Steroid Receptor Coactivator-3 as a Potential Molecular Target for Cancer Therapy

Abstract

Introduction

Steroid receptor coactivator-3 (SRC-3), also called amplified-in-breast cancer-1 (AIB1), is an oncogenic coactivator in endocrine and non-endocrine cancers. Functional studies demonstrate SRC-3 promotes numerous aspects of cancer, through its capacity as a coactivator for nuclear hormone receptors and other transcription factors, and via its ability to control multiple growth pathways simultaneously. Targeting SRC-3 with specific inhibitors therefore holds future promise for clinical cancer therapy.

Areas covered

We discuss critical advances in understanding SRC-3 as a cancer mediator and prospective drug target. We review SRC-3 structure and function and its role in distinct aspects of cancer. In addition, we discuss SRC-3 regulation and degradation. Finally, we comment on a recently discovered SRC-3 small molecular inhibitor.

Expert opinion

Most targeted chemotherapeutic drugs block only a single cellular pathway. In response, cancers frequently acquire resistance by up-regulating alternative pathways. SRC-3 coordinates multiple signaling networks, suggesting SRC-3 inhibition offers a promising therapeutic strategy. Development of an effective SRC-3 inhibitor faces critical challenges. Better understanding of SRC-3 function and interacting partners, in both the nucleus and cytosol, is required for optimized inhibitor development. Ultimately, blockade of SRC-3 oncogenic function may inhibit multiple cancer-related signaling pathways.

1. Introduction

Nuclear hormone receptors (NRs) bind DNA at specific sites to regulate gene expression and impact many physiological processes. Upon binding to its ligand, a NR commonly undergoes a conformational change and forms a dimer. The liganded NR complex then translocates into nucleus, recognizes specific regulatory DNA sequences, and binds response elements upstream of target genes to activate gene transcription. By themselves, NRs cannot initiate optimal transcriptional activation. It is through interaction with coactivator proteins that hormone-activated NRs can direct the assembly and stabilization of a preinitiation complex that ultimately conducts the transcription of the target genes¹. The p160 steroid receptor coactivator (SRC) family, consisting of three members (SRC-1²/NCOA1, SRC-2/TIF2³/GRIP1⁴/NCOA2, and SRC-3/p/CIP⁵, RAC3⁶, AIB1⁷, ACTR⁸, TRAM1⁹ and NCOA3), is a key coactivator group for serving as a bridge between the hormone-activated NRs, other co-regulators, and the basal transcriptional machinery. These SRC proteins can also act as coactivators for non-NR transcription factors to regulate target gene transcription and impact multiple growth factor pathways. In general, SRC proteins sit at the nexus of multiple cancer signaling pathways that impact cancer initiation, growth, migration, invasion, metastasis and chemotherapeutic resistance. In this review, we will focus on SRC-3 and discuss the role of SRC-3 in these distinct aspects of carcinogenesis. We will then evaluate the advantages and feasibility of designing SRC-3 inhibitors as therapeutic agents for cancer.

2. Steroid receptor coactivator-3 (SRC-3)

SRC-3 was initially found to be amplified and overexpressed in human breast cancer cell lines and a subset of breast tumors¹⁰. AIB1 (SRC-3) was identified during a search on the long arm of chromosome 20 for genes with overexpression and increased copy number in breast cancers¹⁰. Serving as an adapter to recruit chromatin remodeling proteins and other transcriptional enzymes, SRC-3 is known to mediate transcriptional activities of NRs such as ER (estrogen receptor) and PR (progesterone receptor). SRC-3 was found to promote hormone-dependent growth of human MCF-7 breast cancer cells by coactivating ERα and PRβ¹¹. SRC-3 also potentiates the transcriptional activities of other TFs (transcription factors) such as E2F-1, PEA3, AP-1 and NF-κB (nuclear factor-kappa B)^12–15. SRC-3 shares a common structure with the other members of the p160 family, containing domains with well-studied functional relevance. The structural and functional domains of SRC-3 are summarized in Figure 1. The N-terminal region is the most conserved among p160 family members and consists of bHLH (basic helix-loop-helix) region and a PAS (Per/ARNT/Sim) motif^{16, 17}. This region is necessary for several protein-protein interactions, including association with TFs such as myogenin and MEF2C^18–20. In addition, the bHLH-PAS region contains multiple nuclear localization signals²¹. A region between the bHLH-PAS domain has been linked to the turnover and degradation of SRC-3²². The central region of the SRC proteins contains three LXXLL (L, Leucine; X, any amino acid) motifs which form amphipathic α-helices and are essential for direct interactions with NRs in a ligand-dependent manner^23–26. Finally, the C-terminal region contains two transcriptional activation domains (AD1 and AD2). The AD1 domain directly binds to CBP (CREB-binding protein) and p300 proteins²⁷. The recruitment of CBP/p300 to the chromatin by SRC proteins results in histone acetylation necessary for SRC-mediated transcriptional activation. The AD2 domain interacts with histone methyltransferases such as CARM1 (co-activator-associated arginine methyltransferase 1) and PRMT1 (protein arginine N-methyltransferase 1) to promote histone methylation and subsequently facilitate chromatin remodeling^{28, 29}. Interestingly, the C termini of SRC-1 and SRC-3 also contain HAT (histone acetyltransferase) activity³⁰. These structural elements allow SRC proteins to provide a platform through which transcription factors can interact with additional coregulators that promote chromatin remodeling and assembly of general transcription machinery. In sum, their structure underlies the ability of SRC proteins to coordinate signals from a myriad of cellular signaling pathways in regulating gene expression output.

