1
ROLE OF HUMAN GROWTH HORMONE- REGULATED HOXA1 IN MAMMARY GLAND NEOPLASTIC PROGRESSION
ZHANG XIN
A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
DEPARTMENT OF MEDICINE
NATIONAL UNIVERSITY OF SINGAPORE
2005
2
ROLE OF HUMAN GROWTH HORMONE- REGULATED HOXA1 IN MAMMARY GLAND NEOPLASTIC PROGRESSION
ZHANG XIN
(B.S., Shangdong Medical University, Jinan, Shangdong, China) (M.S., Shangdong Medical University, Jinan, Shangdong, China)
A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
DEPARTMENT OF MEDICINE
NATIONAL UNIVERSITY OF SINGAPORE
2005
3
Acknowledgement
I would like to thank the following individuals for their support and help during the course of this experience.
Professor K O Lee and Associate Professor Peter Lobie, my supervisors, for their guidance, constant support and encouragement. Also for their advice on my career and different aspects of life.
To all the members of the Endocrine and Signal Transduction labs, for the collaboration, interactions, advice and help.
To all the members in the Department of Medicine, particularly Associate Professor Benjamin Ong and Nancy, for being so helpful and kind.
My parents and my sister and my wife for their deeply love and being my strength.
4
Table of Contents
Acknowledgements…………………………………………………………………….....i
Table of Contents………………………………………………………………………...ii
Summary………………………………………………………………………………….v
List of publications and presentations…………………………………..………….....vii
Abbreviations……………………………………………………………………….….viii
Chapter 1 Introduction………………………………………………………………..1
Chapter 2 Literature review…………………………………………………………..6
2.1 Growth hormone (GH)………………………………………………………....6
2.2 Growth hormone and mammary gland development and neoplasia………15
2.3 Autocrine GH and breast cancer……………………………………………..20
2.4 Homeobox genes………………………………………………………………32
2.5 Homeodomain…………………………………………………………………36
2.6 Homeobox genes and development…………………………………………..38
2.7 Cellular effects and targets of homeobox genes……………………………..44
2.8 Homeobox genes and human diseases………………………………………..51
2.9 Homeobox genes and cancers…………………………………………...……54
2.10 Homeobox genes and mammary gland…………………………….………...63
2.11 E-cadherin………………………………………………………..……………76
2.12 E-cadherin directed signaling…………………………………….…………..82
Chapter 3 Materials and methods…………………………………………..……….88
Chapter 4 Results………………………………………………………….………..113
4.1 Human GH-regulated HOXA1 is a human mammary epithelial oncogene…113
5
4.1.1 Autocrine hGH stimulation of mammary carcinoma cells increases HOXA1
mRNA, protein, and transcriptional activity………………………………..…………115
4.1.2 Forced expression of HOXA1 in human mammary carcinoma cells results in
increased total cell number……………………………………………….…………….120
4.1.3 HOXA1 expression in human mammary carcinoma cells prevents apoptotic
cell death in a Bcl-2 dependent manner…………………………………….…………..132
4.1.4 HOXA1 expression in mammary carcinoma cells protects against
doxorubicin-induced apoptosis………………………………………………………….140
4.1.5 Forced expression of HOXA1 in human mammary carcinoma cells results in
increased anchorage-independent growth in a Bcl-2-dependent manner…………...146
4.1.6 Forced expression of HOXA1 results in oncogenic transformation of
immortalized human mammary epithelial cells in vitro………………………………152
4.1.7 Forced expression of HOXA1 results in oncogenic transformation of
immortalized human mammary epithelial cells in vivo……………………………….163
4.2 E-cadherin-directed signaling upregulates HOXA1 through Rac1……...…..170
4.2.1 HOXA1 mRNA, protein, and transcriptional activity are increased in
human mammary carcinoma cells at full confluence ………...………...……172
4.2.2 The increment in HOXA1 expression in human mammary carcinoma cells at
full confluence is cell-cell adhesion dependent…………………….….……...177
4.2.3 The increment in HOXA1 expression in human mammary carcinoma cells at
full confluence is E-cadherin dependent…………………….………..……...183
4.2.4 The increment in HOXA1 expression in human mammary carcinoma cells at
6
full confluence is regulated by E-cadherin-activated signaling………...…...190
4.2.5 Rac1, but not PI-3 kinase, α-, β-, or γ-catenins is involved in the E-
cadherin-activated increment in HOXA1 expression in human mammary
carcinoma cells……………………………………….………………..…..…..194
4.2.6 Increased HOXA1 expression by E-cadherin-activated signaling has
increased anti-apoptotic effect in human mammary carcinoma cells……..201
Chapter 5 General discussion…………………………………………………....…206
References……………………………………………………………………..……….222
7
Summary
Autocrine production of human growth hormone (hGH) by human mammary carcinoma cells may direct human mammary carcinoma cell behavior to impact on the final clinical prognosis. One major mechanism by which GH affects cellular and somatic function is by regulating the level of specific mRNA species. To investigate the regulation of homeobox gene HOXA1 by autocrine production of hGH in human mammary carcinoma cells (MCF-7), which was identified by previous researchers using cDNA miroarray analyses, the present study used an in vitro model of MCF-7 cells stably transfected with the wild-type hGH gene (MCF7-hGH) and a control stably transfected with a translation- deficient hGH gene (MCF7-MUT). The production of autocrine hGH increased HOXA1 mRNA, protein, and transcriptional activity in human mammary carcinoma cells. Forced expression of HOXA1 in MCF-7 cells resulted in increased total cell number. HOXA1 expression in MCF-7 cells prevented apoptotic cell death in a Bcl-2-dependent manner.
HOXA1 expression in MCF-7 cells also protected against doxorubicin-induced apoptosis.
Forced expression of HOXA1 in MCF-7 cells, in a Bcl-2-dependent manner, resulted in dramatic increase in anchorage-independent proliferation in suspension culture and colony formation in soft agar. HOXA1 overexpression was sufficient to oncogenically transform spontaneously immortalized human mammary epithelial cells (MCF-10A) with resultant colony formation in soft agar. When implanted into the mammary (axillary) fat pad of female severe combined immunodeficient mice (SCID mice), MCF-10A cells with overexpressed HOXA1 (MCF10A-HOXA1 cells) formed aggressive tumor in vivo. Thus, human growth hormone-regulated HOXA1 is a human mammary epithelial oncogene.
8
HOXA1 was found to have increased expression in MCF-7 cells at full confluence. This confluence-dependent expression of HOXA1 was abrogated by incubating the cell with
EGTA which blocked the cell-cell adhesion or with an E-cadherin functional blocking antibody. To distinguish the E-cadherin-dependent regulation from the E-cadherin-
activated signaling regulation of HOXA1 expression, a functional E-cadherin ligand
(hE/Fc) was utilized and this confluence-dependent expression of HOXA1 was confirmed
to be direct downstream consequence of E-cadherin ligation. Furthermore, E-cadherin
increased HOXA1 at full confluence through Rac1. Increased HOXA1 expression at full
confluence had increased anti-apoptotic effect in human mammary carcinoma cells.
These results indicated that a cross link may exist between E-cadherin and HOXA1
signaling pathways and that direct E-cadherin signaling may have anti-apoptotic effect.
9
List of publications and presentations
I. Overexpression of human growth hormone-regulated HOXA1 results in oncogenic transformation of human mammary epithelial cells. Zhang X, Zhu T, Lee KO, Lobie PE (2002) Abstract presented at Endocrine Society meeting and awarded a 2002 Endocrine Society Travel Award
II. Human growth hormone-regulated HOXA1 is a human mammary epithelial oncogene. Zhang X, Zhu T, Chen Y, Mertani HC, Lee KO, Lobie PE (2003) J Biol Chem. 2003 Feb 28; 278(9):7580-90
III. Oncogenic transformation of human mammary epithelial cells by human autocrine growth hormone Zhu T, Starlin-Emerald B, Zhang X, Lee KO, Gluckman PD, Mertani HC, ,Lobie PE (2005) Cancer Res. 2005 Jan 1, 65(1):317-24
IV. E-cadherin-directed signaling upregulates HOXA1 through Rac1. Zhang X, Chen Y, Lee KO, Lobie PE (manuscript in preparation)
10
Abbreviations
ATP adenosine triphosphate bp base pair
β-gal beta-galactosidase
BSA bovine serum albumin
Brdu 5’-bromo-2’deoxyuridine
CAM cell adhesion molecule cAMP cyclic adenosine monophosphate cDNA complementary deoxyribonucleic acid ddH2O double distilled water
DMEM Dulbecco’s modified Eagle’s medium
DMSO dimethyl sulfoxide
DNA deoxyribonucleic acid
DTT 1,4-dithiothreitol
EDTA ethylenediaminetetra acetic acid
EGF epidermal growth factor
EGTA ethyleneglycoltetra acetic acid
FBS fetal bovine serum
HD homeodomain hGH human growth hormone hGH-N hGH-normal gene hGH-V hGH-variant gene
11 hPRL human prolactin
FGF fibroblast growth factor
FITC fluorescein isothiocyanate
FAK focal adhesion kinase
GHA GH antagonist
GHBP GH binding protein
GHR GH receptor
GHRH GH releasing hormone
GTP guanosine triphosphate hr hour
IGF-1 insulin-like growth factor-1
IL interleukin
IU international unit
JAK Janus kinase kb kilobase kDa kilo Dalton
LUC luciferase
M molar
MAPK mitogen activated protein kinase
MEK MAPK kinase min minute mRNA messenger ribonucleic acid
N-CAM neural cell adhesion molecule
12
NP-40 nonidet P-40
PAGE polyacrylamide gel electrophoresis
PBS phosphate buffered saline
PDGF platelet-derived growth factor
PI3-kinase phosphatidylinositol 3’-kinase
Pit-1 pituitary-specific factor-1
PKA protein kinase A
PTKs protein tyrosine kinase
PMSF phenylmethylsulfonyl fluoride
PRL prolactin
RNase ribonuclease
RA retinoic acid rpm revolutions per minute
RPMI 1640 Roosevelt Park Memorial Institute medium
RT room temperature
RT-PCR reverse transcriptase-polymerase chain reaction
SAM substrate adhesion molecule
SDS sodium dodecyl sulphate sec second
SFM serum-free media
STAT signal transducer and activator of transcription
Tris 2-amino-2(hydroxymethyl)-1-3-propanediol
13
UV ultraviolet
1
Chapter 1 Introduction
Growth hormone (GH) is obligatory for normal pubertal mammary gland development in
non-primates (Hull and Harvey, 2001). Specifically, GH acts on both the mammary
stromal and epithelial components to result in ductal elongation and the differentiation of
ductal epithelia into terminal end buds (Walden et al., 1998; Kleinberg, 1997). Expression
of the human growth hormone (hGH) transgene in mice results in precocious
development of the mammary gland (Bchini et al., 1991; Tornell et al., 1991) and the
development of neoplasia (Tornell et al., 1991). Conversely, spontaneous or experimentally engineered functional deficiency of GH (Nagasawa et al., 1985; Swanson and Unterman, 2002; Stavrou and Kleinberg, 2001; Okada and Kopchick, 2001) results in severely impaired mammary gland development and virtual resistance to the spontaneous
development of hyperplastic alveolar nodules (Nagasawa et al., 1985) and chemically induced mammary carcinogenesis (Swanson and Unterman, 2002; Okada and Kopchick,
2001). Similarly, in a primate model, hGH administration results in marked hyperplasia
of the mammary gland with an increased epithelial proliferation index (Ng et al., 1997).
Thus, the somatotropic axis represents one potential and unutilized approach for novel
therapeutic approaches to the treatment of mammary epithelial neoplasia.
The hGH gene is also expressed in epithelial cells of the normal human mammary gland
(Raccurt et al., 2002). Increased epithelial expression of the hGH gene is associated with
the acquisition of pathological proliferation, and the highest level of hGH gene expression
is observed in metastatic mammary carcinoma cells (Raccurt et al., 2002). Human growth
hormone receptor gene expression per mammary epithelial cell remains constant
2
throughout the process of neoplastic progression (Mertani et al., 1998), and therefore,
alterations in the local concentration of ligand are likely to be pivotal in determining the
effects of hGH on mammary epithelial cell behavior. We have recently generated a model
system to study the role of autocrine-produced hGH in mammary carcinoma by stable
transfection of either the hGH gene or a translation-deficient hGH gene into mammary
carcinoma cells (Kaulsay et al., 1999). The autocrine production of hGH by mammary carcinoma cells results in a hyperproliferative state with marked synergism between
trophic agents such as insulin-like growth factor-1 (Kaulsay et al., 1999). The increase in
mammary carcinoma cell number as a consequence of autocrine production of hGH is a
result of both increased mitogenesis and decreased apoptosis (Kaulsay et al., 2001). Thus, autocrine production of hGH by mammary carcinoma cells may direct mammary
carcinoma cell behavior to impact on the final clinical prognosis, and therefore,
systematic analysis of the relevant mechanistic features by which it exerts its cellular
effects is required.
One major mechanism by which GH affects cellular and somatic function is by regulating
the level of specific mRNA species (Isaksson et al., 1985). The previous researchers
utilized cDNA microarray analyses to identify genes regulated by autocrine production of
hGH in human mammary carcinoma cells (MCF-7) (Mertani et al., 2001). One gene
demonstrated to be upregulated by autocrine hGH in MCF-7 cells was the homeobox
containing transcription factor HOXA1 (Mertani et al., 2001).
Homeobox genes are a family of regulatory genes encoding specific nuclear proteins
(homeoproteins) that act as transcription factors. Homeobox genes share a common
3
nucleotide sequence motif (the homeobox) encoding the roughly 61 amino acid
homeodomain (HD). The Drosophila Antennapedia (Antp) HD defines a consensus
sequence referred to as class I HD (Hox genes) (Akam, 1987). In mice (Hox genes) and
human (HOX genes) there are at least 39 genes organized in four genomic clusters of
approximately 100 kb in length, called Hox loci, each localized on a different
chromosome (HOXA at 7p15.3, HOXB at 17p21.3, HOXC at 12q13.3, and HOXD at
2q31) and comprising 9-11 genes each (Apiou et al., 1996; Scott, 1992). The Hox genes
are arranged in the same order along the chromosomes as they are expressed along the
anteroposterior axis of the embryo, i.e. the genes that are located 5’ in the clusters are
expressed most posteriorly, whereas the more 3’ located genes are progressively expressed in more anterior regions (Mark et al., 1997).
Homeobox genes play a key role in a variety of developmental processes including
central nervous system and skeletal development, limb and digit specification, and
organogenesis (Edelman et al., 1993). For example, HOXA1 has been demonstrated to be
required for vertebrate hindbrain segmentation (Barrow et al., 2000). Several reports also suggest the involvement of homeobox-containing genes in the control of cell proliferation
and, when dysregulated, in oncogenesis (Cillo et al., 1992; Abate-Shen, 2002). Gain of
function of certain classes of homeobox genes has been demonstrated to promote the
oncogenic phenotype. For example, it was shown that murine homeobox genes, Pax, can promote oncogenesis in tissue culture cells and in mice. Pax1, Pax2, Pax3, Pax6, and
Pax8 proteins were able to induce transformation of cultured cells and tumor formation in mice (Maulbecker and Gruss, 1993). Moreover, alterations of HOX gene expression have
been detected in a variety of human tumors (Cillo et al., 1992; Abate-Shen, 2002; Celetti
4
et al., 1993), including those of the mammary gland (Care et al., 1998; Raman et al.,
2000). Accordingly, Hoxa1 has been detected in carcinoma of the mammary gland in the mouse but not in normal mammary tissue (Friedmann et al., 1994), and HOXA1
expression has also been detected in neoplastic lesions of the human mammary gland
(Chariot and Castronovo, 1996). Therefore, HOXA1 may be involved in mammary gland
neoplastic progression.
To date, no effect of HOXA1 on the human mammary gland tumorigenesis has been reported. The work presented here sought, therefore, to determine the role of hGH-
regulated HOXA1 in human mammary gland neoplasia in vitro and in vivo.
Autocrine hGH production by human mammary carcinoma cells increased the expression and transcriptional activity of HOXA1. Forced expression of HOXA1 in human
mammary carcinoma cells increased total cell number primarily by prevention of
apoptotic cell death in a Bcl-2-dependent manner. Forced expression of HOXA1 was
sufficient to result in the oncogenic transformation of immortalized human mammary
epithelial cells with resultant anchorage-independent growth and tumor formation in vivo.
In addition, HOXA1 was found to have increased expression at full confluence in human mammary carcinoma cells, MCF-7. HOXA1 confluence-dependent expression was directed by E-cadherin signaling through Rac1. Increased HOXA1 expression at full confluence had increased anti-apoptotic effect in human mammary carcinoma cells.
The aims of this study were as follows:
5
1. To confirm that HOXA1 is up-regulated by human autocrine GH.
2. To investigate the role of HOXA1 in human mammary gland neoplasia in vitro
and in vivo.
3. To determine the mechanism of HOXA1 confluence-dependent expression.
6
Chapter 2 Literature review
2.1 Growth hormone
2.1.1 Growth hormone structure
The pituitary gland was first recognized to have growth-promoting activity in 1921
(Evans and Long, 1921). Later, growth was believed to be regulated by a growth factor.
In 1945, Li et al. isolated growth hormone (GH) from bovine pituitary glands (Li et al.,
1945). Growth hormone belongs to a large family of evolutionarily related protein hormones that includes prolactin (PRL), mouse proliferin (mPLF), mouse proliferin- related protein (mPRP), rat decidual PRL-like protein (rdecPRP), and somatolactin (Roby et al., 1993; Ono et al., 1990; Linzer et al., 1985; Linzer and Nathans, 1984; Duckworth et al., 1986). The human GH (hGH) gene is approximately 2.6 kb in length and is located on the long arm of chromosome 17, q22-24 (Miller and Eberhardt, 1983; Kopchick and
Andry, 2000). The hGH gene is a part of a gene cluster of five related genes (Figure 1).
They are GH-N (growth hormone-normal gene), chorionic somatomammotropin-like
(CS-L) gene, chorionic somatomammotropin-A (CS-A), GH-V (growth hormone- variant), and the chorionic somatomammotropin-B (CS-B) gene (Hirt et al., 1987). In humans, the five genes have more than 92% nucleotide sequence identity in their coding and flanking regions (Miller and Eberhardt, 1983). Human GH-N is expressed only in the pituitary gland and encodes two hGH proteins: a 22-kD and a 20-kD protein. The latter is encoded by an alternative spliced product of the primary hGH-N gene transcript (Cooke et al., 1988).
7
0kb 50 kb
GH-N GS-L GS-A GH-V GS-B
I II III IV V
0kb 2.5kb
Figure 1. Schematic representation of the human GH gene cluster. The gene cluster spans over 50 kb and consists of GH-N, CS-L, CS-A, GH-V, and CS-B. The hGH gene, approximately 2.6 kb in length, consists of five exons (I–V) and four introns (a–d) (modified from Hirt et al., 1987).
In 1987, the three-dimensional structure of porcine GH was first determined (Abdel-
Meguid et al., 1987). In 1992, the crystal structure of hGH was also determined (De Vos et al., 1992). Human GH is a 191 amino acid single chain polypeptide with a molecular weight of 22 kDa. Human GH has four helices in the core. The NH2- and COOH-terminal helices (helices 1 and 4) are longer than the other two. The helices are oriented up-up- down-down. Helix 1 and helix 2 are linked by a connection including residues 35 to 71 and a connection consisting of residues 129 to 154 links helix 3 and helix 4. Helix 2 is linked to helix 3 by a short segment (residues 93 to 105). In addition to the four helices in the core, there are three much shorter helices in the connecting loops. One is located at the beginning and another is at the end of the connection between helices 1 and 2. The third one is located in the connection between helices 2 and 3. Human GH has two disulfide bonds at Cys53-Cys165 and Cys182-Cys189 (De Vos et al., 1992). These four
8
Cys residues are conserved between GH, prolactins (PRL) and placental lactogens (PL)
(Nicoll et al., 1986). The core of hGH is made up of mostly hydrophobic residues and there are other hydrophobic clusters between the four-helix core and the connections (De
Vos et al., 1992).
2.1.2 Growth hormone biological effects
The biological effects of GH are pleiotropic and involve multiple organs and physiological systems. The major biological effect of GH is to stimulate postnatal longitudinal growth. It has been known to promote general body growth, including both organ size and longitudinal bone growth (Isaksson et al., 1985). Hypersecretion of the hormone can lead to gigantism, or acomegaly, in adults, resulting in enlarged bones, especially in the face, as well as oversized organs (Calao et al., 1997). Transgenic rabbits that overexpress GH have also been found to develop acromegaly (Costa et al., 1998). On the other hand, hyposecretion of GH can lead to dwarfism. Growth hormone-deficient children grow taller when treated with GH (Blethen et al., 1997).
In addition to growth effects, GH exerts many metabolic effects that persist throughout life (Le Roith et al., 1992). Through interaction with the GH receptor (GHR), GH is known to regulate lipid, and carbohydrate metabolism (Casanueva, 1992; Kelly et al.,
1994). Growth hormone also promotes nitrogen retention, mineral metabolism, and electrolyte balance (Strobl and Thomas, 1994). Growth hormone has been shown to stimulate the conversion of preadipose 3T3 cell lines to adipocytes (Morikawa et al.,
1982). Growth hormone can enhance the lipolytic activity of adipose tissue and reduce
9
triglyceride accumulation via inhibition of lipoprotein lipase activity (Richelsen et al.,
1994; Johannsson et al., 1997; Lobie et al., 2000). Growth hormone has been shown to
reduce fat mass, especially in individuals who have accumulated excess fat mass during
prolonged periods of GH deficiency (Russell-Jones et al., 1993). Growth hormone deficiency children are usually mildly obese and GH replacement therapy leads to a reduction in body fat (Wabitsch et al., 1995). The anabolic effect of GH induces protein synthesis in muscle resulting in increased lean muscle mass (Corpas et al., 1993). Muscle size is increased in GH-deficient individuals undergoing replacement therapy with recombinant human GH (hGH) at all ages (Manson and Wilmore, 1986; Rudman et al.,
1990). It was demonstrated that GH therapy actually increases whole body protein synthesis (Wolf et al., 1992; Le Roith et al., 2001).
Growth hormone has also several immunomodulatory functions including the proliferation of B and T cells (Postel-Vinay et al., 1997), stimulation of cytokine
(Malarkey et al., 2002) and immunoglobulin production (Kimata and Yoshida, 1994) and regulation of the activity of neutrophils, macrophages and natural killer cells
(Auernhammer and Strasburger, 1995).
2.1.3 Cellular mechanisms of GH signal transduction
The actions of GH are mediated by the binding of GH to the extracellular region of the transmembrane GH receptor (GHR) (Zhu et al., 2001). Growth hormone receptor is a member of the class I hematoproietin or cytokine/growth hormone/prolactin receptor
10
superfamily (Kopchick and Andry, 2000). Site 1 of GH binds to one receptor molecule,
after which a second receptor molecule binds to site 2 on the hormone. Thus a GH-GHR
complex is formed consisting of one molecule of GH and two molecules of GHR
(homodimer of GHR). Upon GH receptor dimerization (or conformational change as a result of ligand binding to a preformed dimmer), the tyrosine kinase JAK2 becomes rapidly phosphorylated and activated (Argetsinger et al., 1993). Studies using mutated and truncated GHRs have indicated that the cytoplasmic proline-rich box 1 region of
GHR is required for GHR-JAK2 interaction and for tyrosyl phosphorylation of JAK2
(Frank et al., 1994; Sotiropoulos et al., 1994). The GHR itself has no intrinsic tyrosine kinase activity and each JAK2 is thought to transphosphorylate one or more tyrosine residues in the kinase domain of the other JAK2, thereby activating both JAK2 molecules
(Argetsinger and Carter-Su, 1996). Once activated, JAK2 then phosphorylates the GHR on multiple intracellular tyrosine residues providing docking sites for other signaling molecules that contain SH2 or other phosphotyrosine-binding motifs (Wang et al., 1996).
Activation of JAK2 leads to the phosphorylation of and activation of a number of cytoplasmic signaling molecules including signal transducer and activator of transcription-1 (STAT1), -3 and -5, the mitogen-activated protein kinase (MAP kinase), extracellular signal-regulated kinase-1 and -2 (Erk1 and 2 or also named p44/p42 MAP kinase), insulin receptor substrate-1 (IRS-1), -2 and -3 and PI-3 kinase, protein kinase C
(PKC), focal adhesion kinase (FAK), p38 and c-Jun amino (N)-terminal kinase (JNK)
(Carter-Su et al., 1996; Moutoussamy et al., 1998; Yamauchi et al., 1998; Kopchick and
Andry, 2000). These molecules are involved in distinct signaling transduction pathways
11 that are activated by GH. As shown in Figure 2, there are three major pathways of GH signaling.
GH GH GHR Extracellular GHR P P Ras Raf JAK2 JAK2 SHC SOS Intracellular P Grb2 STAT5 2 STAT5 P P MAPKK
P 3 11 P P IRS-1 STAT5 STAT5 P p44/42 PI-3 K
P P Metabolism STAT5 STAT5 Growth Proliferation Differentiation Nucleus P P Transcriptional activation of STAT5 STAT5 Target genes
Figure 2. Diagram illustrating the three major signaling pathways of GH: 1) The JAK/STAT pathway; 2) The MAP kinase pathway; 3) The IRS/PI-3 kinase pathway (modified from Carter-Su et al., 1996).
1) The JAK/STAT pathway
12
The Janus family of tyrosine kinases is thought to be the predominant nonreceptor
tyrosine kinases required for the initiation of GH signal transduction upon ligand binding
to the receptor (Argetsinger et al., 1993; Foster et al., 1988; Silva et al., 1993). The JAK family includes JAK1, JAK2, JAK3, and Tyk2 (Argetsinger and Carter-Su, 1996). JAK2 is the tyrosine kinase that is predominantly utilized by the GH receptor (Argetsinger et al., 1993; Zhu et al., 2001). Tyrosine phosphorylation of JAK1 and JAK3 can also be stimulated by GH, but the level of activation is much smaller than that of JAK2 (Carter-
Su et al., 1996).
The main pathway by which GH and many other cytokines activates the JAK kinases and regulates gene transcription involves the STAT proteins. The STATs gene family consists of seven members: STAT1, STAT2, STAT3, STAT4, STAT5a, STAT5b and STAT6
(Copeland et al., 1995). STAT proteins are recruited to the proximity of the activated
JAKs by binding, through their SH2 domains, to phosphotyrosine-containing motifs on the cytoplasmic part of the receptor (Zhu et al., 2001). Activated STAT molecules dissociate from the receptor and translocate to the nucleus, bind to their appropriate DNA response elements and stimulate transcription. STAT5 is the predominant STAT utilized by GH.
2) The mitogen-activated protein kinase (MAPK) pathway
The MAPK superfamily consists of serine-threonine protein kinases that have important functions in the regulation of gene expression, cellular proliferation and prevention of apoptosis (Seger and Krebs, 1995; Peyssonnaux and Eychene, 2001; Pearson et al.,
13
2001). To date, nearly 20 mammalian MAP kinase family members have been identified
and more are anticipated (Pearson et al., 2001). Among them, the p44/42 MAPK (also
named ERKs), the c-jun N-terminal kinases (also named JNK) and the p38 MAPK have
been relatively well characterized (Zhu et al., 2001). Growth hormone has been reported
to activate p44/42 MAPK, JNK and p38 MAPK (Campbell et al., 1992; Moller et al.,
1992; Winston and Bertics, 1992; Zhu et al., 1998; Zhu and Lobie, 2000). One common pathway by which the membrane receptor tyrosine kinases activate p42/p44 MAP kinase involves a signal transduction cascade through the Ras-Raf pathway consisting of molecules like Shc, Grb2, Son-of-sevenless (Sos), Ras and Raf (Crews and Erikson,
1993; Pearson et al., 2001). Growth hormone has also been demonstrated to utilize this cascade (Winston and Hunter, 1995; Vanderkuur et al., 1997). p44/p42 MAPK has been reported to phosphorylate and/or activate a variety of proteins including phospholipase
A2, protein kinases c-Raf-1 and the S6 kinases p70rsk and p90rsk, cytoskeletal proteins, the transcription factors c-jun and p62TCF/Elk1 (ternary complex factor) (Davis, 1993;
Pearson et al., 2001), and the STAT proteins (Davis, 1993; David et al., 1995; Treisman,
1996; Pearson et al., 2001). Growth hormone has been demonstrated to involve all of
these downstream proteins in its signal transduction pathways (Finidori, 2000; Thomas,
1998; Argetsinger and Carter-Su, 1996).
The activation of JNK is another pathway by which GH may affect cellular function.
JNK is involved in many cellular processes including transcriptional regulation and
apoptosis and it is likely that GH utilizes JNK for some of these purposes (Herdegen et
al., 1997; Zhu et al., 2001). It has also been demonstrated that GH phosphorylates and
14
activates p38 MAPK in a JAK2-dependent manner (Zhu and Lobie, 2000). p38 MAPK is
demonstrated to be required for GH stimulation of ATF-2 and CHOP-mediated
transcription, for GH-stimulated reorganization of the actin cytoskeleton and also for GH-
stimulated mitogenesis (Zhu and Lobie, 2000).
3) The PI-3 kinase pathway
Growth hormone–stimulated PI-3 kinase activity is associated with either IRS-1 or IRS-2
in a variety of cell types in vitro and in vivo (Yenush and White, 1997; White and
Yenush, 1998). PI-3 kinase phosphorylates inositol lipids at the 3’ position of the inositol
ring to generate the polyphosphoinositides PtdIns-3-P, PtdIns-3,4-P2 and PtdIns-3,4,5-P3.
These lipids then function in signal transduction and membrane trafficking by interaction with 3-phosphoinositide-binding modules in a broad variety of proteins (Leevers et al.,
1999). PI-3 kinase is pivotal in many cellular processes including cell proliferation and survival, cytoskeletal reorganization and cellular metabolism (Yenush and White, 1997;
White and Yenush, 1998). Growth hormone has been reported to activate in a PI-3 kinase-dependent manner the serine/threonine kinase Akt/PKB to deliver an anti- apoptotic signal (Costoya et al., 1999; Liang et al., 2000). Several substrates of Akt have been identified, including glucose synthase kinase 3 (GSK-3), 6-phospho-fructo-kinase
(PFK2), GLUT4 and p70S6K and GH may utilize these molecules for its cellular effects.
PI-3 kinase has also been demonstrated to play a role in GH-stimulated actin cytoskeletal reorganization (Goh et al., 1997). Another PI-3 kinase-dependent cellular effect described is GH-stimulated lipogenesis (Ridderstrale and Tornqvist, 1994).
15
2.2 Growth hormone and mammary gland development and neoplasia
2.2.1 Mammary gland development
The mammary gland epithelium is an ectodermal derivative. The differentiation of
mammary epithelium occurs in two major stages (Lewis, 2000). In the first stage (about
day 10 of gestation), the mammary streaks are established and these streaks represent the
first morphological evidence of mammary pattern formation and differentiation. The
second stage occurs around day 11 of gestation with the definition of the nipple region.
The mammary epithelium continues to grow and invades the underlying mammary
mesenchyme. As the mammary bud elongates into a mammary sprout, it reaches a second
mesenchyme, the fat pad precursor mesenchyme and forms the rudimentary branched
tubular gland of the neonate. The gland remains rudimentary and relatively growth
quiescent from birth to puberty. At puberty, ovarian hormones stimulate rapid ductal
elongation. Ducts grow, divide, and form club-shaped terminal end buds. Breast size is increased by fat deposition and connective tissue development. Upon reaching the limits of the fat pad at ductal maturity, ductal elongation ceases and terminal end buds regress to leave a branched system of differentiated ducts. These ducts remain relatively quiescent until pregnancy. During pregnancy, hormonal changes initiate a transition from a predominantly ductal to a predominantly lobuloalveolar gland morphology. Near midpregnancy, the gland acquires the ability to produce milk proteins and at partition, the gland begins to secrete milk. Upon weaning, milk secretion ceases and the gland involutes (Tobon and Salazar, 1974; Russo and Russo, 1987; Lewis, 2000).
16
2.2.2 Growth hormone regulation of mammary gland development
In addition to estrogen, GH is obligatory for normal pubertal mammary gland
development, especially in non-primates (Hull and Harvey, 2001). Mammary
development at puberty occurs because of synergy between estrogen and GH on
formation of terminal end buds (Kleinberg, 1998). Pubertal mammary gland development
cannot take place in the absence of the pituitary gland GH in the rat (Kleinberg et al.,
1990; Feldman et al., 1993; Walden et al., 1998). Non-lactogenic and lactogenic growth hormones have been shown to be potent stimulators of mammary development while
GHR has been shown to mediate the action of GH within the mammary gland (Kleinberg,
1998). Specifically, GH acts on both the mammary stromal and epithelial components to result in ductal elongation and the differentiation of ductal epithelia into terminal end buds (Walden et al., 1998; Kleinberg, 1997).
Local implantation of bovine (b) or mouse (m) GH in the mammary gland stimulated end bud formation in female mice (Bradley and Towle, 1992). Feldman et al. have reported that treatment of hypophysectomized and ovariectomized rats with rat GH and estrogens can stimulate mammary gland development as judged by increased numbers of terminal end buds or alveolar structures (Feldman et al., 1993). In addition, in vivo administration of recombinant hGH to hypophysectomized and gonadectomized rats have both been demonstrated to be approximately 10-20 times more potent than hPRL in stimulating mammary development (Kleinberg et al., 1990). Growth hormone administration to peri- pubertal heifers and lambs similarly increase the amount of parenchymal tissue (Sejrsen
17 et al., 1986; Sandles et al., 1987; Purup et al., 1993; McFadden et al., 1990). Treatment of aged female rhesus monkeys with GH resulted in a 3 to 4-fold mitogenic increase in mammary glandular size and epithelial proliferation index, although whether this result from direct effects of GH or induction of local production of IGF-I is unknown (Ng et al.,
1997).
Expression of the hGH transgene in mice results in precocious development of the mammary gland and milk synthesis (Bchini et al., 1992; Tornell et al., 1991). Conversely, spontaneous or experimentally engineered functional deficiency of GH results in severely impaired mammary gland development and virtual resistance to the spontaneous development of hyperplastic alveolar nodules (Nafasawa et al., 1985; Swanson and
Unterman, 2002; Stavrou and Kleinberg, 2001; Okada and Kopchick, 2002). In GH receptor knock out mice, litter size is markedly reduced and ductal outgrowth is greatly retarded (Zhou et al., 1997; Kelly et al., 2002).
2.2.3 GH and mammary gland neoplasia
Many findings have shown that the growth hormone (GH)/insulin-like growth factor I axis plays an important role in mammary gland neoplasia.
2.2.3.1. In vitro studies
18
Growth hormone was shown to have growth-promoting effects on human mammary cancer cells in vitro (Benlot et al., 1997). Antagonists of growth hormone-releasing hormone were effective at inhibiting the growth of estrogen-independent human breast
cancer cells (Kahan et al., 2000).
2.2.3.2. In vivo animal studies
It was demonstrated in monkeys that hGH treatment induced mammary gland hyperplasia in aging animals (Ng et al., 1997). Transgenic mice that overexpress genes encoding GH agonists exhibit both mammary gland epithelial cell hyperplasia and an increased frequency of mammary tumors (Tornell et al., 1992). Conversely, it was demonstrated that mammary gland carcinogenesis was reduced in transgenic mice expressing a growth hormone antagonist (Pollak et al., 2001). In addition, the growth hormone receptor antagonist (G120R) was demonstrated to inhibit the growth of breast cancer xenografts in nude mice (Roshan et al., 1999).
lit/lit mouse is a mouse strain with extremely low circulating GH and low IGF-I levels
because of a mutation in the growth hormone-releasing hormone receptor. Tumor growth
is dramatically reduced after the transplantation of human MCF-7 breast cancer cells into
this model. In addition, serum from lit/lit mice was less mitogenic to breast cancer cells
in vitro than control serum (Yang et al., 1996). The growth hormone-deficient
Spontaneous Dwarf rat (SDR) is resistant to chemically induced mammary carcinogenesis (Swanson and Unterman, 2002).
19
2.2.3.3. Human studies
Tall stature is associated with an increased incidence of breast cancer (Hunter and
Willett, 1993). Acromegaly patients suffer from malignant disorders more frequently than
the normal population (Alexander et al., 1980; Nabarro, 1987; Bengtsson et al., 1988).
Some investigators have found a general increase in the incidence of malignant tumors
(Alexander et al., 1980; Bengtsson et al., 1988) whereas other investigators have reported an over-representation of tumors in the colon of acromegaly patients (Ituarte et al., 1984).
Nabarro (Nabarro, 1987) has even reported a small but statistically significant increase in the prevalence of breast cancer in a personal series of 256 patients with acromegaly, although this has not been substantiated by other (smaller) series. In patients with breast cancer, increased serum levels of GH have also been reported (Emerman et al., 1985;
Peyrat et al., 1993). In addition, in treatment of advanced breast cancer in human it was demonstrated that hypohysectomy had beneficial effects in combination with estrogen removal, compared to estrogen removal alone (VanGilder and Goldenberg, 1975).
2.2.3.4. Expression of GH and GHR in normal breast and breast cancer
Growth hormone mRNA is detectable in normal and neoplastic mammary glands of dogs and cats (Mol et al., 1995a). Similarly, normal, hyperplastic, and neoplastic canine mammary tissue have also been demonstrated to express GH mRNA transcripts identical to canine pituitary GH mRNA (Mol et al., 1995). The expression of hGH in normal and
20
neoplastic human mammary glands identical to pituitary hGH has also been observed
(Mol et al., 1995b). Increased epithelial expression of the hGH gene is associated with
the acquisition of pathological proliferation, and the highest level of hGH gene
expression is observed in metastatic mammary carcinoma cells (Raccurt et al., 2002).
Receptors for GH have been identified in the mammary gland in mouse (Ilkbahar et al.,
1995), rat (Lincoln et al., 1990), cow (Hauser et al., 1990), rabbit, sheep, pigs (Jammes et
al., 1991), monkey (Ng et al., 1997), and human mammary gland as well (Mertani et al.,
1998). Growth hormone receptor (GHR) mRNA expression has been documented in the
mouse mammary epithelium and stroma with expression higher in the stromal versus the
epithelial compartment (Ilkbahar et al., 1999). In addition, GH receptor mRNA
expression in the mouse mammary gland decreases gradually throughout pregnancy
starting on day 8 of gestation and declines further during lactation (Ilkbahar et al., 1995).
In human, the expression levels of hGH receptor mRNA and protein were shown to be increased in both the epithelial and stromal components of breast tissue, with levels being higher in cancer tissue compared with adjacent normal tissue (Mertani et al., 1998).
2.3 Autocrine GH and breast cancer
Growth hormone, identical to pituitary GH, is synthesized locally in different tissues and organs, including the mammary gland and mammary tumors (Mol et al., 1995b).
Receptors for GH have been identified in almost all tissues, including human mammary gland and human breast cancer (Mertani et al., 1998). These data suggest that GH may
21
participate in an autocrine stimulatory loop within breast tissues and that there is a
possible role for autocrine GH in the pathogenesis of breast cancer.
With these data in mind, previous students in our lab have developed an in vitro model and investigated a) the role of autocrine hGH production in the proliferation, spreading and attachment of human mammary carcinoma cells in vitro; b) the mechanism of action of the cellular effects of autocrine hGH production in this in vitro model of human mammary carcinoma cells; and c) the role and specificity of the hGH receptor in the mediation of these effects of autocrine GH in human mammary carcinoma cells in vitro.
(Kaulsay et al., 1999; Kaulsay et al., 2000; Kaulsay et al., 2001).
2.3.1 Autocrine GH and human mammary carcinoma cell proliferation
Kaulsay et al. first demonstrated that the hGH receptor is present in MCF-7 cells which
are human mammary carcinoma cells (Kaulsay et al., 1999). She also examined the
receptor protein localization within the cell by confocal laser scanning microscopy and
found that GH receptor immunoreactivity was in a nucleocytoplasmic distribution in
MCF-7 cells, but with the majority of the immunoreacivity confined to the cytoplasm.
After that, an in vitro model of autocrine hGH production was established by stable
transfection of the hGH gene (MCF7-hGH) and also the translation-deficient hGH gene
as control (MCF7-MUT) into mammary carcinoma cells (MCF-7). The MCF7-hGH cells
synthesized hGH mRNA and hGH protein within the cell and secreted hGH into the
extracellular medium. The predominant localization of hGH was confined to vesicular
22
like structures in the cytoplasm. The MCF7-MUT cells produced hGH mRNA but did not
produce hGH protein and no secreted hGH protein was detected in the medium. Confocal
laser scanning microscopy also did not detect hGH protein in MCF7-MUT cells (Kaulsay
et al., 1999).
Using these two cell lines, a series of investigations on the cellular function of the
autocrine hGH was performed. In serum-free medium, the MCF7-hGH cell number
increased dramatically faster than MCF7-MUT cell number. Ten percent serum increased
proliferation in both cell lines. However, the proportionate increase in cell number was
greater in MCF7-hGH compared to MCF7-MUT. In the presence of hGH-G120, an hGH
receptor antagonist, proliferation of MCF7-hGH cells was suppressed to the level of cell
proliferation observed with MCF7-MUT cells. When the cells were grown in 10% serum,
hGH-G120 also abrogated the proliferative effects of autocrine hGH production in
MCF7-hGH cells (Kaulsay et al., 1999).
It has been demonstrated that exogenous hGH stimulates both the p42/44 mitogen-
activated protein kinase (MAPK) (Moller et al., 1992; Hodge et al., 1998) and the p38
MAPK pathways (Hodge et al., 1998). Both of these pathways have been demonstrated to
mediate mitogenesis in response to various cellular stimuli (Dhanasekaran and
Premkumar, 1998). In serum-free medium with PD98059, which is p42/p44 MAPK
inhibitor, or with SB203580, which is p38 MAPK inhibitor, there was no increase in cell
number of MCF7-MUT. PD98059 and SB203580 both prevented the increase in cell number conferred by autocrine productin of hGH such that there were no differences in
23
cell number between MCF7-MUT and MCF7-hGH in the presence of either inhibitor.
PD98059 or SB203580 completely prevented the synergism between autocrine hGH and
10% serum (Kaulsay et al., 1999). These results demonstrated that in this in vitro model, autocrine GH increased cell proliferation via p42/p44 and p38 MAPK pathway.
The investigators also examined the interaction of autocrine hGH with 17-β-estradiol and
insulin-like growth factor-1 (IGF-1). The proliferation of both MCF7-MUT and MCF7-
hGH was increased by 17-β-estradiol, and the different proliferation rates between
MCF7-MUT and MCF7-hGH were maintained. On the other hand, hIGF-1 stimulated the proliferation of both MCF7-MUT and MCF7-hGH, but the fold stimulation observed for
MCF7-hGH was greater than that observed for MCF7-MUT (Kaulsay et al., 1999).
Kaulsay et al. continued to check JAK-STAT expression level in the two cell lines by
Western blot analysis and found that there were no consistent differences in the expressions of JAK2, JAK1, STAT5, and STAT3 in MCF7-MUT and MCF7-hGH grown in serum-free medium or in the presence of 10% serum. However, they found that autocrine hGH increased the transcriptional response mediated by STAT5 using reporter gene assays. Autocrine hGH did not show any effect on the transcriptional response mediated by STAT1 and STAT3 (Kaulsay et al., 1999).
In conclusion, the results from these in vitro studies suggest that autocrine hGH in human mammary carcinoma cells can promote cell proliferation and transcriptional activation.
24
2.3.2 Autocrine GH and human mammary carcinoma cell spreading
The role of autocrine production of human growth hormone (hGH) in the attachment and
spreading of mammary carcinoma cells in vitro was then investigated (Kaulsay et al.,
2000). The same two cell lines, MCF7-MUT and MCF7-hGH, were used in this study.
MCF7-MUT and MCF7-hGH were cultured in collagen-coated dishes. MCF7-hGH cells
demonstrated a highly polarized morphology, with locomotor activity initiated by
formation of long protrusions as early as 5 min after plating. This was continued at 15
min, resulting in an increased surface area of cellular attachment to the substrate. MCF7-
MUT also demonstrated the formation of protrusions, but beginning at a later time point
(15 min) compared with MCF7-hGH cells and in fewer numbers compared with MCF7-
hGH cells. In contrast to MCF7-hGH, the development of cellular protrusions over the
course of 4 h in MCF7-MUT cells did not result in an elongation of the individual cell
(Kaulsay et al., 2000).
There were no significant differences in the rates of attachment between MCF7-MUT and
MCF7-hGH cells. However, MCF7-hGH cells spread at a faster rate than MCF7-MUT
cells. The increase in the rate of spreading of MCF7-hGH compared with MCF7-MUT
was completely inhibited by an hGH receptor antagonist, hGH-G120R. In addition, the rapid formation of filamentous actin-containing complexes was observed in MCF7-hGH cells. The complex was accumulated most prominently within filipodia extending from marginal edges of the cell. In MCF7-MUT cells, only small patches of filamentous actin
25
were observed and they were distributed in the cytoplasm and at the cell membrane or
short linear deposits of filamentous actin in the cell margins. MCF7-MUT cells also
demonstrated a rounded, symmetrical, and characteristically epithelial morphology.
