The oncoprotein LMO3 interacts with calcium- and integrin-binding protein CIB.

Abstract

The protein LMO3 belongs to the LIM only (LMO) group of transcriptional regulators, which act as molecular adaptors for protein-protein interactions. However, little is known about its interactive proteins and functions. Evaluating LMO3 in a yeast two-hybrid screen, we identified the calcium- and integrin-binding protein CIB as an LMO3-binding protein, which binds via the second LIM domain (LIM2) of LMO3. Cotransfection of LMO3 and CIB resulted in a shift in LMO3 protein from the nucleus to the cytoplasm. In functional assays, LMO3 induced C8 astrocyte proliferation was suppressed by the overexpression of CIB. This study demonstrates one function for LMO3 in C8 cells and suggests that one role of the CIB/LMO3 complex is to inhibit cell proliferation.

Full text

Page 1

Research Report
The oncoprotein LMO3 interacts with calcium- and
integrin-binding protein CIB
Ling Huia,b,1, Chaoneng Jic,1, Bin Huid, Tongde Lva, Xiaoqin Haa,
Jinsheng Yanga,⁎, Wenqin Caib,⁎
aExperimental Center of Medicine, Lanzhou General Hospital, Lanzhou Command, Lanzhou, 730050, China
bDepartment of Neurobiology, The Third Military Medical University, Chongqing, 400038, China
cInstitute of Genetics, College of Life Sciences, Fudan University, Shanghai, 200032, China
dDepartment of Pharmacology, Medical College of Jiaxing University, Jiaxing, 314001, China
A R T I C L E I N F OA B S T R A C T
Article history:
Accepted 10 February 2009
Available online 21 February 2009
The protein LMO3 belongs to the LIM only (LMO) group of transcriptional regulators, which
act as molecular adaptors for protein–protein interactions. However, little is known about its
interactive proteins and functions. Evaluating LMO3 in a yeast two-hybrid screen, we
identified the calcium- and integrin-binding protein CIB as an LMO3-binding protein, which
binds via the second LIM domain (LIM2) of LMO3. Cotransfection of LMO3 and CIB resulted in
a shift in LMO3 protein from the nucleus to the cytoplasm. In functional assays, LMO3
induced C8 astrocyte proliferation was suppressed by the overexpression of CIB. This study
demonstrates one function for LMO3 in C8 cells and suggests that one role of the CIB/LMO3
complex is to inhibit cell proliferation.
© 2009 Elsevier B.V. All rights reserved.
Keywords:
LMO3
CIB
Yeast two-hybrid
Proliferation
Astrocyte
1. Introduction
The nuclear LIM-domain-only proteins (LMOs), which are
composed of only two tandem LIM domains, are a subfamily
of LIM-containing proteins. The LIM domain is characterized
by a double zinc finger structure. Unlike many other zinc
finger structures, the primary function of the LIM domain is
to mediate the interaction of DNA with other proteins but not
to bind to DNA (Retaux and Bachy, 2002). LMOs, which
consist of four transcription regulatory factors LMO1–4, have
roles in determining cell fate, controlling cell growth, and
differentiation. There is also increasing evidence that LMOs
are involved in transcriptional regulation by forming com-
plexes with other protein factors and then altering the
transcription of target genes.
LMO1 and LMO2 were originally identified by their trans-
location in acute T-cell leukemia, and their overexpression in
transgenic mice leads to T-cell tumors. It has been established
thatLMO2interactsdirectlywitha numberof proteinsandhas
a critical function in early hematopoiesis and angiogenesis by
participating in a number of multiprotein complexes (Yamada
et al., 2002, 1998). Like LMO2, LMO4 is upregulated in breast
cancer, and its overexpression in mice leads to hyperplasia
and tumor formation (Visvader et al., 2001). LMO4 also forms a
complex with the corepressor CtIP and the breast tumor
suppressor BRCA1 in breast epithelial cells, and represses
BRCA1-mediated transcriptional activation (Sum et al., 2002).
LMO3 is the least understood member of the LMO family. It
was first cloned on the basis of its sequence homology and
was found to be highly expressed in the brain and some
B R A I N R E S E A R C H 1 2 6 5 ( 2 0 0 9 ) 2 4 – 2 9
⁎ Corresponding authors. W.Q. Cai is to be contacted at fax: +86 23 6531 8230.
E-mail addresses: huiling866@yahoo.com.cn (L. Hui), CWQTMMU@yahoo.cn (W. Cai).
1Both authors contributed equally to this work.
0006-8993/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.brainres.2009.02.021
available at www.sciencedirect.com
www.elsevier.com/locate/brainres

