Development and Characterization of Novel Rat Anti-mERβ Sera

Abstract

Estrogens regulate normal sexual and reproductive development in females. Their actions are mediated mainly by estrogen receptor (ER)α and ERβ. Understanding the function of ERs necessitates knowing their cellular location and protein partners, which, in turn, requires reliable and specific antibodies. Several antibodies are available for ERα; however, discrepancies in immunoreactivity have been reported for ERβ. Here, we have developed antisera for mouse ERβ (mERβ) using a specific C-terminal 18-amino acid peptide conjugated to mariculture keyhole limpet hemocyanin. Sprague Dawley rats were immunized, and the resulting antisera were characterized by Western blot analysis of nuclear extracts from tissues of wild-type (WT) mice, and mice genetically modified to lack either ERα (CERαKO) or ERβ (CERβKO). An approximately 56-kDa protein was detected in the hypothalamus, uterus, ovary, mammary gland, testes, and epididymis of WT mice, consistent with the predicted molecular size of ERβ. In addition, the same protein band was identified in in vitro synthesized mERβ protein and in the mammary glands of CERαKO mice. The approximately 56-kDa protein was not observed in in vitro synthesized mERα protein or in any tissue examined in the CERβKO mice. Immunohistochemistry using the antisera revealed ERβ staining in the granulosa cells of WT ovaries and in the mediobasal hypothalamus, paraventricular nucleus, and cerebral cortex in the WT adult mouse brain. These data suggest that the novel rat anti-mERβ sera are specific to ERβ to allow investigators to explore to cellular and physiological role of ERβ in the brain and other mouse tissues.

The hypothalamic-pituitary-gonadal axis is the principal modulator of reproductive function in both genders of mammals. In females, GnRH neurons in the hypothalamus secrete GnRH, which plays a pivotal role in regulating the cascade of hormonal events necessary for ovarian maturation, estrogen production, and normal reproductive function (1).

Estrogens are steroid hormones with effects nuclear mediated by both of its receptors, estrogen receptor (ER)α and ERβ. There are many differences in tissue-specific distribution and function of the ERs (2–6). Mouse ERβ (mERβ) expression in females is most abundant in the ovary, uterus, lung and neuronal tissues (2–6). In males, mERβ is expressed in reproductive tissues, such as testes and prostate (3–5). In addition, ERα is abundantly expressed in all of these tissues, except for the ovary (3, 7).

Understanding the biological effects of ERβ requires knowing where it is located and what proteins it interacts with, which therefore requires reliable and specific antibodies (Table 1) (8–14). ERα and ERβ contain common structural elements that include a highly conserved DNA-binding domain (located in the C domain) that is responsible for their specific binding to estrogen response elements of target genes (3, 7). The ligand-binding domain (LBD) (located in the E/F domain) is a structural element moderately conserved with 55% homology between ERs that contributes to receptor dimerization (7). Between the DNA-binding domain and the LBD is the hinge region (D domain); this structural region is 30% homologous between the ERs. The D domain may play a role, providing the flexibility present in the 3-dimensional structure of ERs (15). In addition, the A/B and E/F domains are elements contained in the ERs that harbor the activation function 1 and activation function 2, respectively, important for ligand-independent transactivation (16, 17). Similarities between the ERs have made the development of the specific antibodies to ERβ most difficult.

Table 1.

