Identification and functional characterization of a novel
barnacle cement protein
Youhei Urushida
1
, Masahiro Nakano
1
, Satoru Matsuda
1
, Naoko Inoue
2
, Satoru Kanai
2
,
Naho Kitamura
3
, Takashi Nishino
3
and Kei Kamino
1
1 Marine Biotechnology Institute, Iwate, Japan
2 Pharma Design, Inc., Tokyo, Japan
3 Department of Chemical Science & Engineering, Faculty of Engineering, Kobe University, Japan
Living on a boundary brings various advantages for
organisms; such organisms therefore have developed a
variety of molecular systems to hold themselves on the
boundary during their evolution. Marine sessile organ-
isms possess underwater attachment capability as an
indispensable physiologic function, enabling them to
live on a liquid–solid boundary during most of their
life cycle. This underwater attachment is closely related
to other biological functions such as metamorphosis,

2007, accepted 29 June 2007)
doi:10.1111/j.1742-4658.2007.05965.x
Barnacle attachment to various foreign materials in water is guided by an
extracellular multiprotein complex. A 19 kDa cement protein was purified
from the Megabalanus rosa cement, and its cDNA was cloned and
sequenced. The gene was expressed only in the basal portion of the animal,
where the histologically identified cement gland is located. The sequence of
the protein showed no homology to other known proteins in the databases,
indicating that it is a novel protein. Agreement between the molecular mass
determined by MS and the molecular weight estimated from the cDNA
indicated that the protein bears no post-translational modifications. The
bacterial recombinant was prepared in soluble form under physiologic con-
ditions, and was demonstrated to have underwater irreversible adsorption
activity to a variety of surface materials, including positively charged, nega-
tively charged and hydrophobic ones. Thus, the function of the protein was
suggested to be coupling to foreign material surfaces during underwater
attachment. Homologous genes were isolated from Balanus albicostatus and
B. improvisus, and their amino acid compositions showed strong resem-
blance to that of M. rosa, with six amino acids, Ser, Thr, Ala, Gly, Val
and Lys, comprising 66–70% of the total, suggesting that such a biased
amino acid composition may be important for the function of this protein.
Abbreviations
ASW, artificial seawater; Balcp-19k, Balanus albicostatus 19 kDa cement protein; Bicp-19k, Balanus improvisus 19 kDa cement protein;
cp, cement protein; Dopa, 3,4-dihydroxyphenylalanine; GSF1 and GSF2, cement fractions separated by their solubility in a guanidine
hydrochloride solution; Mrcp, Megabalanus rosa cement protein; rMrcp-19k, recombinant 19 ka Megabalanus rosa cement protein in
Escherichia coli; RU, response unit; SPR, surface plasmon resonance.
4336 FEBS Journal 274 (2007) 4336–4346 ª 2007 The Authors Journal compilation ª 2007 FEBS
stiff and tough, and protection from microbial degra-
dation. This multifunctionality, together with the
insoluble ⁄ sticky and complex nature of the adhesive,

of the adhesive, no proteins contributing the necessary
surface functions such as priming, spreading and cou-
pling have been identified. Nor have any direct mea-
surement of these activities been reported for the
cement proteins prepared under physiologic conditions,
and such kinds of measurement have never been
achieved in any biotic underwater adhesive protein
studies.
The holdfast system of the barnacle shows no simi-
larity to that of the mussel, a relatively well-character-
ized one. There are no sequence similarities among the
protein components between the two systems. The
mussel holdfast system [1] depends on several protein
modifications, including 3,4-dihydroxyphenylalanine
(Dopa); however, no involvement of Dopa in the bar-
nacle cement was found [10,12]. Thus, the barnacle
system represents a novel biological adhesive system.
The present study identified a novel cement protein,
cp-19k, in the barnacle holdfast system, and demon-
strated its ability to be adsorbed to a foreign material
surface in seawater using a bacterial recombinant pro-
tein prepared under physiologic conditions. We also
show that the function of the protein is reliant upon
common amino acids, with no specific modifications.
Results
Purification and characterization of Mrcp-19k
Mrcp-19k was detected by SDS ⁄ PAGE in both guani-
dine hydrochloride-soluble fractions 1 (GSF1) and 2
(GSF2) of barnacle cement [9] with the same mobility
(Fig. 1). The molecular mass was estimated to be