Figure 1. Structural and Functional Domains of SRC-3 protein.

The letters within the bar figure indicates functional domains. bHLH, basic helix-loop-helix domain; PAS, Per/ARNT/Sim homologous domain; S/T, serine/threonine-rich region; L, LXLL α -helix motif; AD1, Activation Domain 1; Q, glutamine-rich domain; AD2, Activation Domain 2. AD1 contains CID (CBP/p300 interacting domain) and AD2 contains HAT (histone acetyltransferase domains). The proteins listed below the bar figure are known to interact with SRC-3. This list is incomplete.

SRC proteins exist at limited concentrations in normal physiology. They are considered “master regulators” of differential gene expression and accomplish this through combinatorial codes of post-translational modifications (PTMs). Known SRC-3 PTMs, the responsible modifying proteins, and the modified sites are listed in Table 1. Extracellular stimuli such as hormones, growth factors and cytokines activate signaling pathways that may post-translationally modify SRCs through phosphorylation, ubiquitylation, sumoylation etc. The combinatorial codes of PTMs determine protein stability, interaction specificity and transcriptional activity of SRCs. Deregulation of these PTMs has a significant impact on cellular physiology and results in human diseases such as cancer. Early observation that SRC-3 localization and transcriptional activity could be regulated by IKKβ (IκB kinase β) phosphorylation provided the initial clue that SRC-3 is subject to PTM³¹. SRC-3 contains at least eight specific phosphorylation sites³². Seven Serine/Threonine (Thr24, Ser 505, Ser 543, Ser 601, Ser 857, Ser 860 and Ser 867) phosphorylation sites and one Tyr (Tyr 1357) phosphorylation site have been demonstrated to be functionally important^{33, 34}. These sites are phosphorylated by a number of different kinases including MAPK, IKK, GSK3α, GSK3β, and CK1d. SRC-3 is also a target of ABL tyrosine kinase which can be activated by estrogen and growth factors. Phosphorylation of Tyr1357 on SRC-3 by ABL tyrosine kinase increases binding of SRC-3 to p300 and transcription factors, thus mediating ER, PR and NF-κB-dependent transcription activities³⁴. Furthermore, Tyr1357 phosphorylation has been shown to increase in ERBB2-induced breast tumors in mice, suggesting an oncogenic function of SRC-3 with Tyr 1357 phosphorylation. In conclusion, phosphorylation can convert inactive SRC-3 into a potent transcriptional coactivator, resulting in differential gene expression with relevance to the development of cancer.

Table 1.

SRC-3 is subject to multiple post-translational modifications

PTM	SRC-3 Modifier	Effect on SRC-3	Modified Residue	Ref
Acetylation	CBP/p300	Inactivation	K626, K629, K630	89
Dephosphorylation	PP2A	Inactivation	S505, S543	22
	PP1	Stabilization	S101, S102	22
Methylation	CARM1	Inactivation, then Degradation	R1171	90
Phosphorylation	JNK	Activation	T24, S505, S543, S860, S867	32
	ERK	Activation	S505, S543	32
	p38MAPK	Activation	T24, S505, S543, S860, S867	32
	PKA	Activation	S857	32
	c-Abl	Activation	Y1357	34
	IKK	Activation	S857	32
	GSK3β	Activation, then Degradation	S505	77
	aPKC	Stabilization	S/T sites in a.a. 1031–1130 region	91
Ubiquitinylation	E6-AP	Degradation	Unknown	78
	SCFFbw7α	Degradation	K723, K786	77
	CUL-3 and RBX1	Degradation	Unknown	80
	CHIP	Degradation	Unknown	81
	SPOP	Degradation	N-terminal	83
SUMOylation	SUMO-1	Inactivation	K731	92

Notes for Table 1: Created by proteins from multiple intracellular signaling pathways, these modifications constitute a combinatorial code through which SRC-3 can integrate pathway-specific information to coordinate cellular outcomes. CBP, CREB-binding protein; PP2A, protein phosphatase 2A; PP1, protein phosphatase 1; CARM1, coactivator-associated arginine methyltransferase 1; JNK, c-Jun N-terminal kinase; ERK, extracellular-signal-regulated kinase; p38MAPK, p38 mitogen-activated protein kinases; PKA, protein kinase A; c-Abl, Abelson tyrosine kinase; GSK3β, glycogen synthase kinase 3; IKK, IκB kinase; aPKC, atypical protein kinase C; E6-AP, E6-associated protein; SCF, SKP1–cullin-1–F-box; Fbw7a, F-box and WD repeat domain-containing 7; CUL-3, Cullin-3; RBX1, RING box protein 1; CHIP, carboxyl terminus of Hsc70-interacting protein; SPOP, speckle-type POZ protein.

