**3.3 Amino acid analysis**

The amino acid compositions determined from amino acid analyses of the 6 principal bands are given in Table 1. No data were collected for Band H4 as it appeared to be a doublet (Figure 3). Deamidation during acid hydrolysis means that Asn cannot be distinguished from Asp, nor can Gln be distinguished from Glu; the two pairs are given as Asx and Glx respectively. This prevents an estimation of pI for each of these proteins. Hydroxyproline was not observed in any of the 6 principal bands.

Comparison of the analyses for the various bands did not show signature features for any particular band, and the compositions were broadly similar for all bands. All of the bands had high contents of Gly (7.8-16.8 mol%) and Glx (11.6-16.2 mol%) relative to the average for eukaryotic proteins (6.9 and 9.7 mol%, respectively) (Doolittle, 1986); similarly, DeMoor et al. (2003) observed high Gly contents (16-22 mol%) for the proteins extracted from *H. forskåli.*  As noted above, the SDS-PAGE molecular weight data suggests that it is possible that some bands could be related between the species - H2 and 95 kDa, H6 and 45 kDa and H7 and 33 kDa. Comparison of the amino acid composition data, however, did not show strong similarities. However, it has been suggested (Flammang and Jangoux, 2004), that the protein components present in the adhesives differ between species.


Table 1. Amino acid analysis of individual protein bands after separation by SDS-PAGE. Results are given as Mol %. ND = Not detected. Trp was not determined.

#### **3.4 Adhesion characteristics**

252 Biomaterials – Physics and Chemistry

Previously, Flammang and colleagues (DeMoor et al., 2003) have shown a gel electrophoresis pattern for the tubule print from *H. forskali* samples. In this case, a high background staining was present, and the bands were generally less well defined and more poorly resolved. In some cases the apparent *H. forskali* bands had comparable molecular weights to those observed in the present study. Thus the sharp bands at 95 kDa and 45 kDa may be similar to the H2 (89 kDa) and H6 (44 kDa) bands, while the diffuse bands at 63 kDa and 33 kDa may be similar to the H4 (63 kDa) and the H7 (37 kDa) bands, respectively.

The amino acid compositions determined from amino acid analyses of the 6 principal bands are given in Table 1. No data were collected for Band H4 as it appeared to be a doublet (Figure 3). Deamidation during acid hydrolysis means that Asn cannot be distinguished from Asp, nor can Gln be distinguished from Glu; the two pairs are given as Asx and Glx respectively. This prevents an estimation of pI for each of these proteins. Hydroxyproline

Comparison of the analyses for the various bands did not show signature features for any particular band, and the compositions were broadly similar for all bands. All of the bands had high contents of Gly (7.8-16.8 mol%) and Glx (11.6-16.2 mol%) relative to the average for eukaryotic proteins (6.9 and 9.7 mol%, respectively) (Doolittle, 1986); similarly, DeMoor et al. (2003) observed high Gly contents (16-22 mol%) for the proteins extracted from *H. forskåli.*  As noted above, the SDS-PAGE molecular weight data suggests that it is possible that some bands could be related between the species - H2 and 95 kDa, H6 and 45 kDa and H7 and 33 kDa. Comparison of the amino acid composition data, however, did not show strong similarities. However, it has been suggested (Flammang and Jangoux, 2004), that the protein

 H2 H3 H6 H7 H8 H9 Asx 9.7 8.7 12.1 13.1 9.7 15.0 Ser 9.8 11.6 7.0 6.5 10.2 8.0 Glx 16.2 13.3 13.2 12.7 15.0 11.6 Gly 16.1 16.8 7.8 8.1 13.6 10.5 His 1.1 1.0 1.3 ND 0.7 1.3 Arg 3.6 5.2 4.9 4.1 5.5 4.4 Thr 3.9 5.2 5.1 5.0 4.9 4.6 Ala 7.8 7.0 9.2 9.4 6.8 8.0 Pro 3.2 4.9 3.9 3.2 6.3 3.1 Tyr 1.8 2.5 2.4 1.9 3.0 2.4 Val 4.6 5.5 6.4 6.4 5.5 5.7 Met ND 0.4 3.2 3.8 0.4 5.0 Lys 8.8 4.9 8.3 9.4 5.0 7.0 Ile 4.1 3.8 5.6 5.3 3.5 4.8 Leu 7.0 7.0 7.3 8.3 7.5 5.8 Phe 2.4 2.4 2.5 3.0 2.5 2.6 Table 1. Amino acid analysis of individual protein bands after separation by SDS-PAGE.

