Background
Id4 (inhibitor of differentiation-4), is a member of the inhibitor of differentiation (Id) gene family (Id1, Id2 and Id3) and acts as a transcriptional regulator of basic helix-loop-helix (bHLH) family of transcription factors [
1]. Due to lack of the basic DNA binding domain, Id4 (and all Id proteins) acts as a dominant negative regulator of bHLH transcription factors, most notably E2A (TCF3) [
1,
2].
The interaction repertoire of Id proteins also involves several non-bHLH proteins. Whereas all Id proteins interact with bHLH TCF3, their interaction with non-bHLH proteins appears in large part to be isoform dependent- Id1: CASK, ELK1, GATA4, caveolin; Id2: ELK1, 3 and 4, CDK2, PAX2, 5 and 8, Rb and related pocket proteins, Id3: ELK1 and 4, ADD1 ([
1,
2] and public databases). Specific non-bHLH interaction partners for Id4 are currently not known. Thus Id proteins are capable of regulating the expression of a large number of genes through specific bHLH and non-bHLH interactions that in turn regulates many cellular processes such as cell growth, differentiation, and apoptosis [
3].
Id proteins are expressed by essentially all cell lineages at some point of development. In general, Id expression is highest in undifferentiated, proliferating populations and is down-regulated as cells exit from cell cycle and terminally differentiate (reviewed in [
1‐
3]). Knock out mouse models evaluating Id genes have demonstrated their essential role in development. Id2 null mice displays phenotypic abnormalities of retarded growth and neonatal morbidity due to a lactation defect [
4], impaired chondrogenesis [
5], B cell development [
6] and severe cardiac defects [
7]. Male Id2-/- mice also exhibit defects in spermatogenesis [
8]. Id3 null mice develops primary Sjögren’s syndrome-like symptoms [
9], specific defects in B/T lymphocyte development [
10], and restricted development of the gamma delta lineage during thymopoiesis [
11]. Interestingly, no phenotype is observed in mice lacking only Id1 suggesting that its function can be effectively compensated by the other three Ids. So far embryonic lethality has been observed only in mice homozygously lacking both Id1 and Id3 suggesting that Id1 and Id3 may have many overlapping functions [
12]. Id4 is required for normal brain size and lateral expansion of the proliferative zone in the developing cortex and hippocampus possibly by regulating neural stem cell proliferation and differentiation [
13]. Id4 is also required for normal mammary gland development in p38MAPK dependent pathway [
14] and for spermatogonial stem cell renewal [
15].
Studies have also shown that unlike other Ids, Id4 promotes differentiation in many systems including osteoblast [
16], adipocytes [
17], neurons [
13] and oligodendrocytes [
18]. Paradoxically, Id4 appears to demonstrate both pro-tumor and anti-tumor properties. Epigenetic silencing of Id4 in leukemia [
19], breast [
20,
21], colorectal [
22] mouse and human CLL (chronic lymphocytic leukemia [
23]) and gastric cancer [
24] tend to support its anti-tumor activity. Whereas high Id4 expression is reported in B-cell acute lymphoblastic leukemia [
25] and B-cell precursor acute lymphoblastic leukemia (BCP-ALL) [
26] due to the t(6;14)(p22;q32) chromosomal translocation, and in bladder [
27] and rat mammary gland carcinomas [
28] suggests that it may also have pro-tumor activity.
We and others have recently shown that Id4 is highly expressed in the normal prostate and decreased in prostate cancer due to promoter hypermethylation [
29,
30]. Id4 expression in the prostate thus appears in contrast with the expression of other Id genes (Id1 and Id3) which are expressed at low to negligible levels in the normal prostate although their expression increases significantly in prostate cancer [
31‐
33]. Moreover, Id4 is regulated by androgens in cells that respond to androgen stimulation such as testicular Sertoli cells and prostate epithelial cells [
34]. Id4 also restores androgen receptor expression and activity in the androgen receptor negative prostate cancer cell line DU145 [
35]. These results suggest that Id4 could potentially act within the androgen receptor pathway to regulate the development and function of the prostate. We used the Id4 -/- mouse model to evaluate further the role of Id4 in prostate development and its significance in prostate cancer. Our findings suggest that Id4 is required for normal prostate development. The prostate in Id4-/- mice have a complex phenotype characterized by attenuated growth and development that also mimics subtle features of prostatic intraepithelial neoplasia (PIN).
