Human Molecular Genetics, 2003, Vol. 12, No.

5 575582
DOI: 10.1093/hmg/ddg048

Identication of CDH1 germline missense

mutations associated with functional inactivation
of the E-cadherin protein in young gastric cancer
probands
Gianpaolo Suriano1,{, Carla Oliveira1,2,{, Paulo Ferreira1, Jose C. Machado1,3,
Maria C. Bordin2, Olivier De Wever4, Erik A. Bruyneel4, Nicole Moguilevsky5, Nicola Grehan2,
Timothy R. Porter6, Frances M. Richards6, Ralph H. Hruban7, Franco Roviello8,
David Huntsman9, Marc Mareel4, Fatima Carneiro1,3, Carlos Caldas2,* and Raquel Seruca1,3,*
1
Instituto de Patologia e Imunologia Molecular da Universidade do Porto (IPATIMUP), 4200 Porto, Portugal,
2
Cancer Genomics Program, Department of Oncology, University of Cambridge, Hutchison/MRC

Downloaded from http://hmg.oxfordjournals.org/ by guest on July 11, 2016

Research Centre, Addenbrookes Hospital, Cambridge CB2 2XZ, UK, 3Faculdade de Medicina, Hospital de S. Joao,
4200 Porto, Portugal, 4Laboratory of Experimental Cancerology, UZG, B-9000 Ghent, Belgium, 5Service of
Applied Genetics, Institute of Biology and Molecular Medicine, B-6041 Gosselies, Belgium, 6Section of Medical and
Molecular Genetics, Division of Reproductive and Child Health, University of Birmingham, The Medical School,
Birmingham B15 2TT, UK, 7Department of Pathology, Johns Hopkins Medical Institutions, Baltimore, MD, USA,
8
Istituto Policattedra di Scienze Chirurgiche, Universita degli Studi di Siena, 53100 Siena, Italy and
9
Department of Pathology and Laboratory Medicine, University of British Columbia Vancouver, Canada

Received November 19, 2002; Revised and Accepted December 20, 2002

E-cadherin is involved in the formation of cell-junctions and the maintenance of epithelial integrity. Direct
evidence of E-cadherin mutations triggering tumorigenesis has come from the nding of inactivating germline
mutations of the gene (CDH1) in hereditary diffuse gastric cancer (HDGC). We screened a series of 66 young
gastric cancer probands for germline CDH1 mutations, and two novel missense alterations together with an
intronic variant were identied. We then analysed the functional signicance of the two exonic missense variants
found here as well as a third germline missense variant that we previously identied in a HGDC family. cDNAs
encoding either the wild-type protein or mutant forms of E-cadherin were stably transfected into CHO (Chinese
hamster ovary) E-cadherin-negative cells. Transfected cell-lines were characterized in terms of aggregation,
motility and invasion. We show that a proportion of apparently sporadic early-onset diffuse gastric carcinomas
are associated with germline alterations of the E-cadherin gene. We also demonstrate that a proportion of
missense variants are associated with signicant functional consequences, suggesting that our cell model can
be used as an adjunct in deciding on the potential pathogenic role of identied E-cadherin germline alterations.

INTRODUCTION 110 aa with conserved calcium-binding motifs (EC1-5), a

E-cadherin is a 120 kDa glycoprotein localized at adherens
junctions of epithelial cells, where it mediates homophilic
calcium-dependent cell-adhesion (1). Its modular structure
consists of five extracellular domains of roughly
transmembrane region and a cytoplasmatic domain, which
interacts with filaments of actin through catenins (2).
Disruption of the E-cadherin complex is expected to induce
loss of cell adhesion with a concomitant increased cell motility
(3,4).

*To whom correspondence should be addressed. Tel: 0044 1223 3319 89; Fax: 0044 1223 3317 53; Email: cc234@cam.ac.uk and Tel: 351
225570764; Fax: 351 225570799; Email: rseruca@ipatimup.pt
{
The authors wish it to be known that, in their opinion, the first two authors should be regarded as joint First Authors.