Antibody Table

Antibodies/Antisera		Host sp	Applications				ERβ Positive Sites	ERα Cross-Reaction	Reference
Description	Antigen		WB	CC	IHC	IF
PAI-310B	rC-term (467–485)	Rabbit	NR	NR	+	NR	Mouse granulosa cells	−	2
			NR	+	NR	NR	Rat granulosa cells	NT	17
			+	+	−	NR	HT22 cells* and mouse granulosa cells	+	27
RB2.3/custom made	rC-term (182–485)	Rabbit	+	NR	NR	NR	Rat granulosa cells and rat liver	−	17
PC168	rC-term (467–485)	Rabbit	NR	NR	+	NR	Rat granulosa cells	−	17
Ck5912/custom made	rC-term (467–485)	Chicken	+	+	NR	NR	HT22 cells*, mouse Hyp and hippocampus	+	27
D7N/51-7700	hC-term (NR)	Rabbit	+	NR	+	NR	Human ductal carcinoma cells, MCF-7, T74D, HCT-16, T98G, HUVEC	NT	25
			+	NR	NR	NR	MCF-7, HUVEC	NT	33
			+	NR	NR	NR	HT22 cells*, mouse Hyp and hippocampus	+	27
Y19/sc-6821	mN-term (NR)	Goat	+	NR	NR	NR	NIH 3T3 cells*	−	29
			+	NR	NR	+	C2C12 cells	NT	35
			+	NR	NR	NR	HT22 cells*, mouse Hyp and hippocampus	+	27
L20/sc-6822	mC-term (NR)	Goat	+	NR	NR	+	C2C12 cells	NT	35
			−	NR	NR	NR	HT22 cells*	NT	27
H150/sc-8974	hN-term (1–150)	Rabbit	+	NR	NR	+	C2C12 cells	NT	35
			+	NR	NR	NR	HT22 cells*, mouse Hyp and Hippocampus	+	27
1531/sc-53494	hC-term (256–505)	Mouse	+	NR	NR	+	Human CD4 and CD8 T lymphocytes, B lymphocytes and NK cells	NT	26
			+	NR	NR	NR	HT22 cells*, mouse Hyp and hippocampus	+	27
9.88/Ab16813	Recombinant h-full	Mouse	−	NT	NT	NT	MCF-7, T74D, HCT-16, T98G, HUVEC	NT	25
			+	NR	NT	NT	HT22 cells*, mouse Hyp and hippocampus	+	27
06-0629	r/m (54–71) and h (46–63)	Rabbit	NR	NR	+	+	Human breast fibroadenoma cells (epithelium and stroma)	NT	(32)
			+	NR	NR	NR	HT22 cells*	+	27
Z8P	mC-term (468–485)	Rabbit	NR	NR	+	NR	Mouse PVN and granulosa cells	−	4
			+	NR	+	+	Rat substantia nigra pars compacta cells	−	9
			NR	+	+	NR	Rat hippocampal cells*	−	10
			+	NR	+	NR	Mouse MN hybrid cells*	−	11
			NR	NR	+	+	Mouse SCN cells	−	12
			NR	NR	+	NR	GT1–7 cells, rat PVN, Cpu, Rpa and AVPV, MPO and OLVT cells	−	14
80424/custom made	mC-tem (64–82)	Rabbit	+	NR	+	+	SF9 and COS-7 cells*, mouse HYP tissue		13
ERβ18C/custom made	mC-term (467–485)	Rat	+	NT	+	+	Mouse tissues*, recombinant ERβ	−	This study

Summary of ERβ antibodies. r, rat; h, human; m, mouse; C-term, C terminus; N-term, N terminus; sp, species; WB, Western blotting; IF, immunofluorescence; +, detected protein; −, not detected protein; NR, not reported; T, tested; NT, not tested; HT22 cells*, HT22 cells transiently transfected with ERα (HT22 cells-ERα) and ERβ2-GFP expression vectors; MCF-7, T74D: human breast cancer cells; HCT-116, human colorectal cancer cells; T98G, human glioma cancer cells; HUVEC, human umbilical vein endothelial cells; NIH 3T3 cells*, murine NIH 3T3 cells transfected with mERβ expression vector; C2C12 cells, murine skeletal muscle cells; recombinant ERβ, recombinant mouse ERβ protein synthetized in vitro; rat hippocampal cells*, primary hippocampal neuron cultures from female rat pups; mouse motoneurons (MN) hybrid cells*, embryonic mouse spinal cord motoneurons (MN) fused with neuroblastoma cells (N18TG2); SCN cells, suprachiasmatic hypothalamic nucleus; Cpu, caudate putamen nucleus; AVPV, anteroventral periventricular nucleus; Rpa, rostral preoptic area; MPO, medial preoptic area; OVLT, organum vasculosum laminae terminalis; SF9 cells*, SF9 cells transiently transfected with mouse ERβ expression vector; COS-7 cells*, COS-7 cells transiently transfected with ERα and ERβ expression vectors; mouse different tissues*: hypothalamus (PVN), medial hypothalamus, cerebral cortex, pituitary, MG, uterus, ovary (granulosa cells), and liver tissues.

