The expression of metastasis suppressor MIM/MTSS1 is regulated
by DNA methylation
Jochen Utikal1, Alexei Gratchev1*, Isabelle Muller-Molinet1, Sandra Oerther1, Julia Kzhyshkowska1,
Norbert Arens2, Rainer Grobholz2, Sheila Kannookadan1and Sergij Goerdt1
1Department of Dermatology, University Medical Center Mannheim, Ruprecht-Karl University of Heidelberg, Mannheim, Germany
2Institute of Pathology, University Medical Center Mannheim, Ruprecht-Karl University of Heidelberg, Mannheim, Germany
MIM/MTSS1 was initially described as a gene missing in invasive
bladder cancer cell lines. Functional analysis revealed that MIM
is an actin binding protein involved in the regulation of actin cyto-
skeleton dynamics. MIM was shown to be sonic hedgehog (Shh)
signaling dependent and synergizes with the effects of Gli tran-
scription factors. Overexpression of MIM in cell lines leads to the
inhibition of cell proliferation. In this study, we showed that the in-
hibition of cell growth by MIM is anchorage independent. We iden-
tified and cloned the promoter region of MIM and located the main
promoter activity to 276 bp of 50flanking sequence sited within a
CpG island. Analysis of DNA methylation using bisulphite sequenc-
ing revealed that MIM promoter is methylated in its 50region in
cells and tissue samples with reduced endogenous MIM expression.
Using luciferase reporter assay, we demonstrated that nonmethy-
lated MIM promoter has a similar activity in cell lines with different
endogenous MIM expression. Inhibition of DNA methylation by 5-
Aza-20-deoxycytidine led to upregulation of MIM expression in a
low expressing cell line. In conclusion, we clearly demonstrate here
that the expression of metastasis suppressor MIM is regulated by
DNA methylation of a CpG island within its promoter region.
' 2006 Wiley-Liss, Inc.
Key words: actin; tumor suppressor; carcinoma; 5-Aza-20-deoxycytidine;
Since decades DNA methylation is known to play an essential
role during tumor initiation and progression. Two main directions of
DNA methylation changes are associated with carcinogenesis: gen-
eral DNA hypomethylation and regional DNA hypermethylation.1
Hypomethylation leads to a general destabilization of the genome
and increases the frequency of chromosomal aberrations,2while the
regional DNA hypermethylation is described as an alternative for
mutation or deletion of the tumor suppressor gene leading to the loss
of function.3–5Structurally, the promoters of genes regulated by
DNA methylation may be divided into 2 groups: 1––containing
CpG island (a region with increased density of CpG dinucleotides)
and 2––lacking CpG island. Although the nonmethylated state of a
CpG island does not necessarily indicate the transcriptionally active
promoter, methylated state of a CpG island usually leads to inactiva-
tion of the promoter.3During carcinogenesis de novo methylation of
normally nonmethylated CpG island causes deacetlytion of the his-
tones, and formation of transcriptionally inactive chromatin.3Fol-
lowing tumor suppressors are examples of the genes that can be
inactivated by de novo methylation within the promoter region CpG
island: pRb in retinoblastoma,6p16INK4a in melanoma and others,7
p15INK4b in haematological malignancies,8hMLH1 and APC in
colorectal carcinoma9,10and BRCA1 in breast carcinoma.11Al-
though usually the aberrant methylation of CpG island containing
promoters is associated with carcinogenesis, changes in methylation
pattern of promoters lacking CpG island were also reported.12,13
MIM (Missing in metastasis)––also named BEG4 (basal cell car-
cinoma-enriched gene 4), MTSS1 (mammalian metastasis suppres-
sor 1), or KIAA0429––was first described by Lee et al.14as a poten-
tial metastasis suppressor in bladder cancer. MIM expression was
shown to be significantly reduced in metastatic cells in comparison
to nonmetastatic ones. MIM is located on the chromosome 8q24.1
and is expressed as a 5.3-kb transcript in human spleen, thymus,
prostate, testis, uterus, colon and peripheral blood. Mouse homo-
logue was cloned by Mattila et al.,15who also showed the function-
ality of a predicted actin binding WH2 domain.15The human and
mouse cDNAs show 85% similarity and the proteins share 96%
sequence identity. Beside the WH-2 domain, MIM contains other
function motifs, including a serine riche domain, a proline riche do-
main and a lysine rich domain. MIM also shares a homology with
the insulin receptor substrate p53 (IRSp53) in a region of 250 AA at
the N-terminus,16which is termed IMD (IRSp53/MIM homology
domain). The IMD of MIM induces filopodia in HeLa cells and the
formation of tightly packed parallel F-actin bundles in vitro.17In
addition, MIM was found to colocalize with cortactin, an ARP2/3
complex activator, and interacts directly with the cortactin SH3
The involvement of MIM in carcinogenesis was supported by
the findings of Callahan et al. who showed the association of MIM
with the sonic hedgehog (Shh) signaling pathway.19It was shown
that MIM activity can be controlled by regulating the expression
of Gli-responsive genes via sonic hedgehog (Shh) signaling. MIM
is part of a Gli/Suppressor of Fused complex and potentiates Gli-
dependent transcription independent of actin binding domains.19
While the role of MIM in carcinogenesis is established, the mecha-
nisms of its expression regulation remain unclear.