Results are given as Mol %. ND = Not detected. Trp was not determined.

**3.3 Amino acid analysis**

was not observed in any of the 6 principal bands.

components present in the adhesives differ between species.

In the present study, a 90 Degree Peel Test was used to evaluate the adhesion of freshly expelled Cuvierian tubules. This method was chosen as we had encountered problems when tensile testing the *H. dofleinii* tubules following the approach used by Flammang and colleagues (Flammang et al., 2002). Specifically, when *H. dofleinii* tubules were sandwiched between two materials to which there was good adhesion, e.g. glass or metals, testing could lead to strength values which reflected the structural failure of the Cuvierian tubule rather than the failure of the adhesive, especially if some drying had occurred (data not shown). This method would only allow the determination of a minimum value for the adhesive strength as the latter exceeded the break strength of the tubule material itself.

The present test was suitable for rapidly examining numerous, freshly expelled samples, thus allowing ready comparison between the effects of various treatment solutions. The various treatments (i.e., incubations of tubules in the appropriate wash solutions) prior to adhesive testing were rapid (1 min) as it appeared that the adhesion could decline if tubules were left soaking for lengthy periods (data not shown). With *H. forskåli*, a lag period of about 60 min at 16 C was recorded before adhesion started to decline, decreasing to about 15 min at 26 C (Müller et al., 1972). In another study (Flammang et al., 2002) a longer lag phase was observed, and an initial increase in adhesive strength was reported. Yet others have reported adhesive strength to fall after 20 min (Zahn et al., 1973). The present approach, therefore, used short incubations in order to minimise time-based variations and to mimic the timescale over which tubules would be required to act in the natural environment.

Previous studies (Flammang et al., 2002) have shown that a compressive force of 2–10 N during adhesion led to a 6- to 8-fold increase in the resulting bond strength. In the present case, no compressive load was added so as to better simulate the natural process of ensnaring a predator.

Tubule widths showed little variation between individual samples, the average size being 4.0 mm. Tubules that were not fully expelled, and which therefore had a lesser diameter, were discarded. The observed width is larger than that found for *H. forskåli* (Flammang et al., 2002; Zahn et al., 1973) and *H. leucospilota* (Flammang et al., 2002), the species previously studied in detail, and also larger than for *H. impatiens* and *H. maculosa;* these other species generally have tubule diameters of 1–2 mm (Flammang et al., 2002). Although there are many potential tubules within the body cavity (Figure 1A) *H. dofleinii* expels only a few, typically 8 - 12 for organisms stimulated in the holding tanks compared with the more numerous thin tubules expelled by *H. leucospilota* or *H. forskali* (Flammang et al., 2002).

Adhesive strength was also found to vary when different substrata were examined, all after washing the tubules in 3.5% NaCl 10 mM Na/PO4, pH 7.6. There was a trend for strongest adhesion to be observed with hydrophilic substrata, glass and aluminium (Table 2). Adhesion to polycarbonate, PMMA, and PTFE was very poor; indeed, for PMMA and PTFE, the load required for peel was barely more than the weight of the 50 mm of tubule overhang. Intermediate adhesion values were observed with polyvinyl chloride and crab chitin surfaces (Table 2). The chitin samples were unusual in having a textured surface rather than a smooth one. Previously, Zahn, Flammang and colleagues had shown strong adhesion to hydrophilic surfaces such as glass and stainless steel, and poor adhesion to hydrophobic ones such as paraffin wax, polystyrene and polyethylene (Zhan et al., 1973; Flammang et al, 2002). In general our results are consistent with this trend: the best adhesion was observed with glass whilst the poorest was observed with PTFE.

Solution Force/width S.D. n

Biomimetic Materials as Potential Medical Adhesives – Composition and Adhesive

reversed by washing (Zahn et al., 1973).

substratum for tubule adhesion.

