DMM Advance Online Articles. Posted 18 June 2015 as doi: 10.1242/dmm.019885
Access the most recent version at http://dmm.biologists.org/lookup/doi/10.1242/dmm.019885
Scaffold attachment factor B2 (SAFB2) null mice reveal non-
Accepted manuscript
Shiming Jiang 1,2, Tiffany A Katz 3, Jason P Garee 1,4, Francesco J DeMayo 1, Adrian V Lee 1,3,
Disease Models & Mechanisms DMM
redundant functions compared to its paralog SAFB1
2 – Current Address: Department of Epigenetics and Molecular Carcinogenesis, UT MD Anderson
Steffi Oesterreich 1, 3, 5
1 - Lester and Sue Smith Breast Center, Department of Molecular and Cellular Biology, Baylor
College of Medicine, One Baylor Plaza, Houston, TX 77030, USA.
Cancer Center, Houston, TX
3 – Current Address: Women’s Cancer Research Center, Department of Pharmacology and
Chemical Biology, University of Pittsburgh Cancer Institute and Magee-Women’s Research
Institute, University of Pittsburgh Cancer Center, Department of Pharmacology and Chemical
Biology, Pittsburgh, PA 15213, USA.
4 – Current Address: Hondros College, Westerville, OH 43081 USA
5 – Corresponding author
© 2015. Published by The Company of Biologists Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution and reproduction in any
medium provided that the original work is properly attributed.
Accepted manuscript
Disease Models & Mechanisms DMM
Abstract
Scaffold Attachment Factors SAFB1 and SAFB2 are multifunctional proteins that share >70%
sequence similarity. SAFB1 knockout (SAFB1-/-) mice display a high degree of lethality, severe
growth retardation, and infertility in male mice. To assess the in vivo role of SAFB2, and to identify
unique functions of the two paralogs, we generated SAFB2-/- mice. In stark contrast to SAFB1-/-,
SAFB2-/- offspring were born at expected Mendelian ratios and did not show any obvious defects in
growth or fertility. Generation of paralog-specific antibodies allowed extensive expression analysis
of SAFB1 and SAFB2 in mouse tissues, showing high expression of both SAFB1 and SAFB2 in
the immune system, and in hormonally controlled tissues, with especially high expression of
SAFB2 in the male reproductive tract. Further analysis showed significantly increased testes weight
in SAFB2-/- mice, which was associated with increased number of Sertoli cells. Our data suggest
that this is at least in part caused by alterations in androgen receptor function and expression upon
deletion of SAFB2. Thus, despite a high degree of sequence similarity, SAFB1-/- and SAFB2-/mice do not totally phenocopy each other. SAFB2-/- mice are viable, and do not show any major
defects, and our data suggest a role for SAFB2 in the differentiation and activity of Sertoli cells,
which deserves further study.
Accepted manuscript
Disease Models & Mechanisms DMM
Introduction
Scaffold Attachment Factor B2 (SAFB2) is a member of a protein family consisting of the
paralog SAFB1, a closely related protein scaffold attachment factor-like transcriptional modulator
(SLTM), and SAFB2. SAFB2 was originally identified as KIAA0138 and subsequently named
SAFB2 based on its high homology to SAFB1 (Oesterreich, 2003; Townson et al., 2003; Jiang et
al., 2006; Hammerich-Hille et al., 2009; Hammerich-Hille et al., 2010; Hong et al., 2012). The two
proteins display the highest conservation in the functional SAF-Box and RNA recognition motif
(RRM) domains. In the human genome, SAFB1 and SAFB2 map adjacent to each other to
chromosome 19p13.3 where they are oriented in a bidirectional divergent configuration and
separated by a ~500bp region which can act as a bidirectional promoter (Townson et al., 2003).
Given the high degree of sequence similarity, we hypothesized that the proteins have both shared
but also unique functions.
While less is known about SAFB2, there have been many reports on SAFB1. SAFB1 is a
multifunctional protein that binds both DNA and RNA and is involved in the attachment of
chromatin to the nuclear matrix, and in regulation of transcription, stress response, and splicing
(Garee and Oesterreich, 2010; Garee et al., 2011; Altmeyer et al., 2013). More recently it has been
shown that SAFB1 may play a role in embryonic stem (ES) cell self-renewal as a target of FoxD3
(Plank et al., 2014). Intriguingly, there have been a number of reports that collectively imply an
important role of SAFB1 in DNA damage pathways. Lachepelle et al. demonstrated that SAFB1
binds directly to Werner Syndrome helicase (Wrn), which is important for DNA repair and
replication (Lachapelle et al., 2011). These studies also showed that Wrn protein was required for
immortalization and tumorigenesis in SAFB1-/- models. SAFB1 also interacts with, and
downregulates p53, resulting in suppression of p53-mediated gene expression (Peidis et al., 2011).
Further demonstrating a role in DNA damage, γH2AX spreading is enabled by SAFB1 binding to
acetylated histones (Altmeyer et al., 2013).