Human Molecular Genetics, Vol. 12, No. 5 # Oxford University Press 2003; all rights reserved
576 Human Molecular Genetics, 2003, Vol. 12, No. 5
Table 1. Details of patients, tumours and CDH1 putative deleterious sequence alterations in five cases of early-onset gastric cancer
Case Geographic origin Age/gender Histologic tumour type
293 USA 43/F Diffuse
294 USA 43/F Diffuse
J68 Portugal 30/M Diffuse
54 England 51/M Diffuse
CE137 Portugal 42/F Mixed
a
Nucleotide positions refer to an E-cadherin sequence deposited in the EMBL/GenBank Data Libraries, accession no. Z13009; the A of the ATG of the initiator
Met codon is denoted nt1.
b
Somatic mutation in endometrial and gastric cancer (16,17).
c
Mutation described in a colon cancer cell line (11).
Germline truncating mutations of the CDH1 gene (EMBL/
GenBank Data Libraries no. CDH1-Z13009) resulting in
E-cadherin inactivation have been identified in hereditary
diffuse gastric carcinoma (OMIM no. Gastric cancer-137215)
(5,6). To date, 30 HDGC families harbouring CDH1 germline
mutations have been described (7) and only five are associated
with missense mutations. The pathogenic role of these non-
truncating mutations has not yet been established.
The number of families with germline CDH1 mutations
described to date is limited and the proportion of gastric
cancers due to autosomal inheritance is unknown. One would
predict that the proportion of gastric cancer cases related to
genetic predisposition is higher in young individuals. To date, a
single CDH1 truncating germline mutation in a young person
with apparently sporadic diffuse gastric cancer has been
reported (5).
The current study aimed at screening a series of 66 young
gastric cancer probands for germline CDH1 mutations. We then
analysed the functional significance of three germline sequence
alterations, two identified in this study and one previously
reported by us in a hereditary gastric cancer kindred.
RESULTS
The histological characteristics of the carcinomas in the 66
early-onset gastric cancer patients studied were: 37/66 (56.1%)
diffuse/isolated cell type gastric carcinomas, 17/66 (25.8%)
atypical/mixed type carcinomas and 12/66 (18.2%) intestinal/
glandular type carcinomas. Patients with diffuse/isolated cell
type gastric cancer were significantly younger (41.6
5.8
years) than those with intestinal gastric carcinoma (45.2
3.8
years; P 0.05). The same holds true when the age of patients
with atypical/mixed carcinomas (39.9
7.0 years) was com-
pared with that of patients with intestinal/glandular carcinomas
(P 0.03). No significant difference was observed between the
age of patients with diffuse/isolated cell and atypical/mixed
carcinomas. Patients with diffuse/isolated cell type gastric
carcinomas younger than 40 years showed a nearly equal
male:female ratio (1.3:1), whereas patients older than 40
showed a male preponderance (2.3:1). Atypical/mixed carci-
noma patients showed a male:female ratio of 2:1 in younger
than 40 and 2.7:1 in older than 40 years. Patients with
intestinal/glandular type tumours had a male:female ratio of
2:1. These results confirm previous reports that intestinal
carcinomas are more frequent in older patients, while diffuse/
Ex/In Nucleotide changea Protein change Comments
Ex 12 1849 G!A Ala617Thr (mis) Previously describedb
Ex 12 1849 G!A Ala617Thr (mis) Previously describedb
Ex 12 1901 C!T Ala634Val (mis) Previously describedc
Int 4 532-18 C!T New
Int 4 532-18 C!T New
isolated cell type cancers predominate in young patients.
Furthermore, the diffuse type of gastric carcinoma in young
patients demonstrates a nearly equal sex ratio, compared with a
male preponderance in the intestinal form.
The mutation screening identified five cases (7.6%) with poten-
Downloaded from http://hmg.oxfordjournals.org/ by guest on July 11, 2016
tially deleterious germline sequence variants. Four cases were
diffuse/isolated cell type carcinomas and one was an atypical/
mixed-type gastric carcinoma. Details of patients, tumours and
CDH1 putative deleterious sequence alterations in the five cases
are presented in Table 1. No germline CDH1 sequence alterations
were found in intestinal gastric cancer patients.
Three of the five patients with germline sequence alterations
in CDH1 harboured missense variants in exon 12. Two African-
American female patients, 293 and 294, had the identical
transition mutation in codon 617 (1849 G to A), leading to an
amino acid substitution (Ala to Thr) [Fig. 1A, lane 2, B(2)].
Laboratory contamination was excluded by repeating the
mutation analysis with new samples obtained separately from
both patients, and a familial relationship between them was
excluded by microsatellite analysis (data not shown). We
screened exon 12 in 100 Portuguese blood donors and none of
these controls showed the alteration. We also screened 93
African (sub-Saharans) DNA samples for this alteration and
two showed the same 1849 G to A change (allele frequency
1%). E-cadherin immunohistochemistry in the two cancer cases
revealed membranous staining in neoplastic cells forming
microglandular/trabecular structures, and heterogeneous mem-
branous staining in signet ring cells.
The second missense alteration (Table 1) was identified in
individual J68, with a signet ring cell carcinoma of the stomach
at the age of 30. The alteration, a transition at codon 634 (1901 C
to T), leads to an amino acid substitution (Ala to Val) [Fig. 1A,
lane 1, B(1)]. This missense variant was not present in 100
Portuguese normal controls. Expression of E-cadherin protein in
the cancer was not determined because of lack of tumour tissue.
The remaining germline sequence alteration (Table 1) was a
transition in intron 4, at position 532-18 (C to T transition)
found in two individuals. This alteration was not found in a
series of 100 Portuguese normal controls. The RTPCR as well
as the immunohistochemical analysis, in the cancer cells of one
of these individuals for whom tumour material was available
(CE 137), revealed lack of E-cadherin mRNA and protein
expression, respectively (data not shown). MS-PCR of the CpG
island 3 of the CDH1 promoter in this case showed the
presence of both methylated (M) and unmethylated (U) alleles
in the tumoural DNA extracted from frozen material and

