Báo cáo khoa học: MBP-1 is efficiently encoded by an alternative transcript of the ENO1 gene but post-translationally regulated by proteasome-dependent protein turnover doc

MBP-1 is efficiently encoded by an alternative transcript of
the ENO1 gene but post-translationally regulated by
proteasome-dependent protein turnover
Jrhau Lung
1
, Ko-Jiunn Liu
1
, Jang-Yang Chang
1
, Sy-Jye Leu
2
and Neng-Yao Shih
1
1 National Institute of Cancer Research, National Health Research Institutes, Tainan, Taiwan
2 Department of Microbiology and Immunology, Taipei Medical University, Taiwan
Introduction
The role of c-myc promoter binding protein-1 (MBP-1)
in tumor suppression has been demonstrated, in cer-
tain types of cancer, to be that of a general transcrip-
tional repressor. MBP-1 was originally identified from
a human cervical carcinoma. It has been reported to
bind to a 5¢ sequence adjacent to the TATA box of
the human c-myc P2 promoter and to negatively regu-
late transcription by preventing the formation of a
transcription initiation complex [1]. Exogenously
expressed MBP-1 reportedly suppresses cell growth,
and induces apoptosis and necrosis in breast [2],
neuroblastoma [3], or non-small-cell lung cancer cells
[4] via the transcriptional repression of c-myc or
through physical interplay with its cellular partners
[5,6]. More recently, the results of an in vitro experi-

ment suggested that the physiological level of MBP-1
is modulated by the concentration of glucose and that
a change in the expression of MBP-1 leads to an alter-
ation in cell proliferation [7]. The antitumor activity of
MBP-1 has also been demonstrated in human tumor-
xenografted mice [2,5]. Thus, the expression level of
MBP-1 appears to be a determining factor for cell
Keywords
alternative transcription initiation; a-enolase;
c-myc promoter-binding protein-1; protein
stability; ubiquitination
Correspondence
N Y. Shih, National Institute of Cancer
Research, National Health Research
Institutes, Tainan 704, Taiwan
Fax: +886 6 208 3427
Tel: +886 6 700 0123 ext. 65108
E-mail:
or
S J. Liu, Department of Microbiology and
Immunology, Taipei Medical University,
Taipei 110, Taiwan
Fax: +886 2 2377 862
Tel: +886 2 2736 1661 ext. 3414
E-mail:
(Received 28 June 2010, revised 16 August
2010, accepted 19 August 2010)
doi:10.1111/j.1742-4658.2010.07819.x
The c-myc promoter-binding protein-1 (MBP-1) is a transcriptional sup-
pressor of tumorigenesis and thought to be the product of alternative

translation initiation of the a-enolase (ENO1) transcript. In the present
study, we cloned a 2552-bp novel cDNA with a putative coding sequence
of MBP-1 and functionally examined its ability to encode the MBP-1 pro-
tein. Similarly to ENO1, the obtained MBP-1 was widely and differentially
expressed in a variety of normal tissues and cancer cells. Experiments using
MBP-1 promoter-driven luciferase reporter assays, biochemical cell frac-
tionation followed by RT-PCR detection of the cytoplasmic mRNA, and
transcription ⁄ translation-coupled reactions, consistently demonstrated that
this novel transcript was alternatively transcribed from intron III of the
ENO1 gene and was feasible for MBP-1 production. Hypoxia treatments
significantly increased the transcriptional activation of the MBP-1 gene.
Blocking the proteasomal degradation by MG132 stabilized the MBP-1
protein in cells. Compared with the translation efficiency for production of
the MBP-1 protein, the MBP-1 transcript was 17.8 times more efficient
than the ENO1 transcript. Thus, we suggest that this newly discovered
transcript is a genuine template for the protein synthesis of MBP-1 in cells,
and optimal expression of this gene in tumors may lead to effective clinical
therapies for cancers.
Abbreviations
CHX, cycloheximide; ENO1, a-enolase; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HA, hemagglutinin; HIF1a, hypoxia-inducible
factor 1a; MBP-1, c-myc promoter-binding protein-1; RT-qPCR, reverse transcription-quantitative polymerase chain reaction.
4308 FEBS Journal 277 (2010) 4308–4321 ª 2010 The Authors Journal compilation ª 2010 FEBS
growth, and alterations in its level by tumor microen-
vironmental factors may affect cancer development.
Because MBP-1 shares a high sequence similarity
with the glycolytic enzyme a-enolase (ENO1), this
37 kDa cellular protein is thought to be a short form
of the 48 kDa ENO1 protein. Sequence analysis of
MBP-1 revealed 95% sequence identity with ENO1
cDNA in both the coding region and the 3¢-UTR [8].

Coincidentally, both genes are close to the 1p36
region on chromosome 1 [9,10]. Ghosh et al. [11]
reported that the C-terminal MBP-1 protein, which is
highly homologus to ENO1, exhibited transcriptional
repression activity, and its activity was sufficient to
stimulate regression of prostate tumor growth in nude
mice [5]. Additionally, work from Feo’s laboratory
has shown that ectopic expression of the short form
of ENO1, lacking the first 96 amino acids, functions
in a manner similar to MBP-1 [12]. They also showed
that in vitro transcription and translation of the cod-
ing sequence of ENO1 can yield two polypeptides
with apparent molecular masses of 48 and 37 kDa. In
an RNase protection assay, hybridization of the total
RNA of HeLa cells with a cRNA antisense probe
corresponding to ENO1 gave a single transcript, sug-
gesting that the same transcript may encode both
ENO1 and MBP-1. Moreover, site-directed mutagene-
sis of Met94 and Met97 on the ENO1 cDNA further
supports this single-transcript hypothesis and suggests
that MBP-1 is a product of alternative translation ini-
tiation of the ENO1 transcript [13]. It is unclear,
however, whether alternative translation-initiation
codons of the ENO1 transcript are in the correct
context for eukaryotic ribosomal scanning when a
5¢-UTR is present to mimic the endogenous ENO1
transcript. In the present study, we report that a
novel, naturally occurring, transcript is an alternative
transcriptional product of ENO1 and is widely
expressed in many normal tissues and cancer cells.

Hypoxia treatments induced significant levels of its
mRNA in normal-type lung epithelial cells (BEAS-
2B), and blockade of the ubiquitin-dependent degra-
dation pathways by MG132 markedly elevated the
stability of MBP-1. Experiments using a full-length
ENO1 cDNA (with intact 5¢- and 3¢-UTRs) for pro-
duction of MBP-1 revealed that the amount of pro-
tein generated from the construct was 17.8 times
lower than that produced from the full-length MBP-1
cDNA. Therefore, we conclude that this newly dis-
covered transcript is a genuine template for MBP-1
synthesis in cells. In the future, it will be critical to
elucidate potential factors or conditions to modulate
its optimal expression in tumor cells in order to
develop effective therapies for cancers.
Results
Cloning of full-length putative MBP-1 cDNA
MBP-1 protein is thought to be an alternative transla-
tional product of the ENO1 transcript, produced by
utilizing its internal AUG start codons [12,13]. Here,
we characterized an undefined cDNA (NCBI access
number: AL833741), which comprises a putative ORF
that encodes the MBP-1 protein. Intriguingly, in addi-
tion to a 1486-nucleotide sequence spliced from exon
IV to exon XII of ENO1, the AL833741 cDNA con-
tains 789 nucleotides upstream of this sequence.
Because the latter, 789-nucleotide, sequence was
located in the 3¢-end of intron III (Fig. 1), the
AL833741 may be an alternative transcription product
of ENO1.

