Mar Biotechnol (2008) 10:111–121 DOI 10.1007/s10126-007-9076-3
INVITED REVIEW
Underwater Adhesive of Marine Organisms as the Vital Link Between Biological Science and Material Science Kei Kamino
Received: 10 August 2007 / Accepted: 4 December 2007 / Published online: 16 February 2008 # Springer Science + Business Media, LLC 2008
Abstract Marine sessile organisms naturally attach themselves to diverse materials in water by a technique that has so far remained unreproducible. Recent studies on the holdfast of marine sessile organisms have revealed natural concepts that are currently beyond our understanding with respect to the molecular design and macroscopic range. The combination of valuable and practical natural design of biotic adhesives as biomolecular materials, together with continuing efforts towards mimetic design, hold the promise of revolution for future materials. This review focuses on recent advances in the study of barnacle underwater cement, a protein complex whose constituents and the properties of individual components are being uncovered. A comparison is made with the model systems used by the mussel and tubeworm. Keywords sessile organism . adhesive . multi-protein complex . self-assembly . coupling
Introduction Adhesive technology is indispensable for our daily life as well as industry. Various adhesives are used for a number of different purposes but with the proviso “in air.” Chemically synthesized polymers, which are the basis of recent adhesive technology, are usually produced in an organic solvent, and are, in general, incompatible with water. This may be the reason for aspiring the expansion of this technology to bond two materials together “in water.” K. Kamino (*) Marine Biotechnology Institute, Kamaishi, Iwate 026-0001, Japan e-mail:
[email protected]
There is a diverse array of sessile organisms in the marine environment. Attachment to substrata is essential for these organisms, because they then benefit from access to resources necessary for their life, protection from predators, and also from improvement in gene transfer. This attachment is intimately linked to such other physiological functions, such as metamorphosis, molting, and biomineralization. Thus, a secure holdfast is the basis for their life. The holdfast utilized is a natural bio-molecule and seems to be multiprotein complexes in most cases. This bio-molecule is, in contrast to a synthetic polymer, biosynthesized in an aqueous environment, and therefore water is an essential ingredient of the molecule. This forms the basic difference between the holdfast system of marine organisms and synthetic polymers, the former being capable of undergoing underwater attachment that are usually not achievable by the latter. Thus, understanding the biotic molecular system may lead to future technology in underwater attachment. Although sessile organisms are diverse in their types, studies on the chemical nature of their holdfasts have generally been limited to the mussel, tubeworm, and barnacle. The former two animals are characterized by their reliance on post-translational modification (Sagert et al. 2006), especially 3,4-dihydroxy phenylalanine (DOPA), whereas the third is distinct from these two animals in that it does not involve the “DOPA-system” (Naldrett 1993; Kamino et al. 1996). Mussel byssus has been studied extensively, and results from these studies have revealed the ingenuity of the system and have had several impacts on material science. Researches on mussel byssus have so far identified nine unique proteins and found several posttranslational modifications, including DOPA (Waite and Tanzer 1981), hydroxyproline, dihydroxyproline (Taylor and Waite 1994), 4-hydroxyarginine (HyArg; Papov et al. 1995), O-
112
Phosphoserine (Pho-Ser, Waite and Qin 2001), and glycosylation (Anderson and Waite 1998; Ohkawa et al. 2004). DOPA is actually involved in cross-linking in several ways (McDowell et al. 1999; Zhao and Waite 2006b; Yu et al. 1999; Nagai and Yamamoto 1989) including metal coordination bonds (Sever et al. 2004), and in adsorption to a metal surface (Zhao et al. 2006; Ooka and Garrell 2000; Lee et al. 2006) in byssus formation. Although several intriguing concepts of material design, including property-gradient (Waite et al. 2004) and a solid foam structure (Stewart et al. 2004; Waite 1986) have been inferred from the byssus, versatility of DOPA as a cross-linker and coupling agent has had the greatest impact (Deming 1999; Ninan et al. 2004; Hwang et al. 2007; Lee et al. 2007). This versatility of DOPA may, however, obscure our view on the nature of it as a biological molecule; thus, limited attention has been paid to the significance of protein conformation or molecular interaction based on this conformation, in underwater adhesion. Advances in the study of barnacle have revealed the involvement of a completely different molecular system from that of the mussel and tubeworm (Kamino 2006). The barnacle system has shown the significances of commonly used amino acids, the protein backbone conformation, and the intermolecular noncovalent interaction, in its functioning. The research also revealed that underwater adhesive proteins may form a good model for developing peptidebased materials that self-assemble in a noncovalent fashion. Underwater attachment is multifunctional (Waite 1987), which involves displacing the bound-water layer on a foreign substratum with the adhesive, as well as spreading, coupling the adhesive with a variety of material surfaces, self-assembly of the adhesive, curing to make the holdfast stiff and tough, and protecting from microbial degradation, although the insoluble and sticky nature of the materials involved have obstructed an understanding of the structural relationship among these subfunctions. Nevertheless, recent cooperation between biological science and material science has contributed to a deeper understanding of the system, and continuing efforts to resolve the mysteries of biological molecular system will surely bring richer fruits. In this review, I will mainly focus on recent advances in barnacle cement studies. Although the terminology of underwater adhesives also include sticky glue, such as those used by the echinoderm (Flammang et al. 2005), mollusc (Werneke et al. 2007), cephalopod (von Byern and Klepal 2006), and algal spore (Humphrey et al. 2005), these are not addressed in this article.