Figure 1. Germline missense mutations in the CDH1 gene. (A) CDH1 exon
12 PCR/SSCP analysis, showing a band pattern in normal DNA (N) and
band shifts in DNA from patients with early-onset diffuse gastric cancer
(1, J68; and 2, 293). (B) Sequence chromatograms. (1) Missense mutation
in codon 634 (1901 C to T) in patient J68; (2) Missense mutation in
codon 617 (1849 G to A) in patient 293 (patient 294 had an identical
mutation).
analysis of tumour DNA did not reveal any somatic CDH1
mutation or LOH (data not shown).
We selected for analysis the two distinct germline missense
CDH1 variants here described (Ala634Val; Ala617Thr),
together with a germline missense CDH1 variant previously
described by us in a HDGC kindred (Thr340Ala) (8). CHO
(Chinese hamster ovary) cells do not express E-cadherin and
fail to aggregate homotypically, and were therefore chosen as a
model. For each mutation, two independent transfected clones
were analysed. Using western blotting, the expected 120 kDa
E-cadherin band was detected in the transfected clones and not
in the native CHO cells. Only clones expressing comparable
levels of E-cadherin were chosen for the functional character-
isation. For example of selected clones see Figure 2.
CHO cells failed to aggregate in both the fast and slow
aggregation assays. Cells transfected with the wild-type CDH1
construct displayed cell-to-cell aggregation in both assays
[Table 2; Fig. 3A(1) and B(1)]. The dependence of aggregation
on the presence of E-cadherin was demonstrated by its
inhibition with the MB2 antibody [Table 2; Fig. 3A(1) and
B(2)]. Cells expressing Ala634Val or Thr340Ala E-cadherin
mutants failed to aggregate in both assays [Table 2; Fig. 3A(2)
and B(3) and (4)]. Cells expressing the mutant Ala617Thr
displayed an intermediate phenotype. In the fast assay a 10-fold
decreased aggregation was observed in comparison to the wild-
type protein [Table 2; Fig. 3A(1)], and this reduced aggregation
was shown to be dependent on residual E-cadherin function,
since it was blocked by the MB2 antibody [Table 2; Fig. 3A(1)].
Cells expressing this mutant showed normal aggregation when
tested in the slow assay.
Cells expressing the wild-type E-cadherin protein were not
invasive when tested in a collagen invasion assay (Fig. 4).
These cells, when incubated with the E-cadherin blocking
antibody MB2, acquired invasive properties and were able to
penetrate the collagen matrix. Cells expressing the Ala634Val
and Thr340Ala E-cadherin mutants showed an invasive
phenotype in this assay (see Fig. 4). Cells expressing the
Ala617Thr mutation showed the same non-invasive behaviour
as cells expressing wild-type protein (see Fig. 4).
Human Molecular Genetics, 2003, Vol. 12, No. 5 577
Figure 2. Western blot analysis of E-cadherin transfected cell lines using the
specific antibody HECD1. Equal amounts of protein were loaded in each lane.
Table 2. Median values of particle diameters in fast
aggregation assay
Downloaded from http://hmg.oxfordjournals.org/ by guest on July 11, 2016
T0a T30a
CHO cells 16.52 19.25
E-Cad wild-type 17.95 1067.00
E-Cad wild-typeMB2 11.92
Thr340Ala 18.51 12.22
Ala634Val 18.06 15.13
Ala617Thr 15.30 103.4
Ala617ThrMB2 11.92
a
Particle diameter expressed in micrometres at time 0
(T0) and after 30 min (T30), respectively.
In a wound closure assay (Fig. 5), cells expressing wild-type
E-cadherin showed a conserved epithelial-like architecture and
a compact front of migration, with cells moving unidirectionally.
Cells expressing the Ala634Val and Thr340Val mutants appe-
ared spindle-shaped, with a non-uniform front of migration and
loose cells. Cells expressing the Ala617Thr mutation behaved
similarly to cells expressing wild-type protein.
DISCUSSION
Our findings of sequence variants in the germline of young
patients with apparently sporadic gastric cancer contrast with
those published by Stone et al. (9) (5/66 versus 0/96 with
sequence variants, P < 0.05), but the two series differed
significantly. First, the mean age of the patients reported here
was lower (41.7
6.0 versus 62.0
10.3 years) and secondly it
included a higher proportion of diffuse and atypical/mixed
tumours (82 versus 27%). As was observed in HDGC (10),
sporadic tumours in carriers of CDH1 germline sequence
variants are diffuse or mixed (one case) and never of the
intestinal type.
The potentially deleterious germline sequence variants found
in the present study included two distinct missense coding
sequence alterations and an intronic transition. One of the
missense alterations identified was a transition at codon 634
(1901 C to T) leading to an amino acid substitution (Ala to
Val). This variant was not present in 100 normal controls and
was previously reported as a mutation in a colon carcinoma cell
line and the corresponding primary tumour (11). The second
578 Human Molecular Genetics, 2003, Vol. 12, No. 5