Early studies using commercially developed and custom-made ERβ antisera showed both agreement and discrepancies in the localization of ERβ immunoreactivity (18), raising concerns about the specificity of the ERβ antisera preparation (18, 19). To address this controversy, we produced novel ERβ antisera against a peptide sequence specifically located in the mERβ protein.

Materials and Methods

Development of ERβ antisera

Small compounds (drugs or peptides) could not be sufficiently complex by themselves to induce an immune response or be processed in a manner that elicits production of specific antibodies. Therefore, for antibody production to be successful with small antigens, immunogenic carrier proteins such as keyhole limpet hemocyanin (KLH) can be chemically conjugated with small peptides to increase the intensity of the immune response. For the ERβ antisera (ERβ18C) production, an 18-amino acid peptide, specific to the C terminus of mERβ (present in the mERβ1 and mERβ2) was synthetized in vitro and conjugated to mariculture KLH (77610; Thermo Scientific Pierce) by the Synthesis and Sequencing Facility Laboratory (Johns Hopkins University). Briefly, 10 μL of the conjugated peptide were immobilized onto 5-mm² pieces of nitrocellulose paper and surgically implanted into the peritoneal cavity of Sprague Dawley rats. Rats received booster immunizations at 2-week intervals, for a total of 4 implants. Two weeks after the last booster, rats were exsanguinated by cardiac puncture. The blood was allowed to clot, and the resulting immune antisera were characterized by Western blot and immunofluorescence analyses (Figure 1). Four rats were immunized; the first antiserum analyze was found to not be immunogenistic, the second is describe in this manuscript, and the additional 2 have not be analyzed.

Figure 1. Diagram of methods used to produce the mouse ERβ antibody.

The specific ERβ C-terminal 18-amino acid peptide was identified in the E/F domain, synthetized in vitro, and implanted in rats. The antisera obtained were characterized by Western blotting and IHC. 18aa, C terminus of mERβ protein; mcKLH, maleimide-activated mariculture KLH.

Animals

Generation of ERβ “floxed” mice

Floxed ERβ mice were designed with Lox P sites flanking exon 3 of the ERβ (ERβ) in a targeting construct that also included a Frt flanked neomycin^r selection cassette. A targeting vector was produced by recombineering and was electroporated into embryonic stem cells. Eight positive targeted clones were obtained. Correctly targeted ES cells were injected into C57BL/6J blastocysts to generate chimeras that were backcrossed to C57BL/6J animals to give ERβNEO+ mice. For genotyping, genomic DNA from tail biopsies was used and mice were genotyped by PCR as described. In addition, to remove the FRT-flanked neomycin resistance cassette, ERβNEO+ animals were bred to FLP wild-type (WT) mice to create ERβNEO-mice (20, 21).

Generation of complete-ERβ (CERβKO) knockout

The ERβ floxed animals were crossed to the EIIa-Cre^+/− mice described previously by our laboratory (22, 23). Fl-ERβ mice (ERβ^WT/fl) were crossed with EIIa-Cre^+/− to obtain ERβ^WT/fl/EIIa-Cre^+/− mice, then, this mouse was crossed back with ERβ^WT/fl mouse. Cre recombinase expression in the whole-body produced a complete ERβ (CERβKO) mice. Genomic DNA from tail biopsies was isolated using phenol-chloroform extraction and isopropanol precipitation. To determine the presence of the ERβ floxed allele, WT allele, and Cre transgene by PCR analysis. The genotyping primers designed to detect the presence of the floxed allele, WT allele, or knockout recombinations were as follows: F1 sense, 5′-ACCAAACCAAGCCAAACCAA-3′ and R1 antisense, 5′-ATCACTGTGGAGGTGGAGGA-3′ (Supplemental Figure 1A). Primers F1 and R1 were designed to produce a 1544-bp band to indicate the floxed ERβ allele, a 1366-bp band to indicate the WT allele, and a 539-bp band after excision of the sequence between the Lox P sites (Supplemental Figure 1B). Genotyping primers used to determine the presence of the Cre recombinase gene were Cre sense, 5′-CGACCAAGTGACAGCAATGCT-3′ and Cre antisense, 5′-GGTGCTAACCAGCGTTTTCGT-3′, and have been described elsewhere (23). All mice used in these experiments were maintained on a mixed CD1/129SvJ/C57BL6 genetic background, and each genotypic or experimental group was compared with littermate controls carrying either the floxed ERβ gene without the Cre gene, the Cre gene without the floxed ERβ gene, or neither the Cre gene nor the floxed ERβ allele. These littermate controls are referred to as WT mice throughout this article. All procedures were performed under standard light and dark cycles with approval of the Johns Hopkins Animal Care and Use Committee. Before all experiments, mice were anesthetized with isoflurane (Penn Veterinary Supply), and blood samples were obtained via mandibular blood samples or heart puncture in the case of terminal studies.