Discussion
This study supports a role for Id4 as a key regulator of male genital tract development. Although we focused on the prostate, the size and development of accessory sex glands (seminal vesicles) and testis is also severely impaired. Id4 may not be required to maintain fertility [
15] but it could cooperate with other possibly overlapping regulatory genes to support normal development of various organs within the genital tract.
Genital tract development in general and prostate in particular are androgen dependent. Prostate fetal development, structural and functional maturation at puberty is strictly androgen regulated [
55]. Loss of androgen receptor, specifically in the prostate epithelial cells (PEARKO, prostate epithelial AR knockout) leads to a phenotype [
56,
57] that is very similar to the Id4-/- prostates e.g. increased proliferation, decreased size and number of tubules and lack of differentiated epithelial cells. Based on the chromatin immuno-precipitation studies of the mouse Nkx3.1 promoter and increased NKX.3.1 expression in DU145 + Id4 cells, we propose that Id4 is required to maintain certain facets of androgen receptor activity in the prostate epithelium. In particular, Id4 could support the function of the AR as a suppressor of epithelial proliferation in the mature prostate, which is defective in prostate cancer [
58].
Nkx3.1 regulates early postnatal ductal morphogenesis and maintains normal differentiation of the prostate epithelium including the production of secretory proteins [
38,
39]. Similar to Nkx3.1-/- mice, the Id4-/- mice also display reduced ductal branching morphogenesis, epithelial hyperplasia and dysplasia. But unlike Id4-/- mice, the overall prostate sizes and wet weights in
Nkx3.1 -/- and +/+ mice [
39] are similar. Nevertheless, loss of Nkx3.1, a marker of epithelial differentiation and androgen response is a significant observation that further supports the attenuation of androgen regulatory network post androgen receptor expression in the Id4-/- prostates.
Nkx3.1 also regulates the rate at which proliferating luminal epithelial cells exit the cell cycle and its loss extends the transient proliferative phase of luminal cells [
59] which is consistent with increased expression of ki67, Myc and Id1 in Id4-/- prostate. An increase in the Myc:Nkx3.1 ratio observed in Id4-/- mice could also promote Myc dependent transactivation of pro-tumorigenic target genes [
47]. Conversely, a decrease in Myc:Nkx3.1 ratio may promote Nkx3.1 dependent transactivation of anti-tumorigenic target genes. Mice expressing Myc in the prostate also develop PIN like lesions followed by invasive adenocarcinoma [
60]. Inactivation of Pten also promotes cellular Myc activation [
42] which is consistent with our results. Thus, some of the phenotypes resulting from the loss of Nkx3.1 are consistent with the literature but the smaller prostate size in Id4-/- mice appears to result also from alterations of other regulatory pathways that could be independent of Nkx3.1 such as Akt signaling (see below).
Id1 is also a member of helix-loop-helix family of transcriptional regulators that contributes to cell proliferation and restrains differentiation and apoptosis [
61,
62]. Both Id1 and Id4 share strong sequence homology and interact with similar bHLH proteins for example TCF3, but their expression patterns are largely non-overlapping [
61]. We and others have shown that Id4 and Id1 expression is mutually exclusive in the normal prostate [
35] and prostate cancer [
29‐
31,
33,
50,
63]. Such a mutually exclusive expression pattern is also observed in the Id4-/- mice further suggesting loss of epithelial differentiation and increased proliferation. Sustained Id1 expression also failed to rescue the Id4-/- deficient phenotype supporting the argument that these two structurally similar proteins are functionally divergent and non-compensatory.
