized in vitro using the SuperScript First-Strand Synthesis System (Life

Technologies, Inc). Quantification of cDNA (100 ng total RNA/µl) was performed using the

TaqMan® Low Density Array containing the ready-made primers and probes optimized for TaqMan

PCR (Life Technologies, Inc) with the ABI Prism 7900HT Fast Real-Time PCR System (Life

Technologies, Inc). The gene expression levels of PIAS1 in primary cells were quantified by

quantitative real-time PCR using the ready-made primers and probes optimized for PIAS1 (Life

Technologies, Inc). The total RNA of primary cells (PrECs, RWPE-1 cells, and SkMCs) was

isolated using an RNeasy Mini Kit (QIAGEN Japan, Tokyo, Japan). In this study, 18S rRNA was

used as the internal standard to normalize the amount of mRNA used for PCR. Data were analyzed

by the qualitative standard method using RQ Manager 1.2 software (Life Technologies, Inc).

Statistical analysis

Data are presented as the mean ± S.E.M.. Comparisons between two groups were performed

using Student’s t-test. Comparisons between multiple groups were performed using Dunnett’s t-test.

A P-value < 0.05 (two-tailed) was considered statistically significant.

54

Results

In vitro properties

TSAA-291 and TSAA-272, another steroidal compound with a similar SARM profile as

TSAA-291, had similar binding affinities for androgen receptor, with Ki values of approximately

1.4 and 1.2 nM, respectively, whereas the Ki value of DHT was 0.28 nM (Table 1).

In the reporter assays using COS-7 cells, TSAA-291, and TSAA-272 exhibited antagonist

activity in the presence of 100 nM DHT (Figure 9A), whereas TSAA-291 and TSAA-272 partially

stimulated the reporter, in the absence of DHT, by approximately 50% at maximum relative to the

maximal stimulation induced by 1 μM DHT (Figure 9A).

In the reporter assay in PrECs and SkMCs, which are derived from the prostate and muscle,

respectively, TSAA-291 induced maximum AR activation that was more similar to that of DHT in

SkMCs than in PrECs (Figure 9B). According to the clinical data of TSAA-291, the plasma level of

TSAA-291 after weekly intramuscular administration at 400 mg/kg for 12 weeks is approximately

100 nM (Drug Information). The agonist activity of TSAA-291 at 100 nM was 47.5% of that of

DHT at the same concentration in PrECs (Figure 9B), whereas the agonist activity of TSAA-291

was 87.6% of that of DHT in SkMCs (Figure 9B).

Based on these data, I reasoned that TSAA-291 may have SARM profile in vivo and studied

the SARM activity in mice.

Tissue specific effect of TSAA-291 in castrated mice

First, I monitored the plasma level of TSAA-291 after a single administration. Plasma

TSAA-291 levels were dose-dependently elevated with a subcutaneous dosing at 10 and 30 mg/kg

(Figure 10A). TSAA-291 treatment at 30 mg/kg was sufficient to maintain a plasma concentration

of 100 nM for approximately 15 h (Figure 10A). Daily subcutaneous administration of 10 and 30

mg/kg TSAA-291 markedly increased the weight of the levator ani muscle in a dose-dependent

55

manner in castrated mice (Figure 10B), and TSAA-291 at 30 mg/kg increased the levator ani

muscle weight to 78% of that in the intact animals (P < 0.001, Figure 10B). In contrast, no marked

weight increase in the prostate was observed after the treatment of castrated mice with TSAA-291

at doses of up to 30 mg/kg (Figure 10B). Similar results were obtained after the treatment of

castrated mice with TSAA-272 at a dose of 30 mg/kg (Figure 10C). Treatment with TSAA-291 at

doses of 3 to 30 mg/kg also showed tendency to induce libido and sexual behavior in castrated male

mice (Figure 11). These results indicate that TSAA-291 has tissue-specific effects and functions as

SARM.

Comprehensive cofactor recruitment analysis

The molecular mechanism of the tissue selectivity of SARM has not been completely

clarified. I hypothesized that the selectivity is the result of the different patterns of cofactor

recruitment to AR induced by ligands. I constructed a comprehensive M2H system with FL human

AR and a total of 112 cofactors in 293T cells (Figure 12A, Table 2) and assessed the effects of

TSAA-291, TSAA-272, Ts, and DHT on the cofactor recruitment to AR. These cofactors were

selected from the cofactors in literature information, not only related with androgen receptor but

also related with other nuclear receptors such as estrogen receptor, progesterone receptor, and

peroxisome proliferator-activated receptor. Fifty-seven cofactors did not bind to AR (no binding

type; Table 2, Figure 12B). Among the 55 cofactors that bound to AR, 43 cofactors exhibited no

difference of recruitment to AR in the presence of TSAA-291 and DHT (ligand-independent type;

Table 2, Figures 12C and 13). Typical examples of the cofactor recruitment with the no binding

type and ligand-independent type are shown in Figure 12B, C, respectively. The other 12 cofactors

(PIAS1, PIAS4, PIAS3, FLNA, FHL2, TAF1, NCOA1, PNRC, DAXX, SENP1, PAK6, and

COX5B) were recruited to AR by TSAA-291 and TSAA-272 in a different manner than that

observed for Ts and DHT (ligand-dependent type: Table 2, Figure 14). These results suggest that

56

AR may undergo a different conformational change when bound by SARM compared with

endogenous androgens.

PIAS1 is recruited to AR differentially by TSAA-291 and endogenous androgens

Among the 12 ligand-dependent cofactors, the recruitment of PIAS1 to AR exhibited

strikingly different patterns in the presence of SARM and endogenous androgens (Figure 14). The

dose-response curve of the recruitment of PIAS1 to AR after treatment with TSAA-291 and

TSAA-272 was right-shifted compared with that with endogenous androgens (Figure 14). The

levels of PIAS1 recruitment induced by TSAA-291 and TSAA-272 at 100 nM were 56 and 58%,

respectively, relative to that induced by DHT at 100 nM (Figure 14). I also examined the

recruitment of PIAS1 to AR in RWPE-1 cells and SkMCs, which are derived from the prostate and

skeletal muscle, respectively, and are supposed to possess the intracellular environments of the

respective target organs. TSAA-291 and TSAA-272 recruited PIAS1 to AR at lower concentrations

and to a larger extent in SkMCs than in RWPE-1 cells (Figure 15). PIAS1 recruitment to AR by

TSAA-291 and TSAA-272 at 100 nM represented 13.9 and 18.8%, respectively, of that induced by

DHT at 100 nM in RWPE-1 cells (Figure 15). On the contrary, in SkMCs, PIAS1 recruitment to

AR by TSAA-291 and TSAA-272 at 100 nM represented 102.2 and 92.4%, respectively, of that

induced by DHT at 100 nM (Figure 15). Other ligand-dependent type of cofactors such as PIAS3,

PIAS4, and SENP1 also exhibited different AR recruitment pattern by TSAA-291 in SkMCs and

RWPE-1 cells (Figure 15), although only PIAS1 was the cofactor that was most strikingly recruited

to AR by TSAA-291 at 100 nM in SkMCs (Figure 15). These results suggest that AR may undergo

a different conformational change when bound by SARM versus endogenous androgens and that

the different recruitment of cofactors such as PIAS1 to AR by SARM, and endogenous androgens

as well as its different cofactor recruitment to AR between target tissues may be responsible for the

tissue selectivity of SARM.

57

PIAS1 acts as a coactivator of AR

I confirmed the expression of PIAS1 in RWPE-1 prostate cells and SkMCs. Prostate cells

displayed higher PIAS1 expression than skeletal muscle cells (Figure 16A). In addition, I confirmed

the gene expression of several cofactors in prostate and skeletal muscle tissue. The expression level

of PIAS1 in prostate tissue was higher than that in skeletal muscle (Figure 16B), and the expression

level of COX5B in skeletal muscle was higher than that in prostate tissue (Figure 16B). Finally, I

investigated whether PIAS1 is a coactivator or corepressor of the AR in my reporter assay using

293T cells. Cotransfection of increasing amounts of the PIAS1 vector augmented the transcriptional

activity of the AR in the presence of DHT (Figure 17A). I also confirmed that the transfection of

PIAS1 enhanced the transcriptional activity of the AR in the presence of TSAA-291 (Figure 17A),

although the extent of the increase induced by TSAA-291 was lower than that induced by DHT

(Figure 17A). In addition, silencing of PIAS1 by siRNAs suppressed DHT-induced PSA secretion

in BT-474 cells (Figure 17B). The knockdown efficiency of PIAS1 mRNA levels by each PIAS1

siRNA was confirmed (Figure 17C). PIAS1 siRNA also inhibited the mRNA expression of PSA

(data not shown). These results indicate that PIAS1 acts as a coactivator of the AR. Cofactors such

as PIAS1 might contribute to the tissue selectivity of some SARM compounds.

58

Discussions

SARM is defined as tissue-selective androgen receptor ligands. Several groups recently

discovered a structurally distinct class of orally active nonsteroidal compounds that exert weaker

effects on the prostate while maintaining anabolic effects in muscle and the bone (Allan et al., 2007;

Miner et al., 2007; Vajda et al., 2009). In the present study, I first observed that TSAA-291, a

steroidal antiandrogen previously used to treat BPH in Japan, displayed partial agonist activity in

reporter assays using COS-7 cells (Figure 9A). In addition, I found that the maximum AR agonist

activity of TSAA-291 was almost comparable to that of DHT in the reporter assays using skeletal

muscle cells, whereas in prostate cells, the agonist activity of TSAA-291 was approximately half of

that of DHT (Figure 9B). Based on these findings, I reasoned the different in vitro activities in the

prostate and muscle cells may reflect the in vivo profile of TSAA-291. As expected, I observed that

the administration of TSAA-291 at 10 and 30 mg/kg markedly increased the weight of the levator

ani muscle without increasing that of the prostate and seminal vesicle in castrated mice,

demonstrating that TSAA-291 has SARM properties in mice (Figure 10B, C). In addition, I also

observed that the treatment of castrated male mice with TSAA-291 at 3-30 mg/kg exhibited

tendency to induce sexual behavior (Figure 11). The effective concentrations of TSAA-291 in

clinical situations, mouse models, and cell models are consistent. The plasma level of TSAA-291

after weekly intramuscular administration at 400 mg/kg for 12 weeks in patients is approximately

100 nM (Drug Information), while TSAA-291 administration at 30 mg/kg was sufficient to

maintain the plasma concentrations of 100 nM for approximately 15 h in mice (Figure 10A). A

cellular response in reporter and M2H assays was also observed for 100 nM TSAA-291(Figures 9B,

14, and 15).