Tissues collection

The complete ERβ knockout, genetically modified to lack ERβ (CERβKO) and control WT mice of both genders were developed in our laboratory as described previously in the material and methods section. The adult mice were anesthetized with isoflurane (Penn Veterinary Supply), euthanized and tissues collected for Western blot, immunohistochemistry (IHC), and immunofluorescence analyses. All procedures were performed under standard light and dark cycles with approval of the Johns Hopkins Animal Care and Use Committee. Tissues from complete ERα knockout (CERαKO) mice were a kind gift from Dr Shu-Yuan Yeh (University of Rochester Medical Center, Rochester, NY) (22).

Western blot analysis

Protein extracts were obtained from male and female WT, CERαKO, and CERβKO mice, and Western blot analysis was performed (female tissues were harvested before 10 am in metestrous-diestrous phase). Briefly, nuclear protein lysates were prepared using cell lysis buffer (9803; Cell Signaling Technology) as described previously (24). A total of 20 μg of nuclear proteins were separated on 10% Sodium Dodecyl Sulfate/polyacrylamide gels and transferred onto Protran nitrocellulose membrane (8484140; Whatman GmbH). The membranes were blocked for 1 hour with 5% nonfat milk in Tris-buffered saline containing 0.05% Tween 20 (TBST) and then incubated overnight with rat anti-mERβ serum (1:500 diluted in 5% nonfat milk-TBST). The membranes were washed 3 times for 10 minutes with TBST and incubated for 1 hour at room temperature with goat antirat IgG (2065; Santa Cruz Biotechnology, Inc). Similarly, membranes loaded with the same protein extracts were incubated overnight with rabbit ERα antibody (H184; Santa Cruz Biotechnology, Inc), then washed and incubated for 1 hour at room temperature with goat antirabbit IgG (172-1019; Bio-Rad Laboratory, Inc). The mERβ and mERα proteins were synthetized in vitro using the T7 TNT-coupled reticulocyte lysate system (Promega Corp) from cDNA that were prepared from vector pGEMTeasy-mERβ and mERα, respectively (17), and were used as controls for the respective antibodies. The blots were hybridized with a mouse monoclonal antibody to actin clone C4 (MAB1501; EMD Millipore) as an experimental control. The blots were visualized using ECL Prime Western blotting detection system (RPN 2109; GE Healthcare).