In this study, we demonstrated that the over-expression of MIM
suppresses growth of transformed and cancer cell lines. We cloned
the 50flanking region of MIM gene and identified the functional
promoter of MIM. We established that the increase of DNA meth-
ylation density within the promoter region of MIM correlates with
silencing of endogenous MIM expression and that the inhibition of
DNA methylation by 5-Aza-20-deoxcytidine in cells with low en-
dogenous MIM expression leads to its increase. Our data indicate
that the expression of metastasis suppressor MIM is regulated by
DNA methylation within promoter associated CpG island.
Material and methods
The bladder carcinoma cell lines 5637, RT-4 and TccSuP, pros-
tate carcinoma cell lines PC-3 and LNCaP and mouse fibroblasts
NIH-3T3 were obtained from DSMZ (Braunschweig, Germany).
Culture conditions were: DMEM supplemented with 10% FBS for
NIH-3T3 and TccSuP cells, Ham’s F12: DMEM (1:1) medium sup-
plemented with 10% FBS for PC-3 cells, and RPMI supplemented
with 10% FBS for 5637, RT-4 and LNCaP. All media were from
Invitrogen (Karlsruhe, Germany). FBS was from Biochrom (Berlin,
Germany). Cells were propagated at 37?C with 5% CO2.
Matched pairs of normal and carcinoma prostate tissue samples
were obtained from 5 radical prostatectomy (RP) specimens, all
The first two authors have contributed equally.
*Correspondence to: Klinik f€ ur Dermatologie, Venerologie und Aller-
gologie, Klinikum Mannheim gGmbH – Universit€ atsklinikum – Ruprecht-
Karls-Universit€ at Heidelberg, Theodor-Kutzer-Ufer 1-3, 68167 Mannheim,
Germany. Fax: 149-621-383-733103/149-621-3833815.
Received 4 March 2006; Accepted 19 April 2006
Int. J. Cancer: 119, 2287–2293 (2006)
' 2006 Wiley-Liss, Inc.
Publication of the International Union Against Cancer
histopathologically staged as pT3a and graded as Gleason score
7–9. Accuracy of tumor cell content was inspected by histomor-
phological analysis of H&E-stained tissue section of the freshly
frozen tissue samples.
Plasmids and cosmid library
Full size coding sequences for human and mouse MIM were
obtained from RZPD (Berlin, Germany). Human MIM cDNA was
obtained as a clone FLJ23362/HEP15314, and mouse MIM cDNA
as a clone IMAGp998H1711245. To generate MIM expression
constructs, the coding sequences were first cloned into pDNR2
vector (BD Clontech, Heidelberg, Germany) and then transferred
into pLP-IRESneo acceptor vector using Cre recombination proce-
dure according to the recommendation of the manufacturer.
Obtained constructs were: pLP-IRESneo-hMIM-FLAG containing
FLAG-tagged full size coding sequence of human MIM and pLP-
IRESneo-mMIM-FLAG containing FLAG-tagged full size coding
sequence of mouse MIM.
Chromosome 8 specific cosmid library was obtained as a high
density colony filter from RZPD (Berlin, Germany). Cosmid clones
detected to contain the 1st exon of MIM were as well obtained from
RZPD. Cosmids were isolated using Qiagen Large-construct kit
(Qiagen, Hildern, Germany). The clone used as a source of 50flank-
ing region of MIM was LANc151P0115Q3. All subclonings were
performed in pBluescript-SK1 vector (Stratagene).
Luciferase reporter plasmids were cloned into pGL3-basic vector
(Promega). As an internal control a reporter vector containing syn-
thetic Renilla luciferase gene––phRL-TK––was used (Promega).
Cell growth analysis
For stable transfections, NIH-3T3 and PC-3 cells were seeded
into 6-well plates and grown until 60–80% confluence. Transfec-
tion was carried out using Lipofectamine Reagent (Invitrogen,
Karlsruhe, Germany), according the recommendations of the man-
ufacturer with following constructs: pLP-IRESneo-hMIM-FLAG
for PC-3 cells and pLP-IRESneo-mMIM-FLAG for NIH-3T3 cells
as well as control plasmids pLP-IRESneo-EGFP-FLAG and pLP-
IRESneo. For both cell lines, 1 lg/well plasmids and 4 ll/well lip-
ofectamine reagent were used. Transfected NIH-3T3 and PC-3
cells were then selected in the presence of G418 at a concentration
of 100 lg/ml cell culture medium for 3–4 weeks. For further experi-
ments, only the first 3 passages of transfected cells were used.