3. SRC-3 impacts multiple axes in cancer

The role of SRC-3 as a master regulator in cellular growth and development places it at the nexus of several intracellular signaling pathways critical for cancer (Figure 2). As SRC-3 amplification and overexpression have been correlated in multiple clinical studies with tumor aggressiveness or poor patient outcome (Table 2), it has become imperative to perform detailed mechanistic studies on the role of SRC-3 in tumorigenesis and cancer metastasis.

Figure 2. SRC-3 integrates multiple signaling pathways.

Depending on the cellular context, SRC-3 can coactivate different nuclear receptors and transcription factors to promote multiple hallmarks of cancer. SRC-3 classically facilitates the transcriptional activities of nuclear receptors to promote cell growth. It also coactivates E2F-1 to promote cell division. To impact on tumor expansion, SRC-3 serves as a coactivator for AP1 to upregulate IGF-Akt pathway and may be involved in HER2 pathway. SRC-3 also acts as a pro-metastatic factor by coactivating PEA3 and AP1 to upregulate MMP production, a requisite process in extracellular matrix breakdown that accompanies invasive tumor behavior. A spliced SRC-3 isoform located in cytoplasm can also contribute to metastasis by serving as an adaptor for FAK-Src signal pathway. SRC-3 has been implied to bind NFκB and upregulate MIF through the HIF transcription factor. This, in turn, contributes to chemoresistance, as well as angiogenesis through the FGF signaling pathway. NR, nuclear receptor such as ER, PR or AR. The solid arrow indicates direct interaction or known pathways while the dotted arrow indicates indirect interaction.

Table 2.

SRC-3 overexpression and amplification are associated with aggressive tumor phenotype or poor prognosis in human carcinomas of endocrine and non-endocrine organs.

Cancer type	Change	Detection	Frequency	Clinical Associations	Ref
Endocrine Carcinomas
Breast	Gene amplification	FISH	9.50%	Undetermined	10
	mRNA Overexpression	ISH	64%	Positive correlation between expression and tumor size	10
	Protein Overexpression	WB	25%	Positive correlation between expression and tamoxifen resistance	62
	Protein Overexpression	WB	25%	Highly correlated with proliferation and expression of ERBB2 and PR	62
	Protein Overexpression	IHC	73.8%	Positive correlation between expression and MMP2, MMP9, and PEA3 protein levels	13
Prostate	mRNA Overexpression	ISH	ND	Positive correlation between expression and tumor grade	93
	Protein Overexpression	IHC	ND	Positive correlation between expression and tumor grade	93, 94
Endometrial	mRNA Overexpression	qPCR	17%	Undetermined	95
	Protein Overexpression	IHC	97%	Associated with poor prognosis	96
Ovarian	Gene Amplification	FISH	25%	Associated with DNA amplification at 20q12-q13	97
	Protein Overexpression	IHC	64%	expression of SRC-3 positively correlates with invasiveness of tumor	44
Non-Endocrine Carcinomas
Esophageal squamous cell carcinoma	Gene Amplification	FISH	4.3–4.9%	Associated with chromosome 20q amplification and increased copy number	98, 99
	Protein Overexpression	IHC	46%	Associated with large tumor size and Ki67 proliferation staining	99
Gastric	Gene Amplification	FISH	7%	Amplification positively correlated with metastasis	100
	mRNA Overexpression	Northern, qPCR	40%		100
Colorectal	Gene Amplification	FISH	10%	Nuclear expression positively correlated with metastasis	101
	mRNA Overexpression	IHC	35%		101
Pancreatic	Gene Amplification	FISH	37%	Undetermined	102
	mRNA Overexpression	ISH	>65%		102
	Protein Overexpression	IHC	>65%		102
Hepatocellular	Gene Amplification	FISH	25%	Undetermined	103
	Protein Overexpression	WB	68%	Undetermined	104
Urothelial	Gene Amplification	FISH	7%	Undetermined	105
	Protein Overexpression	IHC	33%	Overexpression positively correlated with Ki67 proliferation staining	105
Nasopharyngeal	Gene Amplification	FISH	7%	Overexpression positively associated with large tumor and Ki67 index	106
	Protein Overexpression	IHC	51%	Undetermined	106

Notes for Table 2: FISH, fluorescence in situ hybridization; ISH, in situ hybridization; WB, western blot; IHC, immunohistochemistry; qPCR, quantitative PCR.