Barnacle as a Unique Sessile Crustacean The barnacle is a unique sessile crustacean that never moves once the cypris larva has settled onto a foreign substratum
Mar Biotechnol (2008) 10:111–121
(Fig. 1). The adult barnacle bonds two different materials in water: its own calcareous base with such foreign substrata as metals, minerals, synthetic polymers, and biotic surfaces (Fig. 2). The underwater adhesive is biosynthesized in a specialized cement gland and is secreted via a duct into the space of a few microns in width to join the base to the substratum (Saroyan et al. 1970; Walker 1970; Lacombe 1970). In other words, barnacle underwater attachment after secretion from the duct is an entirely molecular event. The underwater adhesive, or cement, is a multiprotein complex (Fig. 2). The adhesive layer, however, seems to be macroscopically uniform (Weigemann and Watermann 2003), with all the proteins being ejected together into the interspace for attachment. This might be similar to the cement of the tubeworm (Stevens et al. 2007), whereas different from that of the mussel (Waite et al. 2005) (Fig. 3). The mussel attaches to a foreign substratum by making several tens of threads, byssus, whose distal end is actually bonded to the foreign material. The byssus consists of macroscopically distinct portions: the disk, thread, stem, and coating. The tip of the byssus, called the disk, directly attaches to the foreign material, and the animal is connected to the proximal end of the thread via the stem. The disk is actually composed of three portions: the surface that will be attached to the foreign material, the bulk of the disk, and the joining part to the distal thread. The thread is composed of a graded content of two different proteins along its length. The macroscopic structure of the segment, which is essential for the durability of the byssus, results from the longitudinal arrangement of each secretory gland along the specialized organ—the foot. The holdfast of the barnacle and mussel are thus morphologically different. As stated earlier, underwater attachment is a multifunctional process (Waite 1987). The subfunctions are classified roughly into surface functions and bulk functions. The surface functions include displacement of the water layer, spreading the cement, coupling to diverse materials, and cleaning the biofilm, whereas the bulk-functions include self-assembly, curing to suitable toughness, and protection from microbial degradation. Sessile organisms provide a multiprotein complex for these multiple functions, suggesting that each component has some responsibility for each one of the subfunctions. No macroscopic segment structure is known in barnacle cement or any information about the microscopic localization of a specific component within the cement. Therefore, it is not known whether the category of functions in underwater attachment and microscopic localization of the corresponding molecules in the cement are simply parallel. This needs to be given more consideration. Although the foot-print of a cypris larva in the exploring stage and cypris cement for its settlement (Okano et al. 1996) also are intriguing, these aspects are not covered in this article.
Mar Biotechnol (2008) 10:111–121
113
Several days ~ a few weeks
1 day
Cypris
Substratum Fig. 1 The barnacle as a unique sessile crustacean. The cypris larva never moves once it has settled. Right: Megabalanus rosa attached to titanium oxide and polyethylene (inset)
The Cement The barnacle cement at the attachment site is insoluble, which is a common obstacle for biochemical analyses of an underwater cement or holdfast. The cement is actually proteinaceous with more than 90% of its content occupied by proteins (Kamino et al. 1996). Methods developed to solubilize it have revealed that the cement is a multiprotein complex (Kamino et al. 2000), of which six proteins have been identified and are categorized into four groups: six amino-acid-biased proteins, a charged amino acid-rich protein, remarkably hydrophobic proteins, and an enzyme (Kamino 2006) (Fig. 2). All these cement proteins, except for the enzymatic one, are novel without significant homologues in databases currently available. Two proteins are of simple type, bearing no posttranslational modifications, whereas one protein bears limited glycosylation, and three proteins are unknown for any possible modifications. Molecular systems not relying on post-translational modification has not been found in other underwater adhesives. All mRNAs for the cement proteins, except the enzyme that has not yet been examined, are expressed only in the basal portion of the animal where the histologically identified cement gland is located. The expression levels of the mRNAs are tuned on and off together, and seem to increase upon approaching of the molting stage.
Charged amino acid-rich: cp-20k Hydrophobic: cp-100k, -52k Six amino acid-biased: cp-68k, -19k Unknowns: cp-165k, -130k, -89k, -14k
Enzyme: cp-16k
To complete the underwater attachment successfully, all of its subfunctions must be operated with proper timing in the cement. Although artificial reproduction has yet to be achieved, interpretation has been assisted by the characterization of each of its components, which will be discussed in the subsequent sections.
Cement Protein with a Surface Function: A Multi–surface-coupling Protein The 19-kDa protein (cp-19k) is a minor component of the cement in terms of its amount (Urushida et al. 2007). The molecular mass of the protein purified from the cement agrees with that estimated from the cDNA sequences, indicating that the protein bears no posttranslational modifications. A recombinant protein of Megabalanus rosa (Mr) cp-19k, rMrcp-19k, was produced with E. coli in a soluble form under physiological conditions, enabling us to directly measure its adsorption properties to underwater surfaces. rMrcp-19k was adsorbed to surfaces with various characteristics, including negatively charged, positively charged, and hydrophobic surfaces, in water. The materials that were able to adsorb the protein in seawater include bare gold, alkylated gold, glass, TiO2, polystyrene, polyethylene, positively charged polymers, and hydroxyapatite. Polycrystalline gold and masses adsorbed onto the surfaces after washing were
Calcareous base Cement gland Cement Substratum
Substratum a few microns
Fig. 2 Cross-section of the barnacle. The cement is biosynthesized in the cement gland and is transported via a duct to the narrow gap between the animal’s own calcareous base and the foreign substratum.