was first amplified from M. rosa cDNA by PCR. The
deduced amino acid sequence of the 53 bp DNA com-
pletely matched the N-terminal amino acid sequence
of the mature Mrcp-19k. Subsequent 3¢-RACE and
5¢-RACE resulted in a 750 bp and a 102 bp DNA
fragment, respectively. An 852 bp cDNA fragment
encoding the Mrcp-19k protein was finally determined.
Ten randomly selected clones for the coding region of
Mrcp-19k had one nonsynonymous substitution and
several synonymous substitutions, presumably due to
errors introduced by the PCR amplifications (as each
substitution was found only in one randomly selected
clone but not in any other clones). Both B. albicostatus
(Bal)cp-19k (881 bp) and B. improvisus (Bi)cp-19k
(970 bp) cDNAs were also amplified by 3¢-RACE with
the oligonucleotide primers designed from the N-termi-
nal region of Mrcp-19k.
Structural outline of cp-19ks
The coding region of Mrcp-19k encoded 198 amino
acids (supplementary Fig. S1A). The mature N-termi-
nal sequence was found to start at residue number 26;
thus the first 25 amino acids function as the signal
peptide that has been cleaved off in the mature pro-
tein. The amino acid sequences of the N-terminal and
three internal peptide fragments of Mrcp-19k deter-
mined experimentally were found to be contained in
the deduced sequence and are in complete agreement
with those of the deduced sequence. The cDNA frag-
ments of 881 bp and 970 bp encoding 173 amino acids
each were also determined for Balcp-19k and

a
Mrcp19k ⁄ standard Balcp19k ⁄ standard Bicp19k ⁄ standard
Asp 10.00 14.8
c
5.00 5.00 1.92 0.96 0.96
Asn 8.00 – 6.00 7.00 1.86 1.40 1.63
Ser 18.00 19.30 15.00 17.00 2.61 2.17 2.46
Glu 9.00 11.2
d
8.00 3.00 1.45 1.29 0.48
Gln 4.00 – 3.00 8.00 0.98 0.73 1.95
Gly 27.00 29.60 22.00 25.00 3.65 2.97 3.38
His 1.00 1.10 2.00 2.00 0.43 0.87 0.87
Arg 1.00 2.30 1.00 0.00 0.20 0.20 0.00
Thr 21.00 21.20 25.00 20.00 3.56 4.24 3.39
Ala 18.00 18.20 18.00 21.00 2.34 2.34 2.73
Pro 4.00 4.90 4.00 5.00 0.78 0.78 0.98
Cys 2.00 ND 2.00 2.00 1.00 1.00 1.00
Tyr 0.00 0.60 0.00 0.00 0.00 0.00 0.00
Val 14.00 14.60 17.00 16.00 2.12 2.58 2.42
Met 0.00 0.00 1.00 0.00 0.00 0.42 0.00
Lys 17.00 14.60 24.00 17.00 2.88 4.07 2.88
Ile 5.00 5.20 3.00 9.00 0.94 0.57 1.70
Leu 10.00 9.40 13.00 14.00 1.10 1.43 1.54
Phe 4.00 4.00 4.00 2.00 1.00 1.00 0.50
Trp 0.00 ND 0.00 0.00 0.00 0.00 0.00
a
The amino acid composition calculated from the deduced sequence.
b
The amino acid composition analyzed by amino acid analysis.