3.1 SRC-3 in cancer initiation and tumorigenesis

The proliferative role of SRC-3 in primary tumor formation has been extensively studied. Many mouse models have been employed to investigate the function of SRC-3 in cancer, particularly breast cancer. In transgenic mouse models where overexpression of SRC-3 in mammary epithelial cells was driven by MMTV (mouse mammary tumor virus) promoter, mammary hyperplasia and spontaneous development of mammary tumors were observed, directly supporting the role of SRC-3 in breast cancer initiation³⁵. Hyperactive IGF-1 signaling was found in these MMTV-SRC-3 transgenic mice³⁵. Consistent with results in this model, when SRC-3 knockout mice were crossed with MMTV-v-ras transgenic mice, mammary tumor incidence and growth rate were reduced dramatically in SRC-3−/−;MMTV-v-ras bigenic virgin mice and inhibited completely in ovariectomized SRC-3−/−;MMTV-v-ras bigenic mice³⁶. Further molecular analysis indicated SRC-3 deficiency did not alter estrogen and progesterone-responsive gene levels, but decreased IRS-1 (insulin receptor substrate-1) and IRS-2, resulting in impaired IGF-1 signaling pathway³⁶. Along the same lines, ablation of SRC-3 in mice treated with chemical carcinogen DMBA (7,12-dimethylbenz[α]anthracene) protected against mammary gland tumorigenesis³⁷. Another mouse model was used to assess the impact of SRC-3 ablation on MMTV-Erbb2 (erythroblastosis oncogene B 2)-induced mammary tumors. Knockout of SRC-3 in MMTV-Erbb2 mice completely suppressed tumorigenesis and reduced levels of phosphorylated ERBB2, cyclin D1 and cyclin E³⁸. All these findings suggest that SRC-3 plays an important role in tumor initiation and growth and that targeting SRC-3 can effectively suppress tumor initiation and growth.

In addition to the full-length SRC-3, a splice variant of SRC-3 (SRC-3delta4) which lacks the N-terminal bHLH-PAS domain has been identified and found to be an even more potent coactivator for ER and PR than the full-length SRC-3³⁹. Transgenic mice overexpressing SRC3delta4 develop mammary gland hyperplasia, increased cyclin D1 expression, and increased IGF-1 receptor level⁴⁰. Furthermore, combined overexpression of SRC3delta4 and ERα in mouse mammary gland resulted in increased incidence of hyperplasia and adenocarcinoma with increased stromal collagen deposition⁴¹. SRC-3delta4 also plays an important role in cancer metastasis and its function will be discussed in detail in a later section.

Besides breast cancer, SRC-3 was also found to be important for tumorigenesis in the prostate. SRC-3 knockout mice were crossed with TRAMP (transgenic adenocarcinoma mouse prostate mice), a mouse prostate cancer model. Total ablation of SRC-3 in TRAMP mice arrested prostate tumor growth at well-differentiated stages. Even though initiation of prostate tumors in this mouse model was not delayed, the progression of prostate tumorigenesis substantially declined⁴².

3.2 SRC-3 in cell motility, invasion, and metastasis

In order for metastasis to manifest, cancer cells need to gain motility and invasive potential that will allow them to escape the primary tumor site, invade surrounding stroma and enter the blood stream. The first evidence that SRC-3 plays a role in cell migration and invasion comes from the studies of fruit fly ovary⁴³. In the absence of Taiman, the Drosophila homolog of SRC-3, ecdysone receptor-dependent border cell motility and invasiveness were markedly suppressed and there was an abnormal cellular build-up of E-cadherin, β-catenin, and focal adhesion complexes⁴³. Subsequently, in human ovarian cancer cells, SRC-3 was demonstrated to be important for cellular spreading migration on the substratum⁴⁴. The most relevant in vivo study of SRC-3 function in metastasis originates from MMTV-polyoma middle T antigen (PyMT) transgenic mouse model. Absence of SRC-3 in PyMT transgenic breast cancer mouse model significantly suppressed mammary tumor metastasis to the lung¹³. Molecular studies showed that SRC-3 could impact the expression levels of MMPs (matrix metalloproteinases) that allow tumor cells to break down the extracellular matrix and invade into stromal compartment. In both human (MDA-MB-231) and PyMT tumor cells in culture, SRC-3 regulates MMP2 and MMP9 by directly binding and potentiating activity of the PEA3 transcription factor¹³. SRC-3 also serves as a coactivator for AP-1 (activator protein 1) to drive expression of MMP7 and MMP10 in MDA-MC-231 human breast cancer cells⁴⁵. Furthermore, in prostate cancer, SRC-3 simultaneously coactivates AP-1 and PEA3 to upregulate expression of MMP2 and MMP13¹⁵. By serving as a coactivator for a number of TFs responsible for the expression of MMP family members, nuclear SRC-3 promotes invasion of cancer cells into the surrounding stromal compartment. Recently, in a lung cancer cell line, ERK3 was shown to phosphorylate SRC-3 at S857, a modification essential for the binding of SRC-3 with PEA3 and promotion of MMP gene expression⁴⁶. In sum, SRC-3 clearly promotes cancer invasion by coactivating non-nuclear receptor TFs to regulate MMP gene expression.