The thickness of the cement layer is generally only a few microns. The cement, which is a multiprotein complex, joins the two different materials in water
114 Fig. 3 Byssus as the mussel holdfast. The byssus is formed by a special organ, the foot, and comprises an individual structure. One byssus is produced within 5 min. The distal end of the byssal thread attaches to the foreign material via a disk. See Waite et al. (2002) for details
Mar Biotechnol (2008) 10:111–121
Foot
Removing and reforming to move
Formed within 5min
Byssus
Coating (fp-1)
Thread (several cm) (PreCol-P, -ND, -D)
Disk (fp-4, -2, -6, -3, -5) Substratum
almost the same by repetitively exposing the protein to both surfaces. The adsorption constant, ka, desorption constant, kd, and equilibrium constant, Keq, were calculated as 2.17×105 M1 -1 s , 4.94×10-4 s-1, and 4.39×108 M-1, respectively, for the formation of the rMrcp-19k/Au complex. The values of ka, kd, and Keq for the hydrophobic alkylated gold surface were calculated to be 9.76×104 M-1s-1, 6.67×10-4 s-1, and 1.46×108 M-1, respectively. The adsorption was a rapid process, and the protein formed a stable adhesive layer on the material surfaces. Because the barnacle attaches itself to diverse foreign material surfaces, including metal oxide, glass, plastic, wood, and rock, and naturally occurring surfaces are a patchwork of different surface characteristics but not microscopically homogenous, the cement must adapt the molecular event of adsorption simultaneously to different surfaces. These may support the idea that the cp-19k protein is responsible for the surface functions, at least for the coupling ability of the barnacle cement to foreign materials with different surface characteristics. Genes homologous to that of Mrcp-19k were isolated from Balanus albicostatus and B. improvisus, and their amino acid compositions strongly resembled that of M. rosa, with six amino acids, Ser, Thr, Ala, Gly, Val, and Lys, comprising 66% to 70% of the total (Fig. 4), although the overall sequences similarity was by no means high. This indicates that the function of the protein may be associated with the highly biased amino acid composition. The four amino acids, Ser, Thr, Lys, and Val, in the protein would be useful to couple with diverse foreign material surfaces via hydrogen bonding, electrostatic, hydrophobic interaction, etc. During the initial process of underwater attachment, a cement protein is required to approach the
solid substratum to which water molecules are bound, followed by the displacement of the water molecules before coupling with the substratum surface. Waite (Waite 1987) has suggested the significance of the hydroxyl group on the Ser and Thr residues for this priming process. In a relevant protein, the antifreeze protein, which binds to the ice nucleus to inhibit crystal growth in the cytosolic space of several organisms including bacteria and fish (Fletcher et al. 2001), the Ala and/or methyl group of Thr on the molecular surface of the protein are known to be essential in the process of binding to the ice nucleus (Zhang and Laursen 1998; Jia and Davies 2002), although the exact roles of these amino acids are not yet clearly understood. The requirements of coupling to diverse foreign material surfaces and displacing water molecules bound to a solid substratum may result in the bias of six amino acids in the barnacle cement protein. Bio-informatic analyses have suggested that the primary structures of cp19k are composed of four alternating repetitions of two segments: [S, T, G, and A]-rich, and [V and K]-rich (Fig. 4). The block-copolymer-like primary structure is intriguing. Each segment may separately function as already discussed. The segment structure also is intriguing because of its similarity to another cement protein, cp-68k, whose function in underwater attachment is not yet known. Cp-68k is characterized by a biased amino acid composition, with Ser, Thr, Ala, and Gly comprising 60% of the total residues and an almost even ratio among these four residues (Kamino et al. 1996). Cp-68k has been isolated from both M. rosa and B. amphitrite, each with a similar amino acid composition. These two sequences
Mar Biotechnol (2008) 10:111–121
115
Sample loading 1600
Response Unit (RU)
1400 1200 1000 Asp
800
Trp
150
Asn
Phe
600
Ser
Leu
Glu
Ile
400
Gln
Lys
200
Gly
0
Met
0 0
100
200
300
400
500
His
Val
Arg Tyr
-200
Thr Cys
time (sec)
Ala Pro
Mrcp19k
Balcp19k
Bicp19k
cp-19k C-terminal
N-terminal S,T,G,A-rich
V,K-rich
cp-68k S,T,G,A-rich
V,K-rich
Fig. 4 The six-amino-acid-biased proteins, cp-19k and cp-68k, share a similar segment structure. A typical adsorption trace for rMrcp-19k by surface plasmon resonance on to bare gold in ASW is shown
were 47% identical and 65% similar. The alignment of these two cp-68ks indicates that the primary structure is divided into two regions: a long Ser-, Thr-, Ala-, and Glyrich N-terminal region, and a short C-terminal region composed of much less of these four amino acids and more of amino acids, such as Lys, Pro, Trp, Cys, and hydrophobic types. Further studies on the structure of cp19k and on the functional identification of cp-68k are required.
Cement Protein with a Surface Function: Coupling Protein to the most Frequently Encountered Material for the Animal The 20-kDa protein (cp-20k) is another component of Megabalanus rosa cement and also is minor (1–2%) in terms of its amount (Kamino 2001; Suzuki et al. 2005; Mori et al. 2007). Mrcp-20k does not carry posttranslational modifications. The amino acid composition of cp-20k is characterized by its unusually high abundance of Cys
(18%) and charged amino acids (Asp, 11.5%; Glu, 10.4%; His, 10.4%). All Cys residues of the protein in the adhesive layer are in the intramolecular disulfide form, and the protein functions in a monomeric form. The gene encoding this protein was expressed in E. coli, and the recombinant protein was purified in a soluble form under physiological conditions, which was confirmed to retain almost the same structure as that of the native barnacle protein. Determination of the solution structure using NMR also is in progress. CD spectra and NMR data after a reducing treatment confirmed that the abundant disulfide bonds contributed to the formation of the monomeric structure. In addition, the abundant charged amino acids are suggested to be essential for the function of cp-20k. Characterization of the bacterial recombinant protein has revealed that the protein was adsorbed to limited materials, namely, calcite and metal oxide, in seawater, but not to glass and synthetic polymers. The adsorption isotherm analysis revealed that the adsorption to calcite well fitted the Langmuir model, confirming the protein to be a calcite-specific adsorbent. The cement is always required to couple with the animal’s own calcareous
116
base, which mainly consists of calcite. The opportunity to attach to a calcific exoskeleton, including the peripheral shell of another barnacle, is necessarily frequent because of their gregariousness. Calcific material is one of the most frequent targets for attachment in the molecular system, and the barnacle seems to have a specific protein to couple with this most frequently encountered material. The function of Mrcp-20k is dependent on its structure, and this is the first case of structural dependency confirmed with underwater adhesive proteins, because otherwise, it has been assumed that a flexible backbone structure in a coupling protein would be effective to maximize the surface contact of abundant amino acid residues, especially modified ones (Williams et al. 1989). One possible reason for Mrcp-20k having such a rigid structure may have its origin in molecular evolution. Structural dependency is general among cellular proteins and is a typical molecular feature to display functional specificity. The formation of a structure via increased intramolecular disulfide bonds also is known in extracellular small-protein compounds (Kallio et al. 2007; Iijima et al. 2005; Mandard et al. 1999). Mrcp20k may thus suggest that a structurally dependent coupling agent is a good molecular design for coupling to specific materials. Mrcp-20k is not covalently linked to other proteins in the barnacle cement. The protein is thus retained in the cement complex by molecular interaction in a noncovalent fashion. This may suggest that some surface amino acids on Mrcp20k may be optimized for molecular interaction to another bulk cement protein. Although studies on molecules involved in biotic adhesive have just been started, it is expected that these studies will lead to the finding of unknown functions in underwater attachment and fascinate molecular design to mimic these processes artificially.
