1 HOXC8 Inhibits Androgen Receptor Signaling in Human Prostate Cancer

Author Manuscript Published OnlineFirst on November 2, 2010; DOI: 10.1158/1541-7786.MCR-10-0111 Author manuscripts havePublished been peer reviewed OnlineFirst and accepted on November for publication 2, 2010but have not yet been edited.

HOXC8 inhibits androgen receptor signaling in human prostate cancer cells by inhibiting SRC-3 recruitment to direct androgen target genes

S. Daddario Axlund*†, J.R. Lambert*, S.K. Nordeen*†

*Department of Pathology, †Program in Molecular Biology University of Colorado Denver, School of Medicine, Anschutz Medical Campus, 12800 E. 19th Ave., Aurora, CO, 80045, USA

KEYWORDS

AIB1; androgen receptor; homeobox; HOX; HOXC8; NCOA3; prostate cancer; SRC-3

ABBREVIATIONS

AIB1, amplified in breast cancer-1; ADT, androgen deprivation therapy; AR, androgen receptor; ARE, androgen response element; CBP, CREB binding protein, ChIP, chromatin immunoprecipitation; CoIP, co-immunoprecipitation; CRPC, castration- resistant prostate cancer; FBS, fetal bovine serum; GST, glutathione S-transferase; HAT, histone acetlytransferase; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; HDAC, histone deacetylace; HOX, homeobox; IP, immunoprecipitation; MMTV, mouse mammary tumor virus; PCa, prostate cancer; PBS, phosphate buffered saline; PSA, prostate-specific antigen; S.D., standard deviation; SDS, sodium dodecyl sulfate; SRC, steroid receptor coactivator.

ACKNOWLEDGMENTS

The authors thank Drs. Andrew Bradford and Heide Ford for critical reading of the manuscript. This work was supported by Prostate Cancer Research Program Predoctoral Fellowship W81XWH-06-1-0066 to S.D.A. from the Department of Defense and, in part, by NIH NCI CA84269 to S.K.N.

Downloaded from mcr.aacrjournals.org on September 26, 2021. © 2010 American Association for Cancer Research. Author Manuscript Published OnlineFirst on November 2, 2010; DOI: 10.1158/1541-7786.MCR-10-0111 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

ABSTRACT

HOX genes encode homeodomain-containing transcription factors critical to

development, differentiation and homeostasis. HOX gene dysregulation has been

implicated in a variety of cancers. Previously, we showed that a subset of genes of the

HOXC cluster is up-regulated in primary prostate tumors, lymph node metastases, and

malignant prostate cell lines. In the present study, we show that HOXC8 inhibits

androgen receptor (AR)-mediated gene induction in LNCaP prostate cancer (PCa) cells

and HPr-1 AR, a non-tumorigenic prostate epithelial cell line. Mechanistically, HOXC8

blocks the AR-dependent recruitment of the steroid receptor coactivators SRC-3 and CBP

to the androgen-regulated prostate specific antigen (PSA) gene enhancer and inhibits

histone acetylation of androgen-regulated genes. Inhibition of androgen induction by

HOXC8 is reversed upon expression of SRC-3, a member of the SRC/p160 steroid

receptor cofactor family. Co-immunoprecipitation studies demonstrate that HOXC8

expression inhibits the hormone-dependent interaction of AR and SRC-3. Finally,

HOXC8 expression increases invasion in HPr-1 AR non-tumorigenic cells. These data

suggest a complex role for HOXC8 in PCa, promoting invasiveness while at the same

time inhibiting AR-mediated gene induction at androgen response element (ARE)-

regulated genes associated with differentiated function of the prostate. A greater

understanding of HOXC8 actions in the prostate and its interactions with androgen

signaling pathways may elucidate mechanisms driving the onset and progression of PCa.

INTRODUCTION

Androgen ablation therapy for PCa exploits the characteristic androgen-

dependence of prostate epithelium for survival and growth and is the first line strategy for

treatment of surgically refractory or recurrent PCa. However, despite sustained androgen

ablation, the majority of men eventually cease responding and “castration-resistant

prostate cancer” (CRPC) claims the life of the patient. Thus, a major objective of PCa

research is to better understand the mechanisms underlying the transition from androgen-

dependence to a castration-resistant state. Previously, our laboratory reported that certain

HOX (homeobox) genes are overexpressed in primary and metastatic PCa (1). These

observations led us to explore a potential cross-talk between HOX genes and AR-

mediated signaling.

There are 39 human HOX genes, arranged in four chromosomal clusters

designated as HOXA, HOXB, HOXC and HOXD. HOX genes encode a large family of

homeodomain-containing transcription factors considered “master regulators” in

embryonic development because of their key involvement in body patterning, growth and

differentiation in vertebrates and invertebrates (2-5). As many of these processes critical

to development are also known to be dysregulated in carcinogenesis, it is not unexpected

then that aberrant HOX gene regulation has been implicated in a variety of cancers

including leukemias (6-12), cervical (13), colorectal (14), breast (15;16), brain (17), renal

(18), lung (19;20) and esophageal cancers (21).

In the androgen-responsive, normal mature prostate, little or no expression of

HOXC genes is seen but any of a subset of these genes (HOXC4, HOXC5, HOXC6, or

HOXC8) are upregulated in primary prostate tumors, lymph node metastases, and

malignant prostate cell lines (1). HOXC8 overexpression has been correlated with loss of

differentiation/higher Gleason grade in human PCa cell lines and tissues (22). siRNA-

mediated knockdown of HOXC8 and its transcriptional binding partner, PBX1, results in

a decrease in androgen-independent growth of DU-145 PCa cells (23). Loss of HOXC6

expression in PCa cells by siRNA induces apoptosis, suggesting a protective role for

HOXC6 in PCa cell survival (24). Our present studies show that HOXC8 overexpression

promotes invasiveness in a non-tumorigenic prostate epithelial model suggesting a role

for HOXC genes in cancer progression.

A link between androgen signaling and HOX gene expression is suggested by a

report indicating that a subset of HOX proteins bind to the steroid receptor coactivators

CBP and p300, inhibiting the histone acetyltransferase (HAT) activity of the coactivators

(25). A direct interaction of HOXB13 and AR has been reported (26;27). This

interaction can have positive and negative effects on AR- and HOXB13-regulated

transcription (26). In the present studies directed to understanding the consequences of

the overexpression of HOXC genes in prostate, we show that overexpression of HOXC8

inhibits androgen induction at promoters regulated by AREs. Investigation into the

mechanistic basis for this inhibition indicate that HOXC8 inhibits the loading of SRC-3, a

transcriptional coactivator critical in AR-mediated gene transcription, at direct androgen

target genes and that inhibition of AR-induced transcription is abrogated by SRC-3

overexpression. Significantly, SRC-3 amplification and/or overexpression has been

observed in many hormone-sensitive tumors, including prostate (28;29), breast (30),

endometrial (31), and ovarian cancer (32), as well as many non-steroid targeted tumors

such as pancreatic cancer (33;34), gastric cancer (35), colorectal carcinoma (36), and

hepatocellular carcinoma (37).

SRC-3 is expressed in PCa cell lines and prostate tumors, and its expression is

correlated with tumor grade and stage in PCa patients (38;39). SRC-3 is overexpressed

in tissue samples from PCa patients, and its overexpression correlates with PCa

proliferation and inversely correlates with apoptosis (29). Knockdown of SRC-3 by

siRNA results in decreased cell proliferation, delays in the G1-S transition, and increased

apoptosis in PCa cell lines as well as decreased tumor growth in nude mice (29). Other

studies show that SRC-3, along with the AR, is required for androgen-dependent and –

independent proliferation of PCa cells and for xenograft tumor growth (40). Further, AR

and SRC-3 are recruited to gene promoters involved in cell cycle control in an androgen-

independent manner (40). Taken together, these data suggest that SRC-3 is required for

PCa cell proliferation and survival, and together with the AR, plays a central role in

progression to CRPC.

The present work ties together 3 major players in PCa, AR, SRC-3, and HOXC8

and explores the consequences of their interaction. We also provide evidence implicating

HOXC8 overexpression in PCa progression and discuss the role of HOXC gene

overexpression in PCa despite its inhibitory actions on AR-mediated gene expression.