cp-19ks agreed well with each other, especially in terms
of the content of the six dominant residues, Gly, Thr,
Ser, Ala, Lys and Val (Table 2).
A blast search of the nonredundant database and
a sequence profile-based fold-recognition method for
three-dimensional structural prediction failed to pro-
vide any homologous sequences and meaningful struc-
ture (supplementary Document S1). In particular, no
sequence similarity between cp-19ks and foot proteins
in the mussel was evident. The primary structures of
cp-19ks also showed no homology with cp-100k and
cp-20k. Naldrett & Kaplan [14] have reported the par-
tial amino acid sequences of peptide fragments from
B. eburneus cement. Among these fragments, WCD-21,
a peptide fragment obtained by cyanogen bromide
treatment of B. eburneus cement, showed homology to
the N-terminal region of cp-19ks (supplementary
Fig. S1B), indicating that the protein homologous to
cp-19k should also be present in B. eburneus cement.
Characterization of the recombinant Mrcp-19k
protein
Recombinant (r)Mrcp-19k was expressed in Escherichia
coli as a soluble cytosolic fraction, and was purified to
homogeneity (Fig. 1). rMrcp-19k had a slightly lower
mobility than that of the native Mrcp-19k isolated
from the cement. This was due to the additional N-ter-
minal dipeptide in the recombinant protein as the
result of the vector design. The N-terminal sequence
and molecular mass were determined to be
AMVPPPXDLG and 17 201 Da (predicted molecular

Mrcp19k VPPPCDLGIASKVKQKGVTGGGASVSTTSATQGSGTTNCVTRTPNSVEKKNVAGNTGVTA
Bacp19k VPPPCDLSIKSKLKQVGATAGNAAVTTTGTTSGSGVVKCVVRTPTSVEKKAAVGNTGLSA
Bicp19k VPPPCDFSIKSKQKQVGVTAGGASVSAKGATSGSGSITCITKTPTSVTKKVAAGNAGVSG
70
11020 30 4050 60
80 90 100 110 120
Mrcp19k TSVSAGDGAFGNLAAALTLVEDTEDGLGVKTKNGGKGFSEGTAAISQTAGANGGATVKKA
Bacp19k VSASAANGFFKNLGKATTEVKTTKDGTKVKTKTAGKGKTGGTATTIQIADANGGVSEKSL
Bicp19k AAAAAGNGVFKNLVTALTNISTTDDITKVQTQTIGSGGTGGAATILQLADANGGAALKEV
130 140 150 160 170
Mrcp19k KLDLLTDGEDLFDTKKVEKGTVTSSSSHQGSGAGDSIFEILNEAESKIKKSGD
Bacp19k KLDLLTDGLKFVKVTEKKQGTATSSSGHKASGVGHSVFKVLNEAETELELKGL
Bicp19k KLDLLPIGTGLGVVKQTKQGQVTSSSSHKASGLGNSVLKVLNAHETELKLKGI
Fig. 2. Alignment of the amino acid
sequences of mature cp-19ks.The deduced
amino acid sequences of mature Mrcp-19k,
Balcp-19k and Bicp-19k were aligned by
CLU-
STALW
[34]. The three homologous proteins
have the same amino acid length, and the
two Cys residues are conserved. Identical
amino acids are reversed.
Y. Urushida et al. Barnacle surface-cement protein
FEBS Journal 274 (2007) 4336–4346 ª 2007 The Authors Journal compilation ª 2007 FEBS 4339
software (supplementary Fig. S2). The adsorption con-
stant k
a
and desorption constant k
d

k
d
and K
eq
were calculated to be 9.76 · 10
4
m
)1
Æs
)1
,
6.67 · 10
)4
s
)1
and 1.46 · 10
8
m
)1
, respectively. The
amounts adsorbed to the glass and formaldehyde resin
surfaces in 5 min at 25 °C were estimated, and the
results are shown in Table 3 and supplementary
Fig. S3.
Localization and expression site of Mrcp-19k
M. rosa cement was usually collected by gently scrap-
ing the surface of the calcareous base on the side
attached to the foreign material surface [10], making
the cement proteins vulnerable to contamination by
calcified material during the process of collection. We