Cement Proteins with a Bulk Function: Significance of Noncovalent Molecular Interaction in the Bulk Formation The cement is extremely insoluble; this insoluble nature depends on two major proteins. There are two methods to render M. rosa cement soluble, one is nonproteolytic (Kamino et al. 2000) and the other one is proteolytic (Kamino et al. 1996). The nonproteolytic method involves a heat treatment with relatively high concentrations of a protein denaturant and a reductant, which solubilized more than 90% of the cement. The two major proteins, cp-100k and cp-52k, can only be solubilized by this treatment. Other cement proteins, which are partially soluble in other solutions, could be completely solubilized by destroying the insoluble framework formed by the two major proteins. Treating the
Mar Biotechnol (2008) 10:111–121
two major proteins with the reducing agent alone or the reductant alone did not render them soluble. Moreover, the proteins could not be solubilized by the reductant in combination with a low concentration of the denaturant, indicating that complete denaturation of the proteins was required for their solubilization. The cp-100k and cp-52k proteins are novel from database searches and have low homology to each other. Each of them has a relatively low content of the Cys residue (1% of the total residues in each protein), and there are as yet no data to indicate whether the disulfide bonds in these proteins are intramolecular or intermolecular. Other types of cross-linking did not seem to contribute to the insoluble nature of the proteins, and hydrophobic interaction contributed more to the insoluble nature as would be expected. The non-proteolytic treatment resulted in a complete loss of the adhesive strength. These proteins are therefore the bulk materials in the cement and are thought to tie up other cement proteins and form the framework that provides adhesive strength. Homologous sequences have been found from other species. Among the cement proteins, cp-100k has been the most investigated (Urushida et al. 2005, in preparation), and its homologous cDNA was found in more than seven genera, including Megabalaus sp., Balanus sp., Amphibalanus sp, Fistulobalanus sp., and Semibalanus sp. The expression of the cDNA in Semibalanus cariosus is intriguing, because it has a membranous base, not calcareous. Moreover, Chelonibia sp., which specifically attaches to the shell of the turtle, has provided a cp-100k homologue in the cement (Fujiwara et al. 2007, in preparation). These results suggest that cp-100k is commonly present in the barnacle cement.
Individual and Common Concepts in Sessile Organisms There are many different types of marine sessile invertebrates. The molecular systems for underwater fixation have evolved in the phyla individually. Barnacle cement and the “DOPA-system” of the mussel and tubeworm are compared to identify the concepts for biomolecular material design. The prerequisite subfunctions for an underwater adhesive include displacing the water bound to the foreign substratum by the adhesive, spreading the adhesive over the surface, and coupling to diverse materials. The adhesive proteins corresponding to these surface functions are cp19k and cp-20k for the barnacle and fp-3 and fp-5 (Zhao et al. 2006) for the mussel. The proteins applied to surface functions are hydrophilic and are the smallest of the adhesive proteins in each complex. They use a cocktail of more than two proteins, each with different coupling ability, and one seems to particularly adapt to calcite, which is the most frequently encountered material for sessile organisms.
