**3.1 Tubule structure and microscopy**

Dissection of euthanised *H. dofleinii* showed that the body cavity contained a large number, several dozen, Cuvierian tubules in their compressed form (Figure 1A). In their compressed form the tubules were not sticky, but they rapidly became sticky on mechanical extension, even from the dead animals. The compressed, individual tubules showed a corrugated and folded surface (Figure 1B), which would allow extension when required, like a concertina bellows. In cross-section (Figure 1C) a three-lobed channel could be seen that would allow fluid insertion for expansion of the tubules. When expelled and fully extended *in vivo*, these tubules became instantly sticky and changed from the 25-35 mm compressed length up to around 350-400 mm. The fully extended tubules (Figure 1D) were typically about 4mm flat width when fully inflated, and still showed some patterning from the folding that was present in the compressed state. Individual animals contained several dozen non-inflated tubules, but when the animals were stimulated only a small number were expelled, normally around 8-12.

Fig. 1. Cuvierian tubules for *H. dofleinii*. (A, B, C) After dissection of a euthanised animal, showing, (A) the total mass of tubules, (B) the tips of compressed tubules, and (C) the crosssection of compressed tubules. (D) The surface of an *in vivo* expelled tubule. Bar (A) = 10 mm, Bars (B, C, D) = 1 mm.

Biomimetic Materials as Potential Medical Adhesives – Composition and Adhesive

Samples with little wall material typically had little, if any insoluble material.

Fig. 4. SDS-PAGE of the reduced proteins from the surface of a freshly expelled Cuvierian tubule from *H. dofleinii*. The gel was stained with Coomassie Blue R-250. Key bands are labelled and their estimated molecular weights, interpolated from a standard curve using

globular protein standards (BioRad), are given.

**3.2 Gel electrophoresis analysis**

Properties of the Material Coating the Cuvierian Tubules Expelled by *Holothuria dofleinii* 251

The adhesion print isolated from glass showed a range of proteins of well-defined molecular weights when analysed under reducing conditions by SDS-PAGE (Figure 4). A similar pattern, except as noted below, was observed from >10 separate samples from different animals. Seven bands, H2 at 89 kDa, H3 at 70 kDa, H4 at 61 kDa, H6 at 44 kDa, H7 at 37 kDa, H8 26 kDa and H9 at 17 kDa, were consistently present in all tubule samples that were examined (n > 12). Band H4 has been examined as a single entity, but in some gels (Figure 4) it appeared that it may comprise 2 components. For all other bands, although each appeared to be a single component, it is also possible that more than one component could be present, migrating similarly. In many samples, but not all, an additional band, H5 at 53 kDa, was present. As it was not consistently present it was assumed that it may not be a key component of the adhesive system and hence was not examined further. In a few samples, an additional band H1 was observed at 150-170 kDa. This band seemed to be more prevalent in samples where tubule fragments were present in the adhesion print, and could possibly be related to the collagen that is the main structural component of the tubule (Watson and Silvester, 1959). The collected material sometimes contained a proportion of material that remained insoluble in the sample buffer. Examination of the adhesive prints under a microscope suggested that the samples that subsequently contained more insoluble material also contained more fibrous material that could be from the collagenous wall of the tubule.

The surfaces of naturally extended tubules showed strong binding of FITC-DSA (Figure 2A) and FITC-LEA lectins to the tubule surface, with a series of bands that resemble the original folding of the un-extended tubule. FITC-ConA lectin binding was weak, suggesting that the DSA and LEA binding was specific for carbohydrate or glycosylated protein on the surface, rather than non-specific binding. These 2 lectins recognise very similar carbohydrate entities; (N-acetyl glucosamine)2 by DSA and (N-acetyl glucosamine)3 by LEA, which are distinct from the -mannose or -glucose recognised by ConA. SEM of the surface of expelled tubules (Figure 2B) showed that the surface had fibrous-like structures, suggesting aggregates of the adhesive material overlaying a further fibrous, collagenous layer of the wall of the tubule.

Fig. 2. Cuvierian tubules from *H. dofleinii*. (A) Fluorescence microscopy of FITC-labelled DSA bound to expelled Cuvierian tubules. Bar = 0.5 mm. (B) SEM of expelled Cuvierian tubules. Bar = 2.5 m

When freshly expelled tubules were briefly immersed in 3.5% NaCl containing Bio-Gel P2 particles, and then washed in 3.5% NaCl, particles bound to the tubule (Figure 3). These data showed that the surface of the tubule was generally adhesive, and there was no specific localisation of the particles, for example to the fibrous-like patches seen by SEM.