HPA:ASW
Fig. 3. Typical SPR analyses on polycrystalline gold and alkylated gold. The arrows and thick arrows indicate the starts of sample loading
(2 l
M) and washing by the running buffer, respectively. The processes of sample loading and washing were sequentially repeated three
times. Open circular symbols, squares and triangles indicate changes of resonance after protein adsorption on polycrystalline gold in ASW,
on the same material in a dilute buffer containing 10 m
M Tris (pH 7.4) ⁄ 25 mM NaCl, and on alkylated gold (HPA) in ASW, respectively. DRUs
after each washing process were as follows: first loading on Au in ASW, 1174 RU; second loading on Au in ASW, 1177 RU; third loading on
Au in ASW, 1182 RU; first loading on Au in dilute buffer, 1278 RU; second loading on Au in dilute buffer, 1318 RU; third loading on Au in
dilute buffer, 1345 RU; first loading on alkylated gold in ASW, 768 RU; second loading on alkylated gold in ASW, 827 RU; third loading on
alkylated gold in ASW, 858 RU.
Table 3. Amount of adsorption of rMrcp-19k to several material
surfaces. The adsorbed amount in ASW or dilute buffer solution
was calculated from the change in RU on SPR [36] for gold and
alkylated gold, and from a quantitative amino acid analysis for glass
and the formaldehyde resin (see details in supplementary Fig. S3).
Surface area per molecule was calculated by a assuming full sur-
face monolayer coverage.
Gold
Alkylated
gold Glass
Formaldehyde
resin
Adsorption amount
(ngÆmm
)2
)
0.76
(0.83 in
dilute buffer)

Mrcp-20k, which has been identified previously [11],
were shown to be simple proteins. Thus, the barnacle
seems to manage its underwater attachment activity
well with common amino acids.
The bacterial recombinant protein of Mrcp-19k,
rMrcp-19k, was prepared in soluble form under physi-
ologic conditions, enabling us to directly measure its
adsorption to underwater surfaces. Two Cys residues
in the protein formed an intramolecular disulfide bond,
probably with the help of a thioredoxin-tag in the
vector system. rMrcp-19k was adsorbed to various
characteristic surfaces, including negatively charged,
positively charged and hydrophobic surfaces. The bar-
nacle attaches to various foreign material surfaces,
including metal oxide, glass, plastic, wood, and rock.
Naturally occurring surfaces such as rock are not
microscopically homogeneous, and have a patchwork
of different surface characteristics. The cement is there-
fore required to simultaneously adapt the molecular
event to different surfaces. The ability of Mrcp-19k
to be adsorbed to various surfaces suggests that this
protein may be responsible for the surface func-
tions, at least for the ability of the barnacle cement
to adsorb to foreign materials with different surface
characteristics.
Polycrystalline gold and hydrophobic alkylated gold
were used as the representative surfaces in this study
for evaluating the adsorption isotherm. The surface
attachment area of a protein molecule on the gold sur-
face was calculated to be 37 nm

base plate, respectively, which have been decalcified and rendered
soluble by the dithiothreitol ⁄ guanidine hydrochloride treatment.
Lanes 1–3 correspond to GSF1, GSF2 and the recombinant protein
rMrcp-19k, respectively, as positive controls. (B) Antibody to Mrcp-
100k was used for the analysis. Lane 2 shows the primary cement
with the dithiothreitol ⁄ guanidine hydrochloride treatment. Lanes 3
and 4 show the barnacle peripheral shell and base plate, respec-
tively, which have been decalcified and rendered soluble by the
dithiothreitol ⁄ guanidine hydrochloride treatment. Lane 1 corre-
sponds to GSF2 as a positive control.
Basal
Upper
Basal
Upper
Fig. 5. Site specificity of Mrcp-19k gene expression in the basal
portion of the adult barnacle, where the histologically identified
cement gland is located.Twenty micrograms of total RNA extracted
from the basal or upper portion of the adult barnacle was electro-
phoresed in formaldehyde gel, transferred to a nylon membrane,
and hybridized with a probe. The basal portion mainly comprises
the mantle, muscle, ovariole, cement gland [20–22], and hemo-
lymph, whereas the upper portion contains the cirri, thorax, pro-
soma and hemolymph. Left, northern blot; right, 18S rRNA on gel
stained by ethidium bromide.
Y. Urushida et al. Barnacle surface-cement protein
FEBS Journal 274 (2007) 4336–4346 ª 2007 The Authors Journal compilation ª 2007 FEBS 4341
repulsion on the surface. The amounts adsorbed to
both glass and formaldehyde resin were two-fold to
five-fold the amount adsorbed to bare gold. These
data, however, were obtained with a method that