Sox9 is critical for maintaining the basal epithelial cells in tissues and may have a similar function in prostate epithelium [
64]. In the adult prostate, SOX9 is expressed diffusely in the basal cell layer suggesting that it is required for maintaining basal cell function. These basal cells represent and/or include prostate stem cells also [
65]. Increased Sox9 expression in the prostate epithelial component may suggest the expansion of this basal cell population that remains undifferentiated as evidenced by persistent Id1 expression, increased proliferation (lack of exit from cell cycle) and decreased differentiation markers (Nkx3.1). However direct studies identifying specific basal cell populations (e.g. p63 expression) and/or stem cell markers and there transitions to specific cell types will be required to further consolidate this specific mechanism.
Investigating whether loss of Id4 results in an early defect or is a later post-pubertal effect will be required to fully comprehend the scope of Id4 in the regulation of prostate development. Whether Id4 is vital to maintain a specific Sox9 positive prostate stem cell component that eventually expands to promote normal prostate development is an interesting proposition based on two different studies. First, Id4 is required for neuronal stem cell maintenance but a relatively mild mutant phenotype is observed at post natal day 0 despite the early loss of stem cells due to both premature differentiation and compromised cell cycle transition [
13]. Second, in mice lacking Id4 expression, quantitatively normal spermatogenesis is impaired due to progressive loss of the undifferentiated spermatogonial stem cell population during adulthood [
15]. These studies indicated that Id4 is a distinguishing marker of spermatogonial stem cells in the mammalian germline and plays an important role in the regulation of self-renewal. The observations made in the later study are particularly exciting given the overall impact of Id4-/- on the male reproductive tract and suggests a potential common molecular mechanism of action targeting a stem cell population in various organs of the male reproductive tract. In the prostate, Id4 could also be expressed in a specific stem cell population such as Sca-1
hi, Sca-1
lo, Sca-negative [
66] and/or their progenitors that contribute to the prostate phenotype in Id4-/- mice.
Loss of Id4-/- also impairs mammary gland development [
14]. In the mammary gland, Id4 expression is mainly observed in the cap cells, basal cells and in a subset of luminal cells, whereas in the prostate Id4, expression is primarily in the luminal epithelial cells. Conceptually, reduced ductal branching in prostate is similar to reduced ductal branching/expansion and branching morphogenesis in mammary gland of Id4-/- mice. In mammary gland, loss of Id4-/- is associated with reduced cellular proliferation but in the prostate, loss of Id4 was associated with increased proliferation (Ki67) and decreased differentiation (Nkx3.1) suggesting that the regulatory role of Id4 in mammary gland and prostate are distinct.
The presence of focal hyperplastic regions resembling PIN like lesions is also observed in Id4-/- mice. Many of the genes associated with prostate cancer and their respective knockout/transgenic phenotypes are also recapitulated in the Id4-/- model that support the role of Id4 in prostate cancer. Apart from loss of Nkx3.1 as discussed above, a decrease in Pten specifically in the prostate, sustained androgen receptor expression, increased Myc and Sox9 also promote early stages prostatic intraepithelial neoplasia [
51]. Our results suggest that the above noted genes and their regulated pathways are downstream of Id4. However, in spite of these complex alterations, we did not observe a significantly greater number of pre-neoplastic lesions in Id4-/- prostate suggesting the possibility of mechanisms/pathways that restrains the formation of significant pre-cancerous lesions and prostate cancer. One of these pathways could involve Akt, a kinase on which many of these pathways converge. Akt1 and 2 deficiency is sufficient to markedly reduce the incidence of tumors in Pten(+/-) mice [
67] and Myc also cooperates with Akt1 in promoting prostate tumorigenesis [
68]. Thus loss of Akt could be a key mechanism that negatively regulates the formation of PIN like lesions given the remarkable pro-neoplastic gene signature in Id4-/- mice. Loss of Akt1 also leads to increased apoptosis and general growth retardation that affect the size of organs [
69]. We speculate that the smaller genital tract and prostate in Id4-/- could be in part due to decreased Akt expression.