MCF7-hGH cells displayed numerous microspikes oriented around the ventral margins
perpendicular to the long axis of the cell. MCF7-MUT cells also demonstrated
microspike activity. In contrast to the distribution pattern observed in MCF7-hGH cells,
the microspikes in MCF7-MUT cells were distributed symmetrically around the cell
periphery (Kaulsay et al., 2000).
Tyrosine phosphorylation plays important role in cell spreading to extracellular matrix
adhesion (Guan and Shalloway, 1992; Burridge et al., 1992). In MCF7-hGH cells plated
in collagen-coated substrata, an elevated level of cytoplasmic phosphotyrosine was
observed as the cells spread while in MCF7-MUT cells plated in the same strata, only
low levels of phosphotyrosine was observed. Cytoplasmic phosphotyrosine aggregates
were evident in MCF7-hGH cells while the aggregates were not present in MCF7-MUT
cells, as localization generally occurred in a diffuse pattern throughout the cell with some perinuclear localization. Growth hormone-dependent tyrosine phosphorylation within the cell has been shown to be mediated by the receptor-associated JAK2 kinase (Argetsinger et al., 1993; Lobie, 1999). Expression of JAK2 was demonstrated to be equivalent between MCF7-MUT and MCF7-hGH cells. However, when MCF7-MUT and MCF7- hGH cells were cultured on collagen substrata, the phosphotyrosine content of JAK2 was increased in MCF7-hGH cells. In addition, colocalization of JAK2 with phosphotyrosine during cell spreading was observed in MCF7-hGH cells but not in MCF7-MUT cells. The
26
increment in the rate of cell spreading stimulated by autocrine hGH was abrogated by a
JAK2 tyrosine kinase-specific inhibitor, AG490. Transient transfection of JAK2 cDNA
significantly increased the rate of spreading in MCF7-hGH, but not MCF7-MUT cells,
and this effect was inhibited by AG490 (Kaulsay et al., 2000).
In conclusion, these in vitro studies demonstrated that autocrine production of hGH
enhanced the rate of mammary carcinoma cell spreading in a JAK2-dependent manner.
2.3.3 The effects of autocrine GH on human mammary carcinoma cell are
mediated via the hGH receptor
Human growth hormone has been reported to bind to both the hGH receptor and the
hPRL receptor (Cunningham et al., 1990). To study whether the effects of autocrine hGH
on human mammary carcinoma cell behavior were mediated via the hGH receptor,
Kaulsay et al. used a specific human GH antagonist (B2036) which did not bind, activate,
or antagonize PRL receptors of either rat or human origin (Goffin et al., 1999). They
found that the enhanced JAK2 tyrosine phosphorylation observed in MCF7-hGH cells compared with MCF7-MUT cells was abrogated by B2036 (Kaulsay et al., 2001). B2036
did not affect MCF7-MUT total cell number in comparison to MCF7-MUT cells cultured
in the absence of B2036 (i.e. no effect). In contrast, B2036 prevented the autocrine hGH-
stimulated increase in cell number observed in MCF7-hGH cells cultured in serum-free
media. B2036 also inhibited the increase in cell number due to the synergistic effect of
autocrine hGH with heterologous factors in serum. The increase in cell number observed
27 in MCF7-hGH compared with MCF7-MUT cells could be due to the increase in cell synthesis or the prevention of apoptosis, so they examined the effect of the hGH antagonist, B2036 on the number of cells in S-phase and the number of apoptotic cells in
MCF7-MUT and MCF7-hGH. Treatment of MCF7-hGH cells with B2036 reduced the level of the cells in S-phase to that observed in MCF7-MUT cells. In addition, B2036 inhibited the protection from apoptosis afforded by autocrine hGH. B2036 was also used to check whether autocrine hGH-stimulated transcriptional activation mediated by
STAT5, CHOP and Elk-1 is via interaction with the hGH receptor. Treatment of MCF7- hGH cells with 600nM B2036 abrogated the ability of autocrine hGH to stimulate
STAT5, CHOP and Elk-1 mediated transcription. B2036 also had a dose-dependent inhibition of the ability of autocrine hGH to enhance the rate of cell spreading (Kaulsay et al., 2001).
In conclusion, these studies using the hGH specific antagonist B2036 demonstrated that the effects of autocrine hGH on human mammary carcinoma cells described above are mediated specifically via the hGH receptor.
2.3.4 Identification of genes regulated by autocrine production of hGH in human mammary carcinoma cells
Autocrine production of hGH by mammary carcinoma cells may direct mammary carcinoma cell behavior such as cell proliferation, apoptosis, cell spreading etc (Kaulsay et al., 1999; Kaulsay et al., 2000; Kaulsay et al., 2001). One major mechanism by which
28
GH affects cellular function is by regulating the level of specific mRNA species
(Isaksson et al., 1985). So it is likely that many of the effects of autocrine hGH on
mammary carcinoma cell function are mediated by specific regulation of certain genes.
Continuing the previous work, Mertani et al. used a cDNA microarray to determine the
effect of autocrine productin of hGH on gene expression in human mammary carcinoma
cells (Mertani et al., 2001). Of 588 screened genes, 24 exhibited a 2-fold or greater
increase in their expression in MCF7-hGH cells (Figure 3 and Table 1) (Mertani et al.,
2001). Most of the up-regulated genes were among those in the regulation of DNA
synthesis/repair/recombination proteins. Of note was the dramatic effects of autocrine
hGH on the expression of DNA repair helicase (ERCC3), ATP-dependent helicase II,
DNA repair protein XRCC1, superoxide dismutase, UV excision repair protein RAD23,
growth arrest, and DNA damage-inducible protein (GADD) 45 and GADD153 (otherwise
known as CHOP). Other up-regulated genes of particular interest included the ski
oncogene, homeobox containing transcription factor HOXA1, and β-catenin (Mertani et
al., 2001). Twenty eight genes exhibited a 2-fold or greater decrease in their expression in
the presence of autocrine hGH (MCF7-hGH cells) (range 2-20-fold) compared with the
absence of autocrine hGH (MCF7-MUT cells) (Figure 3 and Table 2) (Mertani et al.,
2001). Most of the down-regulated genes were to be found among those grouped as DNA binding/transcription/transcription factors. The relative expression of housekeeping genes
(ubiquitin, phospholipase A2, glyceraldehyde-3-phosphate dehydrogenase, β-actin, -
tubulin, 23-kDa highly basic protein, ribosomal protein S9) did not differ by more than
10% between MCF7-MUT and MCF7-hGH cells (Mertani et al., 2001).
29
Figure 3 Effect of autocrine production of hGH on relative levels of gene expression in mammary carcinoma cells. cDNA microarray analysis of the relative levels of gene expression in MCF-7 cells stably transfected with the hGH gene but with the start codon mutated to TTG (MCF7-MUT) (A) or in MCF-7 cells stably transfected with the hGH gene (MCF7-hGH) (B) cultured in serum-free medium. 32P-Labeled cDNA probes generated from poly(A)+ RNA isolated from MCF7-MUT and MCF7-hGH cells were hybridized to a cDNA microarray containing 588 known human genes. The left upper box and left lower box encase those genes grouped as oncogenes/tumor suppressors/cell cycle control proteins and DNA binding/transcription/transcription factors, respectively. The right upper box encases those genes grouped as DNA synthesis/repair/recombination proteins. The position of CHOP cDNA is indicated by the arrow. The relative expression level of specific cDNAs was determined by comparison with the expression of a number of housekeeping genes (Mertani et al., 2001).
30
TABLE 1 Identification by cDNA array of genes positively regulated by the autocrine production of hGH in mammary carcinoma cells. Genes were considered positively regulated when they exhibited a 2-fold or greater increase in the presence of autocrine hGH (MCF7-hGH cells) compared to the absence of autocrine hGH (MCF7- MUT cells) in three independently performed experiments. Fold regulation is calculated as the ratio of the level of gene expression in MCF7-hGH cells to that in MCF7-MUT cells (Mertani et al., 2001).
GeneBank number Name -Fold stimulation
M15024 Myb proto-oncogene protein 3.05 X15218 SKI oncogene 4.10 X03484 raf proto-oncogene 2.00 D10924 HM89 (probable G protein-coupled receptor LCR1 homolog) 3.45 M74816 Clusterin 2.25 M31899 DNA repair helicase [ERCC3] 12.50 M32865 ATP-dependent DNA helicase II 6.10 M36089 DNA repair protein XRCC1 5.55 M74524 Ubiquitin-conjugating enzyme E2 3.50 M29971 Methylated DNA protein cysteine methyltransferase 2.85 K00065 Superoxide dismutase 13.45 U07418 DNA mismatch repair protein hMLH1 2.30 D21090 UV excision repair protein RAD23 (HHR23B) 2.45 D21235 UV excision repair protein RAD23 (HHR23A) 3.75 M60974 Growth arrest and DNA damage-inducible protein GADD45 3.95 S40706 Growth arrest and DNA damage-inducible protein GADD153 5.10 M97287 DNA-binding protein SATB1 3.20 U10421 Homeobox protein HOXA1 8.20 X67951 Proliferation-associated protein PAG 2.15 U14722 Activin type 1 receptor 4.10 X52425 Interleukin-4 receptor α chain 2.15 M21097 CD19 B-lymphocyte antigen 4.90 Y00796 Integrin α L 3.70 X87838 β-Catenin 3.60
31
TABLE 2 Identification by cDNA array of genes negatively regulated by the autocrine production of hGH in mammary carcinoma cells. Genes were considered negatively regulated when they exhibited a 2-fold or greater decrease in the presence of autocrine hGH (MCF7-hGH cells) compared with the absence of autocrine hGH (MCF7- MUT cells) in three independently performed experiments. Fold regulation is calculated as the ratio of the level of gene expression in MCF7-hGH cells to that in MCF7-MUT cells (Mertani et al., 2001).
GeneBank number Name -Fold stimulation
J04111 Transcription factor AP-1 (c-jun proto-oncogene) 0.10 L25080 Transforming protein RhoA 0.20 X02751 Transforming protein p21 (n-Ras) 0.25 M26708 Prothymosin-α 0.40 M25753 Cyclin B1 0.30 U10564 CDK tyrosine 15-kinase; wee1Hu 0.30 U41816 C-1 0.40 M33294 Tumor necrosis factor receptor 1 0.50 J03746 Glutathione S-transferase microsomal 0.40 M68891 Endothelial transcription factor GATA-2 0.15 M59079 CCAAT-binding transcription factor subunit 0.10 X69111 DNA-binding protein inhibitor ID-3 0.15 M95809 Basic transcription factor (62kDa subunit; BTF-2) 0.30 M97796 DNA-binding protein inhibitor ID-2 0.05 X69391 60 S ribosomal protein L6 (DNA-binding protein TAX) 0.40 M28372 Sterol regulatory element-binding protein 0.10 L12579 CCAAT displacement protein 0.20 D26156 Transcriptional activator hSNF2b 0.30 D90209 cAMP-dependent transcription factor ATF-4 0.10 L05515 cAMP-response element-binding protein CRE-BPA 0.05 L14922 Activator 1 140-kDa subunit 0.40 D28118 DB1 (zinc finger protein 161) 0.20 M36711 Transcription factor AP-2 0.40 M76541 Transcriptional repressor protein YY1 0.50 U30504 Transcription initiation factor TFIID subunit TAFII31 0.20 M29366 Protein-tyrosine kinase receptor ERBB-3 0.45 S59184 Related to receptor tyrosine kinase (RTK) 0.45 M14200 Acyl-CoA-binding protein 0.35
32
As the work presented in this thesis will be a study involving the homeobox gene,
HOXA1, the literature review will now concentrate on the homeobox gene and mammary
gland carcinoma.
2.4 Homeobox genes
Homeobox genes are a family of regulatory genes encoding specific nuclear proteins
(homeoproteins) that act as transcription factors. Homeobox genes share a common 183-
bp DNA sequence motif (the homeobox) encoding the roughly 61 amino acid
homeodomain (HD). There are different homeobox gene classes based on the sequence
similarities within the HDs and flanking regions (Duboule, 1994). Among these, the
Drosophila Antennapedia (Antp) HD defines a consensus sequence referred to as class I
HD (Hox genes) (Akam, 1987). Other homeobox gene families are Bicoid (Bcd), Caudal
(Cad), Engrailed (En), Even-skipped (Eve), muscle segment homeobox genes (Msx),
Paired (Pax), Pit-Oct-Unc (Pou), empty spiracles (Emx), and Orthodenticle (Otx) etc.
(Duboule, 1994). These families vary in size from the relatively large ones, such as the
Hox group, that comprises 39 members, to the small families, such as the Msx and En groups, which only have two or three members each.
The homeobox was initially identified in genes controlling Drosophila development
(Lewis, 1978). Later, it was identified in evolutionarily distant animal species (Kappen,
2000), plants (Parcy et al., 1998) and fungi (Torres-Guzman and Dominguez, 1997). The sequence identity between vertebrate and Drosophila homeoboxes was astounding. For example, the human HOXB7 homeodomain differs at only 1 out of 60 amino acids from
33
that of Antp in Drosophila (Gehring, 1993). Novel and divergent homeobox genes are
being continuously isolated. There are over 100 homeobox genes identified in the human,
with a comparable number of homologous identified in the mouse (Stein et al., 1996).
Recent indications suggest that homeobox genes constitute as much as 0.1-0.2% of the
whole vertebrate genome (Stein et al., 1996).
Class I homeobox genes are structurally and functionally homologous to the homeotic
complex (HOM-C) of Drosophila (Krumlauf, 1994). Analysis of the mouse and human
Hox complex is the most detailed and indicates that there are at least 39 genes organized
in four different chromosomal complexes approximately 100 kb in length, termed Hox
loci, and that the genes in each cluster are still oriented in the same 5’ to 3’ direction of transcription (Kessel and Gruss, 1990; Duboule, 1992; Krumlauf, 1992). Each Hox loci is localized on a different chromosome (HOXA at 7p15.3, HOXB at 17p21.3, HOXC at
12q13.3, and HOXD at 2q31) (Apiou et al., 1996) and comprises from 9 to 11 genes
arranged in a homologous sequence organization (Scott, 1992) (see Figure 4) (Mark et
al., 1997). Hox genes in the four clusters can be aligned to any one of 13 homology
groups on the basis of sequence similarity and position on the locus (Acampora et al.,
1989). A distinguishing characteristic of the Hox complex is the correlation between the
physical order of genes along the chromosome and their expression/function along the
anteroposterior axis of the embryo. The 3 Hox genes express early in development and
control anterior regions, followed by progressively more 5 genes express later in
development and control more posterior regions (Dekker et al., 1992). In particular, 3
Hox genes in groups 1 to 4 primarily control the development of the branchial area and
34
Head Thorax Abdomen
lab pb Dfd Scr Antp Ubx abd-A Abd-B HOM-C (Drosophila) A1 A2 A3 A4 A5 A6 A7 A9 A10 A11 A13 HOXA B1 B2 B3 B4 B5 B6 B7 B8 B9 B13 HOXB HOXC C4 C5 C6 C8 C9 C10 C11 C12 C13
D1 D3 D4 D8 D9 D10 D11 D12 D13 HOXD cervical thoracic lumbo-sacral 3’anterior 5’ posterior Early Late
Figure 4. Genomic organization and collinear expression patterns of Drosophila HOM genes and mammalian Hox genes. Schematic representation of the Drosophila homeotic complex (HOM-C) and the four human Hox complexes is displayed showing their possible phylogenic relationships. Each gene is represented by a colored box. The expression domains of HOM/Hox genes are schematized in a fly and in the CNS and prevertebrate of a human fetus (extrapolated from data in the mouse). For the sake of clarity, the partial overlap between HOM gene transcripts in thoracic and abdominal segments of the fly and overlapping expression domains of mammalian Hox genes along the body axis are not represented; hence, each color is meant to show the anteriormost expression domain of a given subfamily (modified from Mark et al., 1989).
35
the rhombencephalon, the embryonic region corresponding to the hindbrain (Lumsden
and Krumlauf, 1996). Central Hox genes in groups 5 to 8 control the thoracic portion of
the body, whereas 5 Hox genes in groups 9 to 13 control the lumbo-sacral region.
The nomenclature of clustered vertebrate Hox genes was determined and agreed at a
meeting in Ascona, Switzerland in 1992. The names are a single letter (A,B,C,or D)
followed by a number from 1 to 13, starting from the ends where the genes express most anteriorly are located. Sequence comparisons permit the assignment of the constituent genes of each complex to 1 of 13 groups (Figure 4). “Paralog” is sometimes used to indicate corresponding genes in different complex, e.g., A1 and B1. The complete name
of a gene therefore will be Hoxa1, Hoxc4, etc. in mice, and HOXA1, HOXD2, etc. in
humans. The name Hox should not be used for homeobox genes outside complex. In other words, the divergent homeobox genes are not classified and named as “Hox” genes.
The Hox network, as a decoding system for external signals, allows the activation of specific genetic programs (Cillo et al., 2001). The external signals from growth factors and signal transduction pathways are collected by some collector genes in the network.
The decoder genes transfer these signals to key genes in the decoding network. Through transmitter genes, a response is sent outside the Hox network. This activates specific programs of effector genes, e.g. cell-cycle related proteins to achieve the changes induced by the signal received (Cillo et al., 1999).
36
2.5 Homeodomain (HD)
Homeobox-containing genes were first recognized in the fruitfly Drosophila
melanogaster where duplication caused homeotic transformation such as a fly with four
wings instead of two (Lewis EB. 1978). In 1984, the homeobox motif was discovered in
genes of the Drosophila homeotic (HOM-C) complexes (McGinnis et al., 1984; Scott and
Weiner, 1984). The term “homeodomain” has evolved to define a class of protein
domains that have a 61 amino acid motif (encoded by 183 bp homeobox sequences).
Those homeodomains that have been tested contain sequence-specific DNA binding
activities and are part of larger proteins that function as transcriptional regulators (Levine and Hoey, 1988).
Homeodomains have a high degree of conservation throughout the animal and plant kingdom. Homeodomains form parts of much larger proteins, and are usually located near the C-terminus of the proteins. Generally, there is little similarity between the primary sequences of the proteins outside the homeodomain.
There are three features of homeodomains. a) In primary sequence, all the homeodomains exhibit striking similarities, even without the introduction of gaps in any of the domains.
These structural similarities in large part define the homeodomain. Four amino acid residues are conserved in all non-yeast homeodomains and three of the four are also conserved in the yeast sequence. The invariant residues are tryptophan, phenylalanine, asparagine and arginine at positions 49, 50, 52 and 54, respectively. Another eight positions are very highly conserved, though not invariant. These are positions 6
37
(arginine/glycine), 13(usually glutamine), 17(leucine/valine), 21(phenylalanine/tyrosine),
41(leucine/asparagine), 46(isoleucine/valine), 56(lysine/arginine), and 58(lysine/arginie).
The 12 extreme highly conserved residues, together with the high degree conservation at other positions characterize the homeodomains. b) There is also considerable
conservation of the secondary structure in homeodomains, even in regions that tend to be
relatively variable in primary sequence. There are predominant α-helix structures in some
specific regions of the homeodomains, which has important functional implications. c)
Another distinguishing characteristic is the nuclear location of the homeodomain-
containing proteins. This nuclear distribution is consistent with their role as
transcriptional regulators. Thus, the primary sequence and secondary structure, as well as
the subcellular distribution are three important features in determining whether a
homeodomain-related sequence is in fact a true homeodomain.
Recent nuclear magnetic resonance and crystallographic studies on two homeodomains
have shown their structures to be related to the helix-turn-helix motif of prokaryotic
DNA-binding proteins (Otting et al., 1990; Kissinger et al., 1990). The first helix is eight
amino acids long and the second helix is nine amino acids long. The critical residues in
the helix-turn-helix framework are a hydrophobic residue at the fourth and eighth
positions in helix 2, an alanine in the fifth position in helix 2, a glycine (or sometimes
cysteine or serine) in the first position of the turn followed by a hydrophobic residue, an
isoleucine or valine at the fourth position of the recognition helix, and a hydryophobic
residue at the seventh position of the recognition helix. There are no helix-breaking
38 residues in the predicted helical regions of homeodomain sequences. All of these characteristics are common to the homeodomain sequences.
Homeodomains are responsible for recognizing and binding specific DNA sequences
(McGinnis and Krumlauff, 1992). Many of these sequences contain the motif TAAT. The specificity of this binding allows homeoproteins to activate or repress the expression of batteries of downstream effector target genes (Boersma et al., 1999).
2.6 Homeobox genes and development
Homeobox genes (clustered Hox genes and divergent homeobox genes) play a key role in a variety of developmental processes including central nervous system and skeletal development, limb and digit specification, and organogenesis (Edelman et al., 1993). In mammalian embryos, the earliest expression of Hox genes can be detected at gastrulation.
Hox genes are expressed in all three germ layers with overlapping domains that extend from the caudal end of the embryo to a sharp anterior limit which is specific for each Hox gene. Mutations of these genes cause dramatic developmental defects including loss of specific structures as well as homeotic transformation in which the anatomy characteristic of one segment in a given location is assigned incorrectly to another segment (Edelman et al., 1993).
2.6.1 Hox genes in axial skeleton patterning
39
The homeobox genes are the major regulator genes in the specification of the body plan
along the antero-posterior axis both in invertebrates and in vertebrates.
In Drosophila, ectopic expression of a homeotic gene often results in posterior homeotic
transformation. For example, gain-of-function mutants of Ubx lead to the transformation
of the second into the third thoracic segment with the transformation of the wings into a
second pair of halteres (Gehring, 1994). Similarly, ectopic expression of the
Antennapedia (Antp) gene leads to the conversion of the antenna into mesothoracic
(second thoracic) legs (Gehring, 1994).
The sequence homology between Drosophila homeotic genes and vertebrate Hox genes,
and their ordered gene organization, raised the possibility that these genes may have a
homeotic function in vertebrates as well. The function of Hox genes in vertebrate
development was analyzed by constructing both gain- and loss-of-function mutants and
testing them in transgenic mice, and by insertion of the mammalian genes into
Drosophila. In 1991, Kessel and Gruss constructed an artificial gain-of-function mutant by fusing the Hoxa7 gene to a ubiquitously expressed actin promoter (Kessel and Gruss,
1991). The resulting transgenic mice had an extra cervical vertebra (a proatlas) which arose from conversion of an occipital bone of the skull into an eighth cervical segment
(Kessel and Gruss, 1991). This conversion represented an anterior to posterior transformation. To date, numerous loss-of-function mutations in mammalian Hox genes have been reported. The mice with targeted disruption of Hoxd3 exhibit anterior transformations of the first and second cervical vertebrae, the atlas and the axis (Condie
40 and Capecchi, 1993). Hoxb4 mutant mice show homeotic transformation of a cervical vertebra and defects in the closure of the sternal rudiments (Ramirez-Solis et al., 1993).
Hoxc8 inactivation in mice resulted in the conversion of the first lumbar vertebra to a thoracic vertebra, thus producing a supernumerary 14th pair of ribs (Le Mouellic, 1992).
Hoxd11 mutant mice show transformation of the first sacral into a lumbar vertebra
(Favier et al., 1995; Favier et al., 1996). Null mutant mice for a given Hox gene usually show homeotic transformations in the anteriormost region where that Hox gene is normally expressed, and not within regions where a more 5’ gene is expressed (Mark et al., 1997). Homeobox genes in mammals are thought to specify segmental identity along the antero-posterior axis in much the same way as in flies. This interpretation receives further support from gene-transfer experiments between mouse and Drosophila. For example, the Hoxb6 of the mouse can induce antennal legs when ectopically expressed in
Drosophila (Malicki et al., 1990).
2.6.2 Hox genes in limb morphogenesis and patterning of reproductive and digestive tracts
The mammalian Hox gene family contains 15 genes related to the Drosophila gene, Abd-
B. They are Hoxa9 to Hoxa11, Hoxa13, Hoxb9, Hoxb13, Hoxc9 to Hoxc13, Hoxd9 to
Hoxd13 (Mark et al., 1997). Most of these genes are expressed with overlapping domains in the developing fore and hind limbs consistent with a role in specification of the digit pattern (Duboule, 1991). Knock-out experiments of these murine Abd-B-related genes resulted in size reduction, changes in shape, and/or delayed ossification of skeletal
41
elements of both fore-limb (e.g. Hoxa9, Hoxd9, Hoxd11) (Favier et al., 1995; Fromental-
Ramain et al., 1996; Davis and Capecchi, 1994; Davis and Capecchi, 1996) and hind limb
(e.g. Hoxa10) (Favier et al., 1996). These results indicate that these genes control initially
the allocation and growth of prechon-drogenic condensations and, subsequently, the
ossification sequence of the cartilage models (Favier and Dolle, 1997; Mark et al., 1997).
Compound Hox mutant phenotypes have revealed partial redundant functions in limb
patterning, as for vertebral specification, between paralogous and nonparalogous genes
(Fromental-Ramain et al., 1996; Davis et al., 1995; Favier et al., 1996). For instance,
Hoxa13/Hoxb13 compound mutants displayed alterations of growth and patterning of the
autopods much more severe than those observed in each single mutant (i.e. almost
complete lack of chondrogenic patterning) showing that these gene products can also
partly compensate each other (Fromental-Ramain et al., 1996; Mark et al., 1997).
It was also demonstrated that other Abd-B-related Hox genes play roles in the developing
genitourinary tract and terminal part of the digestive tract. Both male and female Hoxa10
null mutant mice were hypofertile. Mutant male subjects exhibited cryptorchidism,
caused by abnormal formation of the inguinal canal and by a failure of shortening of the
gubernaculums (Rijli et al., 1995; Satokata et al., 1995). In addition, they displayed a
malformation of the vas deferens that resembled a partial homeotic transformation to an epididymis (Benson et al., 1996). In Hoxa11 null mutant mice, partial homeosis of the vas deferens to an epididymis-like morphology was also observed (Hsieh-Li et al., 1995).
The most posterior murine gene, Hoxd12 and Hoxd13, were also demonstrated to have a specific function in the morphogenesis of the terminal part of the digestive tract (Mark et
42 al., 1997). Hoxd12 and Hoxd13 null mutants display a disorganization of the smooth muscles layers forming the internal anal sphincter, resulting in rectal prolapsus in some mutants (Kondo et al., 1996). These data clearly indicate an important function of some
Hox genes in defining regional organ development.
2.6.3 Hox genes in hindbrain patterning and craniofacial development
The hindbrain or rhombencephalon, is transiently divided along its anteroposterior axis into a series of segments (seven in mice and man) called rhombomeres (Keynes and
Lumsden, 1990). Neural crest cells migrate across these segments from the neurectoderm to populate and pattern the pharyngeal arches. It was shown that some Hox genes which are expressed in a given rhombomere are also expressed in the neural crest cells migrating from that rhombomere, suggesting that Hox genes may be important in the patterning of the branchial region of the head (Krumlauf, 1993).
Defects of rhombomeres 4 and 5, cranial nerve and inner ear abnormalities were found in
Hoxa1 null mutant mice (Mark et al., 1993; Carpenter et al., 1993). Hoxa2 mutant mice exhibit homeotic transformation of skeletal elements derived from cranial neural crest
(Gendron-Maguire et al., 1993). Targeted inactivation of Hoxa3 leads to hypoparathyroidism and thymic and thyroid hypoplasia (Manley and Capecchi, 1995). In
Hoxb1 and Hoxb2 knockout mice, selective facial nerve motor nucleus deficiencies were observed (Goddard et al., 1996; Studer et al., 1996; Barrow and Capecchi, 1996).
43
2.6.4 Divergent homeobox genes in normal and abnormal head development
Apart from the clustered Hox genes, divergent homeobox genes are also involved in development. In Drosophila, the control of head specification is dependent on the divergent homeobox genes (genes not present within a cluster) including the orthodenticle and empty spiracles genes (Mark et al., 1997). The vertebrate homologues,
Emx1, Emx2, Otx1 and Otx2 genes, are expressed in discrete, overlapping regions of the
developing forebrain and midbrain in mice, which often coincide with anatomical
landmarks. Null mutant mice with these genes inactivated showed that they play an
important role in the patterning of the forebrain and midbrain (Acampora et al., 1995;
Ang et al., 1996; Pellegrini et al., 1996).
The Engrailed homeobox genes En1 and En2 are both expressed in a domain spanning
the first rhombomere and the midbrain. The knockout experiments have demonstrated
that En1 has a critical role in its entire region of expression, whereas En2 function is restricted to cerebellar foliation (Mark et al., 1997). However, insertion of the En2 cDNA into the En1 locus rescues the En1 null phenotype, agenesis of the tectum and cerebellum, suggesting that the distinct phenotype of the En1 and En2 mutations reflect differences in the temporal expressions of the corresponding proteins, rather than differences in their biochemical activity (Hanks et al., 1995).
Expression of Msx1 and 2 in the mouse fetus and the phenotype displayed by Msx1 null
mice (anodontia, i.e. complete tooth agenesis) supports a critical role for these genes in
44
early stages of odontogenesis (Thesleff and Nieminen, 1996). These and other homeobox
genes such as Distall-less 1 and 2 (Dlx1, Dlx2) and Goosecoid (Gsc), are expressed in
restricted overlapping fields in the developing mandible. Based on these findings an
“odontogenic homeobox code” that specifies tooth position and type (incisor, canine,
premolar, molar) has been proposed (Sharpe, 1995).
2.7 Cellular effects and targets of homeobox genes
2.7.1 Cell cycle
In addition to the functions already described and reviewed, several homeobox genes are
also implicated in the regulation of events in cell cycle progression.
The product of the divergent homeobox gene Msx1 was shown to upregulate cyclin D1 expression and Cdk4 activity, so preventing cells from leaving the cell cycle and undergoing differentiation (Hu et al., 2001). The homeoprotein, HSIX1, was demonstrated to abrogate the DNA-damage-induced G2 cell-cycle checkpoint (Ford et al.,
1998). The divergent homeobox gene HOX11 was demonstrated to cause cell cycle aberrations in transgenic mouse models when overexpressed in the thymus (Kawabe et al.,
1997). HOX11 can bind to the catalytic subunit of both protein serine-threonine
phosphatase 2A and protein phosphatase 1 to promote progression into M-phase in both
G2 arrested xenpus oocytes and in Jurkat cells arrested in G2 following DNA damage
(Kawabe et al., 1997).
45
In addition, it was shown that the CDX2 homeoprotein transactivated and physically
interacted with the promoter of p21 in a p53-independent manner. Moreover,
overexpression of CDX2 increased the mRNA expression of p21 in HT-29 colon
carcinoma cells (Bai et al., 2003). The GAX homeoprotein induces p21 expression,
which results in an increased association of p21 with Cdk2, thereby inhibiting cellular proliferation (Smith et al., 1997). In addition, the Cdk1 gene was shown to contain a homeoprotein binding site in its promoter (Liu and Bird, 1998). HOXA10 was shown to bind directly to the p21 promoter and activated p21 transcription (Bromleigh and
Freedman, 2000).
2.7.2 Apoptosis
Several homeobox genes have been associated with apoptosis. Hoxa9, crucial for the control of hematopoiesis, induces programmed cell death in primitive thymocytes (Izon et al., 1998). Inhibition in the expression of the DLX7 homeobox gene renders erythropoietic cells apoptotic (Shimamoto et al., 1997). A block in the expression of
Cux1, a homeobox gene involved in mammalian kidney development, results in increased apoptosis of embryonal kidney cells (Quaggin et al., 1997). The homeoprotein,
PAX5, is able to bind the 5’ regulatory region of the human p53 gene, at a vital site for
P53 activity (Stuart et al., 1995). Recently, it was reported that HOXA5 is able to regulate p53 transcription in human breast cancer cells through binding at the p53 promoter (Raman et al., 2000). Expression of HOXA5 in epithelial cancer cells
46
expressing wild-type p53, but not in isogenic variants lacking the p53 gene, led to apoptotic cell death (Raman et al., 2000).
2.7.3 Adhesion
Interactions between cytokines and growth factors, homeoproteins, and adhesion molecules are crucial for the living cell (Edelman and Jones, 1993). A link was demonstrated between the expression of homeobox genes and the expression of cell adhesion and substrate adhesion molecules (CAMs and SAMs). CAMs and SAMs have been found to modulate cellular primary processes such as proliferation, movement, interaction, differentiation and death (Edelman, 1988). The promoter region of the cytotactin gene, an extracellular matrix glycoprotein, exhibits many potential transcription factor binding motifs including homeodomain binding sites (Jones et al.,
1990). The gene promoter for the mouse neural cell adhesion molecule (N-CAM) has also been found to contain potential homeodomain binding sites (Hirsch et al., 1990).
Later, it was shown that the product of homeobox gene Evx1, activates the cytotactin gene and the specific homeodomain binding site was also found in the promoter of the gene (Jones et al., 1992). In addition, N-CAM gene expression is controlled by different
Hox gene products. Expression of Evx1, Hoxb8, and Hoxb9 are able to interact with the
promoter region of N-CAM genes by mechanisms involving a growth-factor signal
transduction pathway (Edelman, 1993; Jones et al., 1992). Similarly, a homeoprotein
encoded by the Hoxc6 gene was bound to a specific TAAT-containing sequence
CCTAATTATTAA in the N-CAM promoter (Jones et al., 1993). In addition, both
47
homeobox genes, Msx1 and Msx2, are able to regulate cadherin-mediated cell adhesion
(Lincecum et al., 1998).
2.7.4 Extracellular matrix
Extracellular matrix (ECM) molecules play an important role in determining cell phenotypes through modulation of gene expression (Srebrow et al., 1998). Recent findings support an interaction between homeobox genes and ECM. Expression of
HOXD3 increases 5 3 integrin expression and the matrix-degrading proteinase
urokinase-type plasminogen activator (uPA) (Boudreau et al., 1997). HOXC10, 11, 13
are connected with decreased expression of 2 1, 5 1, and 6 1 (Cillo et al., 1996).
Reduced expression of Hoxb9 has been observed in 5 integrin null mice embryos (Goh
et al., 1997).
2.7.5 Specific target genes of homeoproteins in Drosophila
The selector homeoproteins are conserved throughout the Eumetazoa and include the
Hox (or homeotic) proteins, the Engrailed-like proteins, and the Eve-like proteins
(Burglin, 1994; Raff, 1996; Kuhn et al., 1996). These transcription factors are each
expressed in unique domains along the anterior/posterior axis of animals, and they
coordinate the development of most or all cells within these domains (Garcia-Bellido,
1975; Lawrence, 1992; Biggin and McGinnis, 1997). It is shown that much of the
48
Drosophila genome is under homeobox gene control (Mannervik, 1999). In vivo UV crosslinking suggested that the selector homeoproteins bind at significant levels to a majority of genes in Drosophila (Walter et al., 1994). Later, it was shown that the majority of Drosophila genes, at least 87% of genes, are directly or indirectly regulated by the homeoproteins and each downstream gene shows a unique pattern of expression, indicating that all are under precise and specific control (Liang et al., 1998).
2.7.6 Specific target genes of HOXB7 and Hoxa1
It was demonstrated that HOXB7 directly transactivates the basic fibroblast growth factor
(bFGF) gene through one out of five putative homeodomain binding sites present in its promoter (Care et al., 1996). Treatment of melanoma cell lines with antisense oligomers targeting HOXB7 mRNA specifically abolished expression of bFGF mRNA (Care et al.,
1996). In another study, it was shown that vascular endothelial growth factor, melanoma growth-stimulatory activity/growth-related oncogene α, interleukin 8, and angiopoietin 2 were upregulated by HOXB7 transduction (Care et al., 2001). In the same study, several enzymes were also analyzed including matrix metalloprotease (MMP)-2 and MMP-9 and heparanase, capable of proteolytic degradation of extracellular matrix and basement membranes. Results showed an induction of only MMP-9 (Care et al., 2000).
In order to clone and characterize target genes of Hoxa1, Shen et al. utilized differential hybridization screening and cDNA subtractive hybridization methods to identify genes
49
which are differentially expressed in F9-10, a murine F9 teratocarcinoma stem cell line
which expresses high levels of exogenous Hoxa1, compared to F9 wild-type stem cells,
which do not express endogenous Hoxa1 mRNA in the absence of retinoid acid (Shen et al., 2000). A wide range of genes were isolated, including signaling molecules such as
BMP-4 and BMP-2, the enzyme superoxide dismutase, the cell-cell adhesion molecule cadherin-6, proteins involved in gene transcription such as HMG-1 and SAP18, homeodomain-containing proteins Gbx2 and Evx2, and cell cycle regulatory proteins such as the retinoblastoma binding protein-2 (Table 3) (Shen et al., 2000).
Eph receptor tyrosine kinases and their membrane-bound ligands, ephrins, form a large family of molecules that are involved in diverse developmental and differentiation processes (Frisen et al., 1999; Holder and Klein, 1999). There is accumulating evidence that homeobox genes can regulate Eph family members function. For example, reduced
EphA2 expression was observed in Hoxa1 and Hoxb1 double mutant mice. Co- expression of either Hoxa1 or Hoxb1 with Pbx1 transactivated EphA2 enhancer- dependent reporter gene expression. These results suggest that expression of EphA2 gene
in rhombomere 4 is directly regulated by Hoxa1 and Hoxb1 homeobox transcription
factors (Chen and Ruley, 1998). In addition, Hoxa2 was shown to regulate EphA7
expression (Taneja et al., 1996).
50
Table 3 Clones of genes that are differentially expressed in Hoxa1-overexpressing F9 cells (F9-10 line) vs F9 wild-type (WT) cells (modified from Shen et al., 2000).
Clone number Fold of change of expression Blast search in F9-10 line vs F9 WT (from NCBI)
2 + 3.2 Novel gene 3 + 3.0 Novel gene 5, 26 - 9.3 SOD1 13 - 5.6 BMP-4 21 + 3.2 BMP-2 22 + 4.4 Retinoblastoma binding protein-2 27 + 4.0 Lactate dehydrogenase A4 29, 102 + 3.0 Mitochondrial genome 32 - 2.1 IGF binding protein 6 42 + 4.0 HMG-1 60 - 2.0 Novel gene 62 + 2.0 Novel gene 63 -12.0 eEf-1-α 64 - 4.0 Novel gene 80 - 8.4 Novel gene 100 + 3.2 Cadherin-6 101 + 16.0 CTD-binding SR-like protein rA8 103 + 32.0 Novel gene 104 + 36.0 Novel gene 112 - 6.4 Sin3-associated polypeptide p18 127 + 4.0 Proteasome subunit RC6-1 131 + 27.0 Novel gene 157 + 28.2 Novel gene 168 + 4.0 Phosphoglycerate kinase Pgk1-ps1 172 + 11.2 Gbx2 174 + 13.4 Evx2
51
2.8 Homeobox genes and human diseases
Emerging evidence supports a role of HOX genes in human diseases and development
and growth. In 1996, a human malformation syndrome was first shown to be caused by
mutations in a HOX gene. This syndrome was synpolydactyly and the gene in question
was HOXD13 (Muragaki et al., 1996). Synpolydactyly is a rare, dominantly-inherited
limb malformation in which there is a combination of syndactyly and polydactyly. One to
four limbs can be involved. Typically, patients have syndactyly between the third and the
fourth fingers, and between the fourth and fifth toes. There may also be clinodactyly, camtodactyly or brachydactyly of the fifth fingers, and variable syndactyly of the second to fifth toes, often with middle phalanxhypoplasia. Severely affected males can also have hypospadias. Mutation in HOXD13 was found to be responsible for synpolydactyly
through the expansion of an alanine stretch in the corresponding homeoprotein (Muragaki
et al., 1996).
In 1997, hand-foot-genital (HFG) syndrome was shown to be caused by a mutation in
HOXA13 (Mortlock and Innis, 1997). Hand-foot-genital syndrome is another rare,
dominantly-inherited condition, in which distal limb abnormalities are accompanied by
malformations of the lower genito-urinary tract. In the limbs, the striking feature is first
digit hypoplasia, producing short, proximally placed thumbs with hypoplastic thenar
eminences and short, medially deviated halluces. These abnormalities are fully penetrant,
bilateral and symmetrical with little variation in severity. In contrast, the genito-urinary
tract malformations have incomplete penetrance and vary in the severity. Genital
52
abnormalities include hypospadias in males and Mulllerian duct fusion defects in
females. The mutations which cause hand-foot-genital (HFG) syndrome include the
nonsense mutation in the homeobox and the polyalanine tract expansion (Mortlock and
Innis, 1997).
De novo mutations in the human homeobox gene EMX2 have been reported in patients
with schizencephaly, an extremely rare congenital disorder characterized by a full-
thickness cleft within the cerebral cortex; eventually, large portions of the cerebral
hemispheres may be lacking, resulting in a holohemispheric cleft filled with
cerebrospinal fluid. The phenotype of these patients suggests a requirement of the EMX2
protein for the correct formation of the cerebral cortex (Brunelli et al., 1996).
Many PAX homeobox genes are connected with human diseases. PAX2 is connected to
renal defects (Dressler and Douglass, 1992). In Waardenburg syndrome (WS), a
dominantly-inherited syndrome associated with sensorineural hearing loss and
pigmentary disturbances, loss-of-function mutations of the PAX3 have been found in WS
type I and WS type III (Stuart et al., 1993). Heterozygotic mutations in PAX6 have been
reported in families with eye defects such as 1) aniridia, a panocular disorder in which the
development of the iris, cornea lens, and retina are disturbed; 2) Peters’ anomaly, a defect
of the anterior chamber of the eye with corneal malformations and attachment of the lens
to the central aspect of the cornea; and 3) isolated foveal hypoplasia (Glaser et al., 1994;
Azuma et al., 1996).
53
It was demonstrated that selective tooth agenesis (i.e. lack of all permanent second premolars and third molars) in a family with autosomal dominant tooth agenesis was caused by a single point mutation in the human MSX1 gene (Davideau et al., 1999).
Mutation analysis in families with Rieger’s syndrome (an autosomal-dominant disorder characterized by hypodontia, abnormalities of the anterior chamber of the eye, and a protuberant ombilicus) has led to the identification of a novel homeobox gene, RIEG, whose mutations are responsible for the abnormalities observed in the Rieger syndrome
(Semina et al., 1996).
One member of the POU family of homeoproteins, Pit1, is synthesized only in the anterior pituitary gland. Pit1 regulates expression of growth hormone, prolactin, and β- glycoprotein subunit of the TSH-β. Pit1 mutations have been identified in patients with combined pituitary hormone deficiency characterized by lack of GH, prolactin, and TSH, resulting in mental retardation and growth deficiency (Rhodes et al.,1994). The mutant
Pit1 can still bind to its DNA-binding site in target genes, but unlike the normal transcription factor, it does not activate transcription and it prevents the normal protein from binding to DNA (Rhodes et al.,1994).
Mutations in the POU3F4 homeobox gene cause deafness with fixation of the stapes, which represents the most frequent X-linked form of hearing impairment (Kok et al.,
1995).
Homeobox genes are also related to lipid metabolism and metabolic diseases. The Smith-
Lemli-Opitz syndrome, an inborn error of cholesterol metabolism, is related to the sonic hedgehog embryonic signaling pathway and the expression of homeobox genes (Kelley
54 and Hannekam, 2000). PDX1 homeobox genes can regulate pancreatic development and islet specific gene expression (insulin 1 and 2 and prohormone convertase 1/3) (Ferber et al., 2000).
2.9 Homeobox genes and cancers
2.9.1 Homeobox genes and solid tumors
Following the initial reports that HOX genes are expressed in carcinoma cell lines, the deregulated expression of homeobox genes has been described in many tumors and derivative cell lines (Table 4) (Abate-Shen, 2002).
Table 4 Deregulated homeobox genes in solid tumors (modified from Abate-Shen, 2002).
Homeobox genes Deregulation in cancer Normal expression Functional insights HOX Gain of expression in Expression patterns Overexpression promotes primary tumors and cell during embryogenesis cellular transformation lines from brain, breast, reflect roles in patterning, in culture colon, lung and kidney segmentation and fate determination of many tissues
MSX Gain of expression in Expression during Overexpression leads to in mammary, colon, embryogenesis inhibition of stomach, kidney, thyroid associated with differentiation, correlated and other carcinomas epithelial-mesenchymal with upregulation of interactions and cyclin D1 inversely correlated with differentiation
HSIX1 Gain of expression in Limited expression Overexpression abrogates mammary and other analyses available; the G2 cell-cycle carcinomas homologues expressed checkpoint in in the developing response to X-ray brain, eye, muscle, and irradiation
55
tissues
GBX2 Gain of expression in Expressed in the Downregulation of GBX2 Prostate carcinoma developing nervous via antisense correlated system; limited with reduced information concerning tumorigenicity expression in prostate
PAX Gain of expression in Expression patterns Overexpression promotes Wilms’ tumor, brain during embryogenesis cellular transformation and breast cancer; reflect roles in in culture translocation of PAX8 organogenesis of kidney in thyroid carcinoma and other tissues
CDX2 Loss of protein Expressed during Overexpression promotes expression in colon embryogenesis in differentiation of carcinoma extraembryonic tissues; intestinal cells, while expression in older leading to reduced embryos and adults proliferation and restricted to intestinal tumorigenicity; epithelium heterozygous mutant mice predisposed to colon cancer
NKX3.1 Loss of protein Expressed during Localized to 8p21, expression in embryogenesis in which is frequently prostate cancer somites and other deleted in prostate and pre-neoplastic derivatives; cancer; overexpression lesions expression in older leads to reduced cell embryos and adults growth and is restricted to and tumorigenicity; prostatic epithelium homozygous and heterozygous mutant mice are predisposed to prostate cancer
BARX2 Loss of expression Expressed in normal Localized to 11q24, in ovary carcinoma ovarian surface which is frequently epithelium; limited deleted in ovarian expression analyses cancer; displays available tumor suppression and anti-metastastic activities in cell culture
56
Gain of function of certain classes of homeobox genes has been demonstrated to promote
the oncogenic phenotype.