The SARM activity of TSAA-291 may not be due to differences in tissue distribution because

it is reported that the 3H-TSAA-291 levels in the levator ani muscle and ventral prostate after the

treatment of rats with 3H-TSAA-291 were approximately identical (Sudo et al., 1981).

59

The molecular mechanisms underlying the tissue-specific activity of SARM are unclear. I

hypothesized that the tissue selectivity results from different AR conformational changes induced

by SARM and full agonists and that the different AR conformation could be revealed by differences

in cofactor recruitment profiles. To test these hypotheses, I performed comprehensive M2H cofactor

recruitment analysis. The results illustrated that the recruitment pattern of 12 cofactors to AR was

different between SARM and endogenous androgens (Figure 14). These data suggest that SARM

and endogenous androgens may induce different conformational changes of the AR, and the altered

AR conformation induced by the different ligands might facilitate the recognition of different DNA

motifs, leading to differences in gene transcription levels. Considering the lower affinity of

TSAA-291/272 than DHT/Ts for the AR, the other ligand-independent type cofactors such as

growth arrest and DNA damage-inducible gamma (GADD45G), growth arrest and DNA

damage-inducible alpha (GADD45A), nuclear transcription factor Y gamma (NFYC), peroxisome

proliferative activated receptor gamma coactivator 1 beta (PPARGC1B), spen family transcriptional

repressor (SPEN), and nuclear receptor subfamily 0 group B member 2 (NR0B2), which display

overlapped dose-response curves in M2H assays (Figure 12C and Table 2), might also contribute to

the differences in conformation of the AR.

The 12 cofactors displaying different recruitment profiles between endogenous androgens and

TSAA-291 in the present study were categorized as follows: PIAS family, post-translational

modifications-related molecules, cell cycle regulating protein, cell cytoskeleton-related protein, and

others. PIAS family genes reportedly interact with the DNA-binding domain of androgen receptor

(Moilanen et al., 1999) and regulate the receptor’s transcriptional activity (Gross et al., 2001).

Among them, recruitment profiles of PIAS1 to AR between endogenous androgens and

TSAA-291were significantly different in both M2H assay system using 293T cells and primary

cells (Figures 14 and 15), however, the effects of PIAS family proteins on AR-dependent

transcriptional activity are controversial. It has been reported that PIAS1 functions as an AR

coactivator in prostate cancer cells (Gross et al., 2001), whereas PIAS1 acts as a SUMO-E3 ligase

60

for AR and represses AR-dependent transactivation via sumoylation in U2OS human osteosarcoma

cells (Nishida and Yasuda, 2002). I confirmed that PIAS1 functioned as a coactivator of AR (Figure

17A, B). The more potent AR transactivation induced by TSAA-291 in the reporter assay using

skeletal muscle cells than observed using prostate cells may be because of greater PIAS1

recruitment to AR in response to TSAA-291 in skeletal muscle cells. The interaction between the

AR and PIAS1 as assessed using co-IP assays has recently been reported (Toropainen et al., 2015).

The interaction between the AR and PIAS1 in the presence of TSAA-291 and DHT as well as the

type of intracellular environment factors responsible for the different sensitivity in prostate and

skeletal muscle cells remains to be elucidated.

The differential expression levels of AR cofactor genes might also be responsible for a

potential mechanism of tissue specificity. Prostate cells exhibited higher PIAS1 expression than

skeletal muscle cells. The higher expression of PIAS1 and greater recruitment of PIAS1 to the AR

induced by DHT in prostate cells might explain the prostate-oriented activity of DHT. On the

contrary, the expression level of COX5B in skeletal muscle was higher than that in prostate tissue

(Figure 16B). The recruitment of COX5B to the AR in muscle cells but not prostate cells is not

different in the presence of TSAA-291 and DHT at 100 nM (Figure 14). Cofactors such as COX5B

might account for the agonist activity of the AR in skeletal muscle.

PIAS1 has the characteristics of a scaffold/matrix attachment region-binding protein (Bode et

al., 2000; Kipp et al., 2000). As scaffold attachment factors, PIAS family proteins may control the

modulatory functions of many proteins and mediate interaction between different signaling

pathways. I speculate that, at least in the case of TSAA-291, greater AR activation via higher

recruitment of PIAS1 and other cofactors to AR in skeletal muscle cells than in prostate cells may

lead to high anabolic/androgenic ratio profile in vivo. Further investigation on the functional role of

PIAS1 or PIAS family proteins in the tissue selectivity of SARM is required. In addition, proteomic

analysis of AR-bound coregulators in different cell lines could provide more insight into the

tissue-specific effects of SARM.

61

Another possible mechanism for the tissue selectivity of TSAA-291 is steroidal metabolism

in tissues. The activity of 5-reductase is essential for DHT production from Ts. 5-reductase is

present in the prostate but not in muscle (Yamana and Labrie, 2010). The localization of this

enzyme may be involved in the SARM profile of TSAA-291. A substantial portion of TSAA-291

may be converted to 5-DHTSAA-291, a reduced form of TSAA-291, by 5-reductase in the

prostate (Sudo et al., 1981), whereas TSAA-291 itself may be the dominant form in muscle. I

obtained data illustrating that 5-DHTSAA-291 possesses high AR binding affinity compared with

TSAA-291, markedly lower AR agonistic activity than TSAA-291 in reporter assays, and no effect

on the weight of the prostate and levator ani muscle in castrated mice (data not shown).

It was reported that the coactivator GR-interacting protein 1 (GRIP-1) does not completely

enhance SARM-mediated transcriptional activity but completely enhances full agonist-mediated

transcription (Miner et al., 2007). I also evaluated the recruitment of GRIP-1 by TSAA-291 and

endogenous androgens using my constructed M2H system. However, a clear difference between

each ligand was not observed. The reason for this is currently unclear; however, the discrepancy

may be explained by the different cell lines used for the M2H assay.

Reducing sex hormone levels by treatment with LHRH analogs is currently used in men to

treat hormone-responsive prostate cancer and in women to treat uterine fibroids, endometriosis, and

hormone receptor-positive premenopausal breast cancer. These treatments cause side effects such as

accelerated bone loss with an increased risk of skeletal fractures, loss of libido, sexual behavior,

depression, heart attacks, cardiovascular disease, or hot flashes (Harle et al., 2006; Higano, 2003;

Sharma and Muggia, 2013; Smith et al., 2001). My data suggest that TSAA-291 has anabolic

effects on skeletal muscle and induce sexual interest with limited harmful androgenic effects.

Combination treatment with TSAA-291 and LHRH analogs may represent a useful therapeutic

option for sex hormone-dependent diseases.

62

In conclusion, I found that TSAA-291, a steroidal antiandrogen previously used for BPH

patients in Japan, has SARM activity, and my comprehensive cofactor recruitment analysis suggests

that TSAA-291 may induce different conformational change of AR compared with DHT.

63

Tables

Table 1. Chemical structure and binding affinity to the AR. Radiolabeled mibolerone (3 nM)

and test compounds (TSAA-291, TSAA-272, and DHT) were added to a solution containing the

wild-type AR and incubated at 4C for 3 h. Bound and free forms of mibolerone were separated by

the dextran/charcoal method. The radioactivity of bound mibolerone was measured, and the

inhibitory rate of the test compound was calculated.

64

Table 1

TSAA-291 TSAA-272 DHT

Structure

Ki（nM）

1.4 1.3 0.28

65

Table 2. Cofactor list and recruitment to AR. The cofactors used for the M2H analysis and their

recruitment to AR are summarized. “No binding to AR” indicates that the cofactor did not bind to

AR. “Ligand independent” indicates that the cofactor was recruited to AR in a ligand-independent

manner. “Ligand dependent” indicates that the cofactor was recruited to AR in a ligand-dependent

manner.

66

Table 2

No binding to AR

Gene ID

amino-terminal enhancer of split AES

COP9 signalosome subunit 2 COPS2

amyloid beta precursor protein (cytoplasmic tail) binding protein 2 APPBP2

adaptor protein, phosphotyrosine interaction, PH domain and leucine zipper containing 1 APPL1

adaptor protein, phosphotyrosine interaction, PH domain and leucine zipper containing 2 APPL2

RAN, member RAS oncogene family RAN

nuclear receptor binding SET domain protein 1 NSD1

SWI/SNF related,matrix associated, actin dependent regulator of chromatin, subfamily d,member 3 SMARDC3

bromodomain containing 8 BRD8

calreticulin CALR

RNA binding motif protein 39 RBM39

cyclin D1 CCND1

anti-silencing function 1A histone chaperone ASF1A

C-terminal-binding protein CTBP

nuclear receptor coactivator 7 NCOA7

EF-hand calcium binding domain 6 EFCAB6

growth arrest and DNA-damage-inducible, beta GADD45B

gelsolin GSN

PSMC3 interacting protein PSMC3IP

histone deacetylase 1 HDAC1

histone deacetylase 3 HDAC3

high mobility group AT-hook 1 HMGA1

DDB1 and CUL4 associated factor 6 DCAF6

COP9 signalosome subunit 5 COPS5

ribosomal protein L7 RPL7

ligand dependent nuclear receptor corepressor LCOR

lipin 1 LPIN1

multiprotein bridging factor 1 MBF1

paired box 8 PAX8

67

Table 2

No binding to AR

Gene ID

mediator complex subunit 24 MED24

SNW domain containing 1 SNW1

integrin beta 3 binding protein (beta3-endonexin) ITGB3BP

PRKC, apoptosis, WT1, regulator PAWR

parkinson protein 7 PARK7

protein arginine methyltransferase 1 PRMT1

protein arginine methyltransferase 2 PRMT2

pre-mRNA processing factor 6 PRPF6

splicing factor proline/glutamine-rich SFPQ

proteasome (prosome, macropain) 26S subunit, ATPase, 3 PSMC3

protein tyrosine kinase 2 beta PTK2B

ring finger and CHY zinc finger domain containing 1, E3 ubiquitin protein ligase RCHY1

prohibitin 2 PHB2

squamous cell carcinoma antigen recognized by T cells 3 SART3

SET domain, bifurcated 1 SETDB1

nuclear receptor corepressor 2 NCOR2

speckle-type POZ protein SPOP

sex determining region Y SRY

SCAN domain containing 1 SCAD1

C1D nuclear receptor corepressor CID

supervillin SVIL

transcription factor 4 TCF4

K(lysine) acetyltransferase 5 KAT5

thyroid hormone receptor interactor 13 TRIP13

tubulin, beta class I TUBB

tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein, beta YWHAB

mediator complex subunit 17 MED17

X-ray repair complementing defective repair in Chinese hamster cells 5 XRCC5

68

Table 2

Ligand independent

Gene ID

transcriptional adaptor 3 ADA3

TATA element modulatory factor 1 ARA160

transforming growth factor beta 1 induced transcript 1) ARA55

phosphoprotein membrane anchor with glycosphingolipid microdomains 1 ) CBP

Cbp/p300-interacting transactivator, with Glu/Asp-rich carboxy-terminal domain, 1 CITED1