The in vitro growth rate was assessed in stable pLP-IRESneo-
hMIM-FLAG (pLP-IRESneo-mMIM-FLAG), pLP-IRESneo-EGFP-
FLAG and pLP-IRESneo transfected PC-3 and NIH-3T3 cells. Stable
transfected cells (10,000) were seeded in a 120-mm cell culture
dishes and grown in the presence of 100 lg/ml G418. After 7 days,
cells were harvested and counted using Coulter Counter Z2
(Beckman Coulter, Freiburg, Germany). In vitro growth rate assay
was performed in triplicate for each cell line. Anchorage independent
growth assay was performed in the same way as described earlier,
using 120-mm culture dishes coated with 1.0% agar.
To perform colony forming assay, PC-3 and NIH-3T3 cells were
transfected with pLP-IRESneo-hMIM-FLAG, pLP-IRESneo-mMIM-
FLAG, pLP-IRESneo-EGFP-FLAG and pLP-IRESneo as described.
After 48 hr, the cells from each well were trypsinized and seeded into
3 new 12-cm cell culture plates and cultivated for 24 hr under normal
culture conditions. After 24 hr, G418 was added to final concentration
of 100 lg/ml. G418 supplemented medium was replaced every 4 days
until visible colonies of stable transfected cells formed (3 weeks for
NIH-3T3 and 4 weeks for PC-3 cells). Colonies were stained with
May-Grunwald’s eosine-methylene blue solution (Merck, Darmstadt,
Germany) for 5 min and colonies were counted. Experiments were
performed in triplicate.
RNA isolation and cDNA synthesis
For RNA isolation, the cells were lysed directly in the plastic
petri dishes and the RNA was isolated using RNeasy Mini kit
(Qiagen, Hilden, Germany), according to the recommendations of
The same kit was used to isolate total RNA from serial sections
of the freshly frozen tissue samples. Amount and integrity of the
isolated total RNA samples were verified using the RNA 6000
Nano lab chip kit (Agilent) and the Agilent 2100 bioanalyzer.
Total RNA (2 lg) was treated with 2U RNase free DNase
(Ambion, Austin, TX) and used for Reverse Transcription (RT)
with Superscript II reverse transcriptase (Invitrogen, Karlsruhe,
Germany) using oligo dT primers, according to the recommenda-
tions of the manufacturer. The obtained cDNA was diluted 1:10
with DDH2O and 1 ll was used for PCR reaction.
Real time PCR analysis of human MIM expression was per-
formed using Assay-on-Demand1TaqMan assay Hs00207341_m1
together with TaqMan PCR master mix (all from Applied Biosys-
tems, Darmstadt, Germany) using standard conditions. As internal
control, a predeveloped TaqMan assay for human GAPDH was used
(Applied Biosystems). The experiments were performed on ABI
PRISM17000 sequence detection system (Applied Biosystems)
and analyzed using ABI Prism RQ relative quantification software.
The expression level of MIM was normalized to GAPDH mRNA
Transient transfection and luciferase assay
Cells were transfected using lipofectamine reagent (Invitrogen),
according to the recommendations of the manufacturer. Transfec-
tion was carried out in 6-well plates. For each transfection, 3 lg of
reporter plasmid and 1 lg of internal control plasmid was taken.
After 3 days of transfection, the cells were harvested and lysed in
the passive lysis buffer provided with the luciferase assay kit. Lu-
ciferase and Renilla luciferase were measured using Dual Lucifer-
ase assay kit (Promega). Luminescence measurement was per-
formed with Lumoscan (Thermo labsystems, Frankfurt, Germany).