Reactive oxygen species generation is indirectly
Accepted manuscript
Disease Models & Mechanisms DMM
regulated by SAFB1 with knockdown inducing the expression of xanthine oxoreductase, a key
enzyme in ROS production (Lin et al., 2008).
Both SAFB1 and SAFB2 have been shown to interact with and repress transcriptional
activity of nuclear receptors, including estrogen receptor alpha (ER) (Townson et al., 2003;
Hashimoto et al., 2012). Similar to SAFB1, SAFB2 can block cell growth, as SAFB2
overexpression leads to a decrease in the proportion of cells in S phase and an increase in cells
blocked in G1 (Townson et al., 2003). SAFB2 also plays a role in alternative RNA splicing,
inhibiting the splicing of a tra2 variable exon (Sergeant et al., 2007), and binding to and regulating
activity of the SR protein kinases SRPK1 and SRPK1a (Tsianou et al., 2009). More recently it was
shown that SAFB2 is a substrate for BRCA1/BARDs’ E3 ubiquitin ligase activity, resulting in
regulation of SAFB2 protein expression (Song et al., 2011).
To identify unique physiological functions of SAFB2, we generated SAFB2-/- mice through
targeted deletion in ES cells. Interestingly, SAFB2-/- offspring were born at expected Mendelian
ratios and did not show any obvious defects in growth or fertility. Despite many shared functions,
the presence of severe phenotypes of SAFB1-/- mice, including a high degree of lethality, severe
growth retardation, and infertility in male mice, indicate that SAFB2 can’t fully compensate for
loss of SAFB1 (Ivanova et al., 2005).
Extensive expression analysis revealed high SAFB2
expression in male reproductive system, and subsequent focused analysis showed that SAFB2-/mice have significantly bigger testes, and a significant increase in Sertoli cell number, and we show
data suggesting that this is at least in part a result of altered androgen receptor activity. Our results
reveal physiologically diverse functions of SAFB1 and SAFB2, including a unique role of SAFB2
in Sertoli cells.
Accepted manuscript
Disease Models & Mechanisms DMM
Materials and Methods
Gene targeting and generation of SAFB2-/- mice. The targeting vector resulting in deletion of exons
4 (amino acid 236) through 11 was generated by amplifying a 4.3-kb region (5' arm), or a 3.4 kb
region (3’ arm) from 129Sv mouse brain tissue genomic DNA using primer set 1 or primer set 2
(Supplementary File S1). PCR products were then cloned into pGEM 5Zf(+) using Sac II/Not I
sites (3’ arm) or Xho I site (5’arm). An internal ribosome entry site (IRES)-lacZ (ß-galactosidase)neo-selectable cassette was inserted between the arms. After electroporation of the linearized
targeting construct into 129/Sv ES cells, Southern blot analysis was used to screen for homologous
recombination in 192 G418-resistant ES clones. A 600-bp probe was cloned from 129Sv genomic
DNA and hybridized to Bam HI-digested genomic DNA (Fig. 1A, and Supplementary File S1).
Three independent SAFB-/- ES clones were injected into blastocytes (C57BL/6J) and implanted into
pseudo pregnant females. Three lines of SAFB2-/- mice were generated, which were analyzed in a
C57BL/6J x 129/Sv mixed genetic background. The mice used in this study were maintained on a
12-h daylight cycle and fed a diet of a standard food and water ad libitum in a pathogen-free facility
at BCM. Animal care was performed in accordance with BCM institutional guidelines.
RNA analysis: Total RNA was extracted using the RNeasy Mini Kit (Qiagen), reverse transcribed,
and PCR was performed as previously described (Ivanova et al., 2005). SAFB2 cDNA spanning
exons 4 through 7, and 18 through 21, was amplified with primers shown in Supplemental File S1.
To determine specificity of the Safb2 knockout, SAFB1 cDNA spanning exons 1 to 4, and 9
through 11, was amplified. -actin was used as loading control. For Northern blot analysis, total
RNA was separated by gel electrophoresis, transferred to a membrane, and then hybridized with a
[32P]-dCTP-labeled cDNA probe (619bp) targeting mouse SAFB2 C-terminal sequence (position
2666 bp to 3284 bp) (Fig. 1A, and Supplementary File S1).
Immunoblot analysis. To analyze SAFB2 expression in mouse tissues and confirm loss of
expression of SAFB2 in the SAFB2-/- mice, we generated an antibody targeting the mouse SAFB2
Accepted manuscript
Disease Models & Mechanisms DMM
N-terminus (105 to 199). To generate mouse SAFB2 N-terminal antibody, a cDNA fragment
targeting
amino
acids
105
to
199
(RYGQDGVVILQSSQDRDTMDTGVPDGMEAEDLSVPCLGKADTVNQILHAFDDSK
EYVAAQLGQLPAQLLKHAVDEEVFKNTLEASVSDLKVTLAD) was amplified using primer
pairs with Bam HI and Eco RI restriction sites (see Supplementary File S1 for sequence of primer
pairs), and cloned in frame with GST into Bam HI/Eco RI sites in pGEX-2TK vector. The GSTmB2 N-terminal peptide was expressed in B21 E. coli (Supplementary Fig. S1A) and purified with
Glutathione Sepharose 4 Fast Flow beads (Supplementary Fig. S1B).