Cbp/p300-interacting transactivator, with Glu/Asp-rich carboxy-terminal domain, 2 CITED2

catenin (cadherin-associated protein), beta 1, 88kDa CTNNB1

mediator complex subunit 14 DRIP150

forkhead box O3 FOXO3

growth arrest and DNA-damage-inducible, alpha GADD45A

growth arrest and DNA-damage-inducible, gamma GADD45G

glutamate receptor interacting protein 1 GRIP1

LIM domain only 4 LMO4

proteasome (prosome, macropain) 26S subunit, ATPase, 4 PSMC4

K(lysine) acetyltransferase 7 KAT7

nuclear receptor corepressor 1 NCoR1

nuclear transcription factor Y, gamma NFYC

nemo-like kinase NLK

nuclear receptor subfamily 2, group F, member 2 NR2F2

K(lysine) acetyltransferase 2B CAT2B

SMAD family member 3 (SMAD3) SMAD3

peroxisome proliferative activated receptor, gamma, coactivator 1 beta PPARGC1B

protein inhibitor of activated STAT, 2 PIAS2

proline-rich nuclear receptor coactivator 1 PNRC1

receptor of activated protein kinase C 1 RACK1

nuclear receptor coactivator 6 NCOA6

69

Table 2

Ligand independent

Gene ID

nuclear receptor interacting protein 1 NRIP1

splicing factor 1 SF-1

spen family transcriptional repressor SPEN

nuclear receptor subfamily 0, group B, member 2 NR0B2

SIN3 transcription regulator family member A SIN3A

decapping mRNA 1A DCP1A

nuclear receptor coactivator 3 NCOA3

nuclear receptor subfamily 2, group C, member 2 NR2C2

mediator complex subunit 1 MED1

proteasome (prosome, macropain) 26S subunit, ATPase, 5 PSMC5

thyroid hormone receptor interactor 6 TRIP6

tumor susceptibility 101 TSG101

SAM pointed domain containing ETS transcription factor SPDEF

ubiquitin-conjugating enzyme E2I UBE2I

ubiquitin-like modifier activating enzyme 3 UBA3

ubiquitously-expressed, prefoldin-like chaperone UXT

zinc finger, MIZ-type containing 1 ZMIZ1

70

Table 2

Ligand dependent

Gene ID

Cytochrome c oxidase subunit Vb COX5B

death-domain associated protein DAXX

four and a half LIM domains 2 FHL2

filamin A, alpha FLNA

nuclear receptor coactivator 1 NCOA1

protein inhibitor of activated STAT, 1 PIAS1

protein inhibitor of activated STAT, 3 PIAS3

protein inhibitor of activated STAT, 4 PIAS4

proline-rich nuclear receptor coactivator 2 PNRC

SUMO1/sentrin specific peptidase 1 SENP1

P21 potein-activated kinase 6 PAK6

TAF1 RNA polymerase II,TATA box binding protein (TBP)-associated factor, 250kDa TAF1

71

Figures

Figure 9. In vitro AR transcriptional activity of SARM compounds. (A) The reporter assay

using COS-7 cells. Agonist activity (○) and antagonist activity (●). The agonistic activities are

shown relative to those of the 1 μM DHT treatment as 100%. The antagonistic activities by test

compound are shown relative to those of the 0.1 μM DHT treatment as 100%. Data are shown as

the mean ±S.E.M. (n = 3). (B) The reporter assay using human prostate epithelial and human

skeletal muscle cells. The concentration of DHT (●) or TSAA-291 (■) was from 0.1 pM to 1 μM.

The ordinate values are expressed as RLU, setting the value of the DMSO control as 1. Data are

shown as the mean ± S.E.M. (n = 3).

72

Figure 9

TSAA-291 TSAA-272

-9 -8 -6 -5 -4

100

7550

250

-25

% o

f c

on

tro

l

log[M]-7

125100

755025

0-25

% o

f c

on

tro

l

-9 -8 -6 -5 -4-7log[M]

100755025

0-25

% o

f c

on

tro

l

-10 -9 -7 -6 -5-8log[M]

DHT125

10

5

0

RL

U

-14 -10 -8 -6-12log[M]

15

6

2

0

RL

U

-14 -10 -8 -6-12log[M]

8

4

10PrEC SKMC

A

B

73

Figure 10. Tissue specific effect of TSAA-291 in castrated mice. (A) Pharmacokinetic data of

TSAA-291. Male mice were treated with 10 mg/kg (dotted line) or 30 mg/kg (solid line) of

TSAA-291 subcutaneously for 1, 2, 4, 7, or 24 h. The plasma concentration of TSAA-291 was

determined by HPLC. Data are shown as the mean ± S.E.M. (n = 3). (B) The effect of continuous

treatment with TSAA-291 on the weights of androgen-responsive tissues in castrated male mice.

The castrated male mice (n = 5) were treated with TSAA-291 (3, 10, or 30 mg/kg, subcutaneous) or

vehicle for two weeks. In addition, the intact animals were treated with vehicle during the treatment

period (n = 6). On the day after the final treatment, each tissue was excised and weighed. (C) The

effect of continuous treatment with TSAA-272 on the weights of androgen-responsive tissues in

castrated male mice. The castrated male mice (n = 5) were treated with TSAA-272 or TSAA-291

(30 mg/kg, subcutaneous) or vehicle for seven days. In addition, the intact animals were treated

with vehicle during the treatment period (n = 5). On the day after the final treatment, each tissue

was excised and weighed. Data are shown as the mean ± S.E.M. *P < 0.05, **P < 0.01, ***P <

0.001, versus vehicle by Dunnett’s t-test; ###P < 0.001 versus vehicle by Student’s t-test.

74

Figure 10

A

Time （hour）

co

nc

en

tra

tio

n （

mM）

10mg/kg1

0.1

0.010 10 20

30mg/kg

B levator ani

we

igh

t （

mg）

60 ###

****

50403020100

mg/kg

Castration Intact

103 30

ve

hic

le

ve

hic

le

prostate

14###

121086420

mg/kg

Castration Intact

we

igh

t （

mg）

103 30

ve

hic

le

ve

hic

le

seminal vesicle

160 ###140120100

806040200

mg/kg

Castration Intact

we

igh

t （

mg）

103 30

ve

hic

le

ve

hic

le

C prostate

16 ###

14121086420

Castration Intact

we

igh

t （

mg）

TS

AA

-272

TS

AA

-291

veh

icle

veh

icle

**

seminal vesicle

250 ###

200

150

100

50

0

Castration Intact

we

igh

t （

mg）

TS

AA

-272

TS

AA

-291

veh

icle

veh

icle

******

levator ani

60 ###

40

50

30

2010

0

Castration Intact

we

igh

t （

mg）

TS

AA

-272

TS

AA

-291

veh

icle

veh

icle

75

Figure 11. The effect of TSAA-291 on the sexual behavior in castrated mice. The effect of

continuous treatment with TSAA-291 on the induction of sexual behavior in castrated male mice.

The castrated male mice were treated with TSAA-291 (3, 10, or 30 mg/kg, subcutaneous) or vehicle

for 11-14 days and were mated for 2 consecutive days with female mice in the proestruos stage on a

one-on-one basis (n = 5). The induction rates of pseudopregnancy (black bar) and menstrual

disorder (white bar) were confirmed by the observation of vaginal smears every morning. ΨP < 0.05

versus vehicle by Fisher’s exact test.

76

Figure 11

Ψ

0

20

40

60

80

100

vehicle 3 10 30

TSAA-291 (mg/kg)

Castration Intact

rate

（%）

77

Figure 12. Typical examples of the cofactor recruitment with the no binding type and

ligand-independent type in M2H assays. (A) Schematic representation of M2H assays. (B) The

binding between AR and each cofactor was measured by M2H assays. 293T cells were transfected

with pBIND, pACT, and pG5luc vectors (vector control) or with pBIND-Id, pACT-MyoD and

pG5luc vectors (positive control) or with hAR-pBIND, cofactor-pACT, and pG5luc vectors. (C)

Typical examples of the cofactor recruitment with ligand-independent type (DHT, ● and

TSAA-291, ▲) were shown. (B, C) The ordinate values are expressed as RLU, setting the value of

the vector control as 1. Data are shown as the mean ± S.E.M. (n = 3).

78

Figure 12

A

Ligand

AR

GAL4

GAL4 binding site Luciferase

VP16

Cofactor

+++

B

ve

cto

r c

on

t.p

os

i.co

nt.

HD

AC

1P

AR

K7

AP

PL

2C

CN

D1

AP

PB

P2

CA

LR

HD

AC

3G

AD

D4

5A

GA

DD

45

GC

ITE

D2

No binding Binding

50 93

14

12

10

8

6

4

2

0

RL

U

CDRIP150

1.81.51.20.90.60.30.0

-14 -12 -10 -8 -6 -4log ［M］

RL

U

2.5

2.0

1.5

1.0

0.5

0.0-14 -12 -10 -8 -6 -4

GADD45G

log ［M］

RL

U

79

Figure 13. The ligand-independent cofactor analysis. 293T cells were transfected with

hAR-pBIND, cofactor-pACT, and pG5luc vectors using LipofectamineTM2000. After treatment with

varying concentrations of SARM compounds (TSAA-291, ▲) or endogenous androgens (DHT, ●)

for 24 h, luciferase activity in the cells was measured. Ordinate values are expressed as RLU,

setting the value of the DMSO control as 1. Data are shown as the mean ± S.E.M. (n = 3).