1) It was shown that NIH 3T3 fibroblast cells bearing the activated Hoxb8 gene produced
fibrosarcomas in nude mice (Aberdam and Sachs, 1991).
2) Msx1 is involved in the terminal differentiation of myogenic cells (Song et al., 1992).
Transfection of Msx1 into myoblasts inhibited terminal differentiation and induced cell
transformation (Song et al., 1992).
3) The oncogenic potential of several deregulated murine homeobox genes was
investigated using in vitro and in vivo transformation assays (Maulbecker and Gruss,
1993). It was shown that the overexpression of the murine homeobox genes, Hoxa7,
Hoxa5, Hoxa1, Hoxb7, and Hoxc8 as well as the homeobox genes Evx1 and Cdx1 can
cause transformation and tumorigenesis in mice (Maulbecker and Gruss, 1993).
4) Retroviral-mediated transduction of the sole Hoxb7 cDNA into SkBr3 breast
adenocarcinoma cells resulted in transcription of both Hoxb7 and bFGF and in several
phenotypic changes including increased cell growth rate and independence from serum
withdrawal (Care et al., 1998). NIH3T3 transfectants bearing an activated Hoxb8 gene
exhibited a transformed phenotype and produced fibrosarcomas in nude mice (Aberdam
et al., 1991).
5) The divergent homeobox gene HOX11 was demonstrated to cause tumorigenicity in
transgenic mouse models when overexpressed in the thymus (Kawabe et al., 1997). The transforming properties of this homeodomain-containing protein were further
57
documented by experiments showing that overexpression of HOX11 in NIH3T3 cells led to transformation of the cells (Dube et al., 1992).
6) PAX3 and PAX7 have been involved in some forms of rhabdomyosarcoma
characterized by t(2;13) (q35;q14) and t(1;13) (p26;q14) chromosomal translocations,
respectively (Sorensen and Triche, 1996). It was shown that murine Pax genes can
promote oncogenesis in tissue culture cells and in mice. Pax1, Pax2, Pax3, Pax6, and
Pax8 homeoproteins were able to induce transformation of cell cultures and tumor formation in mice (Maulbecker and Gruss, 1993). The oncogenic potential of the Pax homeoproteins is dependent on the DNA binding function of the paired motif, as the Un-
Pax1 protein, which carries a point mutation in this domain that impairs DNA binding, is also defective in tumor formation (Maulbecker and Gruss, 1993).
It is also likely that HOX genes can regulate vasculogenesis and angiogenesis. Indeed,
HOXD3 has been linked to the mechanism converting endothelial cells from a resting to angiogenic/invasive state (Boudreau et al., 1997), whereas HOXB13 is required for the subsequent capillary morphogenesis of the new vascular sprouts (Myers et al., 2000).
Importantly, HOXB7 was demonstrated as a key factor upregulating a variety of proangiogenic molecules, including vascular endothelial growth factor, melanoma growth-stimulatory activity/growth-related oncogenence α, interleukin 8, and angiopoietin 2 (Care et al., 2001). Furthermore, HOXB7 was found to induce growth and angiogenesis in vitro and in vivo (Care et al., 2001).
Normal adult human organs such as the kidney, colon, and lung display expression patterns of the whole network of HOX genes typical of each organ (Cillo et al., 1992;
58
DeVita et al., 1993). The patterns differ greatly from organ to organ with respect to active or silent genes. HOX genes active in different organs often display differences in transcript size. This indicates that the patterns of HOX gene expression result from the whole expressions of the different cell phenotypes characteristic of a specific organ and that these organ-specific HOX patterns may be responsible, at the same time, for form, structure and correct positioning of the organs along the antero-posterior body axis (Cillo et al., 2001). But, major differences in HOX gene expression are detectable in primary solid tumors (kidney, colon, small cell lung cancer, etc.) compared with the corresponding normal adult organs (Cillo, 1994-1995). These alterations involve HOX genes crucial for the organ development as well as HOX genes apparently unrelated to the organ.
In renal carcinomas and normal human kidneys, several differences in HOX gene expression were detected (Cillo et al., 1992). In primary kidney tumors HOX genes can be turned off (HOXB5, HOXB9) or on (HOXC11) or can display different-sized transcripts (HOXD4) if compared to the expression of normal kidneys, indicating an association between altered HOX gene expression and kidney cancer. In addition, HOX and other homeobox genes may play a role in aberrant kidney differentiation that leads to
Wilm’s tumor (Cillo et al., 1995). The possibility for some HOX genes to act as reliable indicators of the histological subtype of kidney cancers and Wilm’s tumors has been confirmed using HOXD10, HOXA9, and HOXC9 genes (Redline et al., 1994).
59
HOX gene expression was analyzed in normal human colon and in primary and metastatic colorectal carcinomas (De Vita et al., 1993). Twenty-nine HOX genes appear
to be actively expressed in normal colon. The entire HOXA locus, containing 11 genes, is
expressed and virtually all the genes of HOXB cluster are expressed with the exception of
HOXB1. In contrast, only some of the genes of HOXC and HOXD locus are expressed.
The expression of some HOX genes is identical in normal and neoplastic colon including
genes at 5’ end of the HOXA locus, HOXA9 through HOXA13 and five genes at the 5’
end of the HOXC locus, HOXC9 through HOXC13. Virtually all the HOXB genes are
active and their expression does not vary when normal tissues are compared to colorectal
cancer specimens. In contrast, other HOX genes exhibit altered expression in primary
colon cancers and their hepatic metastases. These HOX genes include the 3’ HOXA
genes, HOXA1 through HOXA4 and the 3’ end of HOXC genes, HOXC4 through
HOXC8. The HOXD locus is characterized by a marked variable expression of HOX
genes (De Vita et al., 1993).
The expression of the whole HOX gene network in normal adult human lung was tested.
The expression of 11 genes of the network was detected. Their expression is restricted to
5 genes of the 3 end of the HOXA locus (from HOXA2 to HOXA6), the paralogous genes of the HOXB locus (from HOXB2 to HOXB6) (corresponding to the boundary of the expression of the embryonal lung bud), and the gene HOXD1 (Tiberio et al., 1994).
In contrast, a strikingly different overall pattern of HOX gene expression was observed in
human small-cell lung cancer (SCLC). Between 23 and 30 HOX genes were expressed
and the alteration in HOX gene expression in SCLCs mainly concerns the HOXB and C
60 loci (Tiberioi et al., 1994). Furthermore, in SCLC, the number of actively expressed HOX genes may be lower in metastatic cancer than in primary tumors (Tiberio et al., 1994).
Hox genes were shown to play a role in murine skin morphogenesis (Matthews et al.,
1993; Scott and Goldsmith, 1993). Analysis of cultured fetal melanocytes shows the expression of several genes of the A and C Hox clusters (Bieberich et al., 1992; Scott and
Goldsmith, 1993). On the basis of HOX gene expression, it is possible to group melanoma clonal populations derived from the same metastatic cell line (Me 665/2) into two types (Cillo et al., 1996). In type one, the HOX gene network is expressed according to a very similar pattern of the parental Me 665/2 cell line and in type two, HOX genes display a silent block of genes (HOXC10, HOXC11, and HOXC13) localized at the 5’ end of the HOXC locus (Cillo et al., 1996). The expression and function of HOXB cluster genes in human normal and neoplastic melanocytic cells were investigated (Care et al.,
1996). In this study, HOXB7 was found to be consistently expressed in all the melanoma cell lines and patients’ melanoma samples examined. Conversely, HOXB7 was expressed in proliferating but not quiescent normal melanocytes (Care et al., 1996). Treatment of melanoma cell lines with antisense oligomers targeting HOXB7 mRNA markedly inhibited cell proliferation (Care et al., 1996).
All these findings above have indicated an association between altered HOX gene expression and primary human tumors (Cillo et al., 1992). Besides the involvement of altered homeobox gene expression in primary tumors, major variations in homeobox gene expression are detectable along tumor progression (Cillo et al., 2001). Misexpression of
61
HOX genes was detected in metastatic lesions with respect to the primary tumor of origin
and the corresponding normal tissue (De Vita et al., 1993). Furthermore, in some cases,
HOX gene expression appears to be specifically linked to stage of tumor progression and
histological tumor type (Tiberio et al., 1994).
In conclusion, homeobox genes that are normally expressed in developing tissues and are
downregulated in differentiation are often re-expressed in cancer. Conversely, homeobox
genes that are expressed in differentiated tissues are often downregulated in cancer
progression (Abate-Shen, 2002).
2.9.2 Homeobox genes and leukemia
Role of homeobox genes in malignant processes was first documented in leukemia. Many homeobox genes play an important role in haemopoiesis, including the specification of definitive haemopoietic stem cells and differentiation along the myeloid lineage (Kyba et al., 2002; van Oostveen et al., 1999). The first evidence of the involvement of a homeobox gene in leukemia was the Hoxb8 gene which is transcriptionally activated in the mouse myeloid leukemia cell line WEH1-3B. The constitutive over-expression of this homeobox gene is due to the insertion of its first exon of a retrovirus-like intracisternal A particle (IAP) (Kongsuwan et al., 1989; Perkins et al., 1990). It was shown that the overexpression of HOXB8 inhibited specific pathways of the myeloid differentiation program (Perkins et al., 1990). Several indications have been provided by leukemic mice on the interaction between Hoxa7, Hoxa9, and Hoxa10 (Nakamura et al., 1996). The
62 targeted destruction of Hoxa9 gene is leukemogenic in mice (Lawrence et al., 1997). The overexpression of Hoxa10 induces myeloid leukemia after long periods of latency
(Thorsteinsdottir et al., 1997). The fusion in frame between the nucleoporin gene NUP98 and the HOXA9 gene is achieved through the translocation t(7;11) (p15;p15) in human
AML (Nakamura et al., 1996). The pattern of expression of homeobox genes contained in the four human HOX clusters has been studied in different types of human leukemia
(Celetti et al., 1993). The results suggest that hematopoietic malignant cells express a repertoire of HOX genes characteristic of a particular cell lineage at a specific stage of differentiation. PBX1 is a divergent homeobox gene that was identified as the chromosome 1 partner of the t(1;19) translocation in human pre-B-cell ALL. The t(1;19) results in the fusion of a portion of PBX1 including the homeodomain, with a truncated
EA2 protein. Fusion between PBX1 and E2A proteins alters the sequence-specific binding to the HOX genes directing the homeoprotein on a different target and inducing leukemogenesis (Knoepfler and Kamps., 1997). In addition, cooperation between
HOXB3, HOXB4, and PBX1 homeoproteins plays an important role in blood cell proliferation and transformation (Krosl et al., 1998). HB24 is another divergent human homeobox gene expressed in hematopoietic progenitor cells (Deguchi et al., 1992). High expression of HB24 was shown to confer to a human T cell line the ability to form metastasis in mouse (Deguchi et al., 1993). HOX11 is a divergent homeobox gene originally isolated from a human leukemia. Overexpression of this gene occurs in rare cases of human T cell ALL that have the translocations t(10;14)(q24,q11) or t(7;10)
(q35;q24), in which the HOX11 gene is activated due to juxtaposition with promoter elements from the T cell receptor α and β genes, respectively (Nakamura et al., 1996).
63
Homeobox genes are also involved in lymphomagenesis. The locus HOXC is mainly implicated in lymphomas: while HOXC4 and C6 are expressed in non-Hodgkin's lymphoma, HOXC5 displays a type- and site-restricted expression pattern in both T- and
B- cell non-Hodgkin's lymphomas (Bijl et al., 1997).
In conclusion, the consequences of deregulated homeobox genes for carcinogenesis can be interpreted as an extension of their normal function. In normal tissues, their cumulative activities provide a balance between proliferation and differentiation, whereas perturbations of homeobox gene expression promote the transformed phenotype (Abate-
Shen, 2002).
2.10 Homeobox genes and mammary gland
Early immunohistochemical studies and screens based on polymerase chain reaction first detected homeobox gene expression in mammary epithelial cell lines and tumors (Wewer et al., 1990; Castronovo et al., 1994). This has led to a series of investigations that have indicated that homeobox genes may be important for normal mammary gland and mammary carcinoma development.
2.10.1 Homeobox genes and mice mammary gland and mice breast cancer
In 1994, Friedmann et al. examined Hox gene expression in normal and neoplastic mouse mammary gland. The Hox gene expression detected in mouse mammary gland
64
development included Hoxa7, Hoxb6, Hoxc6, and Hoxd12. Further study found that
Hoxc6 transcripts were detected in mammary glands throughout all stages of
development from pubescent to mature animals. The transcript levels then declined in
glands from animals that were in early stages of pregnancy and could not be detected in
the late stages of pregnancy or during lactation. In contrast, Hoxa1 was not detected in
any of these stages of development (Friedmann et al., 1994). Hoxc6 showed very low
levels of expression in the precancerous mouse mammary tissue but showed no
expression in cancers. In contrast, Hoxa1 showed high levels of expression in mouse
breast cancer but showed no expression in precacerous tissue (Friedmann et al., 1994).
Friedmann and coworkers continued their studies and further identified several other Hox
genes in mouse mammary gland (Friedmann et al., 1995). In situ hybridization
demonstrated that many of these genes are expressed either in the mammary epithelium
or in the periductal stroma, or both (Figure 5). In addition, expression of the normally
expressed homeobox genes examined (Hoxb6, Hoxb7, Hoxc8, Hoxd4, Hoxd8, Hoxd9,
and Hoxd10) was lost on neoplastic progression in a selected population of mouse
hyperplasias and tumors (Friedman, 1995).
Soon after, five Hox genes were identified to be expressed in CID-9 mouse mammary epithelial cells (Srebrow et al., 1998). It was found that Hoxa1 and Hoxb7 were downregulated by extracellular membrane, suggesting that homeobox gene expression may be modulated in vivo by interactions between epithelial cells and components of the basement membrane. It was shown that such interactions are important for mammary
65
differentiation and function and that they tend to be altered on neoplastic progression
(Robinson et al., 1999; Streuli and Gilmore, 1999).
In later studies, Chen and Cappechi examined and analyzed mice with mutant Hoxa9,
Hoxb9, and Hoxd9 genes (Chen and Cappechi, 1999). The genotype of mutant mice is
designated by capital and lower-case letters, where A, B, and D denote the wild-type
alleles and a, b, and d denote the loss-of-function alleles of Hoxa9, Hoxb9, and Hoxd9,
respectively. Single mutant (monoallellic) of Hoxa9 or Hoxb9 showed only a small
decease in newborn survival, while the mutant of Hoxd9 reduced survival to below 50%.
Double mutant (biallellic) showed synergistic effects that reduced newborn survival
below the survival rate compared with single mutation. The mammary glands of Aabbdd, aaBbdd, and aabbdd females showed hypoplasia of the mammary glands during
pregnancy and post parturition. Measurement of serum-estrogen levels in virgin and
pregnant aabbdd females showed that concentrations of this hormone do not differ
significantly from those in age-matched, wild-type female controls, indicating that the
mammary hypoplasia observed in aabbdd mice is not likely to be caused by a deficit in
the production of reproductive hormones. Expression of Hoxa9, Hoxb9, and Hoxd9 was
found in mammary-gland samples from 8-week-old virgin females, pregnant and
lactating females, as well as 12.5 days gestational embryos. The expression of Hoxa9,
Hoxb9, and Hoxd9 in embryonic, pregnant and lactating mammary-gland tissues suggests
that they have direct roles in the development of the mammary gland (Chen and
Cappechi, 1999).
66
E13 Postnatal virgin Early pregnancy Late pregnancy Lactation
Figure 5. Tissue distribution of homeobox gene expression through mouse mammary gland development. Selected stages of mammary gland development are depicted with reference to the expression patterns of several homeobox genes as demonstrated by in situ hybridization. Expression at a given stage is shown by a bar above the stage. Bars are pattern coded to represent a unique tissue compartment or epithelial structure. Transition points affected by a given homeobox gene mutation are denoted by a hatched box above the arrow representing the transition (modified from Friedmann et al., 1995).
Msx is a small family of three related homeobox genes. Only Msx1 and Msx2 have been
examined in the mouse mammary gland (Friedmann and Daniel, 1996; Phippard et al.,
1996). Transgenic mice that express Msx1 under the control of the mouse mammary
tumor virus long terminal repeat (MMTV LTR) display impaired differentiation of the
mammary epithelium during pregnancy (Hu et al., 2001). The expression of Msx1 and
67
Msx2 was present in mammary glands of virgin mice and in mammary glands of mice in
early pregnancy, but the expression decreased dramatically during late pregnancy. In
mammary glands from lactating animals and during the first days of involution, low
levels of Msx1 transcripts were detected. But during lactation or early involution, Msx2
expression was not detected in mammary glands. Expression of both genes increased
gradually as involution progressed (Friedmann and Daniel, 1996). During embryogenesis,
both Msx1 and Msx2 were expressed in the epithelium. In contrast, in postnatal mice,
Msx1 and Msx2 were expressed in reciprocal tissue compartments, with Msx1 expressed
in the epithelium and Msx2 expressed in the periductal stroma (Friedmann and Daniel,
1996). Expression studies also suggested a potential role of Msx2 in mediation of
hormone responses. Msx2 expression was decreased following ovariectomy or following
exposure to anti-estrogen implanted directly into the gland, and returned to normal levels after estrogen was administered to ovariectomized animals (Friedmann and Daniel,
1996).
Both Engrailed-related (En) homeobox genes, En1 and En2, have also been examined in the mouse mammary gland (Friedmann, 1995). En1 was expressed at constant levels through early pregnancy but was not detectable during late pregnancy or lactation. En2
was not detected in any tissue examined (Friedmann, 1995).
2.10.2 Homeobox genes and human mammary gland and human breast cancer
68
Four different human homeobox genes, HOXA1, HOXA10 HOXB6, and HOXC6 were
identified in the human breast cancer cell line, MCF-7 (Castronovo et al., 1994). In
another study, it was shown that HOXA1 can be induced by retinoic acid in MCF-7
breast cancer cells (Chariot et al., 1995). HOXA1 was further demonstrated to be induced by progestin in MCF-7 cells (Chariot and Castronovo, 1996). HOXA1 expression was also detected in a variety of human beast cancer lesions (Chariot and Castronovo, 1996).
Later, more and more homeobox genes were examined in human mammary gland.
In 2003, Cantile et al. analyzed whole HOX gene network expression in human normal
breast tissue and in primary breast cancers by RT-PCR. Seventeen of 39 HOX genes
were expressed in normal adult human breast (Figure 6) (Cantile et al., 2003). There were
no quantitative differences in HOX gene expression detected between normal epithelial
breast tissues. More loci A and C were expressed compared with loci B and D. In locus
A, HOXA3, HOXA7, HOXA9, HOXA10, and HOXA11 were expressed. In locus B,
HOXB3, HOXB5, HOXB6, and HOXB7 were expressed. In locus C, HOXC6, HOXC8,
HOXC10, HOXC11, and HOXC13 were expressed. In locus D, HOXD1, HOXD8 and
HOXD9 were expressed. More thoracic (7/11 active genes) and lumbo-sacral (7/16)
HOX genes were expressed in the normal breast whereas only 3 out of 12 cervical HOX
genes were expressed. Furthermore, no adjacent HOX genes were expressed in the
cervical region. In contrast, adjacent 3' or 5' end HOX genes were expressed both in the
thoracic and lumbo-sacral regions of the network (Cantile et al., 2003).
69
Cantile et al. also compared the whole HOX gene expression pattern in normal breast
(large squares) and in fourteen breast cancer biopsies (Figure 6) (Cantile et al., 2003).
The expression of some HOX genes is similar in normal and neoplastic breast tissue indicating that these genes may be involved in breast organogenesis. In contrast, other
HOX genes had altered expression in breast cancers compared with normal breast indicating their involvement in breast cancer tumorigenesis.
The cervical part of the HOX gene network displayed the largest difference between normal breast and breast cancer. In normal breast HOX gene network, only HOXA3,
HOXB3 and HOXD1 in cervical region were expressed while in breast cancer, 11 of 12 cervical HOX genes became alternatively expressed. HOXC4 was constitutively silent in normal breast and breast cancers. In contrast, HOXD1 was active in all normal tissue and breast cancers. HOXD3 was silent in normal breast and expressed in all the breast cancer tissues examined (Cantile et al., 2003). HOXA1, which is silent in normal breast, became expressed in some breast cancer tissues. This is consistent with the reports of Chariot and
Castronovo, who showed that HOXA1 expression was detected in a variety of human beast cancer lesions and in human breast cancer cell line MCF-7 (Chariot and
Castronovo, 1996). Interestingly, the human HOXA1 expression pattern in mammary gland is also seen in the mice (described above) (Friedmann et al., 1994).
70
5’ Lumbo-sacral Thoracic Cervical 3’
Figure 6. Diagram of HOX gene expression in normal and neoplastic breast. Closed or open symbols indicate active or silent HOX genes, respectively. Large squares represent normal breast; small ovals represent each individual cancer biopsy tested. Small ovals in the same position in the figure refer to the same breast cancer biopsy (modified from Cantile et al., 2003).
Thoracic genes of the HOX network displayed only minor difference between normal breast and breast cancer. HOXA7 was invariantly expressed in all normal and neoplastic tissues. HOXA5 and HOXA6 were silent in the normal breast tissue and became active in the breast cancer biopsies. Three out of 4 thoracic HOXB genes were expressed in normal breast. HOXB5 became silent in most of the breast cancer tissues while the other three HOXB genes did not display any quantitative differences in their expression between normal and neoplastic tissues. The thoracic HOXC genes showed slight differences in their expression between normal and neoplastic breast tissues. Finally,
71
HOXD8 appeared to be constitutively active in both normal and neoplastic breast tissues
(Cantile et al., 2003).
There are 7 out of 16 lumbo-sacral HOX genes expressed in normal breast tissue
including three adjacent genes HOXA9, HOXA10 and HOXA11. HOXA9 and HOXA10
were still active in the cancer biopsies while HOXA11 was silent in all, but one, of the
breast cancer tissue tested by the authors. HOXA13 and HOXB9 were switched off in the
normal breast tissue but active in some breast cancer tissues. The locus C HOX genes did
not display major differences in their expression between normal and neoplastic breast
tissue. HOXD10 and HOXD13 manifested differences in their expression between
normal and neoplastic breast tissue (Cantile et al., 2003).
The HOX gene network manifests expression patterns corresponding to specific cell phenotypes (Cillo et al., 2001). Thus, it is possible that thoracic paralogous HOX gene groups control the thoracic embryonic development and its homeostatic regulation during adult life. In contrast, cervical and lumbo-sacral paralogous HOX groups may be involved in signaling pathways related to the determination of the breast cell phenotype and therefore their expression is altered during breast cancer progression (Cantile et al.,
2003).
In the human mammary gland, a family of Iroquois-class (IRX) homeobox genes was identified (Lewis et al., 1999). One gene, IRX2, is expressed in discrete epithelial cell lineages being found in ductal and lobular epithelium, but not in myoepithelium. During
72
gland development, its expression is increased in terminal end buds and terminal lobules and is reduced in a subset of epithelial cells during lactation. IRX2 expression is also
maintained in human mammary neoplasias (Lewis et al., 1999).
At least 4 POU homeobox genes: OCT1, OCT2, OCT3, and OCT11 were identified in human breast cancer cell line, MCF-7 (Jin et al., 1999). The expression of OCT1 and
OCT2 in other human breast epithelial cell lines was further determined. It was found that OCT3 is expressed in both human breast cancer cell lines and all human primary breast carcinomas examined, but not in normal human breast tissue, indicating its important role in mammary gland carcinogenesis (Jin et al., 1999).
Overexpression of homeobox gene HSIX1 in MCF-7 cells abrogated the G2 cell cycle checkpoint in response to x-ray irradiation. HSIX1 expression was absent or very low in normal mammary tissue, but elevated expression was observed in 44% of primary breast cancers and 90% of metastatic lesions. In addition, HSIX1 was expressed in a variety of cancer cell lines (Ford et al., 1998).
2.10.3 Summary of homeobox gene expression and function in mammary gland development and neoplasia
Thus far, it seems that loss-of-function mutations simply affect developmental progression but that gain-of-function mutations impact neoplastic progression (Lewis
2000). For example, loss-of-function mutations of Hoxa9/Hoxb9/Hoxd9 resulted in
73 developmental failure, rather than tumor formation (Chen and Capecchi, 1999). In terms of gain-of-function mutations, overexpression of murine Hoxb7 in SkBr3 mammary cancer cells caused increased proliferation and decreased growth factor dependency (Care et al., 1998). Table 5 summarizes homeobox gene expression and function in mammary gland development and neoplasia (Lewis, 2000).
Table 5 Summary of homeobox gene expression and function in mammary gland development and neoplasia (modified from Lewis, 2000).
Gene Expression Expression altered Possible Mammary or name detected in in neoplasias regulation? in vitro phenotype?
Hoxa1 MG(UD) OE (freq.) Extracellular matrix Scp2, CID-9 components HOXA1 MCF-7 Retinoic acid Hoxa5 CID-9 Hoxa7 MG HOXA7 MG Hoxa9 MG KO: lactational defects HOXA10 MCF-7 Vitamin D OE in MCF-7: promotes cell cycle arrest in G1 Hoxb6 MG Lost HOXB6 MCF-7 Hoxb7 MG UE Extracellular matrix OE in SkBr3 cells: bFGF components; induction; increased developmental proliferation; decreased growth factor dependency Hoxb8 CID-9 Hoxb9 CID-9 E12.5 KO: lactational defects Embryo (MM) Hoxc6 MG Lost Estrogen (-); developmental HOXC6 MCF-7; Variable MCF-7D Hs578Bst MCF-10F T47D
74
Hoxc8 MG Lost Estrogen (-); Developmental HOXC10 MG Hoxd3 MG (UD) OE (rarely) Hoxd4 MG (E) Lost Hoxd8 MG Lost Developmental Hoxd9 MG (E+S); E12.5 Embryo Lost Developmental KO: lactational defects (MM) Hoxd10 MG (E+S) Lost Developmental KO: lactational defects Hoxd11 MG (UD) No Hoxd12 MG OE (myc tumor) Msx1 E13.5 Embryo (E); Maintained Developmental KO: none detected MG (E) but variable in embryos Msx2 E13.5 Embryo (E); Lost Estrogen; MG (S) developmental KO: MG fail to form IRX1 MG IRX2 MG; primary Maintained. Developmental tumors expression levels or splice form usage may altered in some tumors IRX3 MG IRX4 MG IRX5 MG Alx4 MG (S) Expression KO: none detected associated with actively condensing periductal stroma
HSIX1 MCF-7 OE Expressed in S OE in MCF-7 phase of cell abolishes X-ray cycle induced cell cycle arrest in G2 phase OCT1 MCF-7; MCF-10; SkBr3; MDA-MB453 Oct1 MG (lact.) OCT2 MCF-7; MCF-10; SkBr3 OCT3 MG (UD); OE (freq.) Form with a 5 MCF-7; amino acid deletion
75
MCF-10; detected in MCF-7 SkBr3; cells primary tumors Oct3 MG (lact.) OCT11 MG (UD); OE (freq.) MCF-7; SkBr3; MDA-MB453; primary tumors Pit-rs1 MG (lact.) Pou4f1 MG (lact.) Pou4f-rs1 MG (lact.) Pou3f4 MG (lact.) Pou3f-rs1 MG (lact.) En1 MG UE (freq.) Developmental En2 MG (UD) UD KO: none detected MOX1 MG MEIS2a MG MEIS2d MG MRG1a MG
E, epithelium; freq., frequently; KO, knockout mutation; lact., lactating; M, mammary mesenchyme; MG, mammary gland; OE, overexpressed; S, periductal stroma; UD, undetectable; UE, underexressed.
76
Recently a study in our laboratory demonstrated that autocrine GH effects in MCF-7 cells included changes in cell-cell adhesion and E-cadherin localization (Mukhina et al., manuscript submitted). Recent developments show that E-cadherin functions not only as an adhesion structure to bring cells together, but also may have additional functions as active receptors to direct cell signaling and activate downstream molecules (Kovacs et al.,
2002). Initial results showed that HOXA1 was upregulated in MCF-7 at full confluence and this upregulation was confirmed to be directed by E-cadherin-activated signaling through Rac1 (results detailed in Chapter 4.2).
2.11 E-cadherin
2.11.1 E-cadherin structure
Adjoining epithelial cells are connected by a number of structures between them: (1) tight junctions at the topical part of the cell-cell contact, (2) adherens junctions, (3) desmosomes, (4) gap junctions, (5) other structures including hemidesmosomes and focal adhesion complex (Jiang and Robert, 2000). The appropriate functions of these structures maintain the normal physiology, integrity, and function of the epithelial cells.
Cadherins are a major class of cell adhesion molecules responsible for strong cell-cell adhesion. The main structure of cadherin complex is composed of the following components: cadherin, intracellular components associated with cadherin (such as catenins), and the cytoskeleton (such as actin) (Jiang and Robert, 2000). Cadherins are a
77
family of transmembrane glycoproteins that mediate cell-cell adhesion. They have a
single-span transmembranous domain containing a carboxy terminal in the intracellular
domain, with five tandem repeats in the N-terminal extracellular domain, and anchor to
the cytoskeleton via other proteins, including catenins (Shapiro et al., 1995; Nagar et al.,
1996). The extracellular repeats allow cadherin to form homotypical interactions in the
presence of extracellular calcium, thus forming a zip-like structure between two cells.
However, heterotypic interactions between different cadherin molecules are also possible.
A-CAM-(N-cadherin) expressing cells and L-CAM-(E-cadherin) expressing cells were
found to form heterotypic junctions (Jiang, 1996; Volk et al., 1987).
The classification of cadherins is traditionally based on the tissue distribution or the
origin from which they are discovered (Jiang and Robert, 2000). For example, the
cadherin distributed in epithelial cells is named as E-cadherin, in heart as H-cadherin, in neural tissues as N-cadherin.
E-cadherin is a protein 120 kDa in size and is widely expressed in epithelial cells. Like other cadherins, E-cadherin has three domains, the extracellular, transmembrane, and intracellular domains. Ca2+ binds to the extracellular domain of E-cadherin, thereby
inducing a conformational change that promotes homophilic interaction with E-cadherin
in an adjacent cell (Koch et al., 1997). The cytoplasmic domains of E-cadherin provide
attachment to the actin cytoskeleton via catenins and other cytoskeletal proteins
(Gumbiner, 1996; Takeichi, 1995). β- or γ-catenin binds to E-cadherin and α-catenin and
α-catenin links the complex to the cytoskeleton by associating with actin (Provost and
78
Rimm, 1999). Another protein, p120ctn, binds E-cadherin independently of β- and γ-
catenin and is thought to regulate adhesion separately from the other catenins (Provost
and Rimm, 1999). Furthermore, it has been demonstrated that a homotypical binding
domain is conferred by the HAV sequence, a domain located in the first cadherin repeat
of E-cadherin (Overdiun et al., 1995; Pokutta et al., 1994; Noe et al., 1999). Disruption or
mutation of the HAV domain results in a damage to cell-cell adhesion (Noe et al., 1999).
Once synthesized, E-cadherin has a short half-life (5-10h). The subsequent assembly and
turnover of the adhesion structure are regulated by many external and internal signals
(Jiang and Robert, 2000).
2.11.2 Cellular effects of E-cadherin
The role of E-cadherin in cell-cell adherens junctions is described as above. Continued expression and functional activity of E-cadherin are required for cells to remain tightly associated in the epithelium, and in its absence the many other cell adhesion and cell junction proteins expressed in epithelial cells are not capable of supporting intercellular adhesion (Gumbiner, 1996). E-cadherin-mediated cell-cell adhesion is regulated by a number of pathways. Activation of Wnt pathways was shown to result in an enhanced cell-cell adhesion that is mediated by E-cadherin (Hinck et al., 1994; Bradley and Cowin,
1992). It was also shown that phosphorylation of E-cadherin, β-, and α-catenins alters E- cadherin-mediated adhesion (Hiscox and Jiang, 1999). Several lines of evidence suggest that members of the Rho family are required for the establishment and maintenance of E- cadherin-based adherens junctions (Braga et al., 1997; Hordijk et al., 1997; Takaishi et
79
al., 1997; Zhong et al., 1997; Kuroda et al., 1997). Overexpression of constitutively
active Rac1 or Cdc42 in epithelial cells increases E-cadherin localization and actin
accumulation at cell-cell junctions, whereas these events are inhibited by dominant negative mutants of Rac1 and Cdc42 (Hordijk et al., 1997; Takaishi et al., 1997; Kuroda et al., 1997; Ridley et al., 1995; Jou and Nelson, 1998).
In addition to its effects on adhesion, E-cadherin also has effects on cell growth. In normal epithelial cells, when the cells grow to a certain degree, the rate of growth and proliferation is reduced. This phenomenon is called contact inhibition. This is one of the key mechanisms by which normal epithelial cells control their own fate and is one of the fundamental differences between normal epithelial cells and epithelium-derived tumor cells. The contact inhibition is lost and the cancer cells undergo uncontrolled growth. It is now known that the loss of the contact inhibition in epithelial cell lineage is linked to E- cadherin and a cell cycle regulator, p27 (Jiang and Robert, 2000). p27 is able to interact with cyclins and cyclin-dependent kinases (CDK), resulting in the formation of p27-CDK complex and deactivation of CDKs and the blockade of cycle progression (Hengst et al.,
1994; Hengst et al., 1996). It has been demonstrated that when the E-cadherin positive cells grow to confluence, the close cell-cell contact is mainly mediated by E-cadherin and the p27 level is high in the nucleus (Croix et al., 1998). When the E-cadherin function is disturbed, p27 is lost and cell cycle progresses (Croix et al., 1998).
E-cadherin is involved in the expression of Eph receptors and ephrins (Orsulic and
Kemler, 2000; Zantek et al., 1999). It has been shown that membrane localization and
80 phosphorylation of EphA2 in mammary epithelial cells are dependent on E-cadherin– mediated adhesion (Zantek et al., 1999). In epithelial cells, E-cadherin is required for
EphA2 receptor localization at cell-cell contacts; in the absence of functional E-cadherin,
EphA2 localizes to the perinuclear region (Orsulic and Kemler, 2000). Constitutive ectopic expression of E-cadherin in non-epithelial NIH3T3 cells results in the production of the EphA2 receptor (Orsulic and Kemler, 2000). In another study, it was found that E- cadherin regulates expression of several Eph recettors and ephrins in embryonic stem
(ES) cells (Orsulic and Kemler, 2000). Rescue of E-cadherin null ES cells with E- cadherin cDNA restores the wild-type expression patterns of Eph family members
(Orsulic and Kemler, 2000). Rescue of E-cadherin null ES cells with N-cadherin cDNA does not restore the wild-type expression pattern, indicating that the regulation of differential expression of Eph family members is specific to E-cadherin (Orsulic and
Kemler, 2000).
2.11.3 E-cadherin and cancer
E-cadherin has been demonstrated to be involved in cancer invasion and progression.
There are a number of abnormalities seen in breast cancer cells and breast cancer tissues.
E-cadherin levels can be reduced or even absent totally, and the level of E-cadherin in breast cancer appears to bear no relationship with ER and PR status of the tumor (Perl et al., 1998; Palacios et al., 1995; Guriec et al., 1996; Dillon et al., 1998; DeLeeuw et al.,
1997; Bankfalvi et al., 1999; Jacquemier et al., 1999). Loss of E-cadherin is more common in lobular- than ductal-type carcinomas and presents a common feature for the
81
lobular type (Bankfalvi et al., 1999; Jacquemier et al., 1999; Siitonen et al., 1996;
Zschiesche et al., 1997; Mbalaviele et al., 1996). The reduction of E-cadherin can be in a steady progression pattern from pure intra-ductal carcinomas though invasive ductal to recurrent carcinomas (Siitonen et al., 1996). Lack of E-cadherin-mediated adhesion in human tumors correlates with the loss of epithelial morphology and the acquisition of mesenchymal characteristics. Blocking E-cadherin function in vitro with antibodies or
antisense RNA increases tumor cell motility, invasion and metastatic potential (Behrens
et al., 1989; Vleminckx et al., 1991).
Mutations of E-cadherin have also been reported in breast cancer. E-cadherin mutation
appears to be the loss of heterozygosity (LOH) of chromosomal region 16q22.1
containing the E-cadherin locus (Berx et al., 1996; Berx et al., 1998). It was found that
the LOH was higher in ductal carcinoma than the lobular carcinoma and LOH was
associated with the histological grade of the tumors (Huiping et al., 1999). Germline
mutations of the E-cadherin gene appear to be a prominent feature of inherited cancer
(Guilford et al., 1999).
The mechanisms of loss of expression of E-cadherin in cancer include DNA
hypermethylation of promoter, downregulating of E-cadherin by Snail family, and
shedding of E-cadherin by metalloproteinases (MMPs). It was demonstrated that in breast cancer cells which were negative for E-cadherin expression, there was aberrant methylation of a 5’ promoter region CpG island. This aberrant methylation was also seen in 50% of primary breast cancer tissues. The methylation and expression of E-cadherin
82
can be reversed by demethylation agents (Graff et al., 1995). It was also demonstrated that the density of E-cadherin promoter methylation increases (with E-cadherin level
reduction) when invasion begins to occur (Graff et al., 2000). In addition, E-cadherin transcriptional expression was reported to be downregulated by Snail family (Cano et al.,
2000). The Snail family was known to be involved in embryogenesis and some diseases
(Grau et al., 1984; Boulay et al., 1987). It was found that there was an inverse relationship between the level of E-cadherin and Snail in cancer cells (Cano et al., 2000).
Cancer cells transfected with Snail have reduced expression of E-cadherin and enhanced migration and invasiveness in vitro and in vivo (Jiang and Robert, 2000). Certain proteolytic enzymes, namely metalloproteinases (MMPs) and adamilysins are able to enzymatically degrade the cadherin protein. This degradation destroys E-cadherin
function and enhances invasiveness of cancer cells (Noe et al., 2000; Jiang and Robert,
2000).
2.12 Direct signaling from E-cadherin
Classical cadherin molecules are critical morphogenetic determinants in metazoan organisms (Takeichi, 1995; Yap et al., 1997). At the cellular level, there is increasing evidence that E-cadherin ligation initiates a cascade of molecular and cellular events that ultimately determine cellular recognition. Many signaling molecules are reported to interact with classical cadherins, albeit under conditions that likely depend on cell type and context. The catalogue includes tyrosine kinases and phosphatases, lipid kinases,
83
heterotrimeric GTPases, adaptor proteins, as well as β-catenin itself (Steinberg and
McNutt, 1999; Pece et al., 1999; Meigs et al., 2001; Xu et al., 1997).
2.12.1 E-cadherin signaling and Rho family GTPases
The functional relationship between cadherin adhesion and Rho family GTPases was first
proposed by the observation that several members of this family, including RhoA, Rac1,
and Cdc42, localized to cadherin-based cell-cell contacts (Braga, 2000). Changes in Rho
family GTPase activity can accompany the cadherin-dependent formation of cell-cell
contacts. It was demonstrated that initiation of E-cadherin-mediated cell-cell attachment
significantly increased in a time-dependent manner the amount of active Cdc42 in MCF-7
mammary cancer cells (Kim et al., 2000). By contrast, Cdc42 activity was not increased
under identical conditions in MCF-7 cells incubated with anti-E-cadherin antibodies nor
in MDA-MB-231 (E-cadherin negative) epithelial cells (Kim et al., 2000). In addition, it
was shown that induction of cell-cell junctions increased Rac1 activity and this was
inhibited by E-cadherin function-blocking antibodies (Noren et al., 2001). In another
study, Rac1 was shown to be colocalized with E-cadherin at sites of cell-cell contact and
translocate to the cytosol during disruption of E-cadherin-mediated cell-cell adhesion by
Ca2+ chelation (Nakagawa et al., 2001). The formation of E-cadherin-mediated cell-cell adhesion induced Rac1 activation and this activation was inhibited by treatment of cells with the E-cadherin functional blocking antibody, or with the inhibitor of PI-3 kinase
(Nakagawa et al., 2001). In another study, Kovacs et al. reported that homophilic cadherin ligation recruits Rac to nascent adhesive contacts and specifically stimulates
84
Rac signaling (Kovacs et al., 2001). The E-cadherin ligation also recruits PI3-kinase to
the cadherin complex, leading to the production of phosphatidylinositol 3,4,5-
trisphosphate in nascent cadherin contacts (Kovacs et al., 2001). They also found that Rac
activation involved an early phase, which was PI3-kinase-independent, and a later
amplification phase, which was PI3-kinase dependent (Kovacs et al., 2001).
2.12.2 Discrimination of E-cadherin-dependent juxtacrine signals and direct E- cadherin-activated signals
All the studies and results described above could not, however, distinguish between signals that arise as direct consequences of E-cadherin ligation (direct E-cadherin signaling), and those due to juxtacrine signals that required E-cadherin adhesion to bring cell surfaces together but were not themselves direct downstream consequences of E- cadherin ligation (Yap et al., 1997; Fagotto and Gumbiner, 1996). To overcome this problem, the recombinant E- cadherin-specific adhesive ligands were utilized (Kovacs et al., 2002; Niessen and Gumbiner, 2002). When presented on planar substrata or coated on beads, these ligands support E-cadherin-specific adhesion and recruiting catenins as well as regulating the actin cytoskeleton (Brieher et al., 1996; Lambert et al., 2002; Kovacs et al., 2002). These reagents therefore present a powerful opportunity to discriminate the two cellular effects of E-cadherin. Using this strategy, rapid stimulation of active Rac1 was observed as E-cadherin-containing cells adhered to substrata coated with ligands for human E-cadherin (Kovacs et al., 2002). Rac activation was not seen when cells adhere
85
to poly-l-lysine, indicating that the rapid change in Rac signaling was a specific consequence of E-cadherin ligation, and not due to changes in cell shape concomitant upon spreading on planar sustrata (Kovacs et al., 2002). Notably, Rac signaling increased within minutes of E-cadherin ligation, a time course comparable to those associated with direct pathways activated by growth factors and integrins (Kovacs et al., 2002).
2.12.3 How do E-cadherins activate Rho family GTPases?
The mechanism by which E-cadherin activates Rho family GTPases was investigated by
several studies. The participation of Rho family GTPases in cell signaling depends on
both their intrinsic capacity to interact with downstream effector molecules and their
correct subcellular localization (Symons and Settleman, 2000). Emerging evidence
suggests that classical cadherins can, directly or indirectly, contribute to both these
processes (Yap and Kovacs, 2003) (Figure 7).
Like other GTPases, Rho proteins cycle between two conformational states, active GTP- bound and inactive GDP-bound forms. Guanine nucleotide exchange factors (GEFs) are major molecules responsible for promoting exchange of GDP for GTP (Schmidt and
Hall, 2002). Several GEFs might mediate the early activation of Rho family GTPases by cadherins. VAV2, which activates Rho, Rac, and Cdc42 (Schmidt and Hall, 2002), can interact with p120-ctn, which binds the cytoplasmic tail of classical cadherins (Noren et al., 2000). In addition, Tiam-1 is a Rac-specific GEF capable of affecting E-cadherin
86 expression and the stability of epithelial adherens junctions (Hordijk et al., 1997). Recent evidence suggests that PI3 kinase may act as an upstream activator of Rac in cadherin- activated signaling. PI3 kinase is capable of activating Rac (Hawkins et al., 1995).
Cadherin ligation can recruit PI3 kinase to the cadherin complex and stimulate PI3 kinase activity (Pece et al., 1999; Kovacs et al., 2002). Moreover, inhibition of PI3 kinase activity prevented stimulation of Rac and Cdc42 by E-cadherin (Nakagawa, et al., 2001;
Kovacs et al., 2002; Kim et al., 2000). However, the process by which PI3 kinase activates Rho family proteins remains to be determined. The pleckstrin homology domains, which bind phosphatidylinositol lipids and are found in GEFs, have been postulated as the link (Reif et al., 1996). In addition, the α-isoform of p21-activated kinase-interacting exchange factor, PIX, a GEF for Rac and Cdc42, is activated by PI3- kinase (Yoshii et al., 1999).