Polyclonal serum was
generated (Sigma), affinity-purified using the GST-mB2 N-terminal peptide, and verified with
immunoblotting (Supplementary Fig. S1C, S1D).
For the analysis of SAFB1 protein expression, our previously described (Ivanova et al., 2005)
polyclonal antibody (B1-1753) was used.
Antibodies against -actin (Sigma A5441, mouse
monoclonal) or lamin A/C antibodies (Cell Signaling 2032, rabbit polyclonal) were used to blot for
loading controls. Bands were imaged and analyzed with Alpha Innotech CCD camera
AlphaEaseFC (FluorChem8000) software, respectively.
IGF-I measurements. Serum was stored at –80°C until assayed for hormones. IGF1 was measured
with enzyme-linked immunosorbent assay kits in accordance with the manufacturer's (Diagnostic
Systems Laboratories) protocol.
Histology and immunohistochemistry analysis. With the exception of testes that were fixed in
Bouin's solution, all tissues were fixed in 4% paraformaldehyde, processed, embedded in paraffin,
and sectioned (5 µm). To detect the expression pattern of SAFB2 and SAFB1 in tissues, the newly
generated N-term polyclonal SAFB2 (see above), and polycloncal SAFB1 (B1-1753) antibodies
were used at dilution of 1:150 and 1:500, respectively. Sertoli cells were stained with Wilms tumor
1 (wt1) monoclonal antibody 6F-H2 (DakoCytomation). An androgen receptor antibody (Santa
Cruz sc-816) was used for immunohistochemistry at a dilution of 1:50. LacZ immunofluorescent
Accepted manuscript
Disease Models & Mechanisms DMM
staining was performed as described previously (Ismail et al., 2002). Briefly, testes were embedded
in OCT medium (Sakura Finetechnical Co., Ltd, Tokyo, 103, Japan), cryosectioned at 20 M, and
fixed for 15 min with 2% formaldehyde, 0.2% glutaraldehyde, 0.02% Nonidet P-40, then washed
and incubated overnight at 37°C in x-gal staining solution [1.3mM MgCl2, 15mM NaCl, 5mM
K3Fe(CN)6, 5mM K4Fe(CN6).3H2O, 0.02% Nonidet P-40, 44 mM HEPES (pH 7.9), and 0.05%
(wt/vol) 4-chloro-5-bromo-3-indolyl--D-galactopyranoside]. The sections were washed and
counterstained with Nuclear Fast Red. All images were captured using an Olympus IX70 inverted
scope under a 40x objective with a CCD camera.
AR Reporter Assay. LNCaP and COS7 cells were maintained in RPMI 10% FBS, and 293T cells
were maintained in DEM 5% FBS, and all cells were plated for experiments in phenol red free
RPMI + 10% CSS. Lipofectamine LTX/Plus (Invitrogen) was used to transiently transfect 1 ug of
hSAFB2 or empty vector (pcDNA3.1), 100 ng of ARE, and/or 10 ng AR into LNCaP cells for 24h.
All cells were transfected with 10 ng of null-renilla. Cells were then treated with R1881 (1nM) or
vehicle for 24h.
A dual luciferase reporter assay (Promega) was then conducted as per
manufacturer’s instructions.
Statistical analysis. Number of pups, number of embryos, serum IGF1 levels, AR positive cells, and
body weight were assessed using descriptive statistics, including t-tests, Wilcoxon rank-sum tests,
ANOVA, and Chi-square tests. Number of samples, and statistical analyses are indicated in figures
and/or figure legends.
Accepted manuscript
Disease Models & Mechanisms DMM
Results
Generation of SAFB2-/- mice
SAFB2-/- mice were generated by homologous recombination in 129/Sv ES cells, replacing exons 4
to 10 with an IRES-LaxZ-SV40pA (-galactosidase) marker and a PGK-Neomycin-bGHpA
selection marker (Fig. 1A). The targeting construct was linearized and electroporated into ES cells,
and the neomycin resistant ES cell colonies were screened by Southern blot analysis (Fig. 1B, left
panel). Three recombinant clones were injected into C57BL/6 blastocysts, and two (B11 and G12)
were transmitted to germ line. The chimeric mice were bred with wild type C57BL/6 mice to
produce F1 mixed background (129/Sv and C57BL/6) heterozygous mutant mice (SAFB2+/-).
Heterozygous F1 mice from each line were intercrossed to produce two independent lines of
homozygous mutant mice (SAFB2-/-). Genotyping was confirmed by Southern blot analysis (Fig.
1B, right panel). Loss of SAFB2 mRNA expression was confirmed by Northern blot analysis (Fig.