Downloaded from http://hmg.oxfordjournals.org/ by guest on July 11, 2016

Figure 3. (A) Fast aggregation assay. Panel 1aggregate formation after 30 min (T30). Cells transfected with wild-type E-cadherin (red solid line); Ala617Thr
(blue solid line); incubated with the E-cadherin-blocking antibody MB2 (red dashed line, wild type; blue dashed line, Ala617Thr). Panel 2aggregate formation
for mutant Thr340Ala (red solid line, T0; red dashed line, T30) and for mutant Ala634Val (blue solid line, T0; blue dashed line, T30). *Difference statistically
significant. (B) Slow aggregation assay. Typical examples are shown for wild-type (panel 1) and Thr340Ala (panel 2).

missense variant, present in two African-American individuals,
was a transition in codon 617 (1849 G to A) leading to an
amino acid substitution (Ala to Thr). It occurs in the fifth
extracellular repeat (EC5) of E-cadherin, affecting a conserved
sequence encoding one of the calcium binding motifs (12). As
known, cadherins are functional only in the presence of
calcium ions, which stabilize the protein active conformation.
Adhesion is promoted in a process of cis-dimerization between
neighbour E-cadherin molecules, for which a key role played
by repeats 1 and 2 (EC1 and EC2) was demonstrated (1315).
Although less information is available for the other repeats, a
comparable activity could be predicted on the basis of their
sequence homology, suggesting a putative pathogenic effect
for the mutation we found. Besides, this variant was also
previously reported as a somatic mutation in an endometrial
cancer in association with loss of heterozygosity (16), fulfilling
the classic two-hit model of tumour suppressor gene inactiva-
tion. E-cadherin immunohistochemistry in these two cases
revealed a pattern similar to what is commonly seen in sporadic
tumours with somatic missense mutations. The Ala617Thr
variant has an allele frequency of 1% in Africans, suggesting
it may be an uncommon polymorphism occurring in popula-
tions of African origin. Nevertheless, it would be a very
improbable occurrence (less than 1:10 000 chance) that both
cases of diffuse gastric carcinoma in young African-Americans
would also be carriers of the same rare polymorphism, unless
such polymorphism is associated with predisposition to
stomach cancer. We note with interest that a third case of the
identical germline sequence variant in an African-American
with gastric cancer has now been reported, further supporting a
pathogenic association (17). The remaining germline sequence
alteration was a transition in intron 4, at position 532-18 (C to
T transition), present in two individuals. The results obtained in
one of these individuals, CE 137, were consistent with
disruption of gene expression as a result of inactivation of
one allele by the germline alteration (first hit) and of the second
(wild-type) allele by somatic methylation of the CDH1
promoter in tumour cells (second hit), although this could not
be formally proven. One can only speculate why the germline
sequence variant would not be expressed, although its location
near the two splicing branch sites where the splicing lariat loop
could potentially form (A nucleotides at positions 13 and
27 upstream of the 30 splice acceptor site) could somehow
disrupt transcript stability.
The pathogenic significance of the two distinct missense
variants (Ala634Val and Ala617Thr) identified in these
apparently sporadic diffuse gastric cancer cases was assessed
by analysing the functional consequences in a cell model
system. We also tested a missense variant previously identified
by our group in a HDGC family (Thr340Ala) (8). Interestingly
this variant was shown by computer modelling to be located in
a newly characterized E-cadherin sequence motif, likely to be
involved in the stabilization of the active protein conformation.
Two of the E-cadherin missense variants (Thr340Ala and
Ala634Val) studied resulted in dramatic functional conse-
quences: cell adhesion was disrupted, and the cells acquired a
highly invasive potential associated with enhanced cell motility
and loss of epithelial structure. One of these, Ala634Val, is the