Based on sequence homology and interaction studies, Id4 could still function as a dominant negative inhibitor of bHLH transcription factor of the E2A (TCF3) family. However, its interactions with non-bHLH proteins could be the key to understand it’s pro-differentiation vs. inhibitor of differentiation functions. For example, in response to BMP4, Id4 stabilizes RUNX2 and promotes osteoblast differentiation [
16]. A similar mechanism can be envisioned in the prostate where Id4 could stabilize transcription factors involved in prostate development such as the Homeobox (
Hox cluster, and Nkx3.1) and Forkhead box genes (Fox A1 and A2) in response to secreted signaling molecules (
Wnts, Fgfs, BMPs/TGFβ/Activins) [
70]. These complex interactions and cross-regulation could promote Id4 dependent prostate morphoregulatory gene signature essential for normal prostate development. Id4 could also regulate the correct timing of prostate epithelial cell differentiation, in a mechanism similar to neural differentiation [
71] through complex interplay involving transcription factors (bHLH and non-bHLH) and response to signals from the surrounding mesenchyme.
Materials and methods
Animals
All animal studies were conducted in accordance with federal guidelines and approved by the Institutional Animal Care and Use Committee, Geisel Medical School at Dartmouth. The mice were sedated using a lethal dose of tribromoethanol (TBE) followed by terminal perfusion with 10% acetate buffered formalin. The reproductive tract including prostates from 6-8 weeks old Id4-/-, Id4+/- and Id4+/+ mice were obtained from Dr. Mark A. Israel (Norris Cotton Cancer Center, Lebanon, NH, USA). The Id4-/- mice were generated by targeted replacement of the endogenous Id4 locus with the green fluorescent protein (GFP) coding sequence [
13]. The tissues were fixed in buffered formalin and paraffin embedded.
Histological analysis
Five micron sections were used for all histological and immuno-histochemical analysis. The sections were stained with hematoxylin and eosin using standard procedures. The H&E sections from knockout, heterozygous and wild type mice were examined by veterinary pathologists (Dr. Thomas Graham, DVM, PhD, and Dr. Ebony Gilbreath, DVM, PhD, Department of Pathobiology, School of Veterinary Medicine, Tuskegee University, Tuskegee, AL, USA). All the sections were performed from proximal to distal region with ventral prostate as the most proximal region.
Immuno-histological analysis
Slides were processed through standard protocols. Following antigen retrieval (autoclave in 0.01 M sodium citrate buffer pH 6.0 at 121C/20 psi for 30 min), the peroxidase activity was blocked in 3% H2O2 and non-specific binding sites blocked in 10% Goat serum. The blocked sections were incubated overnight at 4°C with either of the following antibodies: Androgen receptor (Rabbit mAb, Cell Signaling, cat#153P), Akt (11E7, Rabbit mAB, Cell Signaling, cat# 4685), phospho Akt (ser473, Rabbit pAb, Cell Signaling, cat# 9271), Pten (Rabbit mAb, Cell Signaling, cat#9559), Myc (Rabbit mAb, Cell Signaling, cat#5605X), NKX3.1 (mouse mAb, Thermo Scientific, cat#16906), Sox9 (Rabbit pAb, Novus biological, NB-100-2202), Id4 (Rabbit pAb, Aviva, ARP38058-T100), Id1 (Rabbit mAb, cat# BCH-1#195-14), Ki67 (Rabbit polyclonal, AbCam, #ab15580) followed by incubation with secondary antibody (goat anti-rabbit (#32260) or goat anti-mouse (#32230) -HRP, Thermo Scientific) for 1 hour. The slides were stained with DAB for 2 min, counterstained with hematoxylin and mounted with Immuno-mount (Thermo Scientific), examined and photo-micrographs taken using the Zeiss microscope with an AxioVision version 4.8 imaging system. All the antibodies were mono-reactive, that is a single reactive band was observed in western blot using total cell lysate from prostate cancer cell lines LNCaP, DU1545 and PC3. Non-specific binding of the secondary antibodies was evaluated using respective normal IgGs (data not shown).
TUNEL assay
The terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate (dUTP) nick end labeling (TUNEL) assay was used to detect fragmented DNA as marker for apoptosis in FFPE tissue sections using TACS 2 TdT-DAB apoptosis detection kit (Trevigen). The slides were counterstained in hematoxylin and mounted with Immuno-mount (Thermo Scientific).