not accurately determined in this study, it was by no
means a major component. Cement proteins contribut-
ing to surface functions might be minor constituents,
whereas the proteins for bulk functions [9] would be
present in much higher amounts in the adhesive layer.
Northern blot analysis has indicated that the Mrcp-
19k gene is specifically expressed in the basal portion
of the animal, where the histologically identified
cement gland is located [20–22]. This result is consis-
tent with that for Mrcp-100k [9]. The cement proteins
are probably biosynthesized together in the cement
gland and transported by a duct to the narrow inter-
space outside, between the animal’s base and the for-
eign substratum.
In conclusion, this study has identified a novel pro-
tein, cp-19k, in barnacle cement and demonstrated that
it is able to be adsorbed to various underwater sur-
faces, suggesting that this protein is a surface protein
of the cement complex. Our results also revealed that
the function of cp-19k is dependent on common amino
acid residues on the molecular surface. This is in con-
trast to the underwater adhesive proteins of mussel
and tubeworm studied so far, where modified amino
acids have been found to play major roles [23,24]. The
barnacle cement protein characterized in this study
may therefore represent a new mechanism of biological
adhesion, which is likely to be useful in helping the
interdisciplinary links between biotechnology and
material science, e.g. development of adsorbents for
various material surfaces, of support for protein align-

difluoride) membrane (ProBlott; Applied Biosystems, Foster
City, CA, USA) using a Tris ⁄ borate buffer containing 0.1%
SDS [29], and was stained with Coomassie Brilliant Blue
R-250. In order to get peptide fragments of Mrcp-19k, the
band corresponding to Mrcp-19k on the poly(vinylidene di-
fluoride) membrane before Coomassie Brilliant Blue staining
was cut out and subjected to in situ enzymatic digestion [30]
using lysylendopeptidase (Wako Pure Chemical Industries).
The generated peptide fragments were separated and frac-
tionated by RP-HPLC in a 3.9 mm diameter · 150 mm
Barnacle surface-cement protein Y. Urushida et al.
4342 FEBS Journal 274 (2007) 4336–4346 ª 2007 The Authors Journal compilation ª 2007 FEBS
l-Bondasphere column (C18, 100 A
˚
; Waters, Milford, MA,
USA). The amino acid sequence was determined with a
Procise 494 cLC (Applied Biosystems) or PSQ-2 protein
sequencer (Shimadzu, Kyoto, Japan). Mrcp-19k was also
purified from GSF1 by ion exchange chromatography (SP
Sepharose FF; Amersham Biosciences, Uppsala, Sweden).
The column was equilibrated with 50 mm acetic acid, and
eluted with a linear gradient of NaCl from 0 m to 0.6 m in
80 min. The fractions were monitored with a polyclonal anti-
body raised against the bacterial recombinant protein corre-
sponding to the C-terminal 10 kDa portion of Mrcp-19k, as
described in the latter section of recombinant in E. coli except
for using 5¢-TGG CCG CAG
CCA TGG CAT TGG T-3¢
as the 5¢-primer. The fraction containing Mrcp-19k was
concentrated by ultrafiltration (Microcon YM-3; Amicon-