Figure 4. Collagen invasion assay. Bars represent the percentage of invasion for
each of the experimental conditions.
result of a nucleotide substitution that was shown in a colon
carcinoma cell line to result in activation of a cryptic splicing
site leading to a truncated protein (11). Our results showed that
the sequence variant also results in an amino acid alteration that
disrupts the function of the full length E-cadherin protein,
which would be expressed when abnormal splicing is not
complete. Loss of epithelioid organization in cell lines has been
shown to be characteristic of malignant transformation and has
been associated to the acquisition of invasiveness and loss of
differentiation (18,19). The aggressive phenotype observed
for Thr340Ala and Ala634Val is clearly mediated by the
E-cadherin mutations introduced, since untransfected CHO
cells fail to aggregate, but show low motility and reduced
ability to invade (data not shown) when compared with clones
transfected with the two mutants. Based on this functional data,
we propose that these two missense variants are pathogenic and
should be considered as mutations.
The third missense variant (Ala617Thr) did not result in
increased motility or invasiveness but was associated with a
significant reduction of cellular adhesion on the fast aggrega-
tion assay. These mild functional consequences, disruption of
adhesion without an effect on motility, have been previously
noted when single amino acid substitutions in conserved
extracellular domains of E-cadherin were tested (20), and
concur with data showing that adhesion and motility are
mediated by different E-cadherin-dependent mechanisms (21).
The occurrence of this missense substitution as a somatic
mutation associated with LOH and as a germline variant in
three African-American diffuse gastric cancer patients consti-
tutes strong circumstantial evidence for its association with
carcinogenesis. In the assays used here this mutation resulted
only in mild functional consequences and therefore we could
not conclusively determine its pathogenicity.
In conclusion, we report that a small proportion of apparently
sporadic early-onset diffuse gastric carcinomas are associated
with CDH1 germline variants. It has previously been reported that
some families with hereditary diffuse gastric cancer are also
associated with germline CDH1 missense variants (8,2224).
Here we were able to demonstrate that there is a proportion of
missense variants associated with significant functional conse-
quences, suggesting they are pathogenic mutations. We propose
that functional assays can be used as an adjunct in deciding on the
potential pathogenic role of germline variants with significant
potential to help clinical counselling of gene carriers.
Human Molecular Genetics, 2003, Vol. 12, No. 5 579
MATERIALS AND METHODS
Patients
Constitutional DNA was isolated from 66 gastric carcinoma
patients with no familial history of stomach cancer diagnosed
before the age of 51 (65 patients were 50 or younger). Cases
originated from the UK (n 26), Portugal (n 22), Italy
(n 11), USA (n 6) and Greece (n 1). Genomic DNA was
isolated, using standard methods, either from peripheral blood
or from fresh frozen normal gastric mucosa. The gastric
carcinomas were classified according to Laure n and Carneiro
(25,26).
PCRSSCP
All 16 exons including intronexon boundaries and the
promoter region of the CDH1 gene were amplified by PCR.
Downloaded from http://hmg.oxfordjournals.org/ by guest on July 11, 2016
Primer sequences and PCR conditions were based on those
reported previously (5,8,27). Genomic DNA (25100 ng) was
amplified by PCR using as cycling conditions: 30 s at 94 C,
30 s at the appropriate annealing temperature, and 45 s at 72 C
for 35 cycles. Reaction products were subsequently diluted 1:1
with denaturing buffer (formamide with 0.025% xylene cyanol
and 0.025% bromophenol blue) and heated up to 99 C for
10 min prior to loading onto 0.6 and 0.8 mutation detection
enhancement (MDE; Flowgen, Rockland, ME, USA) gels.
Three gel conditions were used: 0.6 MDE/10% glycerol, 8 C;
0.8 MDE, 8 C; 0.8 MDE, 20 C. Gels were run at constant
temperature for 1218 h and stained with silver nitrate.
Sequencing analysis
Abnormal bands as well as corresponding normal bands,
detected by SSCP, were recovered from the gels and submitted