DNA isolation and bisulphite sequencing
Genomic DNA was isolated from the cells using QIAamp DNA
mini kit (Qiagen), according to the recommendations of the manu-
facturer. For bisulphite sequencing, 1 lg DNA was digested with
EcoRI. Obtained digested DNA was diluted and 200 ng were sub-
jected to bisulphite treatment according to the protocol described
elsewhere.13,20Amplification of the bisulfite modified DNA was
performed using PCR with AmpliTaq Gold polymerase (Applied
Biosystems). Nested PCR was performed to increase the specific-
ity of the reaction. If the primer annealing site contained a CpG di-
nucleotide, Y was on the C position in the sense primer and R in
the antisense to avoid preferential amplification of either methyl-
ated or nonmethylated sequence. Primers for the first PCR step
were: forward F640 50-GAA GGT YGT AGT AGY GGT AGT
AG and reverse R640 50-ACA AAC AAC RCT CRA CTC CTA
ACC T, amplification conditions were: 5 min initial denaturation
at 95?C followed by 30 cycles of 30 sec denaturation at 95?C,
45 sec annealing at 53?C, 1 min synthesis at 72?C. Second PCR step
was performed with primers forward F639 50-ATT ATA AGY GGG
TTT TGG GTT AGG and reverse R639 50-CAA ACC ACC ACT
AAT ACC CAC TA for the DNA samples from cell lines and R932
50-AAC CTA CTA ATC TAC TAA CRA ACA AAC for the DNA
samples from tissues, using 1:100 diluted product of the first step
amplification as a template. Amplification conditions were: 5 min ini-
tial denaturation at 95?C followed by 35 cycles of 30 sec denaturation
at 95?C, 45 sec annealing at 50?C, 1 min synthesis at 72?C. Obtained
PCR products were purified using gel filtration columns and cloned
into pCRII-TOPO vector (Invitrogen). Ten independent clones were
sequenced for every PCR product.
For inhibition of DNA methylation, the culture medium was
supplemented with 5-aza-20-deoxycytidine (5-aza-dC) (Sigma) at
UTIKAL ET AL.
a concentration of 1 lM. Control cells were cultured in parallel.
Cells were used for extraction of total RNA or genomic DNA as
MIM suppresses cell growth in anchorage independent manner
To test whether MIM expression may influence the growth of
transformed or tumor cells we analyzed the growth characteristics
of cells over-expressing MIM. Stable transfected PC-3 and NIH-
3T3 were obtained after transfecting them with pLP-IRESneo
based plasmids expressing human or mouse MIM respectively. In-
ternal ribosome entry site (IRES) enabled stable high level of
MIM expression during the long-term cultivation in medium sup-
plemented with G418. Cell growth was analyzed by direct count-
ing of cells after 7 days of subcultivation of stable transfected cells
seeded at the same density. MIM transfected cells exhibited a 2–
3 fold reduced growth in comparison to mock transfected ones in
3 independent experiments indicating that MIM over-expression
inhibits the cell growth (Fig. 1a).
To test whether the growth inhibition by MIM is anchorage de-
pendent, we analyzed the ability of MIM transfected cells to grow
on agar-coated culture plates. Stable transfected PC-3 and NIH-
3T3 cells were seeded on agarose-coated culture plates and culti-
vated in medium supplemented with G418 for further 7 days. The
cell growth was assessed by direct cell counting. We found that in-
dependent on anchorage MIM suppressed the cell growth to the
extent similar to normal growth conditions (Fig. 1a).
To further investigate the effect of MIM over-expression on cell
growth, we performed a colony-forming assay. PC-3 or NIH-3T3
cells were transfected with MIM expressing constructs and seeded
onto the new culture plates after 72 hr of transfection. Further culti-
vation was performed in the medium supplemented with G418. Af-
ter 21 days (NIH-3T3) or 28 days (PC-3), the number of colonies
formed by MIM transfected and mock transfected cell was deter-
mined. The number of colonies formed by MIM transfected cells
was significantly lower than that by mock transfected cells (Fig. 1b).
As well MIM transfected colonies were smaller than mock trans-
fected (not shown). Interestingly, the effect of MIM expression
was stronger in PC-3 than that in NIH-3T3 (Fig. 1b) what can be
explained by low level of endogenous MIM expression in PC-3.
Quantification of MIM expression in cell lines
For further experiments, the cell lines 5636, PC-3, LNCaP, RT-
4 and TccSuP with described differences in MIM expression14
were chosen. We first quantified the MIM expression in these cell
lines using real time RT-PCR. We demonstrated that––although
MIM transcripts were detected in all cell lines analyzed––the
amounts differ significantly. Cell lines 5637 and RT-4 express 22-
and 12-fold amounts of MIM in comparison to LNCaP respec-
tively (Fig. 2). MIM expression in PC-3 was comparable with its
expression in LNCaP. Surprisingly TccSuP, reported not to ex-
press MIM14showed a 4-fold higher number of MIM transcripts in
comparison to LNCaP (Fig. 2). This data allowed us to design a
model system to analyze the mechanisms regulating MIM expression.