SRC-3delta4 also has been implicated in promotion of metastasis. SRC-3delta4 is mainly sequestered in the cytosol and acts as a signal adaptor for EGFR (epidermal growth factor receptor) and FAK (focal adhesion kinase-1) at the plasma membrane⁴⁷. EGF is a critical mediator for cancer cell migration and metastasis⁴⁸. Extracellular EGF binds EGFR on the cell membrane and activates a number of intracellular protein kinases including PAK1 (p21-activated kinase 1)⁴⁹, FAK⁵⁰ and c-Src⁵¹ in a cascade signaling fashion. SRC-3delta4 serves as a bridge between EGFR and FAK to allow optimal activation of EGF-FAK-cSrc signal transduction⁴⁷. Activation of this signal pathway promotes the movement and invasion of cancer cells.

3.3 SRC-3 in inflammation and angiogenesis

Persistent inflammation is a characteristic of the tumor microenvironment and is recognized as a hallmark of cancer. Inflammation can promote proliferation and survival of cancer cells, facilitate angiogenesis and metastasis, destabilize adaptive immunity, and reduce response to hormone therapy and chemotherapy⁵². As in other inflammatory contexts, accumulating evidence shows NF-κB is a key mediator of tumor inflammation. SRC-3 has been shown to interact with and coactivate NF-κB in HeLa cancer cells¹⁴. In response to TNF-α(tumor necrosis factorα), SRC-3 is phosphorylated by IKK in cytosol of HeLa cells³¹. SRC-3 translocates along with NF-κB into the nucleus where, aided by SRC-3, NF-κB can bind promoters of target genes and promote the initiation of inflammatory responses. One important NF-κB target gene is IL-6 (interleukin-6), a pro-inflammatory cytokine that plays an important role in tumor metastasis and inflammation⁵³. IL-6 is elevated in prostate cancer tissues and acts as an autocrine growth factor in prostate cancer^{54, 55}. This suggests SRC-3 can serve as a coactivator for NF-κB to promote inflammation in cancer. Demonstrating the context specificity of inflammatory responses, SRC-3 knockout mice are actually more susceptible to acute inflammatory responses than controls⁵⁶. SRC-3 knockout macrophages are more sensitive to LPS (lipopolysaccharide)-induced endotoxic shock and they produce more proinflammatory cytokines. In these cells, SRC-3 was demonstrated to bind translational repressors TIA-1 (T cell intracellular antigen-1) and TIAR (TIA1-related protein) to inhibit translation of TNF-α, IL-6, and IL-1⁵⁶. In addition, SRC-3 is required for clearing bacteria, and represses inflammatory response in E. Coli-induced septic peritonitis⁵⁷. These findings suggest that the role of SRC-3 in NF-κB-mediated cytokine expression may be specific to cell types.

Angiogenesis is another important process in cancer progression because the growth of a tumor relies on a sufficient blood supply. Many studies have been focused on investigating the function of SRC-3 in growth factor signaling, with the main focus on cell-autonomous regulation of proliferation and invasive capacity. However, less is known about the SRC-3 function in stroma. A recent study elucidates the role of SRC-3 in angiogenesis and wound healing⁵⁸. SRC-3 was shown to promote proliferation and motility of endothelial cells, such that neoangiogenesis was dependent on the presence of SRC-3⁵⁸. The study also demonstrated that both alleles of SRC-3 were required for proper wound healing in vivo and that SRC-3 may cross-talk with FGF (fibroblast growth factor) signaling to regulate wound healing process⁵⁸.

The tumor microenvironment plays a critical role in cancer progression. By immunohistochemistry, SRC-3 protein expression is found in stromal compartment. However, the in vivo function of SRC-3 in tumor microenvironment has not been clearly defined due to lack of appropriate models. Future investigation of SRC-3 in tumor microenvironment can be aided by the generation of mice with floxed SRC-3 alleles⁵⁹, so that SRC-3 may be deleted in specific cell types with relevance to these novel functions.