1C), and by RT-PCR assays using primers spanning Safb2 C-terminus (exons 18-21) (Fig. 1D).
RT-PCR with primers covering exons 4-7 revealed remaining expression of the N-terminus.
SAFB1 mRNA expression was not affected, as shown by RT-PCR using primers spanning Nterminal (exons 1-4) and C-terminal (exons 9-11) regions of the gene (Fig. 1D).
To determine whether the remaining expression of N-terminal RNA product would result in
expression of a truncated protein, we generated polyclonal antibodies against the SAFB2 Nterminus (aa105-199, exons 3-5) (Material and Methods, and Supplementary Fig. S1). We were
unable to detect full-length or truncated SAFB2 in the SAFB2-/- mice, even after prolong exposure
of the film, suggesting that either there is no truncated N-terminal protein expression or the
potential truncated protein is highly unstable (Fig. 1E). As expected, SAFB1 protein expression
was detected in both SAFB2+/+ and SAFB2-/- mice, confirming specificity of the SAFB2 knockout.
Finally, immunofluorescence for LacZ was performed on testes from SAFB2+/+ and SAFB2-/- mice
Accepted manuscript
Disease Models & Mechanisms DMM
indicating the presence of LacZ in SAFB2-/- Sertoli cells, and also in the germ cells although at a
lower level (Fig. 1F). We conclude that our model represents a SAFB2 null mutation resulting in
lack of SAFB2 protein expression.
SAFB2-/- mice do not display any gross phenotypes
We have previously shown that SAFB1-/- mice display both prenatal and neonatal lethality, growth
retardation (caused by defects in the IGF signaling system), infertility in male mice, and reduced
testes weight (Ivanova et al., 2005). To determine if SAFB2-/- mice displayed similar defects we
first assessed if there was some degree of lethality by analyzing genotypes of litters from SAFB2+/intercrosses. All three genotypes of SAFB2 (+/+,
+/-
, and
-/-
) were produced at the expected
Mendelian distribution (1:2:1 ratio), suggesting that SAFB2-/- did not affect viability (Table 1).
There were also no significant differences in body weight (Fig. 2A) or levels of circulating IGF-1
(Fig. 2B) between SAFB2+/+ and SAFB2-/- mice.
To test whether SAFB2-/- caused defects in fertility, we bred SAFB2-/- males with SAFB2+/+
females, and SAFB2-/- females with SAFB2+/+ males. The number of successful pregnancies,
number of litters, and number of pups per litter were not significant different (Table 2), suggesting
that there are no major defects in the development of the reproductive system. Gross histological
analysis of SAFB2-/- mice also did not identify any major structural defects in brain, spine, skin,
tonsil, lung, heart, liver, spleen, stomach, intestine, and kidney (data not show). Thus, in stark
contrast to SAFB1-/- mice, SAFB2-/- does not cause any major phenotypes.
Accepted manuscript
Disease Models & Mechanisms DMM
SAFB2 protein is highly expressed in the male reproductive tract, and SAFB2-/- results in
increased testes weight and number of Sertoli cells
To narrow down which tissue could have potential, less obvious phenotypes in the SAFB2-/- mice,
we surveyed a large panel of mouse tissues for SAFB2 protein, and compared it to that of SAFB1.
Immunoblot (Fig. 3) and IHC (Fig. 4) analysis were performed, using tissue from knockout animals
as controls (Fig. 4A). Immunoblot data showed high expression of both SAFB1 and SAFB2 in the
immune system (spleen, thymus) and in hormonally regulated organs (uterus, ovary). Of interest,
in some tissues of the male reproductive tract, including the prostate, coagulating gland,
epidydimus, and seminal vesicle, SAFB2 expression was significantly higher than that of SAFB1.
In general, protein expression was concordant between immunoblot and IHC analyses, and IHC
data revealed a potential weak cytoplasmic location of SAFB2 in some tissues, such as heart and
kidney. These results indicate that SAFB2 might play a role in the male reproductive system.
Despite lack of gross phenotypes on fertility, further analysis showed that SAFB2-/- mice displayed
a significant increase in testes weight compared to SAFB2+/+ mice (Fig. 5A). This was in contrast
to SAFB1-/- male mice, which we had previously shown to have decreased testes size (Ivanova et
al., 2005). A potential cause for increased testes size in SAFB2-/- mice were changes in Sertoli
cells, known as the “nursing” cells of the testes. Staining for expression of Wilms tumor 1 (WT1),
a marker of Sertoli cells, showed significantly increased number of Sertoli cells in SAFB2-/- testes
compared to SAFB2+/+ testes (Fig. 5B). A representative section showing increased WT1 staining
in the SAFB2-/- mice is shown in Fig. 5C. Since SAFB1 was previously shown to regulate AR
activity (Mukhopadhyay et al., 2014) and AR signaling can effect numbers of Sertoli cells
(Johnston et al., 2004; Tan et al., 2005), we next tested the effect of SAFB2 on androgen-stimulated
AR activity in a number of cell lines (LNCaP, COS7, 239T). We observed significant repression of
AR activity by SAFB2 overexpression in LNCaP (Fig. 5D) and COS7, but not in 293T (data not
Accepted manuscript
Disease Models & Mechanisms DMM
shown). In adult mice (12 months), there was less AR expression in SAFB2-/- testes compared to
SAFB2+/+ testes (Fig. 5E, and Supplementary Fig. S2), possibly as a result of negative feedback as
previously described (Cai et al., 2011). There were no differences in expression of ERbetween
SAFB2-/- and SAFB2+/+ testes (data not shown). Thus, SAFB2-/- mice have increased weight of
testes, associated with an increase in Sertoli cell number, likely at least in part due to altered AR
signaling.