Cloning of the promoter region of MIM and analysis of its activity
The data obtained by us as well as the literature data strongly
suggest the tumor suppressor role of MIM. Therefore, the analysis
of MIM promoter region is essential for the understanding of
mechanisms underlying suppression of its expression during me-
tastasis development. To recover the promoter region of MIM, we
screened the chromosome 8 genomic library with the probe com-
prising the first exon of MIM. Further analysis resulted in cloning
of a fragment of about 8 kb, containing the sequence of the first
exon of MIM. BLAST analysis of the cloned fragment sequence
revealed high degree of similarity with the genomic clone with
Acc. Number AC090922. Cloned fragment contained 3758 bp of
50flanking region of MIM, first exon and 3622 bp of the first intron
of MIM. Computer analysis of the sequence using CGplot soft-
ware (http://www.ebi.ac.uk/emboss/cpgplot/) revealed a CpG
island of 559 bp containing 76 CpG sites in immediate proximity
of MIM transcribed sequence. The nucleotide positions are related
FIGURE 1 – MIM suppresses cell growth. (a) Growth of cells stably
transfected with MIM (grey bars) or EGFP (black bars) expressing
constructs. Growth of pLP-IRESneo-EGFP transfected cells was taken
as 100%. Adherence conditions are: plastic––for normal adherence,
agarose––for inhibited adhesion. Experiments were repeated 3 times.
(b) Colony forming assay with cells transfected with empty vector
(opened bars), EGFP (black bars) or MIM (grey bars) expressing con-
structs. All experiments were repeated 3 times.
FIGURE 2 – Real-time RT-PCR analysis of MIM expression in cell
lines. MIM expression in LNCaP was taken as 1. The experiment was
repeated 4 times.
MIM/MTSS1 EXPRESSION REGULATION
to 50end of annotated MIM mRNA sequence with accession num-
To study the promoter activity, the complete 3758 bp as well as
5 deletion fragments were cloned into a luciferase reporter vector
pGL3basic (Fig. 3a). Obtained constructs were transiently trans-
fected into 5637 cells that showed the highest endogenous MIM
expression (Fig. 2). The promoter activity was then assessed on
the basis of normalized luciferase activity. Three promoter con-
structs comprising 3758, 1660 and 448 bp of MIM upstream
region demonstrated similar transcriptional activity (Fig. 3a), indi-
cating the location of the basic promoter elements within the 448
bp fragment. Deletion of a region from –398 to –67 bp strongly
reduced the transcriptional activity of 1660 bp fragment and com-
pletely abolished the activity of 448 bp fragment. Accordingly, the
67 bp fragment alone did not reveal any transcriptional activity.
As expected, the 3758 bp fragment cloned into pGL3basic in op-
posite direction showed only background activity (Fig. 3a).
To further analyze the structure of MIM promoter, 4 constructs,
containing truncated 448 bp fragment were created (Fig. 3b). Re-
porter analysis using luciferase reporter systems showed that the
fragment comprising nucleotides –276 to 113 has the highest ac-
tivity in 5637 cells. Deletion of further 90 nucleotides on the 50
end led to the reduction of the promoter activity by 50%. The low-
est promoter activity in the test was observed for the fragment
containing 147 bp of the promoter region.
In silico analysis of a 276 bp fragment of MIM promoter using
MatInspector program (Genomatrix software at http://www.
genomatix.de) revealed the presence of several potential transcrip-
tion factor binding sites: ETS family member ELI at –256, –240;
HIC-1 at –186, –147; MAX at –180, –166; E2F at – 166, –150 and
226, –10; N-Myc at –145, –131; HIF-1 at –68, –54 and SP1 at –33,
MIM promoter activity in cell lines with different endogenous
To test whether MIM expression is regulated only on the basis
of DNA sequence we compared MIM promoter activity in cell
lines with different levels of endogenous MIM expression using
luciferase reporter assay. Cell lines 5637 and PC-3 were transi-
ently transfected with the reporter constructs pGLpr3758-13,
pGLpr1660-13, pGLpr448-13 and pGLpr3758-13rev (Fig. 4).
Luciferase analysis revealed no significant differences in tran-
scriptional activity of any fragment analyzed. Similarly, inverted
promoter region was inactive in both cell lines (Fig. 4). We con-
FIGURE 3 – Identification of the MIM promoter. Analysis of MIM promoter activity fragments in 5637 cells using luciferase reporter assay.
(a) Activity of pr3758-13 promoter fragment was taken as 1. (b) Analysis of the deletions of pr448-13 promoter fragment. Activity of pr449-13
was taken as 1. All experiments were repeated at least 3 times. The nucleotide positions are related to the 50end of annotated MIM mRNA
sequence with accession number NM_014751.
UTIKAL ET AL.
cluded that cell lines 5637 and PC-3 express all transcription fac-
tors necessary to activate MIM promoter. Since MIM promoter is
entirely located within a CpG island, the role of DNA methyla-
tion in the regulation of its activity was proposed.