Cadherins may influence the precise sites at the plasma membrane where Rho family signaling occurs. Rho family molecules were observed to be localized to adherens junctions (Braga, 2000). This may be a consequence of cadherin signaling, especially local generation of PIP3 by PI3 kinase (Hansen et al., 2002; Kovacs et al., 2002). In addition, it is also possible that proteins of the cadherin-catenin complex can associate directly with Rho family proteins (Yap and Kovacs, 2003).
87
1
3
5
2
4
Figure 7. A model for cadherin-activated Rac signaling participation in early cell– cell recognition. (A) Productive cadherin ligation in newly forming contacts (1) activates Rac signaling at the plasma membrane via a PI3 kinase–dependent intermediary step (2) and possibly also a pathway independent of PI3 kinase (3). One key consequence of Rac activation is the stimulation of cadherin-directed actin assembly by Arp2/3 (4), thereby leading to protrusion of the cell surface (5). (B) Cadherin-directed actin assembly, coordinated by Rac activation, is predicted to direct the surface-protrusive activity of the actin cytoskeleton toward such nascent contacts, to extend the regions of contact and ultimately stabilize cell–cell adhesion (modified from Yap and Kovacs, 2003).
Apart from the signaling pathways responsible for cadherin-actin cooperation at the plasma membrane, E-cadherin direct signaling pathways may have other effects. For example, E-cadherin activated PI3 kinase signaling induced the recruitment to, and phosphorylation of Akt (PKB) (Watton and Downward, 1999; Kovacs et al., 2002), an enzyme well implicated in controlling apoptosis.
88
Altogether, recent developments firmly establish the principle that E-cadherins function
both as mediators of cell surface adhesion, and as adhesion-activated cell surface
receptors.
The increment in HOXA1 expression at full confluence in human mammary carcinoma cells (MCF-7) by E-cadherin-activated signaling through Rac1 is detailed in the result section (Chapter 4.2).
89
Chapter 3 Materials and methods
3.1 Materials
3.1.1 Growth factor, chemicals and reagents
Recombinant human growth hormone (hGH) was a generous gift of Novo-Nordisk
(Singapore). 17-β-estradiol and bovine serum albumin (BSA) were obtained from Sigma
Chemical Co. (St. Louis, MO). RPMI 1640 medium and heat-inactivated fetal bovine
serum were obtained from GIBCO BRL, Life Technologies (Carlsbad, CA). Dulbecco's
modified Eagle's medium/F-12 medium and horse serum were purchased from Invitrogen
(Carlsbad, CA). Glutamine, streptomycin, penicillin, ampicillin B, cholera toxin,
epidermal growth factor were obtained from Upstate Biotechnology (Lake Placid, NY).
Hydrocortisone was purchased from Calbiochem (La Jolla, CA). Ca2+ - and Mg2+ - free phosphate buffered saline solution (PBS) used for cell dissociation were obtained from
Life Technologies (Grand Is., NY). The EffecteneTM Transfection reagent, RNA isolator
kit, oligonucleotides, and the One Step RT-PCR kit were obtained from QIAGEN GmbH
(Hilden, Germany). N-l(2,3-dioleoyloxy) propyl-NNN-trimethylammonium
methylsulfate (DOTAP) transfection reagent was purchased from Roche Diagnostic
(Mannheim, Germany). The luciferase assay system was purchased from Promega
(Madison WI). ECL detection reagents and HybondTM-N Nylon membranes were
purchased from Amersham Pharmacia Biotech (Little Chalfont, UK) and BrdU staining kit from Zymed Laboratories Inc. (San Francisco, CA). Ethyleneglycoltetra acetic acid
90
(EGTA), phenylmethylsulfonyl fluoride (PMSF), Hoechst 33528, the E-cadherin function blocking antibody, DECMA-1, and 5’-bromo-2´-deoxy-uridine were purchased from
Sigma Chemical Co (St Louis, MO). Doxorubicin, Bcl-2 inhibitor, and wortmannin (PI-3 kinase inhibitor) were obtained from Calbiochem (La Jolla, CA).
3.1.2 Cell lines
The human mammary gland carcinoma cell line, MCF-7, and the immortalized human mammary epithelial cell line, MCF-10A were obtained from the American Type Culture
Collection (ATCC) (Manassas, VA). MCF-7 cell line was stably transfected with an expression plasmid containing the wild-type hGH gene (pMT-hGH) (Liu et al., 1997) under the control of the metallothionein 1a promoter (designated MCF7-hGH) (Kaulsay et al., 1999). For control purposes the ATG start site in pMT-hGH was disabled via a mutation to TTG (stop codon) generated by standard techniques (pMT-MUT) (Liu et al.,
1997), and MCF-7 cells stably transfected with this plasmid were designated MCF7-
MUT (Kaulsay et al., 1999). MCF7-MUT cells therefore transcribe the hGH gene but do not translate the mRNA into protein. A detailed description of the characterization of these cell lines has been published previously (Kaulsay et al., 1999). MCF-7 and MCF-
10A cell lines were stably transfected with a HOXA1 expression plasmid (pSG5-
HOXA1) (Di Rocco et al., 1997) in a ratio of 5:1 by use of Effectene transfection reagent obtained from Qiagen (Hilden, Germany) according to the manufacturer’s instruction.
Positive transfectants were selected in 800 µg/ml G418 in the appropriate culture medium
91
for the respective cell lines. Individual colonies were selected to determine the HOXA1
expression level. Cell lines were established as MCF7-HOXA1 and MCF10A-HOXA1,
respectively, by pooling five individual colonies. As controls, the cell lines were also
stably transfected with the same plasmid which lacks HOXA1 cDNA, named MCF7-
VECTOR and MCF10A-VECTOR respectively. MCF7-HOXA1 and MCF10A-HOXA1
demonstrated higher levels of HOXA1 mRNA, protein, and transcriptional activity in
comparison to MCF7-VECTOR and MCF10A-VECTOR (Zhang et al., 2003). A detailed
description of the characterization of these cell lines has been previously published
(Zhang et al., 2003).
3.1.3 Antibodies
The anti-β-actin, anti-HOXA1, anti-cyclin D1, anti-α-catenin, anti-β-catenin, and anti-γ- catenin antibodies used for Western immunoblotting were obtained from Santa Cruz
Biotechnology (Santa Cruz, CA), anti-p27Kip1 and anti- p21waf/cip antibodies from
Transduction Laboratories (Lexington, KY); and anti-Fas, anti-Fadd, anti-p53, anti-bax,
anti-bak, anti-bcl-2 and anti-bcl-xl were obtained from Oncogene (San Diego, CA).
3.1.4 Plasmids
Plasmids used include a) the expression plasmid containing the hGH gene (pMT-hGH)
under the control of the metallothionein la promoter (Palmiter et al., 1983), b) for control
92
purposes the ATG start site in pMT-hGH was disabled via a mutation to TTG generated
by standard techniques (pMT-MUT) c) pSG5-HOXA1 was a generous gift of Dr.
Vincenzo Zappavigna (Milano, Italy) (Di Rocco et al., 1997). d) EphA2-r42B-LUC was a
generous gift of Dr. Jin Chen (Nashville, TN) (Chen and Ruley, 1998), and e) PBX1
expression vector is a generous gift of Dr. Vincenzo Zappavigna (Milano, Italy) (Di
Rocco et al., 1997). f) the Bcl-2 P1 promoter reporter plasmid was a kind gift of Dr. John
Kurland (Houston, TX) (Kurland et al., 2001). g) the dominantly inhibitory N17 Rac1 mutant Rac1 (DNRac1) was a gift from Dr. Tom Leong, IMCB, NUS, Singapore. f)
HOXA1 RNAi was synthesized in IMCB, NUS, Singapore.
3.2 Experiment procedures
3.2.1 Cell culture
MCF-7, MCF7-MUT, MCF7-hGH, MCF7-VECTOR and MCF7-HOXA1 cells were cultured in RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine
serum (FBS), 100 IU/ml penicillin, 100 µg/ml streptomycin, and 2 mM L-glutamine at
37ºC in 5% CO2. MCF-10A, MCF10A-VECTOR, and MCF10A-HOXA1 cells were
cultured in Dulbecco's modified Eagle's medium/F-12 medium (Invitrogen, Carlsbad,
CA) supplemented with 5% horse serum (HS) (Invitrogen) plus 2 mM glutamine,
100 µg/ml streptomycin, 100 IU/ml penicillin, 0.25 µg/ml ampicillin B, 100 ng/ml cholera toxin, 20 ng/ml epidermal growth factor, 0.5 µg/ml hydrocortisone, and 10 µg/ml
insulin at 37ºC in 5% CO2. To subculture the cells, the cultures were firstly examined
93
carefully for signs of contamination or deterioration. The cell-culture media were
aspirated completely and cells were washed with PBS. Then 0.10–0.25% trypsin was
added to trypsinize cells. After the cells were detached from the dish or flask, media
(containing serum to inactivate the trypsin) were added and cells were dispersed by
repeated pipetting. Finally the cells were seeded in new culture dishes or flasks at the
appropriate concentration.
To freeze and store cells, cells were usually grown to the late log phase and then
trypsinized as above and counted using a hemocytometer. The freezing media were made
by diluting dimethyl sulfoxide (DMSO) to 5-10% in growth media. Cells were
resuspended in freezing media at approximately 1x106-1x107 cells/ml. After the cell
suspension was dispensed into prelabled ampoules, the ampoules were laid on cotton
wool in a polystyrene foam box with a wall thickness of ~ 15 mm. (This box, plus the
cotton wool, provided sufficient insulation such that the ampoules cooled at 1 oC /min when the box was placed at -70oC or -90oC in a regular deep freezer.) When the
ampoules reached -70oC (a minimum of 4-6 h after placing them at -70 oC if starting from
20 oC ambient temprature), they were transferred to a liquid nitrogen freezer. To thaw
frozen cells, the ampoules were retrieved from the freezer, and placed in a waterbath at
37 oC. When the ampoules were thawed, they were swabbed thoroughly with 70%
alcohol and opened. The contents of the ampoules were transferred to a culture flask with
media.
3.2.2 Preparation of total RNA
94
Total RNA extraction was performed using Qiagen mini RNA extraction kit. All steps of
the RNeasy protocol were performed at room temperature. All centrifugation steps were
performed at 20–25°C in a standard microcentrifuge. 1 x 107 cells were usually used for
total RNA extraction. Cells grown in a monolayer in cell-culture vessels were either lysed directly in the culture vessel (up to 10 cm diameter) or trypsinized and collected as cell pellets prior to lysis. For the pelleted cells, 600 µl Buffer RLT (lysis buffer) was added.
The solution was mixed by vortexing or pipeting. β-Mercaptoethanol (β-ME) was added to Buffer RLT before use (10 µl/1ml). For direct lysis of cells grown in a monolayer, 600
µl Buffer RLT was added to the cell-culture dish. Cell lysates were collected with a rubber policeman. Cell lysates were pipeted into a microcentrifuge tube and mixed by vortexing or pipeting to ensure that no cell clumps were visible. Three alternative methods (a, b, or c) were used to homogenize the sample. a. The lysates were pipeted directly onto a QIAshredder spin column placed in a 2 ml collection tube, and centrifuged for 2 min at maximum speed. b. Cells were homogenized for 30 s using a rotor–stator homogenizer. c. Cell lysates were passed at least 5 times through a 20-gauge needle (0.9 mm diameter). 70% (600 µl) ethanol was added to the homogenized lysate, and mixed well by pipetting. Up to 700 µl of the sample, including any precipitate that may have formed, was applied to a RNeasy mini column placed in a 2 ml collection tube. The sample was centrifuged for 15 s at ≥8000 x g (≥10,000 rpm). The flowthrough was
discarded. If the volume exceeded 700 µl, the sample was loaded successively onto the
RNeasy column, and centrifuged as above. The flow-through was discarded after each
centrifugation step. Buffer RW1 (700 µl) was added to the RNeasy column, and centrifuged for 15 s at ≥8000 x g (≥10,000 rpm) to wash the column. The flow-through
95
and collection tube were discarded. Then the RNeasy column was transferred into a new
2 ml collection tube. Buffer RPE (500 µl) was added onto the RNeasy column, and
centrifuged for 15 s at ≥8000 x g (≥10,000 rpm) to wash the column. The flow-through was discarded. Another 500 µl Buffer RPE was added to the RNeasy column, and centrifuged for 2 min at ≥8000 x g (≥10,000 rpm) to dry the RNeasy silica-gel membrane. The RNeasy column was transferred to a new 1.5 ml collection tube. To elute, RNase-free water (30–50 µl) was added to RNeasy silica-gel membrane, and centrifuged for 1 min at ≥8000 x g (≥10,000 rpm). DNase digestion was also performed.
The small residual amounts of DNA remaining were removed using the RNase-Free
DNase Set (Qiagen) for the optional on-column DNase digestion or by a DNase digestion after RNA isolation.
3.2.3 Quantitation, determination of quality, and storage of total RNA
Quantitation of RNA
The concentration of RNA was determined by measuring the absorbance at 260 nm
(A260) in a spectrophotometer. The cuvettes were washed with 0.1M NaOH, 1 mM
EDTA, and then washed with RNase-free water. The water was used to zero the
spectrophotometer. An absorbance of 1 unit at 260 nm corresponds to 40 µg of RNA per
ml. This relation is valid only for measurements in water. Therefore, dilution of the RNA
sample was done in water.
Determination of purity and quality of RNA
96
The ratio of the readings at 260 nm and 280 nm (A260/A280) provides an estimate of the
purity of RNA with respect to contaminants that absorb in the UV, such as protein. Pure
RNA has an A260/A280 ratio of 1.9–2.1. RNA quality was assessed by denaturing
formaldyhyde agarose (FA) gel electrophoresis and ethidium bromide staining.
Formaldehyde agarose (FA) gel electrophoresis
FA gels (1.2% agarose) of size 10 x 14 x 0.7 cm were prepared by mixing 1.2 g agarose,
10 ml 10x FA gel buffer (see composition below), and adding RNase-free water to 100 ml. The mixture was heated to melt the agarose, and cooled to 65°C in a water bath. Then the mixture was mixed with 1.8 ml of 37% (12.3 M) formaldehyde and 1 µl of a 10 mg/ml ethidium bromide stock solution. The agarose was mixed thoroughly and poured onto the gel support. Prior to running the gel, the gel was equilibrated in 1x FA gel running buffer (see composition below) for at least 30 min. One volume of 5x loading buffer (see composition below) was added per 4 volumes of RNA sample (for example
10 µl of loading buffer and 40 µl of RNA). The RNA sample was incubated for 3–5 min at 65°C, chilled on ice, and loaded onto the equilibrated FA gel. Gel was run at 5–7 V/cm in 1x FA gel running buffer. The respective ribosomal bands should appear as sharp bands on the stained gel. Twenty-eight S ribosomal RNA bands should be present with an intensity approximately twice that of the 18S RNA band. If the ribosomal bands in a given lane were not sharp, but appeared as a smear of smaller sized RNAs, it was likely that the RNA sample suffered major degradation during preparation.
97
Composition of FA gel buffers
10x FA Gel buffer
200 mM 3-[N-morpholino]propanesulfonic acid (MOPS) (free acid)
50 mM sodium acetate
10 mM EDTA pH to 7.0 with NaOH
1x FA Gel Running Buffer
100 ml 10x FA gel buffer
20 ml 37% (12.3 M) formaldehyde
880 ml RNase-free water
5x RNA Loading Buffer
16 µl saturated aqueous bromophenol blue solution
80 µl 500 mM EDTA, pH 8.0
720 µl 37% (12.3 M) formaldehyde
2 ml 100% glycerol
3084 µl formamide
4 ml 10 x FA gel buffer
RNase-free water to 10 ml
98
Storage of total RNA
RNA samples with high quality were stored at -80 °C for further analysis.
3.2.4 Reverse transcriptase-polymerase chain reaction (RT-PCR)
Reverse transcription and PCR were carried out sequentially in the same tube by use of
the Qiagen OneStep RT-PCR Kit. All components required for both reactions were added during setup. RT-PCR was performed in a final volume of 50 µl containing 0.2 µg of tRNA template, 0.6 µM gene-specific primers, 2 µl of enzyme mix, 400 µM of each dNTP, 1× reaction buffer, and 1× Q-Solution. All reactions were set up on ice. An
RNase-free environment was maintained during reaction setup. Briefly, RNA template was reverse-transcribed into cDNA for 30 min at 50 °C; Hotstart Taq DNA polymerase was activated by heating for 15 min at 95 °C; the denatured cDNA templates were amplified by the following cycles: 95 °C/30 s, 60 °C/30 s, and 72 °C/60 s. A final extension was performed for 8 min at 72 °C. Fifteen to 40 cycles of PCR were performed, and the amplified β-actin cDNA served as an internal control for cDNA
quantity and quality. All RNA samples were treated with DNaseI to avoid genomic DNA
contaminations.
Sequences of the oligonucleotide primers used for RT-PCR are as follows:
a) HOXA1 sequences:
99
Sense 5'-GGGAAAGTTGGAGAGTACGGC-3';
Antisense 5'-CCTCAGTGGGAGGTAGTCAG-3';
Amplification yielded a 359 bp cDNA product.
b) β-actin sequences:
Sense 5'-ATGATATCGCCGCGCTCG-3';
Antisense 5'-CGCTCGGTGAGGATCTTCA-3'.
Amplification yielded a 564bp cDNA product.
c) Bcl-2 sequences:
Sense 5'-TGCACCTGAC GCCCT TCAC-3';
Antisense 5'-TAGCTGATTCGACGTTTTGCCTGA-3'.
Amplification yielded a 484bp cDNA product.
Amplified PCR products were visualized on a 1% agarose gel.
3.2.5 Agarose gel electrophoresis
The gel (1%, 50mL volume) was made by mixing 0.5g agarose and 50mL 0.5xTBE into a 250mL conical flask. Agarose was dissolved in microwave for about 1 minute. The
100 heated solution was left to cool on the bench for 5 minutes down to about 60°C. While the agarose was cooling, gel tank was prepared on a level surface. Ethidium bromide (1
µl of 10mg/mL) was added to the agarose and swirled to mix. Then the gel was poured slowly into the tank. The bubbles were pushed away to the side by using a disposable tip.
The comb was inserted. The gel was left to set for at least 30 minutes, preferably 1 hour.
Usually 0.5 x TBE buffer (running buffer, see composition below) poured into the gel tank to submerge the gel to 2–5mm depth. The DNA samples were mixed with loading buffer (see composition below) and then loaded to the gel. The first well of the gel was loaded with marker. The gel was stopped when the bromophenol blue (in the loading buffer) ran 3/4 the length of the gel. The gel was carried to the dark-room to look at on the UV light-box, and photographs were taken properly.
Recipe for 2L of 10xTBE
218g Tris base
110g Boric acid
9.3g EDTA
Loading buffer
25mg bromophenol blue or xylene cyanol
4g sucrose
H2O to 10mL
101
3.2.6 Transient transfection and luciferase reporter assay for EphaZ-r42B and Bcl-2
promoter constructs
MCF-7, MCF7-MUT, MCF7-hGH, MCF7-VECTOR, MCF7-HOXA1, MCF10A-
VECTOR, and MCF10A-HOXA1 cells were cultured in 6-well plates the day before
transfection. The cells were incubated at 37ºC in 5% CO2, and grew to 50% or 80%
confluence on the day of transfection. Transient transfection was performed in serum-free
media with EffecteneTM according to the manufacturer's instructions. Firstly, 0.2 µg of
the respective luciferase construct was dissolved in TE buffer, pH 7 to pH 8. Then 8 µl
Enhancer was addedand and mixed by vortexing for 1 s. After incubating at room
temperature (15–25°C) for 2–5 min, the mixture was spun down for a few seconds to
remove drops from the top of the tube. Effectene Transfection Reagent (25 µl) was added to the DNA-Enhancer mixture and mixed by pipetting up and down 5 times, or by vortexing for 10 s. The samples were incubated for 5–10 min at room temperature (15–
25°C) to allow transfection-complex formation. Growth medium (1 ml) was added to the tube containing the transfection complexes, and mixed by pipetting up and down twice.
The transfection complexes were added onto the cells in 6-well plates. Transfected cells were incubated in serum-free media for 12 h before the media were changed to fresh serum-free media with or without 50 nM hGH and 10% FBS or 5% HS. After a further
24 h, cells were washed with PBS and scraped into lysis buffer (25 mM Tris phosphate, pH 7.8, 2 mM DTT, 2 mM 1,2-diaminocyclohexane-N,N,N’,N’,-tetraacetic acid, 10% glycerol, and 1% Triton X-100). Luciferase assays were performed using a luminometer with a pump dispenser. Extract (30 µl) was added to 300 µl of buffer containing 0.12
102
mg/ml D-luciferin, 10 mM DTT, 1 mg/ml bovine serum albumin (BSA), 10 mM
magnesium acetate, 0.1 M Tris acetate, pH 7.75, and 2 mM EDTA. Luciferase
measurements were initiated by injecting 30 µl of 5 mM ATP, 0.5 mM sodium
pyrophosphate and continued for 30 s. The integrated signal was taken as representative of luciferase activity. The protein contents of the samples were normalized and luciferase assays were performed. Results were normalized to the level of β-galactosidase activity to control for transfection efficiency (Wood et al., 1995).
3.2.7 Western Blot analysis
MCF-7, MCF7-MUT, MCF7-hGH, MCF7-VECTOR, MCF7-HOXA1, MCF10A-
VECTOR, and MCF10A-HOXA1 cells were grown to 60% or 100% confluence and serum-deprived for 12 h and then cultured in serum-free media or in serum-free media supplemented with 50 nM hGH, 10% FBS, 5% HS, Doxorubicin, Bcl-2 inhibitor, EGTA, wortmannin or DECMA-1. After 24 hours cells were washed once in ice-cold 1x PBS and lysed at 4°C in lysis buffer (50 mM Tris-HCl, pH 7.4, 1% Triton X-100, 150 mM
NaCl, 1 mM EGTA, 1 mM EDTA, 0.2 mM sodium orthovanadate, 0.5% Nonidet P-40,
0.2% phenylmethylsulfonyl fluoride) for 30 min with regular vortexing. Cell lysates were
then centrifuged at 4°C at 14000 x g for 15 min, the resulting supernatants were
collected, and protein concentration determined in triplicate by the Lowry protein assay
method, using BSA as a standard. Equal amounts of proteins were used in each
experiment. Remaining protein samples could be stored at -80 °C for further analysis. 5x
SDS sample buffer (50 mM Tris-HCl, pH 6.6, 5% SDS, 5% β-mercaptoethanol and
103
bromphenol blue) was added to 20 µg of total protein, boiled for 10 min, and centrifuged
at 14000 x g for 5 min. Proteins were separated by 10% SDS-PAGE (1.5 M Tris-HCL, pH 8.8, 20% (w/v) SDS, acrylamide/bis-acrylamide (30%/0.8% w/v), 10% (w/v) ammonium persulfate, 0.05% TEMED) in 1x Laemlli running buffer (5x Laemllu running buffer: 120 mM tris[hydroxymethyl]methylamine, 960 mM glycine, 17 mM
sodium dodecyl sulphate [SDS]). Proteins were then transferred to nitrocellulose
membrane using a standard semidry electroblotting apparatus in Laemlli buffer (23 mM
Tris[hydroxymethyl]methylamine, 19 mM glycine, 0.64 mM SDS) containing 10%
methanol. Subsequently, nitrocellulose membranes were blocked with 5% non-fat dry
milk in phosphate buffered saline with 0.1% Tween 20 (PBST) for 2 h at 22 °C and then
washed twice for 5 min each in PBST. The blots were then treated for 1 h at 22 °C with
the primary antibodies (HOXA1 at 1:500 and β-actin, cyclin D1, p27Kip1, p21waf/cip, Fas,
Fadd, p53, bax, bak, bcl-2, bcl-xl, α-catenin, β-catenin, γ-catenin at 1:1000) in PBST containing 1% non-fat dry milk. After 3 washes for 15 min each in PBST, membranes were incubated in goat anti-mouse (1:1000), goat anti-rabbit IgG (1:1000) or donkey anti- goat (1:1000) horseradish peroxidase-conjugated second antibodies in PBST containing
1% non-fat dry milk for 1 h at 22 °C. Membranes were further washed 3 times for 15 min each in PBST before immunolabeling was detected using the enhanced chemiluminescence (ECL) system according to the manufacturer’s instruction. Reprobing of the membranes was performed after stripping in buffer containg 20 mM Tris-HCL, pH
6.8, 2% SDS, and 100mM β-mercaptoethanol and incubated at 70°C for 30 min.
3.2.8 5'-Bromo-2'-deoxyuridine incorporation assay (Brdu)
104
Mitogenesis was directly assayed by measuring the incorporation of BrdU (Kaulsay et al.,
2001). For incorporation of BrdU, subconfluent or confluent MCF-7, MCF7-MUT,
MCF7-hGH, MCF7-VECTOR, MCF7-HOXA1, MCF10A-VECTOR, and MCF10A-
HOXA1 cells were washed twice with PBS and seeded to glass coverslips in either
serum-free medium or serum-free medium supplemented with 50 nM hGH, 10% FBS, or
5% HS for 24 h. The cells were pulse-labeled with 20 mM BrdU for 30 min, washed twice with PBS, and fixed in cold 70% ethanol for 30 min. After that, the cells were washed in 3 changes of distilled water, and BrdU detection was performed by using the
BrdU staining kit (Zymed, South San Francisco, CA) according to the manufacturer's instructions. Briefly, the cells were firstly incubated with denaturing solution for 30 min and rinsed in PBS (2 min, 3 times). Then blocking solution was added to the cells (30 min, room temperature) and blotted off. The cells were incubated with biotinylated mouse anti-Brdu (30-60 min, room temperature), and streptavidin-peroxidase (10 min, room temperature) successively, and rinsed with PBS (2 min, 3 times). Sufficient quantity of the DAB mixture was added to cover the cells (4 min), and rinsed well with distilled water. The specimen was counterstained with hematoxylin, and washed in tap water, and then put into PBS until sections turn blue (approx. 30 seconds). Finally, the slides were dehydrated in a graded series of alcohol, and cleared with xylene. The slides were examined after adding 2 drops of histomount and coverslip. A total population of over 3 times 300 cells was analyzed in several arbitrarily chosen microscopic fields to determine the BrdU labeling index (percentage of cells synthesizing DNA).
105
3.2.9 Use of Bcl-2 antisense oligonucleotide to deplete cellular Bcl-2
The Bcl-2 sense and antisense oligonucleotides utilized were obtained from Calbiochem.
Bcl-2 (Sense) 5’-TACCGCGTGCGACCCTCT-3’;
(antisense) 5’-TCTCCCAGCGTGCGCCAT-3’;
Cells were cultured for 24 hours in serum free medium and then transfected with 800 nM
Bcl-2 sense and antisense oligonucleotides in DOTAP (10 µl/ml medium). Briefly, the
oligonucleotides were mixed with DOTAP gently by carefully pipetting several times.
The mixtures were incubated at room temperature for 15-20 min, and added onto the cells
in 6-well plates. The cells were incubated with the complex for 8 hours at 37°C in a CO2 incubator. Cells were either lysed and processed as described above for Western blot analysis or processed for measurement of apoptosis.
3.2.10 Measurement of apoptosis
Apoptotic cell death was measured by fluorescent microscopic analysis of cell DNA staining patterns with Hoechst 33258. MCF-7, MCF7-MUT, MCF7-hGH, MCF7-
VECTOR, MCF7-HOXA1, MCF10A-VECTOR, and MCF10A-HOXA1 cells were trypsinized with 0.5% trypsin and washed twice with serum-free medium. The cells were then seeded to glass cover slips in 6-well plates and incubated in serum-free medium with or without 50 nM hGH, 10% FBS, doxorubicin, Bcl-2 inhibitor, Bcl-2 antisense,
DECMA-1 and HOXA1 RNAi. After a culture period of 24 h, the cells were fixed for 20
106
min in 4% paraformaldehyde in 1X PBS, pH 7.4, at room temperature. The cells were
then rinsed twice in PBS and stained with the karyophilic dye Hoechst 33258 (20 µg/ml)
for 5 min at room temperature. Following washing with 1X PBS, nuclear morphology
was examined under an ultrraviolet-visible fluorescence microscope (Zeiss Axioplan).
Apoptotic cells were distinguished from viable cells by their nuclear morphology characterized by nuclear condensation and fragmentation as well as the higher intensity of the blue fluorescence of the nuclei. For statistical analysis, 3 times 300 cells were counted in eight random microscopic fields at 400× magnification.
3.2.11 Determination of anchorage-dependent and -independent growth of
transfected cells
MCF7-VECTOR, MCF7-HOXA1, MCF10A-VECTOR, and MCF10A-HOXA1 cells (5
x 104) were seeded into 6-well plates in monolayers in serum-free media or in serum-free
media supplemented with either 50 nM hGH or 10% FBS. After 48 h, media were
aspirated, and cells were washed with PBS. Then the cells were trypsinized by adding
0.10–0.25% trypsin in PBS. After cells detached from the dish or flask, media
(containing serum) were added to inactivate the trypsin. Cell numbers were determined
using a hemocytometer. MCF7-VECTOR and MCF7-HOXA1 cells (5 x 104) in suspension culture were grown in 30-mm plastic bacteriological dishes (Sterilin,
Teddington, United Kingdom). On the days indicated, cells were harvested and counted
as for monolayers. For soft agar colony formation assay, three layers were prepared. 1.
Base agar lay. 1% Agar (DNA grade) was melt in microwave, cooled to 40oC in a
107
waterbath. 2X RPMI with 20% FBS was warm to 40oC in waterbath. At least 30 minutes were needed for temperature to equilibrate. Equal volumes of the two solutions were mixed to give 0.5% agar + 1X RPMI + 10% FBS. 1.5 ml base agar was added to each well (6-well plate) and allowed to set. The plates can be stored at 4oC for up to 1 week. 2.
Top agar layer. 0.7% agar (DNA grade agarose) was melted in microwave, and cooled to
40oC in a waterbath. (It is important not to exceed 40oC, otherwise cells can be killed).
2X RPMI with 20% FBS was also warmed to the same temperature. Equal volumes of
the two solutions were mixed to give 0.35% agar + 1X RPMI + 10% FBS. MCF7-
VECTOR and MCF7-HOXA1 cells were trypsinized, counted, and seeded with 1.5 ml
top agar onto the base agar (5 x 103 cells/well). 3. Top layer. Medium was added to prevent drying of the agarose gels. The MCF10A-VECTOR and MCF10A-HOXA1 cells in soft agar colony assay were same with the process above except that the medium changed to DMEM-F12 containing appropriate components described in cell culture above, and 1 x 104 cells were seeded to the top agar layer of the soft agar. The plates
were incubated at 37oC in humidified incubator for 9 days for MCF7-VECTOR and
MCF7-HOXA1 cells and 14 days for MCF10A-VECTOR and MCF10A-HOXA1, after
which the cultures were stained (1.5ml of 0.005% crystal violet for >1 hour), counted,
and photographed.
3.2.12 Xenograft studies
3.2.12.1 Mice
108
Nude mice and severe combined immunodeficient (SCID) mice models were used as host
animals for xenograft studies. The two species were used because the significant
differences in their recipient to tumor cells. Although nude mice have been used for many
years to grow various human tumors, nude mice and thymectomized mice can regain
immune competence and ultimately reject the tumor after some time (Gerald et al., 1991).
SCID mice are characterized both by T- and B-cell deficiency (Bosma et al., 1983). SCID mice have enhanced take rates for tumors compared to genetically athymic nude mice,
which are T-cell deficient (Giovanella et al., 1974), and also compared to neonatally
thymectomized mice (Gerald et al., 1991).
Four- to 6-week-old pathogen-free nude mice and SCID mice were purchased from the
animal holding facility of the National University of Singapore. The mice were
acclimated for 7–10 days after purchase. All the mice were housed and bred in specific
pathogen-free conditions at constant temperature (24-26oC) and humidity (30-50%).
Autoclaved standard chow and water were given ad libitum. On the day of implantation,
the mice were anaesthetised in the animal laboratory.
3.2.12.2 Cells preparation and implantation
MCF10A-VECTOR, and MCF10A-HOXA1 were cultured in Dulbecco's modified
Eagle's medium/F-12 medium supplemented with 5% horse serum plus 2 mM glutamine,
100 µg/ml streptomycin, 100 IU/ml penicillin, 0.25 µg/ml ampicillin B, 100 ng/ml cholera toxin, 20 ng/ml epidermal growth factor, 0.5 µg/ml hydrocortisone, and 10 µg/ml
109
insulin at 37ºC in 5% CO2. After the cells were grown to the log phase, they were
trypsinized as detailed in the cell culture section above and counted using a
hemocytometer. Cells were then re-suspended in cold PBS or matrigel (5 x 107 cells per
ml), and the suspension was dispensed into prelabled ampoules and placed on ice. After
implantation, the viability of the remaining cells in excess ampoules was examined by
trypan blue exclusion. One drop of cell suspension was mixed with one drop of trypan
blue and loaded to the counting chamber of the hemocytometer. After 1-2 min, the cell
viability was examined under a microscope (10 x objective) to ensure that the time lag
from sequential implantation did not significantly affect cell viability and cell number.
In the first series of s.c. implantation, 0.1 ml (5 x 106) vector- or HOXA1-transfected
MCF-10A cells were injected into the first mammary (axillary) fat pad of nude mice using a 100 µl syringe and a 23-gauge needle. In the second series of implantation, 0.1 ml
(5 x 106) vector- or HOXA1-transfected MCF-10A cells were injected into the first mammary (axillary) fat pad of SCID mice using a 100 µl syringe and a 23-gauge needle.
The SCID mice simultaneously received a 60-day release pellet containing 0.72 mg of E2
(Sigma) as previous investigators have shown that without estrogen involvement, breast
tumors fail to grow in immunodeficient mice (Brunner et al., 1987; Lippman et al., 1987;
Phillips et al., 1989). Ten mice were used for each group. The positive controls (nude
mice and SCID mice implanted with human mammary carcinoma cells, MCF-7) and
negative controls (nude mice and SCID mice implanted with untransfected MCF-10A
cells) were included. After implantation, the mice were put back to the animal laboratory.
110
3.2.12.3 Tumor growth and autopsy
The presence of tumours was checked every second day by tactile inspection of the axilla
of every mouse. At about 2 or 3 weeks after implantation, the animals were killed by cervical dislocation. Immediately afterwards, an autopsy of the mice was performed.
Primary tumors and all organs were evaluated macroscopically for the presence of
tumors. Tumor volume was calculated using the formula, volume = (d1 x d2 x d3) x
0.5236, in which dn represents the three orthogonal diameter measurements, measured
with callipers.
3.2.12.4 Histology
Tissue samples of the primary tumor and organs were dissected, fixed in phosphate-
buffered 4% paraformaldehyde, dehydrated, embedded in paraffin casts, cut in 6-µm-
thick sections and stained with haematoxylin and eosin (H-E) for histopathological examination. Haematoxylin and eosin (H-E) staining followed standard procedures
Solutions were prepared for the staining including 95% ethanol, 100% ethanol, xylene,
0.1% ammonium hydroxide, H2O, hematoxylin, and eosin. The developed slides were
stained for 20 to 30 sec in hematoxylin, and rinsed two times in water. After quickly
incubated with 0.1% ammonium hydroxide and washed in water, the slides were stained for 20 to 30 sec in eosin. The slides were then progressively dehydrated in 95% ethanol,
100% ethanol, and xylene. Finally, the slides were mounted and examined microscopically.
111
To freeze the biopsies, the tumors were chopped into about 3-4 mm pieces, and placed in
ampoules. The ampoules with the tumors were then transferred to -70 oC freezer or
liquid-nitrogen freezer.
3.2.13 Calcium switch
For calcium switch experiments, MCF-7 cells were seeded into dishes, cultured for 24 hours, and then the cells were serum starved. After 24 hours incubation, E-cadherin- mediated cell-cell contacts were disrupted by treatment with 4 mM EGTA at 37oC for 30 minutes. Thereafter, intercellular contacts were allowed to reform in the presence of
2+ o normal Ca containing medium (CaCl2~ 1.8 mM) at 37 C at various time points with or
without E-cadherin function-blocking antibodies (DECMA clone-1 at 1:100 dilution)
(Sigma-Aldrich).
3.2.14 E-cadherin cell adhesion assays (hE/Fc experiments)
The recombinant protein (hE/Fc) consisting of the complete ectodomain of human E-
cadherin fused to the Fc region of IgG was the generous gift of Dr. Alpha Yap (Brisbane,
Australia) (Kovacs et al., 2002). Substrata adsorbed with hE/Fc supported E-cadherin- specific adhesion and contact formation (Kovacs et al., 2002). Using this model, E- cadherin engagement is induced without cell-cell contact (Kovacs et al., 2002). For hE/Fc experiment, bacterial 10cm dishes (precoated with nitrocellulose) were coated with hE/Fc
112
(50µg/ml) for 2 h at room temperature. Blocking of non-specific binding sites was performed with 1% BSA in Hanks Balanced Salt solution supplemented with 5 mM
CaCl2 for 1h and 30min at room temperature. Dishes were washed 5 times before addition of cells. MCF-7 cells were washed once with 10mM EDTA in Hanks/ CaCl2 for
1 min and directly trypsinized in 0.01% crystalline trypsin in Hanks/CaCl2 for 15 min at
o 37 C. MCF-7 cells were then washed with Hanks/CaCl2 and resuspended in
Hanks/CaCl2 supplemented with 0.05% FCS, seeded onto the 10 cm dishes and incubated at 37 oC. Thirty minutes after seeding, non-adhered cells were carefully removed by exchange of media. Five hours after seeding the cells onto the dishes, RNA and protein extraction were started.
3.2.15 Statistics
All experiments were repeated at least three to five times. All numerical data are expressed as mean ± SD. Data were analyzed using the two tailed t-test or ANOVA.
113
4.1 Human growth hormone-regulated HOXA1 is a human mammary
epithelial oncogene
The upregulation of HOXA1 gene by autocrine production of hGH was addressed in this
series of studies. The studies also investigated the role of HOXA1 in human mammary
carcinoma cells (MCF-7) and the immortalized but not transformed human mammary epithelial cells (MCF-10A). Increased epithelial expression of the hGH gene is associated
with the acquisition of pathological proliferation, and the highest level of hGH gene
expression is observed in metastatic mammary carcinoma cells (Raccurt et al., 2002).
One major mechanism by which GH affects cellular and somatic function is by regulating
the level of specific mRNA species (Isakasson et al., 1985). cDNA microarray analyses
have previously been utilized to identify genes regulated by autocrine production of hGH
in human mammary carcinoma cells (MCF-7) (Mertani et al., 2001). One gene
demonstrated to be upregulated by autocrine hGH in MCF-7 cells was the homeobox
containing transcription factor HOXA1 (Mertani et al., 2001). Alterations of HOX
expression have been detected in a variety of human tumours (Cillo et al., 1992; Celetti et
al., 1993), including those of the mammary gland (Care et al., 1998; Raman et al., 2000).
Hoxa1 was expressed only in cancerous but not in normal or precancerous mouse
mammary tissue (Friedmann et al., 1994), suggesting that it may play a role in
development of mammary malignancies.
114
Autocrine hGH production by human mammary carcinoma cells increased the expression and transcriptional activity of HOXA1. Forced expression of HOXA1 in human mammary carcinoma cells resulted in increased total cell number primarily by the promotion of cell survival mediated by the transcriptional upregulation of Bcl-2. HOXA1 also abrogated the apoptotic response of mammary carcinoma cells to doxorubicin.
Forced expression of HOXA1 in mammary carcinoma cells, in a Bcl-2-dependent manner, resulted in dramatic enhancement of anchorage-independent proliferation and colony formation in soft agar. Finally, forced expression of HOXA1 was sufficient to result in the oncogenic transformation of immortalized human mammary epithelial cells with aggressive in vivo tumor formation in severe combined immunodeficient mice (SCID mice). Thus, human growth hormone-regulated HOXA1 is a human mammary epithelial oncogene.
115
4.1.1 Autocrine hGH stimulation of mammary carcinoma cells increases
HOXA1 mRNA, protein, and transcriptional activity
HOXA1 had previously been identified by cDNA microarray analysis as an autocrine hGH-regulated gene in human mammary carcinoma cells (Mertani et al., 2001). To verify the autocrine hGH-dependent upregulation of HOXA1 mRNA observed by cDNA array screening in mammary carcinoma cells, the level of HOXA1 mRNA in MCF7-MUT and
MCF7-hGH cells was examined by semi-quantitative RT-PCR.
Figure 4.1.1 A shows the RT-PCR products on an agarose gel. One amplified fragment of the predicted size (359 bp) appropriate for HOXA1 mRNA was detected in MCF7-
MUT and MCF7-hGH cells. Autocrine production of hGH by MCF7-hGH cells resulted in an increased level of HOXA1 mRNA in comparison to MCF7-MUT cells. Exogenous hGH and FBS also increased HOXA1 mRNA in MCF7-MUT cells and potentiated the increased level of HOXA1 mRNA observed in MCF7-hGH cells due to autocrine production of hGH. The level of β-actin mRNA did not differ between the two cell lines under the different treatment conditions and was used as a control for RNA quality
(Figure 4.1.1 A).
To determine if the autocrine hGH-stimulated increase in HOXA1 mRNA also resulted in increased HOXA1 protein, we examined the level of HOXA1 protein in lysates collected
116 from MCF7-MUT and MCF7-hGH cells by Western blot analysis. Figure 4.1.1 B shows that both the 33-kDa and the 35-kDa forms of HOXA1 were detected in lysates from
MCF7-MUT and MCF7-hGH cells with increased HOXA1 protein observed in MCF7- hGH cells. FBS also increased the level of HOXA1 in MCF7-MUT cells and potentiated the level of HOXA1 in MCF7-hGH cells. Exogenous hGH did not stimulate an increase in HOXA1 protein in MCF7-MUT cells and only minimally enhanced the level in
MCF7-hGH cells (Figure 4.1.1 B).
To determine if the autocrine hGH-stimulated increase in HOXA1 protein also resulted in increased HOXA1-mediated transcription, we examined luciferase reporter activity from the EphA2-r42B enhancer which contains four HOX-PBX binding sites (Chen and Ruley,
1998). Figure 4.1.1C shows that autocrine production of hGH by MCF7-hGH cells resulted in increased luciferase activity from the EphA2-r42B enhancer compared with
MCF7-MUT cells indicative of increased HOXA1 transcriptional activity. Transient transfection of HOXA1 cDNA increased HOXA1 transcriptional activity in MCF7-MUT cells and dramatically enhanced the already greater HOXA1 transcriptional response in
MCF7-hGH cells. Transient transfection of MCF7-MUT cells with cDNA for the HOX binding partner PBX1 was without effect, whereas transient transfection of PBX1 cDNA dramatically increased HOXA1-mediated transcription in MCF7-hGH cells. Transient transfection of MCF7-MUT cells with both HOXA1 and PBX1 cDNAs resulted in a robust transcriptional response through the EphA2-r42B enhancer. Transient transfection of MCF7-hGH cells with both PBX1 and HOXA1 cDNAs synergistically increased
HOXA1-mediated transcription above the enhancement observed with either HOXA1 or
117
PBX1 alone (Figure 4.1.1 C). Thus, autocrine hGH production by mammary carcinoma
cells result in increased HOXA1-mediated transcriptional activity.
A
MCF7-MUT MCF7-hGH 500bp 400bp HOXA1 300bp
600bp β-Actin 500bp 400bp hGH - + - - + - FBS - - + - - +
Figure 4.1.1 (A) Effect of autocrine hGH on HOXA1 mRNA in mammary carcinoma cells. MCF7-MUT and MCF7-hGH cells were cultured in serum-free media or in serum-free media supplemented with either 50 nM hGH or 10% FBS. Experiments were performed as described under “Experimental Procedures.” The level of HOXA1 mRNA was determined by RT-PCR as indicated. β-Actin was used as loading control where appropriate.
118
B
MCF7-MUT MCF7-hGH
HOXA1
30 kDa
45 kDa β-Actin
hGH - + - - + - FBS - - + - - +
Figure 4.1.1 (B) Effect of autocrine hGH on HOXA1 protein in mammary carcinoma cells. MCF7-MUT and MCF7-hGH cells were cultured in serum-free media or in serum-free media supplemented with either 50 nM hGH or 10% FBS. Experiments were performed as described under “Experimental Procedures.” The level of HOXA1 protein was determined by Western blot analysis as indicated. β- Actin was used as loading control where appropriate.
119
C VECTOR * HOXA1 * 8 PBX1 HOXA1+PBX1
6 * * * 4 * * 2 ARBITRARY UNIT
0 MCF7-MUT MCF7-hGH
Figure 4.1.1 (C) Effect of autocrine hGH on HOXA1 transcriptional activity in mammary carcinoma cells. MCF7-MUT and MCF7-hGH cells were cultured in serum-free media or in serum-free media supplemented with either 50 nM hGH or 10% FBS. Experiments were performed as described under “Experimental Procedures.” The level of HOXA1 transcriptional activity was determined by reporter assay as indicated. *, p < 0.01.