3.4 SRC-3 in endocrine therapy-resistant cancer and chemoresistant cancer

SRC-3 has been implicated in the development of resistance to chemotherapeutic agents. Tamoxifen is an antagonist that competes with estrogen for binding to ER, resulting in the inhibition of ER-mediated transcription and thus estrogen dependent cancer growth. Tamoxifen has been the standard endocrine therapy for women with ER-positive breast cancer. However, only 50% of ER-positive breast cancer patients respond to tamoxifen therapy⁶⁰. Other patients treated with tamoxifen for long periods tend to acquire resistance to the therapy. Resistance to endocrine therapy often has been associated with activation of growth factor signaling pathways such as EGFR pathway⁶¹. There is a positive correlation between SRC-3 protein expression and the levels of HER family proteins in the breast cancer patients with recurrence after tamoxifen treatment⁶². Recently, it was demonstrated PAX2 (paired box gene 2) competes with SRC-3 for binding and regulation of HER2 transcription. High SRC-3 expression was associated with high recurrence rate in patients with ER-positive tumors and treated with tamoxifen⁶³. Another class of endocrine therapeutic agent, aromatase inhibitors, acts by blocking conversion of testosterone and androstenedione into estrogen. Aromatase inhibitors are used to treat postmenopausal women with ERα-positive breast cancer. However, breast tumors with high HER2 and SRC-3 expression may also develop resistance to aromatase inhibitors, as its family member SRC-1 does⁶⁴.

Bortezomib (PS-314 or Velcade) is a proteasome inhibitor that has anti-cancer activity in various cancer cell lines including prostate cancer cell and prostate cancer xenograft models. In a neoadjuvant clinical trial of bortezomib in men with prostate cancer at high risk of recurrence, unexpected increase in proliferation in treated tissues and cultured cells was found⁶⁵. In these treated tissues and cell lines, SRC-3 level and phosphorylated Akt level were found to be increased. Knockdown of SRC-3 decreased the level of the phosphorylated Akt. These data suggest that SRC-3 may contribute to chemo-resistant prostate cancer⁶⁵.

A recent paper identified MIF (macrophage migration inhibitory factor) as a new target gene of SRC-3 and demonstrated MIF is a suppressor of autophagic cell death⁶⁶. Upregulation of MIF expression by SRC-3 in cancer cells can contribute to chemoresistance. Inhibition of MIF expression can sensitive cancer cells to anti-cancer drugs such as doxorubicin and etoposide⁶⁶.

4. SRC-3 in development, metabolism and other physiological process

Genetically engineered mouse models have been employed to study the physiological relevance of steroid receptor coregulators. Much of our understanding about SRC-3 function in vivo stems from characterization of SRC-3 knockout mice. Targeted deletion of SRC-3 in mice has revealed its critical role for normal somatic growth, mammary gland development and female reproduction^{67, 68}. Circulating IGF-1(insulin-like growth factor-1) level was found to be significantly reduced in SRC-3 knockout mice⁶⁹. All three members of SRC family play a critical role in metabolic regulation. In particular, loss of SRC-3 impairs white adipogenic differentiation through decreased PPARγ2 (peroxisome proliferator-activated receptor-γ activity)⁷⁰. SRC-3 knockout mice are resistant to high-fat diet-induced obesity and have improved insulin sensitivity⁷¹. The phenotype was partly due to the regulation of PGC-1 (peroxisome proliferator-activated receptor-γ coactivator-1) acetylation by SRC-3⁷¹. Furthermore, a knock-in mouse model with mutations at four conserved phosphorylation sites displayed increased body weight and adiposity, and reduced peripheral insulin sensitivity. These mice were also more susceptible to carcinogen-induced liver tumorigenesis. These results support the idea that PTMs are important for the normal function of SRC-3 and that changes in PTMs are sufficient to alter glucose homeostasis and cancer susceptibility⁷². Because all cancer cells rely on changes in metabolism to support growth and survive, targeting metabolism for anti-cancer therapy has been a recent focus in cancer research. Owning to its close relevance to metabolism, the exploration on the function of SRC-3 in the regulation cancer metabolism might provide some insights into successful cancer therapy.

5. Regulation of SRC-3 mRNA/protein levels

SRC-3 expression and protein amount can be regulated at three different levels: Transcription, translation and protein degradation. SRC-3 is a coactivator for E2F-1^{12, 73} and SRC-3 gene contains E2F-1 binding sites at its promoter region suggesting SRC-3 can self-regulate and form a positive feedback loop for its own expression¹². Another binding site on the SRC-3 promoter region is for the SP1 (Specificity Protein 1) transcription factor. Interestingly, E2F-1-dependent transcription of SRC-3 did not require E2F-1 binding to its binding site but rather the binding of SP1 to the SP1 binding site in a proximal SRC-3 promoter region. At translational level, SRC-3 can be regulated by miRNAs (microRNAs). Endogenous miRNAs bind on site-specific sequences within the 3-untranslated regions and inhibit the translation. miRNA Mir-17-5p was found to specifically inhibit the translation of SRC-3 mRNA. In breast cancer cell lines with high levels of SRC-3 protein, Mir-17-5p was found to be at low levels ⁷⁴.