80

Figure 13

ADA3

-12 -10 -8 -60.0

0.2

0.4

0.6

0.8

1.0

log[M]

RL

U

ARA160

-12 -10 -8 -60.0

0.5

1.0

1.5

log[M]R

LU

ARA55

-12 -10 -8 -60.0

0.5

1.0

1.5

log[M]

RL

U

CBP

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

2.0

log[M]

RL

U

CITED1

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

log[M]

RL

U

CITED2

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

log[M]

RL

UCTNNB1

-14 -12 -10 -8 -6 -40

1

2

3

4

log[M]

RL

U

FOXO3

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

2.0

log[M]

RL

U

GADD45A

-14 -12 -10 -8 -6 -40

1

2

3

4

log[M]

RL

U

GRIP1

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

2.0

log[M]

RL

U

LMO4

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

2.0

log[M]

RL

U

PSMC4

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

2.0

log[M]

RL

U

81

Figure 13

NCoR1

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

log[M]

RL

U

NFYC

-14 -12 -10 -8 -6 -40

1

2

3

log[M]

RL

U

NLK

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

log[M]

RL

U

NR2F2

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

2.0

log[M]

RL

U

CAT2B

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

2.0

log[M]

RL

U

PPARGC1B

-14 -12 -10 -8 -6 -40

1

2

3

4

log[M]

RL

U

PIAS2

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

log[M]

RL

U

PNRC1

-12 -10 -8 -60

5

10

15

log[M]

RL

U

RACK1

-12 -10 -8 -60.0

0.5

1.0

1.5

2.0

2.5

log[M]

RL

U

NCOA6

-12 -10 -8 -60.0

0.5

1.0

1.5

log[M]

RL

U

SMAD3

-12 -10 -8 -60.0

0.5

1.0

1.5

2.0

2.5

log[M]

RL

U

ZMIZ1

-12 -10 -8 -60.0

0.2

0.4

0.6

0.8

1.0

log[M]

RL

U

82

Figure 13

NRIP1

-14 -12 -10 -8 -6 -40

2

4

6

8

log[M]

RL

U

SF-1

-12 -10 -8 -60.0

0.5

1.0

1.5

log[M]R

LU

SPEN

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

log[M]

RL

U

NR0B2

-14 -12 -10 -8 -6 -40

1

2

3

4

log[M]

RL

U

SIN3A

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

log[M]

RL

U

DCP1A

-12 -10 -8 -60.0

0.2

0.4

0.6

0.8

1.0

log[M]R

LU

NCOA3

-12 -10 -8 -60.0

0.5

1.0

1.5

2.0

log[M]

RL

U

NR2C2

-12 -10 -8 -60.0

0.5

1.0

1.5

log[M]

RL

U

MED1

-12 -10 -8 -60.0

0.5

1.0

1.5

log[M]

RL

U

PSMC5

-12 -10 -8 -60.0

0.5

1.0

1.5

log[M]

RL

U

TRIP6

-14 -12 -10 -8 -6 -40

1

2

3

4

log[M]

RL

U

TSG101

-12 -10 -8 -60.0

0.5

1.0

1.5

2.0

2.5

log[M]

RL

U

83

Figure 13

UBE2I

-12 -10 -8 -60.0

0.5

1.0

1.5

log[M]

RL

U

UBA3

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

2.0

2.5

log[M]

RL

U

UXT

-14 -12 -10 -8 -6 -40.0

0.5

1.0

1.5

2.0

log[M]

RL

U

ZMIZ1

-12 -10 -8 -60.0

0.2

0.4

0.6

0.8

1.0

log[M]

RL

U

SPDEF

-12 -10 -8 -60.0

0.5

1.0

1.5

log[M]

RL

U

84

Figure 14. Comprehensive cofactor recruitment analysis. 293T cells were transfected with

hAR-pBIND, cofactor-pACT, and pG5luc vectors using LipofectamineTM2000. After treatment

with varying concentrations of SARM compounds (TSAA-291, ▲ and TSAA-272, ▼ ) or

endogenous androgens (DHT, ● and Ts, ■) for 24 h, luciferase activity in the cells was

measured. Ordinate values are expressed as RLU, setting the value of the DMSO control as 1. Data

are shown as the mean ± S.E.M. (n = 3).

85

Figure 14

86

Figure 15. M2H analysis using primary cells. RWPE-1 cells and SkMCs were transfected with

hAR-pBIND, PIAS1-pACT, and pG5luc vectors using LipofectamineTM LTX PLUS. After

treatment with varying concentrations of SARM compounds (TSAA-291, ▲ and TSAA-272, ▼)

or endogenous androgens (DHT, ● and Ts, ■) for 24 h, luciferase activity in the cells was

measured. The ordinate values are expressed as RLU, setting the value of the DMSO control as 1.

Data are shown as the mean ± S.E.M. (n = 3).

87

Figure 15

88

Figure 16. The expression levels of cofactors. (A) PIAS1 expression levels were measured in

PrECs, RWPE-1 cells, and SkMCs by a quantitative real-time PCR method. 18S rRNA was used as

the internal standard to normalize the amount of mRNA. Data were analyzed by the qualitative

standard method and were shown relative to those of PrECs as 1. (B) The cofactors displaying

different recruitment profiles between DHT and TSAA-291 expression levels were measured in

normal prostate and skeletal muscle by a quantitative real-time PCR method. 18S rRNA was used

as the internal standard to normalize the amount of mRNA. Data were analyzed by the qualitative

standard method and were shown cofactor mRNA ratio against 18S rRNA

89

Figure 16

90

Figure 17. PIAS1 as a coactivator of AR. (A) The effect of PIAS1 overexpression on AR

transcriptional activity in the presence of DHT or TSAA-291. 293T cells were transfected with 1 μg

of hAR-pcDNA3.1 and 1 μg of 6kPSA-pGL3 together with PIAS1-pcDNA3.1 at the indicated

concentration using LipofectamineTM 2000. Total plasmid amount was set at 3.6 μg. 293T cells

were treated with DHT (●; 100 nM) or TSAA-291(■; 100 nM) for 24 h, and then, luciferase

activity was measured. The ordinate values are expressed as RLU, setting the value of the each

mock control (without PIAS1 plasmid at DHT_100nM or TSAA-291_100nM) as 1. Data are shown

as the mean ± S.E.M. (n = 3). (B) The effect of knockdown of PIAS1 by siRNA on PSA secretion in

BT-474 cells. Knockdown by 15 nM of siRNA (●; control siRNA , ▲; PIAS1 siRNA#1, ■;

PIAS1 siRNA#2, ◆; PIAS1 siRNA#3) was performed using LipofectamineTM RNAiMAX by the

reverse transfection method. Twenty-four hours after siRNA transfection, DHT (0.1, 1, 10, and 100

nM) was added. After 6 days of culture, the amount of PSA in the conditioned medium was

determined. (C) The knockdown efficiency was examined using the mRNA sample after 48 h of

transfection by the quantitative real-time PCR TaqMan method. 18S rRNA was used as the internal

standard to normalize the amount of mRNA. The ordinate values are shown relative to those of the

control siRNA treatment as 1. Data are shown as the mean ± S.E.M. (n = 3).

91

Figure 17

3A

2.5

2

1.5

1

0.5

00 0.4 0.8 1.2 1.6

RL

U

PIAS1 （μg）

30B

25

20

15

10

5

0

PS

A （

ng

/ml）

0 0.1 1 10 100

DHT （nM）

1.2C

1

0.8

0.6

0.4

0.2

0

PIA

S1

mR

NA

/ 1

8S

rR

NA

（F

old

of c

on

tro

l siR

NA）

92

General Discussion

93

In this study, I focused attention on the transcriptional-mediated cell signaling which alters or

modifies cellular functions through induction or suppression of gene expression. TNF and AR are

representative players of it and disruption or dysfunction of pathway components including them is

well known to cause various diseases including cancers. Huge efforts have been made to target

these molecules and pathways for new medication, but there still remain unidentified mechanisms

underlying them. Identification of anti-proliferative effect of LIGHT on RMS (part I) and SARM

profile of AR ligand TSAA-291 (part II) served as initial steps toward elucidation of novel

mechanisms related with TNF and AR signaling as discussed below.

In part I, I firstly identified that LIGHT binds with two known receptors, HVEM/TR2 and

LTR, and induces dramatic morphological change in a human RMS cell line, RD. The change in

the morphology was characterized further as growth delay with elongated cytoplasm hypertrophy

(Figures 3 and 4) and increased SM -actin expression (Figure 7). Interestingly, LT12, which is

specifically bound to LTR showed similar results, but TNF or LTdid not (Figures 3, 4, and 7).

Both LIGHT and LT12 stimulated the transcriptional activity of NF-B and the expression of

NF-B-responsive proteins, such as IL-8 and RANTES. Those effects were relatively weak in

compared with TNF or LT (Figures 5 and 6), indicating that the activated NF-B plays only a

minor role in the morphological change and growth delay of RD cells by LIGHT or LT12

treatment. My data showed that LIGHT treatment extremely induced SM -actin expression on RD

cells. To explain increased SM -actin expression, other transcriptional factor would need to be

explored (Figure 18). The transcriptional activation of the SM -actin gene is known to involve

distinct transcriptional control mechanisms in different cell types and developmental stages. MCAT

elements within promoter region is one of possible molecular basis for differential expression of

SM -actin, since transcriptional enhancer factor-1 (TEF-1) family members show distinct binding

patterns to MCAT among the cell types (Gan et al., 2007). TEF-1 and C/EBPβ transcriptional

activities are regulated by p38 MAP kinase (Ambrosino et al., 2006） and TNF activates p38

MAP kinase (Zhou et al., 2006）. Tissue-specific expression of SM -actin is required for the

94

principal force-generating capacity of the vascular smooth muscle cells. The SM -actin gene is

also well known to be activated in cardiac primordium, switched off at later stage, and replaced by

cardiac and skeletal -actins. In RD cells, the type of transcriptional factor involved in the

modulation of SM -actin expression and transdifferentiation from a skeletal to smooth muscle

lineage via LTR is still unclear at this moment. The application of knock-down experiment with

siRNAs for representative transcriptional factors might be effective in addressing this point.

Some groups have reported that LIGHT induces apoptosis through LTR alone (Rooney et al.,

2000) or both LTR and HVEM/TR2 (Zhai et al., 1998) in human colorectal cancer cell line HT29.