Discussion
SAFB2 was originally identified due to high homology with its paralog SAFB1 (Townson et al.,
2003). Similar to SAFB1, SAFB2 is a multi-functional protein with roles in transcriptional
repression, RNA splicing and SR protein kinase signaling, and with effects on cell growth (Tsianou
et al., 2009; Garee and Oesterreich, 2010; Garee et al., 2011; Song et al., 2011; Altmeyer et al.,
2013). Here we describe a novel mouse model deficient in SAFB2. In contrast to SAFB1-/- mice,
which have major phenotypes including prenatal and neonatal lethality, growth retardation, and
fertility problems, there were no gross phenotypes in SAFB2-/- mice. Comprehensive expression
analysis of both paralogs shows high expression of SAFB2 in the male reproductive tract, and our
results indicate a role for SAFB2 in the testes.
SAFB1-/- mice displayed severe defects including a high rate of lethality and dwarfism, and
additionally, the males are infertile (Ivanova et al., 2005). These dramatic phenotypes were not
recapitulated with genetic deletion of SAFB2, suggesting that either SAFB1 has unique critical
functions not shared with SAFB2, or that SAFB1 can fully compensate for loss of SAFB2, whereas
the reverse is not true. This may indicate that while SAFB1 and SAFB2 are paralogs, having arisen
from a common ancestor, they may not have retained all their shared functions, and new functions
Accepted manuscript
Disease Models & Mechanisms DMM
have been gained. Previous work has demonstrated that SAFB1 and SAFB2 can heterodimerize but
only SAFB1 is able to homodimerize (Townson et al., 2003). Elegant biochemical work by
Sergeant et al. (Sergeant et al., 2007) has clearly shown that SAFB1 is a component of a small
nuclear complex, whereas SAFB2 is part of higher molecular mass nuclear protein complexes.
There are also differences in subnuclear localization between SAFB1 and SAFB2 that are cell type
dependent, and additional studies are necessary to gain a more detailed understanding shared and
unique functions of the paralogs.
In our study, we provide the first comprehensive expression analysis of SAFB1 and SAFB2,
showing high expression of both proteins in the immune system and in hormonally regulated
organs. SAFB1-/- mice have defects in the immune system (Oesterreich et al, manuscript in
preparation), but again, this phenotype is not mirrored in SAFB2-/- mice. Expression of SAFB2
was especially high in the male reproductive tract, in agreement with previous studies showing high
SAFB2 expression in Sertoli cells (Sergeant et al., 2007). SAFB1 is also highly expressed in the
testes, and has been shown to be involved in the regulation of AR function indicating that SAFB1
may also play a role in the testes (Mukhopadhyay et al., 2014). Upon detailed analysis, we found
an increased weight of the testes in the SAFB2-/- male mice. We hypothesize that this is at least in
part the result of SAFB2 playing a role in Sertoli cells, since genetic deletion of SAFB2 led to an
increase in the number of Sertoli cells.
Androgen signaling controls the regulation of Sertoli cell number and activity (Johnston et al.,
2004; Atanassova et al., 2005; Tan et al., 2005; Walker and Cheng, 2005), and since we show here
that SAFB2 can regulate AR activity and levels, it is feasible that this cofactor role of SAFB2
mediates the effect on Sertoli cells. Our recent preliminary studies have shown that AR expression
was increased and proliferation was decreased in prepubertal testes from SAFB2-/- compared to
SAFB1+/+ mice (Data not shown), suggesting early maturation in SAFB2-/- mice, and this will be
Accepted manuscript
Disease Models & Mechanisms DMM
the focus of future studies. We can not exclude the involvement of other pathways, such as altered
FSH and estrogen signaling, since both pathways have been implicated in Sertoli cell functions
(Smith et al., 1982; Johnston et al., 2004; Atanassova et al., 2005; Tan et al., 2005; Walker and
Cheng, 2005; Lucas et al., 2011). Of note, we did not detect altered ERβ levels in testes comparing
the two genotypes. Finally, SAFB2 may also have specific splicing functions in Sertoli cells, as
previously suggested (Sergeant et al., 2007), that contribute to the observed phenotypes. More
detailed mechanism-oriented questions are the topic of future studies.