120
4.1.2 Forced expression of HOXA1 in human mammary carcinoma cells
results in increased total cell number
To determine the functional consequences of the autocrine hGH-stimulated increased
HOXA1 expression in mammary carcinoma cells we generated a mammary carcinoma
cell line (MCF-7) with increased expression of HOXA1 by stable transfection of HOXA1
cDNA (termed MCF7-HOXA1). Figure 4.1.2 A shows the RT-PCR results which
demonstrated clearly the increased expression of HOXA1 mRNA in MCF7-HOXA1 cells
compared with vector-transfected (MCF7-VECTOR) cells. Figure 4.1.2 B shows the
Western blot results of the same cells and demonstrated higher levels of HOXA1 protein
in MCF7-HOXA1 compared with MCF7-VECTOR cells.
Analysis of HOXA1 transcriptional activity by use of the EphA2-r42B enhancer
demonstrated increased HOXA1-mediated transcriptional activity in MCF7-HOXA1
cells compared with MCF7-VECTOR cells under serum-free conditions or when
stimulated with exogenous hGH or FBS (Figure 4.1.2C). Transient transfection of cDNA
for the HOXA1 binding partner PBX1 in MCF7-VECTOR cells did not alter the level of
HOXA1-mediated transcriptional activity under serum free conditions or exogenous hGH
stimulation and only slightly enhanced HOXA1-mediated transcriptional activity of
MCF7-VECTOR cells in serum. In contrast, transient transfection of PBX1 cDNA in
MCF7-HOXA1 cells synergistically amplified under all conditions, the increased
121
HOXA1-mediated transcription as a result of forced expression of HOXA1 (Figure
4.1.2C).
We therefore examined the effect of forced expression of HOXA1 in mammary carcinoma cells on total cell number. MCF7-VECTOR and MCF7-HOXA1 cells were plated in identical number under serum-free conditions and with either 50 nM exogenous hGH or 10% FBS, and the cell number was determined after 48 hours. Exogenous hGH stimulated an increase in MCF7-VECTOR cell number as did FBS (although to a greater extent). MCF7-HOXA1 cells cultured in serum-free media increased in number dramatically more than MCF7-VECTOR cells (Figure 4.1.2 D). Exogenous hGH and
FBS also increased cell number in MCF7-HOXA1 above that observed in serum free conditions but similar to the percentage increases observed with hGH and FBS stimulation of MCF7-VECTOR cells. In any case, forced expression of HOXA1 in mammary carcinoma cells resulted in a significant increase in cell number.
Increased cell number is achieved by either increased proliferation or decreased apoptotic cell death, and we therefore proceeded to determine the relative contribution of these processes to the observed increase in cell number as a consequence of forced expression of HOXA1.
The D family of cyclins is pivotal to initiate progression through the G1 phase of the cell
cycle (Sherr, 1995). Cyclin D1 is the predominant member of this family expressed in
mammary gland (Musgrove et al., 1996) and it was also demonstrated that cyclin D1 is
required for autocrine hGH-stimulated mammary carcinoma cell cycle progression
122
(Graichen et al., 2002). Therefore the level of cyclin D1 protein was examined in MCF7-
VECTOR compared with MCF7-HOXA1 cells. The level of cyclin D1 protein was
increased in MCF7-HOXA1 compared with MCF7-VECTOR cells (Figure 4.1.2 E)
under serum-free conditions. Both exogenous hGH and FBS increased cyclinD1 protein in MCF7-VECTOR cells and accentuated the increase in cyclinD1 observed upon forced expression of HOXA1 in MCF7-HOXA1 cells (Figure 4.1.2 E). We consequently also examined the expression of the cyclin-dependent kinase (cdk) inhibitor p21waf1/cip1 in
MCF7-VECTOR and MCF7-HOXA1 cells. MCF7-HOXA1 cells exhibited a marked decrease in p21waf1/cip1 protein in comparison to MCF7-VECTOR cells (Figure 4.1.2 E).
We next examined the protein level of the cdk inhibitor p27Kip1. Decreased expression of
Kip1 p27 is associated with G1/S phase transition (Sgambato et al., 2000). The level of
p27Kip1 was not altered between MCF7-VECTOR and MCF7-HOXA1 cells (Figure 4.1.2
E) under the conditions studied. Equal loading of the cell extracts was verified by
reprobing the stripped membrane for β-actin (Figure 4.1.2 E). To determine the effects of
forced expression of HOXA1 on cell cycle progression in MCF7-VECTOR and MCF7-
HOXA1 cells, we examined the nuclear incorporation of 5’-bromo-2’-deoxyuridine
(Brdu) in these cells. Although MCF7-HOXA1 cells exhibited a significantly higher
percentage of nuclear BrdU incorporation compared with MCF7-VECTOR cells (Figure
4.1.2 F) the differences were small and could not account for the observed differences in
total cell number between MCF7-VECTOR and MCF7-HOXA1 cells.
We therefore examined the effect of forced expression of HOXA1 on a variety of
proteins involved in the apoptotic process. The level of Bax, Bak, Bcl-xL, Fas and FADD
123
did not differ between MCF7-VECTOR and MCF7-HOXA1 cells under serum-free
conditions nor with stimulation by either exogenous hGH or FBS (Figure 4.1.2 G). The
level of p53 did not differ between MCF7-VECTOR and MCF7-HOXA1 cells in serum-
free media nor by stimulation with exogenous hGH but was dramatically reduced in both cell lines in the presence of FBS. Interestingly however, the level of Bcl-2 was dramatically increased in MCF7-HOXA1 cells compared with MCF7-VECTOR cells in serum-free media, and the level of Bcl-2 was even further increased in MCF7-HOXA1 cells in the presence of FBS. Apoptotic cell death was also dramatically reduced in
MCF7-HOXA1 cells compared with MCF7-VECTOR cells in serum-free conditions
(Figure 4.1.2 H). Exogenous hGH slightly reduced apoptotic cell death in both cell lines.
FBS functioned as a powerful survival stimulus for both cell lines, although apoptosis
was still less in MCF7-HOXA1 cells compared with MCF7-VECTOR cells (Figure 4.1.2
H).
124
A
MCF7-VECTOR MCF7-HOXA1 500b 400b HOXA1 300b
600b 500b β-Actin 400b
Figure 4.1.2 (A) Demonstration of overexpression of HOXA1 mRNA upon stable transfection of MCF-7 cells with HOXA1 cDNA. MCF-7 cells were stably transfected with either the empty vector (MCF7-VECTOR) or the vector containing HOXA1 cDNA (MCF7-HOXA1). The level of HOXA1 mRNA was determined by RT-PCR. β-Actin was used as loading control.
125
B
MCF7-VECTOR MCF7-HOXA1
HOXA1 30 kDa
45 kDa β-Actin
Figure 4.1.2 (B) Demonstration of overexpression of HOXA1 protein upon stable transfection of MCF-7 cells with HOXA1 cDNA. MCF-7 cells were stably transfected with either the empty vector (MCF7-VECTOR) or the vector containing HOXA1 cDNA (MCF7-HOXA1). The level of HOXA1 protein was determined by Western blot analysis. β-Actin was used as loading control.
126
C * VEHICLE 50 * hGH FBS 40 * * 30 * *
20
ARBITRARY UNIT 10
0 PBX1 - + - + - + - + - + - + MCF7-VECTOR MCF7-HOXA1
Figure 4.1.2 (C) Demonstration of functional overexpression of HOXA1 protein upon stable transfection of MCF-7 cells with HOXA1 cDNA. MCF-7 cells were stably transfected with either the empty vector (MCF7-VECTOR) or the vector containing HOXA1 cDNA (MCF7-HOXA1). HOXA1 transcriptional activity (in the presence and absence of its binding partner PBX1) was determined by reporter assay. *, p < 0.01.
127
D
VEHICLE
hGH
FBS ) 3 200 * (x10
150 * * 100 * 50 TOTAL CELL NUMBER
0 MCF7-VECTOR MCF7-HOXA1
Figure 4.1.2 (D) Effect of forced expression of HOXA1 in mammary carcinoma cells on cell number. MCF7-VECTOR and MCF7-HOXA1 cells were cultured in serum-free media or in serum-free media supplemented with either 50 nM hGH or 10% FBS as indicated. Total cell number was determined in both cell lines under the indicated conditions as detailed under “Experimental Procedures.” Results represent means ± S.D. of triplicate determinations. *, p < 0.01.
128
E
MCF7-VECTOR MCF7-HOXA1
Cyclin D1 30kDa
p21 20kDa
30kDa p27
β-Actin 45kDa
hGH - + - - + - FBS - - + - - +
Figure 4.1.2 (E) Effect of forced expression of HOXA1 in mammary carcinoma cells on cell cycle progression. MCF7-VECTOR and MCF7-HOXA1 cells were cultured in serum-free media or in serum-free media supplemented with either 50 nM hGH or 10% FBS as indicated. Western blot analysis was utilized to determine the level of proteins involved in cell cycle progression (cyclinD1, p21waf1/cip1, and p27Kip1). β-Actin was used as loading control.
129
F VEHICLE
hGH 25 FBS * *
20 * *
15
10
5 PROPORTION CELLS WITH NUCLEAR 0 MCF7-VECTOR MCF7-HOXA1
Figure 4.1.2 (F) Effect of forced expression of HOXA1 in mammary carcinoma cells on cell cycle progression. MCF7-VECTOR and MCF7-HOXA1 cells were cultured in serum-free media or in serum-free media supplemented with either 50 nM hGH or 10% FBS as indicated. Cell cycle progression (BrdU incorporation) was determined in both cell lines under the indicated conditions as detailed under “Experimental Procedures.” Results represent means ± S.D. of triplicate determinations. **, p < 0.05.
130
G
MCF7-VECTOR MCF7-HOXA1
Fas 45 kDa
Fadd 25 kDa
p53 45 kDa Bax 20 kDa
Bcl-xL 30 kDa 30 kDa Bak
30 kDa Bcl-2
45 kDa β-Actin
hGH - + - - + - FBS - - + - - +
Figure 4.1.2 (G) Effect of forced expression of HOXA1 in mammary carcinoma cells on apoptosis. MCF7-VECTOR and MCF7-HOXA1 cells were cultured in serum-free media or in serum-free media supplemented with either 50 nM hGH or 10% FBS as indicated. Western blot analysis was utilized to determine the level of proteins involved in apoptosis (Fas, Fadd, p53, Bax, Bak, Bcl-xL, and Bcl-2). β-Actin was used as loading control.
131
H
VEHICLE 60 hGH * FBS * 40
20 PERCE NTAGE APOPTOTIC NUCLEI 0 MCF7-VECTOR MCF7-HOXA1
Figure 4.1.2 (H) Effect of forced expression of HOXA1 in mammary carcinoma cells on apoptosis. MCF7-VECTOR and MCF7-HOXA1 cells were cultured in serum-free media or in serum-free media supplemented with either 50 nM hGH or 10% FBS as indicated. Apoptotic cell death was determined in both cell lines under the indicated conditions as detailed under “Experimental Procedures.” Results represent means ± S.D. of triplicate determinations. *, p < 0.01.
132
4.1.3 HOXA1 expression in human mammary carcinoma cells prevents
apoptotic cell death in a Bcl-2-dependent manner
The previous studies showed that forced expression of HOXA1 in mammary carcinoma
cells resulted in decreased apoptotic cell death associated with a specific increase in Bcl-
2 protein. To determine if HOXA1 regulated Bcl-2 at the transcriptional level, we examined the effect of forced expression of HOXA1 in mammary carcinoma cells on the level of Bcl-2 mRNA. Increased Bcl-2 mRNA was observed in MCF7-HOXA1 in comparison to MCF7-VECTOR cells in serum-free media (Figure 4.1.3 A). FBS stimulation of MCF7-VECTOR cells also increased Bcl-2 mRNA in comparison to the serum-free state, and FBS enhanced the level of Bcl-2 mRNA in MCF7-HOXA1 cells.
Examination of Bcl-2 promoter activity demonstrated that the effect of forced expression of HOXA1 on Bcl-2 mRNA expression was mediated at the transcriptional level. Thus,
Bcl-2 promoter activity was significantly higher in MCF7-HOXA1 cells in comparison to
MCF7-VECTOR cells in serum-free conditions (Figure 4.1.3 B). Exogenous hGH was without effect on Bcl-2 promoter activity, but FBS stimulated Bcl-2 gene transcription in
MCF7-VECTOR cells and dramatically enhanced Bcl-2 promoter activity in MCF7-
HOXA1 cells. Thus HOXA1 regulates Bcl-2 at the transcriptional level.
To determine if Bcl-2 was responsible for the dramatic decrease in apoptosis as a consequence of forced expression of HOXA1, we prevented expression of Bcl-2 by use of antisense oligonucletides (Chi et al., 2000) as observed in Figure 4.1.3 C. Transfection of antisense oligonucleotides to Bcl-2 prevented the HOXA1-stimulated expression of
133
Bcl-2 in MCF7-HOXA1 cells, whereas sense control oligonucleotides did not
significantly affect the level of Bcl-2 (Figure 4.1.3 C). Transfection of Bcl-2 antisense
oligonucleotides in MCF7-HOXA1 increased apoptotic cell death above that in control or
sense oligonucleotide-transfected MCF7-VECTOR cells (Figure 4.1.3 D), and Bcl-2
antisense also increased apoptotic cell death in MCF7-VECTOR cells. To further verify the Bcl-2 dependence of the survival effect of forced expression of HOXA1 in mammary
carcinoma cells, we utilized a novel Bcl-2 inhibitor which antagonizes Bcl-2 function by
blocking the BH3 binding pocket in Bcl-2 (Enyedy et al., 2001). Bcl-2 expression was
not altered by cellular treatment with the Bcl-2 inhibitor (Figure 4.1.3 E). Use of the Bcl-
2 inhibitor also abrogated the protection from apoptotic cell death as a consequence of forced expression of HOXA1 (Figure 4.1.3 F) similar to the effect observed with Bcl-2 antisense oligonucleotides. The anti-apoptotic effects of forced expression of HOXA1 in mammary carcinoma cells are therefore mediated by increased Bcl-2 gene transcription.
134
A
MCF7-VECTOR MCF7-HOXA1
600bp
500bp Bcl-2
400bp
600bp β-Actin 500bp 400bp hGH - + - - + - FBS - - + - - +
Figure 4.1.3 (A) HOXA1 expression in human mammary carcinoma cells increases Bcl-2 mRNA. Bcl-2 mRNA levels in MCF7-VECTOR and MCF7-HOXA1 cells were determined by RT-PCR under the conditions indicated (serum-free, 50 nM hGH, and 10% FBS) and β-actin was used as loading control. The experiment was performed as described under “Experimental Procedures.”
135
B
8 MCF7-VECTOR *
MCF7-HOXA1 6
4
* *
ARBITRARY UNIT 2
0 hGH - + - FBS - - +
Figure 4.1.3 (B) HOXA1 expression in human mammary carcinoma cells increases Bcl-2 promoter activity. Bcl-2 promoter activities in MCF7-VECTOR and MCF7-HOXA1 cells were determined by reporter assay using the Bcl-2 promoter under the conditions indicated (serum-free, 50 nM hGH, and 10% FBS). The experiment was performed as described under “Experimental Procedures.” *, p < 0.01.
136
C
VEHICLE SENSE ANTISENSE 30kDa
Bcl-2
45kDa β-Actin
MCF7-VECTOR + - + - + - MCF7-HOXA1 - + - + - +
Figure 4.1.3 (C) Bcl-2 antisense oligonucleotides prevent the HOXA1-stimulated expression of Bcl-2. Bcl-2 protein levels in MCF7-VECTOR and MCF7-HOXA1 cells were determined by Western blot analysis under the conditions indicated (serum-free, serum free with Bcl-2 sense and antisense oligonucleotide (800nM)). β- Actin was used as loading control. The experiment was performed as described under “Experimental Procedures.”
137
D
100 VEHICLE
SENSE
75 ANTISENSE
50
25 PERCENTAGE APOPTOTIC NUCLEI
0 MCF7-VECTOR + - + - + - MCF7-HOXA1 - + - + - +
Figure 4.1.3 (D) HOXA1 expression in human mammary carcinoma cells prevents apoptotic cell death in a Bcl-2-dependent manner. Apoptotic cell death was determined in MCF7-VECTOR and MCF7-HOXA1 cells under the conditions indicated (serum-free, serum free with Bcl-2 sense and antisense oligonucleotide (800nM)). Bcl-2 antisense oligonucleotide abrogated protection from apoptotic cell death as a consequence of forced expression of HOXA1 as indicated. The experiment was performed as described under “Experimental Procedures.”
138
E
VEHICLE Bcl-2 INHIBITOR
30kDa
Bcl-2
45kDa β-Actin
MCF7-VECTOR + - + -
MCF7-HOXA1 - + - +
Figure 4.1.3 (E) Bcl-2 inhibitor does not alter Bcl-2 expression. Bcl-2 protein levels in MCF7-VECTOR and MCF7-HOXA1 cells were determined by Western blot analysis under the conditions indicated (serum-free, serum free with Bcl-2). β- actin was used as loading control. The experiment was performed as described under “Experimental Procedures.”
139
F
VEHICLE 100 Bcl-2 INHIBITOR
75
50
25 PERCENTAGE APOPTOTIC NUCLEI
0 MCF7-VECTOR + - + - MCF7-HOXA1 - + - +
Figure 4.1.3 (F) HOXA1 expression in human mammary carcinoma cells prevents apoptotic cell death in a Bcl-2-dependent manner. Apoptotic cell death was determined in MCF7-VECTOR and MCF7-HOXA1 cells under the conditions indicated (serum-free, serum free with Bcl-2 inhibitor). Bcl-2 inhibitor abrogated protection from apoptotic cell death as a consequence of forced expression of HOXA1 as indicated. The experiment was performed as described under “Experimental Procedures.”
140
4.1.4 HOXA1 expression in mammary carcinoma cells protects against
doxorubicin-induced apoptosis
Since forced expression of HOXA1 in mammary carcinoma cells offered dramatic
protection from apoptosis, we reasoned that increased HOXA1 expression would also
result in decreased sensitivity to cell death as a result of exposure to anti-neoplastic
agents. We therefore examined the effect of forced expression of HOXA1 on the
apoptotic response of mammary carcinoma cells to doxorubicin. Doxorubicin decreased
Bcl-2 promoter activity in MCF7-VECTOR cells and also decreased the HOXA1-
enhanced promoter activity in MCF7-HOXA1 cells, although Bcl-2 promoter activity in
MCF7-HOXA1 cells treated with doxorubicin remained higher than in vehicle-treated
MCF7-VECTOR cells (Figure 4.1.4 A). The level of Bcl-2 mRNA as determined by RT-
PCR demonstrated concordance with Bcl-2 promoter activity (Figure 4.1.4 B).
Furthermore, Western blot analysis for Bcl-2 demonstrated an equivalent decrease in Bcl-
2 protein upon treatment of MCF7-VECTOR cells and relative maintenance of Bcl-2
protein in doxorubicin-treated MCF7-HOXA1 cells. Bcl-xL protein was also slightly
decreased after doxorubicin treatment but to equivalent levels in both cell lines. The level
of Bak and Bax did not differ between the two cell lines, and no effect of doxorubicin
was observed. A dramatic increase in p53 in both MCF7-VECTOR and MCF7-HOXA1 demonstrated the efficacy of doxorubicin treatment (Figure 4.1.4 C). Doxorubicin dramatically increased apoptosis in MCF7-VECTOR cells, and in comparison MCF7-
HOXA1 exhibited marked resistance to the apoptotic effects of doxorubicin (Figure
141
4.1.4 D). Therefore, increased expression of HOXA1 in human mammary carcinoma cells produces a chemoresistant phenotype.
142
A
VEHICLE DOXORUBICIN
10 * * 8
6
4
ARBITRARY UNIT 2
0 MCF7-VECTOR MCF7-HOXA1
Figure 4.1.4 (A) Forced expression of HOXA1 in mammary carcinoma cells (MCF7-HOXA1) maintains Bcl-2 promoter activity above that observed in vector- transfected control cells upon treatment of both cell lines with doxorubicin. Bcl-2 promoter activities in MCF7-VECTOR and MCF7-HOXA1 cells were determined by reporter assay using the Bcl-2 promoter under the conditions indicated (1 µg/ml doxorubicin). The experiment was performed as described under “Experimental Procedures.” *, p< 0.01.
143
B
MCF7-VECTOR MCF7-HOXA1 600bp 500bp Bcl-2 400bp
600bp β-Actin 500bp 400bp DOXORUBICIN - + - +
Figure 4.1.4 (B) Forced expression of HOXA1 in mammary carcinoma cells (MCF7-HOXA1) maintains Bcl-2 mRNA levels above that observed in vector- transfected control cells upon treatment of both cell lines with doxorubicin. Bcl-2 mRNA levels in MCF7-VECTOR and MCF7-HOXA1 cells were determined by RT- PCR under the conditions indicated (1 µg/ml doxorubicin) and β-actin was used as loading control. The experiment was preformed as described under “Experimental Procedures.”
144
C
MCF7-VECTOR MCF7-HOXA1
p53 45 kDa 30 kDa Bak
Bax 20 kDa
Bcl-xL 30 kDa 30 kDa Bcl-2
45 kDa β-Actin
DOXORUBICIN - + - +
Figure 4.1.4 (C) Western blot analysis demonstrates dramatic doxorubicin- induced expression of p53 in both MCF7-VECTOR and MCF7-HOXA1 cells and maintenance of Bcl-2 expression in MCF7-HOXA1 cells compared with MCF7- VECTOR cells. p53, Bak, Bax, Bcl-xL, and Bcl-2 protein levels in MCF7-VECTOR and MCF7-HOXA1 cells were determined by Western blot under the conditions indicated (1 µg/ml doxorubicin) and β-actin was used as loading control. The experiment was preformed as described under “Experimental Procedures.”
145
D
VEHICLE
80 DOXORUBICIN *
60
40
20 PERCENTAGE APOPTOTIC APOPTOTIC PERCENTAGE NUCLEI
0 MCF7-VECTOR MCF7-HOXA1
Figure 4.1.4 (D) Forced expression of HOXA1 in mammary carcinoma cells protects against doxorubicin-induced apoptosis. Doxorubicin-induced apoptotic cell death was abrogated by forced expression of HOXA1 in mammary carcinoma cells as indicated. The experiment was preformed as described under “Experimental Procedures.” *, p < 0.01.
146
4.1.5 Forced expression of HOXA1 in human mammary carcinoma cells
results in increased anchorage-independent growth in a Bcl-2-
dependent manner
Anchorage-independent growth is one pivotal characteristic of malignant transformation
(Hanahan and Weinberg, 2000). To determine if forced expression of HOXA1 would
alter the oncogenicity of mammary carcinoma cells, we examined the anchorage-
independent growth of MCF7-VECTOR and MCF7-HOXA1 in both soft agar and
suspension culture. MCF7-VECTOR cells formed colonies in soft agar as expected.
However, forced expression of HOXA1 in MCF7-HOXA1 cells dramatically enhanced colony formation in soft agar (Figure 4.1.5A). The size of the individual colonies formed by MCF7-HOXA1 was also significantly larger than the small cell aggregates observed
with MCF7-VECTOR cells (Figure 4.1.5D). Similarly, when cultured in suspension
condition, MCF7-HOXA1 cells also increased in number significantly greater than
MCF7-VECTOR cells and after 10 days were approximately tripled in number compared
with MCF7-VECTOR cells (Figure 4.1.5B).
To determine if Bcl-2 expression was also required for the observed enhancement of
anchorage-independent growth as a result of forced expression of HOXA1, we examined
soft agar colony formation of MCF7-VECTOR and MCF7-HOXA1 cells in the presence
of the Bcl-2 inhibitor. As is observed in Figure 4.1.5C, use of the Bcl-2 inhibitor
completely prevented HOXA1 enhancement of soft agar colony formation by mammary
147 carcinoma cells (Figure 4.1.5D). Thus, forced expression of HOXA1 enhances oncogenicity of human mammary carcinoma cells in a Bcl-2-dependent manner.
148
A
1500 *
R 1000
500 COLONY NUMBE
0 MCF7-VECTOR MC7-HOXA1
Figure 4.1.5 (A) Forced expression of HOXA1 enhances anchorage-independent growth of mammary carcinoma. Soft agar colony formation by MCF7-VECTOR and MCF7-HOXA1 cells was performed as described under “Experimental Procedures.” The results are given as means ±S.D. of triplicate determinations. *, p < 0.001.
149
B
400 MCF7-VECTOR MCF7-HOXA1 ) 3 300
200
100 TOTAL CELL NUMBER (x 10
0 0 2 4 6 8 10 TIME (DAYS)
Figure 4.1.5 (B) Forced expression of HOXA1 enhances anchorage-independent growth of mammary carcinoma cells. Anchorage-independent growth of MCF7- VECTOR and MCF7-HOXA1 in bacteriological dishes was performed as described under “Experimental Procedures.” The results are given as means ± S.D. of triplicate determinations.
150
C
* 1500
MCF7-VECTOR
MCF7-HOXA1
R 1000
500 COLONY NUMBE
0 Bcl-2 INHIBITOR - - + +
Figure 4.1.5 (C) Forced expression of HOXA1 enhances anchorage-independent growth of mammary carcinoma cells in a Bcl-2-dependent manner. Soft agar colony formation by MCF7-VECTOR and MCF7-HOXA1 cells was performed as described under “Experimental Procedures.” Bcl-2 inhibitor was used at a concentration of 1 µM to demonstrate Bcl-2 dependence of HOXA1-stimulated enhancement of soft agar colony formation. The results are given as means ± S.D. of triplicate determinations. *, p <0.01.
151
D
MCF7-VECTOR MCF7-HOXA1 D
VEHICLE
Bcl-2 INHIBITOR
Figure 4.1.5 (D) Forced expression of HOXA1 enhances anchorage-independent growth of mammary carcinoma cells in a Bcl-2-dependent manner. Soft agar colony formation by MCF7-VECTOR and MCF7-HOXA1 cells was performed as described under “Experimental Procedures.” Bcl-2 inhibitor was used at a concentration of 1 µM to demonstrate Bcl-2 dependence of HOXA1-stimulated enhancement of soft agar colony formation.
152
4.1.6 Forced expression of HOXA1 results in oncogenic transformation
of immortalized human mammary epithelial cells in vitro
To further determine if forced expression of HOXA1 would result in oncogenic transformation of human mammary epithelial cells, we utilized the immortalized but not transformed human mammary epithelial cell line MCF-10A (Soule et al., 1990). When grown attached to a plastic substrate these cells display a typical epithelial morphology, do not form colonies in soft agar. They also are not capable of growth in immunocompromised mice (Soule et al., 1990).
To examine the potential oncogenic capacity of HOXA1, we therefore stably transfected
HOXA1 cDNA in MCF-10A cells (MCF10A-HOXA1). These cells expressed higher levels of both HOXA1 mRNA (Figure 4.1.6A) and protein (Figure 4.1.6B) compared with vector-transfected controls (MCF10A-VECTOR). Forced expression of HOXA1 in
MCF10A-HOXA1 cells also resulted in enhanced HOXA1 transcriptional activity as evidenced by reporter activity from the EphA2-r42B enhancer (Figure 4.1.6C). Transient transfection of the cDNA for HOXA1 binding partner PBX1 further dramatically enhanced HOXA1 transcriptional activity in MCF10A-HOXA1 cells as expected (Figure
4.1.6C).
We therefore proceeded to examine the growth characteristics of both MCF10A-
VECTOR and MCF10A-HOXA1 cells. MCF10A-HOXA1 total cell number was increased significantly greater under all examined conditions (serum free, hGH and FBS)
153
than MCF10A-VECTOR cell number (Figure 4.1.6D). Comparison of nuclear BrdU incorporation between the two cell lines under serum-free conditions, 50 nM exogenous hGH, or FBS demonstrated that forced expression of HOXA1 significantly increased cell cycle progression of mammary carcinoma cells (Figure 4.1.6E). Apoptotic cell death was
also dramatically reduced in MCF10A-HOXA1 cells compared with MCF10A-VECTOR
cells in serum-free conditions (Figure 4.1.6F) as previously described for forced
expression of HOXA1 in MCF-7 cells. Exogenous hGH slightly reduced apoptotic cell
death in both MCF10A-VECTOR and MCF10A-HOXA1 cells but proportionate to the
serum-free condition for each cell. FBS functioned as a powerful survival stimulus for
both cell lines, although apoptosis was still less in MCF10A-HOXA1 cells compared
with MCF10A-VECTOR cells (Figure 4.1.6F).
MCF10A-HOXA1 cells formed large numerous colonies in soft agar, whereas MCF10A-
VECTOR cells were largely ineffective in colonization of soft agar (Figure 4.1.6G and
4.1.6H). Thus forced expression of HOXA1 conferred tumorigenic potential upon human
mammary epithelial cells. This result was also verified with a second independently
generated pool of HOXA1 cDNA-transfected cell clones (Figure 4.1.6I). MCF10A-
HOXA1 colonization of soft agar was also Bcl-2-dependent since it was entirely
prevented by the Bcl-2 inhibitor described above (Figure 4.1.6G).
154
A MCF10A-VECTOR MCF10A-HOXA1 500bp
400b HOXA1 p 300bp
600bp β-Actin 500bp
400bp
Figure 4.1.6 (A) Demonstration of overexpression of HOXA1 mRNA upon stable transfection of MCF-10A cells with HOXA1 cDNA. MCF-10A cells were stably transfected with either the empty vector (MCF10A-VECTOR) or the vector containing HOXA1 cDNA (MCF10A-HOXA1). The level of HOXA1 mRNA was determined by RT-PCR under serum-free conditions. β-Actin was used as loading control.
155
B
MCF10A-VECTOR MCF10A-HOXA1
HOXA1
30 kDa
45 kDa β-Actin
Figure 4.1.6 (B) Demonstration of overexpression of HOXA1 protein upon stable transfection of MCF-10A cells with HOXA1 cDNA. MCF-10A cells were stably transfected with either the empty vector (MCF10A-VECTOR) or the vector containing HOXA1 cDNA (MCF10A-HOXA1). The level of HOXA1 protein was determined by Western blot analysis under serum-free conditions. β-Actin was used as loading control.
156
C
VEHICLE 30 hGH * FBS * * * * 20 *
10 ARBITRARY UNIT
0 PBX1 - + - + - + - + - + - + MCF10A-VECTOR MCF10A-HOXA1
Figure 4.1.6 (C) Demonstration of functional overexpression of HOXA1 protein upon stable transfection of MCF-10A cells with HOXA1 cDNA. MCF-10A cells were stably transfected with either the empty vector (MCF10A-VECTOR) or the vector containing HOXA1 cDNA (MCF10A-HOXA1). HOXA1 transcriptional activity (in the presence and absence of its binding partner PBX1) was determined by reporter assay. *, p < 0.01.
157
D
VEHICLE * 150 hGH FBS ) 3 * (x10
R * 100
50 TOTAL CELL NUMBE TOTAL CELL
0 MCF10A-VECTOR MCF10A-HOXA1
Figure 4.1.6 (D) Effect of forced expression of HOXA1 in immortalized mammary epithelial cells (MCF-10A) on cell number. MCF10A-VECTOR and MCF10A-HOXA1 cells were cultured in serum-free media or in serum-free media supplemented with either 50 nM hGH or 10% FBS as indicated. Total cell number was determined in both cell lines under the indicated conditions as detailed under “Experimental Procedures.” Results represent means ± S.D. of triplicate determinations. *, p < 0.01.
158
E VEHICLE * hGH 15 FBS *
*
10
5 PROPORTION CELLS WITH NUCLEAR NUCLEAR WITH CELLS PROPORTION INCORPORATION Brdu 0 MCF10A-VECTOR MCF10A-HOXA1
Figure 4.1.6 (E) Effect of forced expression of HOXA1 in immortalized mammary epithelial cells (MCF-10A) on cell cycle progression. MCF10A-VECTOR and MCF10A-HOXA1 cells were cultured in serum-free media or in serum-free media supplemented with either 50 nM hGH or 10% FBS as indicated. Cell cycle progression (BrdU incorporation) was determined in both cell lines under the indicated conditions as detailed under “Experimental Procedures.” Results represent means ± S.D. of triplicate determinations. *, p < 0.01.
159
F
VEHICLE * 60 hGH
* FBS
40 *
20 PERCENTAGE APOPTOTIC NUCLEI
0 MCF10A-VECTOR MCF10A-HOXA1
Figure 4.1.6 (F) Effect of forced expression of HOXA1 in immortalized mammary epithelial cells (MCF-10A) on apoptosis. MCF10A-VECTOR and MCF10A-HOXA1 cells were cultured in serum-free media or in serum-free media supplemented with either 50 nM hGH or 10% FBS as indicated. Apoptotic cell death was determined in both cell lines under the indicated conditions as detailed under “Experimental Procedures.” Results represent means ± S.D. of triplicate determinations. *, p < 0.01.
160
G
MCF10A-VECTOR MCF10A-HOXA1
VEHICLE
Bcl-2 INHIBITOR
Figure 4.1.6 (G) Forced expression of HOXA1 in immortalized mammary epithelial cells results in oncogenic transformation in vitro. Forced expression of HOXA1 in immortalized mammary epithelial cells permits soft agar colony formation. Assays were performed as described under “Experimental Procedures.” Bcl-2 inhibitor (0.5µg/ml) was utilized to demonstrate Bcl-2 dependence of oncogenic transformation.
161
H
300
*
R 200
100 CONLONY NUMBE
0
MCF10A-VECTOR MCF10A-HOXA1
Figure 4.1.6 (H) Forced expression of HOXA1 in immortalized mammary epithelial cells results in oncogenic transformation in vitro. Forced expression of HOXA1 in immortalized mammary epithelial cells permits soft agar colony formation. Assays were performed as described under “Experimental Procedures”. *, p < 0.01.
162
I
300
R * 200
100 CONLONY NUMBE
0 MCF10A-VECTOR MCF10A-HOXA1
Figure 4.1.6 (I) Forced expression of HOXA1 in immortalized mammary epithelial cells (a second independently generated pool of HOXA1 cDNA-transfected cell clones) results in oncogenic transformation in vitro. Forced expression of HOXA1 in immortalized mammary epithelial cells permits soft agar colony formation Assays were performed as described under “Experimental Procedures”. *, p < 0.01.
163
4.1.7 Forced expression of HOXA1 results in oncogenic transformation
of immortalized human mammary epithelial cells in vivo
In vitro analyses of oncogenic transformation are not always concordant with tumorigenic potential in vivo. Results from in vitro assays may oversimplify many factors. For example, in vitro models lack the interaction between cancer cells, nonmalignant cells of the tumor supporting stroma, cells of the endocrine system, and the extracellular matrix. The ability to grow human tumors in immune-deficient mice offers research opportunities to investigate tumorigenicity in vivo (Sharkey and Fogh, 1984).
We therefore implanted both MCF10A-VECTOR and MCF10A-HOXA1 cells into the
first mammary (axillary) fat pad of female nude mice with use of either PBS or matrigel
as vehicle. Both cells failed to grow and develop tumors in the nude mice (Table I, A).
Interestingly, the positive controls, nude mice implanted with human mammary
carcinoma cells, MCF-7, also failed to grow and develop tumors (Table I, A).
In the next series of experiments, severe combined immunodeficient (SCID) mice were
used instead of nude mice and the SCID mice simultaneously received a 60-day release
pellet containing 0.72 mg of E2 (Sigma) as previous investigators have shown that without
estrogen involvement, breast tumors fail to grow in immunodeficient mice (Brunner et
al., 1987; Lippman et al., 1987; Phillips et al., 1989). Although nude mice have been used
for many years to grow various human tumors, SCID mice are better recipients for
164 growth of human tumors because of their additional B-cell deficiency compared with the single T-cell deficiency of nude mice (Bosma et al., 1983; Topley et al., 1993).
We therefore implanted both MCF10A-VECTOR and MCF10A-HOXA1 cells into the first mammary (axillary) fat pad of female SCID mice with use of either PBS or matrigel as vehicle. The mice simultaneously received a 60-day release pellet containing 0.72 mg of E2 (Sigma). MCF10A-HOXA1 cells formed large palpable and visible tumors in SCID mice injected with the cells, whereas MCF10A-VECTOR cells did not (Figure 4.1.7A,
4.1.7B, and Table I, B). The tumor volume is got as follows: tumor volume = (d1 x d2 x d3) x 0.5236, in which dn represents the three orthogonal diameter measurements, measured with callipers (Figure 4.1.7C and 4.1.7D). The average volume of the tumors formed by MCF10A-HOXA1 cells was related to the injection vehicle with matrigel resulting in larger tumor volume (Table I, B). This is consistent with the previous studies which showed that matrigel, when co-injected with human tumor cells, can improve the growth of subcutaneous xenografts in immunodeficient mice (Fridman et al., 1990;
Pretlow et al., 1991; Noel et al., 1992).
Necropsy revealed that the tumors were attached to the underlying axillary muscle and surrounded by a vascular fibrous capsule. Histologically the neoplastic cells were locally invasive and associated with fibrous connective tissue (Figure 4.1.7E and 4.1.7F). The cells exhibited moderate cytoplasmic and nuclear pleomorphism and formed a solid mass often with areas of central necrosis. Necropsy of SCID mice injected with MCF10A-
VECTOR cells failed to identify any growth. Macroscopic and histological examination
165 of lung and liver of SCID mice injected with MCF10A-HOXA1 cells failed to identify metastatic extension of the implanted MCF10A-HOXA1 cells.
166
A B
VECTOR
VECTORVECTOR HOXA1
VECTOR HOXA1
Figure 4.1.7 (A,B) Forced expression of HOXA1 in immortalized mammary epithelial cells results in oncogenic transformation and tumor formation in vivo. MCF10A-HOXA1 cells implanted into the first mammary (axillary) fat pad of female severe combined immunodeficient mice formed a large visible tumor mass.
167
C D
Figure 4.1.7 (C,D) Forced expression of HOXA1 in immortalized mammary epithelial cells results in oncogenic transformation and tumor formation in vivo. MCF10A-HOXA1 cells implanted into the first mammary (axillary) fat pad of female severe combined immunodeficient mice formed a large visible tumor mass. Tumor volume = (d1 x d2 x d3) x 0.5236, in which dn represents the three orthogonal diameter measurements, measured with callipers.
168
E F
Figure 4.1.7 (E,F) Forced expression of HOXA1 in immortalized mammary epithelial cells results in oncogenic transformation and tumor formation in vivo. MCF10A-HOXA1 cells implanted into the first mammary (axillary) fat pad of female severe combined immunodeficient mice formed a large visible tumor mass and histological appearance of the tumor visualized with hematoxylin and eosin (E), and local invasion of MCF10A-HOXA1 tumor cells into surrounding skeletal muscle (F).
169
Table I
HOXA1 overexpression induces in vivo tumor formation
A, 5 x 106 exponentially growing MCF-7, MCF10A-HOXA1, and vector- transfected control cells were implanted in the absence of estrogen into the mammary fat pad of nude mice.
Incidence of tumorigenicity Mean tumor volume Cell PBS Matrigel PBS Matrigel (mm3)
MCF10A-VECTOR 0/10 0/10 NAα NA
MCF10A-HOXA1 0/10 0/10 NA NA
MCF-7 0/10 0/10 NA NA
B, 5 x 106 exponentially growing MCF10A-HOXA1, and vector-transfected control cells were implanted in the presence of estrogen into the mammary fat pad of SCID mice, and tumors were harvested 3 weeks after inoculation.
Incidence of tumorigenicity Mean tumor volume Cell PBS Matrigel PBS Matrigel
(mm3)
MCF10A-VECTOR 0/10 0/10 NA NA
MCF10A-HOXA1 9/10 10/10 734 ± 55 903 ± 286
αNA, not applicable
170
4.2 E-cadherin-directed signaling upregulates HOXA1 through Rac1
The confluence-dependent expression of HOXA1 was examined in this study. This study also investigated the upregulation of HOXA1 by E-cadherin-directed signaling. When cells grow to confluence, they are connected by forming cell-cell adhesion. In E-cadherin positive cells, this close cell-cell contact at confluence is mainly mediated by E-cadherin
(Croix et al., 1998). Recent developments show that E-cadherin functions not only as mediator of cell surface adhesion, but also as adhesion-activated cell surface receptor, i.e.
E-cadherin itself can direct cell signaling when it is activated as cell-cell adhesion forms.
For example, it was demonstrated that Rac1 was activated as direct downstream consequence of E-cadherin ligation (Kovacs et al., 2002). However, the downstream targets of E-cadherin-activated signaling are far from identified.
HOXA1 was found to have increased expression at full confluence in human mammary carcinoma cells MCF-7. Using a calcium switch method to manipulate junction assembly, we found that induction of cell-cell junctions increased HOXA1 expression, and this was inhibited by E-cadherin function-blocking antibodies. To distinguish the E- cadherin-dependent regulation from the E-cadherin-activated signaling regulation of
HOXA1 expression, a functional E-cadherin ligand (hE/Fc) was utilized and this confluence-dependent expression of HOXA1 was confirmed to be direct downstream consequence of E-cadherin ligation. Furthermore, E-cadherin-activated increment of
HOXA1 was through a Rac1 mechanism. Increased HOXA1 expression by E-cadherin- activated signaling had increased anti-apoptotic effect in human mammary carcinoma
171 cells. Taken together, these results indicated that a cross link may exist between E- cadherin and HOXA1 signaling pathways and that direct E-cadherin signaling may have anti-apoptotic effect.
172
4.2.1 HOXA1 mRNA, protein, and transcriptional activity are increased in
mammary carcinoma cells at full confluence
The level of HOXA1 mRNA in mammary carcinoma cells MCF-7 at full confluence and
at subconfluence was first examined by semi-quantitative RT-PCR. Figure 4.2.1 A
demonstrated an amplified fragment of the predicted size (359 bp) appropriate for
HOXA1 mRNA in MCF-7 cells. HOXA1 mRNA was increased in MCF-7 cells at full
confluence (100% confluence) in comparison to subconfluences (20%, 40%, 60%, and
80% confluences). The level of β-actin mRNA did not differ in MCF-7 cells under the different conditions and was used as a control for RNA quality (Figure 4.2.1 A).
To determine if the increased HOXA1 protein was also observed in MCF-7 cells at full confluence, we examined the level of HOXA1 protein in lysates collected from MCF-7 cells under the different confluence conditions by Western blot analysis. Both the 33 kDa and the 35 kDa form of HOXA1 were detectable (Figure 4.2.1 B) with increased
HOXA1 protein observed at 100% confluence.
To determine if HOXA1-mediated transcription was also increased in MCF-7 cells at full confluence, we examined luciferase reporter activity from the EphA2 r42B enhancer which contains four HOX-PBX binding sites (Chen and Ruley, 1998). MCF-7 cells at full confluence (100% confluence) had increased luciferase activity from the EphA2-r42B enhancer compared with MCF-7 cells at subconfluences (20%, 40%, 60%, 80% confluences) (Figure 4.2.1 C) indicative of increased HOXA1 transcriptional activity.
173
Thus, confluent MCF-7 cells had elevated HOXA1-mediated transcriptional activity when compared with low density cultures.
174
A
500bp 400bp HOXA1 300bp
600bp β-Actin 500bp 400bp 20% 40% 60% 80% 100%
Figure 4.2.1 (A) HOXA1 mRNA is increased in mammary carcinoma cells at full confluence. MCF-7 cells were cultured at different confluences (20%-100% confluences). The level of HOXA1 mRNA was determined by RT-PCR. β-Actin was used as loading control.
175
B
HOXA1 30 kDa
45 kDa β-Actin
20% 40% 60% 80% 100%
Figure 4.2.1 (B) HOXA1 protein is increased in mammary carcinoma cells at full confluence. MCF-7 cells were cultured at different confluences (20%-100% confluences). The level of HOXA1 protein was determined by Western blot analysis. β-Actin was used as loading control.
176
C
20% CON 5 40% CON 4 60% CON 80% CON 3 100% CON
2
1 ARBITRARY UNIT
0
Figure 4.2.1 (C) HOXA1 transcriptional activity is increased in mammary carcinoma cells at full confluence. MCF-7 cells were cultured at different confluences (20%-100% confluences). The level of HOXA1 transcriptional activity was determined by reporter assay.
177
4.2.2 The increment in HOXA1 expression in mammary carcinoma cells at full
confluence is cell-cell adhesion dependent
The marked elevation in HOXA1 expression in confluent cultures of MCF-7 cells
prompted us to examine the role of E-cadherin-mediated cell-cell junctions in the
increment in HOXA1 expression that occurs with high cell density. To achieve this, we
used EGTA which can chelate Ca2+, thereby disrupting E-cadherin-mediated cell-cell
contact (Kim et al., 2000). As shown in Figure 4.2.2 A, use of EGTA in MCF-7 cells
completely abrogated the increased expression of HOXA1 mRNA at full confluence.
Similarly, analysis of HOXA1 transcriptional activity by use of the EphA2-r42B
enhancer demonstrated the abrogation of increased HOXA1-mediated transcriptional
activity in MCF-7 cells at full confluence by EGTA (Figure 4.2.2 B).