MATERIALS AND METHODS

Cell lines and culture- The human prostate cell lines were obtained from the

following laboratories: LNCaP (41;42), Dr. J. Horoszewicz (Roswell Park Memorial

Institute, Buffalo, NY); HPr-1-AR (43), Dr. K. Yamamoto (University of California San

Francisco, San Francisco, CA). LNCaP cells were maintained in RPMI 1640

supplemented with 10% FBS (Hyclone, Logan UT), penicillin G (100 U/ml), and

streptomycin (100 μg/ml). HPr-1-AR cells were maintained in keratinocyte serum-free

media (Invitrogen, Carlsbad, CA) supplemented with penicillin G (100 U/ml),

streptomycin (100 μg/ml), gentamicin (50 μg/ml), and the manufacturer provided

supplements, epidermal growth factor and bovine pituitary extract.

HOXC8 retrovirus construction and infection- pMSCV-IRES-GFP-HOXC8 was

constructed by standard subcloning techniques. HOXC8 cDNA was cloned into the

EcoRI site within the multiple cloning site of pMSCV-IRES-GFP (gift from Dr. James

DeGregori, University of Colorado Denver) and confirmed by sequencing. Virus

preparation and cell infection performed as previously reported (44). Expressing cell

lines were selected by sorting for GFP expression and represent many independent

integrants.

RNA isolation and PCR analysis- Total RNA was isolated using RNeasy Plus

(Qiagen, Valencia, CA) according to the manufacturer’s instructions. 1 μg total RNA

was reverse transcribed using MMLV reverse transcriptase (Promega, Madison, WI).

Real time reverse transcriptase-PCR for HOXC8, SRC-3, PSA and 18s was performed on

1/10 (HOXC8, SRC-3, PSA) or 1/20 (18s) of the synthesized cDNA using primer and

probe sets and TaqMan Fast Universal Master Mix (Applied Biosystems, Foster City,

CA). Amplification signals were detected with an ABIprism 7500 Fast Sequence

Detection system (Applied Biosystems, Foster City, CA). Fold change in expression was

calculated using the comparative Ct method (45) following confirmation that the

amplification efficiencies of the target (HOXC8, SRC-3 or PSA) and endogenous

reference control (18s) were approximately equal.

Luciferase Reporter Assays- Cells were seeded at 3 x 105 cells per well in 3mL

RPMI 1640 supplemented with 10% FBS in 6 well plates. 24 hours later, cells were co-

transfected with 2 μg probasin-, MMTV- or PSA- luciferase reporter constructs (46) (gift

from Dr. Carlos Perez-Stable, University of Miami School of Medicine) and pCMV6c-

HOXC8 expression plasmid and/or empty vector control pCMV6c to ensure that the

same amount of DNA was transfected into each well, using Mirus Bio TransIT LT-1

transfection reagent (Mirus Bio, Madison, WI). After 24 hours, cells were treated with

10nM R1881 (New England Nuclear, Boston, MA) in phenol red-free RPMI (Invitrogen,

Carlsbad, CA) supplemented with 5% dextran-coated charcoal treated FBS for 18 hours.

In all experiments the total amount of plasmid transfected was equalized by adding empty

expression vectors as necessary. Luminescence was normalized to total protein. For

coactivator co-transfection reporter assay (Fig. 4), the net induction (I) for each condition

was the normalized luciferase activity with hormone minus that without. The average

inhibition of hormone induction mediated by HOXC8 expression alone [1 – (IHOX/ICON) x

100] was ~58%. Percent control HOXC8 inhibition is the ratio of the inhibition of

androgen induction in HOXC8 + coactivator transfections to the inhibition with HOXC8

alone.

Western blot analysis- Cell extracts were separated on NuPAGE 4-12% Bis-Tris

gel (Invitrogen, Carlsbad, CA) and transferred to PVDF membrane. Following

incubation with antibody, signals were detected by enhanced chemiluminescence

(Perkin–Elmer, Waltham, MA) and autoradiography. Bands were quantified using Image

J software (NIH) and normalized to β-actin loading control.

Co-immunoprecipitation (co-IP)- 7 x 106 LNCaP were seeded on 15cm dishes in

phenol red-free RPMI 1640 supplemented with 5% dextran-coated charcoal-treated FBS

and grown to near confluency. Cells were treated with 10 nM R1881 or ethanol control

for 24 hours. Co-immunoprecipitation analysis was performed as previously described

by Gnanapragasam et al. (39) with an additional added step to preclear the lysate with

100 μl protein A agarose. Immune complexes were collected by centrifugation, washed,

and analyzed by Western blot as described.

Chromatin immunoprecipitation (ChIP)- Cells were grown to near confluency

in 15cm dishes in phenol red-free RPMI (Invitrogen, Carlsbad, CA) supplemented with

5% dextran-coated charcoal-treated FBS for 3-5 days. For histone acetylation analyses,

cells were transfected with 2 μg/mL MMTV-luciferase and 200 ng/mL pCMV6c-

HOXC8 or pCMV6c empty vector 18 hours before hormone treatment. Cells were

treated with 10 nM R1881 or vehicle control for 30 (SRC-3 and CBP ChIP) or 60

(histone acetylation ChIP) minutes. ChIP analysis was performed as previously

described by Lambert et al. (47). DNA was purified using QIAprep Spin Miniprep Kit

(Qiagen, Valencia, CA), and eluted in water. Control experiments using serial dilutions

of input DNA were performed to affirm that the PCR reactions were within the linear

range of amplification cycles. For histone acetylation ChIP experiments, PCR for

MMTV-luciferase was performed as previously described (47), and samples were

analyzed by agarose gel electrophoresis. For PSA ChIP experiments, real time reverse

transcriptase-PCR for PSA enhancer region was performed on 1/10 (PSA enhancer) or 8

1/20 (18s) of the samples using primer and probe set (PSA enhancer forward primer: 5’-

ACACCTTTTTTTTTCTGGATTGTTG-3’, PSA enhancer reverse primer 5’-GCCTGG

ATCTGAGAGAGATATCAT-3’, PSA enhancer probe: 5’-TGCAAGGAT

GCCTGCTTTACAAACATCC-3’) and TaqMan Fast Universal Master Mix (Applied

Biosystems, Foster City, CA).

Antibodies- HOXC8 mouse monoclonal antibody was used to detect HOXC8

protein (Covance, Princeton, NJ). For Western blot analysis, SRC-3 (AIB-1, NCoA3)

protein was detected using either mouse AIB-1 monoclonal antibody (BD Transduction

Laboratories, San Jose, CA) or NCoA3 N-16 goat polyclonal antibody (Santa Cruz

Biotechnology, Santa Cruz, CA); for ChIP and Co-IP experiments, SRC-3 was

immunoprecipitated using either mouse AIB-1 monoclonal antibody (BD Transduction

Laboratories, San Jose, CA) or NCoA3 N-16 goat polyclonal antibody (Santa Cruz

Biotechnology, Santa Cruz, CA). For Western blot analysis, AR protein was detected

using either rabbit N-20 polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz,

CA) or mouse monoclonal antibody (BD Biosciences, San Jose, CA). For chromatin

immunoprecipitation, acetylated histone H3 or H4 were immunoprecipitated using the

appropriate anti-acetylated histone antibody (Upstate Cell Signaling Solutions, Billerica,

MA). For Western blot analysis, β-actin was detected using mouse monoclonal IgG1

clone AC-15 antibody (Sigma, St. Louis, MO).

GST pull-down assay- HOXC8 cDNA was cloned into pGEX-4T1 (GE

Healthcare, Piscataway, NJ) and purified using the GST Purification Kit (Clontech,

Mountain View, CA) according to the manufacturer’s instructions. To assess interaction

between GST-HOXC8 and SRC-3, baculovirus-expressed, flag-tagged SRC-3 protein

was used. To assess interaction between GST-HOXC8 and AR, cell extract from a

confluent 15 cm dish of LNCaP cells treated for 24 hours with 10 nM R1881 was used.

Cells were harvested in cold PBS and lysed in 1 mL GST lysis buffer. GST-pull down

was performed as previously described (48).

Invasion- Invasion was assayed using Cell Biolabs Cytoselect 96-well Cell

Migration and Invasion Assay (Cell Biolabs, Inc., San Diego, CA) according to

manufacturer’s directions. LNCaP or HPr-1 AR cells were starved for 18 hours then 5 x

104 cells in serum-free RPMI (LNCaP) or keratinocyte serum-free medium without

supplements (HPr-1 AR) were added to the upper invasion chamber. Cells were

incubated for 48 hours at 37ºC, and cell invasion was assessed. Invasion values are

reported as mean RFU (relative fluorescence units) of quadruplicate samples.