To determine whether there were more sequences
upstream of the AL833741 transcript, 5¢-RACE reac-
tions were performed. The cytoplasmic RNA was
extracted from the CL1-5 lung adenocarcinoma cell
line, then primed with oligo-d(T)
18
primer and reverse
transcribed into cDNA. Three gene-specific reverse
primers – P1, P2 and P3 (as indicated in Fig. 1) – were
used to align the putative ORF and 5¢-UTR sequences
of AL833741. The resultant PCR product shown in
Fig. S1 was directly sequenced. We identified 279 addi-
tional nucleotides upstream of this transcript and
deposited this sequence in the NCBI database (access
number: GU170215). To confirm this finding, amplifi-
cation using primers specific to the 5¢-end of the 279-
nucleotide sequence and to the 3¢-end of the ENO1
cDNA was performed to clone out a 2552-bp full-
length cDNA. Collectively, these data confirm the exis-
tence of this transcript in the cells.
Distribution of the MBP-1 transcript in normal
tissues and cancers
To determine the distribution of this putative MBP-1
transcript in normal tissues and cancer cells, commer-
cially available cDNAs of a variety of normal tissues,
and the cDNAs generated from normal lung primary
cells as well as lung cancer and breast cancer cells,
were analyzed by semiquantitative RT-PCR using
gene-specific primers to align the putative MBP-1 or
the ENO1 cDNAs. The results in Fig. 2A show wide

expression of MBP-1 in many normal tissues. In con-
trast to the ENO1 transcript, which was highly
expressed in brain, liver and kidney tissues, the MBP-1
mRNA was exclusively highly abundant in spleen, and
was rarely found in stomach and colon tissues. These
results suggest that the expressions of ENO1 and
J. Lung et al. Expression and regulation of MBP-1
FEBS Journal 277 (2010) 4308–4321 ª 2010 The Authors Journal compilation ª 2010 FEBS 4309
MBP-1 are differentially regulated in a tissue-depen-
dent manner. In various lung cancer and breast cancer
cells it was found that although the MBP-1 transcript
level was slightly lower in normal human lung tissue
than in other cells, there were no significant differences
between normal and lung cancer cells, between poorly
and highly invasive cells, such as CL1-0 and CL1-5,
respectively, or between benign and malignant breast
cancer cells, such as MCF-7 and SK-BR3, respectively.
These data suggest that the MBP-1 transcript level is
not related to cell malignancy under normal culture
conditions.
Upregulation of the MBP-1 transcript level in
hypoxia
Upregulation of glycolytic enzymes under hypoxia has
been previously documented [14]. A hypoxia-inducible
factor 1a (HIF1a)-binding site (GCGTG) is also
found in a region 956 nucleotides upstream from the
transcription start site of MBP-1, as indicated in
Fig. 1A. To test whether the expressions of MBP-1
and ENO1 could be modulated by hypoxia, BEAS-2B
and A549 cells were incubated in 2% O

2
or treated
with CoCl
2
for 24 h to mimic hypoxic conditions. The
mRNA transcript and protein levels of MBP-1 and
ENO1 were measured using reverse transcription-
quantitative polymerase chain reaction (RT-qPCR)
and immunoblotting analyses, respectively (Fig. 3). In
response to treatment with CoCl
2
or 2% O
2
, HIF1a
expression was markedly increased in BEAS-2B cells
followed by upregulation of ENO1, as reflected by the
increased levels of protein (Fig. 3A) and of mRNA
transcripts (Fig. 3B). However, this phenomenon was
not consistently observed in A549 cells, as elevated
HIF1a resulted in an increased steady-state level of
the ENO1 transcript, but barely affected the level of
ENO1 protein. Thus, induction of HIF1a by hypoxia
A
B
Fig. 1. The putative MBP-1 transcript is an
alternative transcriptional product of the
ENO1 gene locus. (A) The AL833741 tran-
script comprises a 789-bp sequence tran-
scribed from the intron III of the ENO1 gene
and a sequence spliced from exons IV to

XII. The translational start codon for the
ENO1 protein is located in exon II (gray
arrow) and that for the MBP-1 protein is
located in exon V (black arrow). The
5¢-RACE experiments using three gene-
specific reverse primers (P1, P2 and P3)
cloned out an additional 279-nucleotide
sequence upstream from the AL833741
sequence. This nucleotide sequence is
listed at the bottom of panel A. The c-Myc
and HIF1a binding sites are marked with
solid (
•
) and open (s) circles, respectively.
(B) The gene constructs transcribed into a
transcript with intact 5¢- and 3¢-UTRs, as
well as the coding sequences of ENO1 or
MBP-1, were designated as ENO1 ⁄ FL or
MBP-1 ⁄ FL, respectively. The ENO1 ⁄ ORF
and MBP-1 ⁄ ORF constructs contained only
the coding regions of the ENO1 and MBP-1
genes, respectively. The ENO1 cDNA with
the first translational start codon mutated or
lacking the 5¢-UTR, was named ENO1-
mut ⁄ FL or ENO1 ⁄ 3¢-UTR, respectively. The
MBP-1 ⁄ s-UTR lacked the first 206 nucleo-
tides in its 5¢-UTR.
Expression and regulation of MBP-1 J. Lung et al.
4310 FEBS Journal 277 (2010) 4308–4321 ª 2010 The Authors Journal compilation ª 2010 FEBS
was apparently insufficient to promote the production

of ENO1 protein in this cell line. By contrast, the
expression pattern of the putative MBP-1 transcript
was similar to that of the ENO1 transcript in both cell
lines. However, endogenous MBP-1 protein remained
undetectable under these hypoxic conditions even
though the transcriptional activation of its gene was
markedly enhanced. This may be a consequence of the
remarkably low abundance of the MBP-1 mRNA
transcripts in those cells compared with the mRNA
transcripts of ENO1 (Fig. 3A). Surprisingly, we later
found that the MBP-1 protein was also undetected in
cells transfected with an exogenous MBP-1 ⁄ FL-HA
construct, suggesting that MBP-1 is a stability-labile
protein. Thus, failure to detect the MBP-1 protein in
BEAS-2B and A549 cells under normoxic and hypoxic
conditions can be explained by the low transcript
abundance and the low stability of the MBP-1 protein
in these cells.
The MBP-1 transcript is functionally translatable
Initially, we constructed a luciferase reporter driven by
the 1 kb proximal promoter of MBP-1 containing the
HIF-1a-binding site to determine whether the pro-
moter was transcriptionally functional. BEAS-2B cells
were transfected with the MBP-p-Luc plasmid, and
12 h later the cells were treated with or without CoCl
2
for an additional 24 h then with luciferase reporter
assays. As shown in Fig. 4A, the luciferase activity dri-
ven by the MBP-1 promoter was significantly induced
by 4.3-fold compared with its vector control. The

A
B
Fig. 2. Abundance of putative MBP-1 and ENO1 transcripts in vari-
ous tissues and cell lines. (A) The ENO1 and putative MBP-1 tran-
script levels were determined by semiquantitative RT-PCR using a
variety of commercially available normal tissue cDNAs as templates
and resolved on 1% agarose gels containing ethidium bromide. (B)
Total RNA was extracted from normal lung bronchial epithelial
(NHBE) cells and WI38 normal-like lung cells, as well as from lung
cancer and breast cancer cells, as indicated. After generation of the
first-strain cDNA in the oligo-(dT)
18
-primed reverse-transcription
reactions, the 5¢-UTRs of ENO1 and of putative MBP-1 cDNAs
were amplified using their individual gene-specific primers (Table 1).
The cDNA pool of normal lung tissue (N. Lung) was used as a con-
trol for normal lung cells, and the b-actin gene served as a loading
control.
A
B
Fig. 3. Increasing expression of MBP-1 transcription in hypoxia.
BEAS-2B lung normal-like epithelial or A549 lung adenocarcinoma
cells were seeded onto six-well plates, incubated at 37 °C over-
night and subjected to incubation in 20% O
2
and 5% CO
2
normoxia
(Nm) or 2% and 5% CO
2

hypoxia (Hp) conditions, or to treatment
with 150 m
M CoCl
2
(Co) at 37 °C for 24 h. (A) The total RNA was
extracted from each type of cell and reverse transcribed into cDNA.
The levels of endogenous MBP-1 and ENO1 gene transcripts were
quantified by RT-qPCR, normalized to the b-actin control levels and
calculated as ) 4C
t
= )[C
t(ENO1 or MBP-1)
) C
t(b-actin)
]. The ratios of
ENO1 or MBP-1 mRNA to b-actin mRNA were then expressed as
2
ÀDC
t
 K (K, constant). Data obtained from three independent
experiments are shown as the mean ± SEM of the relative levels
of MBP-1 or ENO1 to the b-actin control level, as indicated. (B) The
lysates were resolved on 10% SDS ⁄ PAGE, and the protein levels
of endogenous MBP-1 and ENO1 were determined by western blot
analyses using an antibody cross-reactive to ENO1 and MBP-1.
After antibody stripping, the same blots were reprobed with anti-
HIF1a or b-actin IgG, as indicated. The immunocomplexes were
detected using the SuperSignal chemiluminescent system.
J. Lung et al. Expression and regulation of MBP-1
FEBS Journal 277 (2010) 4308–4321 ª 2010 The Authors Journal compilation ª 2010 FEBS 4311