Specific post translational modifications such as phosphorylation and methylation serve as a code that mediates SRC-3 interaction, function and degradation⁷⁵. SRC-3 protein turnover is mediated by proteasomal degradation pathways⁷⁶ and the NLS (nuclear localization signal) within the bHLH domain of SRC-3 is critical for this proteasome-dependent turnover ²¹. In the ubiquitin-dependent proteasome degradation, ubiquitin molecules are linked to the target proteins by E3 ligases. The ubiquitinylated proteins are then degraded by the 26S proteasome in an ATP –dependent manner. Both SCF^Fbw7α and E6-AP are examples of E3 ligase that can interact with SRC-3, targeting SRC-3 for degradation^{77, 78}. REGγ, 20S proteasome regulator, can also interact with SRC-3 and mediates its turnover in an ubiquitin-independent manner⁷⁹. Recently, the components of E3 ligase, CUL1 and RBX1, were shown to be involved in SRC-3 ubiquitinylation and degradation, in response to retinoid acid treatment⁸⁰. CHIP (carboxyl terminus of Hsc70-interacting protein) is a U-box-type ubiquitin ligase that induces ubiquitinylation and degradation of its substrates. SRC-3 was also found to be a target of CHIP and knockdown of SRC-3 reduces Smad and Twist expression⁸¹. In human hepatocellular carcinoma, Hepatitis B virus X protein (HBx) stabilizes SRC-3 so SRC-3 cannot be targeted for degradation by E3 ligase⁸². Recently, speckle-type POZ protein (SPOP), a cullin 3 (CUL3)-based ubiquitin ligase, was found to promote SRC-3 ubiquitinylation and degradation⁸³. Interestingly, loss-of-function mutations of SPOP were identified in 6–13% of human prostate cancers that do not contain PTEN mutation or TMPRSS2:ERG fusion rearrangement⁸⁴. In addition, a recent study of exome sequencing in 112 human prostate tumor and normal tissue pairs identified SPOP as one of the most frequently mutated gene in 13% of prostate tumors and all the SPOP mutation affected in the structure that involves in substrate-binding function⁸⁵. Thus, it would be interesting to find out whether SPOP mutation is associated with SRC-3 protein elevation and human prostate carcinogenesis.

6. Development of an SRC-3 inhibitor

A specific SRC-3 inhibitor has not yet been generated. This is largely due to an incomplete understanding of the protein’s structure and lack of crystallography data. By employing high throughput screening assays, the O’Malley lab recently identified gossypol as a small molecule inhibitor of SRC-1 and SRC-3⁸⁶. Gossypol, a compound derived from the cotton plant, has been shown to partially inhibit SRC-3 function in cell culture, stemming the growth of cancerous but not non-cancerous cells. Gossypol selectively reduces SRC-1 and SRC-3 protein levels in cancer cell lines including breast, prostate, lung, and liver cells. In fact, gossypol could reduce cell viability while also sensitizing lung and breast cancer cell lines to other chemotherapeutic agents such as MEK (MAPK kinase) inhibitor and EGFR inhibitors. Identified as an inhibitor for Bcl-2, gossypol was already demonstrated to be a proapoptotic agent for cancer cells and is currently being evaluated as a therapeutic agent for prostate cancer and lung cancer in clinical trials^{87, 88}. O’Malley’s group identified that gossypol could directly bind to SRC-3 and reduce its protein levels without affecting its mRNA level. They also demonstrated gossypol could reduce mRNA levels of Bcl2. To further explain, genomic study of SRC-3 cistrome in MCF-7 identified several SRC-3 binding sites within Bcl-2 and Bcl-X genes. Knockdown of SRC-3 in MCF-7 cells decreased both Bcl-2 mRNA and protein levels. These findings further support that the anti-cancer mechanism of gossypol is through downregulation of SRC-3 protein levels, impairing anti-apoptotic pathways in cancer cells.

In pre-clinical and clinical studies, gossypol has had mixed results as a chemotherapeutic agent in the treatment of serious malignancies like small cell lung and adrenal cancers. Although gossypol has its own weaknesses including in vivo toxicity to serve as an effective therapeutic cancer drug in human^{87, 88}, the identification of gossypol to be a small molecule inhibitor of SRC-1 and SRC-3 is a proof-of-principle study that oncogenic coactivators can be directly targeted for inhibiting cancer growth. Taken together, gossypol studies suggest direct inhibition of SRC-3 by small molecular inhibitors is possible and has specific impact on cancerous cells. Nonetheless, the failure to achieve near complete SRC-3 inhibition, even in culture, indicates we do not yet understand the full potential for SRC-3 inhibition in modulating disease. Given promising results of SRC-3 knockout in animal models, it will be critical to search for improved inhibitors whose efficacy against real tumors can be evaluated.