Mar Biotechnol (2008) 10:111–121
The function of cp-20k is based on commonly used amino acids optimized by the conformation of the protein backbone, as seen in typical cellular proteins which are, in general, capable of specific interactions with other molecules. On the other hand, the fp-5 protein provides a sequence motif incorporating Pho-Ser that can be found in several proteins for calcification. Although the molecular design of adhesive proteins for coupling with a specific material shares some common features with other types of proteins, such as cellular globular proteins and proteins for biomineralization, the molecular design for coupling to diverse foreign materials would be unique to adhesive proteins. The fp-3 protein is posttranslationally modified to HyArg and DOPA, which are never found in other types of proteins. Some studies have shown that the DOPA residue, even alone, has coupling activity towards several metals and minerals (Lee et al. 2006). The degree of modification to fp-3 is the highest among all fps, so it had been presumed that the coupling action of fp-3 was based on modified amino acid residues. A recent study using surface force apparatus has, however, indicated that sequences, including other amino acid residues also contributed to the function (Lin et al. 2007). The cp-19k protein lacks any remarkable post-translational modifications shown in fp-3; therefore, the distribution of common surface amino acids in the protein would be essential for the function, although whether the structure of cp-19k is rigid like a typical globular protein or more flexible as a random polymer (Williams et al. 1989) is not yet known. Comprehensive studies on cp-19k and fp-3 would lead to elucidation of the molecular design required for coupling to a variety of surfaces. There is another view for the ability to couple with diverse foreign materials. Heterogeneity resulting from both genetic variation and the degree of modification is apparent in fp-3 (Papov et al. 1995); the latter can also be seen in fp5 (Waite and Qin 2001). Such wide diversity may be useful to adapt to the various characteristics of substrata, although controlling the variant gene expression corresponding to different types of substrata is unlikely (Floriolli et al. 2000). On the other hand, no meaningful variant has been found in Mrcp-20k and Mrcp-19k, making adaptation based on variants, such as fp-3 unlikely in the barnacle system. Priming and spreading are currently the vaguest functions in underwater attachment. There has been no study on underwater adhesive proteins in which these functions have been directly measured. A possible mechanism was only suggested in tubeworm cement. The formation of a denser concentrated phase (coacervate) by a mixture of oppositely charged polyelectrolytes separated from the more dilute equilibrium phase has been suggested to exhibit spontaneous spreading (Stewart et al. 2004). Although characterizing these functions is a
117
difficult research task, it is likely that the experimental approach to proteins with surface functions will lead to novel functions and molecular design. The adhesive proteins are required to condense the components or to self-assemble after their secretion into the interspace between the organism and substratum, and to cure to achieve durability. The proteins with bulk functions are cp-100k and cp-52k in the barnacle, and fp-2 (Inoue et al. 1995), fp-4 (Zhao and Waite 2006a) and fp-6 (Zhao and Waite 2006b) in the disk of the mussel. The formation in the mussel disk and tubeworm cement of a complex coacervate also has been speculated to be useful to maintain fluidity without being dispersed into the seawater environment. The major bulk proteins in barnacle cement are remarkably hydrophobic, whereas no hydrophobic protein has been found in the mussel byssus or tubeworm cement. Typical globular proteins have the hydrophobic residues crowded into the core of the molecule, thereby forming the molecular structure stably. The abundant hydrophobic residues in barnacle cement seem to be used for intermolecular interaction (Kamino and Nishino 2007, in preparation). It has been found recently that amyloid-like sequences are present in the primary structure of the bulk cement protein (Nakano and Kamino 2007, in preparation). The secondary structures of the amyloid-like peptides, which are part of a bulk-cement protein, are drastically changed and form a selfassembly under such stimuli as pH and salt concentration. The control of hydrophobic interaction via conformational change of the bulk protein is suggested to be a possible mechanism for the self-assembly of barnacle cement. Curing of the byssus depends heavily on intermolecular cross-linking, which has been demonstrated from the extreme resistance to rendering the byssus soluble. The DOPA residue is chemically versatile, as shown by the fact that three types of cross-linking (DOPA-Lys, DOPA-DOPA, and DOPA-Cys) and coordination bonding via metal have been identified. The fp-2 protein is rich in Cys residues, which also has been suggested to contribute to the bulk formation of the disk by intermolecular disulfide bonding. It has not, however, been confirmed whether cystines, which have been detected in the hydrolysate of the protein, form intermolecularly or intra-molecularly. The mussel is required to complete the formation of the byssus quickly in order to achieve full strength (Fig. 3). Because the byssus can, at any moment, sustain serious impact caused by wave action, its formation must be completed quickly to resist repeated stretching from every direction. Each byssus is an individual product. This also is the case with tubeworm cement. The animal attaches an individual particulate to the leading edge of the tube by a kissing action in an entirely external environment. Full attachment is required with extraordinary speed, which requires rapid hardening of the cement. Curing by crosslinking is thought to be likely with these organisms.
118
Mar Biotechnol (2008) 10:111–121
Fig. 5 Possible applications of the barnacle cement
Estimated M.W. (kDa)
In contrast, barnacle cement attaches the enlarged marginal area of the calcareous base to the substratum as the animal grows or repairs the cement layer that has already been formed. Curing may not be urgent with barnacle cement, because the adhesive layer that has already been formed would assist the holdfast to withstand physical impact. Therefore, a new adhesive layer may not need to be quick-curing to acquire full strength. No intermolecular cross-linking has been found yet in barnacle cement. The fact that the cement can be rendered almost soluble by a nonproteolytic treatment (Kamino et al. 2000) supports this. Noncovalent proteininteraction, however, seems to be stronger than would be expected, as can be seen typically in amyloid (Dobson 1999). Amyloid plaque is formed by microspicules, each being a self-assembled form of amyloid-beta peptide by noncovalent molecular interaction. This form of selfassembly is well-known for its extremely resistance against various types of solubilizing agents. Although the author does not always discount the involvement of intermolecular cross-linking in the curing of barnacle cement, the
1 µm
significance of noncovalent molecular interaction should be noted. The different context of each organism might have an influence on the different bio-molecular material design. The durability of the holdfast is not only derived from the strength of molecular interactions. In the cases of the mussel disk and tubeworm cement, the bulk consists of a solid foam structure which is thought to be useful for acquiring the necessary durability of the holdfast. These organisms seem to provide the required mechanism in the microscopic structure. From the macroscopic point of view, the byssus, as soft matter, attaches to foreign substrata, as hard matter. The byssus is composed of macroscopically separate modular structures, from the tip of the disk to the distal end of the thread, in the longitudinal direction at a distance of a few centimeters long. The complete mechanical property also seems to depend on the sequential gradient of stiffness in the longitudinal direction (Waite et al. 2002). In contrast, the barnacle calcareous base is attached directly to the substratum via cement at a distance of a few microns. The barnacle joins two different items of hard material via a layer of soft material: the cement (Fig. 2). The joining characteristics of two different hard particulates by tubeworm cement is similar to that of barnacle cement. It is intriguing how such cement retains the necessary durability to join two hard items of material by an adhesive layer of a few microns in thickness, which has different stiffness. These are architectural aspects rather than the design of a building block. Sessile organisms design their holdfast in several classes: the molecular design of each, the variation of molecular species, and the construction of the building block. Parallel progress in understanding the different
150000 100000 50000 0 0
0.2
0.4
0.6
0.8
1
NaCl [M]
Fig. 6 Salt–concentration-dependent peptide self-assembly designed from cp-20k. The self-assembly has a threshold level close to the salt concentration of seawater
Fig. 7 The cp-19k protein as a multisurface coupling tag. The polymer particles were quickly covered with the bacterial recombinant protein of Mrcp19k fused with green fluorescent protein and emitted fluorescence
Mar Biotechnol (2008) 10:111–121
molecular systems are expected to lead to fruitful material design in the future.