Page 2

vestibular ganglion cells (Boehm et al., 1991; Foroni et al., 1992;
Hinks et al., 1997). Then the expression of LMO3 was found to
be upregulated in astrocytes treated with dopamine (Shi et al.,
2001). A recent study has shown that LMO3 is expressed at
higher levels in unfavorable subsets of neuroblastoma and is a
neuronal-specific oncogene in neuroblastoma cells (Aoyama
et al., 2005). Despite the identification of a variety of binding
proteins for LMOs, including LMO1, LMO2, and LMO4, no
protein that binds LMO3 has been identified with yeast two-
hybrid analysis. To investigate the function of LMO3, a yeast
two-hybrid study was performed to identify interacting
protein partners that may suggest the biological role of this
protein. The calcium- and integrin-binding protein (CIB) was
identified and the interacting domains were delineated.
2.Results
2.1.
protein
Library screening identifies CIB as an LMO3-binding
The LIM domain of LIM proteins is believed to function as a
specific protein-binding interface. To identify proteins that
interact with LMO3, a yeast two-hybrid assay was performed
using the second LIM domain of LMO3 as bait. CIB was
identified as a putative protein-binding candidate. To quantify
the binding specificity of CIB to LMO3, we tested the
interaction in a yeast mating assay. As shown in Fig. 1, the
second LIM domain of LMO3, but not the first LIM domain
(amino acids 1–73), associated with CIB in yeast. The full-
length LMO3 coding region was also subcloned into the pLexA
vector to produce the construct pLexA-LMO3. When this
construct was tested in the yeast two-hybrid assay, full-length
LMO3 was also shown to bind to CIB.
2.2.
LMO3 interacts with CIB in vivo
Coimmunoprecipitation experiments confirm that
A coimmunoprecipitation assay was used to investigate
whether CIB and LMO3 interact in vivo. Plasmids encoding
myc-CIB and His-LMO3 were cotransfected into C8 cells, and
cell lysates were prepared after 48 h incubation. The lysates
were subjected to immunoprecipitation using monoclonal
antibody specific for a Myc or His epitope, and His-LMO3 and
myc-CIB were detected in the immunocomplexes by western
blotanalysiswithanti-Hisandanti-mycantibody,respectively.
Western blot analysis showed that myc-CIB was present in the
immunocomplex precipitated with anti-His antibody, and that
His-LMO3 was present in the anti-Myc-precipitated immuno-
complex (Fig. 2). These results suggest an association between
the two proteins in C8 cells.
2.3. Subcellullar colocalization of LMO3 and CIB
The physiological relevance of LMO3 and CIB was also
assessed by colocalization assays in C8 cells and NIH3T3
cells. Immunofluorescence revealed that exogenously
expressed His-LMO3 localized primarily in the nucleus of
C8 cells (Fig. 3A), whereas intense cytoplasmic staining with
some nuclear staining was observed in NIH3T3 cells (data
not shown). Unlike LMO3, myc-CIB showed predominant
cytoplasmic staining, with less staining in the nuclei of the
C8 cell lines (Fig. 3B). After cotransfection of LMO3- and CIB-
expressing constructs, LMO3 was restricted to the cytoplasm
and CIB also localized to the cytoplasm (Fig. 3C). The
Fig. 1 – Specific interaction between LMO3 and CIB. The full-length CIB cDNA fused in frame with the transcription activator B42
cDNA in the pB42AD vector was transformed into the EGY48 yeast strain and mating assays were performed, with YM4271
containing one of the following baits fused to the pLexA DNA-binding domain: LMO3 (full-length LMO3), LIM1 (the first LIM
domain of LMO3), or LIM2 (the second LIM domain of LMO3). Transformants were tested for specific interaction on galactose
X-galplates.Theblue colourindicatespositiveprotein–proteininteractions.
Fig. 2 – Coimmunoprecipitation experiment confirmed that
LMO3 and CIB interact in vivo. C8 cells were transfected with
pCMV-myc-CIB and/or pcDNA4/HisA-LMO3. The expression
of Myc-CIB and His-LMO3 in the cotransfected cells was
analyzed by western blotting (A, left lane, and B, left lane). (A)
Cell lysates were immunoprecipitated with anti-His antibody
and immunoblotted with anti-Myc antibody. (B) Cell lysates
were immunoprecipitated with anti-Myc antibody and
immunoblotted with anti-His antibody.
25
B R A I N R E S E A R C H 1 2 6 5 ( 2 0 0 9 ) 2 4 – 2 9