MIM promoter methylation in cell lines and clinical samples
To test whether the differences in endogenous MIM expression
are associated with the differences in MIM promoter methylation,
we performed bisulphite sequencing of the –420 to 13 promoter
region in cell lines 5637, RT4, PC-3 and LNCaP. Genomic DNA
isolated from cells was treated with bisulphite and amplified using
nested PCR with primers specific for bisulphite treated MIM pro-
moter sequence. Obtained PCR products were cloned and 10 ran-
domly selected clones were sequenced. We found overall DNA
methylation in MIM expressing cell 5637 and RT4 to be 1.01 and
3.19%, respectively. In contrast, both cell lines demonstrating low
MIM expression, showed a significant increase in overall DNA
methylation within analyzed region. In PC-3, overall methylation
was 34.93% and in LNCaP––18.84%. Interestingly methylated
cytosines were not evenly distributed throughout the whole CpG
island (Fig. 5a). The concentration of methylated cytosines was
higher in 50part of the region. 30part remained free of methylation
with the exception of a –47 site methylated in LNCaP (Fig. 5a). As
well LNCaP demonstrated a highly homogenous methylation pat-
tern in contrast to that of PC-3. We concluded that DNA methyla-
tion of 50part of MIM promoter region is sufficient for inactivation
of MIM expression in PC-3 and LNCaP cell lines.
To test whether DNA methylation plays a role in inactivation of
MIM in vivo, we analyzed MIM expression and MIM promoter
methylation in 5 samples of prostate carcinoma and respective
normal tissues. The analysis of MIM expression by real-time RT-
PCR revealed a 90% reduction of MIM expression in the tumor
sample PC2 as compared to matched normal tissue sample. On the
other hand, no significant change of MIM expression was observed
in other samples (Fig. 5b). DNA methylation analysis was per-
formed for the methylation sites from –420 to –240––the region
showing increased DNA methylation in the cells lines. It was
found that in the tumor sample PC2 the overall DNA methylation
within the analyzed region was about 24%, while in the corre-
sponding normal tissue the level of methylation was about 5%. No
change in the level of DNA methylation within MIM promoter
was observed in other samples (Fig. 6). This data indicate that
DNA methylation may be responsible for the reduction of MIM
expression in vivo.
Treatment with 5-Aza-20-deoxycytidine activates MIM expression
Next we tested whether the expression of MIM may be influenced
by inhibiting DNA methylation by 5-Aza-20-deoxycytidine (5-Aza-
dC). To test whether suppression of DNA methylation may up-regu-
late the endogenous MIM expression, we treated the cell line with
homogenous methylation pattern––LNCaP by the methylation in-
hibitor 5-Aza-dC. Cells were cultivated in medium supplemented
with 5-Aza-dC for 24, 48 and 72 hr. MIM expression was quantified
using real-time RT-PCR. We found that inhibition of methylation
efficiently upregulated the expression of MIM mRNA already after
24 h treatment (Fig. 6). Further incubation with Aza led to an addi-
tional increase of the number of MIM transcripts (Fig. 6). Analysis
of DNA methylation in the MIM promoter region after Aza treat-
ment revealed the appearance of nonmethylated alleles (data not
shown). This data supported the finding that MIM expression is
DNA methylation dependent.
MIM was shown to be important for the tumor development and
several mechanisms of its action were proposed. In this study, we
provided an additional evidence for the tumor suppression func-
tion of MIM. Growth suppression observed by us may be a result
of secondary effects induced by the effect of overexpressed MIM
on Shh signaling.19Observed inhibition of cell growth is in agree-
ment with the data of Loberg et al. who showed the inhibition of
cell proliferation in PC-3 cells stably transfected with MIM.21
Although we did not observe adherence dependence, this does not
contradict with the data on cytoskeleton remodeling function of
MIM15,16,18since adherence is not the only parameter that may be
regulated via cytoskeleton changes. Similarly to our data the over-
expression of MIM was shown not to influence cell adhesion to
different matrixes or cell motility.21The mechanism of action in-
dependent on cell adhesion properties was also suggested by Mat-
tila et al.15who observed MIM expression in highly polarized cell
types such as neurons and myoblasts. It was proposed that the loss
of MIM expression during tumor development leads to the loss of
polarity and the gain of growth advantage.15
While the role of MIM for tumor development is well supported
by diverse experimental evidences, little is known about the regu-
lation of MIM expression and the mechanisms underlying its loss
during tumor development. The presence of a CpG island within
the 50-flanking region of MIM was already considered as an indi-
cation for methylation dependent regulation of MIM expression.22
The attempt was undertaken to activate MIM expression by inhibi-
ting DNA methylation by 5-Aza-dC. In the model system used this
treatment had no effect on MIM expression, so the conclusion was
drawn that DNA methylation is not responsible for MIM inactiva-
tion. The authors did not analyze the methylation status of MIM
promoter in tissues or cell lines as well the promoter activity was
not analyzed in reporter assays.22In the other study, however, it
was shown that 5-Aza-dC treatment of gastric cancer cell line
AGS caused more than 4-fold upregulation of MIM expression.23
In support of the hypothesis of methylation dependent MIM ex-
pression, we demonstrated here that MIM promoter is entirely
located within a CpG island. Moreover, the level of MIM expres-
sion correlates with the methylation state of its promoter region in
cell lines as well as in primary tissue samples providing a strong
evidence for DNA methylation dependence of MIM expression. In
accord with our data, methylation of MIM promoter region was
also shown for 3 out of 10 samples of primary gastric carcino-
mas.23Intriguing the DNA methylation was not evenly distributed
along the CpG island, but was concentrated within a 50part of the
CpG island while about 300 bp in the immediate proximity of
transcribed sequence stayed methylation free. Typically CpG
island is entirely methylated as it was shown for example for p15
in acute myeloid leukemia24or RASSF1A in bladder carcinoma.25
The profile of DNA methylation within the CpG island associated
with MIM promoter overlaid with the promoter structure indicate
that the part of the CpG island found to be hypermethylated in the
FIGURE 4 – MIM promoter activity in PC-3 and 5637 cell lines.