Practical use of the Structural Unit for the Subfunction in Underwater Attachment: Self-assembly and Adsorbent The natural design of a biomolecular material is currently beyond our understanding; elucidating the concepts would lead to novel practical materials. Although underwater attachment will undoubtedly have many applications in several ways, it remains a challenge to achieve. Each subfunction in underwater attachment has a particular impact on material science. The mussel molecular system has been mainly studied to identify the structural units for crosslinking and surface coupling. Peptide chemists have been intrigued by the remarkable characteristics of DOPA as the cross-linking unit (Nagai and Yamamoto, 1989; Yu and Deming 1998) and as a surface coupler, and by that of PhoSer (Yamamoto et al. 2003). Messersmith et al. continue their study on combining the DOPA functionality with PEG (Dalsin et al. 2003) or with nanotechnology (Lee et al. 2007). The use of protein extracts from the mussel foot (Loizou et al. 2006) and of bacterial recombinants (Hwang et al. 2005) also has been promoted in medical applications. Neither evidence for cross-linking nor any remarkable functional unit, such as DOPA, has been found in barnacle cement, and the complex molecular system in barnacle cement may give the impression that learning from the system is difficult. Although there has been a very limited study that was inspired by the barnacle system, developments in material science and its influence on other branches of science and technology might revitalize this inspiration. Peptide- and protein-based materials could be an example of this (Fig. 5). An intermolecular cross-linking system is a double-edged sword, especially in control and turnover as a cellular process, and its biological use is, in general, limited in an extracellular environment. Biomolecules are essentially capable of controlling noncovalent molecular interaction. A self-assembled peptide with noncovalent interaction represents a novel class of practical material. Such material has benefit from its huge molecular diversity, control of a self-assembled nano-structure, controlled biodegradability, easy insertion of a biological motif, and freedom from contamination of viral and prion origins in its production. Molecular self-organization is also setting a trend in supramolecular chemistry (Whitesides and Boncheva 2002), nano-scale technology (Zhao and Zhang 2004), and tissue engineering. The structural unit for selfassembly in a biomolecule should be a target to learn about, and it has recently been shown that an underwater adhesive naturally includes this functional unit.
119
A peptide (Nakano et al. 2007) designed from the repetitive sequence of Mrcp-20k has shown a dependence on salt concentration for self-assembly in noncovalent fashion, and resulted in the formation of a macroscopically observable and handlable membranous material (Fig. 6). This peptide was a soluble monomer in water, and selfassembly was suddenly triggered at a threshold of approximately 0.5 mol/l of NaCl concentration, which is close to that of seawater. AFM and SEM imaging indicated that the peptide assembly had a rod shape of 100 nm in width. The consistent diameter of the peptide suggests its ordered selfassembly. The peptide had no simple amphiphilicity and no beta-sheet structure, putting it in a novel class of selfassembled peptide. Peptide self-assembly systems also have been designed from two other cement proteins. A mixed solution of two different peptides only resulted in uniform self-assembly of each peptide, indicating that each selfassembly system seemed to be constructed for specific peptide interaction. It is suggested that improving the peptide sequence by molecular engineering and optimizing the conditions for self-assembly may lead to the development of practical peptide-based materials that could be used for such an application as the scaffolding for tissue engineering. A protein-based material is another approach for practical use. The cp-19k protein might be useful as a multisurface coupling tag to deposit functional proteins on to a solid material. Such a tool is, for instance, necessary in protein-chip technology (Chen et al. 2002; Hyun et al. 2004; Das et al. 2004), especially in the controlled orientation of an immobilized protein, avoidance of denaturing on the solid material, control of surface density, and prolonged stability under the applied conditions. Several materials, including polymers and metals, could be quickly covered with cp-19k, and this adsorption to each changed the surface contact angle to a common value independent of the material. The cp-19k protein fused with another protein retained almost same ability for multisurface coupling. For instance, green fluorescent protein (GFP) fused with Mrcp-19k quickly covered polymer particles in water, and gave fluorescence to the particles (Fig. 7). A peptide corresponding to the antibody-binding domain of protein A fused with Mrcp-19k resulted in improved functionality of the antibody on to a solid material. The protein-based immobilization tag studied here may be suitable for controlling the orientation and providing prolonged stability on a solid surface. Mrcp-20k also has proved to be a good coupling agent to titanium and hydroxyapatite which are typical materials used for implantation. This protein has many intramolecular disulfide bonds, resulting in a so high stability that the coupling ability was retained even in pure water. This property is particularly intriguing in its application to
120
electronics, because, whereas a biomolecule usually needs some salt and buffering of the pH value, salt is usually incompatible with electronics. The 3-D structure of Mrcp20k has almost completely been determined by NMR spectroscopy. Decoding the optimization of the functional group on the molecular surface may give further momentum to the design of mimetic materials. The combination of valuable and practical natural design of biomolecular materials of various types promises to provide improvements to our future lives.
Conclusions Biotic holdfasts incorporate a diverse array of molecular systems and are a class of biomolecular material that is a treasure chest only just opened. Although it might initially be difficult to use directly, the exciting link between material science and biological science might lead to the use of elements of this natural design in many applications. Elucidating these concepts of nature by a close combination of scrupulous study and mimetic practice with more than a little imagination offers great promise for the future. Acknowledgment The author thank Dr. T. Innes and Prof. J-R. Shen for their critical reading of the manuscript.