RESULTS

HOXC8 inhibits AR-mediated transcription- It has been previously demonstrated

that HOX proteins bind to the steroid receptor coactivators CBP and p300, inhibiting the

HAT activity of the coactivators (25). We therefore asked whether overexpression of

HOXC8 might inhibit AR-mediated transcription via inhibition of CBP/p300. These

experiments were performed in LNCaP cells as they express functional AR and very low

endogenous HOXC8 mRNA (Supplementary Fig. S1A) and protein levels

(Supplementary Fig. S1B). A progressive increase in expression of HOXC8 (Fig. 1A)

decreased androgen-dependent activity of a probasin-luciferase reporter vector (Fig. 1B).

This cannot be ascribed to a general squelching or inhibitory effect of HOXC8

overexpression as neither expression of pCMV-β-gal (Fig.1C) or pRSV-luciferase (Fig.

1D) is affected.

Many HOX proteins share overlapping function. Our laboratory previously

observed overexpression of HOXC genes including HOXC6 in primary tumors,

metastases, and in PCa cell lines including LNCaP (1). We asked whether further

overexpression of HOXC6 in LNCaP cells also inhibits androgen-mediated gene

induction. Similar to the results of overexpressing HOXC8, HOXC6 expression inhibits

androgen-induced transcription in a dose-dependent manner (Supplementary Fig. S2A).

HOXB13, important in normal prostate function as well as in PCa, also exhibited a

similar inhibitory effect (Supplementary Fig. S2B). These data support the notion that,

among the subset of HOX proteins tested, there may exist functional redundancy with

respect to inhibition of AR-mediated signaling. The availability of an adequately specific

commercial antibody for detection of HOXC8 and the null background in LNCaP

prompted us to focus further studies on HOXC8.

To validate the transient transfection reporter assays and to investigate additional

activities of HOXC8, an LNCaP cell line stably overexpressing HOXC8 was constructed

by retroviral transduction. HOXC8 mRNA (Supplementary Fig. S3A) and protein

(Supplementary Fig. S3B) expression levels are increased in LNCaP-HOXC8 compared

with the negative control LNCaP-Empty cell line. While many PCa cell lines exhibit low

HOXC8, expression of HOXC8 in LNCaP-HOXC8 cells is comparable to the levels seen

in the PCa cell line, PC-3, that endogenously expresses HOXC8 (Supplementary Fig. S1).

These HOXC8- and empty vector-transduced LNCaP stable cell lines were transiently

transfected with the androgen-responsive MMTV-luciferase or PSA-luciferase reporter

plasmids. The HOXC8 expressing lines exhibit diminished AR-mediated promoter

activity of both reporters (Fig. 2A, 2B).

In order to determine whether this effect was specific to PCa cells or to LNCaP

cells in particular, we performed similar experiments in the immortalized, non-

tumorigenic prostate epithelial cell line HPr-1 AR. HPr-1-AR cells were transduced with

a retrovirus to stably overexpress HOXC8 or empty vector control. Real time reverse

transcriptase-PCR analysis and Western blot analysis confirm that HOXC8 mRNA

(Supplementary Fig. S3C) and protein (Supplementary Fig. S3D) expression levels are

increased in HPr-1 AR-HOXC8 cells compared with the negative control HPr-1 AR-

Empty cells. Androgen-induced activity directed by the MMTV-luciferase (Fig. 2C) and

PSA-luciferase (Fig. 2D) reporter plasmids was diminished in HPr-1 AR cells stably

overexpressing HOXC8. These data indicate that the inhibitory effect of HOXC8 on

androgen-mediated induction of different direct androgen target promoters is observed in

both tumorigenic and non-tumorigenic prostate cell lines and when HOXC8 is transiently

or stably overexpressed.

We also tested whether androgen induction of an endogenous gene was inhibited

by HOXC8 expression. We compared the androgen induction of the endogenous PSA

gene in LNCaP-HOXC8 and LNCaP-Empty as well as in HPr-1 AR-HOXC8 and HPr-1

AR-Empty cells. In both comparisons, the androgen induction of the gene was reduced

in the HOXC8-expressing line (Fig. 3).

HOXC8 inhibition of AR-mediated transcription is not due to down-regulation

of the androgen receptor- In order to rule out the possibility that HOXC8 inhibits AR-

mediated transcription by downregulating AR expression levels, real time reverse

transcriptase-PCR and Western blot analysis of AR mRNA and protein levels was

performed in LNCaP-HOXC8 and compared with that of the empty vector control cell

line. LNCaP-HOXC8 and LNCaP-Empty cells express nearly identical AR mRNA

(Supplementary Fig. S4A) and protein (Supplementary Fig. S4B) levels, indicating that

the inhibition of AR-mediated transcription by HOXC8 cannot be attributed to

downregulation of AR expression levels. AR protein levels are also unchanged in HPr-1

AR-HOXC8 cells when compared with the empty vector control cell line (data not

shown).

Overexpression of SRC-3, but not CBP, reverses the HOXC8 induced inhibition

of AR-mediated transcription- The report of Shen et al. (25) that HOX genes interact

with the transcriptional coactivator, CBP, and inhibit its HAT activity suggest a possible

mechanistic basis for the HOXC8-mediated inhibition of androgen-induced gene

expression. If suppression of CBP coactivator function were responsible for the inhibition

of AR action, we reasoned that overexpression of CBP might overcome the inhibition by

HOXC8. We overexpressed CBP and other cofactors that are major coregulators of

steroid signaling, the CBP relative, p300, and the three members of the p160 steroid

receptor coactivator family of coregulators, SRC-1, SRC-2, and SRC-3.

LNCaP PCa cells were co-transfected with the probasin-luciferase reporter, 100

ng HOXC8 expression vector (or empty vector control), and 100 or 500 ng of the

indicated coactivator expression vector (or empty vector control). Western blot analysis

of coactivator protein expression levels following transfection is shown in Supplementary

Figure S5. The results of the coactivator co-expression experiment are presented in

Figure 4. The average inhibition of the hormone induction by HOXC8 from duplicate

experiments is 58% without added coactivators. This is the control inhibition (Fig. 4,

leftmost bar) and the inhibition imposed by HOXC8 when coactivators are expressed is

compared to that standard. Contrary to our expectation, overexpression of CBP only

modestly counters the HOXC8-mediated inhibition of the hormone induction (Fig. 4,

second and third bars from the left). Both p300 and SRC-1 enhance the HOXC8-

mediated inhibition of AR-mediated transcription somewhat (Fig. 4, from left, bars 4-7)

increasing the inhibition from 58% to about 70-80%. Overexpression of SRC-2 weakly

diminishes the HOXC8 inhibition (Fig. 4, from left, bars 8-9). In contrast to the other

coactivators, overexpression of SRC-3 reverses the HOXC8-mediated inhibition. At the

highest level of expression, SRC-3 completely restores the hormone induction despite the

overexpression of HOXC8 (Fig. 4, from left, last two bars). These data suggest that

SRC-3, rather than CBP, may be the critical target for HOXC8 inhibition of AR-mediated

signaling. This result is of particular interest, as will be discussed, and directed

subsequent efforts toward elucidating the interplay between HOXC8 and SRC-3 in the

context of AR signaling in prostate cells.

SRC-3 expression in LNCaP PCa cell lines- In consideration of the finding that

overexpression of SRC-3, but not CBP, abrogates the HOXC8-mediated inhibition of AR

transcriptional activity, we asked whether SRC-3 mRNA or protein levels are altered

with HOXC8 overexpression. SRC-3 mRNA and protein levels were compared in

LNCaP-Empty and LNCaP-HOXC8 PCa cell lines by real time reverse transcriptase-

PCR and Western blot analysis. LNCaP-Empty and LNCaP-HOXC8 cells express very

similar SRC-3 mRNA (Supplementary Fig. S6A) and protein (Supplementary Fig. S6B)

levels. Therefore, HOXC8-mediated changes in overall SRC-3 expression levels are not a

contributing factor in the HOXC8 inhibition of AR-mediated signaling.

HOXC8 overexpression inhibits the hormone-dependent interaction of AR and

SRC-3- Gnanapragasam et al. demonstrated that SRC-3 binds AR in a ligand-dependent

manner in LNCaP PCa cells and that this interaction is important in enhancing

transcriptional activity in PCa cells (39). We, therefore, tested whether overexpression of

HOXC8 might alter the physical interaction between AR and SRC-3.