to PCR re-amplification with the original primer sets.
Re-amplified products were purified and sequenced on both
strands on an ABI Prism 377 automated sequencer (Perkin-
Elmer) using the ABI Prism Dye Terminator Cycle Sequencing
Kit (Perkin-Elmer, Foster City, CA, USA) and the original
primers. The CDH1 amplicon including exon 1 was directly
sequenced as reported by Oliveira et al. (8). All sequence
alterations detected were confirmed in a second independent
PCR.
Polymorphism analysis
The germline mutations detected and polymorphisms were
tested in at least 100 Caucasian blood donors using the same
primer sets as described before. In addition to the group of
Caucasian controls, 93 DNA samples from black Africans
(sub-Saharans), without history of neoplasia, were screened for
an exon 12 missense alteration (1849 G to A).
Immunohistochemical analysis
Formalin-fixed paraffin embedded tissues were used for the
study of E-cadherin immunohistochemical (IHC) expression
using a monoclonal antibody HECD-1 (diluted 1:200; Zymed,
580 Human Molecular Genetics, 2003, Vol. 12, No. 5
Figure 5. Wound closure assay. (A) Results of the wound closure assay graphically represented as distance migrated between wound edges (mm in function of time
after wound). Typical representative wounds after 5 h (20) are shown for Ala617Thr (B), and Ala634Val (C).
San Francisco, CA, USA). A modification of the avidin
biotinperoxidase complex method was used.
Methylation assay
CDH1 promoter methylation analysis was performed using
methylation-specific PCR (MSPCR) and primers described by
Graff et al. (28) for CpG island 3 of the CDH1 promoter. This
assay entails initial modification of DNA by sodium bisulphite,
converting all unmethylated, but not methylated, cytosines to
uracil, and subsequent amplification with primers specific for
methylated versus unmethylated DNA.
Statistical analysis
The statistical analysis was performed using the Students t-test
and w2 test. Differences were taken to be significant at P < 0.05.
The KolmogrovSmirnov test was used for the statistical
analysis of the aggregation curves in fast aggregation assay.
Construction of the plasmids encoding wild-type and
mutant E-cadherins
According to the human E-cadherin nucleotide sequence
reported in literature (29), oligonucleotide primers were
designed and purchased from GIBCO-BRL (forward: 50 -AAA
GCT TAC CAT GGG CCC TTG GA-30 ; reverse: 50 -AAA CTC
GAG CTA GTC GTC CTC GC-30 ). A human colon cDNA
library (human colon 50 -stretch plus cDNA library, 1 ng/ml,
colon cells pooled from 50 Caucasian males) was purchased
from Clontech (Clontech Laboratories Paolo Alto, CA, USA)
and the full-length E-cadherin cDNA (2649 bp) amplified by
using the Hot Start Method (30). Amplification was performed
for 30 cycles using the following settings: denaturation at 95 C
Downloaded from http://hmg.oxfordjournals.org/ by guest on July 11, 2016
for 1 min, annealing at 58 C for 1 min and extension at 72 C for
1 min and 30 s per cycle. Sequencing was performed as
aforementioned. The cDNA was inserted into the mammalian
expression vector pcDNA3 (Invitrogen) as a HindIII
XhoI cassette, leading to the plasmid pECAD1. Mutant
plasmids were obtained by nested PCR, by using specific
primers carrying the desired mutations (Thr340Ala: Ecad1012F,
50 -TTC CCT GCG TAT ACC CTG G-30 and Ecad1031R,
50 -ACC AGG GTA TAC GCA GGG AA-30 ; Ala634Val:
Ecad1893F, 50 -ACA CGG GGT GAG TGC CAA C-30 and
Ecad1911R, 50 -GTT GGC ACT CAC CCC GTG T-30 ; Ala
617Thr: Ecad1841F, 50 -TCA TTG ATA CAG ACC TTC CTC-30
and Ecad1859R, 50 -GGA AGG TCT GTA TCA ATG ATG-30 )
and pECAD1 as DNA template.
Cell culture and cDNA transfection
CHO (Chinese hamster ovary) DG44 dhfr cells were chosen
for stable transfection. Cells were tested to exclude presence of
E-cadherin expression before transfection, and cultured at 37 C
under 5% CO2 in humidified air, in a-MEM () medium
(GIBCO-BRL) supplemented with 5% fetal bovine serum,
2 mM L-glutamine, 1% penicillin/streptomycin. For the selec-
tion of clones 1000 mg/ml geneticin was added to the culture
medium.
Cells (5  106) were stably transfected with the different
cDNA constructs and with the plasmid alone (25 mg each) by
electropermeabilization, and neomycin selection was carried on
for the following 3 weeks. Single cell clones were selected and
analysed for E-cadherin expression by western blotting.