activity was further augmented in cells treated with
CoCl
2
, indicating that the MBP-1 promoter is a
functional element and can be modulated by hypoxic
conditions.
To determine whether the MBP-1 transcript could
be exported from the nucleus to the cytoplasm, the
MBP-1 and ENO1 transcripts were analyzed by bio-
chemical cell fractionation followed by RT-PCR.
Western blot analyses of the fractionated lysates of
BEAS-2 and A549 cells were first carried out using
antibodies against the cytosolic and nuclear markers,
glyceraldehyde-3-phosphate dehydrogenase (GAPDH)
and Histone 3, respectively, to show that the cyto-
plasmic fractions were clearly separated from the
nuclear fractions (Fig. 4B). Subsequently, total RNAs
of the fractionated lysates of both cell lines were
extracted and reverse transcribed into cDNAs using
the oligo-(dT)
18
primer. The results shown in Fig. 4C
demonstrate that both MBP-1 and ENO1 mRNA
species are present in the cytoplasm and nucleus, indi-
cating that the nuclear export of MBP-1 mRNA is
similar to that of other mature mRNA species, such as
that from COX-2 [15].
To determine whether the putative MBP-1 transcript
was translationally functional, experiments using tran-
scription ⁄ translation-coupled reactions were per-

formed. To mimic endogenous mRNA, we constructed
a recombinant plasmid, MBP-1 ⁄ FL, which contains
putative MBP-1 cDNA along with its 5¢- and 3¢-UTRs
as well as its ORF fused with the hemagglutinin (HA)-
tag sequence. Using T7 RNA polymerase in the
coupled reticulocyte lysate system, we detected a radio-
labeled protein with an apparent molecular mass of
37 kDa (Fig. 4D), indicating that this transcript was
AB
CD
Fig. 4. The MBP-1 transcript is functionally translatable. (A) The 1-kb promoter sequence of the putative MBP-1 gene was amplified by PCR
and cloned into the pGL3-Basic reporter vector. The resultant MBP-1-p-Luc reporter plasmid (1 lg) and pRL-TK plasmid (0.05 lg), which was
used to correct the differences in transfection efficiency between experiments, were co-transfected into BEAS-2B cells and cultured for
12 h. The transfected cells were treated with (+) or without ()) 150 m
M CoCl
2
for 24 h. The luciferase activity of each lysate was deter-
mined using a dual-luciferase reporter system (Promega). Data are the mean ± SEM luciferase activities of four independent experiments.
VC, control vector. (B) The cytoplasmic (C) and nuclear (N) fractions of BEAS-2B and A549 cells were fractionated as described in the Materi-
als and methods. Western blot analysis of one-tenth of the original volume of each fractionated lysate was carried out to examine the frac-
tionation efficiency by probing with antibody specific to a cytosolic or a nuclear protein marker (GAPDH or Histone 3, respectively). (C) The
total RNA of the remaining lysates was extracted and reverse transcribed into cDNA using the oligo-(dT)
18
primer. The MBP-1 and ENO1
cDNAs were detected by PCR amplification, and then the amplicons were resolved on 1% agarose gels containing ethidium bromide. A
blank control with no cDNA is marked ‘B’. (D) In vitro transcription ⁄ translation-coupled reactions were performed using the putative full-
length MBP-1 (MBP-1 ⁄ FL) plasmid or a control vector (VC) as templates (lower panel). The inserted gene was transcribed by T7 RNA
polymerase, expressed in a [
35
S] methionine-containing reticulocyte lysate system with a protease inhibitor cocktail, and visualized by

autoradiography.
Expression and regulation of MBP-1 J. Lung et al.
4312 FEBS Journal 277 (2010) 4308–4321 ª 2010 The Authors Journal compilation ª 2010 FEBS
translatable in vitro. Collectively, our data confirm that
the MBP-1 transcript is a product of alternative tran-
scriptional initiation of ENO1 and is transcribed from
the intron III of the gene. The transcript can
be exported to the cytoplasm, and can be translated
in vitro.
Susceptibility of the MBP-1 protein to
ubiquitin-dependent degradation
Surprisingly, when the same construct was transfected
into Cos-7 and HEK293 cells, there was no immunore-
activity against HA. To investigate whether this out-
come was because of the poor stability of the MBP-1
protein, the transfected cells were treated with a vari-
ety of protease and proteasome inhibitors. Only the
ubiquitin-associated proteasome inhibitor, MG132,
could effectively stabilize MBP-1 protein in the cells;
lactacystin, a 20S proteasome inhibitor, provided only
a minor protective effect (Fig. 5A), and chymotrypsin-
and trypsin-like serine protease inhibitors (tosyl-l-
lysine chloromethyl ketone and tosyl-l-phenylalanine
chloromethyl ketone), a lysosome protease inhibitor
(chloroquine), and a serine or cysteine protease inhibi-
tor (leupeptin) had no effect on MBP-1 stability. Thus,
although the MBP-1 transcript is translatable in vitro,
the MBP-1 protein is susceptible to proteolysis, requir-
ing the presence of the proteasome inhibitor MG132
for stability in cells.

To further verify the poor protein stability of
MBP-1 under normal culture conditions, its half-life
and its ubiquitination state in cells were determined by
AB
C
D
E
Fig. 5. MBP-1 is a stability-liable protein regulated by ubiquitin-dependent degradation. (A) Cos-7 or HEK293 cells were transfected with the
MBP-1 ⁄ FL gene and incubated for 24 h. Then, the transfected cells were treated with tosyl-
L-lysine chloromethyl ketone (TL; 100 lM), tosyl-
L
-phenylalanine chloromethyl ketone (TP; 20 lM), leupeptin (LU; 10 lgÆmL
)1
), chloroquine (CQ; 150 lM), MG132 (MG; 10 lM), lactacystin
(LC; 10 l
M), or dimethylsulfoxide vehicle control (C), for 24 h. (B) Cos-7 cells were transfected with the MBP-1 ⁄ ORF plasmid. After 24 h of
incubation, the transfected cells were treated with various doses of MG132 or dimethylsulfoxide vehicle control (C), or untreated ()), as
indicated, for 12 h. The lysates were resolved on 10% SDS ⁄ PAGE, blotted onto nitrocellulose membrane and probed with antibody against
HA-tagged MBP-1. The protein b-actin served as a loading control. (C) After transfection with MBP-1 ⁄ FL or ENO1 ⁄ FL genes for 24 h, the
cells were treated with (+) or without ())10l
M MG132 for an additional 24 h, lysed, then immunoblotted for HA-tagged MBP-1 or ENO1,
respectively. MBP-1 (37 kDa), alternatively translated from the ENO1 ⁄ FL gene, is marked with an arrow at the right. (D) The half-life of
MBP-1 was measured following treatment with CHX. Cos-7 cells were transfected with the MBP-1 ⁄ ORF gene, incubated for 20 h, then pre-
treated with (+) or without ()) MG132 (10 l
M) for an additional 6 h. After washing, translation of the MG132-pretreated cells was blocked
with 100 lgÆmL
)1
of CHX for the designated periods of time in the presence or absence of MG132, as indicated at the top. The levels of
MBP-1-HA and endogenous ENO1 proteins were determined by western blotting analyses using antibodies specific to HA or ENO1, respec-
tively. (E) Cos-7 cells were co-transfected with MBP-1 ⁄ ORF, or the vector control (VC), and the pcDNA-Flag-Ub plasmid encoding flag-tagged

ubiquitin. After 24 h of incubation, the transfected cells were treated (+) or untreated ()) with 10 l
M MG132 for an additional 24 h, as indi-
cated, and lysed. The ubiquitinization of MBP-1 was determined by precipitating with an anti-HA IgG, followed by resolution of the precipitants
in SDS ⁄ PAGE, and immunoprobing for flag-tagged ubiquitin. The heavy chain of the antibody loaded for immunoprecipitation is indicated
by an arrow on the right.
J. Lung et al. Expression and regulation of MBP-1
FEBS Journal 277 (2010) 4308–4321 ª 2010 The Authors Journal compilation ª 2010 FEBS 4313
treatment with cycloheximide (CHX) and immunopre-
cipitation of ubiquitinated MBP-1 protein, respectively.
Previous studies have shown that MBP-1 protein can
be expressed in cells and function as a c-myc pro-
moter-binding protein [2,16] by ectopic expression of
its ORF conjugated with the 5¢-Kozak sequence
(MBP-1 ⁄ ORF). In Fig. 5B, we demonstrated that the
MBP-1 level could be further elevated fivefold follow-
ing treatment with ‡ 2 lm MG132 when compared
with the level in cells treated with a control vehicle.
The same observations were obtained from cells trans-
fected with newly discovered full-length MBP-1 (MBP-
1 ⁄ FL) (Fig. 5C, left panel) or with ENO1 (ENO1 ⁄ FL)
(Fig. 5C, right panel), showing that MG132 could
stabilize MBP-1 protein generated by either of these
constructs. To measure the half-life of MBP-1, cells
were transfected with MBP-1 ⁄ ORF for 20 h, then pre-
treated, or not, with MG132, as indicated in Fig. 5D.
After removal of MG132, the transfected cells were
incubated in medium containing either a combination
of MG132 + CHX or CHX alone, for different peri-
ods of time. The half-life of MBP-1 in cells treated
with CHX alone was about 2.5 h, whereas the half-life