Co-IP experiments were performed to examine binding interactions between

SRC-3 and AR using LNCaP cells stably overexpressing HOXC8 and LNCaP control

cells. LNCaP-Empty cells exhibit an increase in the interaction of SRC-3 and AR upon

hormone treatment (Fig. 5A, top left). In contrast, LNCaP-HOXC8 cells exhibit

diminished interaction of SRC-3 and AR following treatment with R1881 (Fig. 5A, top

right bands), demonstrating that HOXC8 expression disrupts the ligand-enhanced

interaction between SRC-3 and AR. This is not due to a decrease in SRC-3 or AR

protein levels (Fig. 5A, middle, bottom panel, respectively). While a slight decrease was

observed in SRC-3 protein levels after treatment with androgen (Fig. 5A, middle panel),

this decrease is seen in both the LNCaP-Empty and LNCaP-HOXC8 cell lines.

Quantitation of these data normalized to input is presented in Figure 5B.

HOXC8 expression inhibits recruitment of SRC-3 and CBP and acetylation of

histones- ChIP analyses were performed to assess whether the recruitment of SRC-3 to

the endogenous AR target gene, PSA, is also inhibited by HOXC8 overexpresssion. It

has been previously demonstrated that in the presence of hormone, AR predominately

occupies the distal enhancer region, not the proximal promoter region, of the PSA gene in

LNCaP cells (49-51). In LNCaP-HOXC8 cells, the occupancy by SRC-3 at the PSA

enhancer region is greatly inhibited when compared to that observed in the LNCaP-

Empty control cell line (Fig. 6A). The SRC coactivators interact directly with steroid

receptors bound to target DNA elements, and serve as a platform to recruit additional

coregulators including the HAT coactivator CBP (52;53). Recruitment of these factors

by SRC/p160 proteins is critical for steroid receptor-mediated local chromatin

remodeling and assembly of transcriptional machinery around target gene promoters

(54;55). We, therefore, also assessed recruitment of CBP and found its androgen-

dependent recruitment to be diminished in LNCaP-HOXC8, consistent with the idea that

the inhibition of SRC-3 recruitment in turn diminishes CBP recruitment (Fig 6B).

A potential explanation for reduced recruitment of SRC-3 is that HOXC8 inhibits

recruitment of AR to target response elements. However, we observe no significant

change in either the basal or hormone-induced occupancy of AR at the PSA enhancer

region in LNCaP-HOXC8 cells when compared with that of the LNCaP-Empty control

cell line (Fig. 6C). This suggests that HOXC8 imposes its inhibitory effects on AR

signaling independent of AR binding to target genes by inhibiting further recruitment of

cofactors including SRC-3 and CBP to AR target genes.

The SRC-3 loading experiments show that basal as well as hormone-induced

levels of SRC-3 loading are diminished by HOXC8 expression. SRC-3 is a coactivator

for many factors acting at many promoters and HOXC8 may be inhibiting SRC-3 at

many or all of these. This suggestion implies that HOXC8 would impose pleiotropic

actions beyond that of androgen-regulated genes through its inhibitory effect on SRC-3.

Both CBP and SRC-3 possess intrinsic HAT activity (52;56;57). Coactivator-

dependent chromatin modification is a critical step in hormone receptor-mediated

induction of gene expression. We predicted that the HOXC8-mediated inhibition of

coactivator recruitment would be reflected in reduced acetylation of histones at target

promoters. As can be seen in Fig. 7A, androgen-induced acetylation of histones H3 and

H4 at the MMTV promoter is indeed suppressed by HOXC8. Quantitation of these data

normalized to input is shown in Figure 7B.

HOXC8 does not directly interact with SRC-3, AR or CBP- To further assess the

mechanism of HOXC8-mediated inhibition of AR signaling, experiments were conducted

to test for direct interactions between HOXC8 and SRC-3. In vitro GST-pull down

assays were performed using the fusion protein GST-HOXC8 as bait along with flag-

tagged SRC-3 purified from baculovirus expression vector-infected insect cells. No

interaction was detected between GST-HOXC8 and purified SRC-3 above the minimal

levels observed with the GST control (Supplementary Fig. S7A). Interaction between

HOXC8 and AR was also examined, but no binding above background was detected

between GST-HOXC8 and AR using extracts from hormone-treated LNCaP as a source

of AR (Supplementary Fig. S7B) or extracts from untreated cells (data not shown). We

also looked for binding between HOXC8 and AR, SRC-3 or CBP by co-IP using LNCaP

cellular extracts but no binding interactions were detected either in the presence or

absence of hormone (data not shown). All in vitro binding assays have limitations,

therefore, the possibility of interactions between HOXC8 and SRC-3 or AR in the context

of the cell cannot be completely discounted.

HOXC8 effects on in vitro surrogate markers of tumorigenesis and progression-

Overexpression of HOXC genes is associated with PCa (1) and, in the case of HOXC8,

increased Gleason grade (22). Recent studies have implicated HOXC8 in the regulation

of cell cycle and proliferation (58-61), apoptosis (59;62), cell differentiation (63-68),

cell adhesion (59;62;69), cytoskeleton/motility/migration (62) and tumorigenesis

(59;61;62;69;70). Based on these reports and the present lack of understanding as to

HOXC8 function in prostate carcinogenesis, its effects on in vitro surrogate markers of

prostate cell tumorigenicity were examined.

Malignant transformation is associated with certain phenotypic changes including

dysegulation of cell cycle and cell death pathways, loss of contact inhibition and

anchorage-independence, and increased motility and invasiveness. We observed no

dramatic changes in anchorage-independent growth, apoptosis, or migration in

conjunction with HOXC8 overexpression in PCa cells or non-tumorigenic prostate cell

lines. A mild inhibition of growth of LNCaP and HPr-1 AR cells was seen with an

accompanying increase in G1 phase of the cell cycle (data not shown). None of these

assays suggested a critical role of HOXC8 in PCa tumorigenesis or progression.

HOXC8 expression increases invasiveness in HPr-1 AR prostate epithelial cells,

but not LNCaP PCa cells- Invasive capacity is the single most important trait that

distinguishes benign from malignant lesions. Invasion assays in vitro require cells to

migrate through an extracellular matrix or basement membrane extract barrier by first

enzymatically degrading the matrix barrier in order to become established in a new

location. HOXC8 has been implicated as a regulator of cadherin 11 and osteopontin

(59;62). Cadherin 11 has been shown to produce significant changes in the invasive

capacity of cancer cells (71), while cumulative evidence suggests that osteopontin

functions in the regulation of tumor metastasis and invasion (72;73).

The ability of LNCaP PCa and HPr-1 AR non-tumorigenic prostate epithelial

cells overexpressing HOXC8 to invade through a coated membrane was assessed.

LNCaP-HOXC8 cells invade to a similar degree as LNCaP-Empty control cell line (Fig.

8A). HPr-1 AR-Empty cells invade to a level less than 20% that of LNCaP cells, but in

contrast to paired LNCaP cell lines, invasion by HPr-1 AR-HOXC8 cells is nearly double

that of HPr-1 AR-Empty control cells (Fig. 8B). These data suggest that HOXC8

expression may direct pro-tumorigenic actions in prostate carcinogenesis by promoting

increased invasive capacity.

DISCUSSION

Inappropriate expression of HOX genes is associated with a number of cancers (6-

10;14-20). In particular, any of four genes of the HOXC cluster may be overexpressed in

PCa (1). The present study suggests a complex role for HOXC8 in PCa, promoting

invasiveness while at the same time inhibiting AR-mediated gene induction at ARE-

regulated genes associated with differentiated function of the prostate. HOXC8

overexpression may well affect the participation of SRC-3 at promoters in addition to

ARE-driven promoters. SRC-3 serves as a cofactor for many transcription factors and a

general inhibition of its activity by HOXC8 would be expected to have pleiotropic

disruptive effects on cellular homeostasis.

Mechanism of HOXC8-dependent inhibition of androgen-mediated gene

expression- Although a report that HOX genes inhibited the HAT activity of CBP (25)

suggested that HOXC8 might inhibit AR-meditated gene induction through this

mechanism, we find evidence that it is SRC-3 rather than CBP, p300, SRC-1, or SRC-2

that seems to be the key target of HOXC8. SRC-3 is an intriguing candidate because, like

HOX genes, SRC-3 is overexpressed in many cancers. Indeed, one of its many names

(NCOA-3/AIB-1/pCIP/ACTR/RAC3/TRAM-1) AIB-1, amplified in breast cancer-1,

came from its identification as one of three gene amplifications frequently observed in

human breast carcinomas (74). The involvement of SRC-3 overexpression in PCa

appears to be especially critical in AR-mediated gene induction. Among the p160 family

of coactivators, SRC-3 is the preferred coactivator for hormone-activated AR. SRC-3 is

selectively phosphorylated when cells are treated with androgen (75) the binding affinity

of SRC-3 for AR is 10-100-fold stronger than that of SRC-1 or SRC-2 (76), and SRC-3

activates AR more potently than other SRCs in the presence of 5α-dihydrotestosterone

(76). Thus, SRC-3 represents an attractive target for HOXC8-inhibition of AR function.