In this study, I examined growth inhibition activity of LIGHT on RD cells. LIGHT decreased the

number of viable RD cells up to one-third and slightly increased resting cells at G0/G1 phase (Figure

3). The complete growth inhibition and marked increment in sub G1 population by LIGHT

treatment were not observed as typical characteristics of apoptotic cells. Therefore, the effect of

LIGHT on RD cells seemed to involve a delay of cell proliferation rather than a cell death. Similar

results were obtained by LT12 treatment on RD cells. Common receptor for LIGHT and LT12

is LTR, so growth delay on RD cells by these ligands might be mediated through only LTR.

These discrepancies between cell lines used might be attributable to cellular characteristics, such as

expression levels receptors, sensitivity to ligands, and signal transduction machineries.

RMS is the most common soft tissue sarcoma in childhood and adolescence. In spite of

clinical efforts conducted in worldwide, outcomes for high risk patients of this disease have not

been improved and survival rates of relapsed or metastatic patients remain extremely low (Pappo et

al., 1999). Currently, there are 18 embryonal and 12 adult alveolar human RMS cell lines. They

differ in their origins, karyotypes, histology, and methods of validation (Hinson et al., 2013). I

examined additional RMS cell lines available, A673 and A204, to confirm whether LIGHT can also

induce differentiation with growth delay, but significant effects were not observed in these cell lines.

Since these cells lack expression of myogenin, MyoD, or desmin, they might not be RMS origin or

loose myogenic potential through passaging in culture condition (Morton and Potter, 1998). The

95

other cell line TE671, which was originally thought to be a medulloblastoma line, was later shown

to be a subclone of RD cells by cytogenetic analysis and DNA fingerprinting (Stratton et al., 1989).

TE671 would be valuable to confirm the effect of LIGHT on RMS in addition to RD cells. The

most important point of this study is to show efficiency of cytokines, LIGHT and LT12, for

growth suppression and transdifferentiation of RMS cell for the first time. This finding is expected

to be a significant base in exploring effective medication for RMS.

In part II, I demonstrated the SARM profile of TSAA-291 in mouse model. TSAA-291 is a

steroidal anti-androgen previously used to treat BPH in Japan. Subsequent analysis suggested to me

that divergent tissue specificity of TSAA-291 and endogenous androgens may stem from difference

in recruitment of cofactors, especially PIAS1, to AR. The result is consistent with the previous

finding that TSAA-291 has anabolic effects on muscle while having limited effects on the prostate

and seminal vesicles in immature castrated rats (Masuoka et al., 1979). Administration of S-40542,

novel non-steroidal SARM with clinical benefit against BPH, also decreased weight of prostate in

dose-dependent manner by repeated administration to BPH rat models (Nejishima et al., 2012). In

my study, administration of TSAA-291 at 10 and 30 mg/kg significantly increased the weight of the

levator ani muscle without increasing that of the prostate and the seminal vesicle in castrated mice.

The result was consistent with that of S-40542 and demonstrated that TSAA-291 has SARM

properties in mice (Figure 10). In addition to them, I also observed that the treatment of castrated

male mice with TSAA-291 at 3 to 30 mg/kg exhibited tendency to induce sexual behavior (Figure

11). In humans, plasma levels of TSAA-291 after weekly intramuscular administration at 400

mg/kg for 12 weeks is approximately 100 nM (Prescribing Information). A cellular response in

reporter and M2H assays was observed at 100 nM TSAA-291 (Figures 9, 14, and 15). Therefore,

there seems to be no discrepancy in effective concentrations of TSAA-291 between clinical

situations, mouse models, and cellular models.

In reporter assays with COS-7 cells, TSAA-291 displayed partial agonist activity (Figure 9A).

Similar results were obtained with 293T cells (data not shown). The maximum AR agonist activity

96

of TSAA-291 was comparable to that of DHT in skeletal muscle cells. On the other hand, it was

approximately half of DHT in prostate cells (Figure 9B). These discrepancies in in vitro activities in

the prostate and muscle primary cells may reflect the in vivo profile of TSAA-291, which increases

the weight of the levator ani muscle without having a stimulatory effect on the prostate. Moreover,

an appropriate balance between agonistic and antagonistic activities of compounds might be

necessary in exerting SARM profiles in vivo.

The authentic mechanism underlying tissue-specificity of SARM is still uncertain. I

hypothesized that it might result from difference in conformational change of AR induced by

SARM and endogenous androgens. Since profiling of cofactor recruitment one of feasible

approaches to confirm the hypothesis, I initially conducted comprehensive M2H cofactor

recruitment analysis using 293T cells. As a result, the recruitment pattern of 12 cofactors to AR was

different between SARM and endogenous androgens (Figure 14). These data suggest that SARM

and endogenous androgens may induce different conformational changes of AR to recruit different

cofactors. Whole AR molecule has not been successfully crystallized, but direct evidence for the

difference in conformation between SARM-bound AR and endogenous androgens-bound AR may

be obtained through protein engineering approaches. The 12 cofactors displaying different

recruitment profiles between TSAA-291 and endogenous androgens in this study were categorized

as follows: PIAS family, post-translational modifications-related molecules, cell cycle regulating

protein, cytoskeleton-related protein, and others. Among those12 cofactors, not only PIAS1, but

also PIAS3, PIAS4, and SENP1exhibited different recruitment pattern depending on ligands (Figure

14). Therefore, further M2H analysis using primary cells was conducted to identify key cofactor.

Interestingly, recruitment of PIAS1 in the presence of 100 nM SARM or endogenous androgens

were markedly different between prostate and muscle cells, but the other cofactors showed no

considerable differences (Figure 15). Intracellular environment might have some effect on these

cell-dependent discrepancies in cofactor recruitment. I confirmed that PIAS1 functioned as a

coactivator of AR (Figure 17). AR transactivation induced by TSAA-291 was detected in reporter

97

assay with skeletal muscle cells and it was more potent than that of prostate cells. It may be due to

intensive PIAS1 recruitment to AR in skeletal muscle cells in response to TSAA-291. Based on

these results, I presumed that PIAS1 is the most important cofactors to discuss the tissue selectivity

of SARM and it is intracellular environment factor responsible for the different sensitivity in

prostate and skeletal muscle cells. The discovery regarding the involvement of PIAS1 in SARM

tissue selectivity has biological significance and is very meaningful for development of novel

therapeutic agents.

PIAS family genes are reported to interact with the DNA-binding domain of AR (Moilanen et

al., 1999) and regulate the transcriptional activity for AR (Gross et al., 2001). However, the effects

of PIAS family proteins on AR-dependent transcription are controversial. PIAS1 functions as an

AR coactivator in prostate cancer cells (Gross et al., 2001). In contrast, PIAS1 binds p300 and

behaves as a coactivator or corepressor of the transcription factor c-Myb dependent on SUMO

status (Ledsaak et al., 2016). These inconsistencies may arise from distinct expression patterns of

cellular factors implicated in the regulation of AR-dependent transcription. PIAS1 has the

characteristics of a scaffold/matrix attachment region-binding protein (Bode et al., 2000; Kipp et al.,

2000). As a scaffold attachment factor, PIAS family proteins may control the modulatory functions

of many proteins and mediate crosstalk between different signaling pathways. In addition to PIAS1,

PIAS4 also exhibited the different recruitment pattern in the presence of TSAA-291 and

endogenous androgens (Figure 14). It was reported that PIAS4 represses the transcriptional activity

of the AR and inhibits AR by recruiting histone deacetylases, independent of its SUMO ligase

activity (Gross et al., 2004). Further investigation would be required for detailed understanding of

functional roles of PIAS1 or PIAS family proteins in tissue selectivity of SARM.

Some groups showed that different ligands induce distinct AR and estrogen receptor (ER)

conformations leading to their association with different coactivator peptides using combinatorial

peptide-phage display (Chang et al., 1999; Chang and McDonnell, 2002) and they may support my

hypothesis. The other SARMs, that are RTI-018 and RTI-001, had potency as agonist and changed

98

kinetics of transcriptional activation response. These SARM may cause structural change of AR to

lead association of SARM-AR with coactivator peptides which is distinct from DHT-AR complex

(Kazmin et al., 2006). These findings also support my hypothesis that conformational change of AR

induced by SARM lead association and recruitment of different co-regulators.

It was reported that orally active non-steroidal SARM inhibited growth of prostate and breast

tumors (Narayanan et al., 2014; Chisamore et al., 2016). MK-4541, known as a novel SARM,

exerts anti-androgenic activity in xenograft models of prostate cancer (Chisamore et al., 2016).

Enobosarm (GTx-024) is the most advanced SARM in clinical development. In the Phase II clinical

trial of GTx-024, the drug was demonstrated to exert anti-tumor activity in breast cancer patients,

such as inhibition of interaction between epithelial and mesenchymal stem cells and subsequent

invasion and metastasis (Narayanan et al., 2014). In this study, I did not examined whether

TSAA-291 itself shows anti-tumor activity in in vitro and in vivo or not, but non-steroidal SARM

synthesized exhibited in vitro growth inhibition activities in various cancer cell lines (data not

shown).

Structurally distinct classes of orally active non-steroidal SARMs which are currently under

development are resistant to 5α-reduction or aromatization. The resistance to these enzymes may be

the other possible mechanism for the tissue selectivity of SARM (Buijsman et al., 2005). The most

potent androgen for the prostate is DHT, which is formed by the 5α-reduction of Ts, and it is the

most abundant circulating androgen. Inhibition of 5α-reductase by finasteride, a 5α-reductase

inhibitor, leads to shrinkage of prostate size without any effect on muscle or bone mass. The result

indicates that lack of possible 5α-reduction in Ts may cause different response in the prostate from

the muscle and the bone (Wright et al., 1999). The expression of 5α-reductase gene is high in the

prostate and the skin but low in the bone and the muscle. This discrepancy can explain the

significance of DHT in the prostate and Ts in the muscle and the bone. SARMs cannot interact with

5α-reductase and this is considered to be a one of logical explanations for their tissue selectivity. A

substantial portion of TSAA-291 was reported to be converted into 5-DHTSAA-291, a reduced

99

form of TSAA-291, in the prostate (Sudo et al., 1981), but TSAA-291 would be dominant in

skeletal muscle due to low expression of 5-reductase. A series of my data was aligned well with

these findings. AR binding affinity of 5-DHTSAA-291 is high in compared with TSAA-291 but

AR agonistic activity is lower than TSAA-291 in reporter assays. The metabolite did not showed

both recruitment of PIAS1 in M2H assay and weight change in the prostate and the levator ani

muscle of castrated mice (data not shown).