Activity of pr3758-13 in 5637 cells was taken as 1.
MIM/MTSS1 EXPRESSION REGULATION
FIGURE 5 – MIM promoter methylation in cell lines (a) and tissue samples (b). (a) X-axis indicates the position of methylation site. Z-axis
indicates the number of clones showing methylated cytosine in corresponding position. Red bars represent LNCaP, yellow bars––PC-3, green
bars––RT-4, blue bars––5637. (b) Analysis of MIM expression and DNA methylation of the sites 2420 to 2240 in tissue samples. Ten inde-
pendent clones were sequenced for each sample.
UTIKAL ET AL.
cell lines with reduced MIM expression does not contain any criti-
cal promoter elements. This finding suggests that in these cell
lines MIM expression inactivation depends on chromatin conden-
sation or methylation specific transcription inhibitors rather than
on the hindrance of transcription factor binding.
Since the genes inactivated by DNA methylation within the pro-
moter region usually can be reactivated by treatment of the cells
with methylation inhibitors,13,26we demonstrated that 5-Aza-dC
treatment leads to up-regulation of MIM expression. These data
strongly support the paradigm of methylation dependent regula-
tion of MIM expression.
The theory that MIM is a tumor suppressor gene is now strongly
supported by its functional analysis as well as by our data on the
mechanisms regulating MIM expression. To complete the para-
digm of tumor suppressor function of MIM, its level of expression
should be analyzed not only in relation to the methylation status of
the promoter region, but also in relation to loss of heterozygocity
(LOH) and inactivating mutations.27Further analysis of MIM
expression or inactivation in tissue samples and its association
with different human malignancies will define a novel candidate
to be used as a marker of primary tumors or metastasis.
In summary, we found that MIM is capable to suppress cell
growth independently on the adherence state of the cell. We estab-
lished that entire MIM promoter is located within a CpG island. We
clearly demonstrated that MIM expression is inactivated by DNA
methylation of the CpG island and may be reactivated by DNA
We thank Ms. B€ arbel Schleider, Ms. Christel Herbst and
Ms. Christina Schmuttermaier for excellent technical assistance.
1.Robertson KD, Jones PA. DNA methylation: past, present and future
directions. Carcinogenesis 2000;21:461–7.
Chen W, Zouboulis CC, Fritsch M, Blume-Peytavi U, Kodelja V,
Goerdt S, Luu-The V, Orfanos CE. Evidence of heterogeneity and
quantitative differences of the type 1 5a-reductase expression in cul-
tured human skin cells—evidence of its presence in melanocytes.
J Invest Dermatol 1998;110:84–9.
Issa JP. CpG island methylator phenotype in cancer. Nat Rev Cancer
Momparler RL, Bovenzi V. DNA methylation and cancer. J Cell
Szyf M, Pakneshan P, Rabbani SA. DNA methylation and breast can-
cer. Biochem Pharmacol 2004;68:1187–97.
Stirzaker C, Millar DS, Paul CL, Warnecke PM, Harrison J, Vincent PC,
Frommer M, Clark SJ. Extensive DNA methylation spanning the Rb
promoter in retinoblastoma tumors. Cancer Res 1997;57:2229–37.
Ruas M, Peters G. The p16INK4a/CDKN2A tumor suppressor and its
relatives. Biochim Biophys Acta 1998;1378:F115–F177.
Herman JG, Jen J, Merlo A, Baylin SB. Hypermethylation-associated
inactivation indicates a tumor suppressor role for p15INK4B. Cancer
Herman JG, Umar A, Polyak K, Graff JR, Ahuja N, Issa JP,
Markowitz S, Willson JK, Hamilton SR, Kinzler KW, Kane MF,
Kolodner RD, et al. Incidence and functional consequences of hMLH1
promoter hypermethylation in colorectal carcinoma. Proc Natl Acad Sci
10. Hiltunen MO, Alhonen L, Koistinaho J, Myohanen S, Paakkonen M,
Marin S, Kosma VM, Janne J. Hypermethylation of the APC (adeno-
matous polyposis coli) gene promoter region in human colorectal car-
cinoma. Int J Cancer 1997;70:644–8.