Page 3

superimposition of the LMO3 and CIB immunostaining
patterns indicated that their cytoplasmic distributions were
largely identical (Fig. 3C). The shift in the localization of
LMO3 when coexpressed with CIB also indicated that LMO3
and CIB interact in vivo.
2.4. LMO3-induced cell proliferation is attenuated by CIB
A recent study has shown that the overexpression of CIB and
its interacting protein PS2 promotes cell death (Stabler et al.,
1999).BecauseCIBisabindingpartnerofLMO3,itseffectoncell
proliferationwasstudied.StablytransfectedsublinesC8-LMO3
and C8-HisA were established, and MTT assays were
performed. The overexpression of LMO3 in the C8 subline
was confirmed by RT-PCR and western blotting (Fig. 4A, top
panel). In the MTT assay, C8-LMO3 cells were shown to have a
higher proliferation rate than C8-HisA cells (2nd–4th day:
p<0.05;5thday:p≤0.05)(Fig.4A,bottompanel),suggestingthat
the overexpression of LMO3 promoted cell growth. The effect
of LMO3 on proliferation may be explained by its role in cell-
cycle progression. In a flow-cytometric cell-cycle analysis, the
overexpression of LMO3 increased the population of cells in S
phasefrom26.19%to32.1%andreducedthepopulationofcells
in G0/G1 phase from 65.74.6% to 60.47% in C8-LMO3 cells
compared with C8-HisA control cells (Fig. 4B). LMO3 over-
expression increased the population of U251 cells in S phase
from16.19%to25.77%andreducedthepopulationofU251cells
in G0/G1 phase from 74.63% to 62.49% (Fig. 4B).To further
investigate the effect of CIB on LMO3-induced cell prolifera-
tion, the stably transfected sublines C8-LMO3 and C8-HisA
were transfected transiently with the pCMV-myc-CIB or
pCMV-myc vector and subjected to the MTT assay. The
expression of CIB in the sublines was examined by western
blotting (Fig. 4C, left panel). The C8-HisA cells transfected with
CIBhadalower proliferationrate thanthatof theC8-HisAcells
transfected with empty vector (p<0.05). Although C8-LMO3
cells transfected with empty vector still displayed accelerated
cell growth, the proliferation rate of the C8-LMO3 cells
transfected with CIB was significantly reduced (2nd day:
p<0.05;3rd–5thday:p<0.01)(Fig.4C,rightpanel).Theseresults
indicate that CIB overexpression attenuates LMO3-induced
proliferation.
3.Discussion
Increasing evidence suggests that the LIM domain acts as a
docking site for the assembly of multiprotein complexes,
including transcription factors and other modeling protein
(Retaux and Bachy, 2002). In this work, we provide evidence
that LMO3 interacts with the EF-hand-containing protein
calcium- and integrin-binding protein CIB (also known as
CIB1, calmyrin, and KIP1). First, LMO3 interacted with CIB in a
yeast two-hybrid assay. Second, the two proteins were
coimmunoprecipitated. Third, the two full-length proteins
colocalized when coexpressed in vivo. Fourth, LMO3-induced
cell proliferation was suppressed by CIB. The interaction
between CIB and LMO3 is noteworthy because, to the best of
our knowledge, CIB is the first protein demonstrated to
interact with LMO3 by yeast two-hybrid screening.
The LMO proteins are composed of two LIM domains,
which provide surfaces for protein interactions. The binding
affinities of the two LIM domains of the LMOs for different
target proteins are distinct. A recent study showed that both
LIMdomainsof LMO4can interact withits bindingpartner, the
glycoprotein 130 subunit (Novotny-Diermayr et al., 2005).
Furthermore, the interaction of LMO2 with GATA appears to
require both LIM domains, but especially the LIM2 domain
(Terano et al., 2005). However, other studies have found that
LMO4 mediated its interaction with cofactor CtBP-interacting
protein and 5S-labeled metastasis tumor antigen 1 (MTA1) via
its first LIM domain but not its second LIM domain (Singh
et al., 2005; Sum et al., 2002). A similar result was reported for
LMO2 and LMO4 when interacting with Ldb1 (Ryan et al., 2006).
Our data show that CIB interacts directly with LMO3 and that
the second LIM domain is sufficient to bind CIB. It is notable
Fig. 3 – Location of immunofluorescent LMO3 and CIB in transfected C8 cell. Optical sections of singly and doubly transfected C8
cell were analyzed with confocal microscopy or fluorescence microscopy for the expression of enhanced green fluorescent
protein (EGFP)-LMO3 (green) or myc-tagged CIB detected with TRITC-coupled secondary antibody (red). (A) Singly transfected C8
cell expressing the EGFP-LMO3 fusion protein (left panel). Nucleus was labeled by hochest33258 (right panel). (B) Singly
transfected C8cell expressing the myc-tagged CIBfusion protein(left panel).Nucleus waslabeledby hochest33258 (rightpanel).
(C) Doubly transfected C8 cell expressing the EGFP-LMO3 fusion protein and the myc-tagged CIB fusion protein. The localization
of the EGFP-LMO3 fusion protein is shown (left panel). The localization of the myc-tagged CIB fusion protein is shown (middle
panel). Superimposition of the images was shown in the right panel.
26
B R A I N R E S E A R C H 1 2 6 5 ( 2 0 0 9 ) 2 4 – 2 9