Aromatase is another enzyme that plays a pivotal role in steroid metabolism. The enzyme

converts Ts to estradiol, which is the most physiologically important form of estrogen. This

enzymatic reaction is known to be crucial in several physiological and pathological processes.

Aromatase is ubiquitously expressed throughout the male reproductive organ and it contributes to

local conversion of Ts to estradiol for prostate growth (Matzkin and Soloway, 1992; Tsugaya et al.,

1996). Excess amount of estradiol increases size of the prostate to be incidence risk of prostate

cancer. Therefore, Ts may increase the prostate size through conversion into both estradiol and

DHT. Since SARM cannot be aromatized, all of its effects will be explained by AR binding but not

metabolic conversion into active androgens/estrogens in prostate.

The other possible mechanism for the tissue selectivity of SARM is tissue-specific expression

of cofactors. Prostatic cells exhibited higher PIAS1 expression than skeletal muscle cells (Figure

16A). Similar trend was obtained in the prostate and the skeletal muscle tissues (Figure 16B). The

high expression of PIAS1, greater recruitment of PIAS1 to the AR, and its co-activation of AR

induced by DHT in prostate cells might be able to explain prostate-oriented activity of DHT. On the

other hand, expression level of COX5B in skeletal muscle was higher than that in prostate tissue

(Figure 16B). In the skeletal muscle, difference in recruitment of COX5B to the AR was not

detected between TSAA-291 and DHT at 100 nM. Cofactors such as COX5B might account for the

agonist activity of the AR in skeletal muscle. I need to examine whether COX5B acts as a

coactivator of AR or not in my reporter assay with 293T cells. The result can be support

responsibility of COX5B in AR co-activation in the skeletal muscle (Figure 19). Additional studies,

100

such as exploration for mutation or amplification status of these cofactors in target organs, will

support elucidation for more promising mechanisms for tissue selectivity of SARM.

LHRH analogs are widely used in cancer treatment. The mechanism of action of these

analogs is mainly based on the inhibition of pituitary and gonadal functions (Emons and Schally,

1994; Schally, 1999). The decrease in circulating LH and FSH by administration of LHRH analogs

and concomitant down regulation of gonadal receptors for LH and FSH inhibit testicular or ovarian

function and reduce sex-steroid levels. This state is called “chemical castration" or medical

castration”. In addition to the pituitary gland, LHRH receptor is expressed in the prostate, the breast,

the uterus, the ovaries, as well as many tumors. The function of these tissues might also be

influenced by inhibitory actions of LHRH analogs (Emons and Schally, 1994, Schally, 1999,

Schally et al., 2001). So, therapies with LHRH analogs cause a variety of adverse effects, such as

accelerated bone loss with an increased risk of skeletal fractures, loss of libido, sexual behavior,

depression, heart attacks, cardiovascular disease, and hot flashes (Smith et al., 2001; Higano, 2003;

Harle et al., 2006; Sharma and Muggia, 2013). My data suggest that TSAA-291 has anabolic effects

on skeletal muscle with limited and less harmful androgenic effects on the prostate. Therefore,

combination treatment with TSAA-291 and LHRH analogs might be one of feasible therapeutic

options for sex hormone-dependent diseases. Taken together, I confirmed that TSAA-291 is SARM

and revealed that greater AR activation via higher recruitment of PIAS1 and other cofactors, such as

COX5B, to AR in response to SARM in skeletal muscle cells than in prostate cells lead to higher

anabolic/androgenic ratio profile of SARM in vivo (Figure 19).

My studies showed that the discovery and development of medicine which was focused on the

signal transduction is very useful due to their diverse physiological responses. Confirmation of

anti-tumor activity or application for hormone-related disease of SARM (TSAA-291) and

development of functional modulators for LIGHT or its receptor could lead further to a novel,

effective and safe therapeutic approach for various diseases including cancer.

101

Figures

Figure 18. Summary of my study regarding LIGHT-mediated growth inhibition activity in

cancer cells. LIGHT activates the transcriptional machinery necessary for transdifferentiation from

a skeletal- to smooth-muscle lineage via LTR. In my studies, the activated NF-B play only a

minor role in significant morphological change and growth delay by LIGHT or LT12 treatment

on RD cells. Other transcriptional factors might be involved in the cellular response.

102

Figure 18

Growth delayDifferentiation

Transcription

DNA

nucleus

LT12 LIGHT

HVEM/TR2

LTR

?

TNF LT

p55TNFR

p75TNFR

Transcription

DNA

nucleus

NF-B

Various physiological functions

103

Figure 19. Summary of my study regarding the tissue selectivity of SARM. The higher

expression of PIAS1, greater recruitment of PIAS1 to the AR and the effect as coactivator of AR

induced by DHT in prostate cells might explain the prostate-oriented activity of DHT. In skeletal

muscle, the higher expression of COX5B and similar recruitment of COX5B to the AR between

TSAA-291 and DHT at 100nM might account for the agonist activity of the AR. To confirm the

responsibility of COX5B in skeletal muscle, the further examination whether or not COX5B acts as

a coactivator of the AR is required.

104

Figure 19

105

Acknowledgements

106

I am most grateful to Professors Osamu Numata, Kazuto Nakata, and Yuji Inagaki and

Associate Professors Kentaro Nakano, University of Tsukuba, for their continuous guidance and

valuable discussions through my doctoral program.

I also thank Drs. Syuichi Furuya, Masahiko Fujino, Yasuhiro Sumino, Yukio Fujisawa, Osamu.

Nishimura, Hidekazu Sawada, Yasushi Shintani, Hideki Matsui, Kazunori Nishi, and my many

colleagues in Takeda Pharmaceutical Company Limited for interest and encouragement throughout

this work; Drs. Masami Kusaka, and Masuo Yamaoka for helpful suggestions; and Drs. Maki

Hasegawa, Yumiko Akinaga, Takashi Santou, and Masaharu Nakayama for the excellent technical

support.

I specially thank Drs. Masaki Hosoya, and Takahito Hara for useful discussions, and

comments.

I also express my gratitude to Drs. Tsudoi Miyoshi, and Masayuki Sumida, current supervisor

at Japan oncology business unit in Takeda Pharmaceutical Company Limited, for endorsement of

participation in the graduate program.

Finally, I would like to appreciate my family who kindly supported my life in University of

Tsukuba.

107

References

108

Aggarwal BB, and Natarajan K. (1996). Tumor necrosis factors: Developments during the last

decade. Eur. Cytokine Netw. 7, 93–124.

Aguanno S, Bouche M, Adamo S, Molinaro, M. (1990).

12-O-Tetradecanoylphorbol-13-acetate-induced differentiation of a human rhabdomyosarcoma cell

line. Cancer Res. 50, 3377–3382.

Allan GF, Tannenbaum P, Sbriscia T, Linton O, Lai MT, Haynes-Johnson D, Bhattacharjee S,

Zhang X, Sui Z, Lundeen SG. (2007). A selective androgen receptor modulator with minimal

prostate hypertrophic activity enhances lean body mass in male rats and stimulates sexual behavior

in female rats. Endocrine 32, 41–51.

Ambrosino C, Iwata T, Scafoglios C, Mallardo M, Klein R, Nebreda AR. (2006). TEF-1 and

C/EBPβ are major p38 MAPK-regulated transcription factors in proliferating cardiomyocytes.

Biochem. J. 396, 163–172.

Baker SJ, and Reddy EP. (1996). Transducers of life and death: TNF receptor superfamily and

associated proteins. Oncogene 12, 1–9.

Bassey EJ, Fiatarone MA, O’Neill EF, Kelly M, Evans WJ, Lipsitz LA. (1992). Leg extensor

power and functional performance in very old men and women. Clin. Sci. (Lond). 82, 321–327.

Baumgartner RN, Waters DL, Gallagher D, Morley JE, Garry PJ. (1999). Predictors of skeletal

muscle mass in elderly men and women. Mech. Ageing Dev.107, 123–136.

Baumgartner RN. (2000). Body composition in healthy aging. Ann. N.Y. Acad. Sci. 904, 437–448.

Bhasin S, Calof OM, Storer TW, Lee ML, Mazer NA, Jasuja R, Montori VM, Gao W, Dalton

JT. (2006). Drug insight: Testosterone and selective androgen receptor modulators as anabolic

therapies for chronic illness and aging. Nat. Clin. Pract. Endocrinol. Metab. 2, 146–159.

Bhasin S, Cunningham GR, Hayes FJ, Matsumoto AM, Snyder PJ, Swerdloff RS, Montori

VM. (2006). Testosterone therapy in adult men with androgen deficiency syndromes: an endocrine

society clinical practice guideline. J. Clin. Endocrinol. Metab. 91, 1995–2010.

109

Bode J, Benham C, Knopp A, Mielke C. (2000). Transcriptional augmentation: modulation of

gene expression by scaffold/matrix-attached regions (S/MAR elements). Crit. Rev. Eukaryot. Gene

Expr. 10, 73–90.

Bosscher KD, Berghe WV, Haegeman G. (2006). Cross-talk between nuclear receptors and

nuclear factor B. Oncogene 25, 6868–6886.

Bouche M, Senni MI, Grossi AM, Zappelli F, Polimeni M, Arnold HH, Cossu G, Molinaro M.

(1993). TPA-induced differentiation of human rhabdomyosarcoma cells: Expression of the

myogenic regulatory factors. Exp. Cell Res. 208, 209–217.

Brady ME, Ozanne DM, Gaughan L, Waite I, Cook S, Neal DE, Robson CN. (1999). Tip60 is a

nuclear hormone receptor coactivator. J. Biol. Chem. 274, 17599–17604.

Buijsman RC, Hermkens PH, van Rijn RD, Stock HT, Teerhuis NM. (2005). Non-steroidal

steroid receptor modulators. Curr. Med. Chem. 12, 1017-1075.

Celik S, Shankar V, Richter A, Hippe HJ, Akhavanpoor M, Bea F, Erbel C, Urban S, Blank N,

Wambsganss N, Katus HA, Dengler TJ.(2009). Proinflammatory and prothrombotic effects on

human vascular endothelial cells of immune-cell-derived LIGHT. Eur. J. Med. Res. 14, 147-156.