Properties of the Material Coating the Cuvierian Tubules Expelled by *Holothuria dofleinii* 255

clear whether this is due to the multiple carboxyl groups of this salt, to its strong metal ion chelating capability, or to some other property. However, certain other marine adhesives, such as that from *Mytilus*, do require metal activity (Hwang et al., 2010). A previous study which tested 15 different amino acids at 0.5% w/v solutions on adhesion by *H. forskåli* tubules (Müller et al., 1972) showed that most had little, if any, effect. The exceptions were the hydrophobic amino acids leucine (20% loss) and phenylalanine (57% loss). For phenylalanine, the loss was slow to develop (taking several minutes) and could not be

Solution Force/width S.D. n

3.5% NaCl > 0.050 >8 Tris/chloride 0.050 0.008 6 Sodium formate 0.047 0.012 6 Sodium oxalate 0.046 0.011 8 Ammonium chloride 0.036 0.008 8 Sodium citrate 0.035 0.009 7 Sodium acetate 0.030 0.006 8 Sodium EDTA <0.003 8

Table 3. Comparison of the effects of different salt solutions on adhesion of Cuvierian tubules onto glass. All salts were 50mM in 3.5% NaCl, 10 mM sodium phosphate, pH 7.6

The effect of pH on the adhesive strength of the *H. dofleinii* adhesive showed that for Tris/chloride buffer, the best observed strength of those tested was at pH 7.6, and that the observed strength decreased at both lower and higher pH values (Figure 6). For citrate and acetate buffers, adhesive strength declined progressively as the pH was lowered from pH 7.6, with little adhesion remaining at pH 5.0 (Figure 6). A loss of adhesive strength at acidic pH values was also observed by Müller et al. (1972), who used paraffin wax as a (poor)

Fig. 6. The effect of different washing solutions on the adhesiveness of *H. dofleinii* Cuvierian tubules for glass. The effect of changes in pH, where ■ indicates acetate buffers, ♦ indicates

citrate buffers and ● indicates Tris/Chloride buffers.

(N/mm)


Table 2. Force required to peel Cuvierian tubules off various substrata to determine adhesive strength.

Fig. 5. The effect of different washing solutions on the adhesiveness of *H. dofleinii* Cuvierian tubules for glass. The effect of NaCl concentration; where ▲ indicates conditions where the force per unit width exceeded 0.05 N/mm. SW is natural sea water.

Adhesive strength decreased with decreasing NaCl concentration (Figure 5). At ≥3% NaCl the adhesion exceeded 0.05 N/mm. Reducing the NaCl concentration incrementally from 2.5% to 1.0% NaCl led to a steady decline in adhesive strength (Figure 5). The adhesive strength at 1% NaCl, which is comparable in concentration to physiological saline, was significantly weaker than in 3.5% NaCl simulated seawater. This is consistent with the previous observations on *H. forskåli* tubules (Flammang et al., 2002). It suggests that hydrophobic interactions may be important in the adhesive mechanism.

The effects on tubule-glass adhesion of other chloride or sodium salts (50 mM) (Table 3) showed that in all cases there was a loss of adhesive strength. For chloride salts, the loss was smaller when Tris rather than ammonium was the cation (Table 3). The other salts examined were all sodium salts of carboxylic acids, for whose action no simple mechanism could be proposed. Thus while formate (a monocarboxylate) and oxalate (a dicarboxylate) both showed similar adhesion, that observed with acetate (another monocaboxylate) was ~35% below the value observed for formate. However, the values presented in Table 3 show only a trend as the errors in measurement are such that the different systems are not necessarily distinguishable. Supplementation with EDTA (a tetracarboxylate) was the most effective at disrupting bond strength, and essentially led to complete loss of adhesion (Table 3). It is not

 (N/mm) Glass > 0.050 - 10 Aluminium > 0.050 - 6 Polyvinyl chloride 0.024 0.006 11 Chitin 0.021 0.005 8 Polycarbonate 0.010 0.002 8 PMMA 0.009 0.003 7 PTFE 0.008 0.002 6

Table 2. Force required to peel Cuvierian tubules off various substrata to determine

Fig. 5. The effect of different washing solutions on the adhesiveness of *H. dofleinii* Cuvierian tubules for glass. The effect of NaCl concentration; where ▲ indicates conditions where the

Adhesive strength decreased with decreasing NaCl concentration (Figure 5). At ≥3% NaCl the adhesion exceeded 0.05 N/mm. Reducing the NaCl concentration incrementally from 2.5% to 1.0% NaCl led to a steady decline in adhesive strength (Figure 5). The adhesive strength at 1% NaCl, which is comparable in concentration to physiological saline, was significantly weaker than in 3.5% NaCl simulated seawater. This is consistent with the previous observations on *H. forskåli* tubules (Flammang et al., 2002). It suggests that