+
RNA
was isolated using Oligo(dT)-Latex Super (Takara Shuzo
Co.). cDNA was prepared from mRNA with a Zap-cDNA
synthesis kit (Stratagene, La Jolla, CA, USA) according to
the instructions of the supplier. DNA fragments of Mrcp-
19k were first amplified by PCR (ExTaq; Takara) with fully
degenerated PCR primers designed from the N-terminal
amino acid sequence of Mrcp-19k: 5¢-GTN
CCN CCN CCN TGY GA-3¢ and 5¢-CAN CCY TTY TGY
TTN ACY TT-3¢. The PCR products were resolved by 3%
NuSieve 3 : 1 agarose (Takara) gel electrophoresis, and a
53 bp DNA fragment from M. rosa was purified from
the gel. The DNA fragment was subcloned in pT7 Blue
T-Vector (Novagen, EMD Biosciences, Madison, WI,
USA), and the insert was sequenced using a Prism Dye
Deoxy sequencing kit and 3700-DNA analyzer (Applied
Biosystems). 3¢-RACE was then carried out with a specific
3¢-RACE primer designed from the 53 bp DNA and using
a3¢-RACE core kit (Takara). The 3¢-RACE primer used
was 5¢-CTG ATC TAG AGG TAC CGG ATC CGT TCC
CCC ACC ATG CGA CCT TGG CAT-3¢. The PCR pro-
duct was subcloned and then sequenced. To obtain the
full-length cDNA, 5¢-RACE was carried out with oligo-
nucleotide primers designed from the sequence of 750 bp
DNA and using a 5¢-RACE core kit (Takara). The 5¢-
RACE primers used were as follows: 5¢-G#CC GTC CCC
GGC CGA C-3¢, where G# is phosphorylated, for reverse
transcription; 5¢-GTG CCG GAG CCC TGC GTG GC-3¢
and 5¢-AAC TCC GTG GAG AAG AAG AA-3¢ for the

created NcoI and BamHI restriction sites. The primers used
were 5¢-ACCGGCCATGGGCAAGGCCGT-3¢ and 5¢-AT
GGTCACGGGATCCCTCCGGTGGTCTTA, whereby the
Y. Urushida et al. Barnacle surface-cement protein
FEBS Journal 274 (2007) 4336–4346 ª 2007 The Authors Journal compilation ª 2007 FEBS 4343
recombinant was designed to have the N-terminal sequence
of AMGKAVTV, in which the mature N-terminal sequence
of Mrcp-19k with an additional dipeptide sequence, AM,
was created after removing the fused tag by enterokinase
cleavage, and with the original C-terminal end. The ampli-
fied DNA was subcloned in pT7 Blue T-Vector (Novagen),
and the sequence was confirmed. Insert DNA was gene-
rated by digestion with the NcoI and BamHI restriction
enzymes, and then subcloned into pET32b (Novagen) with
the same restriction sites. The pET32 vector system pro-
duces fusion proteins with a thioredoxin-tag, which
enhances disulfide bond formation of the target protein in
the cytoplasm of the host strain. The created vector was
transformed into the expression host strain Oligami (DE3)
(Novagen). The recombinant protein was purified with a
metal-chelating column according to the affinity of the His-
tag fused into Mrcp-19k. The cells were inoculated in LB
medium [35] containing ampicillin at 37 °C for 3 h, and
transferred to freshly prepared medium, and inoculated for
another 3 h; protein expression was induced by 0.2% iso-
propyl-thio-b-d-galactoside for an additional 4 h. The cyto-
solic fraction was prepared by sonicating on ice in 20 mm
Tris (pH 7.4) ⁄ 500 mm NaCl ⁄ 40 mm imidazole, and purified
with an Ni
2+