Fig. 3. Cuvierian tubules from *H. dofleinii*. Adhesion of Bio-Gel P2 beads to freshly expelled Cuvierian tubules from *H. dofleinii*. Bar = 0.5 mm

#### **3.2 Gel electrophoresis analysis**

250 Biomaterials – Physics and Chemistry

The surfaces of naturally extended tubules showed strong binding of FITC-DSA (Figure 2A) and FITC-LEA lectins to the tubule surface, with a series of bands that resemble the original folding of the un-extended tubule. FITC-ConA lectin binding was weak, suggesting that the DSA and LEA binding was specific for carbohydrate or glycosylated protein on the surface, rather than non-specific binding. These 2 lectins recognise very similar carbohydrate entities; (N-acetyl glucosamine)2 by DSA and (N-acetyl glucosamine)3 by LEA, which are distinct from the -mannose or -glucose recognised by ConA. SEM of the surface of expelled tubules (Figure 2B) showed that the surface had fibrous-like structures, suggesting aggregates of the adhesive material overlaying a further fibrous, collagenous layer of the

Fig. 2. Cuvierian tubules from *H. dofleinii*. (A) Fluorescence microscopy of FITC-labelled DSA bound to expelled Cuvierian tubules. Bar = 0.5 mm. (B) SEM of expelled Cuvierian

localisation of the particles, for example to the fibrous-like patches seen by SEM.

When freshly expelled tubules were briefly immersed in 3.5% NaCl containing Bio-Gel P2 particles, and then washed in 3.5% NaCl, particles bound to the tubule (Figure 3). These data showed that the surface of the tubule was generally adhesive, and there was no specific

Fig. 3. Cuvierian tubules from *H. dofleinii*. Adhesion of Bio-Gel P2 beads to freshly expelled

Cuvierian tubules from *H. dofleinii*. Bar = 0.5 mm

wall of the tubule.

tubules. Bar = 2.5 m

The adhesion print isolated from glass showed a range of proteins of well-defined molecular weights when analysed under reducing conditions by SDS-PAGE (Figure 4). A similar pattern, except as noted below, was observed from >10 separate samples from different animals. Seven bands, H2 at 89 kDa, H3 at 70 kDa, H4 at 61 kDa, H6 at 44 kDa, H7 at 37 kDa, H8 26 kDa and H9 at 17 kDa, were consistently present in all tubule samples that were examined (n > 12). Band H4 has been examined as a single entity, but in some gels (Figure 4) it appeared that it may comprise 2 components. For all other bands, although each appeared to be a single component, it is also possible that more than one component could be present, migrating similarly. In many samples, but not all, an additional band, H5 at 53 kDa, was present. As it was not consistently present it was assumed that it may not be a key component of the adhesive system and hence was not examined further. In a few samples, an additional band H1 was observed at 150-170 kDa. This band seemed to be more prevalent in samples where tubule fragments were present in the adhesion print, and could possibly be related to the collagen that is the main structural component of the tubule (Watson and Silvester, 1959). The collected material sometimes contained a proportion of material that remained insoluble in the sample buffer. Examination of the adhesive prints under a microscope suggested that the samples that subsequently contained more insoluble material also contained more fibrous material that could be from the collagenous wall of the tubule. Samples with little wall material typically had little, if any insoluble material.


Fig. 4. SDS-PAGE of the reduced proteins from the surface of a freshly expelled Cuvierian tubule from *H. dofleinii*. The gel was stained with Coomassie Blue R-250. Key bands are labelled and their estimated molecular weights, interpolated from a standard curve using globular protein standards (BioRad), are given.

Biomimetic Materials as Potential Medical Adhesives – Composition and Adhesive

strength as the latter exceeded the break strength of the tubule material itself.

**3.4 Adhesion characteristics**

ensnaring a predator.

Properties of the Material Coating the Cuvierian Tubules Expelled by *Holothuria dofleinii* 253

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

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

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

Solution Force/width S.D. n

was observed with glass whilst the poorest was observed with PTFE.

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.