The effects on tubule-glass adhesion of other chloride or sodium salts (50 mM) (Table 3) showed that in all cases there was a loss of adhesive strength. For chloride salts, the loss was smaller when Tris rather than ammonium was the cation (Table 3). The other salts examined were all sodium salts of carboxylic acids, for whose action no simple mechanism could be proposed. Thus while formate (a monocarboxylate) and oxalate (a dicarboxylate) both showed similar adhesion, that observed with acetate (another monocaboxylate) was ~35% below the value observed for formate. However, the values presented in Table 3 show only a trend as the errors in measurement are such that the different systems are not necessarily distinguishable. Supplementation with EDTA (a tetracarboxylate) was the most effective at disrupting bond strength, and essentially led to complete loss of adhesion (Table 3). It is not

force per unit width exceeded 0.05 N/mm. SW is natural sea water.

hydrophobic interactions may be important in the adhesive mechanism.

adhesive strength.

clear whether this is due to the multiple carboxyl groups of this salt, to its strong metal ion chelating capability, or to some other property. However, certain other marine adhesives, such as that from *Mytilus*, do require metal activity (Hwang et al., 2010). A previous study which tested 15 different amino acids at 0.5% w/v solutions on adhesion by *H. forskåli* tubules (Müller et al., 1972) showed that most had little, if any, effect. The exceptions were the hydrophobic amino acids leucine (20% loss) and phenylalanine (57% loss). For phenylalanine, the loss was slow to develop (taking several minutes) and could not be reversed by washing (Zahn et al., 1973).


Table 3. Comparison of the effects of different salt solutions on adhesion of Cuvierian tubules onto glass. All salts were 50mM in 3.5% NaCl, 10 mM sodium phosphate, pH 7.6

The effect of pH on the adhesive strength of the *H. dofleinii* adhesive showed that for Tris/chloride buffer, the best observed strength of those tested was at pH 7.6, and that the observed strength decreased at both lower and higher pH values (Figure 6). For citrate and acetate buffers, adhesive strength declined progressively as the pH was lowered from pH 7.6, with little adhesion remaining at pH 5.0 (Figure 6). A loss of adhesive strength at acidic pH values was also observed by Müller et al. (1972), who used paraffin wax as a (poor) substratum for tubule adhesion.

Fig. 6. The effect of different washing solutions on the adhesiveness of *H. dofleinii* Cuvierian tubules for glass. The effect of changes in pH, where ■ indicates acetate buffers, ♦ indicates citrate buffers and ● indicates Tris/Chloride buffers.

Biomimetic Materials as Potential Medical Adhesives – Composition and Adhesive

these bands do not conform to such a regular series of increases.

concentrations are present.

**5. Acknowledgments**

**6. References** 

Wealth from Oceans National Research Flagship.

Vol.159, pp. 39-47, ISSN 1064-3745

(February), pp. 25-34, ISSN 0021-9673

45-57, ISSN 1436-2228

9932

Properties of the Material Coating the Cuvierian Tubules Expelled by *Holothuria dofleinii* 257

on the tubule surface and release their contents on contact with a surface (VandenSpiegel and Jangoux, 1987) leading (in whole or in part) to the observed adhesion. Histology has shown that these granules contain protein and lipid, but lack polysaccharide (VandenSpiegel and Jangoux, 1987). Biochemical studies have indicated that the granules contain a protein of around 10 kDa, and it has been suggested that polymers of this protein account for the higher molecular weight proteins that are seen in the adhesive prints (Flammang and Jangoux, 2004), but this seems highly unlikely in *H. dofleinii* as the protein bands are very well resolved by gel electrophoresis and the calculated molecular weights of

Our present study emphasises that the adhesives of natural systems are optimised for the specific environments in which they have evolved, such as the present marine environment. An analogue intended for medical use would need to be optimised to yield maximum adhesion in the physiological conditions that prevail in mammalian tissues. In the present case, the adhesion works better at higher NaCl concentrations that found in medical applications so understanding more about the mechanism and the protein structures and properties will be needed in order to adapt this system for applications where lower NaCl

We wish to thank Nicole Murphy for assistance with Holothurian collection and Dr Anita Hill for helpful discussions. This study was facilitated by access to the Australian Proteome Analysis Facility supported under the Australian Government's National Collaborative Research Infrastructure Strategy (NCRIS). The project received support from the CSIRO

Cohen, S.A. (2001). Amino acid analysis using precolumn derivatisation with 6-

Cohen, S.A. & DeAntonis, K.M. (1994). Applications of amino acid analysis derivatisation