SDSPAGE and western immunoblotting
A 5 104 cell sample was lysed with 50 ml TRITON 114 buffer
(EDTA 2 mM, PBS Mg2, Ca2-free 1 , Triton 114 1%, DTT
1 mM, Protease Inhibitor Cocktail Roche 1 tablet/50 ml buffer)
and the extracted protein quantified by following the Bradford
dye-binding procedure (31). Ten micrograms of protein were
then separated on a 7.5% SDSPAGE, followed by transfer to
a nitrocellulose membrane (C-bond, Millipore). Human
E-cadherin monoclonal antibody HECD1 (R&D System,
1/3500 dilution) was used.
Cell aggregation assays
Fast cell aggregation assays were done as reported (32). Briefly,
single cell-suspensions were prepared and incubated in an
isotonic buffer containing 1.25 mM Ca2. Particle diameters
were measured at the start (T0) and after 30 min (T30), using a
particle size counter LS200 (Coulter, Electronics). Results were
plotted against the percentage of volume distribution.
For the slow aggregation assay (32), cells were first
trypsinized and then transferred to an agar gel (0.66% w/v)
in a 96-well plate. Aggregates formation was evaluated after
24 h using an inverted microscope. For both assays, cells were
incubated with the E-cadherin-blocking antibody MB2 (1/20
dilution) to control for inhibition of aggregation.
Collagen invasion and wound closure assays
The invasion assay was performed as previously described (33).
Invasion was expressed as function of the cell ability to
penetrate a matrix of collagen (type I solution, Seromed,
Bichrom KG, Berlin, Germany) and evaluated using a
microscope interfaced with a computer-controlled step motor.
For the wound closure assay (34), cells were cultivated to
confluence in six-well plates. An artificial wound was then
created with a plastic pipette tip (1 mm diameter) and migration
assessed by measuring the distance between wound edges as a
function of time (0, 2, 5 and 9 h).
ACKNOWLEDGEMENTS
This study was funded by grants from: Fundac a o para a Cie ncia 19. Thiery, J.P. (2002) Epithelialmesenchymal transitions in tumour
e a Tecnologia, Portugal (projects: POCTI/35374/CBO/2000
and POCTI/CBO/40820/2001); FORTIS Verzekerngen and the
Fund for Scientific Research, Flanders (FWO), Brussels,
Belgium; and by Cancer Research UK.
REFERENCES
1. Shore, E.M. and Nelson, W.J. (1991) Biosynthesis of the cell adhesion
molecule uvomorulin (E-cadherin) in Madin-Darby canine kidney epithelial
cells. J. Biol. Chem., 266, 1967219680.
2. Leckband, D. and Sivasankar, S. (2000) Mechanism of homophilic cadherin
adhesion. Curr. Opin. Cell Biol., 12, 587592.
3. Christofori, G. and Semb, H. (1999) The role of the cell-adhesion
molecule E-cadherin as a tumour-suppressor gene. Trends Biochem. Sci.,
24, 7376.
4. Van Aken, E., De Wever, O., Correia da Rocha, A.S. and Mareel, M. (2001)
Defective E-cadherin/catenin complexes in human cancer. Virchows Arch.,
439, 725751.
Human Molecular Genetics, 2003, Vol. 12, No. 5 581
5. Guilford, P., Hopkins, J., Harraway, J., McLeod, M., McLeod, N.,
Harawira, P., Taite, H., Scoular, R., Miller, A. and Reeve, A.E. (1998)
E-cadherin germline mutations in familial gastric cancer.
Nature, 392, 402405.
6. Gayther, S.A., Gorringe, K.L., Ramus, S.J., Huntsman, D., Roviello, F.,
Grehan, N., Machado, J.C., Pinto, E., Seruca, R., Halling, K. et al. (1998)
Identification of germline E-cadherin mutations in gastric cancer families of
European origin. Cancer Res., 58, 40864089.
7. Oliveira, C., Seruca, R. and Caldas, C. (2003) Genetic screening for
hereditary diffuse gastric cancer. Expert Rev. Mol. Diag. (in press).
8. Oliveira, C., Bordim, M.C., Grehan, N., Huntsman, D., Suriano, G.,
Machado, J.C., Aaltonen, L., Jackson, C.E., Seruca, R. et al. (2002)
Screening E-cadherin in gastric cancer families reveals germline mutations
only in hereditary diffuse gastric cancer. Hum. Mutat., 19, 510517.
9. Stone, J., Bevan, S., Cunningham, D., Hill, A., Rahman, N., Peto, J.,
Marossy, A. and Houlston, R.S. (1999) Low frequency of germline
E-cadherin mutations in familial and nonfamilial gastric cancer.
Br. J. Cancer, 79, 19351937.
10. Caldas, C., Carneiro, F., Lynch, H.T., Yokota, J., Wiesner, G.L.,
Powell, S.M., Lewis, F.R., Huntsman, D.G., Pharoah, P.D., Jankowski, J.A.
et al. (1999) Familial gastric cancer: overview and guidelines for
management. J. Med. Genet., 36, 873880.
11. Vecsey-Semjen, B., Becker, K.F., Sinski, A., Blennow, E., Vietor, I.,
Downloaded from http://hmg.oxfordjournals.org/ by guest on July 11, 2016
Zatloukal, K., Beug, H., Wagner, E. and Huber, L.A. (2002) Novel colon
cancer cell lines leading to better understanding of the diversity of
respective primary cancers. Oncogene, 21, 46464662.
12. Shapiro, L., Fannon, A.M., Kwong, P.D., Thompson, A., Lehmann, M.S.,
Grubel, G., Legrand, J.F., Als-Nielsen, J., Colman, D.R. and
Hendrickson, W.A. (1995) Structural basis of cellcell adhesion by
cadherins. Nature, 374, 327337.