Chang CY, Norris JD, Gron H, Paige LA, Hamilton PT, Kenan DJ, Fowlkes D, McDonnell

DP. (1999). Dissection of the LXXLL nuclear receptor-coactivator interaction motif using

combinatorial peptide libraries: discovery of peptide antagonists of estrogen receptors α and β.

Mol.Cell Biol. 19, 8226-8239.

Chang CY, and McDonnell DP. (2002). Evaluation of ligand-dependent changes in AR structure

using peptide probes Mol. Endocrinol. 16, 647-660.

Chisamore MJ, Gentile MA, Dillon GM, Baran M, Gambone C, Riley S, Schmidt A, Flores O,

Wilkinson H, Alves SE. (2016). A novel selective androgen receptor modulator (SARM) MK-4541

exerts anti-androgenic activity in the prostate cancer xenograft R-3327G and anabolic activity on

skeletal muscle mass &amp; function in castrated mice. J. Steroid Biochem. Mol. Biol. 163, 88-97.

Crouch GD, Kalebic T, Tsokos M, Helman LJ. (1993). Ara-C treatment leads to differentiation

110

and reverses the transformed phenotype in a human rhabdomyosarcoma cell line. Exp.Cell Res. 204,

210–216.

Crowe PD, VanArsdale TL, Walter BN, Ware CF, Hession C, Ehrenfels B, Browning JL, Din

WS, Goodwin RG, Smith CA. (1994). A lymphotoxin-beta-specific receptor. Science 264, 707–

710.

Dangsheng Li. (2012). A special issue on cell signaling, disease, and stem cells. Cell Res. 22, 1-2.

Degli-Esposti MA, Davis-Smith T, Din WS, Smolak PJ, Goodwin RG, Smith CA. (1997).

Activation of the lymphotoxin beta receptor by cross-linking induces chemokine production

and growth arrest in A375 melanoma cells. J. Immunol. 158, 1756–1762.

Dotzlaw H, Moehren U, Mink S, Cato AC, Iniguez Lluhi JA, Baniahmad A. (2002). The amino

terminus of the human AR is target for corepressor action and antihormone agonism. Mol.

Endocrinol. 16, 661–673.

El-Badry OM, Minniti C, Kohn EC, Houghton PJ, Daughaday WH, Helman LJ. (1990).

Insulin-like growth factor II acts as an autocrine growth and motility factor in human

rhabdomyosarcoma tumors. Cell Growth Differ. 1, 325–331.

Emons G, and Schally AV. (1994). The use of luteinizing hormone releasing hormone agonists

and antagonists in gynecological cancers. Hum. Reprod. 9, 1364–1379.

Ettinger R, Browning JL, Michie SA, van Ewijk W, McDevitt HO. (1996). Disrupted splenic

architecture, but normal lymph node development in mice expressing a soluble lymphotoxin-beta

receptor-IgG1 fusion protein. Proc. Natl. Acad. Sci. U.S.A. 93, 13102–13107.

Fu M, Wang C, Reutens AT, Wang J, Angeletti RH, Siconolfi-Baez L, Ogryzko V,

Avantaggiati,ML, Pestell RG. (2000). p300 and p300/cAMP-response element-binding

protein-associated factor acetylate the androgen receptor at sites governing hormone-dependent

transactivation. J. Biol. Chem. 275, 20853–20860.

Gan Q, Yoshida T, Li J, Owens GK. (2007). Smooth Muscle Cells and Myofibroblasts Use

Distinct Transcriptional Mechanisms for Smooth Muscle -Actin Expression. Circ. Res. 101,

111

883-892.

Germani A, Fusco C, Martinotti S, Musaro A, Molinaro M, Zani BM. (1994). TPA-induced

differentiation of human rhabdomyosarcoma cells involves dephosphorylation and nuclear

accumulation of mutant p53. Biochem. Biophys. Res. Commun. 202, 17–24.

Gooren L. (2003). Testosterone supplementation: why and for whom? Aging Male. 6, 184–199.

Goto G, Yoshioka K, Hiraga K, Masuoka M, Nakayama R, Miki T. (1978). Synthesis and

antiandrogenic activity of 16beta-substituted-17 beta-hydroxysteroids. Chem. Pharm. Bull. 26,

1718–1728.

Gross M, Liu B, Tan J, French FS, Carey M, Shuai K. (2001). Distinct effects of PIAS proteins

on androgen-mediated gene activation in prostate cancer cells. Oncogene 20, 3880–3887.

Gross M, Yang R, Top I, Gasper C, Shuai K. (2004). PIASy-mediated repression of the androgen

receptor is independent of sumoylation. Oncogene 23, 3059–3066.

Harle LK, Maggio M, Shahani S, Braga-Basaria M, Basaria S. (2006). Endocrine complications

of androgen-deprivation therapy in men with prostate cancer. Clin. Adv. Hematol. Oncol. 9, 687–

696.

Harrop JA, McDonnell PC, Brigham-Burke M, Lyn SD, Minton J, Tan, KB, Dede K,

Spampanato J, Silverman C, Hensley P, DiPrinzio R, Emery JG, Deen K, Eichman C,

Chabot-Fletcher M, Truneh A, Young PR. (1998). Herpesvirus entry mediator ligand (HVEM-L),

a novel ligand for HVEM/TR2, stimulates proliferation of T cells and inhibits HT29 cell growth. J.

Biol. Chem. 273, 27548–27556.

Higano CS. (2003). Side effects of androgen deprivation therapy: monitoring and minimizing

toxicity. Urology 61, 32–38.

Hikichi Y, Matsui H, Tsuji I, Nishi K, Yamada T, Shintani Y, Onda H. (2001). LIGHT, a

member of the TNF superfamily, induces morphological changes and delays proliferation in the

human rhabdomyosarcoma cell line RD. Biochemical and Biophysical Research Communications,

289, 670–677.

112

Hikichi Y, Yamaoka M, Kusaka M, Hara T. (2015). Selective androgen receptor modulator

activity of a steroidal antiandrogen TSAA-291and its cofactor recruitment profile. Eur. J.

Pharmacol. 765, 322–331.

Hinson AR, Jones R, Crose LE, Belyea BC, Barr FG, Linardic CM. (2013). Human

rhabdomyosarcoma cell lines for rhabdomyosarcoma research: utility and pitfalls. Front. Oncol. 3,

1-12.

Kazmin D, Prytkova T, Cook CE, Wolfinger R, Chu TM, Beratan D, Norris JD, Chang CY,

McDonnell DP. (2006). Linking ligand-induced alterations in androgen receptor structure to

differential gene expression: a first step in the rational design of selective androgen receptor

modulators, Mol. Endocrinol. 20, 1201-1217.

Kipp M, Gohring F, Ostendorp T, van Drunen CM, van Driel R, Przybylski M, Fackelmayer

FO. (2000). SAF-Box, a conserved protein domain that specifically recognizes scaffold attachment

region DNA. Mol. Cell Biol. 20, 7480–7489.

Koni PA, Sacca R, Lawton P, Browning JL, Ruddle NH, Flavell RA. (1997). Distinct roles in

lymphoid organogenesis for lymphotoxins alpha and beta revealed in lymphotoxin beta-deficient

mice. Immunity 6, 491–500.

Knudsen E, Pazzagli C, Born TL, Bertolaet BL, Knudsen KE, Arden KC, Henry RR,

Feramisco JR. (1998). Elevated cyclins and cyclin-dependent kinase activity in the

rhabdomyosarcoma cell line RD. Cancer Res. 58, 2042–2049.

Kwon BS, Tan KB, Ni J, Oh KO, Lee ZH, Kim KK, Kim YJ, Wang S, Gentz R, Yu GL,

Harrop J, Lyn SD, Silverman C, Porter TG, Truneh A, Young PR. (1997). A newly identified

member of the tumor necrosis factor receptor superfamily with a wide tissue distribution and

involvement in lymphocyte activation. J. Biol. Chem. 272, 14272–14276.

113

Ledsaak M, Bengtsen M, Molværsmyr AK, Fuglerud BM, Matre V, Eskeland R,

Gabrielsen OS. (2016). PIAS1 binds p300 and behaves as a coactivator or corepressor of the

transcription factor c-Myb dependent on SUMO-status. Biochim. Biophys. Acta. 1859,

705-718.

Locksley RM, Killeen N, Lenardo MJ. (2001). The TNF and TNF receptor superfamilies:

Integrating mammalian biology. Cell 104, 487–501.

Marchal JA, Prados J, Melguizo C, Gómez JA, Campos J, Gallo MA, Espinosa A, Arena N,

Aránega A. (1999). GR-891: A novel 5-fluorouracil acyclonucleoside prodrug for differentiation

therapy in rhabdomyosarcoma cells. Br. J. Cancer 79, 807–813.

Marsters SA, Ayres TM, Skubatch M, Gray CL, Rothe M. Ashkenazi A. (1997). Herpesvirus

entry mediator, a member of the tumor necrosis factor receptor (TNFR) family, interacts with

members of the TNFR-associated factor family and activates the transcription factors NF-kappaB

and AP-1. J. Biol. Chem. 272, 14029–14032.

Martin SG. (2003). Cell signaling and cancer. Cancer Cell 4, 167-174.

Masuoka M, Masaki T, Yamazaki I, Hori T, Nakayama R. (1979). Anti-androgen TSAA-291.

III. Hormonal spectra of anti-androgen TSAA-291 (16 beta-ethyl-17 beta-hydroxy-4-oestren-3-one)

and its derivatives. Acta. Endocrinol. Suppl. (Copenh). 229, 36–52.

Matsui H, Hikichi Y, Tsuji I, Yamada T, Shintani Y. (2002). LIGHT, a member of the tumor

necrosis factor ligand superfamily, prevents tumor necrosis factor-alpha-mediated human primary

hepatocyte apoptosis, but not Fas-mediated apoptosis. J Biol Chem. 277, 50054-50061.

Matzkin H, and Soloway MS. (1992). Immunohistochemical evidence of the existence and

localization of aromatase in human prostatic tissues. Prostate 21, 309-314.

Mauri DN, Ebner R, Montgomery RI, Kochel, KD, Cheung TC, Yu GL, Ruben S, Murphy M,

Eisenberg RJ, Cohen GH, Spear PG, and Ware CF. (1998). LIGHT, a new member of the TNF

superfamily, and lymphotoxin alpha are ligands for herpesvirus entry mediator. Immunity 8, 21–30.