was > 6 h in cells treated with a combination of
MG132 and CHX. By contrast, no change was found
in the endogenous ENO1 protein level during those
treatments. To determine whether the stabilizing effect
of MG132 occurred through the blockade of ubiqu-
itin-mediated degradation pathways, the MBP-1-trans-
fected cells were treated or untreated with MG132.
The ubiquitinated MBP-1 was detected exclusively in
cells treated with MG132 (Fig. 5E). Collectively, the
data confirm that MBP-1 is a stability-labile protein
and is susceptible to ubiquitin-dependent degradation
pathways; however, its long form, ENO1, is substan-
tially stable under normal culture conditions.
Protein synthesis of MBP-1 is tightly regulated
In addition to protein stability, the low abundance of
a protein has been attributed to translational regula-
tion of its transcript via the 5¢-UTR [17]. To investi-
gate whether this might be a cause of the low
abundance of MBP-1 in cells, we generated three plas-
mids: one with the MBP-1 ORF insert only (MBP-
1 ⁄ ORF); one with the full-length MBP-1 cDNA
(MBP-1 ⁄ FL); and one with the full-length cDNA lack-
ing the first 206 nucleotides (MBP-1 ⁄ s-UTR). The
MBP-1 ⁄ ORF construct contains the Kozak sequence
followed by the AUG translation start codon. Western
blot analysis of the MBP-1 protein level in cells trans-
fected with each plasmid demonstrated that the
translation efficiency of MBP-1 ⁄ FL with the 5¢-UTR
was decreased by as much as 14.8-fold when compared
with that of cells transfected with MBP-1 ⁄ ORF (Fig. 6).

Moreover, deletion of the first 206 nucleotides in the
5¢-UTR entirely ablated the ability of the MBP-1 tran-
script to encode its protein, suggesting that this
sequence is required for ribosome recognition or bind-
ing. Thus, the low abundance of MBP-1 in cells appears
to be a result of its high turnover rate and the tightly
regulated translation control of the 5¢-UTR of its
transcript.
The ENO1 transcript preferentially encodes ENO1
protein in cells
The results of previous reports showed that ectopic
expression of the ENO1 ORF in cells produced the
48 kDa ENO1 and its short form, the 37 kDa protein
MBP-1 [7,12,13]. To examine whether ENO1 contain-
ing intact UTRs could efficiently encode MBP-1, we
expressed an ENO1 transcript with its 5¢- and 3¢-UTRs
(ENO1 ⁄ FL) to mimic its endogenous native form, with
the first coding AUG mutated (ENO1-mut ⁄
FL), or
with the 3¢-UTR alone (ENO1 ⁄ 3¢-UTR). The results in
Fig. 7 basically agree with previous reports, showing
Fig. 6. Translational regulation of MBP-1 synthesis. Cells were
transfected with MBP-1 ⁄ ORF, MBP-1 ⁄ FL, the full-length MBP-1
construct lacking the first 206 nucleotides in the 5¢-UTR (MBP-1 ⁄ s-
UTR), or a control vector (VC), in the presence or absence of 10 l
M
MG132, as indicated. The sample lysates (80 lgÆsample
)1
) were
resolved by 10% SDS ⁄ PAGE, blotted onto nitrocellulose membrane

and probed with antibody against HA-tag. The protein bands were
quantified using
QUANTITY ONE software (Bio-Rad Laboratories). Data
obtained from three independent experiments were normalized to
b-actin and are expressed as mean ± SEM of the MBP-1 protein
levels compared to cells transfected with MBP-1 ⁄ FL.
Expression and regulation of MBP-1 J. Lung et al.
4314 FEBS Journal 277 (2010) 4308–4321 ª 2010 The Authors Journal compilation ª 2010 FEBS
that MBP-1 can be synthesized from a transcript with
the ORF of ENO1 alone (ENO1 ⁄ ORF). Comparable
MBP-1 expression was also observed in cells transfect-
ed with ENO1⁄ 3¢-UTR or ENO1-mut ⁄ FL plasmids.
However, a very low amount of MBP-1 was synthe-
sized in the ENO1 ⁄ FL-transfected cells, indicating that
the endogenous ENO1 transcript preferentially encodes
the 48 kDa ENO1, and not the 37 kDa MBP-1.
The MBP-1 transcript generates MBP-1 protein
more efficiently
To determine which transcript generated the MBP-1
protein more efficiently, cells were transfected with
full-length ENO1 (ENO1 ⁄ FL) or MBP-1 (MBP-1 ⁄ FL)
cDNA containing their intact UTR sequences.
Additionally, a full-length ENO1 cDNA construct with
a mutated first ATG codon (ENO1 ⁄ mut-FL) served as
a control. Co-transfection of green fluorescence protein
into the cells was used to examine the transfection effi-
ciency of each construct. No significant differences
regarding transfection were found for these constructs,
with their transfection efficiencies ranging from 68.3%
to 76.8%. The mRNA and protein levels of ENO1

and MBP-1 were examined using RT-PCR and western
blot analyses, respectively. The results in Fig. 8 dem-
onstrate that, in the presence of similar levels of each
transcript, 17.8 times more MBP-1 protein is synthe-
sized from the newly discovered MBP-1 transcript
(MBP-1 ⁄ FL) than from the ENO1 ⁄ FL transcript.
Hence, we suggest that this newly discovered transcript
is a genuine template for the synthesis of the MBP-1
tumor suppressor in cells.
Discussion
The role of MBP-1 in cell death and tumor suppres-
sion has been previously demonstrated, and MBP-1 is
believed to be produced through an alternative transla-
tion process that utilizes the internal translation-initia-
tion sites located 374 and 383 nucleotides downstream
on the ENO1 cDNA [13]. In the present study, we
found a novel transcript containing the MBP-1 ORF,
an alternative transcriptional product of ENO1 that is
expressed in a variety of tissues. After cloning its full-
length gene and examining its expression in vitro and
in vivo, we demonstrated that the transcript is func-
tional and has 17.8-fold higher efficiency for coding
the MBP-1 protein than the full-length ENO1 cDNA
with intact UTRs. Moreover, the ENO1 transcript
preferentially generates its own protein rather than
MBP-1, indicating that the first AUG codon of the
ENO1 transcript is optimized for ribosome scanning,
while the alternative start codons are not. Thus, our
data suggest that this transcript could be a genuine
template for the MBP-1 protein in cells.

Most mammalian genes do not appear to conform
to a simple model in which a TATA box directs tran-
scription from a single defined nucleotide position, but
instead they contain multiple transcription-initiation
sites. Intriguingly, it appears that these alternative sites
are generally used in different cell contexts or tissues.
The UDP-glucuronosyltransferase locus, for example,
has at least seven promoters with different tissue-
expression profiles [18]. The human platelet-derived
growth-factor receptor gene has two transcription-initi-
ation sites [19]. One is in the intron 1, and the resulting
transcript contains 363 bp in its 5¢-UTR. Similarly, the
MBP-1 transcript described here contains a sequence
Fig. 7. The native ENO1 transcript preferentially encodes its own
protein. Cells were transfected with a plasmid containing full-length
ENO1 cDNA (ENO1 ⁄ FL), full-length ENO1 with the first ATG start
codon mutated (ENO1-mut ⁄ FL), its ORF plus the 3¢-UTR (ENO1 ⁄ 3¢-
UTR) with Kozak’s sequence, or its ORF alone with Kozak’s
sequence (ENO1 ⁄ ORF), in the presence of 10 l
M MG132. The
levels of exogenous ENO1 and its translational alternative MBP-1
proteins were measured by western blotting analysis using anti-HA
IgG. Cells transfected with MBP-1 ⁄ ORF or a control vector (VC)
served as positive and negative controls, respectively. The protein
bands were quantified using
QUANTITY ONE software (Bio-Rad Labora-
tories). Data obtained from three independent experiments were
normalized to b-actin and are presented as means ± SEM of the
protein levels of MBP-1-HA compared to cells transfected with
ENO1 ⁄ ORF (lower panel).