Inhibition of androgen action- a conundrum? Since prostate growth and

maintenance are androgen dependent, why then doesn’t HOXC8 expression serve a

tumor suppressor function rather than being associated with PCa and/or worse prognosis?

One possibility is that any inhibition of androgen signaling by HOXC gene

overexpression is sufficiently countermanded by concomitant overexpression of SRC-3

to maintain androgen-driven growth of PCa. Another possibility is that maximal AR

activity is not favorable to growth promotion in PCa. LNCaP cells, widely used as a

model of androgen-dependent PCa growth, exhibit a biphasic response to androgens.

Maximal growth promotion is seen at levels of androgen not only below the Kd for the

receptor but also below those considered to be normal physiologic levels of androgens

(Daddario Axlund and Nordeen, unpublished results). Thus to the extent that LNCaP

growth responses reflect PCa in vivo, partial suppression of AR activity at physiologic

androgen levels may effectively shift the growth response to a more growth-promoting

place on the dose response curve.

Alternatively, HOXC genes may be acting differentially on critical genes in the

prostate. PSA is a gene representing a differentiated function of the prostate. Perhaps

HOXC expression inhibits genes associated with prostate differentiation, but has no

effect, or a stimulatory effect, on genes involved in tumorigenesis. Our data

demonstrating that HOXC8 expression inhibits AR-mediated induction of the PSA gene,

coupled with the finding that HOXC8 expression increases invasiveness in prostate

epithelial cells, taken together with recent work on HOXB13 (26) suggest this may in fact

be the case.

Norris et al. have recently demonstrated that HOXB13 regulates AR target gene

activity in both a positive and a negative manner (26). HOXB13 interacts with AR and

inhibits the transcription of genes that contain an ARE, whereas, in complex together,

AR-HOXB13 confers androgen responsiveness to genes whose promoters contain a

HOXB13 response element. Further, they demonstrate that in the case of genes that

contain a HOXB13 element adjacent to an ARE, HOXB13 and AR synergistically

enhance the transcription of these genes (26). These observations support the notion that

HOX proteins may play a dynamic role in the normal development and tumorigenesis in

the prostate through both the repression and activation of AR-mediated signaling. Like

HOXC8 or HOXC6, we see that HOXB13 inhibits AR-mediated transcription at ARE-

driven promoters. While it is plausible that HOXB13 and HOXC8 act in a similar

fashion to inhibit direct AR target genes, we have been unable to detect an interaction

between HOXC8 and AR in contrast to what has been previously reported for HOXB13

and AR (27). To assess whether HOXC8 modulates AR function in both a positive and

negative manner via direct or indirect interaction with AR, SRC-3, or other coregulators

important in AR signaling, further investigations will be required.

A final mechanism, which is not mutually exclusive, to explain a role of HOXC

overexpression in PCa in the face of a suppressive activity on androgen action invokes

the concept that HOXC overexpression is critical to disruption of cellular homeostasis at

an early stage of prostate tumorigenesis. The developing tumor is forced to adapt to

diminished androgen signaling that accompanies other pro-tumorigenic actions of HOXC

expression, including increased invasion by HOXC8 expression, thereby predisposing the

tumor to survive in the face of a subsequent androgen withdrawal through androgen

deprivation therapy (ADT). This model could account for the relatively rapid failure of

ADT by suggesting that an adaptation to diminished androgen signaling has already

occurred before androgen ablation by ADT is imposed.

There is precedence for the notion that dysregulated HOXC expression may be an

early event in carcinogenesis. Oct-4, a homeobox gene critical in the genesis of testicular

germ cell tumors, is expressed in precursor lesions (77;78), while homeobox genes

NKX3.1 and CDX2 are downregulated at the earliest stages of carcinogenesis of the

prostate and colon, respectively (79;80).

HOXC gene expression and prostate cancer- Additional studies are required to

distinguish between the different explanations to reconcile the apparent paradox that

HOXC8 overexpression is associated with PCa yet can inhibit androgen-dependent

signaling. Our studies reveal that HOXC8 promotes invasiveness in prostate epithelial

cells but does not further promote invasiveness in tumorigenic PCa cells consistent with

an early role of HOXC genes in promoting PCa progression and suggesting that HOX

genes play an important role in the disruption of cellular homeostasis leading to cancer.

The findings of this work are summarized in the model presented in Fig. 9 depicting the

inhibition of direct AR target genes by HOXC8 through its inhibition of SRC-3

recruitment. The model also suggests that effects of HOXC8 acting as a transcription

factor in its own right along with its transcription partner PBX may be mediating

increased invasion and other activities promoting tumorigenesis. Preliminary microarray

studies implicate HOXC8 in the regulation of various genes involved in matrix

degradation including MMP1, MMP2, TIMP1, PLAU and SERPINE 1 (data not shown).

Other studies identifying putative transcriptional targets of HOXC8 have implicated it in

the regulation of cell cycle (58-61), apoptosis (59;62), cell adhesion (59;62;69), migration

(62) and tumorigenesis (59;61;62;69;70). Confirmation of these and other HOXC8

targets implicated in cancer may prove extremely valuable and help achieve a greater

understanding of the role of HOXC8 in prostate tumorigenesis.

An estimated 91% of men with newly diagnosed PCa cases are likely to have

stage 1 or stage 2 disease, in which case the 5 year relative survival is nearly 100% (81).

With such a favorable prognosis, many question the benefit of exposing men with early

stage and low or moderate grade PCa to radical surgery or radiation therapy (82). It is

therefore critical to be able to identify men whose histologically early stage PCa will

progress and who would, therefore, benefit from aggressive therapy (83;84). Hence,

there is a great need for reliable PCa markers that can predict cancers requiring

aggressive therapy thereby enabling better informed treatment options. Given the

correlation of HOXC8 with increased Gleason grade (22) and data shown herein

demonstrating that HOXC8 increases invasiveness of non-tumorigenic prostate cells,

further analysis of the consequences of HOXC gene expression and its use as a

prognostic marker is warranted.

REFERENCES

(1) Miller GJ, Miller HL, van BA, Lambert JR, Werahera PN, Schirripa O et al. Aberrant HOXC expression accompanies the malignant phenotype in human prostate. Cancer Res 2003; 63(18):5879-5888.

(2) Lewis EB. A gene complex controlling segmentation in Drosophila. Nature 1978; 276(5688):565-570.

(3) Pearson JC, Lemons D, McGinnis W. Modulating Hox gene functions during animal body patterning. Nat Rev Genet 2005; 6(12):893-904.

(4) Mann RS, Morata G. The developmental and molecular biology of genes that subdivide the body of Drosophila. Annu Rev Cell Dev Biol 2000; 16:243-271.

(5) McGinnis W, Krumlauf R. Homeobox genes and axial patterning. Cell 1992; 68(2):283-302.

(6) Lawrence HJ, Stage KM, Mathews CH, Detmer K, Scibienski R, MacKenzie M et al. Expression of HOX C homeobox genes in lymphoid cells. Cell Growth Differ 1993; 4(8):665-669.

(7) Celetti A, Barba P, Cillo C, Rotoli B, Boncinelli E, Magli MC. Characteristic patterns of HOX gene expression in different types of human leukemia. Int J Cancer 1993; 53(2):237-244.

(8) Perkins A, Kongsuwan K, Visvader J, Adams JM, Cory S. Homeobox gene expression plus autocrine growth factor production elicits myeloid leukemia. Proc Natl Acad Sci U S A 1990; 87(21):8398-8402.

(9) Lawrence HJ, Sauvageau G, Ahmadi N, Lopez AR, LeBeau MM, Link M et al. Stage- and lineage-specific expression of the HOXA10 homeobox gene in normal and leukemic hematopoietic cells. Exp Hematol 1995; 23(11):1160-1166.

(10) Borrow J, Shearman AM, Stanton VP, Jr., Becher R, Collins T, Williams AJ et al. The t(7;11)(p15;p15) translocation in acute myeloid leukaemia fuses the genes for nucleoporin NUP98 and class I homeoprotein HOXA9. Nat Genet 1996; 12(2):159-167.