Immunofluorescence staining

Ovaries were obtained from female WT and CERβKO mice after perfusion with heparinized saline followed by a 4% buffered paraformaldehyde solution. Ovaries were immersed in Tissue-Tek CRYO-OCT Compound (Fisher Scientific) and they were sectioned to 7-μm thickness using cryostat (HM550; MICROM GmbH) at −20°C. Ovarian sections were then fixed in cold acetone for 5 minutes and rinsed in PBS (pH 7.4), and subsequently incubated with blocking solution containing 5% BSA and 5% normal goat serum in PBS for 1 hour at room temperature. Slices were then incubated with the rat anti-mERβ sera (1:100 in PBS containing 3% BSA, 0.1% Triton X-100, and 0.05% Tween 20) overnight at 4°C. After being washed 3 times for 5 minutes with PBS, the sections were incubated with an goat antirat IgG conjugated to Alexa Fluor 488 (dilution 1:200, A11006; Molecular Probes) in PBS for 1 hour at room temperature. Sections were washed 3 times with PBS for 5 minutes, stained with 4′,6-diamidino-2-phenylindole (DAPI) (dilution, 1 mg/100 mL), and mounted with Vectashield (H-1000; Vector Laboratories). Images were acquired using a fluorescence inverted microscope (Axioskop 2; Zeiss) equipped with a charge-coupled device digital camera for image capture and processing with Axiovision (Zeiss) software. Images were captured using appropriate fluorescence filters for DAPI (blue), Green Fluorescent Protein (GFP) (green), and Cyanine-3 detection (red). Overlay images of different fluorescence signals were generated using Axiovision software. Total magnification is given in each image (see figure 3 below).

Immunohistochemistry

Mouse sections (30 μm) from 4 adult C57Bl/6J female mice fixed with acrolein (2.5%), paraformaldehyde (4%), stored in cryoprotectant antifreeze at 4°C were removed from our tissue archives and used to validate staining of ERβ with the new antibody. Brain sections at the level of the hypothalamus containing the arcuate nucleus and paraventricular nucleus (PVN) (known to possess ERβ) were stained using a BioWave Pro Laboratory Microwave System (Ted Pella Instruments) and a PELCO SteadyTemp Pro Digital Thermoelectric Recirculator (Ted Pella Instruments). The ERβ18C titration was performed to determine the antibody concentration resulting in the highest signal of positive population and lowest signal in negative population (all sections were taken from regions of the arcuate nucleus, serial dilutions tested: 1:300, 1:1000, and 1:3000) (Supplemental Figure 2). The programs for the microwave are outlined in Table 2. The microwave temperature was set at 35°C for all steps. After removal of sections from antifreeze and placement in phosphate buffered saline (0.05M [pH 7.4] and 0.2M NaCl, PBS) for selection of the best levels, all washes and reactions other than the act of transferring sections, incubation of sections with the primary antibody, and development of the nickel diaminobenzidine product took place within the microwave. Briefly, the sections were rinsed in 5 changes of PBS in the microwave and then reacted with a 10-mg/mL solution of sodium borohydride (Sigma) for 8 minutes to remove excess free aldehydes from the tissue. They were rinsed again in PBS and then incubated for 60 minutes at room temperature followed by 48 hours at 4°C with the rat anti-ERβ diluted 1:1000 in PBS containing 0.4% Triton X-100 (Sigma). After incubation, the sections were washed 3 times in PBS, incubated with biotinylated goat antirat (Vector Laboratories) diluted 1:600 in PBS with 0.4% Triton X-100, washed again in PBS, and then incubated with highly diluted avidin biotin kit reagents (ABC Elite; Vector Laboratories) with the A and B reagents used at a concentration of 2.25 μL of each per 10-mL PBS and 0.4% Triton X-100. Normal dilutions of the A and B reagents are 45 μL/10 mL, but the ABC reagents in 1 kit were more concentrated than usual. The reaction was visualized by using 3,3′-diaminobenzidine-HCl (0.2 mg/mL) and 0.0025% H₂O₂ enhanced with nickel sulfate (25 mg/mL) diluted in 0.175M sodium acetate for 20 minutes on the bench.

Table 2.

Immunolabeling Protocol Using the BioWave Pro Laboratory Microwave System

Step	Microwave Settings
	Time	Wattage
PBS rinse	5 × 1 mina	150
NaBH4 treatment	8 minb	150
PBS rinse	5 × 1 mina	150
Rat anti-ERβ 1:1000	[1 h at RT/48 h at 4 C, not in MW]
PBS rinse	3 × 1 mina	150
Goat antirat biotin 1:600	19 minc	150
PBS rinse	3 × 1 mina	150
ABC reagent (2.25 μL/10 mL)	19 minc	150
PBS rinse	3 × 1 mina	150
0.175M sodium acetate rinse	3 × 1 mina	150

^aUnit will pause between each rinse so that sections can be moved to fresh buffer.