11. Dobrovic A, Simpfendorfer D. Methylation of the BRCA1 gene in
sporadic breast cancer. Cancer Res 1997;57:3347–50.
12. Gratchev A, Bohm C, Riede E, Foss HD, Hummel M, Mann B, Backert S,
Buhr HJ, Stein H, Riecken EO, Hanski C. Regulation of mucin MUC2
gene expression during colon carcinogenesis. Ann N Y Acad Sci
13. Gratchev A, Siedow A, Bumke-Vogt C, Hummel M, Foss HD, Hanski ML,
Kobalz U, Mann B, Lammert H, Mansmann U, Stein H, Riecken EO
et al. Regulation of the intestinal mucin MUC2 gene expression in vivo:
evidence for the role of promoter methylation. Cancer Lett 2001;168:
14. Lee YG, Macoska JA, Korenchuk S, Pienta KJ. MIM, a potential
metastasis suppressor gene in bladder cancer. Neoplasia 2002;4:
15. Mattila PK, Salminen M, Yamashiro T, Lappalainen P. Mouse MIM, a
tissue-specific regulator of cytoskeletal dynamics, interacts with ATP-
actin monomers through its C-terminal WH2 domain. J Biol Chem 2003;
16. Yamagishi A, Masuda M, Ohki T, Onishi H, Mochizuki N. A novel
actin bundling/filopodium-forming domain conserved in insulin re-
ceptor tyrosine kinase substrate p53 and missing in metastasis protein.
J Biol Chem 2004;279:14929–36.
17. Millard TH, Bompard G, Heung MY, Dafforn TR, Scott DJ, Machesky
LM, Futterer K. Structural basis of filopodia formation induced by the
IRSp53/MIM homology domain of human IRSp53. EMBO J 2005;24:
18. Lin J, Liu J, Wang Y, Zhu J, Zhou K, Smith N, Zhan X. Differential
regulation of cortactin and N-WASP-mediated actin polymerization
by missing in metastasis (MIM) protein. Oncogene 2005;24:2059–66.
19. Callahan CA, Ofstad T, Horng L, Wang JK, Zhen HH, Coulombe PA,
Oro AE. MIM/BEG4, a Sonic hedgehog-responsive gene that potenti-
ates Gli-dependent transcription. Genes Dev 2004;18:2724–9.
20. Olek A, Oswald J, Walter J. A modified and improved method for bisul-
phite based cytosine methylation analysis. Nucleic Acids Res 1996;24:
21. Loberg RD, Neeley CK, Adam-Day LL, Fridman Y, St John LN, Nix-
dorf S, Jackson P, Kalikin LM, Pienta KJ. Differential expression
analysis of MIM (MTSS1) splice variants and a functional role of
MIM in prostate cancer cell biology. Int J Oncol 2005;26:1699–705.
22. Nixdorf S, Grimm MO, Loberg R, Marreiros A, Russell PJ, Pienta KJ,
Jackson P. Expression and regulation of MIM (missing in metastasis),
a novel putative metastasis suppressor gene, and MIM-B, in bladder
cancer cell lines. Cancer Lett 2004;215:209–20.
23. Yamashita S, Tsujino Y, Moriguchi K, Tatematsu M, Ushijima T.
Chemical genomic screening for methylation-silenced genes in gastric
cancer cell lines using 5-aza-20-deoxycytidine treatment and oligonu-
cleotide microarray. Cancer Sci 2006;97:64–71.
24. Dodge JE, List AF, Futscher BW. Selective variegated methylation of the
p15 CpG island in acute myeloid leukemia. Int J Cancer 1998;78:561–7.
25. Lee MG, Kim HY, Byun DS, Lee SJ, Lee CH, Kim JI, Chang SG, Chi SG.
Frequent epigenetic inactivation of RASSF1Ain human bladder carcinoma.
Cancer Res 2001;61:6688–92.
26. Gonzalgo ML, Hayashida T, Bender CM, Pao MM, Tsai YC, Gon-
zales FA, Nguyen HD, Nguyen TT, Jones PA. The role of DNA meth-
ylation in expression of the p19/p16 locus in human bladder cancer
cell lines. Cancer Res 1998;58:1245–52.
27. Jones PA, Laird PW. Cancer epigenetics comes of age. Nat Genet 1999;
FIGURE 6 – Inhibition of methylation enhances endogenous MIM
expression. LNCaP cells were treated with 5-Aza-dC for time indi-
cated. MIM expression was then analyzed by real-time RT-PCR.
MIM/MTSS1 EXPRESSION REGULATION