Page 4

that we did not identify Ldb1 protein in our two-hybrid screen.
It does not seem to be consistent with the fact that Ldb1 binds
LMO1, LMO2, and LMO4. However, instead of the whole LMO3
protein, only the second LIM domain (LIM2) was used as the
bait to screen for LMO3-binding proteins. Consequently, only
the C-terminal-LIM-binding proteins were isolated in our
study and not the N-terminal partners.
Although LMO proteins are thought to function predomi-
nantly in the nucleus, mediating protein interactions with
various transcription factors, they do not contain a nuclear
localization sequence. Therefore, the subcellular localization
of LMOs is different. Previous work has shown that LMO2 and
LMO4 localize in both the nucleus and cytoplasm in normal
tissues (Arber and Caroni, 1996; Neale et al., 1995b). However,
under conditions of forced LMO2 expression in T-cells, LMO2
was detected predominantly in the nucleus (Neale et al.,
1995a). The research of Aoyama et al. has shown that LMO3
physically interacts with HEN2, a neuronal basic helix-loop-
helixprotein,andthat the two proteinscolocalizein the nuclei
of COS7 cells in which endogenous LMO3 and HEN2 are
coexpressed. Their results suggest that LMO3 markedly
promotes the growth of SH-SY5Y neuroblastoma cells,
probably by interacting with HEN2 (Aoyama et al., 2005). Our
study shows that overexpressed LMO3 primarily localized in
the nuclei of C8 cells. Therefore, it seems that HEN2 and LMO3
are assembled into functional protein complexes to promote
C8 cell proliferation. It is interesting that the overexpression of
LMO3 in NIH3T3 cells results in a cytoplasmic localization
pattern. The discrepancy in the localization of LMO3 in C8 and
NIH3T3 cells may be attributed to the nature of the different
cell lines. Variations in the nuclear and cytoplasmic localiza-
tion of LMO2 are also dependent on cell type (Neale et al.,
1995a). Future work on the subcellular localization of LMO3
should clarify this.
Fig. 4 – Cell proliferation was analyzed with MTT and flow-cytometric assays. (A) LMO3 promoted C8 cell proliferation, as
shown in the MTT assay. The overexpression of LMO3 in the stably transfected subline C8-LMO3 and C8-HisA was analyzed by
RT-PCR (top panel). The MTT assay was performed as six replicates, and the symbol error bars indicate the standard deviations.
(B) Comparison of the cell-cycle distribution of C8 cells and U251 cells transfected with pcDNA4/HisA empty vector or
pcDNA4/HisA-LMO3 by flow-cytometric cell-cycle analysis. The assay was performed in triplicate, and the symbol error bars
indicate the standard deviations. The statistical significance of the increase in the S phase population and the reduction in the
G0/G1phase population in the LMO3-transfected cells compared with those of the vector-transfected cells was confirmed with a
paired Student's t test. (*p<0.01 versus vector, **p<0.05 versus vector,#p<0.01 versus vector,##p<0.05 versus vector)
(C) LMO3-induced cell proliferation was suppressed by CIB. The stably transfected sublines C8-LMO3 and C8-HisA were
transiently transfected with pCMV-myc-CIB or the empty pCMV-myc vector. The expression of CIB was analyzed on the days
indicated by western blotting using anti-Myc antibody, with actin as the control (left penal). At 24 h posttransfection, the cells
were seeded to perform the MTT assay. The MTT assay was performed as six replicates and the symbol error bars indicate the
standard deviations.
27
B R A I N R E S E A R C H 1 2 6 5 ( 2 0 0 9 ) 2 4 – 2 9