^bThree minutes on, 2 minutes off, 3 mi minutes on.

^cFive minutes on, 2 minutes off, 5 minutes on, 2 minutes off, 5 minutes on.

Results

Immunological specificity of the novel ERβ antisera in female tissues

Western blot analysis of nuclear protein extracts from tissues obtained from WT, CERαKO and CERβKO mice was used to characterize the immunoreactivity and specificity of the new ERβ antisera. Female WT mice revealed the presence of an approximately 56-kDa protein in the hypothalamus, uterus, ovary, mammary gland (MG), and liver, consistent with the predicted molecular size of ERβ (Figure 2A). In addition, the same protein band was identified in the in vitro synthesized mERβ protein preparation and in the MG of mice genetically modified to lack ERα (CERαKO) (Figure 2C). However, it was not present in either the in vitro synthesized mERα protein preparation or in MG of female CERβKO mice (Figure 2D).

Figure 2. Western blot analysis of ERα and ERβ protein expression in WT, CERαKO, and CERβKO mice.

A, Representative tissue panel of protein bands detected in WT and CERβKO female mice in Hyp (hypothalamus), uterus, ovary, MG, and liver, respectively. B, Representative tissue panel of protein bands detected in WT and CERβKO male mice in Hyp, testes, and epididymis. C and D, Appropriate controls were used to test their specificity including: in vitro synthetized mERβ and mERα proteins; MG of CERαKO and CERβKO mice, respectively. Bands detected using the new ERβ antisera (∼56 kDa), ERα (∼66 kDa), and actin (∼42 kDa) antibodies.

Blots hybridized with the ERα antibody detected an approximately 66-kDa protein band in the hypothalamus, uterus, ovary, MG, and liver of female WT and CERβKO mice (Figure 2A). However, the ERα band was not identified in either the in vitro ERβ recombinant protein preparation or MG of CERαKO mice (Figure 2C). Actin protein (42 kDa) was used as loading sample control (Figure 2, A, C, and D). A representative entire blot of the tissue panel of protein bands detected in WT and CERβKO female mice is shown in Supplemental Figure 3A. In addition, an overexposed blot of WT ovaries is shown in Supplemental Figure 4.

Immunological specificity of the novel ERβ antisera in male tissues

The novel ERβ antisera revealed the presence of an approximately 56-kDa protein in the hypothalamus, testes, and epididymis of male WT mice, consistent with the predicted molecular size of ERβ. In addition, this band was not present in tissues of homozygous ERβ CERβKO male mice (Figure 2B).

Western blot analysis hybridized with the ERα antibody detected an approximately 66-kDa protein band in the hypothalamus, testes, epididymis of male WT and CERβKO mice (Figure 2B). Actin protein (42 kDa) was used as loading sample control (Figure 2B). A representative entire blot of the tissue panel of protein bands detected in WT and CERβKO male mice is shown in Supplemental Figure 3B.

Immunofluorescence specificity of the novel ERβ antisera

Immunofluorescence analysis of ovaries of WT and CERβKO mice were also used to characterize the immunoreactivity and specificity of the new ERβ antisera. The immunofluorescence findings revealed the presence of ERβ primarily in granulosa cells of antral follicles from WT ovaries (Figure 3, B and B2). However, very low or undetectable levels of immunoreactivity were present in interstitial (Figure 3, B and B1) and thecal cells (Figure 3, B and B3). Similarly, no signal was detectable in cells of the corpus luteum from WT ovaries (Figure 3C). In addition, no immunostaining was detected in granulosa cells of antral follicles (Figure 3, F and F2) and corpus luteum (Figure 3G) in ovarian sections from the CERβKO mice.

Figure 3. Immunofluorescence analysis of ERβ in ovaries from WT and CERβKO mice.