13. Noe, V., Willems, J., Vandekerckhove, J., Roy, F.V., Bruyneel, E. and
Mareel, M. (1999). Inhibition of adhesion and induction of epithelial cell
invasion by HAV-containing E-cadherin-specific peptides. J. Cell. Sci., 112,
127135.
14. Pertz, O., Bozic, D., Koch, A.W., Fauser, C., Brancaccio, A. and
Engel, J. (1999) A new crystal structure, Ca2 dependence and
mutational analysis reveal molecular detail of E-cadherin
homoassociation. EMBO, 18, 17381747.
15. Kitagawa, M., Natori, M., Murase, S., Hirano, S., Taketani, S. and
Suzuki, S.T. (2000) Mutation analysis of cadherin-4 reveals amino acid
residues of EC1 important for the structure and function. Biochem. Biophys.
Res. Commun., 271, 358363.
16. Risinger, J.I., Berchuck, A., Kohler, M.F. and Boyd, J. (1994) Mutations of
the E-cadherin gene in human gynecologic cancers. Nat. Genet., 7, 98102.
17. Ascano, J.J., Frierson, H., Moskaluk, C.A., Harper, J.C., Roviello, F.,
Jackson, C.E., El-Rifai, W., Vindigni, C., Tosi, P. and Powell, S.M. (2001)
Inactivation of the E-cadherin gene in sporadic diffuse-type gastric cancer.
Mod. Pathol., 14, 942949.
18. Behrens, J., Mareel, M.M., Van Roy, F.M. and Birchmeier, W. (1989)
Dissecting tumor cell invasion: epithelial cells acquire invasive properties
after the loss of uvomorulin-mediated cellcell adhesion. J. Cell Biol., 108,
24352447.
progression. Nature, 2, 442452.
20. Handschuh, G., Luber, B., Hutzler, P., Hofler, P., Hofler, H. and Becker, K.
(2001) Single amino acid substitutions in conserved extracellular domains
of E-cadherin differ in their functional consequences. J. Mol. Biol., 314,
445454.
21. Chen, H., Paradies, N.E., Fedor-Chaiken, M. and Brackenbury, R. (1997)
E-cadherin mediates adhesion and suppresses cell motility via distinct
mechanisms. J. Cell Sci., 110, 345356.
22. Shinmura, K., Kohno, T., Takahashi, M., Sasaki, A., Ochiai, A., Guilford, P.,
Hunter, A., Reeve, A.E., Sugimura, H., Yamaguchi, N. and Yokota, J.
(1999) Familial gastric cancer: clinicopathological characteristics, RER
phenotype and germline p53 and E-cadherin mutations. Carcinogenesis,
20, 11271131.
23. Yoon, K.A., Ku, J.L., Yang, H.K., Kim, W.H., Park, S.Y. and Park, J.G.
(1999) Germline mutations of E-cadherin gene in Korean familial gastric
cancer patients. J. Hum. Genet., 44, 177180.
24. Yabuta, T., Shinmura, K., Tani, M., Yamaguchi, S., Yoshimura, K.,
Katai, H., Nakajima, T., Mochiki, E., Tsujinaka, T., Takami, M. et al.
(2002) E-cadherin gene variants in gastric cancer families whose probands
are diagnosed with diffuse gastric cancer. J. Int. J. Cancer, 101, 434441.
582 Human Molecular Genetics, 2003, Vol. 12, No. 5
25. Carneiro, F. (1997) Classification of gastric carcinomas. Curr. Diag.
Pathol., 4, 5159.
26. Laure n, P. (1965) The two histological main types of gastric carcinoma:
diffuse and so-called intestinal-type carcinoma. An attempt at a
histo-clinical classification. Acta Pathol. Microbiol. Scand., 64, 3149.
27. Li, L.C., Chui, R.M., Sasaki, M., Nakajima, K., Perinchery, G., Au, H.C.,
Nojima, D., Carroll, P. and Dahiya, R. (2000) A single nucleotide
polymorphism in the E-cadherin gene promoter alters transcriptional
activities. Cancer Res., 60, 873876.
28. Graff, J.R., Herman, J.G., Myohanen, S., Baylin, S.B. and Vertino, P.M.
(1997) Mapping patterns of CpG island methylation in normal and
neoplastic cells implicates both upstream and downstream regions in
de novo methylation. J. Biol. Chem., 272, 2232222329.
29. Berx, G., Staes, K., van Hengel, J., Molemans, F., Bussemakers, M.J.,
van Bokhoven, A. and van Roy, F. (1995) Cloning and characterization of
the human invasion suppressor gene E-cadherin (CDH1). Genomics,
26, 281289.
30. Chou, Q., Russels, M., Birch, D.E., Raymond, J. and Bloch, W. (1992)
Prevention of pre-PCR mis-priming and primer dimerization improves
low-copy-number amplifications. Nucl. Acids Res., 20, 17171723.
31. Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye
binding. Anal. Biochem., 72, 248254.
32. Boteborg, T., Bracke, M.E., Bruyneel, E.A. and Mareel, M.M. (2000)
Cell aggregation assays. In Brooks S.A. and Shumacher U. (eds).
Methods in Molecular Medicine, Vol. 58: Metastasis Research Protocols,
Vol. 2: Cell Behaviour in vitro and in vivo. Humana Press, Totowa, NJ,
pp. 3345.
33. Brake, M.E., Boteborg, T., Bruyneel, E.A. and Mareel, M.M. (2000)
Collagen invasion assay. Brooks, S.A. and Shumacher, U. (eds). Methods
in Molecular Medicine, Vol. 58: Metastasis Research Protocols, Vol. 2:
Cell Behaviour in vitro and in vivo. Humana Press, Totowa NJ, USA,
pp 8189.
34. Andre , F., Rigot, V., Thimonier, J., Montixi, C., Parat, F., Pommier, G.,
Marvaldi, J. and Luis, J., (1999) Integrins and E-cadherin cooperate
with IGF-I to induce migration of epithelial colonic cell. Int. J. Cancer,
83, 497505.
Downloaded from http://hmg.oxfordjournals.org/ by guest on July 11, 2016