J. Lung et al. Expression and regulation of MBP-1
FEBS Journal 277 (2010) 4308–4321 ª 2010 The Authors Journal compilation ª 2010 FEBS 4315
of 1063 nucleotides located in the intron III of
the ENO1 gene locus, indicating that this transcript
is an alternative transcription initiation product of
ENO1.
Although the MBP-1 transcript is constitutively
expressed in a variety of cell types, its encoded protein
was barely detectable. Previous studies have demon-
strated the role of MBP-1 in cell apoptosis and tumor
suppression in an overexpression system [2–4]. Hence,
the level of MBP-1 appears to have an impact on cell
fate. During normal cell culture, cells grow continually,
and a low level of MBP-1 is expected. Ray and Miller
[1] characterized this MBP-1 protein as 37 kDa in
HeLa and HL-60 cells using a radiolabeled c-myc P2
promoter DNA fragment as a probe in a Southwestern
blot. Consistently, endogenous MBP-1 protein has not
been detected in the cell lines examined so far, such as
A549, HeLa, CL1-0 and CL1-5, as well as in the breast
cancer cells used in the present study, with our
previously published ENO1-specific antibody [20], anti-
serum raised against C-terminal ENO1, or with a com-
mercially available mAb in our chemiluminescence
system (data not shown). Experiments using MG132
and CHX, as well as immunoprecipitation of MBP-1
and blotting to determine its polyubiquitinization
(Fig. 5), confirmed that MBP-1 was a stability-liable
protein with a half-life of about 2.5 h in a cell. In
addition to its susceptibility to ubiquitin-dependent

degradation processes, the 5¢-UTR of the MBP-1 tran-
script also tightly regulated the expression of its pro-
tein, resulting in a low abundance of MBP-1 under
normal culture conditions. Although a detectable level
of endogenous MBP-1 has been shown in some cell
lines [12], the two factors contributing to its low abun-
dance (i.e. labile protein stability and tightly regulated
translation of its transcript) may vary among cell
AB
C
Fig. 8. The translational efficiency of the MBP-1 transcript to generate MBP-1 protein is markedly higher than that of the ENO1 transcript.
Cells were transfected with ENO1 ⁄ FL, ENO1-mut ⁄ FL, MBP-1 ⁄ FL, or vector control (VC) in the presence of 10 l
M MG132, as indicated.
Each sample was divided into two aliquots. One was used for RT-PCR analysis and the other was used for western blot analysis. Three
independent experiments were carried out. One representative result is shown here. (A) For RT-PCR analysis, the cytoplasmic RNA of the
transfected cells was reverse transcribed into cDNA. The level of each exogenous gene transcript was determined by PCR amplification
using a pair of primers probing for the common region of ENO1 and MBP-1 cDNA as well as for the HA-tag sequence. The ENO1-HA and
MBP-1-HA PCR amplicons was quantified using
QUANTITY ONE software (Bio-Rad Laboratories). (B) For western blot analysis, the lysate of
each transfected cell type was resolved by 10% SDS ⁄ PAGE, blotted onto nitrocellulose membrane and probed with anti-HA IgG. The level
of b-actin protein served as a control. (C) After being quantified and normalized to b-actin levels, data were expressed as fold increase in the
protein ⁄ mRNA ratio of MBP-1-HA compared to cells transfected with ENO1 ⁄ FL.
Expression and regulation of MBP-1 J. Lung et al.
4316 FEBS Journal 277 (2010) 4308–4321 ª 2010 The Authors Journal compilation ª 2010 FEBS
types. Furthermore, sensitivity of the antibody and
detection systems used in different laboratories may
also affect apparent differences in abundance.
The MBP-1 protein can be produced from at least
two templates that are generated through distinct cellu-
lar processes. One template is derived from alternative

translation initiation of the ENO1 transcript, and the
other is transcribed from an alternative transcription
start site in the ENO1 gene locus. In a comparison of
their relative mRNA abundance in the same cell, the
RT-qPCR data in Fig. 3 support those in a previous
study [13] and show that the level of the MBP-1 tran-
script is much lower than that of the ENO1 transcript.
Using the radiolabeled ENO1 riboprobe in an RNase
protection assay, Subramanian et al. [13] also detected
only very low levels of the endogenous MBP-1 tran-
script in HeLa cells. At that time, they did not find
evidence of a transcript for MBP-1, and suggested that
MBP-1 was exclusively synthesized by utilization of
the alternative AUG codons of the ENO1 transcript.
In the present study, we demonstrated that this newly
discovered MBP-1 transcript has a translation effi-
ciency 17.8-fold higher for MBP-1 than that of the
ENO1 transcript, suggesting that a subtle change in
the cellular level of the MBP-1 transcripts may have a
more profound effect on cell-growth regulation than
that of the ENO1. Furthermore, our data indicate that
the endogenous ENO1 transcript appears to preferen-
tially produce its own protein in cells, rather than the
MBP-1 protein (Figs 7 and 8). Despite the results dis-
cussed above, the MBP-1 transcript is probably still a
minor template for MBP-1 protein production in grow-
ing cells because its high translation efficiency is insuffi-
cient to compensate for its low abundance. In general,
the transcript level is normally more than 100-fold lower
than that of ENO1 in a variety of normal and cancer

cells (Figs 2 and 3). In response to hypoxia, the MBP-1
transcript level can be elevated by up to 4.3-fold
(Fig. 3A). Therefore, it is reasonable to speculate
whether an as-yet-undefined physiological condition,
individually or in combination with hypoxia, favors the
activation of MBP-1 gene transcription. The increased
levels of this high translation-efficient MBP-1 transcript
may, in turn, become one of major templates responsi-
ble for MBP-1 synthesis in the cells. Undoubtedly, this
hypothesis will be an interesting subject of future study.
For optimal expression of MBP-1 in cells, attenua-
tion of its degradation is prerequisite. In the present
study we showed that the level of MBP-1 was dramat-
ically elevated in cells transfected with MBP-1 ⁄ ORF
or MBP-1 ⁄ FL plasmids in the presence of the ubiqu-
itin-dependent proteasome inhibitor, MG132 (Figs 5
and 6), suggesting the presence of a high MBP-1
protein-turnover rate in cells. Similarly to our obser-
vation with MBP-1, HIF1a is also susceptible to
ubiquitin-mediated degradation under normoxic condi-
tions, in which key proline residues of HIF1a are
hydroxylated and then bound by von Hipple-Lindau
protein, leading to its ubiquitination and degradation
[21,22]. However, it is stabilized in hypoxic conditions.
Therefore, we examined whether MBP-1 could also be
stabilized under hypoxia. Experiments using 2% O
2
or
CoCl
2

to mimic hypoxia showed that both treatments
failed to rescue the endogenous MBP-1 protein in
BEAS-2B and A549 cells (Fig. 3B). More recently,
researchers from Miller’s laboratory reported that
MBP-1 expression was increased in cells grown in
low-glucose concentrations [7]. However, our prelimin-
ary studies using 1 nm glucose or serum-free medium
for MCF7 and MBP-1 ⁄ ORF-transfected cells did not
support their results (data not shown). In spite of
those inconsistencies, it is reasonable that the expres-
sion of MBP-1 may be modulated by unique environ-
mental factors, which are different from those
modulating HIF1a expression, presumably because of
their distinct roles in tumorigenesis. Thus, identifica-
tion of physiological factors that increase MBP-1 pro-
tein synthesis and attenuate the MBP-1 turnover rate
is an important challenge for future research related
to cancer therapy.
Materials and methods
Cell culture and cell transfection
Normal human bronchial epithelium primary cells were
grown in bronchial ⁄ tracheal epithelial growth medium sup-
plemented with essential growth factors, as recommended
by the supplier (Cell Applications, San Diego, CA, USA).
WI-38 was cultured in Eagle’s-modified minimum essential
medium (Hyclone, Logan, UT, USA) containing 10% fetal
bovine serum (Hyclone) and 1 · penicillin ⁄ streptomy-
cin ⁄ glutamine (Gibco, Carlsbad, CA, USA). NCI-H23 and
A549 lung adenocarcinoma cells were cultured in RPMI-
1640 (Hyclone) and F12K (Invitrogen, Carlsbad, CA,