ERβ protein was revealed with the new ERβ antisera followed by a secondary antibody Alexa Fluor 488 conjugated (green). The nuclei were identified with DAPI staining (blue) in WT ovaries (A, A1, A2, A3, and D) and CERβKO ovaries (E, E1, E2, E3, and H). ERβ immunolabeling was detected in WT ovaries (B) in interstitial (B1), granulosa (B2), and theca cells of antral follicles (B3). ERβ immunostaining was not detected in CL from WT mice (C). ERβ immunostaining was not detected in CERβKO ovaries (F) in the interstitial (F1), granulosa (F2), and theca cells of antral follicles (F3). ERβ immunostaining was also not detected in CL from CERβKO mice (G). GC, granulosa cells; Th, thecal cells; IC, interstitial cells; CL, corpus luteum. Scale bars, 50 μm.

IHC specificity of the novel ERβ antisera

The specificity of the ERβ18C antibody in mouse brain tissue is demonstrated in Figure 4. The immunostaining was detected specifically in the mediobasal hypothalamus, with the arcuate nucleus showing mostly nuclear ERβ immunoreactivity (Figure 4A), ERβ staining was also detected in the PVN (Figure 4B), and cerebral cortex (Figure 4C), all areas known to express ERβ. Importantly, no staining was present when tissue was incubated with antibody that had been preadsorbed with the antigenic peptide. Immunostaining with the ERβ18C antibody demonstrated a region-specific distribution of ERβ protein through the rostral-caudal extent of the WT adult mouse brain.

Figure 4. ICC staining of ERβ.

A, Arcuate nucleus (ARC) shows mostly nuclear ERβ immunoreactivity; (B) ERβ immunoreactivity is also present in hypothalamus PVN and (C) cerebral cortex, all areas known to express ERβ. Scale bars, 100 μm (for lower-power inset) and 25 μm (for high-power inset).

Discussion

Divergent results from IHC analysis using ERβ antibodies, developed both commercially and using custom methods, have been reported (2, 25–27). Here, we report the development and characterization of novel rat anti-mERβ sera to address the lack of specific antisera against mERβ protein (Table 1). First, a 18-amino acid sequence, specific to the C-terminal domain of the mERβ protein, was identified and synthetized in vitro. The peptide was conjugated to mariculture KLH protein to enhance immunogenicity and implanted into the peritoneal cavity of rats. The resulting ERβ antisera were characterized by Western blot and immunofluorescence analyses using appropriate controls.

The ERβ gene generates various splice variants that have been identified in human, rat, and mouse tissues (28, 29). In the mouse, the originally described WT form has been named mERβ1, and the presence of a splice variant with an 18-amino acid insertion in the LBD domain, referred to as mERβ2, was also described (28, 29). The greatest sequence differences between ER structures are present in 3 major regions: the N-terminal domain, the hinge D domain, and in the C-terminal domain (30, 31). Therefore, preparations of specific ERβ cDNA probes, primers, and antibodies have tended to be targeted to these areas (30, 31). For these reasons, and after systematic study of mERβ structure/sequence, the C-terminal 18-amino acid peptide of the mERβ protein was chosen for antibody production (Figure 1).

The role of the estrogens via ERβ in the control of reproductive development and function in the hypothalamic-pituitary-gonadal axis is complex and not totally understood (1, 2). The actions of the estrogens can occur by the balance between ER expression and tissue-specific location, although discrepancies of ERβ immunoreactivity using commercially and custom-made antibodies make the interpretation of the findings more difficult (32–34).

Several studies have shown that ERβ protein expression in different mammalian tissues is associated with a consistent physiologic role. However, the lack of specificity of ERβ immunoreactivity by different antisera make the interpretation of previously reported findings and conclusions difficult and impairs our understanding of ERβ functions (2, 3, 7, 27, 35).

Studies by Snyder et al (27) using the ERβ antibodies PA1-310B (Affinity Bioreagents), D7N (51-7700; Zymed), 9.88 (ab16813), 1531 (sc-53494), L20 (sc-6822), Y19 (sc-6821), and the ERβ antisera ck5912 (custom made) (27) showed that many ERβ immunoglobulins (commercially and/or custom made) recognized the in vitro synthetized recombinant ERβ protein by Western blot analysis. In addition, the overexpressed ERβ protein in cultured cells was also identified by Western blot analysis and immunocytochemistry (ICC) (27). Interestingly, these antibodies, when the hypothalamus, hippocampus, and ovary of the null ERβ CERβKO mouse were tested by Western blot and IHC analyses, also recognized protein(s) other than known ERβ variants (IHC) (27). Hence, the authors suggested that experiments performed with these antisera should be interpreted with appropriate caution (Table 1) (27).