References Anderson KE, Waite JH (1998) A major protein precursor of xebra mussel (Dreissena polymorpha) byssus: deduced sequence and significance. Biol Bull 194:150–160 Chen X, Ferrigno R, Yang J, Whitesides GM (2002) Redox properties of cytochrome c adsorbed on self-assembled monolayers: a probe for protein conformation and orientation. Langmuir 18:7009– 7015 Dalsin JL, Bh H, Lee BP, Messersmith PB (2003) Mussel adhesive protein mimetic polymers for the preparation of nonfouling surfaces. J Am Chem Soc 125:4253–4258 Das R, Kiley PJ, Segal M, Norville J, Yu AA (2004) Integration of photosynthetic protein molecular complexes in solid-state electronic devices. Nano Lett 4:1079–1083 Deming TJ (1999) Mussel byssus and biomolecular materials. Curr Opin Chem Biol 3:100–105 Flammang P, Santos R, Haesaerts D (2005) Echinoderm adhesive secretions: from experimental characterization to biotechnological applications. In: Matranga V (ed) Progress in molecular and subcellular biology subseries marine molecular biotechnology. Springer-Verlag, Berlin, pp 201–220 Dobson CM (1999) Protein misfolding, evolution and disease. Trends Biochem Sci 24:329–332 Fletcher GL, Hew CL, Davies PL (2001) Antifreeze proteins of teleost fishes. Annu Rev Physiol 63:359–390 Floriolli R, von Langen J, Waite JH (2000) Marine surfaces and the expression of specific byssal adhesive protein variants in Mytilus. Mar Biotech 2:352–363
Mar Biotechnol (2008) 10:111–121 Humphrey AJ, Finlay JA, Pettitt ME, Stanley MS, Callow JA (2005) Effect of Ellman’s reagent and dithiothreitol on the curing of the spore adhesive glycoprotein of the green alga Ulva. J Adhesion 81:791–803 Hwang DS, Gim Y, Cha HJ (2005) Expression of functional recombinant mussel adhesive protein type 3A in Escherichia coli. Biotechnol Prog 21:965–970 Hwang DS, Gim Y, Yoo HJ, Cha HJ (2007) Practical recombinant hybrid mussel bioadhesive fp-151. Biomaterials 28:3560–3568 Hyun J, Lee WK, Nath N, Chilkoti A, Zauscher S (2004) Capture and release of proteins on the nanoscale by stimuli-responsive elastinlike polypeptide “switches.” J Am Chem Soc 126:7330–7335 Iijima M, Hashimoto T, Matsuda Y, Nagai T, Yamano Y, Ichi T, Osaki T, Kawabata S (2005) Comprehensive sequence analysis of horseshoe crab cuticular proteins and their involvement in transglutaminase-dependent cross-linking. FEBS J 272:4774–4786 Inoue K, Takeuchi Y, Miki D, Odo S (1995) Mussel adhesive plaque protein gene is a novel member of epidermal growth factor-like gene family. J Biol Chem 270:6698–6701 Jia Z, Davies PL (2002) Antifreeze proteins: an unusual receptorligand interaction. Trends Biochem Sci 27:101–106 Kallio JM, Linder MB, Rouvinen J (2007) Crystal structures of hydrophobin HFBII in the presence of detergent implicate the formation of fibrils and monolayer films. J Biol Chem 282:28733– 28739 Kamino K (2001) Novel barnacle underwater adhesive protein is a charged amino acid-rich protein constituted by a Cys-rich repetitive sequence. Biochem J 356:503–507 Kamino K (2006) Barnacle underwater attachment. In: Smith AM, Callow JA (eds) Biological adhesives. Springer-Verlag, Berlin, pp 145–166 Kamino K, Odo S, Maruyama T (1996) Cement proteins of the acorn barnacle, Megabalanus rosa. Biol Bull 190:403–409 Kamino K, Inoue K, Maruyama T, Takamatsu N, Harayama S, Shizuri Y (2000) Barnacle cement proteins. Importance of disulfide bonds in their insolubility. J Biol Chem 275:27360–27365 Lacombe D (1970) A comparative study of the cement glands in some balanid barnacles (cirripedia, balanidae). Biol Bull 139:164–179 Lee H, Scherer NF, Messersmith PB (2006) Single-molecule mechanics of mussel adhesion. Proc Natl Acad Sci USA 103:12999– 13003 Lee H, Lee BP, Messersmith PB (2007) A reversible wet/dry adhesive inspired by mussels and geckos. Nature 448:338–341 Lin Q, Gourdon D, Sun C, Holten-Andersen TH, Waite JH, Israelachvili JN (2007) Adhesion mechanisms of the mussel foot proteins mfp-1 and mfp-3. Proc Natl Acad Sci USA 104:3782– 3786 Loizou E, Weisser JT, Dundigalla A, Porcar L, Schmidt G, Wilker JJ (2006) Structural effects of crosslinking a biopolymer hydrogel derived from marine mussel adhesive protein. Macromol Biosci 6:711–718 Mandard N, Sy D, Maufrais C, Bonmatin JM, Bulet P, Hetru C, Vovelle F (1999) Androctonin, a novel antimicrobial peptide from scorpion Androctonus austrails: solution structure and molecular dynamics in the presence of a lipid monolayer. J Biomol Struct Dyn 17:367–380 McDowell LM, Burzio LA, Waite JH, Schaefer J (1999) Rotational echo double resonance detection of cross-links formed in mussel byssus under high-flow stress. J Biol Chem 274:20293–20295 Mori Y, Urushida Y, Nakano M, Uchiyama S, Kamino K (2007) Calcite-specific coupling protein in barnacle underwater cement. FEBS J 274:6436–6446 Nagai A, Yamamoto H (1989) Insolubilizing studies of water-soluble poly(Lys Tyr) by tyrosinase. Bull Chem Soc Jpn 62:2410–2412