USA) media containing 5% or 10% fetal bovine serum,
respectively. HEK293, Cos-7, BEAS-2B, MCF7, BT-474,
MDA-MB-231 and SK-BR3 cells were cultured in
Dulbecco’s modified Eagle’s medium (DMEM) (Hyclone)
containing 25 m m glucose and 10% fetal bovine serum (FBS).
CL1-0 and CL1-5 lung adenocarcinoma cells, generously
provided by Dr Pan-Chyr Yang [23], were cultured in 10%
FBS-containing RPMI-1640. All cell lines were grown in a
humidified atmosphere at 37 °C with 5% CO
2
. Transfection
of HEK293 and Cos-7 cells was performed using
Lipofectamine and Lipofectamine 2000 (Invitrogen), respec-
tively, according to the manufacturers’ instructions.
J. Lung et al. Expression and regulation of MBP-1
FEBS Journal 277 (2010) 4308–4321 ª 2010 The Authors Journal compilation ª 2010 FEBS 4317
Gene cloning and plasmid construction
We used 5¢-RACE to clone additional sequences upstream
of the AL833741 site. CL1-5 cells were lysed in lysis buffer
[20 mm Tris ⁄ HCl, 100 mm KCl, 5 mm MgCl
2
, 0.3% Noni-
det P-40, 100 unitsÆmL
)1
of RNase inhibitor (Invitrogen)
and a protease inhibitor cocktail (Roche Applied Science,
Indianapolis, IN, USA), pH 7.5] and incubated on ice for
5 min. The cytoplasmic fraction was obtained by centrifu-
gation (10 000 g, 10 min, 4 °C). Then, the cytoplasmic
polyA

+
-RNA was purified with Oligtex dT beads (Qiagen,
Valencia, CA, USA) and reverse transcribed into a 5¢- and
3¢-adaptor-tagged cDNA pool using the SMART RACE
kit (Clontech, Palo Alto, CA, USA). The sequence
upstream of AL833741 was obtained via a sequential nest-
ing-PCR using P1, P2 and P3 reverse primers (Fig. 1 and
Table 1) and cloned into a pCR4TOPO vector (Invitrogen).
To verify the existence of the full-length cDNA in the cells,
PCR amplification used a 5¢-RACE adaptor forward pri-
mer as well as a reverse primer to align the 3¢ end of the
ENO1 cDNA (GDB access number: NM_001428). Follow-
ing gel purification, the fragment was cloned and
sequenced. This construct was designated as pCR-MBP-
1 ⁄ FL.
The MBP-1 and ENO1 cDNA constructs, with or with-
out their individual UTRs, were engineered as follows.
First, to express full-length ENO1, a construct was gener-
ated from CL1-5 cDNA using primers that annealed with
the 5¢ and 3¢ ends of the ENO1 complete cDNA (the primer
sequences are shown in Table 1). After amplification with
PfuUltra DNA polymerase (Stratagene, La Jolla, CA,
USA), the PCR product was digested with BamHI and
EcoRV and ligated into the pcDNA3.1 vector. The inser-
tion of an HA-tag sequence into the 3¢ end of the ENO1
ORF was performed by PCR-based site-directed mutagene-
sis to obtain the ENO1 ⁄ FL plasmid. Second, using this
plasmid as a template, the first ATG codon of the ENO1
cDNA was mutated to AAG by site-directed mutagenesis
using two antiparallel primers (Table 1). The resulting con-

struct was designated ENO1-mut ⁄ FL. Third, after PCR
amplification, the fragment lacking the 5¢-UTR of ENO1
was digested with BamHI and EcoRV and cloned into
pcDNA3.1 to obtain the ENO1 ⁄ 3¢-UTR plasmid. Fourth,
using the ENO1 ⁄ FL plasmid, created earlier, as a template,
the ORF of ENO1 was amplified, digested with BamHI and
NotI, and ligated into a vector to obtain the ENO1 ⁄ ORF
plasmid. Fifth, the MBP-1 ⁄ FL plasmid was constructed by
PCR amplification using the pCR-MBP-1 ⁄ FL plasmid as a
template, which was followed by ligation of the 5¢-HindIII
and 3¢-EcoRV-cut fragment into pcDNA3.1. Sixth, the
MBP-1 ⁄ s-UTR plasmid was engineered similarly, but
lacked the first 206 nucleotides in the 5¢-UTR of MBP-1.
Seventh, after PCR amplification and restriction enzyme
digestion, the BamHI and NotI-digested fragment contain-
ing the ORF of MBP-1 was cloned into the pcDNA3.1
vector to obtain MBP-1 ⁄ ORF. Eighth, the insert encoding
N-terminal flag-tagged ubiquitin was generated by PCR
using HeLa cDNA as template, as described by Sadeh et al.
[24], and cloned into the 5¢-BamHI and 3¢-EcoRV sites of
pcDNA3.1. The resultant plasmid was designated as
pcDNA-Flag-Ub. Finally, the 1 kb sequence upstream of
the newly discovered MBP-1 transcript was amplified by
PCR, using BEAS-2B genomic DNA as the template, and
cloned into the 5¢-BamHI and 3¢-EcoRV sites of the pGL3-
Basic vector (Promega, Madison, WI, USA). The resultant
plasmid was designated as MBP-1-p-Luc reporter. All gene
constructs were verified by direct sequencing.
Cell fractionation and RNA isolation
BEAS-2B and A549 cells were treated with 150 lm CoCl

2
or left untreated. After washing with NaCl ⁄ P
i
, the cells
were centrifuged at 250 g for 5 min at room temperature,
resuspended in RSB buffer (10 m m NaCl, 3 mm MgCl
2
,
10 mm Tris ⁄ HCl, pH7.5) containing protease inhibitor
cocktail (Roche Applied Science, Penzberg, Germany) and
1000 unitsÆmL
)1
of RNaseOUT (Invitrogen) and incubated
on ice for 15 min. Igepal CA630 (Sigma, St Louis, MO,
USA) was then added to a final concentration of 0.05%,
mixed by inversion and the cells were placed on ice for an
additional 10 min. Cell fractionation was carried out by
centrifugation at 200 · g for 5 min at room temperature to
separate the cytoplasm from the nuclear fraction. The pellet
containing the nuclear fraction was washed twice with RSB
buffer and resuspended in a volume of buffer equal to that
of the cytoplasmic fraction. One-tenth of the original vol-
ume was analyzed by western blotting to confirm the purity
of each fractionation. The total RNA of the remaining
cytoplasmic and nuclear fractions was extracted using the
RNeasy purification kit (Qiagen), primed with oligo-(dT)
18
primer and reverse transcribed into cDNA.
RT-PCR analysis
Cells were lysed in the aforementioned lysis buffer and

incubated on ice for 5 min. The cytoplasmic polyA
+
-RNA
was purified with Oligtex dT beads and reverse transcribed
into cDNA using SuperScript III reverse transcriptase
according to the manufacturer’s instructions (Invitrogen).
To detect the endogenous putative MBP-1 and ENO1 tran-
scripts, semi- and RT-qPCR analyses of the cDNA pools
generated from various cell lines, obtained as indicated
above or purchased from Clontech Laboratories (Mountain
View, CA, USA), were performed using the primers shown
in Table 1, specific for their corresponding 5¢-UTR
sequences. Normal lung tissue cDNA was purchased from
ResGen, Invitrogen (Huntsville, AL, USA). For semi-qPCR
analyses, the resultant amplicons were resolved in 1% aga-
rose gels containing ethidium bromide and quantified using
quantity one software in the Bio-Rad Gel Doc system
Expression and regulation of MBP-1 J. Lung et al.
4318 FEBS Journal 277 (2010) 4308–4321 ª 2010 The Authors Journal compilation ª 2010 FEBS
(Bio-Rad Laboratories, Hercules, CA, USA). b-actin was
used as a loading control. For RT-qPCR analyses, the tran-
script levels of putative MBP-1 and ENO1 in each sample
were determined using SRBR green qPCR master mix
(Stratagene) in the Bio-Rad DNA Engine OPTICON2
detection system. All reactions were performed under the
following PCR conditions: denaturation at 95 °C for
10 min, followed immediately by 40 cycles at 95 °C for 10 s
and at 60 °C for 1 min. After being normalized to expres-
sion of the b-actin control gene, the changes in the compar-
ative threshold (C