The ERβ antisera PA1-310B (Affinity Bioreagents) and ck5912 (custom made) (27) appropriately detected ERβ expression in murine hippocampal cells (HT22 cells) cotransfected with both ERα-GFP and ERβ-GFP by Western blot analysis and ICC (27). Similarly, in NIH 3T3 cells transiently cotransfected with mERα, mERβ1 and mERβ2, the ERβ antisera Y19 (sc-6821) identified mERβ1 and mERβ2 by Western blot analyses, giving a molecular sizes ranging between 56 to 58 kDa (29). However, the ERβ antisera Y19 detected very faint bands in HT22 cells cotransfected with both ERα-GFP and ERβ-GFP by Western blot analysis (27). Our novel rat anti-mERβ antisera characterized herein was able to identify an approximately 56-kDa protein of the synthetized recombinant mERβ by Western blot analysis. Tissues from both genders of WT mice and MG tissue from the null ERα CERαKO mouse revealed a protein band of the same size. In addition, to confirm the specificity between ERs, our rat anti-mERβ antisera did not recognize the synthetized recombinant mERα protein (Figure 2).

Studies performed by Krege et al (2) using the PA1-310B antibody (Affinity Bioreagents) showed no visible staining in the ovary from the homozygous mutant ERβ−/− mice by IHC, as expected. Similarly, no visible staining in ovaries from the ERβ null mutant mice (ERβKO) was observed by IHC using the same polyclonal antibody (3). However, experiments performed by Snyder et al (27) using the PA1-310B ERβ antisera (Affinity Bioreagents) showed immunolabeling in the granulosa cells from WT mice, as expected; however, as not expected, positive immunostaining was also observed in the null ERβ CERβKO mice (27), highlighting the lack of immunospecificity of this antibody. In contrast, our novel rat anti-mERβ antisera detected immunostaining in the granulosa cells from WT ovaries and produced no visible signal in granulosa cells from CERβKO ovaries. In addition, the antisera revealed low levels of ERβ signal in theca and interstitial cells from WT ovaries (Figure 3). Similarly, previous studies have demonstrated the presence of discrete ERβ immunoreactivity in thecal and interstitial cells of monkey ovary (36). However, because the granulosa cells are the primary site of ERβ expression, coexpression of ERβ and ERα in theca and interstitial cells suggests that both subtypes of ERs may be necessary for normal ovarian function (37).

In the mouse brain, as described by Mitra et al (13), the abundance of both ERs in the caudal area of the amygdala suggest that the 2 receptors may be expressed in the same cells and perhaps form functional heterodimers as demonstrated in vitro (38, 39). In addition, both ER subtypes have been reported to localize in the rat medial amygdala (18, 39) as well as the anteroventral perivantricular nucleus of the preoptic area (18, 38). The ERβ also has been reported to be the predominant ER subtype present in the cerebral cortex of the mouse and rat (40). Here, we demonstrated using the new ERβ18C antibody developed in this study, the presence of ERβ immunoreactivity specifically in the mediobasal hypothalamus, PVN, and cerebral cortex, all areas of the mouse brain known to express ERβ.

In conclusion, our results provide evidence that the novel rat anti-mERβ sera specific to an 18-amino acid region from the C-terminal mERβ peptide recognizes an approximately 56-kDa protein from both genders of WT mouse tissues while showing no cross-reactivity with protein extracts from tissues from the complete null ERβKO mouse (both genders), or with synthetized recombinant mERα protein. In addition, ERβ expression was detected in granulosa cells of WT ovaries by immunofluorescence and was not observed in granulosa cells of null ERβKO (CERβKO) mice. Thus, the novel mERβ antisera specifically detect mERβ protein expression in vitro and in vivo using Western blotting and immunofluorescence assays. This antibody will allow investigators to explore the cellular and physiological role of ERβ.