Mar Biotechnol (2008) 10:111–121 Nakano M, Shen J-R, Kamino K (2007) Self-assembling peptide inspired by a barnacle underwater adhesive protein. Biomacromol 8:1830–1835 Naldrett MJ (1993) The importance of sulphur cross-links and hydrophobic interactions in the polymerization of barnacle cement. J Mar Bio Assoc UK 73:689–702 Ninan L, Monahan J, Stroshine RL, Wilker JJ, Shi R (2004) Adhesive strength of marine mussel extracts on porcine skin. Biomaterials 24:4091-4099 Okano K, Shimizu K, Satuito C, Fusetani N (1996) Visualization of cement exocytosis in the cypris cement gland of the barnacle Megabalanus rosa. J Exp Boil 199:2131–2137 Ohkawa K, Nishida A, Yamamoto H, Waite JH (2004) A glycosylated byssal precursor protein from the green mussel Perna viridis with modified DOPA side-chains. Biofouling 20:101–115 Ooka AA, Garrell RL (2000) Surface-enhanced Raman spectroscopy of DOPA-containing peptides related to adhesive protein of marine mussel, Mytulus edulis. Biopolymers 57:92–102 Papov VV, Diamond TV, Biemann K, Waite JH (1995) Hydroxyarginine-containing polyphenolic proteins in the adhesive plaques of the marine mussel Mytilus edulis. J Biol Chem 270:20183–20192 Sagert J, Sun C, Waite JH (2006) Chemical subtleties of mussel and polychaete holdfasts. In: Smith AM, Callow JA (eds) Biological adhesives. Springer-Verlag, Berlin, pp 125–140 Saroyan JR, Lindner E, Dooley CA (1970) Repair and reattachment in the balanidae as related to their cementing mechanism. Biol Bull 139:333–350 Sever MJ, Weisser JT, Monahan J, Srinivasan S, Wilker JJ (2004) Metal-mediated cross-linking in the generation of a marinemussel adhesive. Angew Chem Int Ed 43:448–450 Stevens MJ, Steren RE, Hlady V, Stewart RJ (2007) Multiscale structure of the underwater adhesive of Phragmatopoma californica: a nanostructured latex with a steep microporosity gradient. Langmuir 23:5045–5049 Stewart RJ, Weaver JC, Morse DE, Waite JH (2004) The tube cement of Phragmatopoma californica: a sold foam. J Exp Biol 207:4727–4734 Suzuki R, Mori Y, Kamino K, Yamazaki T (2005) NMR assignment of the barnacle cement protein mrcp-20k. J Biomol NMR 32:257 Taylor SW, Waite JH (1994) trans-2,3-cis-3,4-Dihydroxyproline, a new naturally occurring amino acid, is the sixth residue in the tandemly repeated consensus decapeptides of an adhesive protein from Mytilus edulis. J Am Chem Soc 116:10803–10804 Urushida Y, Nakano M, Matsuda S, Inoue N, Kanai S, Kitamura N, Nishino T, Kamino K (2007) Identification and functional characterization of a novel barnacle cement protein cp-19k. FEBS J 274:4336–4346 Von Byern J, Klepal W (2006) Adhesive mechanisms in cephalopods: a review. Biofouling 22:329–338 Waite JH (1986) Mussel glue from Mytilus californianus Conrad: a comparative study. J Comp Physiol B 156:491–496
121 Waite JH (1987) Nature’s underwater adhesive specialist. Int J Adhes 7:9–14 Waite JH, Tanzer ML (1981) Polyphenolic substance of Mytilus edulis: novel adhesive containing L-Dopa and hydroxyproline. Science 212:352–354 Waite JH, Qin XX (2001) Polyphosphoprotein from the adhesive pads of Mytilus edulis. Biochemistry 40:2887–2893 Waite JH, Vaccaaro E, Sun C, Lucas JM (2002) Elastomeric gradients: a hedge against stress concentration in marine holdfast? Phil Trans R Soc Lond B 357:143–153 Waite JH, Lichtenegger HC, Stucky GD, Hansma P (2004) Exploring molecular and mechanical gradients in structural bioscaffolds. Biochemistry 43:7653–7662 Waite JH, Anderson NH, Jewhurst S, Sun C (2005) Mussel adhesion: finding the tricks worth mimicking. J Adhes 81:297–317 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 Weigemann M, Watermann B (2003) Peculiarities of barnacle adhesive cured on non-stick surfaces. J Adhes Sci Technol 17:1957–1977 Werneke SW, Swann C, Farquharson LA, Hamilton KS, Smith AM (2007) The role of metals in molluscan adhesive gels. J Exp Biol 210:2137–2145 Whitesides GM, Boncheva M (2002) Beyond molecules: selfassembly of mesoscopic and macroscopic components. Proc Natl Acad Sci USA 99:4769–4774 Williams T, Marumo K, Waite JH, Henkens RW (1989) Mussel glue protein has an open conformation. Arch Biochem Biophys 269:415–422 Yamamoto H, Saitoh A, Ohkawa K (2003) Synthesis of sequential polypeptides containing O-phospho-L-serine. Macromol Biosci 3:354–363 Yu M, Deming TJ (1998) Synthetic polypeptide mimics of marine adhesives. Macromol 31:4739–4745 Yu M, Hwang J, Deming TJ (1999) Role of L-3,4-dihydroxyphenylalanine in mussel adhesive proteins. J Am Chem Soc 121:5825–5826 Zhang W, Laursen RA (1998) Structure-function relationships in a type I antifreeze polypeptide. The role of threonine methyl and hydroxyl groups in antifreeze activity. J Biol Chem 273:34806–34812 Zhao H, Waite JH (2006) Proteins in load-bearing junctions: the histidine-rich metal-binding protein of mussel byssus. Biochemistry 45:14223–14231 Zhao H, Waite JH (2006) Linking adhesive and structural proteins in the attachment plaque of Mytilus californianus. J Biol Chem 281:26150–26158 Zhao X, Zhang S (2004) Fabrication of molecular materials using peptide construction motifs. Trends Biotech 22:470–476 Zhao H, Robertson NB, Jewhurst SA, Waite JH (2006) Probing the adhesive footprints of Mytilus californianus byssus. J Biol Chem 281:11090–11096