7. Expert opinion

Most chemotherapeutic drugs have been designed to target one particular growth factor pathway. For examples, tamoxifen is an antagonist for ERα in breast cancer while Herceptin and Lapatinib target to inhibit ERBB2. Even though these drugs are effective initially, cancer cells eventually upregulate different growth factor pathways to acquire resistance. Therefore, it is important to rationally design a drug that targets multiple pathways simultaneously. SRC-3 is a potential therapeutic target that impacts multiple growth pathways. SRC-3, like its family members, has been shown to play important roles in many aspects of cancer. SRC-3 is an oncogene in that its overexpression is associated with cancer initiation, progression, invasion, metastasis, and chemoresistance. Detailed knowledge of transcription regulation mediated by SRC-3 has been acquired through many in vitro studies, while the functional relevance of SRC-3 in multiple cancer pathways has been illustrated in vivo. It is well established that SRC-3 functions as a coactivator for NR and promotes NR-dependent cell proliferation. SRC-3 can also affect NR-independent cancer motility and invasion by serving as a coregulator for other TFs, such as E2F-1, AP1 and PEA3. Functions of SRC-3 outside its capacity as a transcription regulator have also been demonstrated. SRC-3 was identified as a translational corepressor for TIA-1/TIAR, which inhibits the production of pro-inflammatory cytokines, while SRC-3delta 4 interacts with EGFR and FAK1 to regulate cell invasion and migration. Together, these findings suggest SRC-3 functions to promote many aspects of carcinogenesis and impact multiple cancer pathways (Figure 2). Furthermore, SRC-3 knockout mice have elucidated the physiological relevance of SRC-3 in multiple cancers and served as models for preclinical trials of SRC-3 inhibition. Genetic ablation of SRC-3 in both breast and prostate cancer mouse models inhibits tumorigenesis and blocks metastasis. Given the fact that some SRC-3 knockout mice survived through embryonic and neonatal stages and all survived beyond these early stages have a nearly normal life span, specific inhibition of SRC-3 function may be an ideal approach to control cancer growth without severe side effect.

Ideal therapeutic agents should selectively kill tumor cells while sparing surrounding normal cells. SRC-3 is present at limiting concentration in normal cells and overexpressed in cancer cells. Overexpression of SRC-3 in cancer cells provides a growth advantage, such that cancer cells become “addicted” to SRC-3. Therefore, an SRC-3 inhibitor can theoretically target cancer cells to a greater degree than normal cells. Inhibition of SRC-3 may upregulate cytokine production in some cells of the innate immune, rendering increased risk of cytokine storm⁵⁶. While altered immunity is a potentially serious side effect, the normal viability and health of SRC-3 knockout mice suggests SRC-3 inhibition may be relatively safe.

Many important questions remain to be addressed. First, SRC-3 functions outside transcriptional regulation appear multiple, but remain poorly understood. Second, many in vivo studies of SRC-3 have been carried out in SRC-3 total knockout mice. However, the roles of SRC-3 in specific cell types such as epithelial vs. stromal cells remain to be investigated. It is important to utilize conditional knockout mice to study cell-specific functions of SRC-3 in carcinogenesis. Temporal deletion of overexpressed SRC-3 in mouse models may also provide insights as the stages of cancer progression for which SRC-3 is most relevant. In addition, clearly mapping genes regulated by SRC-3 is important. Even though a small molecule inhibitor of SRC-3 has been identified, better and improved inhibitors for SRC-3 are still necessary. To develop such inhibitors, a crystal structure of SRC-3 protein may be advantageous (although analysis of large proteins like SRC-3 in this manner is extremely challenging). Other detailed research on the regulation of SRC-3 function and degradation may also provide insights for rational design of SRC-3 inhibitors.

The ultimate goal of future research is to block SRC-3 oncogenic function and inhibit multiple cancer-related signaling pathways. However, better knowledge of structure, interaction partners and the manner in which these interaction partners change during cancer progression will be important areas of research.

Article Hightlights Box.

• SRC-3 has been linked to various aspects of carcinogenesis ranging from cancer initiation, progression, cell motility and invasion, inflammation, angiogenesis and resistance to chemotherapy.

• SRC-3 can undergo post-translational modification (PTM) and PTM is responsible for its diverse functions.

• SRC-3 promotes cancer by activating nuclear receptors (NRs) such as ER and AR and facilitating transcription of multiple transcription factors. Thus SRC-3 sits at the nexus of multiple cellular pathways to promote cancer.

• SRC-3delta4, a splicing isoform of SRC-3, has been identified to play an important role in cell invasion and metastasis.

• SRC-3 can be regulated at the mRNA level by transcriptional regulation and microRNA and at protein levels by protein degradation pathways. Knowledge of SRC-3 regulation provides insights into designing small molecules for inhibiting SRC-3 function.

• Gossypol was recently identified to be an SRC-3 inhibitor that can degrade SRC-3 at protein level.