tron, Waltham, MA, USA), either with or without pretreat-
ment with dithiothreitol.
Adsorption of the recombinant protein to underwater
material surfaces was analyzed by: (a) quantitative amino
acid analysis; and (b) SPR.
Protein adsorption to glass and a positively charged poly-
mer were evaluated by quantification of the bound protein
and unbound protein, respectively, by amino acid analysis
after hydrolysis (see details in supplementary Fig. S3). The
substrates to be analyzed were the inner surface of small
glass test tube (5 mm in diameter and 29 mm in length,
73.6 mm
2
for covering the surface area of a 20 lL solution)
and benzoguanamine ⁄ formaldehyde resin particles (Epostar
L15, 11.6 lm in diameter, 73.6 mm
2
for the surface area
test; Nippon Shokubai, Osaka, Japan). The amount
adsorbed in 5 min at 25 °C in ASW was measured with
several protein concentrations and fitted using a Langmuir
adsorption isotherm.
The SPR measurements were performed with a BIA-
core 3000 system (Biacore AB, Uppsala, Sweden) at 25 °C
and with a flow rate of 10 lLÆmin
)1
. The sensor chips of
polycrystalline gold-coated and octadecanethiol-terminated
gold, HPA, were purchased from BIAcore. The running
buffer was 10 mm Tris (pH 7.4) ⁄ 25 mm NaCl with or with-

using biaevaluation version 3.1 software that was sup-
plied with the instrument.
Localization and expression site of Mrcp-19k
To confirm that cp-19k was a cement component, the local-
ization of Mrcp-19k in the primary cement and in the pro-
tein fractions of both the base shell and peripheral shell of
the animal were investigated by western blotting. Polyclonal
antibodies were raised using bacterial recombinants of the
respective C-terminal regions of approximately 10 kDa in
Mrcp-19k and Mrcp-100k as antigens in rabbits with serial
subcutaneous injections. The recombinants were prepared
as described earlier. The primers used for amplifying the
Mrcp-19k portion were 5¢-TGG CCG CAG CCA TGG
CAT TGG T-3¢ and 5¢-ACC TCA GGA TCC AGG TCG
AGA AAA-3 ¢. The primers used for amplifying the Mrcp-
100k portion were 5¢-AGT GCA GCC CAT GGG GGC
AGC CAT-3 ¢ and 5¢-TTG CCT AGG TGG ATC CTC
AGC ATC TGA A-3¢. M. rosa primary cement was
collected as previously reported [10]. The base and peri-
pheral shell were separately collected from living M. rosa
Barnacle surface-cement protein Y. Urushida et al.
4344 FEBS Journal 274 (2007) 4336–4346 ª 2007 The Authors Journal compilation ª 2007 FEBS
specimens, and physically cleaned to remove all contamina-
tion by the animals’ soft tissue. Each shell was decalcified
by dialyzing against 2% acetic acid at 4 °C, and the precip-
itate was recovered. Although the supernatant was also
analyzed, no signal was detected by western blotting. The
precipitate was evaporated to dryness, denatured, separated
by SDS ⁄ PAGE (a Tris ⁄ Tricine buffer system, 16.5% T ⁄ 3%
C for Mrcp-19 k, and 8% T [37] with 6 m urea for Mrcp-

Technology Development Organization (NEDO).
References
1 Sagert J, Sun C & Waite JH (2006) Chemical subtleties
of mussel and polychaete holdfasts. In Biological Adhe-
sives (Smith AM & Callow JA, eds), pp. 125–140.
Springer-Verlag, Berlin.
2 Taylor SW & Waite JH (1997) Marine adhesives: from
molecular dissection to application. In Protein-Based
Materials (McGrath K & Kaplan D, eds), pp. 217–248.
Birkhauser, Boston, MA.
3 Zhao H & Waite JH (2006) Linking adhesive and struc-
tural proteins in the attachment plaque of Mytilus cali-
fornianus. J Biol Chem 281, 26150–26158.
4 Kamino K (2006) Barnacle underwater attachment. In
Biological Adhesives (Smith AM & Callow JA, eds),
pp. 145–166. Springer-Verlag, Berlin.
5 Waite JH (1987) Nature’s underwater adhesive special-
ist. Int J Adhes 7, 9–14.
6 Tay FR & Pashley DH (2002) Dental adhesives of the
future. J Adhes Dent 4, 91–103.
7 Walker G (1972) The biochemical composition of
the cement of two barnacle species, Balanus hameri
and Balanus crenatus. J Mar Biol Assoc UK 52,
429–435.
8 Saroyan JR, Linder E, Dooley CA & Bleile HR (1970)
Repair and reattachment in the Balanidae as related to
their cementing mechanism. Ind Eng Chem Prod Res
Dev 9, 122–133.
9 Kamino K, Inoue K, Maruyama T, Takamatsu N,
Harayama S & Shizuri Y (2000) Barnacle cement pro-