In summary, we have described the mouse phenotype with genetic loss of the multifunctional transcriptional repressor SAFB2. Genetic deletion of SAFB2 shows a mild phenotype
with no gross structural defects, growth retardation, or reproductive defects. Our data suggest a
role for SAFB2 in the male reproductive tract, especially in Sertoli cells, which deserve further
investigation. The results of this study provide the basis for future studies delineating the unique
and overlapping physiological function of two SAFB paralogs with important roles in normal and
pathobiology.
Translational Impact Box
Clinical Issue: Scaffold attachment binding factor (SAFB) -2 is a component of a multifaceted
family of proteins including SAFB1, and scaffold attachment factor-like transcriptional modulator.
SAFBs have been shown to play a role in RNA splicing, DNA damage, and transcription. SAFB1-/mice display severe growth retardation, as well as deficits in reproductive function. Male SAFB1 -/mice were sterile and a significant amount of lethality was observed in both prenatal and neonatal
pups. While many mechanisms of reproductive dysfunction exist, there still remains unexplained
reproductive incompetence in humans.
reproductive function and infertility.
The SAFB family may play an important role in
Accepted manuscript
Disease Models & Mechanisms DMM
Results: In this report the authors designed SAFB2-/- mice to determine if the two closely related
family members had redundant or distinct functions. They find that indeed the two family members
likely have distinct functions. SAFB2-/- mice do not display the severe growth retardation or
significant neonatal lethality of the SAFB1-/-. SAFB2 is likely to play a role in male reproduction
as it was found to be highly expressed throughout the male reproductive track and involved in
regulation of the androgen receptor (AR).
Additionally, the SAFB2-/- mice had significantly
decreased testes weight and a reduction in the number of Sertoli cells in the testes. Finally, the
study includes a comprehensive expression analysis of both family members, showing both shared
but also unique target tissues.
Implications and Future Directions: SAFB1 and SAFB2 play important roles in a number of
normal and pathophysiological processes.
This study shows that loss of SAFB2 has fewer
deleterious effects compared to SAFB1, but analysis of the phenotypes suggests a role in the male
reproductive system. This SAFB2-/- mouse model provides an eliquent system to study these
potential effects on the normal male reproductive system, as well as in pathophysiology such as
cancer.
Acknowledgement:
The authors would like to acknowledge outstanding technical support from Mrs. Ora Britton.
Funding:
The study was in part funded by NIH R01CA097213.
TK is supported by T32 CA186873.
Research reported in this publication was supported by the National Cancer Institute of the National
Institutes of Health under award number T32CA186783. The content is solely the responsibility of
the authors and does not necessarily represent the official views of the National Institutes of Health.
Accepted manuscript
Disease Models & Mechanisms DMM
Competing Interests:
The authors report no competing interests on this project.
Author Contributions:
S.J. and J.S. conducted experiments, and S.J. and S.O. wrote the paper. T.A.K. helped to analyze
the data, and contributed to writing the paper. S.O. and A.V.L. conceived and designed the overall
study, and S.O., A.V.L, and S.J. designed the experiments. FM was instrumental in the generation
of the mouse model.
Table 1. SAFB2 genetic deletion does not cause lethality. P-value was calculated using Chisquare test.
Number
offspring
Genotype
Ratio
p value
+/+
+/−
−/−
+/+ : +/− : −/−
Total, 239
57
127
55
1.0 : 2.2 : 0.96
n.s
Females, 111
27
59
25
1.0 : 2.2 : 0.93
n.s
Males, 128
30
68
30
1.0 : 2.3 : 1.0
n.s
Table 2. Lack of breeding phenotypes in SAFB2-/- mice. P-value was calculated using t-test.
Genotype
(Males × Females)
Number of mice used
in breeding
Males
Number (%) of
pregnant females
Females
Number of pups
per litter (mean ±
SD)
p value
+/+ × +/+
3
3
3 (100)
6.2 ± 2.6
−/− × +/+
5
5
5 (100)
7.4 ± 2.5
n.s.
+/+ × −/−
5
5
5 (100)
5.3±2.3
n.s.
Accepted manuscript
Disease Models & Mechanisms DMM
References
Altmeyer, M., Toledo, L., Gudjonsson, T., Grofte, M., Rask, M. B., Lukas, C., Akimov, V.,
Blagoev, B., Bartek, J. and Lukas, J. (2013). The chromatin scaffold protein SAFB1 renders
chromatin permissive for DNA damage signaling. Molecular cell 52, 206-220.
Atanassova, N. N., Walker, M., McKinnell, C., Fisher, J. S. and Sharpe, R. M. (2005).
Evidence that androgens and oestrogens, as well as follicle-stimulating hormone, can alter Sertoli
cell number in the neonatal rat. The Journal of endocrinology 184, 107-117.
Cai, C., He, H. H., Chen, S., Coleman, I., Wang, H., Fang, Z., Chen, S., Nelson, P. S., Liu, X.
S., Brown, M. et al. (2011). Androgen receptor gene expression in prostate cancer is directly
suppressed by the androgen receptor through recruitment of lysine-specific demethylase 1. Cancer
cell 20, 457-471.