Page 5

CIB itself was cloned in a yeast two-hybrid study using the
integrin α-IIb subunit as bait (Naik et al., 1997). Subsequently,
CIB has been shown to interact with a number of other
proteins (Kauselmann et al., 1999; Lee et al., 2004; Tahara et al.,
2005). Both transfected and endogenous CIB protein have been
shown to accumulate in the nucleus and the cytoplasm (Lee
et al., 2004; Stabler et al., 1999; Tahara et al., 2005). In fact, CIB
may play several roles, including in hemostasis, apoptosis,
and the response to DNA damage in diverse or intersecting
pathways, by binding to different proteins and shuttling
between the nucleus and the cytoplasm (Yamniuk et al.,
2007). The overexpression of CIB and its interacting protein
PS2 in HeLa cells, which colocalized in the cytoplasm,
promoted cell apoptosis in vitro (Stabler et al., 1999). With the
exogenously expression of CIB and its binding protein 6–16 in
TMK-1 and MKN-28 cells, both proteins colocalized in the
cytoplasm, and also induced apoptosis (Tahara et al., 2005).
Conversely, when CIB interacted with the catalytic subunit of
hTERT, they colocalized in the cell nucleus, positively
regulating telomerase activity and telomere length (Lee
et al., 2004). In this study, CIB suppressed LMO3-induced
proliferation and the transfection of the CIB cDNA into C8 and
SHG44 cells also induced apoptosis (L. Hui, unpublished
observations). Furthermore, the coexpression of LMO3 and
CIB in C8 cells did not cause LMO3 to remain the nucleus,
where LMO3 might have interacted with HEN2 and/or other
factors to accelerate cell growth, but resulted in the accumula-
tion of both proteins in the cytoplasm. It is possible that the
overexpression of CIB suppresses cell growth by interfering
with the formation of the physiological nuclear complex
HEN2/LMO3 and causing LMO3 to leave the complex and
enter the cytoplasm, where LMO3 and CIB act synergistically
to alter the transcription of target genes by forming a new
complex with other factors. Overall, the interaction between
LMO3 and CIB is interesting because LMO3 may function as a
cell proliferation protein, whereas CIB is a cell death protein.
Further study of LMO3 and its relationship with CIB will
increase our understanding of proliferation and apoptosis in
both normal and pathological conditions, including tumors.
4.Experimental procedures
4.1.Yeast two-hybrid library screening
The yeast two-hybrid assay was performed using the Match-
maker pLexA Two-Hybrid System (Clontech). The sequence
encoding the second LIM domain (residues 73–145) of mouse
LMO3 was amplified by PCR and cloned into the EcoRI and
BamHI restriction sites of the pLexA vector to act as the bait in
the yeast two-hybrid screen. The nucleotide sequence of the
DNAinsertwasconfirmedbysequenceanalysistoverifythatit
did not contain mutations. This bait construct and the LacZ
reporter plasmid p8op-lacZ were cotransformed into yeast
strain EGY48, which was then transformed with a mouse fetal
brain cDNA library in pB42AD. The transformants were
selectedonplateswithouthistidine,leucine,oruracil.Putative