(11) Frohling S, Scholl C, Bansal D, Huntly BJ. HOX gene regulation in acute myeloid leukemia: CDX marks the spot? Cell Cycle 2007; 6(18):2241-2245.

(12) Shimamoto T, Ohyashiki K, Toyama K, Takeshita K. Homeobox genes in hematopoiesis and leukemogenesis. Int J Hematol 1998; 67(4):339-350.

(13) Alami Y, Castronovo V, Belotti D, Flagiello D, Clausse N. HOXC5 and HOXC8 expression are selectively turned on in human cervical cancer cells compared to normal keratinocytes. Biochem Biophys Res Commun 1999; 257(3):738-745.

(14) De VG, Barba P, Odartchenko N, Givel JC, Freschi G, Bucciarelli G et al. Expression of homeobox-containing genes in primary and metastatic colorectal cancer. Eur J Cancer 1993; 29A(6):887-893.

(15) Chariot A, Castronovo V. Detection of HOXA1 expression in human breast cancer. Biochem Biophys Res Commun 1996; 222(2):292-297.

(16) Chariot A, Castronovo V, Le P, Gillet C, Sobel ME, Gielen J. Cloning and expression of a new HOXC6 transcript encoding a repressing protein. Biochem J 1996; 319 ( Pt 1):91-97.

(17) Buccoliero AM, Castiglione F, Degl'Innocenti DR, Ammanati F, Giordano F, Sanzo M et al. Hox-D genes expression in pediatric low-grade gliomas: real-time-PCR study. Cell Mol Neurobiol 2009; 29(1):1-6.

(18) Cillo C, Barba P, Freschi G, Bucciarelli G, Magli MC, Boncinelli E. HOX gene expression in normal and neoplastic human kidney. Int J Cancer 1992; 51(6):892-897.

(19) Abe M, Hamada J, Takahashi O, Takahashi Y, Tada M, Miyamoto M et al. Disordered expression of HOX genes in human non-small cell lung cancer. Oncol Rep 2006; 15(4):797-802.

(20) Lechner JF, Wang Y, Siddiq F, Fugaro JM, Wali A, Lonardo F et al. Human lung cancer cells and tissues partially recapitulate the homeobox gene expression profile of embryonic lung. Lung Cancer 2002; 37(1):41-47.

(21) Gu ZD, Chen XM, Zhang W, Gu J, Chen KN. [Expression of 39 HOX genes in esophageal cancer cell lines]. Zhonghua Wei Chang Wai Ke Za Zhi 2007; 10(4):365-367.

(22) Waltregny D, Alami Y, Clausse N, de LJ, Castronovo V. Overexpression of the homeobox gene HOXC8 in human prostate cancer correlates with loss of tumor differentiation. Prostate 2002; 50(3):162-169.

(23) Kikugawa T, Kinugasa Y, Shiraishi K, Nanba D, Nakashiro K, Tanji N et al. PLZF regulates Pbx1 transcription and Pbx1-HoxC8 complex leads to androgen-independent prostate cancer proliferation. Prostate 2006; 66(10):1092-1099.

(24) Ramachandran S, Liu P, Young AN, Yin-Goen Q, Lim SD, Laycock N et al. Loss of HOXC6 expression induces apoptosis in prostate cancer cells. Oncogene 2005; 24(1):188-198.

(25) Shen WF, Krishnan K, Lawrence HJ, Largman C. The HOX homeodomain proteins block CBP histone acetyltransferase activity. Mol Cell Biol 2001; 21(21):7509-7522.

(26) Norris JD, Chang CY, Wittmann BM, Kunder RS, Cui H, Fan D et al. The homeodomain protein HOXB13 regulates the cellular response to androgens. Mol Cell 2009; 36(3):405-416.

(27) Jung C, Kim RS, Zhang HJ, Lee SJ, Jeng MH. HOXB13 induces growth suppression of prostate cancer cells as a repressor of hormone-activated androgen receptor signaling. Cancer Res 2004; 64(24):9185-9192.

(28) Zhou HJ, Yan J, Luo W, Ayala G, Lin SH, Erdem H et al. SRC-3 is required for prostate cancer cell proliferation and survival. Cancer Res 2005; 65(17):7976-7983.

(29) Zhou HJ, Yan J, Luo W, Ayala G, Lin SH, Erdem H et al. SRC-3 is required for prostate cancer cell proliferation and survival. Cancer Res 2005; 65(17):7976-7983.

(30) Anzick SL, Kononen J, Walker RL, Azorsa DO, Tanner MM, Guan XY et al. AIB1, a steroid receptor coactivator amplified in breast and ovarian cancer. Science 1997; 277(5328):965-968.

(31) Balmer NN, Richer JK, Spoelstra NS, Torkko KC, Lyle PL, Singh M. Steroid receptor coactivator AIB1 in endometrial carcinoma, hyperplasia and normal endometrium: Correlation with clinicopathologic parameters and biomarkers. Mod Pathol 2006; 19(12):1593-1605.

(32) Tanner MM, Grenman S, Koul A, Johannsson O, Meltzer P, Pejovic T et al. Frequent amplification of chromosomal region 20q12-q13 in ovarian cancer. Clin Cancer Res 2000; 6(5):1833-1839.

(33) Henke RT, Haddad BR, Kim SE, Rone JD, Mani A, Jessup JM et al. Overexpression of the nuclear receptor coactivator AIB1 (SRC-3) during progression of pancreatic adenocarcinoma. Clin Cancer Res 2004; 10(18 Pt 1):6134-6142.

(34) Ghadimi BM, Schrock E, Walker RL, Wangsa D, Jauho A, Meltzer PS et al. Specific chromosomal aberrations and amplification of the AIB1 nuclear receptor coactivator gene in pancreatic carcinomas. Am J Pathol 1999; 154(2):525-536.

(35) Sakakura C, Hagiwara A, Yasuoka R, Fujita Y, Nakanishi M, Masuda K et al. Amplification and over-expression of the AIB1 nuclear receptor coactivator gene in primary gastric cancers. Int J Cancer 2000; 89(3):217- 223.

(36) Xie D, Sham JS, Zeng WF, Lin HL, Bi J, Che LH et al. Correlation of AIB1 overexpression with advanced clinical stage of human colorectal carcinoma. Hum Pathol 2005; 36(7):777-783.

(37) Wang Y, Wu MC, Sham JS, Zhang W, Wu WQ, Guan XY. Prognostic significance of c-myc and AIB1 amplification in hepatocellular carcinoma. A broad survey using high-throughput tissue microarray. Cancer 2002; 95(11):2346-2352.

(38) Nessler-Menardi C, Jotova I, Culig Z, Eder IE, Putz T, Bartsch G et al. Expression of androgen receptor coregulatory proteins in prostate cancer and stromal-cell culture models. Prostate 2000; 45(2):124-131.

(39) Gnanapragasam VJ, Leung HY, Pulimood AS, Neal DE, Robson CN. Expression of RAC 3, a steroid hormone receptor co-activator in prostate cancer. Br J Cancer 2001; 85(12):1928-1936.

(40) Zou JX, Zhong Z, Shi XB, Tepper CG, deVere White RW, Kung HJ et al. ACTR/AIB1/SRC-3 and androgen receptor control prostate cancer cell proliferation and tumor growth through direct control of cell cycle genes. Prostate 2006; 66(14):1474-1486.

(41) Horoszewicz JS, Leong SS, Chu TM, Wajsman ZL, Friedman M, Papsidero L et al. The LNCaP cell line--a new model for studies on human prostatic carcinoma. Prog Clin Biol Res 1980; 37:115-132.

(42) Horoszewicz JS, Leong SS, Kawinski E, Karr JP, Rosenthal H, Chu TM et al. LNCaP model of human prostatic carcinoma. Cancer Res 1983; 43(4):1809-1818.

(43) Ling MT, Chan KW, Choo CK. Androgen induces differentiation of a human papillomavirus 16 E6/E7 immortalized prostate epithelial cell line. J Endocrinol 2001; 170(1):287-296.

(44) Lambert JR, Kelly JA, Shim M, Huffer WE, Nordeen SK, Baek SJ et al. Prostate derived factor in human prostate cancer cells: gene induction by vitamin D via a p53-dependent mechanism and inhibition of prostate cancer cell growth. J Cell Physiol 2006; 208(3):566-574.

(45) Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001; 25(4):402-408. 28

(46) Perez-Stable CM, Pozas A, Roos BA. A role for GATA transcription factors in the androgen regulation of the prostate-specific antigen gene enhancer. Mol Cell Endocrinol 2000; 167(1-2):43-53.