Garee, J. P. and Oesterreich, S. (2010). SAFB1's multiple functions in biological control-lots still
to be done! Journal of cellular biochemistry 109, 312-319.
Garee, J. P., Meyer, R. and Oesterreich, S. (2011). Co-repressor activity of scaffold attachment
factor B1 requires sumoylation. Biochemical and biophysical research communications 408, 516522.
Hammerich-Hille, S., Bardout, V. J., Hilsenbeck, S. G., Osborne, C. K. and Oesterreich, S.
(2009). Low SAFB levels are associated with worse outcome in breast cancer patients. Breast
Cancer Res Treat.
Hammerich-Hille, S., Kaipparettu, B. A., Tsimelzon, A., Creighton, C. J., Jiang, S., Polo, J.
M., Melnick, A., Meyer, R. and Oesterreich, S. (2010). SAFB1 mediates repression of immune
regulators and apoptotic genes in breast cancer cells. The Journal of biological chemistry 285,
3608-3616.
Accepted manuscript
Disease Models & Mechanisms DMM
Hashimoto, T., Matsuda, K. and Kawata, M. (2012). Scaffold attachment factor B (SAFB)1 and
SAFB2 cooperatively inhibit the intranuclear mobility and function of ERalpha. Journal of cellular
biochemistry 113, 3039-3050.
Hong, E. A., Gautrey, H. L., Elliott, D. J. and Tyson-Capper, A. J. (2012). SAFB1- and
SAFB2-mediated transcriptional repression: relevance to cancer. Biochemical Society transactions
40, 826-830.
Ismail, P. M., Li, J., DeMayo, F. J., O'Malley, B. W. and Lydon, J. P. (2002). A novel LacZ
reporter mouse reveals complex regulation of the progesterone receptor promoter during mammary
gland development. Mol Endocrinol 16, 2475-2489.
Ivanova, M., Dobrzycka, K. M., Jiang, S., Michaelis, K., Meyer, R., Kang, K., Adkins, B.,
Barski, O. A., Zubairy, S., Divisova, J. et al. (2005). Scaffold attachment factor b1 functions in
development, growth, and reproduction. Molecular and cellular biology 25, 2995-3006.
Jiang, S., Meyer, R., Kang, K., Osborne, C. K., Wong, J. and Oesterreich, S. (2006). Scaffold
attachment factor SAFB1 suppresses estrogen receptor alpha-mediated transcription in part via
interaction with nuclear receptor corepressor. Mol Endocrinol 20, 311-320.
Johnston, H., Baker, P. J., Abel, M., Charlton, H. M., Jackson, G., Fleming, L., Kumar, T. R.
and O'Shaughnessy, P. J. (2004). Regulation of Sertoli cell number and activity by folliclestimulating hormone and androgen during postnatal development in the mouse. Endocrinology 145,
318-329.
Lachapelle, S., Oesterreich, S. and Lebel, M. (2011). The Werner syndrome helicase protein is
required for cell proliferation, immortalization, and tumorigenesis in Scaffold attachment factor B1
deficient mice. Aging 3, 277-290.
Lin, J., Xu, P., LaVallee, P. and Hoidal, J. R. (2008). Identification of proteins binding to EBox/Ku86 sites and function of the tumor suppressor SAFB1 in transcriptional regulation of the
human xanthine oxidoreductase gene. The Journal of biological chemistry 283, 29681-29689.
Accepted manuscript
Disease Models & Mechanisms DMM
Lucas, T. F., Pimenta, M. T., Pisolato, R., Lazari, M. F. and Porto, C. S. (2011). 17betaestradiol signaling and regulation of Sertoli cell function. Spermatogenesis 1, 318-324.
Mukhopadhyay, N. K., Kim, J., You, S., Morello, M., Hager, M. H., Huang, W. C.,
Ramachandran, A., Yang, J., Cinar, B., Rubin, M. A. et al. (2014). Scaffold attachment factor
B1 regulates the androgen receptor in concert with the growth inhibitory kinase MST1 and the
methyltransferase EZH2. Oncogene 33, 3235-3245.
Oesterreich, S. (2003). Scaffold attachment factors SAFB1 and SAFB2: Innocent bystanders or
critical players in breast tumorigenesis? Journal of cellular biochemistry 90, 653-661.
Peidis, P., Voukkalis, N., Aggelidou, E., Georgatsou, E., Hadzopoulou-Cladaras, M., Scott, R.
E., Nikolakaki, E. and Giannakouros, T. (2011). SAFB1 interacts with and suppresses the
transcriptional activity of p53. FEBS letters 585, 78-84.
Plank, J. L., Suflita, M. T., Galindo, C. L. and Labosky, P. A. (2014). Transcriptional targets of
Foxd3 in murine ES cells. Stem cell research 12, 233-240.
Sergeant, K. A., Bourgeois, C. F., Dalgliesh, C., Venables, J. P., Stevenin, J. and Elliott, D. J.