t
) data were calculated as )4C
t
=
)[C
t(ENO1 or MBP-1)
) C
t(b-actin)
]. The ratio of ENO1 or
MBP-1 mRNA to the b-actin mRNA was expressed as
2
ÀDC
t
 K (K, constant) [25] and the results are shown as
mean ± SEM. The means and SEM values were derived
from three independent experiments.
Table 1. Primer sequences for the gene construct or detection. AS, antisense; S, sense.
Construct or gene Sequence
5 ¢-RACE
AL833741 P1 AS: 5¢-GCCAAGTCAGCGATGTGGCGG-3¢
P2 AS: 5¢-GGCTGTTCAATCAGTTACCTGAGT-3¢
P3 AS: 5¢-TCAAGGCACAGCATCATCCTGCTCC-3¢
ENO1 ⁄ FL S: 5¢-AAGGATCCTAGCTAGGCAGGAAGTCGGCG-3¢
AS: 5¢-CTCATGGGTCACTGAGGCTTTTTATTTTG-3¢
ENO1 ⁄ 3¢-UTR S: 5¢-AAGGATCCGTTCACCATGTCTATTCTCAAGATCCATG-3¢
AS: 5¢-CTCATGGGTCACTGAGGCTTTTTATTTTG-3¢
ENO1 ⁄ ORF S: 5¢-AAGGATCCGTTCACCATGTCTATTCTCAAGATCCATG-3¢
AS: 5¢-CACAGCGGCCGCTTAAGCGTAATCTGGAACATCG-3¢
MBP-1 ⁄ FL S: 5¢-CCCAAGCTTAACTAAAGAAAAGTTTCCCCATCTCCC -3¢
AS: 5¢-CTCATGGGTCACTGAGGCTTTTTATTTTG-3¢

MBP-1 ⁄ s-UTR S: 5¢-CCCAAGCTTTATTCAAATTGACTTAGTGTGTGTG-3¢
AS: 5¢-CTCATGGGTCACTGAGGCTTTTTATTTTG-3¢
MBP-1 ⁄ ORF S: 5¢
-AAGGATCCGTTCACCATGGATGGAACAGAAAATAAATC-3¢
AS: 5¢-CACAGCGGCCGCTTAAGCGTAATCTGGAACATCG-3¢
Flag-Ub S: 5¢-CCCGGATCCATGGATTACAAGGATGACGACGATAAGATGCAAATCTTCGTGAAGACCCTGACTG-3¢
AS: 5¢-TTAACCCCCCCTCAAGCGCAGGACCAAG-3¢
Mutagenesis
ENO1-mut ⁄ FL S: 5¢-AAGTCTATTCTCAAGATCCATGC-3¢
AS: 5¢-GGTGAACTTCTAGCCACTGG-3¢
HA-tag insertion S: 5¢-AGTACCCATACGATGTTCCAGATTACGCTTAAGCTG TGGGCAGGCAAGCC-3¢
AS: 5¢-TAAGCGTAATCTGGAACATCGTATGGGTACTTGGCC AAGGGGTTTCTGAA-3¢
Semi-qPCR
b-actin S: 5¢-GGTCACCCACACTGTGCCCATCTA-3¢
AS: 5¢-GAAGCATTGCGGTGGACGATGGAG-3¢
ENO1 S: 5¢-CACAGTGACCAACCCAAAGAG-3¢ (exon 9)
AS: 5¢-CTCATGGGTCACTGAGGCTTTTTATTTTG-3¢ (exon 12)
Putative MBP-1 S: 5¢-GTCTCTCACTCGCAGTCCTTAATC-3¢ (intron 3)
AS: 5¢-GCCAAGTCAGCGATGTGGCGG-3¢ (exon 6)
ENO1-HA or
MBP-1-HA S: 5¢-ACTGATGATCGAGATGGATGGAACA-3¢ (exon 5)
AS: 5¢-TTAAGCGTAATCTGGAACATCG-3¢
Cytoplasmic and nuclear MBP-1
and ENO1 mRNA levels
S: 5¢-GTCTCTCACTCGCAGTCCTTAATC-3¢ (in 5¢-UTR of MBP-1)
S: 5¢-AAGGATCCTAGCTAGGCAGGA AGTCGGCG-3¢ (in 5¢-UTR of ENO1)
As: 5¢-AATGAATTCTTACTTGGCCAAGGGGTTTCT-3¢
Real-time qPCR
b-actin S: 5¢-CCCTTTTTGTCCCCCAAC-3¢
AS: 5¢-CTGGTCTCAAGTCAGTGTACAGGT-3¢

ENO1 S: 5¢-GATCTCTTCACCTCAAAAGG-3¢ (exon 2)
AS: 5¢-TTCCATCCATCTCGATCATC-3¢ (exon 5)
Putative MBP-1 S: 5¢-CAACAGTGGTTCTCTCTAAC-3¢ (intron 3)
AS: 5¢-TTCCATCCATCTCGATCATC-3¢ (exon 5)
Reporter assay
MBP-1-p-Luc S: 5¢-GAAGGTACCGACCAGCCTGGCCAACATGG-3¢
AS: 5¢-AACAGGCCCTGTTCTCATGC
-3¢
J. Lung et al. Expression and regulation of MBP-1
FEBS Journal 277 (2010) 4308–4321 ª 2010 The Authors Journal compilation ª 2010 FEBS 4319
Western blotting analysis and immuno-
precipitation
HEK293 or Cos-7 cells were transfected for 24 h with
ENO1 or MBP-1 tagged with an HA sequence, then treated
with MG132 or other protease inhibitors, as indicated, for
an additional 24 h. The transfected cells were harvested and
lysed in 1% Triton X-100, 0.5% sodium deoxycholate, 1%
SDS, 0.15 m NaCl, 10 mm Na
2
HPO
4
and 10 mm NaF (pH
7.2), supplemented with a protease inhibitor cocktail
(Roche). Approximately 100 lg of each lysate was loaded
onto and resolved in 10% SDS-containing polyacrylamide
gels, and then blotted onto nitrocellulose membranes
(Whatman Schleicher and Schuell, London, UK). Those
membranes were incubated with monoclonal anti-HA.11
(Covance, Princeton, NJ, USA) (1 : 1500) and probed with
a secondary antibody conjugated with horseradish peroxi-

dase. To detect the cytosolic protein marker GAPDH and
the nuclear marker Histone 3, antibodies from Santa Cruz
(Santa Cruz, CA, USA) and Cell Signaling (Danvers, MA,
USA), respectively, were used. As a loading control, b-actin
was detected by an antibody purchased from Sigma.
To detect ubiquitinized MBP-1, Cos-7 cells (4–5 ·
10
5
cells per well) were seeded into six-well plates 1 day
before transfection. The cells were co-transfected with 4 lg
of MBP-1 ⁄ ORF, or with 2 lg of pcDNA-Flag-Ub and 2 lg
of the control vector using Lipofectamine 2000 (Invitro-
gen), according to the manufacturer’s instructions, and
incubated for 24 h. The transfected cells were treated with
10 lm MG-132, or untreated, for 24 h and then lysed in
lysis buffer [50 mm Tris (pH 7.4), 150 mm NaCl, 0.1%
SDS, 1% Triton X-100, 1 m m EDTA, protease inhibitor
cocktail (Roche), 10 lm MG-132 and proteasome inhibitor
IV (Calbiochem, San Diego, CA, USA)]. Twenty micro-
grams of each sample was resolved in 10% SDS-containing
polyacrylamide gels, blotted onto nitrocellulose membranes
and immunodetected with antibodies against HA-tagged
MBP-1 or b-actin. The lysates (450 lg per sample) were
precleaned with 50% protein A ⁄ G slurry (Pierce Biotech-
nology, Rockford, IL, USA) and immunoprecipitated with
anti-HA.11 IgG (Covance). Subsequently, the precipitants
were resolved on SDS-containing polyacrylamide gels, blotted
onto nitrocellulose membranes and probed for flag-tagged
ubiquitin. The antigen–antibody immunocomplexes were
visualized using SuperSignal enhanced chemiluminescence