(47) Lambert JR, Nordeen SK. CBP recruitment and histone acetylation in differential gene induction by glucocorticoids and progestins. Mol Endocrinol 2003; 17(6):1085-1094.

(48) Detection of protein-protein interactions using the GST fusion protein pull- down technique. Nat Meth 2004; 1(3):275-276.

(49) Louie MC, Yang HQ, Ma AH, Xu W, Zou JX, Kung HJ et al. Androgen- induced recruitment of RNA polymerase II to a nuclear receptor-p160 coactivator complex. Proc Natl Acad Sci U S A 2003; 100(5):2226-2230.

(50) Jia L, Coetzee GA. Androgen receptor-dependent PSA expression in androgen-independent prostate cancer cells does not involve androgen receptor occupancy of the PSA locus. Cancer Res 2005; 65(17):8003-8008.

(51) Jia L, Kim J, Shen H, Clark PE, Tilley WD, Coetzee GA. Androgen receptor activity at the prostate specific antigen locus: steroidal and non- steroidal mechanisms. Mol Cancer Res 2003; 1(5):385-392.

(52) Chen H, Lin RJ, Schiltz RL, Chakravarti D, Nash A, Nagy L et al. Nuclear receptor coactivator ACTR is a novel histone acetyltransferase and forms a multimeric activation complex with P/CAF and CBP/p300. Cell 1997; 90(3):569-580.

(53) Torchia J, Rose DW, Inostroza J, Kamei Y, Westin S, Glass CK et al. The transcriptional co-activator p/CIP binds CBP and mediates nuclear- receptor function. Nature 1997; 387(6634):677-684.

(54) Chen D, Ma H, Hong H, Koh SS, Huang SM, Schurter BT et al. Regulation of transcription by a protein methyltransferase. Science 1999; 284(5423):2174-2177.

(55) McKenna NJ, O'Malley BW. Combinatorial control of gene expression by nuclear receptors and coregulators. Cell 2002; 108(4):465-474.

(56) Kamei Y, Xu L, Heinzel T, Torchia J, Kurokawa R, Gloss B et al. A CBP integrator complex mediates transcriptional activation and AP-1 inhibition by nuclear receptors. Cell 1996; 85(3):403-414.

(57) Chakravarti D, LaMorte VJ, Nelson MC, Nakajima T, Schulman IG, Juguilon H et al. Role of CBP/P300 in nuclear receptor signalling. Nature 1996; 383(6595):99-103.

(58) Kamel S, Kruger C, Salbaum JM, Kappen C. Morpholino-mediated knockdown in primary chondrocytes implicates Hoxc8 in regulation of cell cycle progression. Bone 2009; 44(4):708-716.

(59) Lei H, Juan AH, Kim MS, Ruddle FH. Identification of a Hoxc8-regulated transcriptional network in mouse embryo fibroblast cells. Proc Natl Acad Sci U S A 2006; 103(27):10305-10309.

(60) Min H, Lee JY, Bok J, Chung HJ, Kim MH. Proliferating cell nuclear antigen (Pcna) as a direct downstream target gene of Hoxc8. Biochem Biophys Res Commun 2010; 392(4):543-547.

(61) Kwon Y, Ko JH, Byung-Gyu K, Kim MH. Analysis of plausible downstream target genes of Hoxc8 in F9 teratocarcinoma cells. Putative downstream target genes of Hoxc8. Mol Biol Rep 2003; 30(3):141-148.

(62) Lei H, Wang H, Juan AH, Ruddle FH. The identification of Hoxc8 target genes. Proc Natl Acad Sci U S A 2005; 102(7):2420-2424.

(63) Juan AH, Lei H, Bhargava P, Lebrun M, Ruddle FH. Multiple roles of hoxc8 in skeletal development. Ann N Y Acad Sci 2006; 1068:87-94.

(64) Yueh YG, Gardner DP, Kappen C. Evidence for regulation of cartilage differentiation by the homeobox gene Hoxc-8. Proc Natl Acad Sci U S A 1998; 95(17):9956-9961.

(65) Yang X, Ji X, Shi X, Cao X. Smad1 domains interacting with Hoxc-8 induce osteoblast differentiation. J Biol Chem 2000; 275(2):1065-1072.

(66) Zheng YJ, Chung HJ, Min H, Kang M, Kim SH, Gadi J et al. In vitro osteoblast differentiation is negatively regulated by Hoxc8. Appl Biochem Biotechnol 2010; 160(3):891-900.

(67) Magli MC, Largman C, Lawrence HJ. Effects of HOX homeobox genes in blood cell differentiation. J Cell Physiol 1997; 173(2):168-177.

(68) Shimamoto T, Tang Y, Naot Y, Nardi M, Brulet P, Bieberich CJ et al. Hematopoietic progenitor cell abnormalities in Hoxc-8 null mutant mice. J Exp Zool 1999; 283(2):186-193.

(69) Tomotsune D, Shoji H, Wakamatsu Y, Kondoh H, Takahashi N. A mouse homologue of the Drosophila tumour-suppressor gene l(2)gl controlled by Hox-C8 in vivo. Nature 1993; 365(6441):69-72.

(70) Lei H, Juan AH, Kim MS, Ruddle FH. Mouse naked cuticle 2 (mNkd2) as a direct transcriptional target of Hoxc8 in vivo. J Exp Zool A Ecol Genet Physiol 2007; 307(1):1-6. 30

(71) Feltes CM, Kudo A, Blaschuk O, Byers SW. An Alternatively Spliced Cadherin-11 Enhances Human Breast Cancer Cell Invasion. Cancer Research 2002; 62(22):6688-6697.

(72) Tuck AB, Hota C, Wilson SM, Chambers AF. Osteopontin-induced migration of human mammary epithelial cells involves activation of EGF receptor and multiple signal transduction pathways. Oncogene 0 AD; 22(8):1198-1205.

(73) Philip YW, Paul CK. The role of Osteopontin in tumor metastasis. J Surg Res 2004; 121(2):228-241.

(74) Guan XY, Xu J, Anzick SL, Zhang H, Trent JM, Meltzer PS. Hybrid selection of transcribed sequences from microdissected DNA: isolation of genes within amplified region at 20q11-q13.2 in breast cancer. Cancer Res 1996; 56(15):3446-3450.

(75) Wu RC, Qin J, Yi P, Wong J, Tsai SY, Tsai MJ et al. Selective phosphorylations of the SRC-3/AIB1 coactivator integrate genomic reponses to multiple cellular signaling pathways. Mol Cell 2004; 15(6):937- 949.

(76) Zhou XE, Suino-Powell KM, Li J, He Y, Mackeigan JP, Melcher K et al. Identification of SRC3/AIB1 as a preferred coactivator for hormone- activated androgen receptor. J Biol Chem 2010; 285(12):9161-9171.

(77) Looijenga LH, Stoop H, de Leeuw HP, de Gouveia Brazao CA, Gillis AJ, van Roozendaal KE et al. POU5F1 (OCT3/4) identifies cells with pluripotent potential in human germ cell tumors. Cancer Res 2003; 63(9):2244-2250.

(78) Gidekel S, Pizov G, Bergman Y, Pikarsky E. Oct-3/4 is a dose-dependent oncogenic fate determinant. Cancer Cell 2003; 4(5):361-370.

(79) Shen MM, Abate-Shen C. Roles of the Nkx3.1 homeobox gene in prostate organogenesis and carcinogenesis. Dev Dyn 2003; 228(4):767-778.

(80) Abate-Shen C. Homeobox genes and cancer: New OCTaves for an old tune. Cancer Cell 2003; 4(5):329-330.

(81) Gelmann EP, Henshall SM. Clinically relevant prognostic markers for prostate cancer: the search goes on. Ann Intern Med 2009; 150(9):647-649.

(82) Scardino PT. Localized prostate cancer is rarely a fatal disease. Nat Clin Pract Urol 2008; 5(1):1.

(83) Swanson GP, Riggs M, Hermans M. Pathologic findings at radical prostatectomy: risk factors for failure and death. Urol Oncol 2007; 25(2):110-114.

(84) Thompson IM, Tangen CM, Paradelo J, Lucia MS, Miller G, Troyer D et al. Adjuvant radiotherapy for pathological T3N0M0 prostate cancer significantly reduces risk of metastases and improves survival: long-term followup of a randomized clinical trial. J Urol 2009; 181(3):956-962.