Id4 over-expression and silencing in prostate cancer cell lines
The prostate cancer cell lines LNCaP, DU145 and PC3 were purchased from ATCC and cultured as per ATCC recommendations. Human Id4 was over-expressed in DU145 cells as previously described [
35]. Id4 was stably silenced in LNCaP cells using a gene specific shRNA retroviral vector (Open Biosystems #RHS1764-97196818). Successful Id4 over-expression and gene silencing was confirmed by qRT-PCR and Western blot analysis.
Western blot analysis
Total cellular protein was prepared from cultured prostate cancer cell lines using M-PER (Thermo Scientific). 30ug of total protein was size fractionated on 4-20% SDS-polyacrylamide gel (Novex) and subsequently blotted onto a nitrocellulose membrane (Whatman). The blotted nitrocellulose membrane was subjected to western blot analysis using protein specific antibodies as mentioned above. After washing with 1x PBS with 0.5% Tween 20, the membranes were incubated with a horseradish peroxidase (HRP) coupled secondary antibody against rabbit or mouse IgG and visualized using the Super Signal West Dura Extended Duration Substrate (Thermo Scientific) on Fuji Film LAS-3000 Imager.
Chromatin immuno-precipitation (ChIP) assay
Formalin-fixed paraffin-embedded (FFPE) samples from wild type and Id4 knockout mice were used for ChIP based analysis of androgen receptor binding on the mouse Nkx3.1 promoter. For this analysis, 40 μm thick FFPE sections with more that 75% prostatic ducts were used from Id4-/- and WT mice. Genomic DNA was isolated from these sections by the method of Fanelli et al., [
72] except that tissue samples were de-paraffinized with xylene instead of histolemon. The chromatin extracted from tissue samples was sheared (Covaris S220), subjected to immuno-precipitation with either androgen receptor (Millipore, #06-680), mouse IgG (Active motif # 102302) or RNA polI (Millipore, #05-623) antibodies, reverse cross linked and subjected to qRT- PCR [
72]. The androgen receptor binding site (AAA TTA TGG ATG TTC TTT TAA GTC TT) in the first intron of mouse Nkx3.1 [
40] (311 bp from start of first intron) was quantitated by real time PCR (BioRad CFX96) using forward (5′GCC CAC TCT TAA GTT CCC TT) and reverse (5′CAT GAA AAG TGG TTG GGG CC) primers (130 bp amplicon).
LNCaP and LNCaP-Id4 cells cultured in 10% Fetal bovine serum were used to analyze androgen receptor binding on consensus ARE sites in NKX3.1 promoter using primer pairs described previously [
73] with EZ CHiP kit (Millipore). The reagents for PolA CHiP on GAPDH were included in the EZ CHiP kit as internal standards.
Data and statistical analysis
The NIH Image J [
74] was used for counting, calculation of area and diameter of H&E stained prostatic ducts (for description see respective figure legends). Quantitative real time data was analyzed using the ∆∆Ct method: the Ct values of IgG were used to first calculate ∆Ct. Following this normalization step, the ∆∆Ct was then calculated with ∆Ct of wild type set to 1. Within group Student’s
t-test was used for evaluating the statistical differences between groups. One-way ANOVA and Dunnett’s multiple tests were used to test for differences between more than two groups.
Acknowledgements
The authors also wish to thank Dr. Mark A. Israel and Susan Kasper for critically reviewing the manuscript. The authors also wish to thank Dr. Thomas Graham, DVM, PhD, and Dr. Ebony Gilbreath, DVM, PhD, Department of Pathobiology, School of Veterinary Medicine, Tuskegee University, Tuskegee, AL, USA, for histological analysis of the prostates. The work was supported by NIH/NCI CA128914 (JC) and in part by NIH/NCRR/RCMI G12RR03062.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
PS: Immuno-histochemistry, data analysis and first draft of manuscript. AEK: Development of LNCaP-Id4 cell lines. SC: Chomatin immuno-precipitation assays. SK: Western blot analysis. PN: Histology. DP: Prepared samples for Chromatin immuno-precipitations. MCH: Maintenance of Id4 KO mice and dissection of prostates. JC: Conceived the study and final draft of the manuscript. All authors read and approved the final manuscript.