(Pierce Biotechnology).
In vitro-coupled transcription and translation
Coupled transcription and translation reaction systems
(Promega) were used to detect the translational product of
the putative MBP-1 transcript in vitro. The gene product
was generated from the MBP-1 ⁄ FL plasmid or from a con-
trol vector in a reticulocyte lysate system containing T7
RNA polymerase and [
35
S] methionine (> 1000 CiÆmmole
)1
at 10 mCiÆmL
)1
; Amersham Biosciences, Piscataway, NJ,
USA). Subsequently, the radiolabeled proteins were
resolved using 10% SDS ⁄ PAGE and visualized by autora-
diography.
Protein stability of MBP-1
The half-life of MBP-1 was determined by treatment with
CHX. Briefly, Cos-7 cells were transfected with the MBP-
1 ⁄ ORF construct using Lipofectamine 2000 (Invitrogen)
and incubated at 37 °C for 20 h. Then, the transfected cells
were, or were not, pretreated with 10 lm MG132 for an
additional 6 h. After washing in NaCl ⁄ P
i
, the translational
activity of the transfected cells pretreated with MG-132 was
blocked with 100 lgÆmL
)1
of CHX for specific periods of

time, in the presence or absence of MG132, as indicated in
Fig. 5. The levels of MBP-1-HA and endogenous ENO1
were determined by western blotting using antibodies spe-
cific to HA-tag or ENO1.
Acknowledgements
We thank Dr Peter Goedegebuure and Dr Julia Wang
at Washington University, St Louis, MO, USA for
their critical appraisal of the manuscript and are also
grateful to Dr Pan-Chyr Yang for providing the CL1-5
cells for this study. This work was supported by the
Intramural Research Funding Program of the National
Health Research Institutes, Taiwan, by the NRPGM
grant from the Department of Health, Executive Yuan,
Taiwan [DOH97-TD-G-111-022 (N.Y.S.)] and by the
National Science Council, Taiwan [NSC96-2314-B-038-
022 ⁄ NSC97-2314-B-038-004-MY3 (S.J.L.) and NSC98-
3112-B- 400-012 (K.J.L.)].
References
1 Ray R & Miller DM (1991) Cloning and characteriza-
tion of a human c-myc promoter-binding protein. Mol
Cell Biol 11, 2154–2161.
2 Ray RB, Steele R, Seftor E & Hendrix M (1995)
Human breast carcinoma cells transfected with the gene
encoding a c-myc promoter-binding protein (MBP-1)
inhibits tumors in nude mice. Cancer Res 55, 3747–
3751.
3 Ejeskar K, Krona C, Caren H, Zaibak F, Li L,
Martinsson T & Ioannou PA (2005) Introduction of in
vitro transcribed ENO1 mRNA into neuroblastoma
cells induces cell death. BMC Cancer 5, 161.

4 Ghosh AK, Steele R, Ryerse J & Ray RB (2006)
Tumor-suppressive effects of MBP-1 in non-small cell
lung cancer cells. Cancer Res 66, 11907–11912.
5 Ghosh AK, Steele R & Ray RB (2005) c-myc
Promoter-binding protein 1 (MBP-1) regulates prostate
Expression and regulation of MBP-1 J. Lung et al.
4320 FEBS Journal 277 (2010) 4308–4321 ª 2010 The Authors Journal compilation ª 2010 FEBS
cancer cell growth by inhibiting MAPK pathway. J Biol
Chem 280, 14325–14330.
6 Perconti G, Ferro A, Amato F, Rubino P, Randazzo
D, Wolff T, Feo S & Giallongo A (2007) The kelch
protein NS1-BP interacts with alpha-enolase ⁄ MBP-1
and is involved in c-Myc gene transcriptional control.
Biochim Biophys Acta 1773, 1774–1785.
7 Sedoris KC, Thomas SD & Miller DM (2007) c-myc
promoter binding protein regulates the cellular response
to an altered glucose concentration. Biochemistry 46,
8659–8668.
8 Giallongo A, Feo S, Moore R, Croce CM & Showe LC
(1986) Molecular cloning and nucleotide sequence of a
full-length cDNA for human alpha enolase. Proc Natl
Acad Sci USA 83, 6741–6745.
9 Onyango P, Lubyova B, Gardellin P, Kurzbauer R &
Weith A (1998) Molecular cloning and expression anal-
ysis of five novel genes in chromosome 1p36. Genomics
50, 187–198.
10 White RA, Adkison LR, Dowler LL & Ray RB (1997)
Chromosomal localization of the human gene encoding
c-myc promoter-binding protein (MPB1) to chromo-
some 1p35-pter. Genomics 39, 406–408.

11 Ghosh AK, Steele R & Ray RB (1999) Functional
domains of c-myc promoter binding protein 1 involved
in transcriptional repression and cell growth regulation.
Mol Cell Biol 19, 2880–2886.
12 Feo S, Arcuri D, Piddini E, Passantino R & Giallongo
A (2000) ENO1 gene product binds to the c-myc pro-
moter and acts as a transcriptional repressor: relation-
ship with Myc promoter-binding protein 1 (MBP-1).
FEBS Lett 473, 47–52.
13 Subramanian A & Miller DM (2000) Structural analysis
of alpha-enolase. Mapping the functional domains
involved in down-regulation of the c-myc protoonco-
gene. J Biol Chem 275, 5958–5965.
14 Shaw RJ (2006) Glucose metabolism and cancer. Curr
Opin Cell Biol 18, 598–608.
15 Jang BC, Munoz-Najar U, Paik JH, Claffey K, Yoshida
M & Hla T (2003) Leptomycin B, an inhibitor of the
nuclear export receptor CRM1, inhibits COX-2 expres-
sion. J Biol Chem 278, 2773–2776.
16 Ray RB & Steele R (1997) Separate domains of MBP-1
involved in c-myc promoter binding and growth sup-
pressive activity. Gene 186, 175–180.
17 Vogel HJ & Vogel RH (1967) Regulation of protein
synthesis. Annu Rev Biochem 36, 519–538.
18 Carninci P, Sandelin A, Lenhard B, Katayama S,
Shimokawa K, Ponjavic J, Semple CA, Taylor MS,
Engstrom PG, Frith MC et al. (2006) Genome-wide
analysis of mammalian promoter architecture and
evolution. Nat Genet 38, 626–635.
19 Minato Y, Tashiro E, Kanai M, Nihei Y, Kodama Y &

Imoto M (2007) Transcriptional regulation of a new
variant of human platelet-derived growth factor recep-
tor alpha transcript by E2F-1. Gene 403, 89–97.
20 Chang GC, Liu KJ, Hsieh CL, Hu TS, Charoenfu-
prasert S, Liu HK, Luh KT, Hsu LH, Wu CW, Ting
CC et al. (2006) Identification of alpha-enolase as an
autoantigen in lung cancer: its overexpression is asso-
ciated with clinical outcomes. Clin Cancer Res 12,
5746–5754.
21 Epstein AC, Gleadle JM, McNeill LA, Hewitson KS,
O’Rourke J, Mole DR, Mukherji M, Metzen E, Wilson
MI, Dhanda A et al. (2001) C. elegans EGL-9 and
mammalian homologs define a family of dioxygenases
that regulate HIF by prolyl hydroxylation. Cell 107 ,
43–54.
22 Maranchie JK, Vasselli JR, Riss J, Bonifacino JS,
Linehan WM & Klausner RD (2002) The contribution
of VHL substrate binding and HIF1-alpha to the
phenotype of VHL loss in renal cell carcinoma. Cancer
Cell 1, 247–255.
23 Chu YW, Yang PC, Yang SC, Shyu YC, Hendrix
MJ, Wu R & Wu CW (1997) Selection of invasive
and metastatic subpopulations from a human lung
adenocarcinoma cell line. Am J Respir Cell Mol Biol
17, 353–360.
24 Sadeh R, Breitschopf K, Bercovich B, Zoabi M, Kravts-
ova-Ivantsiv Y, Kornitzer D, Schwartz A & Ciechanov-
er A (2008) The N-terminal domain of MyoD is
necessary and sufficient for its nuclear localization-
dependent degradation by the ubiquitin system. Proc

Natl Acad Sci USA 105, 15690–15695.
25 Yuan A, Yu CJ, Shun CT, Luh KT, Kuo SH, Lee YC
& Yang PC (2005) Total cyclooxygenase-2 mRNA lev-
els correlate with vascular endothelial growth factor
mRNA levels, tumor angiogenesis and prognosis in
non-small cell lung cancer patients. Int J Cancer 115,
545–555.
Supporting information
The following supplementary material is available:
Fig. S1. Cloning of the 5¢-untranslated region (UTR)
sequence of the full-length MBP-1 gene by rapid
amplification of cDNA ends (RACE)-coupled PCR.
This supplementary material can be found in the
online version of this article.
Please note: As a service to our authors and readers,
this journal provides supporting information supplied
by the authors. Such materials are peer-reviewed and
may be re-organized for online delivery, but are not
copy-edited or typeset. Technical support issues arising
from supporting information (other than missing files)
should be addressed to the authors.
J. Lung et al. Expression and regulation of MBP-1
FEBS Journal 277 (2010) 4308–4321 ª 2010 The Authors Journal compilation ª 2010 FEBS 4321