FIGURE LEGENDS

Figure 1. HOXC8 overexpression specifically inhibits AR-mediated gene induction in LNCaP PCa cells (A, B) LNCaP cells were transfected with a probasin-luciferase reporter vector and increasing amounts of HOXC8 expression vector. Total transfected plasmid was held constant by balancing with empty expression vector. Cells were treated with 10 nM R1881 for 18 hours. (A) Western blot for HOXC8 (60 s exposure) and actin (1 s exposure). (B) Luciferase activity normalized to total protein. LNCaP cells were transfected with (C) pRSV-luciferase or (D) pCMV-βgal along with the HOXC8 expression vector or the empty vector control. Luciferase data shown are from a single experiment and are representative of three (B) or two (C,D) separate experiments. Values represent mean ± S.D. (B, n=4). (C, D, n=3).

Figure 2. AR-mediated transcription is inhibited in prostate cells stably overexpressing HOXC8 Retroviral transduction was employed to stably overexpress HOXC8 in PCa (LNCaP, panels A, B) and non-tumorigenic prostate (HPr-1 AR, panels C, D) cell lines. Paired control cell lines were created by transduction of the empty vector. HOXC8-expressing and control cell lines were transfected with an MMTV-luciferase (A, C) or PSA- luciferase (B, D) reporter and treated with 10 nM R1881 for 18 hours. Luciferase activity is normalized to total protein. Data shown are from a single experiment and are representative of three independent experiments. Values represent mean ± S.D. (n=4); *P< 0.05.

Figure 3. Induction of the endogenous PSA gene is inhibited when HOXC8 is stably overexpressed (A) LNCaP and (B) HPr-1 AR cells stably expressing HOXC8 and paired control cells were serum starved for 48 hours then treated with 10 nM R1881 for 6 hours. PSA mRNA levels were analyzed by real time reverse transcriptase-PCR analysis. Data shown are from a single experiment and are representative of three separate experiments. Values represent mean ± S.D. (n=3); * P< 0.01, ** P< 0.05.

Figure 4. SRC-3 overexpression reverses the HOXC8-induced inhibition of AR- mediated transcription LNCaP cells were co-transfected with a probasin-luciferase reporter, 100 ng HOXC8 expression vector or empty vector control and 100 or 500 ng of the indicated coactivator expression vector. Total transfected plasmid was held constant by balancing with empty expression vector. Cells were treated with 10 nM R1881 for 18 hours. The inhibition of the hormone induction by HOXC8 was 58% without added coactivators. This is the control inhibition (leftmost bar) and the inhibition imposed by HOXC8 when coactivators are expressed is compared to that standard. Data shown are the average from two independent experiments where each condition was tested in duplicate. Values represent mean ± range.

Figure 5. The hormone-dependent interaction of AR/SRC-3 is inhibited in LNCaP- HOXC8 PCa cells (A) LNCaP-empty and LNCaP-HOXC8 cells were cultured in 10 nM R1881 or ethanol for 24 hours. SRC-3 was immunoprecipitated and the association with AR assessed by Western blot analysis using an antibody directed against AR. AR protein levels were similar in all samples as shown in the input (bottom panel). (B) Quantitation of AR bound to SRC-3 (data from (A) top panel, normalized to SRC-3 input, middle panel). Data shown are from a single experiment and are representative of three separate experiments.

Figure 6. Recruitment of SRC-3 and CBP to the endogenous PSA enhancer is inhibited in LNCaP-HOXC8 cells ChIP assays were performed to analyze the occupancy of SRC-3 (A), CBP (B) and AR (C) at the endogenous PSA enhancer region. Cells were treated with R1881 or ethanol for 60 minutes. Occupancy was normalized to that of empty vector control cells treated with ethanol. Data shown are from a single experiment and are representative of three separate experiments. Values represent mean ± S.D. (n=3).

Figure 7. HOXC8 expression inhibits histone acetylation at an androgen-responsive gene promoter in PCa cells (A) MMTV-luciferase and 100ng HOXC8 or empty control expression vectors were transiently transfected into LNCaP cells. Cells were treated with hormone as indicated and ChIP was performed to assess acetylation of histone H4 and histone H3 at the MMTV hormone-responsive promoter. (B) Quantitation of IP samples normalized to input is indicated. Data shown are from a single experiment and are representative of three separate experiments.

Figure 8. HOXC8 expression increases invasiveness in non-tumorigenic HPr-1 AR cells, but not in LNCaP PCa cells Invasion by LNCaP PCa cells (A) and HPr-1 AR cells (B) overexpressing HOXC8 when compared to empty vector control cells. Invasion was measured at 48 hours. Data shown in (A,B) are pooled from two separate experiments. Values represent mean ± S.D. (n=4); * P< 0.01.

Figure 9. Model of HOXC8 action in human prostate cells HOXC8 acts as a transcription factor to control gene expression in conjunction with its partner Pbx (left) possibly to mediate increased invasiveness and other pro-tumorigenic actions of HOXC8. HOXC8 simultaneously inhibits the androgen-dependent recruitment of SRC-3 and subsequent recruitment of CBP to ARE-regulated genes (right)., resulting in decreased transcription of direct AR target genes.

Figure S1. HOXC8 expression in PCa cells lines (A) Real-time reverse transcriptase-PCR analysis of HOXC8 mRNA in PCa cell lines. Data shown are from a single experiment and are representative of three separate experiments. Values represent mean ± S.D. (n=3). HOXC8 mRNA levels were normalized to 18s. (B) Western blot analysis of HOXC8 protein levels in PCa cell lines.

Figure S2. HOXC6 and HOXB13 overexpression inhibit AR-mediated gene induction in LNCaP PCa cells The experiment was done exactly as described for figure 1B but HOXC6 or HOXB13 expression vectors replaced HOXC8 vector. Data shown are from a single experiment and are representative of three separate experiments. Values represent mean ± S.D. (n=4).

Figure S3. HOXC8 expression in LNCaP and HPr-1 AR cell lines stably overexpressing HOXC8 Real time reverse transcriptase-PCR analysis of HOXC8 mRNA expression in LNCaP PCa (A) and non-tumorigenic HPr-1 AR (C) cells transduced with a HOXC8 retroviral vector or a similar vector without coding sequences (empty). HOXC8 mRNA levels were normalized to 18s. Data shown are from a single experiment and are representative of two separate experiments. Values represent mean ± S.D. (n=3); *P<0.005. Western blot analysis of HOXC8 (60 s exposure) and actin (1 s exposure) in LNCaP (B) and HPr-1 AR (D) cells. Quantification of HOXC8 bands normalized to actin is shown.

Figure S4. HOXC8 inhibition of AR-mediated signaling is not due to downregulation of AR Expression of AR mRNA (A) and protein (B) in LNCaP-Empty and LNCaP-HOXC8 cells lines.. Data shown are from a single experiment and are representative of two separate experiments. (A) Values represent mean ± S.D. (n=3). (B) Quantification of AR bands normalized to actin is indicated.

Figure S5. Western blot analysis of coactivator overexpression in LNCaP PCa cells LNCaP cells were co-transfected with a probasin-luciferase reporter, 100 ng HOXC8 expression vector (or empty vector control) and 100 or 500 ng of the indicated coactivator expression vector. Total transfected plasmid was held constant by balancing with empty expression vector. 100ng total cell lysate was loaded per lane.

Figure S6. SRC-3 expression levels are unchanged in LNCaP PCa cells stably overexpressing HOXC8 (A) Real time reverse transcriptase-PCR analysis of SRC-3 mRNA in LNCaP-HOXC8 and empty vector control cell lines. SRC-3 mRNA levels are normalized to 18s. Values represent mean ± S.D. (n=3). (B) Western blot analysis of SRC-3 (45 s exposure) and actin (1 s exposure) protein levels. Data shown are from a single experiment and are representative of three separate experiments.

Figure S7. HOXC8 does not bind SRC-3 or AR in vitro

GST pull-down experiments were performed using purified SRC-3 protein (A) or LNCaP cell extract (B) to examine direct binding interactions between HOXC8 and SRC-3 or AR. Data shown are from a single experiment and are representative of three independent experiments.

HOXC8 inhibits androgen receptor signaling in human prostate cancer cells by inhibiting SRC-3 recruitment to direct androgen target genes

Sunshine Daddario Axlund, James R Lambert and Steven K Nordeen

Mol Cancer Res Published OnlineFirst November 2, 2010.

Updated version Access the most recent version of this article at: doi:10.1158/1541-7786.MCR-10-0111

Author Author manuscripts have been peer reviewed and accepted for publication but have not yet been Manuscript edited.

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/early/2010/11/02/1541-7786.MCR-10-0111. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.