threonine methyl and hydroxyl groups in antifreeze
activity. J Biol Chem 273, 34806–34812.
19 Jia Z & Davies PL (2002) Antifreeze proteins: an
unusual receptor–ligand interaction. Trends Biochem Sci
27, 101–106.
20 Walker G (1970) The histology, histochemistry and
ultrastructure of the cement apparatus of three adult
sessile barnacles, Elminius modestus, Balanus balanoides
and Balanus haemri. Mar Biol 7, 239–248.
21 Lacombe D (1970) A comparative study of the cement
glands in some balanid barnacles (cirripedia, balanidae).
Biol Bull 139, 164–179.
Y. Urushida et al. Barnacle surface-cement protein
FEBS Journal 274 (2007) 4336–4346 ª 2007 The Authors Journal compilation ª 2007 FEBS 4345
22 Saroyan JR, Lindner E & Dooley CA (1970) Repair
and reattachment in the balanidae as related to their
cementing mechanism. Biol Bull 139, 333–350.
23 Statz AR, Meagher RJ, Barron AE & Messersmith PB
(2005) New peptidomimetic polymers for antifouling
surfaces. J Am Chem Soc 127, 7972–7973.
24 Deming TJ (1999) Mussel byssus and biomolecular
materials. Curr Opin Chem Biol 3, 100–105.
25 Chen X, Ferrigno R, Yang J & Whitesides GM (2002)
Redox properties of cytochrome c adsorbed on self-
assembled monolayers: a probe for protein conforma-
tion and orientation. Langmuir 18, 7009–7015.
26 Hyun J, Lee WK, Nath N, Chilkoti A & Zauscher S
(2004) Capture and release of proteins on the nanoscale
by stimuli-responsive elastin-like polypeptide ‘switches’.
J Am Chem Soc 126, 7330–7335.

34 Higgins D, Thompson J, Gibson T, Thompson JD, Hig-
gins DG & Gibson TJ (1994) CLUSTAL W: improving
the sensitivity of progressive multiple sequence align-
ment through sequence weighting, position-specific gap
penalties and weight matrix choice. Nucleic Acids Res
22, 4673–4680.
35 Luria SE & Burrous JW (1957) Hybridization between
Escherichia coli and Shigella. J Bacteriol 74, 461–476.
36 Hook F, Kasemo B, Nylander T, Fant C, Sott K &
Elwing H (2001) Variations in coupled water, viscoelas-
tic properties, and film thickness of a Mefp-1 protein
film during adsorption and cross-linking: a quartz crys-
tal microbalance with dissipation monitoring, ellipsome-
try, and surface plasmon resonance study. Anal Biochem
73, 5796–5804.
37 Laemmli UK (1970) Cleavage of structural proteins
during the assembly of the head of bacteriophage T4.
Nature 227, 680–685.
38 Enami I, Murayama H, Ohta H, Kamo M, Nakazato K
& Shen J-R (1995) Isolation and characterization of a
photosystem II complex from the red alga Cyanidium
caldarium. Association of cytochrome c-550 and a 12
kDa protein with the complex. Biochim Biophys Acta
1232, 208–216.
Supplementary material
The following supplementary material is available
online:
Fig. S1. (A) cDNA and deduced amino acid sequences
of Mrcp-19k, Balcp-19k and Bicp-19k. (B) Alignment of
the mature N-terminal region of cp-19ks with WCD-21,