Using the calcium switch model in MCF-7 cells (Nakagawa et al., 2001), we examined
the dynamics of HOXA1 expression by time-lapse analysis (Figure 4.2.2 C). E-cadherin-
mediated cell-cell adhesion was abrogated by chelating Ca2+ and then re-initiated by re-
introducing Ca2+ (Volberg et al., 1986). This approach induced a time-dependent
accumulation of E-cadherin and formation of adherens junctions in MCF-7 cells (Kim et al., 2000). Induction of E-cadherin homophilic adhesion by Ca2+ significantly increased
the expression of HOXA1 mRNA and protein in MCF-7 cells (Figure 4.2.2 C and D).
The increment in HOXA1 mRNA and protein expression was time-dependent. Increased
HOXA1 expression was detected at 15 min (HOXA1 mRNA) and 30 min (HOXA1
protein), peaked at 60 min (HOXA1 mRNA) and 90 min (HOXA1 protein) (Figure 4.2.2
178
C and D). Thus, the increment in HOXA1 expression in mammary carcinoma cells at full confluence is mediated by cell-cell adhesion.
179
A
60% 100% 60% 100% 500bp 400bp HOXA1 300bp
600bp 500bp β-Actin 400bp EGTA - - + +
Figure 4.2.2 (A) The increment in HOXA1 mRNA in mammary carcinoma cells at full confluence is cell-cell adhesion dependent. MCF-7 cells were cultured at subconfluence (60% confluence) and full confluence (100% confluence) in the presence and absence of EGTA. The level of HOXA1 mRNA was determined by RT-PCR. β-Actin was used as loading control.
180
B
6
4
2 ARBITRARY UNIT
0 EGTA - - + + 60% 100% 60% 100%
Figure 4.2.2 (B) The increment in HOXA1 transcriptional activity in mammary carcinoma cells at full confluence is cell-cell adhesion dependent. MCF-7 cells were cultured at subconfluence (60% confluence) and full confluence (100% confluence) in the presence and absence of EGTA. The level of HOXA1 transcriptional activity was determined by reporter assay.
181
C
0 15 30 60 (min) 500bp 400bp HOXA1 300bp
600bp β-Actin 500bp 400bp
Figure 4.2.2 (C) The increment in HOXA1 mRNA in mammary carcinoma cells at full confluence is cell-cell adhesion dependent. Confluent MCF-7 cells were incubated with 4mM EGTA. After 30 min, the medium was replaced with RPMI containing Ca2+ for the indicated time periods (in min). At the indicated times, cells were lysed and the level of HOXA1 mRNA was determined by RT-PCR. β-Actin was used as loading control.
182
D
0 30 60 90 (min)
HOXA1
30 kDa
45 kDa β-Actin
Figure 4.2.2 (D) The increment in HOXA1 protein in mammary carcinoma cells at full confluence is cell-cell adhesion dependent. Confluent MCF-7 cells were incubated with 4mM EGTA. After 30 min, the medium was replaced with RPMI containing Ca2+ for the indicated time periods (in min). At the indicated times, cells were lysed and the level of HOXA1 protein was determined by Western blot analysis. β-Actin was used as loading control.
183
4.2.3 The increment in HOXA1 expression in mammary carcinoma cells at full
confluence is E-cadherin-dependent
The results above are, however, not sufficient to conclude that E-cadherin-mediated cell-
cell adhesion induced the increased HOXA1 expression at full confluence. Because
MCF-7 cells adhere to each other through not only E-cadherin but also other adhesion
molecules, it is possible that cell adhesion molecules besides E-cadherin increased
HOXA1. Moreover, influx of Ca2+ into cells might induce the increment of HOXA1. To exclude the possibility that increased HOXA1 in mammary carcinoma cells at full
confluence was due to factors other than E-cadherin, we tested the effect of DECMA-1, a
function-blocking antibody against E-cadherin (Vestweber and Kemler, 1985). As is
observed in Figure 4.2.3 A and B, use of DECMA-1 in MCF-7 cells completely
abrogated the increased expression of HOXA1 mRNA and protein at full confluence. In
addition, use of DECMA-1 in MCF-7 cells also abrogated the increased HOXA1-
mediated transcriptional activity at full confluence (Figure 4.2.3 C).
Calcium switch was performed with DECMA-1. MCF-7 cells were incubated with
EGTA, and then Ca2+ was restored to the medium in the presence of DECMA-1. The
increase in HOXA1 expression induced by calcium restoration was blocked by DECMA-
1 (Figure 4.2.2 C and D and Figure 4.2.3 D and E), suggesting that E-cadherin is
critical for increased HOXA1 expression at full confluence.
184
Altogether, the results demonstrated that the increment in HOXA1 expression in mammary carcinoma cells at full confluence is E-cadherin-dependent.
185
A
60% 100% 60% 100% 500bp 400bp HOXA1 300bp
600bp 500bp β-Actin 400bp DECMA-1 - - + +
Figure 4.2.3 (A) The increment in HOXA1 mRNA in mammary carcinoma cells at full confluence is E-cadherin-dependent. MCF-7 cells were cultured at subconfluence (60% confluence) and full confluence (100% confluence) in the presence and absence of DECMA-1, a function-blocking antibody against E- cadherin. The level of HOXA1 mRNA was determined by RT-PCR. β-Actin was used as loading control.
186
B 60% 100% 60% 100%
HOXA1 30 kDa
45 kDa β-Actin
DECMA-1 --+ +
Figure 4.2.3 (B) The increment in HOXA1 protein in mammary carcinoma cells at full confluence is E-cadherin-dependent. MCF-7 cells were cultured at subconfluence (60% confluence) and full confluence (100% confluence) in the presence and absence of DECMA-1, a function-blocking antibody against E- cadherin. The level of HOXA1 protein was determined by Western blot analysis. β- Actin was used as loading control.
187
C 5
4
3
2
ARBITRARY UNIT 1
0 DECMA-1 - - + + 60% 100% 60% 100%
Figure 4.2.3 (C) The increment in HOXA1 transcriptional activity in mammary carcinoma cells at full confluence is E-cadherin-dependent. MCF-7 cells were cultured at subconfluence (60% confluence) and full confluence (100% confluence) in the presence and absence of DECMA-1, a function-blocking antibody against E- cadherin. The level of HOXA1 transcriptional activity was determined by reporter assay.
188
D
0 15 30 60 (min)
500bp 400bp β-Actin 300bp
600bp 500bp HOXA1 400bp
DECMA-1 + + + +
Figure 4.2.3 (D) The increment in HOXA1 mRNA in mammary carcinoma cells at full confluence is E-cadherin-dependent. Confluent MCF-7 cells were incubated with 4mM EGTA. After 30 min, the medium was replaced with RPMI containing Ca2+ in the presence of DECMA-1, a function-blocking antibody against E- cadherin, for the indicated time periods (in min). At the indicated times, cells were lysed and the level of HOXA1 mRNA was determined by RT-PCR. β-Actin was used as loading control.
189
E
0 30 60 90 (min)
HOXA1 30 kDa
45 kDa β-Actin
DECMA-1 + + + +
Figure 4.2.3 (E) The increment in HOXA1 protein in mammary carcinoma cells at full confluence is E-cadherin-dependent. Confluent MCF-7 cells were incubated with 4mM EGTA. After 30 min, the medium was replaced with RPMI containing Ca2+ in the presence of DECMA-1, a function-blocking antibody against E-cadherin, for the indicated time periods (in min). At the indicated times, cells were lysed and the level of HOXA1 protein was determined by Western blot analysis. β-Actin was used as loading control.
190
4.2.4 The increment in HOXA1 expression in mammary carcinoma cells at full
confluence is regulated by E-cadherin-activation and not just E-cadherin-
dependent signaling
The results detailed above do not distinguish whether the increment in HOXA1 is direct downstream consequence of E-cadheirn ligation or just a result of juxtacrine signals that required E-cadherin adhesion to bring cell surfaces together. In other words, it should be determined whether HOXA1 increment results from direct E-cadherin-activated signals or E-cadherin-dependent juxtacrine signals. To achieve this, we utilized a recombinant protein (hE/Fc) consisting of the complete ectodomain of human E-cadherin fused to the
Fc region of IgG. Substrata adsorbed with hE/Fc supported E-cadherin-specific adhesion and contact formation (Kovacs et al., 2002). Thus, using this model, E-cadherin engagement is induced without cell-cell contact.
HOXA1 mRNA was increased in MCF-7 cells adhered to hE/Fc-coated substrata in comparison to MCF-7 cells adhered to poly-L-lysine (PLL) which is used as control to hE/Fc. The level of β-actin mRNA did not differ under the different conditions and was used as a control for RNA quality (Figure 4.2.4 A). We further examined the level of
HOXA1 protein in MCF-7 cells adhered to hE/Fc-coated substrata and poly-L-lysine
(PLL). As is observed in Figure 4.2.4 B, both the 33- and the 35-kDa forms of HOXA1 were detected with increased HOXA1 protein observed in MCF-7 cells adhered to hE/Fc- coated substrata. All these experiments were performed in MCF-7 cells at subconfluence
(60% confluence) without cell-cell contact. Thus, HOXA1 increment in mammary
191 carcinoma cells results from direct E-cadherin-activated signals and not E-cadherin- dependent justacrine signals, which require E-cadherin adhesion to bring cell together.
192
A
500bp 400bp HOXA1 300bp
600bp 500bp β-Actin 400bp
PLL hE/Fc
Figure 4.2.4 (A) The increment in HOXA1 mRNA in mammary carcinoma cells at full confluence is regulated by E-cadherin-activated signaling. Subconfluent MCF-7 cells were plated onto poly-L-lysine-coated substrate (PLL) or hE/Fc-coated substrata. The level of HOXA1 mRNA was determined by RT-PCR. β-Actin was used as loading control.
193
B
HOXA1 30 kDa
45 kDa β-Actin
PLL hE/Fc
Figure 4.2.4 (B) The increment in HOXA1 protein in mammary carcinoma cells at full confluence is regulated by E-cadherin-activated signaling. Subconfluent MCF-7 cells were plated onto poly-L-lysine-coated substrate (PLL) or hE/Fc-coated substrata. The level of HOXA1 protein was determined by Western blot analysis. β- Actin was used as loading control.
194
4.2.5 Rac1, but not PI-3 kinase, α-, β-, and γ-catenins is involved in the E-
cadherin-activated increment in HOXA1 expression in mammary
carcinoma cells
α-, β-, and γ-catenins are the key molecules involved in E-cadherin signaling (Provost and Rimm, 1999; Jiang and Robert, 2000). To determine whether these molecules are involved in the E-cadherin-activated increment in HOXA1 expression, we examined their protein levels in MCF-7 cells at subconfluence and confluence. The levels of α-, β-, and
γ-catenins were not altered in the cells at suconfluence (60% confluence) and full confluence (100% confluence) (Figure 4.2.5 A). Equal loading of the cell extracts was verified by reprobing the stripped membrane for β-actin (Figure 4.2.5 A).
Apart from the catenins, PI-3 kinase as well as Rac1 has been demonstrated to be involved in E-cadherin signaling (Pece et al., 1999; Kovacs et al., 2002; Nakagawa et al.,
2001). To determine if PI-3 kinase or Rac1 was responsible for the E-cadherin-increased
HOXA1 expression, we prevented function of PI-3 kinase and Rac1 in MCF-7 cells by use of wortmannin (Vanhaesebroeck et al., 2001) and a dominantly inhibitory N17 Rac1 mutant (Kovacs et al., 2002) respectively. Transient expression of the dominant negative
N17 Rac1 (DNRac1) mutant significantly inhibited the E-cadherin-increased HOXA1 mRNA expression and HOXA1-mediated transcriptional activity at full confluence
(Figure 4.2.5 B and C), whereas wortmannin (60 nM) did not significantly affect the level of HOXA1 expression (Figure 4.2.5 B and C). To further verify the involvement of
Rac1 in direct E-cadherin-activated HOXA1 increment, we examined HOXA1 mRNA
195
and protein levels in MCF-7 cells adhered to hE/Fc-coated substrata and PLL in the presence of DNRac1. As observed in Figure 4.2.5 D and E, use of DNRac1 completely
prevented the increased HOXA1 expression observed in MCF-7 cells adhered to hE/Fc-
coated substrata. Thus, Rac1, but not PI-3 kinase, α-, β-, and γ-catenins is involved in the
E-cadherin-activated increment in HOXA1 expression in mammary carcinoma cells.
196
A
α-catenin 90 kDa
β-catenin
90 kDa
γ-catenin 80 kDa
45 kDa β-Actin
60% 100%
Figure 4.2.5 (A) α-, β-, and γ-catenin are not involved in the E-cadherin- activated increment in HOXA1 expression in mammary carcinoma cells. MCF-7 cells were cultured at subconfluence (60% confluence) and full confluence (100% confluence). The protein levels of α-, β-, and γ-catenin were determined by Western blot analysis. β-Actin was used as loading control.
197
B
60% 100% 60% 100% 60% 100% 500bp 400bp HOXA1 300bp
600bp 500bp β-Actin 400bp Wortmannin - - + + - - DNRac1 - - - - + +
Figure 4.2.5 (B) Rac1, but not PI-3 kinase is involved in the E-cadherin- activated increment in HOXA1 mRNA expression in mammary carcinoma cells. MCF-7 cells were cultured at subconfluence (60% confluence) and full confluence (100% confluence) in the presence and absence of wortmannin (PI-3 kinase inhibitor) or DNRac1 (dominantly inhibitory N17 Rac1 mutant). The level of HOXA1 mRNA was determined by RT-PCR. β-Actin was used as loading control.
198
C
5
4
3
2
ARBITRARY UNIT 1
0 DNRac1 - - - - + + Wortmannin - - + + - - 60% 100% 60% 100% 60% 100%
Figure 4.2.5 (C) Rac1, but not PI-3 kinase is involved in the E-cadherin- activated increment in HOXA1 transcriptional activity in mammary carcinoma cells. MCF-7 cells were cultured at subconfluence (60% confluence) and full confluence (100% confluence) in the presence and absence of wortmannin (PI-3 kinase inhibitor) or DNRac1 (dominantly inhibitory N17 Rac1 mutant). The level of HOXA1 transcriptional activity was determined by reporter assay.
199
D
500bp 400bp HOXA1 300bp
600bp 500bp β-Actin 400bp PLL + - - hE/Fc - + + DNRac1 - - +
Figure 4.2.5 (D) Rac1 is involved in the E-cadherin-activated increment in HOXA1 mRNA expression in mammary carcinoma cells. Subconfluent MCF-7 cells were plated onto poly-L-lysine-coated substrate (PLL) or hE/Fc-coated substrata in the presence and absence of DNRac1 (dominantly inhibitory N17 Rac1 mutant). The level of HOXA1 mRNA was determined by RT-PCR. β-Actin was used as loading control.
200
E
HOXA1 30 kDa
45 kDa β-Actin
PLL + - - hE/Fc - + + DNRac1 - - +
Figure 4.2.5 (E) Rac1 is involved in the E-cadherin-activated increment in HOXA1 protein expression in mammary carcinoma cells. Subconfluent MCF-7 cells were plated onto poly-L-lysine-coated substrate (PLL) or hE/Fc-coated substrata in the presence and absence of DNRac1 (dominantly inhibitory N17 Rac1 mutant). The level of HOXA1 protein was determined by Western blot analysis. β-Actin was used as loading control.
201
4.2.6 Increased HOXA1 expression by E-cadherin-activated signaling has
increased anti-apoptotic effect in human mammary carcinoma cells
HOXA1 has been demonstrated to offer dramatic protection from apoptosis in human
mammary carcinoma cells MCF-7 (Zhang et al., 2003). We therefore examined whether
the increased HOXA1 expression by the E-cadherin-activated signaling would offer
protection from apoptosis in MCF-7 cells. MCF-7 cells were incubated in the presence of
the HOXA1 inhibitor, HOXA1 RNAi and DECMA-1, and the level of apoptotic cell
death was determined. As expected, MCF-7 cells at full confluence had reduced
apoptosis compared with MCF-7 cells at subconfluence (Figure 4.2.6 A). Transient
transfection of HOXA1 RNAi completely prevented the reduced apoptosis at full
confluence (Figure 4.2.6 A), indicating that increased HOXA1 at full confluence is
responsible for this anti-apoptotic effect. If E-cadherin-activated increment in HOXA1
offered this anti-apoptotic effect at full confluence, we then reasoned that blocking of E-
cadherin function would abrogate the anti-apoptotic effect and that expression of HOXA1
should rescue this abrogation. Indeed, MCF-7 cells at subconfluence and full confluence
had the same apoptotic level in the presence of DECMA-1 (Figure 4.2.6 B). Transfection
of HOXA1 expression plasmid (Di Rocco et al., 1997) completely rescued the anti-
apoptosis prevention offered by DECMA-1 (Figure 4.2.6 B). Transfection of HOXA1
expression plasmid also offered anti-apoptosis in MCF-7 cells at subconflunce (Figure
4.2.6 B). Altogether, the above results demonstrated that increased HOXA1 expression
by E-cadherin-activated signaling had increased anti-apoptotic effect in human mammary
carcinoma cells.
202
To determine the effect of increased HOXA1 expression by E-cadherin-activated signaling on cell cycle progression in MCF-7 cells, we examined the nuclear incorporation of 5’-bromo-2’-deoxyuridine (BrdU) in the cells. MCF-7 cells at full confluence did not exhibit significant higher percentage of nuclear BrdU incorporation compared with MCF-7 cells at subconfluence in the presence or without the presence of
HOXA1 RNAi (Figure 4.2.6 C).
203
A
60
50
40
30
20
10
PERCE NTAGE APOPTOTIC NUCLEI 0 HOXA1 RNAi - - + + 60% 100% 60% 100%
Figure 4.2.6 (A) Increased HOXA1 expression by E-cadherin-activated signaling has increased anti-apoptotic effect in human mammary carcinoma cells. MCF-7 cells were cultured at subconfluence (60% confluence) and full confluence (100% confluence) in the presence and absence of HOXA1 RNAi and apoptotic cell death was determined as detailed under “Experimental Procedures”.
204
B
40
30
20
10 PERCE NTAGE APOPTOTIC NUCLEI 0 DECMA-1 - - + + + + HOXA1 - - - - + + 60% 100% 60% 100% 60% 100%
Figure 4.2.6 (B) Increased HOXA1 expression by E-cadherin-activated signaling has increased anti-apoptotic effect in human mammary carcinoma cells. MCF-7 cells were cultured at subconfluence (60% confluence) and full confluence (100% confluence) in the presence and absence of HOXA1 expression plasmid or DECMA- 1 (E-cadherin function-blocking antibody) and apoptotic cell death was determined as detailed under “Experimental Procedures”.
205
C
10
8
6
4
2
PROPORTION CELLS WITH NUCLEAR Brdu INCORPORATION 0 HOXA1 RNAi - - + + 60% 100% 60% 100%
Figure 4.2.6 (C) Effect of increased HOXA1 expression by E-cadherin-activated signaling on cell cycle progression in human mammary carcinoma cells. MCF-7 cells were cultured at subconfluence (60% confluence) and full confluence (100% confluence) in the presence and absence of HOXA1 RNAi and cell cycle progression (BrdU incorporation) was determined under “Experimental Procedures”.
206
Chapter 5 General discussion
The present work has investigated the various properties, and their mechanisms, of the
autocrine human growth hormone (hGH)-regulated homeobox gene product, HOXA1, in
human mammary carcinoma cells. Forced expression of HOXA1 in mammary carcinoma
cells resulted in increased total cell number. This was primarily by the promotion of cell survival mediated by the transcriptional upregulation of Bcl-2. HOXA1 also abrogated the apoptotic response of mammary carcinoma cells to the chemotherapy drug, doxorubicin. HOXA1 was also identified as a powerful oncogene for human mammary epithelial cells in vitro and in vivo. It was further demonstrated that HOXA1 expression in mammary carcinoma cells was increased by cell to cell contact at full confluence. This increase in HOXA1 at full confluence was demonstrated to be mediated actively by E- cadherin.
5.1 Human growth hormone-regulated HOXA1 is a human mammary
epithelial oncogene
Forced expression of HOXA1 was sufficient to result in the oncogenic transformation of immortalized human mammary epithelial cells with aggressive in vivo tumor formation.
Primary rodent cells may be transformed by two concomitantly introduced oncogenes
(Land et al., 1983; Ruley, 1983). Only recently have human cells been transformed by a combination of the genomic version of the SV40 Large T antigen, the hTERT gene that
encodes the telomerase catalytic subunit, and an oncogenic allele of the H-Ras gene, H-
207
RasV12 (Hahn et al., 1999). hTERT allows mammary epithelial cells to bypass both
senescence (M1) and crisis (M2) with resultant immortalization of cells. However,
immortalization of cells is not sufficient to create an oncogenically transformed cell
(Hahn et al., 1999; Jiang et al., 1999). In the present study, oncogenic transformation in
vitro and in vivo was demonstrated in immortalized human mammary epithelial cells by
introduction of only the HOXA1 cDNA. Other members of the homeobox gene family have also been reported to be oncogenic (Kroon et al., 1998; Maulbecker and Gruss,
1993). Such examples include Hoxb8, which was shown to transform NIH3T3 cells
(Aberdam et al., 1991), and Cdx1, which caused transformation and tumorigenesis of intestinal epithelial cells (Soubeyran et al., 2001). Interestingly, certain homeobox family members have also been demonstrated to result in cellular immortalization (Schnabel et
al., 2000). This raises the intriguing possibility that HOXA1 may both immortalize human
mammary epithelial cells and result in their oncogenic transformation (shown here).
Indeed, in preliminary experiments from the same laboratory (Chen et al., manuscript in
preparation), it was shown that forced expression of HOXA1 does increase transcription
of the hTERT gene with a resultant increase in telomerase activity in primary human
mammary epithelial cells. It is possible that HOXA1 may therefore both immortalize and
oncogenically transform human mammary epithelial and potentially constitute a
"complete oncogene" for human mammary epithelial cells. This exciting possibility will
require further investigation.
The early portion of this study demonstrated that HOXA1 gene expression and
transcriptional activity is regulated by hGH. Since forced expression of HOXA1 resulted
in oncogenic transformation of human mammary epithelial cells, it raises the possibility
208
that hGH itself is oncogenic. However, other studies in our laboratory have shown that exogenous hGH does not support the anchorage-independent growth of immortalized
human mammary epithelial cells (Zhu et al., unpublished data), despite the requirement of
endocrine GH for optimal growth of mammary tumor xenografts (Yang et al., 1996). This
is concordant with the fact that, although exogenous hGH increased HOXA1 mRNA, it
did not result in an increase in HOXA1 protein nor transcriptional activity (chapter 4.1.1).
Interestingly, however, and consistent with the previous studies (Raccurt et al., 2002),
autocrine-produced hGH does indeed confer immortalized mammary epithelial cells with
the capacity for anchorage-independent growth and tumor formation in vivo (Zhu et al.,
manuscript submitted). It is a likely possibility that autocrine hGH may therefore utilize
HOXA1 to execute its oncogenic program. The observation that exogenous hGH
regulates HOXA1 mRNA but not protein nor transcriptional activity is suggestive that
autocrine hGH production in human mammary epithelial cells also regulates HOXA1 at
the translational or post-translational level. Such potential regulation of HOXA1 other
than at the transcriptional level may be required for its transforming activity.
By use of a functional Bcl-2 inhibitor, we have demonstrated that HOXA1-dependent
transformation is dependent on Bcl-2 (Enyedy et al., 2001) (chapter 4.1.6). It is unlikely
that this pathway constitutes the sole mechanism by which HOXA1 confers oncogenic
transformation of human mammary epithelial cells and further pathways presumably will
be delineated by microarray and promoter mining experiments in progress. It may be that
Bcl-2 is simply required for human mammary epithelial cell survival, and functional
inhibition of Bcl-2 would consequently drive the cell to apoptosis. Indeed, high
concentrations of the Bcl-2 inhibitor prevented soft agar colony formation by both
209
MCF7-VECTOR and MCF7-HOXA1 cells (see chapter 4.1.5). However, titration of the
Bcl-2 inhibitor allowed selective inhibition of the enhanced colony formation as a consequence of forced expression of HOXA1 with no alteration in colony formation by
MCF7-VECTOR cells. Thus, it is apparent that Bcl-2 is required specifically for HOXA1
enhancement of mammary carcinoma cell colony formation in soft agar.
5.2 HOXA1 abrogated the apoptotic response of mammary carcinoma cells
to the chemotherapy drug, doxorubicin
Chemotherapeutic agents such as doxorubicin result in the induction of p53 to execute an apoptotic program (Hayakawa et al., 2000; Fuchs et al., 1998). Bcl-2 functions to prevent
mitochondrial cytochrome c release and subsequent cell death (Reed, 1998; Gross et al.,
1999), and therefore, increased Bcl-2 expression could be expected to result in
chemoresistance (Gutierrez-Puente et al., 2002). Indeed, Bcl-2 has been observed to be
overexpressed in a large percentage of human neoplasias (Gutierrez-Puente et al., 2002), and Bcl-2 expression is associated with resistance to chemotherapy in many cancers,
including mammary carcinoma (Bonetti et al., 1998; Decaudin et al., 1997). In the present
study, it was demonstrated that forced expression of HOXA1 resulted in resistance to
doxorubicin-induced cell death as a consequence of increased Bcl-2 expression (chapter
4.1.4). Inhibition of HOXA1 transactivation may therefore constitute a novel therapeutic
target to abrogate resistance to chemotherapeutic agents. However, as Bcl-2 (and
HOXA1 induction) are not the only mechanisms utilized by hGH to prevent apoptotic cell
210
death of mammary carcinoma cells, simple antagonism of hGH signal transduction may
represent a more efficacious approach to enhance the response of mammary carcinoma cells to chemotherapy. Previous studies from this laboratory have previously described
that autocrine hGH production promotes mammary carcinoma cell survival by at least
two other mechanisms (Mertani et al., 2001; Graichen et al., 2002). First, autocrine hGH
increases transcription of the CHOP gene, and increased p38 mitogen-activated protein
kinase-dependent CHOP-mediated transcription results in mammary carcinoma cell
survival (Mertani et al., 2001). Autocrine hGH production in mammary carcinoma cells
also conversely results in transcriptional repression of the p53-regulated placental
transforming growth factor-β gene with subsequent decreases in its protein product,
Smad-mediated transcription, and its cellular effects, which include cell cycle arrest and apoptosis (Graichen et al., 2002). Several avenues exist for antagonism of the
somatotropic axis. Non-dimerizing hGH receptor antagonists have been used clinically to
treat acromegaly (Sesmilo et al., 2002) and constitute one potential therapeutic approach.
Indeed, previous studies from this laboratory have utilized this antagonist in vitro to
demonstrate that the effects of autocrine hGH on mammary carcinoma cell proliferation
(both mitogenesis and apoptosis) are mediated via the hGH receptor (Kaulsay et al.,
2001). Growth hormone-releasing hormone antagonists have also been utilized to
decrease mammary GH production and, consequently, decrease the growth of mouse
mammary carcinoma xenografts (Schally et al., 2001). It is therefore conceivable that antagonism of hGH, utilized as adjuvant therapy, will increase the efficacy of
conventional chemotherapeutic regimes currently utilized to treat mammary carcinoma.
211
5.3 E-cadherin-directed signaling upregulates HOXA1 through Rac1 in human
mammary carcinoma cells
E-cadherin is a well-known cell-cell adhesion molecule that mediates cell-cell adhesion by Ca2+-dependent homophilic interactions. It has been demonstrated that when the E- cadherin positive cells grow to confluence, the close cell-cell contact is mainly mediated by E-cadherin (Croix et al., 1998). Recently, E-cadheirn-mediated cell-cell adhesion is postulated to initiate a cascade of cellular signals. Indeed, signaling by PI-3 kinase as well as Rho family GTPases has been observed to be activated as E-cadherin-mediated cell-cell adhesion forms (Kim et al., 2000; Noren et al., 2001; Nakagawa et al., 2001).
However, the downstream targets of the signaling pathways activated by E-cadherin binding are far from identified. In this study (chapter 4.2), the data presented demonstrated that HOXA1 is one of the downstream molecules that was regulated by E- cadherin-mediated cell-cell adhesion. The increased HOXA1 expression observed in confluent MCF-7 cells was lost during disruption of E-cadherin-mediated cell-cell adhesion by Ca2+ chelation. Using a calcium switch method to manipulate junction assembly, it was shown that induction of cell-cell junctions increased HOXA1 expression, and this was inhibited by E-cadherin function-blocking antibodies. In addition, using a functional recombinant E-cadherin adhesive ligand (hE/Fc), it was demonstrated that E-cadheirn homophilic ligation to hE/Fc increased HOXA1 expression.
Adhesion to Poly-L-Lysine did not increase HOXA1, indicating that signaling was a specific response to E-cadherin homophilic ligation and not a nonspecific consequence of cell spreading. Recent reports that PI-3 kinase as well as Rho family GTPases are
212
activated as cultured cells form E-cadherin-dependent contacts with one another could
not exclude the possibility that they were activated by juxtacrine signals that required E-
cadherin adhesion to bring cell surfaces together but were not themselves direct
downstream consequences of E-cadheirn ligation (Kim et al., 2000; Noren et al., 2001;
Nakagawa et al., 2001; Yap et al., 1997; Fagotto and Gumbiner, 1996). The present work
in this thesis now clearly demonstrates that HOXA1 is increased as a direct downstream
consequence of E-cadherin ligation. The present work in this thesis also showed that
Rac1 is responsible for the E-cadherin-increased HOXA1 expression. Since Xenopus C- cadheirn has also recently been shown to directly activate Rac (Noren et al., 2001), it will now be important to assess whether HOXA1 is a downstream molecule upregulated by all classical cadherins.
5.4 Increased HOXA1 expression by E-cadherin-activated signaling
increases anti-apoptotic effect in human mammary carcinoma cells
Although E-cadherin is central to the cell adhesion, it has also been implicated in physiologic roles beyond the mere mechanical interconnection of cells. More recent evidence suggests that E-cadherin may also be associated with regulatory pathways involved in various aspects of cell fate including developmental decisions, cellular differentiation, and apoptosis (Miller and Moon, 1996; Peifer, 1997). It is now believed that just as integrins function to mediate cell-extracellular matrix interactions in
anchorage-dependent survival, cadherins may also act in such a capacity, possessing a
functional role in the regulation of intercellular adhesion-dependent survival. Several studies have reported that N-cadherin-mediated intercellular adhesions inhibit apoptosis
213
of gut and ovarian epithelium (Hermiston and Gordon, 1995; Peluso et al., 1996), and
that E-cadherin-mediated intercellular adhesions suppress apoptosis and promote
anchorage-independent growth of oral squamous carcinoma cells (Kantak and Kramer,
1998). These studies suggested that homophilic binding of cadherin molecules on adjacent cells could transduce anti-apoptotic signals.
Interestingly, several studies showed that the apoptotic cell death increases if the cells are unable to form aggregation and cell-cell adhesion by E-cadehrin (Kantak and Kramer,
1998; Day et al., 1999). For example, Day et al. (1999) found that activation of protein kinase C resulted in 60% apoptosis of mammary and prostate epithelial cells. The surviving cells had undergone dramatic aggregation concurrent with increased E-cadherin
expression. When aggregation was inhibited by the addition of an E-cadherin-blocking antibody, apoptosis increased synergistically (Day et al., 1999). The above results that the formed cell aggregation and E-cadheirn-mediated cell-cell adhesion are able to protect cells against apoptosis are very similar to the results in the present study. In the present study, it was also shown that the confluent MCF-7 cells (cell aggregation) and the cell adhesion mediated by E-cadherin signaling have decreased apoptotic cell death.
However, the previous studies and ours identified different molecular mechanisms responsible for E-cadherin signaling-mediated cell survival. In the study of Day et al.
(1999), they indicated E-cadherin-mediated aggregation results in Rb activation and G1 arrest that is critical for the survival of prostate and mammary epithelial cells. In the study of Kantak and Kramer (1998), Bcl-2 may be involved in the E-cadherin-mediated anti-apoptosis effect (Kantak and Kramer, 1998). In the present study, HOXA1 was
214 demonstrated to be the executor of the anti-apoptosis effect mediated by E-cadehrin- activated signaling. However, different cell lines were utilized. Day et al. (1999) used non-transformed cells, Kantak and Kramer (1998) used oral squamous carcinoma cells, while in this thesis’s study, human mammary gland carcinoma cells were used. So it is possible that the differences between the cell lines may account for the different mechanisms. In addition, Day et al. (1999) used protein kinase C to treat the cells and
Kantak and Kramer cultured the cells in suspension. These different cell conditions may also be a reason for the difference in mechanism seen. Collectively, these studies lead to the conclusion that E-cadherin signaling-mediated cell-cell adhesion is an important and novel mechanism by which epithelial cells can regulate apoptosis.
5.5 Conclusions
1. Autocrine human growth hormone (hGH) production by human mammary carcinoma cells increased the expression and transcriptional activity of HOXA1. Growth hormone has been demonstrated to be involved in breast cancer. For example, transgenic mice that overexpress gene encoding GH agonist exhibit both mammary gland epithelial cell hyperplasia and an increased frequency of mammary tumors (Tornell et al., 1992).
Previous researchers in my supervisor’s laboratory have investigated the role of autocrine hGH in mammary carcinoma cells and provided evidence that autocrine production of hGH by mammary carcinoma cells may direct the cell behavior to impact on the final clinical prognosis (Kaulsay et al., 1999; Kaulsay et al., 2000; Kaulsay et al., 2001), and therefore, systematic analysis of the relevant mechanistic features by which it exerts its
215
cellular effects is required. One major mechanism by which GH affects cellular function
is by regulating the level of specific mRNA species (Isaksson et al., 1985). The previous
investigators in our lab utilized cDNA microarray analyses to identify genes regulated by autocrine hGH in mammary carcinoma cells (MCF-7) (Mertani et al., 2001). One gene demonstrated to be upregulated by autocrine hGH in MCF-7 cells was HOXA1. The present work demonstrated that autocrine production of hGH in MCF-7 cells increases
HOXA1 mRNA protein, and transcriptional activity and thus verified the hGH-dependent
upregulation of HOXA1 observed by cDNA array screening. Thus, autocrine hGH may
therefore utilize HOXA1 to execute its cellular effects in mammary carcinoma cells. For
example, autocrine GH affords dramatic protection from apoptosis (Kaulsay et al., 2001).
HOXA1 was also demonstrated to have marked anti-apoptosis effect in the present work
and thus HOXA1 may be directly involved in autocrine GH anti-apoptosis effect.
Interestingly, autocrine production of hGH in immortalized human mammary epithelial
cells was shown to result in oncogenic transformation and tumor formation in vivo (Zhu
et al., manuscript submitted). So it is possible that HOXA1, acting as the downstream
target of autocrine hGH, is responsible for the oncogenic program of autocrine hGH.
2. Forced expression of HOXA1 in an immortalized human mammary epithelial
cell line (MCF-10A) resulted in oncogenic transformation and the development of a
rapidly growing carcinoma in vivo. This was quite remarkable, given that forced
expression of other known oncogenes (e.g. Ras, Her2/neu, and TC21) have been reported in several laboratories to be insufficient to convey tumorigenic potential on MCF-10A
cells (Clark et al., 1996; Giunciuglio et al., 1995). Furthermore, even cyclin D1
216
(overexpressed in over 50% of human mammary carcinomas (McIntosh et al., 1995)),
although able to support anchorage-independent growth of MCF-10A cells, did not prove
sufficient to induce tumor formation in vivo (Zhou et al., 2000). However, one gene
(EphA2), which interestingly is directly regulated by HOXA1, has also been reported to
transform MCF-10A cells with consequent in vivo tumor formation (Zelinski et al., 2001).
In the present study, we utilized the HOX binding region of the EphA2 promoter (EphA2-
r42B) to measure HOXA1 transactivation, and therefore, EphA2 may also constitute a
pivotal component of the oncogenic program of HOXA1. The aberrant expression of homeobox-containing genes has been reported in a variety of neoplasias (see review by
Abate-Shen, 2002), including those of the mammary gland (Care et al., 1998; Raman et
al., 2000). In addition, Hoxa1 has been detected in carcinoma of the mammary gland in
the mouse but not in normal mammary tissue (Friedmann et al., 1994), and HOXA1
expression has also been detected in neoplastic lesions of the human mammary gland
(Chariot and Castronovo, 1996). However, it is debated whether homeobox-containing
genes should be classified as oncogene/tumor suppressor genes per se or only as "tumor
modulators" that tip the balance of tumor progression (Abate-Shen, 2002; Celetti et al.,
1993). In the present work, the results therefore provide the first direct and novel
evidence that a homeobox-containing gene is a bona fide mammary gland oncogene, the
enhanced expression of which is sufficient for tumorigenesis of human mammary
epithelial cells.
3. HOXA1 is upregulated by E-cadherin-activated signaling in human mammary
carcinoma cells. Homeobox genes encode transcriptional regulators involved in cell fate
217 and pattern formation during development (McGinnis and Krumlauf, 1992). Although the coordinated spatial and temporal expression of HOX genes is critical to their regulatory function, the vast majority of HOX genes upstream regulators have yet to be elucidated.
It has been reported that in the mouse mammary gland, Hoxc6, Msx1, and Msx2 expression are regulated by mammogenic hormones (Friedmann et al., 1994, 1996).
Hoxa1 and Hoxb7 were showed to be regulated by extracellular matrix-dependent signals in mammary epithelial cells (Srebrow et al., 1998). In this present study, we found that
HOXA1 expression is increased in human mammary carcinoma cells MCF-7 at full confluence. The increment of HOXA1 expression at full confluence in MCF-7 cells is directed by E-cadherin signaling through Rac1. Thus, E-cadherin, as well as Rac1, is the novel upstream regulators of HOXA1. To the best of our knowledge, E-cadheirn has not been shown to regulate HOX genes. Herein, the present study provides the first evidence that E-cadherin is a regulator of one of the HOX genes, HOXA1. The study is thus the first to demonstrate the cross link between E-cadherin and HOX signaling pathways. It is likely that E-cadherin also regulates other homeobox genes and future study is needed to address this point.
4. Increased HOXA1 expression by E-cadherin-activated signaling offers protection from apoptosis in human mammary carcinoma cells. The E-cadherin-activated signals have been shown to mediate adhesive stabilization and determine cellular recognition
(Kovacs et al., 2002; Yap and Kovacs, 2003). But the other functions of E-cadherin- activated signals are largely remained unknown. In the present studies, our data identify that the increased HOXA1 by E-cadheirn-activated signals protects mammary carcinoma
218
cells MCF-7 from apoptotic cell death. 1) MCF-7 cells at full confluence had reduced
apoptosis compared with cells at subconfluence. Transient transfection of the HOXA1
inhibitor, HOXA1 RNAi, completely prevented the reduced apoptosis at full confluence,
indicating that increased HOXA1 at full confluence is responsible for this anti-apoptotic
effect. 2) MCF-7 cells at subconfluence and full confluence had the same apoptotic level
in the presence of E-cadherin function-blocking antibody, DECMA-1. 3) Transfection of
HOXA1 expression plasmid ( a gift from Di Rocco et al., 1997) completely rescued the
anti-apoptosis prevention offered by DECMA-1. Taken together, the above results
demonstrated that increased HOXA1 expression by E-cadherin-activated signaling had
increased anti-apoptotic effect in human mammary carcinoma cells.
5.6 Future work
1. In this project, we demonstrated that HOXA1 is a gene which is regulated by
autocrine hGH by use of this in vitro cell model. But the downstream genes of autocrine
hGH are far from identified. Recently, it was demonstrated that autocrine production of
hGH in immortalized human mammary epithelial cells enhances proliferation, offers
protection from apoptosis, and results in oncogenic transformation and tumor formation
in vivo (Zhu et al., manuscript submitted). Therefore, systematic analysis of the downstream genes of autocrine hGH and the relevant mechanistic features by which it exerts its cellular effects is required. It is likely that autocrine production of hGH also
transcriptionally regulates other homeodomain-containing proteins. To identify them,
microarray is a powerful tool to be used.
219
In deed, another homeobox family member, EMX1, was also identified to be an autocrine hGH-regulated gene by microarray analysis in the laboratory of my supervisors. Emx1 gene is expressed in discrete, overlapping regions of the developing forebrain and midbrain in mice. Null mutant mice with this gene inactivation showed that it plays an important role in the patterning of the forebrain and midbrain (Acampora et al., 1995;
Ang et al., 1996; Pellegrini et al., 1996). The regulation and the role of EMX1 in
mammary carcinoma cells may therefore be a useful direction for further studies.
2. In the present study, it was shown that HOXA1 increased the protein level of
Bcl-2, an important molecule involved in cell apoptosis. The mechanism of upregulation
of Bcl-2 by HOXA1 may be at the transcriptional or the translational level or both. As
long as HOXA1 is a transcriptional factor, we are more interested in the transcriptional
regulation. Indeed, we demonstrate that HOXA1 increases Bcl-2 gene transcription.
Interestingly, another homeobox gene, HOXA5, has been shown recently to regulate p53
transcription in human breast cancer cells through direct binding at the HOX-core
binding sequences (ATTA) in the promoter of p53, another important molecule involved
in cell apoptosis (Raman et al., 2000).
This raises the question that whether HOXA1 can direct bind at the Bcl-2 promoter to
regulate its transcription. Interestingly, there are also several ATTA sequences in the
promoter of Bcl-2. To examine the direct binding of HOXA1 to the ATTA-containing
site in the Bcl-2 promoter, electrophoretic mobility shift (EMSA) and supershift assays
220
may need to be performed. Therefore, whether HOXA1 direct binds Bcl-2 promoter can
be examined by EMSA and supershift assay using HOXA1 protein, HOXA1 antibody,
and the DNA fragment which contains the ATTA sequence in Bcl-2 promoter.
3. Chemotherapy is an important option in the treatment of breast cancer. Signal
transduction pathways that regulate mammary epithelial cell proliferation, differentiation,
apoptosis, and transformation have been delineated and utilized as therapeutic targets for
the treatment of mammary gland neoplasia (Pavelic and Gall-Troselj, 2001; Hanahan and
Weinberg, 2000). Examples of such include selective estrogen modulators (Osborne et
al., 2000) and herceptin (Pegram and Slamon, 2000) targeting estrogen receptor and
HER-2 pathways, respectively. However, a significant proportion of carcinomas of the
mammary gland neither express the estrogen receptor (Osborne, 1998) nor overexpress
HER-2 (Slamon et al., 1987) and are, therefore, not responsive to these specific
therapeutic strategies. Clinical data for the treatment of cancer has clearly demonstrated
the superiority of combinatorial therapy over a single-agent approach (Frei and Antman,
2000). Delineation, characterization, and development of therapeutic regimes targeting
novel signaling pathways involved in oncogenic transformation of mammary epithelial cells are therefore required.
In this project, we identified a novel pathway which may be used for development of the
chemotherapy program. Resistance to chemotherapy in many cancers, including
mammary carcinoma, has been associated with increased expression of Bcl-2 (Bonetti et
al., 1998; Decaudin et al., 1997). This increase in Bcl-2, which was demonstrated in the
221
MCF7-HOXA1 cells, resulted in resistance to doxorubicin-induced cell death. Therefore, inhibition of HOXA1 transactivation may be a novel therapeutic target to overcome the development of resistance to chemotherapeutic drugs.
Both in vitro and in vivo models can be used for testing therapeutic approaches. Usually in vivo models are better than in vitro models. Results from in vitro assays have to be interpreted cautiously when attempting to predict an in vivo response, since many factors may not be directly transferable. Immunodeficient mice have become a very important model for the development of new therapeutic approaches (Povlsen and Jacobson, 1975;
Braakhuis et al., 1984). The main condition for an in vivo chemosensitivity test is the possibility of transplanting human tumors into animals. In the present study, we successfully transplanted human mammary carcinoma cells, MCF-7, and MCF10A-
HOXA1 cells into an in vivo model (SCID mice), resulting in tumor formation. In further studies, this same in vivo model can be used to test the effects of chemotherapeutic agents for breast cancer in the presence of HOXA1 inhibitor such as the antisense to
HOXA1, checking whether the signaling pathways involved in HOXA1 oncogenic transformation of mammary epithelial cells can be used as a therapeutic approach.