114

Melton LJ 3rd, Khosla S, Crowson CS, O’Connor MK, O’Fallon WM, Riggs BL. (2000)

Epidemiology of sarcopenia. J. Am. Geriatr. Soc. 48, 625–630.

Miner JN, Chang W, Chapman MS, Finn PD, Hong MH, Lopez FJ, Marschke KB, Rosen J,

Schrader , Turner R, van Oeveren A, Viveros H, Zhi L, Negro-Vilar A. (2007). An orally active

selective androgen receptor modulator is efficacious on bone, muscle, and sex function with

reduced impact on prostate. Endocrinology 148, 363–373.

Moilanen AM, Karvonen U, Poukka H, Yan W, Toppari J, Janne OA, Palvimo JJ. (1999). A

testis-specific androgen receptor coregulator that belongs to a novel family of nuclear proteins. J.

Biol. Chem. 274, 3700–3704.

Mooradian AD, Morley JE, Korenman SG. (1987). Biological actions of androgens. Endocr. Rev.

8, 1–28.

Morton CL, and Potter PM. (1998) Rhabdomyosarcoma-specific expression of the herpes

simplex virus thymidine kinase gene confers sensitivity to ganciclovir. J. Pharmacol. Exp. Ther.

286, 1066–1073.

Narayanan R, Mohler ML, Bohl CE, Miller DD, Dalton JT. (2008). Selective androgen receptor

modulators in preclinical and clinical development. Nucl. Recept. Signal. 6, 1-26.

Narayanan R, Coss CC, Yepuru M, Kearbey JD, Miller DD, Dalton JT. (2008). Steroidal

androgens and nonsteroidal, tissue-selective androgen receptor modulator, S-22, regulate androgen

receptor function through distinct genomic and nongenomic signaling pathways. Mol. Endocrinol.

22, 2448–2465.

Narayanan R, Ahn S, Cheney MD, Yepuru M, Miller DD, Steiner MS, Dalton JT. (2014).

Selective androgen receptor modulators (SARMs) negatively regulate triple-negative breast cancer

growth and epithelial: mesenchymal stem cell signaling. PLoS One. 9, e103202.

Negro-Vilar A. (1999). Selective androgen receptor modulators (SARMs): a novel approach to

androgen therapy for the new millennium. J. Clin. Endocrinol. Metab. 84, 3459–3462.

115

Nejishima H, Yamamoto N, Suzuki M, Furuya K, Nagata N, Yamada S. (2012).

Anti-androgenic effects of S- 40542, A novel non-steroidal selective androgen receptor modulator

(SARM) for the treatment of benign prostatic hyperplasia. The Prostate. 72, 1580-1587.

Nishida T, and Yasuda H. (2002). PIAS1 and PIASxalpha function as SUMO-E3 ligases toward

androgen receptor and repress androgen receptor-dependent transcription. J. Biol. Chem. 277,

41311–41317.

Pappo AS, Shapiro DN, Crist WM, Maurer, HM. (1995). Biology and therapy of pediatric

rhabdomyosarcoma. J. Clin. Oncol. 13, 2123–2139.

Pappo AS, Anderson JR, Crist WM, Wharam MD, Breitfeld PP, Hawkins D, Raney RB,

Womer RB, Parham DM, Qualman SJ, Grier HE. (1999). Survival after relapse in children and

adolescents with rhabdomyosarcoma: a report from the intergroup rhabdomyosarcoma study group.

J. Clin. Oncol. 17, 3487–3493.

Parham DM. (1994). The molecular biology of childhood rhabdomyosarcoma. Semin. Diagn.

Pathol. 11, 39–46.

Purves D, Augustine GJ, Fitzpatrick D. (2001). Neuroscience 2nd edition.

Rooney IA, Butrovich KD, Glass AA, Borboroglu S, Benedict CA, Whitbeck JC, Cohen GH,

Eisenberg J, Ware CF. (2000). The lymphotoxin- receptor is necessary and sufficient for

LIGHT-mediated apoptosis of tumor cells. J. Biol.Chem. 275, 14307–14315.

Roy TA, Blackman MR, Harman SM, Tobin JD, Schrager M, Metter EJ. (2002).

Interrelationships of serum testosterone and free testosterone index with FFM and strength in aging

men. Am. J. Physiol. Endocrinol. Metab. 283, 284–294.

Sasada S, Kodaira M, Shimoi T, Shimomura A, Yunokawa M, Yonemori K, Shimizu C,

Fujiwara Y, Tamura K. (2016). Ifosfamide and etoposide chemotherapy in the treatment of

recurrent/refractory rhabdomyosarcoma in adults. Anticancer Res. 36, 2429-2432.

116

Schally AV. (1999). LH-RH analogues: their impact on the control of tumorigenesis. Peptides 20,

1247–1262.

Schally AV. (1999). LH-RH analogues: I. their impact on reproductive medicine. Gynecol.

Endocrinol. 13, 401–409.

Schally AV, Halmos G, Rekasi A, Arencibia JM. (2001). The actions of LH-RH agonists,

antagonists, and cytotoxic analogs on the LH-RH receptors on the pituitary and tumors. In: Devroey

P, editor. Infertility and Reproductive Medicine Clinics of North America. GnRH Analogs 12,

17-44.

Scrable HJ, Witte DP, Lampkin BC, Cavenee WK. (1987). Chromosomal localization of the

human rhabdomyosarcoma locus by mitotic recombination mapping. Nature 329, 645–647.

Shang Y, and Brown M. (2002). Molecular determinants for the tissue specificity of SERMs.

Science 295, 2465–2468.

Sharma SP, and Muggia F. (2013). Supraventricular tachycardia and urticaria complicating

leuprolide-induced ovarian suppression in a young woman with breast cancer: a case report.

ecancermedicalscience 7, 1–4.

Smith CA, Farrah T, Goodwin RG. (1994). The TNF receptor superfamily of cellular and viral

proteins: Activation, costimulation, and death. Cell 76, 959–962.

Smith MR, Finkelstein JS, McGovern FJ, Zietman AL, Fallon MA, Schoenfeld DA, Kantoff

PW. (2001). Changes in body composition during androgen deprivation therapy for prostate cancer.

J. Clin. Endo. Metab. 345, 599–603.

Srinath R, and Dobs A. (2014). Enobosarm (GTx-024, S-22): a potential treatment for cachexia.

Future Oncol. 10, 187-194.

Stratton MR, Darling J, Pilking-ton GJ, Lantos PL, Reeves BR, Cooper CS, (1989).

Characterization of the human cell line TE671. Carcinogenesis 10, 899–905.

117

Sudo K, Yoshida K, Akinaga Y, Nakayama R. (1981). 5alpha-reduction of an anti-androgen

TSAA-291, 16beta-ethyl-17beta-hydroxy-4-estren-3-one, by nuclear 5alpha-reductase in rat

prostates. Steroids 38, 55–71.

Tamada K, Shimozaki K, Chapoval AI, Zhu G, Sica G, Flies D, Boone T, Hsu H, Fu YX,

Nagata S, Ni J, Chen L. (2000). Modulation of T-cell-mediated immunity in tumor and

graft-versus-host disease models through the LIGHT co-stimulatory pathway. Nat. Med. 6, 283–

289.

Tamada K, Shimozaki K, Chapoval AI, Zhai Y, Su J, Chen S-F, Hsieh S-L, Nagata S, Ni J,

Chen L. (2000). LIGHT, a TNF-like molecule, costimulates T cell proliferation and is required for

dendritic cell-mediated allogeneic T cell response. J. Immunol. 164, 4105–4110.

Toropainen S, Malinen M, Kaikkonen S, Rytinki M, Jaaskelainen T, Sahu B, Janne OA,

Palvimo, JJ. (2015). SUMO ligase PIAS1 functions as a target gene selective androgen receptor

coregulator on prostate cancer cell chromatin. Nucleic. Acids. Res. 43, 848–861.

Tsugaya M, Harada N, Tozawa K, Yamada Y, Hayashi Y, Tanaka S, Maruyama K, Kohri K.

(1996). Aromatase mRNA levels in benign prostatic hyperplasia and prostate cancer. Int. J. Urol. 3,

292-296.

Urashima M, Teoh G, Akiyama M, Yuza Y, Anderson KC, Maekawa K. (1999). Restoration of

p16INK4A protein induces myogenic differentiation in RD rhabdomyosarcoma cells. Br. J.Cancer

79, 1032–1036.

Vajda EG, López FJ, Rix P, Hill R, Chen Y, Lee KJ, O'Brien Z, Chang WY, Meglasson MD,

Lee YH. (2009). Pharmacokinetics and pharmacodynamics of LGD-3303. J. Pharmacol. Exp. Ther.

328, 663–670.

Wright AS, Douglas RC, Thomas LN, Lazier CB, Rittmaster RS. (1999). Androgen-induced

regrowth in the castrated rat ventral prostate: role of 5alpha-reductase Endocrinology 140,

4509-4515.

118

Yamana K, and Labrie F. (2010). Human type 3 5α-reductase is expressed in peripheral tissues at

higher levels than types 1 and 2 and its activity is potently inhibited by finasteride and dutasteride.

Horm. Mol. Biol. Clin. Investig. 2, 293–299.

Yang F, Li X, Sharma M, Sasaki CY, Longo DL, Lim B, Sun Z. (2002). Linking beta-catenin to

androgen-signaling pathway. J. Biol. Chem. 277, 11336–11344.

Zhai Y, Guo R, Hsu TL, Yu GL, Ni J, Kwon BS, Jiang GW, Lu J, Tan J, Ugustus M, Carter

K, Rojas L, Zhu F, Lincoln C, Endress G, Xing L, Wang S, Oh KO, Gentz R, Ruben S,

Lippman ME, Hsieh SL, Yang D. (1998). LIGHT, a novel ligand for lymphotoxin beta receptor

and TR2/HVEM induces apoptosis and suppresses in vivo tumor formation via gene transfer. J. Clin.

Invest. 102, 1142–1151.

Zhou FH, Foster BK, Zhou XF, Cowin AJ, Xian CJ. (2006). TNF-α mediates p38 MAP

kinase activation and negatively regulates bone formation at the injured growth plate in rats. J.

Bone Miner. Res. 21, 1075-1088.