DeMoor, S.; Waite, J.H.; Jangoux, M. & Flammang, P. (2003). Characterization of the

Dimas, D.A.; Dallas, P.P.; Rekkas, D.D. & Choulis, N.H. (2000) Effect of several factors on

Doolittle, R.F. (1986). *Of URFs and ORFs: a primer on how to analyse derived amino acid sequences,* University Science Books, ISBN 0-935702-54-7, Mill Valley, CA, USA

Flammang, P. & Jangoux, M. (2004). *Final Report ONR Grant* N00014-99-1-0853

aminoquinolyl-*N*-hydroxysuccinimidyl carbamate. *Methods in Molecular Biology*,

with 6-aminoquinolyl-*N*-hydroxysuccinimidyl carbamate: Analysis of feed grains, intravenous solutions and glycoproteins. *Journal of Chromatography,* Vol.661, No.1-2,

adhesive from cuvierian tubules of the sea cucumber *Holothuria forskali* (Echinodermata, Holothuroidea). *Marine Biotechnology,* Vol.5, No.1, (January), pp.

the mechanical properties of pressure-sensitive adhesives used in transdermal therapeutic systems. *AAPS PharmSciTech*, Vol.1, No.2, (June), pp. 80-87, ISSN 1530-

Some reports suggest that proteins may play an important role in the adhesion of Cuvierian tubules from *H. forskåli* (DeMoor et al., 2003; Flammang & Jangoux, 2004; Müller et al., 1972). For example, the adhesive residue left when tubules are peeled from a surface consists mainly of protein (DeMoor et al., 2003), and the treatment of tubules with proteases causes loss of adhesion (Müller et al., 1972). However, it has been reported that the proteins most likely differ between species (Flamman & Jangoux, 2004), making further comparative biochemical surveys important for elucidating the mechanism.

The effect of urea on tubule-glass adhesion showed that bond strength decreased progressively with increasing urea concentration until it was completely lost at 2 M urea (Figure 7). However, if tubules that had been incubated for about 60 sec in 2 M urea were then rinsed for about 60 sec in simulated sea water, some adhesion was restored, although the extent was rather variable. Urea disrupts hydrogen bonding, and its effect on adhesion may reflect some partially reversible protein unfolding (Zahn et al., 1972). The rapidity and partial reversibility of the effect indicates that there is not a complete urea-mediated release of proteins from the tubule surface.

Fig. 7. The effect of different washing solutions on the adhesiveness of *H. dofleinii* Cuvierian tubules for glass. (C) The effect of urea concentration, where ▲ indicates conditions where the force per unit width exceeded 0.05 N/mm.

#### **4. Conclusion**

The distinct features of the Cuvierian tubule adhesion mechanism, especially its rapid action under water, are unique. If the mechanism can be understood, then it may be possible to design a synthetic system with analogous properties. An adhesive that provided instant grip in an aqueous environment would be very valuable, especially in medical applications, as the majority of existing adhesives bind well only to dry surfaces.

It appears that although the tubules of *H.dofleinii* are distinct from those of other species, especially in their size and the number expelled, the adhesive properties of these Cuvierian tubules (including preferences for hydrophilic surfaces, pH optima, etc.) are similar to those found in other species, even if mechanistic details may differ between species as has been proposed previously (Flammang and Jangoux, 2004). It is thought that during expulsion and tubule elongation, granular cells that are internal in the pre-release tubule become located on the tubule surface and release their contents on contact with a surface (VandenSpiegel and Jangoux, 1987) leading (in whole or in part) to the observed adhesion. Histology has shown that these granules contain protein and lipid, but lack polysaccharide (VandenSpiegel and Jangoux, 1987). Biochemical studies have indicated that the granules contain a protein of around 10 kDa, and it has been suggested that polymers of this protein account for the higher molecular weight proteins that are seen in the adhesive prints (Flammang and Jangoux, 2004), but this seems highly unlikely in *H. dofleinii* as the protein bands are very well resolved by gel electrophoresis and the calculated molecular weights of these bands do not conform to such a regular series of increases.

Our present study emphasises that the adhesives of natural systems are optimised for the specific environments in which they have evolved, such as the present marine environment. An analogue intended for medical use would need to be optimised to yield maximum adhesion in the physiological conditions that prevail in mammalian tissues. In the present case, the adhesion works better at higher NaCl concentrations that found in medical applications so understanding more about the mechanism and the protein structures and properties will be needed in order to adapt this system for applications where lower NaCl concentrations are present.