222
References
Abate-Shen C 2002 Deregulated homeobox gene expression in cancer: cause or consequence? Nat Rev Cancer 2: 777-785
Abdel-Meguid SS, Shieh HS, Smith WW, Dayringer HE, Violand BN, Bentle LA 1987 Three- dimensional structure of a genetically engineered variant of porcine growth hormone. Proc Natl Acad Sci USA 84:6434–6437
Aberdam D, Negreanu V, Sachs L, Blatt C 1991 The oncogenic potential of an activated Hox-2.4 homeobox gene in mouse fibroblasts. Mol Cell Biol 11:554-557
Acampora D, D'Esposito M, Faiella A, Pannese M, Migliaccio F, Morelli F, Stornaiuolo A, Nigro V, Simeone A, Boncinelli E 1989 The human HOX gene family. Nucleic Acids Res 17: 10385- 10402
Acampora D, Mazan S, Lallemand Y, Avantaggiato V, Maury M, Simeone A, Brulet P 1995 Forebrain and midbrain regions are deleted in Otx2-/- mutants due to a defective anterior neuroectoderm specification during gastrulation. Development 121:3279-3290
Akam ME 1987 The molecular basis for metameric pattern in the Drosophila embryo. Development 101:1-22
Alexander L, Appleton D, Hall R, Ross WM, Wilkinson R 1980 Epidemiology of acromegaly in the Newcastle region. Clin Endocrinol (Oxf) 12:71-79
Ang SL, Jin O, Rhinn M, Daigle N, Stevenson L, Rossant J 1996 A targeted mouse Otx2 mutation leads to severe defects in gastrulation and formation of axial mesoderm and to deletion of rostral brain. Development 122:243-52
Apiou F, Flagiello D, Cillo C, Malfoy B, Poupon MF, Dutrillaux B 1996 Fine mapping of human HOX gene clusters. Cytogenet Cell Genet 73: 114-115
Argetsinger LS, Campbell GS, Yang X, Witthuhn BA, Silvennoinen O, Ihle JN, Carter-Su C 1993 Identification of JAK2 as a growth hormone receptor-associated tyrosine kinase. Cell 74:237-244
Argetsinger LS, Carter-Su C 1996 Mechanism of signaling by growth hormone receptor. Physiol Rev 76:1089-1107
Azuma N, Nishina S, Yanagisawa H, Okuyama T, Yamada M 1996 PAX6 missense mutation in isolated foveal hypoplasia. Nat Genet 13:141-142
Bai YQ, Miyake S, Iwai T, Yuasa Y 2003 CDX2, a homeobox transcription factor, upregulates transcription of the p21/WAF1/CIP1 gene. Oncogene 22:5998-6005
Bankfalvi A, Terpe HJ, Breukelmann D, Bier B, Rempe D, Pschadka G, Krech R, Lelle RJ, Boecker W 1999 Immunophenotypic and prognostic analysis of E-cadherin and beta-catenin
223 expression during breast carcinogenesis and tumour progression: a comparative study with CD44. Histopathology 34:25-34
Barnes TM, Jin Y, Horvitz HR, Ruvkun G, Hekimi S 1996 The Cenorhabditis elegans behavioral gene unc-24 encodes a novel bipartite protein similar to both erythrocyte band 7.2 (stomatin) and nonspecific lipid transfer protein. J Neurochem 76: 46-57
Barrow JR, Capecchi MR 1996 Targeted disruption of the Hoxb-2 locus in mice interferes with expression of Hoxb-1 and Hoxb-4. Development 122:3817-3828
Basu A, Haldar S 1998 The relationship between Bcl2, Bax and p53: consequences for cell cycle progression and cell death. Mol Hum Reprod 4: 1099-1109
Bateson, W 1994 Materials for the Study of Variation (New York: Macmillan).
Behrens J, Mareel MM, Van Roy FM, Birchmeier W 1989 Dissecting tumor cell invasion: epithelial cells acquire invasive properties after the loss of uvomorulin-mediated cell-cell adhesion. J Cell Biol 108:2435-2447
Bengtsson BA, Eden S, Ernest I, Oden A, Sjogren B 1988 Epidemiology and long-term survival in acromegaly. A study of 166 cases diagnosed between 1955 and 1984. Acta Med Scand 223:327-335
Benlot C, Levy L, Fontanaud P, Roche A, Rouannet P, Joubert D 1997 Somatostatin and growth hormone-releasing hormone in normal and tumoral human breast tissue: endogenous content, in vitro pulsatile release, and regulation. J Clin Endocrinol Metab 82:690-696
Benson GV, Lim H, Paria BC, Satokata I, Dey SK, Maas RL 1996 Mechanisms of reduced fertility in Hoxa-10 mutant mice: uterine homeosis and loss of maternal Hoxa-10 expression. Development 122:2687-2696
Berx G, Cleton-Jansen AM, Strumane K, de Leeuw WJ, Nollet F, van Roy F, Cornelisse C 1996 E-cadherin is inactivated in a majority of invasive human lobular breast cancers by truncation mutations throughout its extracellular domain. Oncogene 13:1919-1925
Berx G, Nollet F, van Roy F 1998 Dysregulation of the E-cadherin/catenin complex by irreversible mutations in human carcinomas. Cell Adhes Commun 6:171-184
Bieberich CJ, Ruddle FH, Stenn KS 1991 Differential expression of the Hox 3.1 gene in adult mouse skin. Ann N Y Acad Sci 642:346-353
Biggin MD, McGinnis W 1997 Regulation of segmentation and segmental identity by Drosophila homeoproteins: the role of DNA binding in functional activity and specificity. Development 124:4425-4433
Bijl JJ, van Oostveen JW, Walboomers JM, Horstman A, van den Brule AJ, Willemze R, Meijer CJ 1997 HOX C4, HOX C5, and HOX C6 expression in non-Hodgkin's lymphoma: preferential expression of the HOX C5 gene in primary cutaneous anaplastic T-cell and oro-gastointestinal tract mucosa-associated B-cell lymphoma. Blood 90: 4116-4125
Bishop AL, Hall A 2000 Rho GTPases and their effector proteins. Biochem J 348 Pt 2:241-255
224
Blethen SL, Baptista J, Kuntze J, Foley T, LaFranchi S, Johanson A 1997 Adult height in growth hormone (GH)-deficient children treated with biosynthetic GH. J Clin Endocrinol Metab 82:418- 420
Boersma CJ, Bloemen M, Hendriks JM, van Berkel EA, Olijve W, van Zoelen EJ 1999 Homeobox proteins as signal transduction intermediates in regulation of NCAM expression by recombinant human bone morphogenetic protein-2 in osteoblast-like cells. Mol Cell Biol Res Commun 1: 117-124
Bogue CW, Lou LJ, Vasavada H, Wilson CM, Jacobs HC 1996 Expression of Hoxb genes in the developing mouse foregut and lung. Am J Respir Cell Mol Biol 15:163-171
Bonetti A, Zaninelli M, Leone R, Cetto GL, Pelosi G, Biolo S, Menghi A, Manfrin E, Bonetti F, Piubello Q 1998 bcl-2 but not p53 expression is associated with resistance to chemotherapy in advanced breast cancer. Clin Cancer Res 4:2331-2336
Boudreau N, Andrews C, Srebrow A, Ravanpay A, Cheresh DA 1997 Induction of the angiogenic phenotype by Hox D3. J Cell Biol 139:257-264
Boulay JL, Dennefeld C, Alberga A 1987 The drosophila developmental gene snail encodes a protein with nucleic acid binding fingers. Nature 330:395–398
Braakhuis BJM, Sneeuwloper G, Snow GB 1984 The potential of the nude mouse xenograft model for the study of head and neck cancer. Arch Otorhinolaryngol 239:69-79
Bradley RS, Cowin P 1992 Expression of wnt-1 in pc 12 cells results in modulation of plakoglobin and E-cadherin and increased cellular adhesion. J Cell Biol 123:1857–1865
Bradley DJ, Towle HC, Young WS 1992 Spatial and temporal expression of alpha- and beta- thyroid hormone receptor mRNAs, including the beta 2-subtype, in the developing mammalian nervous system. J Neurosci 12:2288-2302
Braga VM, Machesky LM, Hall A, Hotchin NA 1997 The Small GTPases Rho and Rac Are Required for the Establishment of Cadherin-dependent Cell-Cell Contacts J Cell Biol 137: 1421- 1431
Bromleigh VC, Freedman LP 2000 p21 is a transcriptional target of HOXA10 in differentiating myelomonocytic cells. Genes Dev 14:2581-2586
Brunelli S, Faiella A, Capra V, Nigro V, Simeone A, Cama A, Boncinelli E 1996 Germline mutations in the homeobox gene EMX2 in patients with severe schizencephaly. Nat Genet 12:94-96
Brunner N, Osborne CK, Spang-Thomsen M 1987 Endocrine therapy of human breast cancer grown in nude mice. Breast Cancer Res Treat 10:229-242
Burglin TR 1994 A comprehensive classification of homeobox genes. In Guidebook to the Homeobox genes. Oxford, England: Oxford Unoversity Press: 27-269
225
Burridge K, Turner CE, Romer LH 1992 Tyrosine phosphorylation of paxillin and pp125FAK accompanies cell adhesion to extracellular matrix: a role in cytoskeletal assembly. J Cell Biol 1194:893-903
Colao A, Di Sarno A, Landi ML, Cirillo S, Sarnacchiaro F, Facciolli G, Pivonello R, Cataldi M, Merola B, Annunziato L, Lombardi G 1997 Long-term and low-dose treatment with cabergoline induces macroprolactinoma shrinkage. J Clin Endocrinol Metab 82:3574-3579
Campbell GS, Pang L, Miyasaka T, Saltiel AR, Carter-Su C 1992 Stimulation by growth hormone of MAP kinase activity in 3T3-F442A fibroblasts. J Biol Chem 267: 6074–6080
Campbell GS, Meyer DJ, Raz R, Levy DE, Schwartz J, Carter-Su C 1995 Activation of acute phase response factor (APRF)/Stat3 transcription factor by growth hormone. J Biol Chem 270:3974-3979
Cano A, Perez-Moreno MA, Rodrigo I, Locascio A, Blanco MJ, del Barrio MG, Portillo F, Nieto MA 2000 The transcription factor Snail controls epithelial–mesenchymal transitions by repressing E-cadherin expression. Nat Cell Biol 2 :76–83
Care A, Felicetti F, Meccia E, Bottero L, Parenza M, Stoppacciaro A, Peschle C, Colombo MP 2001 HOXB7: a key factor for tumor-associated angiogenic switch. Cancer Res 61:6532-6539
Care A, Silvani A, Meccia E, Mattia G, Peschle C, Colombo MP 1998 Transduction of the SkBr3 breast carcinoma cell line with the HOXB7 gene induces bFGF expression, increases cell proliferation and reduces growth factor dependence. Oncogene 16:3285–3289
Carpenter EM, Goddard JM, Chisaka O, Manley NR, Capecchi MR 1993 Loss of Hox-A1 (Hox- 1.6) function results in the reorganization of the murine hindbrain. Development 118:1063-1075
Carter-Su C, King AP, Argetsinger LS, Smit LS, Vanderkuur J, Campbell GS 1996 Signalling pathway of growth hormone. Endocr J 43 Suppl:S65-70
Carter-Su C, Rui L, Herrington J 2000 Role of the tyrosine kinase JAK2 in signal transduction by growth hormone. Pediatr Nephrol 14:550–557
Casanueva FF 1992 Physiology of growth hormone secretion and action. Endocrinol Metab Clin North Am 21:483-517
Castronovo V, Kusaka M, Chariot A, Gielen J, Sobel M 1994 Homeobox genes: potential candidates for the transcriptional control of the transformed and invasive phenotype. Biochem Pharmacol 47:137-143
Celetti A, Barba P, Cillo C, Rotoli B, Boncinelli E, Magli MC 1993 Characteristic patterns of HOX gene expression in different types of human leukemia. Int J Cancer 53: 237-244
Chang PY, Kozono T, Chida K, Kuroki T, Huh N 1998 Differential expression of Hox genes in multistage carcinogenesis of mouse skin. Biochem Biophys Res Commun 248:749-752
Chariot A, Moreau L, Senterre G, Sobel ME, Castronovo V 1995 Retinoic acid induces three newly cloned HOXA1 transcripts in MCF7 breast cancer cells. Biochem Biophys Res Commun 215:713-720
226
Chariot A, Castronovo V 1996 Detection of HOXA1 expression in human breast cancer. Biochem Biophys Res Commun 222:292-297
Chen F, Capecchi MR 1999 Paralogous mouse Hox genes, Hoxa9, Hoxb9, and Hoxd9, function together to control development of the mammary gland in response to pregnancy. Proc Natl Acad Sci U S A 96:541-546
Cillo C, Cantile M, Faiella A, Boncinelli E 2001 Homeobox genes in normal and malignant cells. J Cell Physiol 188:161–169
Cillo C, Cantile M, Mortarini R, Barba P, Parmiani G, Anichini A 1996 Differential patterns of HOX gene expression are associated with specific integrin and ICAM profiles in clonal populations isolated from a single human melanoma metastasis. Int J Cancer 66:692-697
Clark GJ, Kinch MS, Gilmer TM, Burridge K, Der CJ 1996 Overexpression of the Ras-related TC21/R-Ras2 protein may contribute to the development of human breast cancers. Oncogene 12:169-176
Condie BG, Capecchi MR 1993 Mice homozygous for a targeted disruption of Hoxd-3 (Hox-4.1) exhibit anterior transformations of the first and second cervical vertebrae, the atlas and the axis. Development 119:579-595
Cooke NE, Ray J, Watson MA, Estes PA, Kuo BA, Liebhaber SA 1988 Human growth hormone gene and the highly homologous growth hormone variant gene display different splicing patterns. J Clin Invest 82:270-275
Copeland NG, Gilbert DJ, Schindler C, Zhong Z, Wen Z, Darnell JE Jr, Mui AL, Miyajima A, Quelle FW, Ihle JN 1995 Distribution of the mammalian Stat gene family in mouse chromosomes. Genomics 29:225-228
Costa C, Solanes G, Visa J, Bosch F 1998 Transgenic rabbits overexpressing growth hormone develop acromegaly and diabetes mellitus. FASEB J 12:1455-1460
Crews CM, Erikson RL 1993 Extracellular signals and reversible protein phosphorylation: what to Mek of it all. Cell 74:215-217
Croix B, Sheehan C, Rak JW, Florenes VA, Slingerland JM, Kerbel RS 1998 E-Cadherin- dependent growth suppression is mediated by the cyclin-dependent kinase inhibitor p27(KIP1). J Cell Biol 142:557-571
Cunningham BC, Wells JA 1991 Rational design of receptor-specific variants of human growth hormone. Proc Natl Acad Sci U S A 88:3407-3411
Cunningham BC, Bass S, Fuh G, Wells JA 1990 Zinc mediation of the binding of human growth hormone to the human prolactin receptor. Science 250:1709-1712
David M, Petricoin E, Benjamin C, Pine R, Weber MJ, Larner AC 1995 Requirement for MAP kinase (ERK2) activity in interferon alpha- and interferon beta-stimulated gene expression through STAT proteins. Science 269:1721-1723
227
Davis RJ 1993 The mitogen-activated protein kinase signal transduction pathway. J Biol Chem 268:14553-14556
Davis AP, Capecchi MR 1994 Axial homeosis and appendicular skeleton defects in mice with a targeted disruption of hoxd-11. Development 120:2187-2198
Davis AP, Capecchi MR 1996 A mutational analysis of the 5' HoxD genes: dissection of genetic interactions during limb development in the mouse. Development 122:175-185
Davis AP, Witte DP, Hsieh-Li HM, Potter SS, Capecchi MR 1995 Absence of radius and ulna in mice lacking hoxa-11 and hoxd-11. Nature 375:791-795
Davideau JL, Demri P, Hotton D, Gu TT, MacDougall M, Sharpe P, Forest N, Berdal A 1999 Comparative study of MSX-2, DLX-5, and DLX-7 gene expression during early human tooth development. Pediatr Res 46:650-656
Day ML, Zhao X, Vallorosi CJ, Putzi M, Powell CT, Lin C, Day KC 1999 E-cadherin mediates aggregation-dependent survival of prostate and mammary epithelial cells through the retinoblastoma cell cycle control pathway. J Biol Chem 274:9656-9664
De Leeuw WJ, Berx G, Vos CB, Peterse JL, Van de Vijver MJ, Litvinov S, Van Roy F, Cornelisse CJ, Cleton-Jansen AM 1997 Simultaneous loss of E-cadherin and catenins in invasive lobular breast cancer and lobular carcinoma in situ. J Pathol 183:404-411
De Vita G, Barba P, Odartchenko N, Givel JC, Freschi G, Bucciarelli G, Magli MC, Boncinelli E, Cillo C 1993 Expression of homeobox-containing genes in primary and metastatic colorectal cancer. Eur J Cancer 6:887-893
De Vita G, Zannini M, Cirafici AM, Melillo RM, Di Lauro R, Fusco A, Santoro M 1998 Expression of the RET/PTC1 oncogene impairs the activity of TTF-1 and Pax-8 thyroid transcription factors. Cell Growth Differ 9:97-103
de Vos AM, Ultsch M, Kossiakoff AA 1992 Human growth hormone and extracellular domain of its receptor: Crystal structure of the complex. Science 255:306–312
Decaudin D, Geley S, Hirsch T, Castedo M, Marchetti P, Macho A, Kofler R, Kroemer G 1997 Bcl-2 and Bcl-XL antagonize the mitochondrial dysfunction preceding nuclear apoptosis induced by chemotherapeutic agents. Cancer Res 57:62-67
Deguchi Y, Kirschenbaum A, Kehrl JH 1992 A diverged homeobox gene is involved in the proliferation and lineage commitment of human hematopoietic progenitors and highly expressed in acute myelogenous leukemia. Blood 79:2841-2848
Deguchi Y, Kehrl JH 1993 High level expression of the homeobox gene HB24 in a human T-cell line confers the ability to form tumors in nude mice. Cancer Res 53:373-377
Dekker EJ, Pannese M, Houtzager E, Timmermans A, Boncinelli E, Durston A 1992 Xenopus Hox-2 genes are expressed sequentially after the onset of gastrulation and are differentially inducible by retinoic acid. Development Supplement :195-202
228
Dhanasekaran N, Premkumar Reddy E 1998 Signaling by dual specificity kinases. Oncogene 17:1447-55
Di Rocco G, Mavilio F, Zappavigna V 1997 Functional dissection of a transcriptionally active, target-specific Hox-Pbx complex. EMBO J 16:3644-3654
Dillon DA, D'Aquila T, Reynolds AB, Fearon ER, Rimm DL 1998 The expression of p120ctn protein in breast cancer is independent of alpha- and beta-catenin and E-cadherin. Am J Pathol 152:75-82
Dressler GR, Douglass EC 1992 Pax-2 is a DNA-binding protein expressed in embryonic kidney and Wilms' tumor. Proc Natl Acad Sci USA 89:1179-1183
Duboule D 1994 Guidebook to the Homeobox Genes. Oxford: Oxford University Press.
Duboule D 1991 Patterning in the vertebrate limb. Curr Opin Genet Dev 1:211-216
Duboule D 1992 The vertebrate limb: a model system to study the Hox/HOM gene network during development and evolution. Bioessays 14:375-384
Duckworth ML, Peden LM, Friesen HG 1986 Isolation of a novel prolactin-like cDNA clone from developing rat placenta. J Biol Chem 261:10879–10884
Edelman GM, Jones FS 1993 Outside and downstream of the homeobox. J Biol Chem 268:20683-20686
Edelman GM, Jones FS 1998 Gene regulation of cell adhesion: a key step in neural morphogenesis. Brain Res Brain Res Rev 26:337-352
Emerman JT, Leahy M, Gout PW, Bruchovsky N 1985 Elevated growth hormone levels in sera from breast cancer patients. Horm Metab Res 17:421-424
Enyedy IJ, Ling Y, Nacro K, Tomita Y, Wu X, Cao Y, Guo R, Li B, Zhu X, Huang Y, Long YQ, Roller PP, Yang D, Wang S 2001 Discovery of small-molecule inhibitors of Bcl-2 through structure-based computer screening. J Med Chem 44: 4313-4324
Evans HM, Long JA 1921 The effect of the anterior lobe administered intraperitoneally upon growth, maturity, and oestrus cycles of the rat. Anat Rec 21:62
Fagotto F, Gumbiner BM 1996 Cell contact-dependent signaling. Dev Biol 180:445-454
Favier B, Dolle P 1997 Developmental functions of mammalian Hox genes. Mol Hum Reprod 3:115-31
Favier B, Le Meur M, Chambon P, Dolle P 1995 Axial skeleton homeosis and forelimb malformations in Hoxd-11 mutant mice. Proc Natl Acad Sci U S A 92:310-314
Favier B, Rijli FM, Fromental-Ramain C, Fraulob V, Chambon P, Dolle P 1996 Functional cooperation between the non-paralogous genes Hoxa-10 and Hoxd-11 in the developing forelimb and axial skeleton. Development 122:449-460
229
Feldman M, Ruan W, Cunningham BC, Wells JA, Kleinberg DL 1993 Evidence that the growth hormone receptor mediates differentiation and development of the mammary gland. Endocrinology 133:1602-1608
Ferber S, Halkin A, Cohen H, Ber I, Einav Y, Goldberg I, Barsack I, Sejffers R, Kopolovic J, Kaiser N, Karasik A 2000 Pancreatic and duodenal homeobox gene 1 induces expression of insulin genes in liver and ameliorates streptozoticin-induced hyperglycemia. Nat Med 6: 568-572
Finidori J 2000 Regulators of growth hormone signaling. Vitam Horm 59:71-97
Ford HL, Kabingu EN, Bump EA, Mutter GL, Pardee AB 1998 Abrogation of the G2 cell cycle checkpoint associated with overexpression of HSIX1: a possible mechanism of breast carcinogenesis. Proc Natl Acad Sci U S A 95:12608-12613
Foster CM, Shafer JA, Rozsa FW, Wang XY, Lewis SD, Renken DA, Natale JE, Schwartz J, Carter-Su C 1988 Growth hormone promoted tyrosyl phosphorylation of growth hormone receptors in murine 3T3-F442A fibroblasts and adipocytes Biochemistry 27:326–334
Frank SJ, Gilliland G, Kraft AS, Arnold CS 1994 Interaction of the growth hormone receptor cytoplasmic domain with the JAK2 tyrosine kinase. Endocrinology 135:2228-2239
Frei E, Antman KH 2000 Cancer Medicine (BC Decker, London) 5th Ed:556-568
Fridman R, Kibbey MC, Royce LS, Zain M, Sweeney M, Jicha DL, Yannelli JR, Martin GR, Kleinman HK 1991 Enhanced tumor growth of both primary and established human and murine tumor cells in athymic mice after coinjection with Matrigel. J Natl Cancer Inst 83:769-774
Fridman R, Giaccone G, Kanemoto T, Martin GR, Gazdar AF, Mulshine JL 1990 Reconstituted basement membrane (matrigel) and laminin can enhance the tumorigenicity and the drug resistance of small cell lung cancer cell lines. Proc Natl Acad Sci U S A 87:6698-6702
Friedmann Y 1995 Expression and developmental role of homeobox containing genes during mouse mammary gland morphogenesis. In Biology. Santa Cruz, CA: University of California
Friedmann Y, Daniel CA, Strickland P, Daniel CW 1994 Hox genes in normal and neoplastic mouse mammary gland. Cancer Res 54:5981-5985
Friedmann Y, Daniel CW 1996 Regulated expression of homeobox genes Msx-1 and Msx-2 in mouse mammary gland development suggests a role in hormone action and epithelial-stromal interactions. Dev Biol 177:347-355
Frisen J, Holmberg J, Barbacid M 1999 Ephrins and their Eph receptors: multitalented directors of embryonic development. EMBO J 18:5159-5165
Fromental-Ramain C, Warot X, Lakkaraju S, Favier B, Haack H, Birling C, Dierich A, Doll e P, Chambon P 1996 Specific and redundant functions of the paralogous Hoxa-9 and Hoxd-9 genes in forelimb and axial skeleton patterning. Development 122:461-472
Fromental-Ramain C, Warot X, Messadecq N, LeMeur M, Dolle P, Chambon P 1996 Hoxa-13 and Hoxd-13 play a crucial role in the patterning of the limb autopod. Development 122:2997- 3011
230
Galili N, Davis RJ, Frederics WJ, Mukhopadhyay S, Rauscher FJ, Emanuel BS, Rovera G, Barr FG 1993 Fusion of a fork head domain gene to PAX3 in the solid tumor alveolar rhabdomyosarcoma. Nat Genet 5: 230-235
Garcia-Bellido AG 1975 Genetic control of wing disk development in Drosophila. In Cell Patterning Ciba Found Symp 29:161-178
Gehring WJ, Ikeo K 1999 Pax 6: mastering eye morphogenesis and eye evolution. Trends Genet 15: 371-377
Gendron-Maguire M, Mallo M, Zhang M, Gridley T 1993 Hoxa-2 mutant mice exhibit homeotic transformation of skeletal elements derived from cranial neural crest. Cell 75:1317-1331
Giunciuglio D, Culty M, Fassina G, Masiello L, Melchiori A, Paglialunga F, Arand G, Ciardiello F, Basolo F, Thompson EW 1995 Invasive phenotype of MCF10A cells overexpressing c-Ha-ras and c-erbB-2 oncogenes. Int J Cancer 63:815-822
Glaser T, Jepeal L, Edwards JG, Young SR, Favor J, Maas RL 1994 PAX6 gene dosage effect in a family with congenital cataracts, aniridia, anophthalmia and central nervous system defects. Nat Genet:463-471
Goddard JM, Rossel M, Manley NR, Capecchi MR 1996 Mice with targeted disruption of Hoxb- 1 fail to form the motor nucleus of the VIIth nerve. Development 122:3217-3228
Goffin V, Bernichtein S, Carriere O, Bennett WF, Kopchick JJ, Kelly PA. 1999 The human growth hormone antagonist B2036 does not interact with the prolactin receptor. Endocrinology 140:3853-3856
Goffin V, Kelly PA 1997 The prolactin/growth hormone receptor family: structure/function relationships. J Mammary Gland Biol Neoplasia 2:7–17
Goh EL, Pircher TJ, Wood TJ, Norstedt G, Graichen R, Lobie PE 1997 Growth hormone-induced reorganization of the actin cytoskeleton is not required for STAT5 (signal transducer and activator of transcription-5)-mediated transcription. Endocrinology 138:3207-3215
Goh EL, Zhu T, Yakar S, LeRoith D, Lobie PE 2000 CrkII participation in the cellular effects of growth hormone and insulin-like growth factor-1. Phosphatidylinositol-3 kinase dependent and independent effects. J Biol Chem 275:17683-17692
Goh Kl, Yang JT, Hynes RO 1997 Mesodermal defects and cranial neural crest apoptosis in alpha5 integrin-null embryos. Development 124:4309-4319
Gong TW, Meyer DJ, Liao J, Hodge CL, Campbell GS, Wang X, Billestrup N, Carter-Su C, Schwartz J 1998 Regulation of glucose transport and c-fos and egr-1 expression in cells with mutated or endogenous growth hormone receptors. Endocrinology 139:1863-1871
Gonzalez MA, Pinder SE, Wencyk PM, Bell JA, Elston CW, Nicholson RI, Robertson JF, Blamey RW, Ellis IO, 1999 An immunohistochemical examination of the expression of E- cadherin, alpha- and beta/gamma-catenins, and alpha2- and beta1-integrins in invasive breast cancer. J Pathol 187:523-529
231
Graff JR, Gabrielson E, Fujii H, Baylin SB, Herman JG 2000 Methylation patterns of the E- cadherin 5' CpG island are unstable and reflect the dynamic, heterogeneous loss of E-cadherin expression during metastatic progression. J Biol Chem 275:2727-2732
Graff JR, Herman JG, Lapidus RG, Chopra H, Xu R, Jarrard DF, Isaacs WB, Pitha PM, Davidson NE, Baylin SB 1995 E-cadherin expression is silenced by DNA hypermethylation in human breast and prostate carcinomas. Cancer Res 55:5195-5199
Graichen R, Liu D, Sun Y, Lee KO, Lobie PE 2002 Autocrine human growth hormone inhibits placental transforming growth factor-beta gene transcription to prevent apoptosis and allow cell cycle progression of human mammary carcinoma cells. J Biol Chem 277:26662-26672
Grau Y, Carteret C, Simpson P 1984 Mutations and chromosomal rearrangements affecting the expression of snail, a gene involved in embryonic patterning. Genetics 108:347–360
Gross A., McDonnell JM, Korsmeyer SJ 1999 BCL-2 family members and the mitochondria in apoptosis. Genes Dev 13:1899-1911
Guan JL, Shalloway D 1992 Regulation of focal adhesion-associated protein tyrosine kinase by both cellular adhesion and oncogenic transformation. Nature 358:690-692
Guilford PJ, Hopkins JB, Grady WM, Markowitz SD, Willis J, Lynch H, Rajput A, Wiesner GL, Lindor NM, Burgart LJ, Toro TT, Lee D, Limacher JM, Shaw DW, Findlay MP, Reeve AE 1999 E-cadherin germline mutations define an inherited cancer syndrome dominated by diffuse gastric cancer. Hum Mutat 14:249-255
Gumbiner BM 1996 Cell adhesion: the molecular basis of tissue architecture and morphogenesis. Cell 84:345-357
Guriec N, Marcellin L, Gairard B, Calderoli H, Wilk A, Renaud R, Bergerat JP, Oberling F 1996 E-cadherin mRNA expression in breast carcinomas correlates with overall and disease-free survival. Invasion Metastasis 16:19-26
Gutierrez-Puente Y, Zapata-Benavides P, Tari AM, Lopez-Berestein G 2002 Bcl-2-related antisense therapy. Semin Oncol 29: 71-77
Hahn WC, Counter CM, Lundberg AS, Beijersbergen RL, Brooks MW, Weinberg RA 1999 Creation of human tumour cells with defined genetic elements. Nature 400: 464-468
Hall BK, Miyake T 2000 All for one and one for all: condensations and initiation of skeletal development. Bioessays 22:138-147
Han Y, Leaman DW, Watling D, Rogers NC, Groner B, Kerr IM, Wood WI, Stark GR 1996 Participation of JAK and STAT proteins in growth hormone-induced signaling J Biol Chem 271: 5947–5952
Hanahan D, Weinberg RA 2000 The hallmarks of cancer. Cell 100:57-70
Hanks M, Wurst W, Anson-Cartwright L, Auerbach AB, Joyner AL 1995 Rescue of the En-1 mutant phenotype by replacement of En-1 with En-2. Science 269:679-682
232
Hansen LH, Wang X, Kopchick JJ, Bouchelouche P, Nielsen JH, Galsgaard ED, Billestrup N 1996 Identification of tyrosine residues in the intracellular domain of the growth hormone receptor required for transcriptional signaling and Stat5 activation. J Biol Chem 271:12669– 12673
Hansen MD, Ehrlich JS, Nelson WJ 2002 Molecular mechanism for orienting membrane and actin dynamics to nascent cell-cell contacts in epithelial cells. J Biol Chem 277:45371-45376
Harpur AG, Andres AC, Ziemiecki A, Aston RR, Wilks AF 1992 JAK2, a third member of the JAK family of protein tyrosine kinases. Oncogene 7:1347-1353
Hauser SD, McGrath MF, Collier RJ, Krivi GG 1990 Cloning and in vivo expression of bovine growth hormone receptor mRNA. Mol Cell Endocrinol 72:187-200
Hawkins PT, Eguinoa A, Qiu RG, Stokoe D, Cooke FT, Walters R, Wennstrom S, Claesson- Welsh L, Evans T, Symons M 1995 PDGF stimulates an increase in GTP-Rac via activation of phosphoinositide 3-kinase. Curr Biol 5:393-403
Hayakawa K, Hasegawa M, Kawashima M, Nakamura Y, Matsuura M, Toda H, Hayakawa K, Mitsuhashi N, Niibe H 2000 Comparison of effects of doxorubicin and radiation on p53- dependent apoptosis in vivo. Oncol Rep 7:267-270
Hellgren G, Jansson JO, Carlsson LM, Carlsson B 1999 The growth hormone receptor associates with Jak1, Jak2 and Tyk2 in human liver. Growth Horm IGF Res 9:212-218
Hengst LV, Dulie V, Slingerland JM, Lees E, Read SI 1994 A cell cycle-regulated inhibitor of cyclin-dependent kinases. Proc Natl Acad Sci USA 91: 5291–5295
Hengst L, Read SI 1996 Translational control of p27kip1 accumulation during theh cell cycle. Science 271: 1861–1864
Herdegen T, Skene P, Bahr M 1997 The c-Jun transcription factor – bipotential mediator of neuronal death, survival and regeneration. Trends Neurosci 205: 227–231
Hermiston ML, Gordon JI 1995 In vivo analysis of cadherin function in the mouse intestinal epithelium: essential roles in adhesion, maintenance of differentiation, and regulation of programmed cell death. J Cell Biol 129:489-506
Hinck L, Nathke IS, Papkoff J, Nelson WJ 1994 Dynamics of cadherin/catenin complex formation: novel protein interactions and pathways of complex assembly. J Cell Biol 125: 1327– 1340
Hirsch MR, Gaugler L, Deagostini-Bazin H, Bally-Cuif L, Goridis C 1990 Identification of positive and negative regulatory elements governing cell-type-specific expression of the neural cell adhesion molecule gene. Mol Cell Biol 10:1959-1968
Hirt H, Kimelman J, Birnbaum MJ, Chen EY, Seeburg PH, Eberhardt NL, Barta A 1987 The human growth hormone gene locus: Structure, evolution, and allelic variations. Dna 6:59–70
Hiscox S, Jiang WG 1999 Hepatocyte growth factor/scatter factor disrupts E-cadherin mediated cell–cell adhesion by induction of the phosphorylation of -catenin. Anticancer Res 19:509–517
233
Hodge C, Liao J, Stofega M, Guan K, Carter-Su C, Schwartz J 1998 Growth hormone stimulates phosphorylation and activation of elk-1 and expression of c-fos, egr-1, and junB through activation of extracellular signal-regulated kinases 1 and 2. J Biol Chem 273:31327-31336
Holder N, Klein R 1999 Eph receptors and ephrins: effectors of morphogenesis. Development 126:2033-2044
Hordijk PL, ten Klooster JP, van der Kammen RA, Michiels F, Oomen LC, Collard JG 1997 Inhibition of Invasion of Epithelial Cells by Tiam1-Rac Signaling. Science 278:1464-1467
Horvath CM, Darnell JE 1997 The state of the STATs: recent developments in the study of signal transduction to the nucleus. Curr Opin Cell Biol 9:233–239
Hsieh-Li HM, Witte DP, Weinstein M, Branford W, Li H, Small K, Potter SS 1995 Hoxa 11 structure, extensive antisense transcription, and function in male and female fertility. Development 121:1373-1385
Hu G, Lee H, Price SM, Shen MM, Abate-Shen C 2001 Msx homeobox genes inhibit differentiation through upregulation of cyclin D1. Development 128:2373-2384
Huiping C, Sigurgeirsdottir JR, Jonasson JG, Eiriksdottir G, Johannsdottir JT, Egilsson V, Ingvarsson S 1999 Chromosome alterations and E-cadherin gene mutations in human lobular breast cancer. Br J Cancer 81:1103-1110
Hull KL, Harvey S 2001 Growth hormone: roles in female reproduction. J Endocrinol 168:1-23
Hunter DJ, Willett WC 1993 Diet, body size, and breast cancer. Epidemiol Rev 15:110-132
Ilkbahar YN, Wu K, Thordarson G, Talamantes F 1995 Expression and distribution of messenger ribonucleic acids for growth hormone (GH) receptor and GH-binding protein in mice during pregnancy. Endocrinology 136:386-392
Isaksson OG, Eden S, Jansson JO 1985 Mode of action of pituitary growth hormone on target cells. Annu Rev Physiol 47:483-499
Isaksson OG, Jansson JO, Gause IA 1982 Growth hormone stimulates longitudinal bone growth directly. Science 216:1237-1239
Ituarte EA, Petrini J, Hershman JM 1984 Acromegaly and colon cancer. Ann Intern Med 101:627-628
Izon DJ, Rozenfeld S, Fong ST, Komuves L, Largman C, Lawrence HJ 1998 Loss of function of the homeobox gene Hoxa-9 perturbs early T-cell development and induces apoptosis in primitive thymocytes. Blood 92:383-393
Jacobs HC, Bogue CW, Pinter E, Wilson CM, Warshaw JB, Gross I 1998 Fetal lung mRNA levels of Hox genes are diferentially altered by maternal diabetes and butyrate in rats. Pediatr Res 44: 99-104
234
Jacquemier J, Eisinger F, Nogues C, Sun ZZ, Guinebretiere JM, Peyrat JP, Geneix J, Lidereau R, Birnbaum D, Sobol H 1999 Histological type and syncytial growth pattern affect E-cadherin expression in a multifactorial analysis of a combined panel of sporadic and BRCA1-associated breast cancers. Int J Cancer 83:45-49
Jammes H, Gaye P, Belair L, Djiane J 1991 Identification and characterization of growth hormone receptor mRNA in the mammary gland. Mol Cell Endocrinol 75:27-35
Jansen B, Uckun FM, Jaszcz WB, Kersey JH 1992 Establishment of a human t(4;11) leukemia in severe combined immunodeficient mice and successful treatment using anti-CD19 (B43)- pokeweed antiviral protein immunotoxin. Cancer Res 52:406-412
Jiang WG 1996 E-cadherin and its associated protein catenins, cancer invasion and metastasis. Br J Surg 83:437–446
Jiang WG, Mansel RE 2000 E-cadherin complex and its abnormalities in human breast cancer. Surg Oncol 9:151-171
Jiang XR, Jimenez G, Chang E, Frolkis M, Kusler B, Sage M, Beeche M, Bodnar AG, Wahl GM, Tlsty TD, Chiu CP 1999 Telomerase expression in human somatic cells does not induce changes associated with a transformed phenotype. Nat Genet 21:111-114
Jin T, Branch DR, Zhang X, Qi S, Youngson B, Goss PE 1999 Examination of POU homeobox gene expression in human breast cancer cells. Int J Cancer 81:104-112
Johannsson G, Marin P, Lonn L, Ottosson M, Stenlof K, Bjorntorp P, Sjostrom L, Bengtsson BA 1997 Growth hormone treatment of abdominally obese men reduces abdominal fat mass, improves glucose and lipoprotein metabolism, and reduces diastolic blood pressure. J Clin Endocrinol Metab 82:727-734
Jones FS, Crossin KL, Cunningham BA, Edelman GM 1990 Identification and characterization of the promoter for the cytotactin gene. Proc Natl Acad Sci U S A. 87:6497-6501
Jones FS, Holst BD, Minowa O, De Robertis EM, Edelman GM 1993 Binding and transcriptional activation of the promoter for the neural cell adhesion molecule by HoxC6 (Hox-3.3). Proc Natl Acad Sci U S A 90:6557-6561
Jones FS, Prediger EA, Bittner DA, de Robertis EM, Edelman GM 1992 Cell adhesion molecules as targets for Hox genes: neural cell adhsion molecule promoter activity is modulated by cotransfection with Hox- 2.5 and 2.4. Proc Natl Acad Sci USA 89:2086-2090
Jou TS, Nelson WJ 1998 Effects of regulated expression of mutant RhoA and Rac1 small GTPases on the development of epithelial (MDCK) cell polarity. J Cell Biol 142:85-100
Kahan Z, Varga JL, Schally AV, Rekasi Z, Armatis P, Chatzistamou L, Czompoly T, Halmos G 2000 Antagonists of growth hormone-releasing hormone arrest the growth of MDA-MB-468 estrogen-independent human breast cancers in nude mice. Breast Cancer Res Treat 60:71-79
Kantak SS, Kramer RH 1998 E-cadherin regulates anchorage-independent growth and survival in oral squamous cell carcinoma cells. J Biol Chem 273:16953-16961
235
Kapp U, Dux A, Schell-Frederick E, Banik N, Hummel M, Mucke S, Fonatsch C, Bullerdiek J, Gottstein C, Engert A 1994 Disseminated growth of Hodgkin's-derived cell lines L540 and L540cy in immune-deficient SCID mice. Ann Oncol 5 Suppl 1:121-126
Kappen S 2000 Analysis of a complete homeobox gene repertoire: implications for the evolution of diversity. Proc Natl Acad Sci USA 97:4481-4486
Kaulsay KK, Mertani HC, Lee KO, Lobie PE 2000 Autocrine human growth hormone enhancement of human mammary carcinoma cell spreading is Jak2 dependent. Endocrinology 141:1571-1584
Kaulsay KK, Mertani HC, Tornell J, Morel G, Lee KO, Lobie PE 1999 Autocrine stimulation of human mammary carcinoma cell proliferation by human growth hormone. Exp Cell Res 250:35- 50
Kaulsay KK, Zhu T, Bennett W, Lee KO, Lobie PE 2001 The effects of autocrine human growth hormone (hGH) on human mammary carcinoma cell behavior are mediated via the hGH receptor. Endocrinology 142:767-777
Kawabe T, Muslin AJ, Korsmeyer SJ 1997 HOX11 interacts with protein phosphatases PP2A and PP1 and disrupts a G2/M cell-cycle checkpoint. Nature 385:454-458
Kawata A, Han T, Dadey B, Weier HU, Okazaki M, Yokota S, Fukiage T, Xiao H, Block AM, Barcos M 1993 Establishment and characterization of the tumors of chronic lymphocytic leukemia cell line in nude and SCID mice. Leuk Res 17:883-894
Kelly PA, Bachelot A, Kedzia C, Hennighausen L, Ormandy CJ, Kopchick JJ, Binart N 2002 The role of prolactin and growth hormone in mammary gland development. Mol Cell Endocrinol 197:127-131
Kelly PA, Edery M, Finidori J, Postel-Vinay MC, Gougon L, Ali S, Dinerstein H, Sotiropoulos A, Lochnan H, Ferrag F 1994 Receptor domains involved in signal transduction of prolactin and growth hormone. Proc Soc Exp Biol Med 206:280-283
Kelley RI, Hannekam RC 2000 The Smith-Lemli-Opitz syndrome. J Med Genet 37:321-335
Kemler R 1993 From cadherins to catenins: cytoplasmic protein interaction and regulation of cell adhesion. Trends Genet 9:317–321
Kessel M, Gruss P 1990 Murine developmental control genes. Science 249:374-379
Kessel M, Gruss P 1991 Homeotic transformations of murine vertebrae and concomitant alteration of Hox codes induced by retinoic acid. Cell 67:89-104
Keynes R, Lumsden A 1990 Segmentation and the origin of regional diversity in the vertebrate central nervous system. Neuron 4:1-9
Kilgour E, Gout I, Anderson NG 1996 Requirement for phosphoinositide 3-OH kinase in growth hormone signalling to the mitogen-activated protein kinase and p70s6k pathways. Biochem J 315:517-522
236
Kim SH, Li Z, Sacks DB 2000 E-cadherin-mediated cell-cell attachment activates Cdc42. J Biol Chem 275:36999-37005
Kissinger CR, Liu BS, Martin-Blanco E, Kornberg TB, Pabo CO 1990 Crystal structure of an engrailed homeodomain-DNA complex at 2.8 A resolution: a framework for understanding homeodomain-DNA interactions. Cell 63:579-590
Kleinberg DL, Ruan W, Catanese V, Newman CB, Feldman M 1990 Non-lactogenic effects of growth hormone on growth and insulin-like growth factor-I messenger ribonucleic acid of rat mammary gland. Endocrinology 126:3274-3276
Kleinberg DL 1998 Role of IGF-I in normal mammary development. Breast Cancer Res Treat 47:201-208
Knoepfler PS, Kamps MP 1997 The highest affinity DNA element bound by Pbx complexes in t(1;19) leukemic cells fails to mediate cooperative DNA-binding or cooperative transactivation by E2a-Pbx1 and class I Hox proteins - evidence for selective targetting of E2a-Pbx1 to a subset of Pbx-recognition elements. Oncogene 14:2521-2531
Koch AW, Pokutta S, Lustig A 1997 Calcium binding and homoassociation of E-cadherin domains. Engel J Biochemistry 36:7697-7705
Kok YJ, van der Maarel SM, Bitner-Glindzicz M, Huber I, Monaco AP, Malcolm S, Pembrey ME, Ropers HH, Cremers FP 1995 Association between X-linked mixed deafness and mutations in the POU domain gene POU3F4. Science 267:685-688
Kondo T, Dolle P, Zakany J, Duboule D 1996 Function of posterior HoxD genes in the morphogenesis of the anal sphincter. Development 122:2651-2659
Kongsuwan K, Allen J, Adams JM 1989 Expression of Hox-2.4 homeobox gene directed by proviral insertion in a myeloid leukemia. Nucleic Acids Res 17:1881-1892
Kovacs EM, Ali RG, McCormack AJ, Yap AS 2002 E-cadherin homophilic ligation directly signals through Rac and phosphatidylinositol 3-kinase to regulate adhesive contacts. J Biol Chem 277:6708-6718