positive colonies were restreaked onto fresh master plates for
β-galactosidase assays, according to standard protocols. The
pB42AD library plasmid DNA from candidate clones was
recovered and the cDNA inserts were analyzed by nucleotide
sequencing to ensure that the inserts were in frame with the
B42 activation domain. The positive clones were analyzed and
verified by yeast mating, with standard protocols.
4.2.Delineation of interacting regions by yeast mating
The full-length cDNA of LMO3 (encoding 145 amino acids) and
the cDNA encoding the first LIM domain, LIM1 (1–73 amino
acids),weresubclonedintotheEcoRI–BamHIsitesofthepLexA
vector to generate pLexA-LMO3 and pLexA-LIM1, respectively.
All constructs were sequenced before testing with yeast
mating and the interactions were scored as ++ (moderately
positive),+(slightlypositive),or−(negative),where++denoted
growth on leu-trp-his-ura plates for four days.
4.3.Plasmid construction and transfection
Eukaryotic expression vectors pEGFP-LMO3, pcDNA4/HisA-
LMO3, and pCMV-myc-CIB were constructed using molecular
biological techniques. All constructs were sequenced before
use. Then C8-D1A (C8) astrocytes or NIH3T3 cells were
transfected using Lipofectamine 2000 (Invitrogen) according
to the manufacturer's instruction. To construct stably trans-
fected sublines, C8 cells transfected with pcDNA4/HisA-LMO3
(C8-LMO3) or pcDNA4/HisA (C8-HisA) were cultured in
medium containing 0.4 mg/mL Zeocin (Invitrogen) until
clone formation. Well-separated clones were picked three
weeks later and screened for the expression of LMO3 by
western blotting using mouse monoclonal anti-His antibody
(Santa Cruz Biotechnology). The stably transfected sublines
were incubated in medium containing 0.2 mg/mL Zeocin.
4.4.Immunoprecipitation
C8 cells were transfected singly with pcDNA4/HisA-LMO3 or
pCMV-myc-CIB or cotransfected with pcDNA4/HisA-LMO3 and
pCMV-myc-CIB. Whole-cell extracts were prepared 48 h after
transfection. Proteins were immunoprecipitated with either
anti-His or anti-Myc antibody (9E10; Santa Cruz Biotechnology)
and protein A/G-agarose (Santa Cruz Biotechnology), and
separatedbySDS-PAGE.Aftertheproteinshadbeentransferred
to PVDF membranes (Amersham Biosciences, Inc.), the filters
wereblockedandincubatedwithanti-Mycoranti-Hisantibody.
The filters were then incubated with horseradish-peroxidase-
coupled secondary antibodies and developed with enhanced
chemiluminescence (ECL; Amersham Biosciences, Inc.).
4.5.
colocalization assay
Subcellular localization and immunofluorescence
C8 or NIH3T3 cells grown on coverslips were transfected with
pEGFP-LMO3 or pCMV-myc-CIB. At 48 h posttransfection, the
cells were fixed and stained with mouse anti-Myc antibody
and tetramethylrhodamine isothiocyanate (TRITC)-conju-
gated goat anti-mouse secondary antibody (1:50; Sigma).
After a PBS wash, cells were counterstained in hochest33258
(Beyotime) for 10 min. Cell immunofluorescence was analyzed
with a TCS-NT laser scanning confocal microscope or a
fluorescence microscope.
28
B R A I N R E S E A R C H 1 2 6 5 ( 2 0 0 9 ) 2 4 – 2 9