(2007). Alternative RNA splicing complexes containing the scaffold attachment factor SAFB2.
Journal of cell science 120, 309-319.
Smith, R. G., Murthy, L. and Lipshultz, L. I. (1982). The effect of estrogen on Sertoli cell
function. The Journal of urology 128, 642-644.
Song, M., Hakala, K., Weintraub, S. T. and Shiio, Y. (2011). Quantitative proteomic
identification of the BRCA1 ubiquitination substrates. Journal of proteome research 10, 51915198.
Tan, K. A., De Gendt, K., Atanassova, N., Walker, M., Sharpe, R. M., Saunders, P. T.,
Denolet, E. and Verhoeven, G. (2005). The role of androgens in sertoli cell proliferation and
functional maturation: studies in mice with total or Sertoli cell-selective ablation of the androgen
receptor. Endocrinology 146, 2674-2683.
Accepted manuscript
Disease Models & Mechanisms DMM
Townson, S. M., Dobrzycka, K. M., Lee, A. V., Air, M., Deng, W., Kang, K., Jiang, S., Kioka,
N., Michaelis, K. and Oesterreich, S. (2003). SAFB2, a New Scaffold Attachment Factor
Homolog and Estrogen Receptor Corepressor. Journal of Biological Chemistry 278, 20059-20068.
Tsianou, D., Nikolakaki, E., Tzitzira, A., Bonanou, S., Giannakouros, T. and Georgatsou, E.
(2009). The enzymatic activity of SR protein kinases 1 and 1a is negatively affected by interaction
with scaffold attachment factors B1 and 2. The FEBS journal 276, 5212-5227.
Walker, W. H. and Cheng, J. (2005). FSH and testosterone signaling in Sertoli cells.
Reproduction 130, 15-28.
Disease Models & Mechanisms DMM
Accepted manuscript
Figures
Figure 1: Generation of SAFB2-/- mice. A) SAFB2 mouse allele, and targeting construct for
deletion of SAFB2 genomic fragment from exons 4 to 10. SB – Southern blot probe. B) Southern
blot analysis of genomic DNA from embryonic stem (ES) cells (left panel) and mice (right panel).
SAFB2+/+ and SAFB2-/- bands have sizes of 12kb and 5kb, respectively. C) Northern blot analysis
of RNA from SAFB2+/+, SAFB2+/- and SAFB2-/- mice. D) RT-PCR using RNA from SAFB2+/+,
SAFB2+/- and SAFB2-/- mice and primers targeting both the N and C terminal regions of SAFB2
and SAFB1. NTC – non-template control. E) Immunoblot of lysates from either SAFB2+/+ or
SAFB2-/- mice with antibodies targeting SAFB2 and SAFB1. Lower panel shows an entire SAFB2
immunoblot to provide proof of expression of N-terminal fragment. F) LacZ staining of SAFB2+/+
or SAFB2-/- Sertoli cells. Blue – DAPI; Red –Alexa 546 (LacZ).
Accepted manuscript
Disease Models & Mechanisms DMM
Figure 2: Body weight and IGF-I levels in SAFB2-/- mice. A) Body weight measurements in male
(N=5) and female (N=6) SAFB2+/+ and SAFB2-/- mice, presented as mean ± SE. B) Serum IGF-1
levels (mean ± SE) in SAFB2+/+ (n=10) and SAFB2-/- (n=10) mice.
Figure 3: SAFB2 and SAFB1 protein expression levels in mouse tissues. Immunoblot analysis
using antibodies against SAFB2, SAFB1, and Lamin A/C (as loading control).
Accepted manuscript
Disease Models & Mechanisms DMM
Figure 4: Immunohistochemical (IHC) staining of SAFB2 and SAFB1 in mouse tissues. A)
SAFB2 and SAFB1 IHC using mouse testes from wildtype, SAFB2-/-, and SAFB1-/- mice. Tissues
were counterstained with hematoxylin. B) IHC for SAFB2 and SAFB1 in mouse tissues, as
indicated (scale bar = 20 um).
Accepted manuscript
Disease Models & Mechanisms DMM
Figure 5: Increased number of Sertoli cells in SAFB2-/- mice. A) Bars represent testes weight as
percent (mean ± SD) of total body weigh (BW), measured in SAFB2+/+ (n=12) and SAFB2-/- mice
(n=22). B) Number of Sertoli cells (mean ± SD) in SAFB2+/+ (n=4) and SAFB2-/- (n=14). C)
Representative picture of IHC for Wilms Tumor 1 (WT1) in SAFB2+/+ and SAFB2-/- mouse testes
(scale bar =10 um). D) Reporter assay in LNCap cells (mean ± SEM, n=4) with fold change
compared to ARE alone. A Two-way ANOVA with a Bonferoni post-test was used to analyze the
data (*=p<0.05). E) Number of AR positive cells/tubule (mean ± SEM, n=6). A t-test was
conducted for statistical analysis (**=p<0.01).