Kroon E, Krosl J, Thorsteinsdottir U, Baban S, Buchberg AM, Sauvageau G 1998 Hoxa9 transforms primary bone marrow cells through specific collaboration with Meis1a but not Pbx1b. EMBO J 17:3714-3725
Krosl J, Baban S, Krosl G, Rozenfeld S, Largman C, Sauvageau G 1998 Cellular proliferation and transformation induced by HOXB4 and HOXB3 proteins involves cooperation with PBX1. Oncogene 16:3403-3412
Krumlauf R 1994 Hox genes in vertebrate development. Cell 70:191-201
Krumlauf R 1993 Hox genes and pattern formation in the branchial region of the vertebrate head. Trends Genet 9:106-112
237
Kuhn K, Streit B, Schierwater B 1996 Homeobox genes in the cnidarian Eleutheria dichotoma: evolutionary implications for the origin of Antennapedia-class (HOM/Hox) genes. Mol Phylogenet Evol 6:30-38
Kurland JF, Kodym R, Story MD, Spurgers KB, McDonnell TJ, Meyn RE 2001 NF-kappaB1 (p50) homodimers contribute to transcription of the bcl-2 oncogene. J Biol Chem 276:45380- 45386
Kuroda S, Fukata M, Fujii K, Nakamura T, Izawa I, Kaibuchi K 1997 Regulation of cell-cell adhesion of MDCK cells by Cdc42 and Rac1 small GTPases. Biochem Biophys Res Commun 240:430-435
Kyba M, Perlingeiro RC, Daley GQ 2002 HoxB4 confers definitive lymphoid-myeloid engraftment potential on embryonic stem cell and yolk sac hematopoietic progenitors. Cell 109:29-37
Land H, Parada LF, Weinberg RA 1983 Tumorigenic conversion of primary embryo fibroblasts requires at least two cooperating oncogenes. Nature 304:596-602
Lawrence HJ, Helgason CD, Sauvageau G, Fong S, Izon DJ, Humphries RK, Largman C 1997 Mice bearing a targeted interruption of the homeobox gene HOXA9 have defects in myeloid, erythroid, and lymphoid hematopoiesis. Blood 89:1922-1930
Lawrence PA 1992 The making of a Fly. Oxford, England: Blackwell Scientific Publications
Le Mouellic H, Lallemand Y, Brulet P 1992 Homeosis in the mouse induced by a null mutation in the Hox-3.1 gene. Cell 69:251-264
Le Roith D, Bondy C, Yakar S, Liu JL, Butler A 2001 The somatomedin hypothesis: 2001. Endocr Rev 22:53-74
Leaman DW, Pisharody S, Flickinger TW, Commane MA, Schlessinger J, Kerr IM, Levy DE, Stark GR 1996 Roles of JAKs in activation of STATs and stimulation of c-fos gene expression by epidermal growth factor. Mol Cell Biol 16:369–375
Levine M, Hoey T 1988 Homeobox proteins as sequence-specific transcription factors. Cell 55:537-540
Li CH, Evans HM, Simpson ME 1945 Isolation and properties of the anterior hypophysical growth hormone. J Biol Chem 159:353–366
Liang Z, Biggin MD 1998 Eve and ftz regulate a wide array of genes in blastoderm embryos: the selector homeoproteins directly or indirectly regulate most genes in Drosophila. Development 125:4471-4482
Liao J, Hodge C, Meyer D, Ho PS, Rosenspire K, Schwartz J 1997 Growth hormone regulates ternary complex factors and serum response factor associated with the c-fos serum response element. J Biol Chem 272:25951-25958
238
Lichty BD, Ackland-Snow J, Noble L, Kamel-Reid S, Dube ID 1995 Dysregulation of HOX11 by chromosome translocations in T-cell acute lymphoblastic leukemia: a paradigm for homeobox gene involvement in human cancer. Leuk Lymphoma 16:209-215
Lincecum JM, Fannon A, Song K, Wang Y, Sassoon DA 1998 Msh homeobox genes regulates cadherin-mediated cell adhesion and cell-cell sorting. J Cell Biochem 70:22-28
Lincoln DT, Temmin L, al-Jarallah MA, Mathew TC, Dashti H 1995 Primary Ki-1 lymphoma of the skin: expression of growth hormone receptors. Nutrition 11(5 Suppl):627-631
Lincoln DT, Water MJ, Breipohl W, Sinowatz F, Lobie PE 1990 Growth hormone receptors expression in the proliferating rat mammary gland. Acta Histochem Suppl 40:47-49
Linzer DI, Lee SJ, Ogren L, Talamantes F, Nathans D 1985 Identification of proliferin mRNA and protein in mouse placenta. Proc Natl Acad Sci USA 82:4356–4359
Linzer DI, Nathans D 1984 Nucleotide sequence of a growthrelated mRNA encoding a member of the prolactin-growth hormone family. Proc Natl Acad Sci USA 81:4255–4259
Lippman ME, Dickson RB, Gelmann EP, Rosen N, Knabbe C, Bates S, Bronzert D, Huff K, Kasid A 1987 Growth regulation of human breast carcinoma occurs through regulated growth factor secretion. J Cell Biochem 35:1-16
Liu H, Bird RC 1998 Characterization of the enhancer-like okadaic acid response element region of the cyclin-dependent kinase 1 (p34cdc2) promoter. Biochem Biophys Res Commun 246:696- 702
Lobie PE, Sadir R, Graichen R, Mertani HC, Morel G 1999 Caveolar internalization of growth hormone. Exp Cell Res 246:47-55
Lobie PE, Zhu T, Graichen R, Goh EL 2000 Growth hormone, insulin-like growth factor I and the CNS: localization, function and mechanism of action. Growth Horm IGF Res 10 Suppl B:S51-56
Logan C, Wizenmann A, Drescher U, Monschau B, Bonhoeffer F, Lumsden A.1996 Rostral optic tectum acquires caudal characteristics following ectopic engrailed expression. Curr Biol 6:1006- 1014
Lu M, Gong ZY, Shen WF, Ho AD 1991 The tcl-3 proto-oncogene altered by chromosomal translocation in T-cell leukemia codes for a homeobox protein. EMBO J 10:2905-2910
Lu MF, Pressman C, Dyer R, Johnson RL, Martin JF 1999 Function of Rieger syndrome gene in left-right asymmetry and craniofacial development. Nature 401:276-278
Lumsden A, Krumlauf R 1996 Patterning the vertebrate neuraxis. Science 274:1109-1115
MacKenzie S, Fleming I, Houslay MD, Anderson NG, Kilgour E 1997 Growth hormone and phorbol esters require specific protein kinase C isoforms to activate mitogen-activated protein kinases in 3T3-F442A cells. Biochem J 324:159–165
239
Madsen K, Friberg U, Roos P, Eden S, Isaksson O 1983 Growth hormone stimulates the proliferation of cultured chondrocytes from rabbit ear and rat rib growth cartilage. Nature 304:545-547
Malicki J, Schughart K, McGinnis W 1990 Mouse Hox-2.2 specifies thoracic segmental identity in Drosophila embryos and larvae. Cell 63:961-967
Manley NR, Capecchi MR 1995 The role of Hoxa-3 in mouse thymus and thyroid development. Development 121:1989-2003
Mark M, Lufkin T, Vonesch JL, Ruberte E, Olivo JC, Dolle P, Gorry P, Lumsden A, Chambon P 1993 Two rhombomeres are altered in Hoxa-1 mutant mice. Development 119:319-338
Mark M, Rijli FM, Chambon P 1997 Homeobox genes in embryogenesis and pathogenesis. Pediatr Res 42:421-429
Mathews CH, Detmer K, Lawrence HJ, Largman C 1993 Expression of the Hox 2.2 homeobox gene in murine embryonic epidermis. Differentiation 52:177-184
Maulbecker CC, Gruss P 1993 The oncogenic potential of Pax genes. EMBO J 12:2361-2367
Maulbecker CC, Gruss P 1993 The oncogenic potential of deregulated homeobox genes. Cell Growth Differ 4:431-441
Mbalaviele G, Dunstan CR, Sasaki A., Williams PJ, Mundy GR, Yoneda T 1996 E-cadherin expression in human breast cancer cells suppresses the development of osteolytic bone metastases in an experimental metastasis model. Cancer Res 56: 4063–4070
McFadden TB, Daniel TE, Akers RM 1990 Effects of plane of nutrition, growth hormone and unsaturated fat on growth hormone, insulin and prolactin receptors in prepubertal lambs. J Anim Sci 68:3180-3189
McGinnis W, Krumlauf R 1992 Homeobox genes and axial patterning. Cell 68:283-302
McIntosh GG, Anderson JJ, Milton I, Steward M, Parr AH, Thomas MD, Henry JA., Angus B, Lennard TW, Horne CH 1995 Determination of the prognostic value of cyclin D1 overexpression in breast cancer. Oncogene 11:885-891
Meigs TE, Fields TA, McKee DD, Casey PJ 2001 Interaction of Galpha 12 and Galpha 13 with the cytoplasmic domain of cadherin provides a mechanism for beta -catenin release. Proc Natl Acad Sci U S A 98:519-524
Mertani HC, Zhu T, Goh EL, Lee KO, Morel G, Lobie PE 2001 Autocrine human growth hormone (hGH) regulation of human mammary carcinoma cell gene expression. Identification of CHOP as a mediator of hGH-stimulated human mammary carcinoma cell survival. J Biol Chem 276:21464-21475
Meyer DJ, Campbell GS, Cochran BH, Argetsinger LS, Larner AC, Finbloom DS, Carter-Su C, Schwartz J 1994 Growth hormone induces a DNA binding factor related to the interferon- stimulated 91-kDa transcription factor. J Biol Chem 269:4701-4704
240
Miller JR, Moon RT 1996 Signal transduction through beta-catenin and specification of cell fate during embryogenesis. Genes Dev 10:2527-39
Miller WL, Eberhardt NL 1983 Structure and evolution of the growth hormone gene family. Endocr Rev 4:97-130
Mol JA, Henzen-Logmans SC, Hageman P, Misdorp W, Blankenstein MA, Rijnberk A 1995 Expression of the gene encoding growth hormone in the human mammary gland. J Clin Endocrinol Metab 80:3094-3096
Mol JA, van Garderen E, Selman PJ, Wolfswinkel J, Rijinberk A, Rutteman GR 1995 Growth hormone mRNA in mammary gland tumors of dogs and cats. J Clin Invest 95:2028-2034
Moller C, Hansson A, Enberg B, Lobie PE, Norstedt G 1992 Growth hormone (GH) induction of tyrosine phosphorylation and activation of mitogen-activated protein kinases in cells transfected with rat GH receptor cDNA . J Biol Chem 267: 23403–23408
Morikawa M, Nixon T, Green H 1982 Growth hormone and the adipose conversion of 3T3 cells. Cell 29:783-789
Mortlock DP, Innis JW 1997 Mutation of HOX A13 in hand-foot-genital syndrome. Nat Genet 15: 179-180
Moutoussamy S, Kelly PA, Finidori J 1998 Growth-hormone-receptor and cytokine-receptor- family signaling. Eur J Biochem 255:1-11
Muragaki Y, Mundlos S, Upton J, Olsen BR 1996 Altered growth and branching patterns in synpolydactyly caused by mutations in HOX D13. Science 272: 548-551
Myers C, Charboneau A, Boudreau N 2000 Homeobox B3 promotes capillary morphogenesis and angiogenesis. J Cell Biol 148:343-351
Nabarro JD 1987 Acromegaly. Clin Endocrinol (Oxf) 26:481-512
Nagar B, Overduin M, Ikura M, Rini JM 1996 Structural basis of calcium-induced E-cadherin rigidification and dimerization. Nature 380:360-364
Nakagawa M, Fukata M, Yamaga M, Itoh N, Kaibuchi K 2001 Recruitment and activation of Rac1 by the formation of E-cadherin-mediated cell-cell adhesion sites. J Cell Sci 114:1829-1838
Nakamura T, Jenkins NA, Copeland NG 1996 Identification of a new family of Pbx-related homeobox genes. Oncogene 13:2235-2242
Nakamura T, Largaespada DA, Lee MP, Johnson LA, Ohyashiki K, Toyama K, Chen SJ, Willman CL, Chen IM, Feinberg AP, Jenkins NA, Copeland NG, Shaughnessy JD 1996 Fusion of the nucleoporin gene NUP98 to HOXA9 by the chromosome translocation t(7;11)(p15;p15) in human myeloid leukaemia. Nat Genet 12:154-158
Neubauer H, Cumano A, Muller M, Wu H, Huffstadt U, Pfeffer K 1998 Jak2 deficiency defines an essential developmental checkpoint in definitive hematopoiesis. Cell 93:397-409
241
Ng ST, Zhou J, Adesanya OO, Wang J, LeRoith D, Bondy CA 1997 Growth hormone treatment induces mammary gland hyperplasia in aging primates. Nat Med 3:1141-1144
Nicoll CS, Mayer GL, Russell SM 1986 Structural features of prolactins and growth hormones that can be related to their biological properties. Endocr Rev 7:169–203
Noe V, Bruyneel E, Mareel M, Bracke M 2000 The E-cadherin/catenin complex in invasion: the role of ectodomain shedding. In: W.G. Jiang and R.E. Mansel, Editors, Cancer metastasis, molecular and cellular mechanisms and clinical intervention, Kluwer Academic Publishers, Dordrecht: 73–119
Noe V, Willems J, Vandederckhove J, van Roy F, Bruyneel E, Mareel M 1999 Inhibition of ahdeison an dinduction of epithelial cell invasion by HAV-containing E-cadherin specific peptides. J Cell Sci 112:127–135
Noel A, Simon N, Raus J, Foidart JM 1992 Basement membrane components (matrigel) promote the tumorigenicity of human breast adenocarcinoma MCF7 cells and provide an in vivo model to assess the responsiveness of cells to estrogen. Biochem Pharmacol 43:1263-1267
Noren NK, Niessen CM, Gumbiner BM, Burridge K 2001 Cadherin engagement regulates Rho family GTPases. J Biol Chem 276:33305-33308
Noren NK, Liu BP, Burridge K, Kreft B 2000 p120 catenin regulates the actin cytoskeleton via Rho family GTPases. J Cell Biol 150:567-580
Okazaki K, Sagata N 1995 The Mos/MAP kinase pathway stabilizes c-Fos by phosphorylation and augments its transforming activity in NIH 3T3 cells. EMBO J 14:5048-5059
Ono M, Takayama Y, Rand-Weaver M, Sakata S, Yasunaga T, Noso T, Kawauchi H 1990 cDNA cloning of somatolactin, a pituitary protein related to growth hormone and prolactin. Proc Natl Acad Sci USA 87:4330–4334
Orsulic S, Kemler R 2000 Expression of Eph receptors and ephrins is differentially regulated by E-cadherin. J Cell Sci 113:1793–1802
Osborne CK 1998 Steroid hormone receptors in breast cancer management. Breast Cancer Res Treat 51:227-238
Osborne CK, Zhao H, Fuqua SA 2000 Selective Estrogen Receptor Modulators: Structure, Function, and Clinical Use J Clin Oncol 18:3172-3186
Otting G, Qian YQ, Billeter M, Muller M, Affolter M, Gehring WJ, Wuthrich K 1990 Protein-- DNA contacts in the structure of a homeodomain--DNA complex determined by nuclear magnetic resonance spectroscopy in solution. EMBO J 9:3085-3092
Overdiun M, Harvey TS, Bagby S, Tong KL, Yau P, Takeichi M, Ikura M 1995 Solutin structure of the epithelial cadherin domain responsible for selective cell adhesion. Science 267: 386–389
Palacios J, Benito N, Pizarro A, Limeres MA, Suarez A, Cano A, Gamallo C 1995 Relationship between erbb2 and E-cadherin expression in human breast-cancer. Virch Arch 427: 259–263
242
Palacios J, Benito N, Pizarro A, Suarez A, Espada J, Cano A, Gamallo C 1995 Anomalous expression of P-cadherin in breast-carcinoma––correlation with E-cadherin expression, pathological features. Am J Pathol 146:605–612
Pantelouris EM, Hair J 1970 Thymic dysgenesis in nude (nu/nu) mice. J Embryol Exp Morph 24:615-623
Pantelouris EM 1968 Absence of thymus in a mouse mutant. Nature 217:370-371
Parcy F, Nilsson O, Bish MA, Lee I, Weigel D 1998 A genetic framework for floral patterning. Nature 395:561-566
Parganas E, Wang D, Stravopodis D, Topham DJ, Marine JC, Teglund S, Vanin EF, Bodner S, Colamonici OR, van Deursen JM, Grosveld G, Ihle JN 1998 Jak2 is essential for signaling through a variety of cytokine receptors. Cell 93:385-395
Park SH, Liu X, Hennighausen L, Davey HW, Waxman DJ 1999 Distinctive roles of STAT5a and STAT5b in sexual dimorphism of hepatic P450 gene expression. Impact of STAT5a gene disruption. J Biol Chem 274:7421–7430
Pavelic K, Gall-Troselj K 2001 Recent advances in molecular genetics of breast cancer. J Mol Med 79:566-573
Pearson G, Robinson F, Beers Gibson T, Xu BE, Karandikar M, Berman K, Cobb MH 2001 Mitogen-activated protein (MAP) kinase pathways: regulation and physiological functions. Endocr Rev 22:153-183
Pece S, Chiariello M, Murga C, Gutkind JS 1999 Activation of the protein kinase Akt/PKB by the formation of E-cadherin-mediated cell-cell junctions. Evidence for the association of phosphatidylinositol 3-kinase with the E-cadherin adhesion complex. J Biol Chem 274:19347- 19351
Pegram M, Slamon D 2000 Biological rationale for HER2/neu (c-erbB2) as a target for monoclonal antibody therapy. Semin Oncol 27:13-19
Peifer M 1996 Beta-catenin as oncogene: the smoking gun. Science 275:1752-1753
Pellegrini M, Mansouri A, Simeone A, Boncinelli E, Gruss P 1996 Dentate gyrus formation requires Emx2. Development 122:3893-3898
Peluso JJ, Pappalardo A, Trolice MP 1996 N-cadherin-mediated cell contact inhibits granulosa cell apoptosis in a progesterone-independent manner. Endocrinology 137:1196-1203
Perkins A, Kongsuwan K, Visvader J, Adams JM, Cory S 1990 Homeobox gene expression plus autocrine growth factor production elicits myeloid leukemia. Proc Natl Acad Sci U S A 87:8398- 8402
Perl AK, Wilgenbus P, Dahl U, Semb H, Christofori G 1998 A causal role for E-cadherin in the transition from adenoma to carcinoma. Nature 392:190-193
243
Peverali FA, D'Esposito M, Acampora D, Bunone G, Negri M, Faiella A, Stornaiuolo A, Pannese M, Migliaccio E, Simeone A 1990 Expression of HOX homeogenes in human neuroblastoma cell culture lines. Differentiation 45:61-69
Peyrat JP, Bonneterre J, Hecquet B, Vennin P, Louchez MM, Fournier C, Lefebvre J, Demaille A 1993 Plasma insulin-like growth factor-1 (IGF-1) concentrations in human breast cancer. Eur J Cancer 29:492-497
Peyssonnaux C, Eychene A 2001 The Raf/MEK/ERK pathway: new concepts of activation. Biol Cell 93:53-62
Pfeffer LM, Mullersman JE, Pfeffer SR, Murti A, Shi W, Yang CH 1997 STAT3 as an adapter to couple phosphatidylinositol 3-kinase to the IFNAR1 chain of the type I interferon receptor. Science 276:1418–1420
Phillips RM, Bibby MC, Double JA 1990 A critical appraisal of the predictive value of in vitro chemosensitivity assays. J Natl Cancer Inst 82:1457-1468
Phippard DJ, Weber-Hall SJ, Sharpe PT 1996 Regulation of Msx-1, Msx-2, Bmp-2 and Bmp-4 during foetal and postnatal mammary gland development. Development 122:2729–2737
Pircher TJ, Petersen H, Gustafsson JA, Haldosen LA 1999 Extracellular signal-regulated kinase (ERK) interacts with signal transducer and activator of transcription (STAT) 5a. Mol Endocrinol 13:555-565
Pokutta S, Herrenknecht K, Kemler R, Engel J 1994 Conformational changes of the recombinant extracellular domain of E-cadherin upon calcium bidning. Eur J Biochem 223:1019–1026
Pollak M, Blouin MJ, Zhang JC, Kopchick JJ 2001 Reduced mammary gland carcinogenesis in transgenic mice expressing a growth hormone antagonist. Br J Cancer 85:428-430
Povlsen CO, Jacobsen GK 1975 Chemotherapy of a human malignant melanoma transplanted in the nude mouse. Cancer Res 35:2790-2796
Pretlow TG, Delmoro CM, Dilley GG, Spadafora CG, Pretlow TP 1991 Transplantation of human prostatic carcinoma into nude mice in Matrigel. Cancer Res 51:3814-3817
Provost E, Rimm DL 1999 Controversies at the cytoplasmic face of the cadherin-based adhesion complex. Curr Opin Cell Bio 11:567-572
Purup S, Sejrsen K, Foldager J, Akers RM 1993 Effect of exogenous bovine growth hormone and ovariectomy on prepubertal mammary growth, serum hormones and acute in-vitro proliferative response of mammary explants from Holstein heifers. J Endocrinol 139:19-26
Quaggin SE, Yeger H, Igarashi P 1997 Antisense oligonucleotides to Cux-1, a Cut-related homobox gene, cause increased apoptosis in mouse embryonic kidney cultures. J Clin Invest 99:718-724
Raccurt M, Lobie PE, Moudilou E, Recher S, Garcia-Caballero T, Frappart L, Morel G, Mertani HC 2002 High stromal and epithelial human gh gene expression is associated with proliferative disorders of the mammary gland. J Endocrinol 175:307-318
244
Raff RA 1996 The Shape of Life: Genes, Development and the Evolution of Animal Form. Chicago, USA: The University of Chicago Press
Raman V, Martensen SA, Reisman D, Evron E, Odenwald WF, Jaffee E, Marks J, Sukumar S 2000 Compromised HOX A5 function can limit p53 expression in humn breast tumors. Nature 405:974-978
Ramirez-Solis R, Zheng H, Whiting J, Krumlauf R, Bradley A 1993 Hoxb-4 (Hox-2.6) mutant mice show homeotic transformation of a cervical vertebra and defects in the closure of the sternal rudiments. Cell 73:279-294
Redline RW, Hudock P, MacFee M, Patterson P 1994 Expression of AbdB-type homeobox genes in human tumors. Lab Invest 71:663-670
Reed JC 1998 Bcl-2 family proteins. Oncogene 17:3225-3236
Reif K, Nobes CD, Thomas G, Hall A, Cantrell DA 1996 Phosphatidylinositol 3-kinase signals activate a selective subset of Rac/Rho-dependent effector pathways. Curr Biol 6:1445-1455
Rhodes SJ, DiMattia GE, Rosenfeld MG 1994 Transcriptional mechanisms in anterior pituitary cell differentiation. Curr Opin Genet Dev 4:709-717
Richelsen B, Pedersen SB, Borglum JD, Moller-Pedersen T, Jorgensen J, Jorgensen JO 1994 Growth hormone treatment of obese women for 5 wk: effect on body composition and adipose tissue LPL activity. Am J Physiol 266:E211-216
Ridderstrale M, Tornqvist H 1994 PI-3-kinase inhibitor Wortmannin blocks the insulin-like effects of growth hormone in isolated rat adipocytes. Biochem Biophys Res Commun 203:306- 310
Ridley AJ, Comoglio PM, Hall A. 1995 Regulation of scatter factor/hepatocyte growth factor responses by Ras, Rac, and Rho in MDCK cells. Mol Cell Biol 15:1110-1122
Rijli FM, Matyas R, Pellegrini M, Dierich A, Gruss P, Dolle P, Chambon P 1995 Cryptorchidism and homeotic transformations of spinal nerves and vertebrae in Hoxa-10 mutant mice. Proc Natl Acad Sci U S A 92:8185-8189
Rimm DL, Morrow JS 1994 Molecular cloning of human E-cadherin suggests a novel subdivision of the cadherin superfamily. Biochem Biophys Res Commun 200:1754–1761
Rivera VM, Miranti CK, Misra RP, Ginty DD, Chen RH, Blenis J, Greenberg ME 1993 A growth factor-induced kinase phosphorylates the serum response factor at a site that regulates its DNA- binding activity. Mol Cell Biol 13:6260-6273
Robinson GW, Karpf AB, Kratochwil K 1999 Regulation of mammary gland development by tissue interaction. J Mammary Gland Biol Neoplasia 4:9–19
Roby KF, Deb S, Gibori G, Szpirer C, Levan G, Kwok SC, Soares MJ 1993 Decidual prolactin- related protein: Identification, molecular cloning, and characterization. J Biol Chem 268:3136– 3142
245
Rodig SJ, Meraz MA, White JM, Lampe PA, Riley JK, Arthur CD, King KL, Sheehan KC, Yin L, Pennica D, Johnson EM Jr, Schreiber RD 1998 Disruption of the Jak1 gene demonstrates obligatory and nonredundant roles of the Jaks in cytokine-induced biologic responses. Cell 93:373-383
Roshan SY 1999 The endocrine society Annual meeting Abstract P2-122
Ruley HE 1983 Adenovirus early region 1A enables viral and cellular transforming genes to transform primary cells in culture. Nature 304:602-606
Russell-Jones DL, Weissberger AJ, Bowes SB, Kelly JM, Thomason M, Umpleby AM, Jones RH, Sonksen PH 1993 The effects of growth hormone on protein metabolism in adult growth hormone deficient patients. Clin Endocrinol (Oxf) 38:427-431
Rygaard J and Povlsen CO 1969 Heterotransplantation of a human malignant tumour to “nude” mice. Acta Pathol Microbiol Scand 77:758-760
Saharinen P, Takaluoma K, Silvennoinen O 2000 Regulation of the Jak2 tyrosine kinase by its pseudokinase domain. Mol Cell Biol 20:3387–3395
Sandstedt J, Tornell J, Norjavaara E, Isaksson OG, Ohlsson C 1996 Elevated levels of growth hormone increase bone mineral content in normal young mice, but not in ovariectomized mice. Endocrinology 137:3368-3374
Satokata I, Benson G, Maas R 1995 Sexually dimorphic sterility phenotypes in Hoxa10-deficient mice. Nature 374:460-463
Schally AV, Comaru-Schally A M, Nagy A, Kovacs M, Szepeshazi K, Plonowski A, Varga JL, Halmos G 2001 Hypothalamic hormones and cancer. Front Neuroendocrinol 22:248-291
Schmidt A, Hall A 2002 Guanine nucleotide exchange factors for Rho GTPases: turning on the switch. Genes Dev 16:587-609
Schnabel CA, Jacobs Y, Cleary ML 2000 HoxA9-mediated immortalization of myeloid progenitors requires functional interactions with TALE cofactors Pbx and Meis. Oncogene 19:608-616
Scott GA 1993 Goldsmith LA. Homeobox genes and skin development: a review. J Invest Dermatol 101:3-8
Scott MP 1992 Vertebrate homeobox gene nomenclature. Cell 71:551-553
Scott MP 1997 Hox genes, arms and the man. Nature 15:117-118
Sefton M., Sanchez S, Nieto MA. 1998 Conserved and divergent roles for members of the snail family of transcription factors in chick and mouse embryo. Development 125:3111–3121
Seger R, Krebs EG 1995 The MAPK signaling cascade. FASEB J 9:726-735
246
Sejrsen K, Foldager J, Sorensen MT, Akers RM, Bauman DE 1986 Effect of exogenous bovine somatotropin on pubertal mammary development in heifers. J Dairy Sci 69:1528-1535
Semina EV, Reiter R, Leysens NJ, Alward WL, Small KW, Datson NA, Siegel-Bartelt J, Bierke- Nelson D, Bitoun P, Zabel BU, Carey JC, Murray JC 1996 Cloning and characterization of a novel bicoid-related homeobox transcription factor gene, RIEG, involved in Rieger syndrome. Nat Genet 14:392-399
Sesmilo G, Fairfield WP, Katznelson L, Pulaski K, Freda PU, Bonert V, Dimaraki E, Stavrou S, Vance ML, Hayden D, Klibanski A 2002 Cardiovascular risk factors in Acromegaly before and after normalization of serum IGF-I levels with the GH antagonist pegvisomant. J Clin Endocrinol Metab 87:1692-1699
Shapiro L, Fannon AM, Kwong PD, Thompson A, Lehmann MS, Grubel G, Legrand JF, Alsnielsen J, Coleman DR, Hendrickson WA 1995 Structural basis of cell–cell adhesion by cadherins. Nature 374: 327–337
Sharpe PT 1995 Homeobox genes and orofacial development. Connect Tissue Res 32:17-25
Shigetani Y, Funahashi JI, Nakamura H 1997 En-2 regulates the expression of the ligands for Eph type tyrosine kinases in chick embryonic tectum. Neurosci Res 27:211-217
Shimamoto T, Nakamura S, Bollekens J, Ruddle FH, Takeshita K 1997 Inhibition of DLX-7 homeobox gene causes decreased expression of GATA-1 and c-myc genes and apoptosis. Proc Natl Acad Sci USA 94:3245-3249
Shuai K 1997 The STAT family of proteins in cytokine signaling. Prog Biophys Mol Biol. 71:405-422
Siitonen SM, Kononen JT, Helin HJ, Rantala IS, Holli KA, Isola JJ 1996 Reduced E-cadherin expression is associated with invasiveness and unfavorable prognosis in breast cancer. Am J Clin Pathol 105:394-402
Silva CM, Day RN, Weber MJ, Thorner MO 1993 Human growth hormone (GH) receptor is characterized as the 134-kilodalton tyrosine-phosphorylated protein activated by GH treatment in IM-9 cells. Endocrinology 133:2307–2312
Slack JM 1985 Homoeotic transformations in man: implications for the mechanism of embryonic development and for the organization of epithelia. J Theor Biol 114:463-490
Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A., McGuire WL 1987 Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 235:177-182
Smith RC, Branellec D, Gorski DH, Guo K, Perlman H, Dedieu JF, Pastore C, Mahfoudi A, Denefle P, Isner JM, Walsh K 1997 p21CIP1-mediated inhibition of cell proliferation by overexpression of the gax homeodomain gene. Genes Dev 11:1674-1689
Song K, Wang Y, Sassoon D 1992 Expression of Hox-7.1 in myoblasts inhibits terminal differentiation and induces cell transformation. Nature 360:477-481
247
Sorensen PH, Triche TJ 1996 Gene fusions encoding chimaeric transcription factors in solid tumours. Semin Cancer Biol 7:3-14
Sotiropoulos A, Perrot-Applanat M, Dinerstein H, Pallier A, Postel-Vinay MC, Finidori J, Kelly PA. 1994 Distinct cytoplasmic regions of the growth hormone receptor are required for activation of JAK2, mitogen-activated protein kinase, and transcription. Endocrinology 135:1292-1298
Srebrow A, Friedmann Y, Ravanpay A, Daniel CW, Bissell MJ 1998 Expression of Hoxa-1 and Hoxb-7 is regulated by extracellular matrix-dependent signals in mammary epithelial cells. J Cell Biochem 69:377-391
Stein S, Fritsch R, Lemaire L, Kessel M 1996 Checklist: Vertebrate homeobox genes. Mech Dev 55:91-108
Steinberg MS, McNutt PM 1999 Cadherins and their connections: adhesion junctions have broader functions. Curr Opin Cell Biol 11:554-560
Streuli CH, Gilmore AP 1999 Adhesion-mediated signaling in the regulation of mammary epithelial cell survival. J Mammary Gland Biol Neoplasia 4:183–191
Strobl JS, Thomas MJ 1994 Human growth hormone. Pharmacol Rev 46:1-34
Stuart ET, Kioussi C, Gruss P 1994 Mammalian Pax genes. Annu Rev Genet 28:219-236
Stuart ET, Haffner R, Oren M, Gruss P 1995a Loss of p53 function through PAX-mediated transcription repression. EMBO J 14:5638-5645
Stuart ET, Kioussi C, Aguzzi A, Gruss P 1995b PAX 5 expression correlates with increasing malignancy in human astrocytomas. Clin Cancer Res 1:207-214
Studer M, Lumsden A, Ariza-McNaughton L, Bradley A, Krumlauf R 1996 Altered segmental identity and abnormal migration of motor neurons in mice lacking Hoxb-1. Nature 384:630-634
Symons M, Settleman J 2000 Rho family GTPases: more than simple switches. Trends Cell Biol 10:415-419
Takaishi K, Sasaki T, Kotani J, Nishioka H, Takai Y 1997 Regulation of cell-cell adhesion by Rac and Rho small G proteins in MDCK cells. J Cell Biol 139:1047-1059
Takeichi M 1991 Cadherin cell adhesion receptors as a morphogenetic regulator. Science 251: 1451–1455
Takeichi M 1995 Morphogenetic roles of classic cadherins. Curr Opin Cell Biol 7:619-627
Tanaka S, Hanafusa H 1998 Guanine-nucleotide exchange protein C3G activates JNK1 by a ras- independent mechanism. JNK1 activation inhibited by kinase negative forms of MLK3 and DLK mixed lineage kinases. J Biol Chem 273:1281-1284
Taneja R, Thisse B, Rijli FM, Thisse C, Bouillet P, Dolle P, Chambon P 1996 The expression pattern of the mouse receptor tyrosine kinase gene MDK1 is conserved through evolution and requires Hoxa-2 for rhombomere-specific expression in mouse embryos. Dev Biol 177:397-412
248
Tannenbaum GS 1991 Neuroendocrine control of growth hormone secretion. Acta Paediatr Scand Suppl 372:5–16
Teglund S, McKay C, Schuetz E, van Deursen JM, Stravopodis D, Wang D, Brown M, Bodner S, Grosveld G, Ihle JN 1998 Stat5a and Stat5b proteins have essential and nonessential, or redundant, roles in cytokine responses. Cell 93:841–850
Thesleff I, Nieminen P 1996 Tooth morphogenesis and cell differentiation. Curr Opin Cell Biol 8:844-850
Thomas MJ 1998 The molecular basis of growth hormone action. Growth Horm IGF Res 8:3-11
Tiberio C, Barba P, Magli MC, Arvelo F, Le Chevalier T, Poupon MF, Cillo C 1994 HOX gene expression in human small-cell lung cancers xenografted into nude mice. Int J Cancer 58:608-615
Tollet P, Hamberg M, Gustafsson JA, Mode A. 1995 Growth hormone signaling leading to CYP2C12 gene expression in rat hepatocytes involves phospholipase A2. J Biol Chem 270:12569-12577
Tornell J, Carlsson B, Pohjanen P, Wennbo H, Rymo L, Isaksson O 1992 High frequency of mammary adenocarcinomas in metallothionein promoter-human growth hormone transgenic mice created from two different strains of mice. J Steroid Biochem Mol Biol 43:237-242
Torres-Guzman JC, Dominguez A 1997 HOY1, a homeo gene required for hyphal formation in Yarrowia lipolytica. Mol Cell Biol 17:6283-6293
Treisman R 1996 Regulation of transcription by MAP kinase cascades. Curr Opin Cell Biol 8:205-215
Twigg SR, Wilkie AOM 1999 Charactrisation of the human snail (SNAI1) gene and exclusion as a major disease gene in craniosynostosis. Hum Genet 105:320–326
Uckun FM, Downing JR, Gunther R, Chelstrom LM, Finnegan D, Land VJ, Borowitz MJ, Carroll AJ, Crist WM 1993 Human t(1;19)(q23;p13) pre-B acute lymphoblastic leukemia in mice with severe combined immunodeficiency. Blood 81:3052-3062
Udy GB, Towers RP, Snell RG, Wilkins RJ, Park SH, Ram PA, Waxman DJ, Davey HW 1997 Requirement of STAT5b for sexual dimorphism of body growth rates and liver gene expression. Proc Natl Acad Sci USA 94:7239–7244
Urbanek P, Wang Z-Q, Fetka I, Wagnar EF, Busslinger M 1994 Complete block of early B cell differentiation and altered patterning of the posterior midbrain in mice lacking Pax5/BSAP. Cell 79:901-912 van Oostveen J, Bijl J, Raaphorst F, Walboomers J, Meijer C 1999 The role of homeobox genes in normal hematopoiesis and hematological malignancies. Leukemia 13:1675-1690
VanGilder JC, Goldenberg IS 1975 Hypophysectomy in metastatic breast cancer. Arch Surg 110:293-295
249
Vanhaesebroeck B, Leevers SJ, Ahmadi K, Timms J, Katso R, Driscoll PC, Woscholski R, Parker PJ, Waterfield MD 2001 Synthesis and function of 3-phosphorylated inositol lipids. Annu Rev Biochem 70:535-602
Vestweber D, Kemler R 1985 Identification of a putative cell adhesion domain of uvomorulin. EMBO J 4:3393-3398
Vider BZ, Zimber A, Hirsch D, Estlein D, Chastre E, Prevot S, Gespach C, Yaniv A, Gazit A 1997 Human colorectal carcinogenesis is associated with deregulation of homeobox gene expression. Biochem Biophys Res Commun 232:742-748
Vignais ML, Gilman M 1999 Distinct mechanisms of activation of Stat1 and Stat3 by platelet- derived growth factor receptor in a cell-free system. Mol Cell Biol 19:3727–3735
Vleminckx K, Vakaet L Jr, Mareel M, Fiers W, van Roy F 1991 Genetic manipulation of E- cadherin expression by epithelial tumor cells reveals an invasion suppressor role. Cell 66:107-119
Volberg T, Geiger B, Kartenbeck J, Franke WW 1986 Changes in membrane-microfilament interaction in intercellular adherens junctions upon removal of extracellular Ca2+ ions. J Cell Biol 102:1832-1842
Volk T, Cohen O, Geiger B 1987 Formation of heterotypic adherens-type junctions between L- CAM-containing liver cells and A-CAM-containing lens cells. Cell 50:987-994
Wabitsch M, Hauner H, Heinze E, Teller WM 1995 The role of growth hormone/insulin-like growth factors in adipocyte differentiation. Metabolism 44:45-49
Walden PD, Ruan W, Feldman M, Kleinberg DL 1998 Evidence that the mammary fat pad mediates the action of growth hormone in mammary gland development. Endocrinology 139:659- 662
Walter J, Dever CA, Biggin MD 1994 Two homeo domain proteins bind with similar specificity to a wide range of DNA sites in Drosophila embryos. Genes Dev 8:1678-1692
Wang X, Darus CJ, Xu BC, Kopchick JJ 1996 Identification of growth hormone receptor (GHR) tyrosine residues required for GHR phosphorylation and JAK2 and STAT5 activation. Mol Endocrinol 10:1249-1260
Wang YD, Wood WI 1995 Amino acids of the human growth hormone receptor that are required for proliferation and Jak-STAT signaling. Mol Endocrinol 9:303-311
Wang YZ, Wharton W, Garcia R, Kraker A, Jove R, Pledger WJ 2000 Activation of Stat3 preassembled with platelet-derived growth factor beta receptors requires Src kinase activity. Oncogene 19:2075–2085
Watton SJ, Downward 1999 Akt/PKB localisation and 3' phosphoinositide generation at sites of epithelial cell-matrix and cell-cell interaction. Curr Biol 9:433-436
Wen Z, Darnell JE 1997 Mapping of Stat3 serine phosphorylation to a single residue (727) and evidence that serine phosphorylation has no influence on DNA binding of Stat1 and Stat3. Nucleic Acids Res 25:2062-2067
250
Wewer UM, Mercurio AM, Chung SY, Albrechtsen R 1990 Deoxyribonucleic-binding homeobox proteins are augmented in human cancer. Lab Invest 63:447-454
White MF, Yenush L 1998 The IRS-signaling system: a network of docking proteins that mediate insulin and cytokine action. Curr Top Microbiol Immunol 228:179-208
Winkler U, Gottstein C, Schon G, Kapp U, Wolf J, Hansmann ML, Bohlen H, Thorpe P, Diehl V, Engert A 1994 Successful treatment of disseminated human Hodgkin's disease in SCID mice with deglycosylated ricin A-chain immunotoxins. Blood 83:466-475
Winston LA, Bertics PJ 1992 Growth hormone stimulates the tyrosine phosphorylation of 42- and 45-kDa ERK-related proteins. J Biol Chem 267: 4747–4751
Wolf RF, Heslin MJ, Newman E, Pearlstone DB, Gonenne A, Brennan MF 1992 Growth hormone and insulin combine to improve whole-body and skeletal muscle protein kinetics. Surgery 112:284-291
Wood TJ, Sliva D, Lobie PE, Pircher TJ, Gouilleux F, Wakao H, Gustafsson JA, Groner B, Norstedt G, Haldosen LA 1995 Mediation of growth hormone-dependent transcriptional activation by mammary gland factor/Stat 5. J Biol Chem 270: 9448-9453
Wood TJ, Sliva D, Lobie PE, Goullieux F, Mui AL, Groner B, Norstedt G, Haldosen LA 1997 Specificity of transcription enhancement via the STAT responsive element in the serine protease inhibitor 2.1 promoter. Mol Cell Endocrinol 130:69-81
Xu Y, Guo DF, Davidson M, Inagami T, Carpenter G 1997 Interaction of the adaptor protein Shc and the adhesion molecule cadherin. J Biol Chem 272:13463-13466
Yamauchi T, Kaburagi Y, Ueki K, Tsuji Y, Stark GR, Kerr IM, Tsushima T, Akanuma Y, Komuro I, Tobe K, Yazaki Y, Kadowaki T 1998 Growth hormone and prolactin stimulate tyrosine phosphorylation of insulin receptor substrate-1, -2, and -3, their association with p85 phosphatidylinositol 3-kinase (PI3-kinase), and concomitantly PI3-kinase activation via JAK2 kinase. J Biol Chem 273:15719-15726
Yamauchi T, Ueki K, Tobe K, Tamemoto H, Sekine N, Wada M, Honjo M, Takahashi M, Takahashi T, Hirai H, Tsushima T, Akanuma Y, Fujita T, Komuro I, Yazaki Y, Kadowaki T 1998 Growth hormone-induced tyrosine phosphorylation of EGF receptor as an essential element leading to MAP kinase activation and gene expression. Endocr J 45 Suppl:S27-31
Yang XF, Beamer WG, Huynh H, Pollak M 1996 Reduced growth of human breast cancer xenografts in hosts homozygous for the lit mutation. Cancer Res 56:1509-1511
Yap AS, Brieher WM, Gumbiner BM 1997 Molecular and functional analysis of cadherin-based adherens junctions. Annu Rev Cell Dev Biol 13:119-146
Yap AS, Kovacs EM 2003 Direct cadherin-activated cell signaling: a view from the plasma membrane. J Cell Biol 160:11-16
Yenush L, White MF 1997 The IRS-signalling system during insulin and cytokine action. Bioessays 19:491-500
251
Yoshii S, Tanaka M, Otsuki Y, Wang DY, Guo RJ, Zhu Y, Takeda R, Hanai H, Kaneko E, Sugimura H 1999 alphaPIX nucleotide exchange factor is activated by interaction with phosphatidylinositol 3-kinase. Oncogene 18:5680-5690
Zantek ND, Azimi M, Fedor-Chaiken M, Wang B, Brackenbury R, Kinch MS 1999 E-cadherin regulates the function of the EphA2 receptor tyrosine kinase. Cell Growth Differ 10:629-638
Zelinski DP, Zantek ND, Stewart JC, Irizarry AR, Kinch MS 2001 EphA2 overexpression causes tumorigenesis of mammary epithelial cells Cancer Res 61:2301-2306
Zhao JJ, Lazzarini RA, Pick L 1993 The mouse Hox-1.3 gene is functionally equivalent to the Drosophila Sex combs reduced gene. Genes Dev 7:343-354
Zhong C, Kinch MS, Burridge K 1997 Rho-stimulated contractility contributes to the fibroblastic phenotype of Ras-transformed epithelial cells. Mol Biol Cell 8: 2329-2344
Zhou Q, Wulfkuhle J, Ouatas T, Fukushima P, Stetler-Stevenson M, Miller F R, Steeg PS 2000 Cyclin D1 overexpression in a model of human breast premalignancy: preferential stimulation of anchorage-independent but not anchorage-dependent growth is associated with increased cdk2 activity. Breast Cancer Res Trea. 59:27-39
Zhou Y, Xu BC, Maheshwari HG, He L, Reed M, Lozykowski M, Okada S, Cataldo L, Coschigamo K, Wagner TE, Baumann G, Kopchick JJ 1997 A mammalian model for Laron syndrome produced by targeted disruption of the mouse growth hormone receptor/binding protein gene (the Laron mouse). Proc Natl Acad Sci U S A 94:13215-13220
Zhu T, Goh EL, LeRoith D, Lobie PE 1998 Growth hormone stimulates the formation of a multiprotein signaling complex involving p130(Cas) and CrkII. Resultant activation of c-Jun N- terminal kinase/stress-activated protein kinase (JNK/SAPK). J Biol Chem 273:33864–33875
Zhu T, Lobie PE 2000 Janus kinase 2-dependent activation of p38 mitogen-activated protein kinase by growth hormone. Resultant transcriptional activation of ATF-2 and CHOP, cytoskeletal re-organization and mitogenesis. J Biol Chem 275:2103–2114
Zhu T, Goh EL, Graichen R, Ling L, Lobie PE 2001 Signal transduction via the growth hormone receptor. Cell Signal 13:599-616
Zschiesche W, Schonborn I, Behrens J, Herrenknecht K, Hartveit F, Lilleng P, Birchmeier W 1997 Expression of E-cadherin and catenins in invasive mammary carcinomas. Anticancer Res 17:561–567