Cell-derived stiffening and plasticizing factors in sea cucumber (Cucumaria frondosa) dermis

Echinoderms utilize mutable connective tissues in physiological responses to a variety of environmental and mechanical cues (Motokawa, 1984, 1988; Wilkie, 1984, 1988). The ubiquitous development of these tissues throughout the phylum suggests that they are of fundamental importance to the life history of echinoderms. Well-studied examples of mutable connective tissues include sea urchin spine ligaments (Takahashi, 1967a,b; Maeda, 1978; Hidaka, 1983; Hidaka and Takahashi, 1983; Diab and Gilly, 1984; Shadwick and Pollock, 1988; Morales et al., 1989), holothurian dermis (Eylers, 1982; Hayashi and Motokawa, 1986; Motokawa, 1982a,b, 1987; Trotter and Koob, 1995) and intervertebral ligaments in brittlestar arms (Wilkie, 1978a,b, 1979; Wilkie and Emson, 1987). Other echinoderm tissues with attributes of mutable connective tissues that have been less well studied are crinoid stalks (Holland and Grimmer, 1981; Wilkie, 1983), oral arm plate ligaments in brittlestars (Wilkie, 1992) and the ligaments connecting the teeth to the jaw in sea urchins (Birenheide and Motokawa, 1996; Birenheide et al., 1996). These tissues have in common the ability to change length and, at any length, to vary their tensile properties on a physiological time scale.

The capacity reversibly to alter tissue compliance derives principally from the presence of discontinuous collagen fibrils which, in compliant tissues, are able to slide past one another. Phase contrast and electron microscopic analyses of native fibrils isolated from starfish body wall, from sea urchin spine ligaments and from sea cucumber dermis have shown that, while they vary in overall dimensions, their shapes are similar (Matsumura, 1973, 1974; Trotter and Koob, 1989; Trotter et al., 1994). Recent analyses of isolated intact, native collagen fibrils from C. frondosa dermis have established that they are molecularly bipolar (Thurmond and Trotter, 1994), spindle-shaped with tapered tips (Trotter et al., 1994) and composites assembled from collagen molecules with three identical alpha chains and a covalently bound chondroitin-sulfate-based glycosaminoglycan (Trotter et al., 1995). The molecular mechanisms responsible for mediating interactions between fibrils in vivo are poorly defined, but at least one matrix macromolecule, stiparin, has been shown to cause isolated fibrils to aggregate (Trotter et al., 1996).

Several distinct hypotheses have been proffered to account for tissue mutability in echinoderm catch connective tissues. Ca²⁺-activated transglutaminase crosslinking of collagen fibrils has been suggested to play a role in the stiffening response (catch) of sea urchin spine ligaments, since the inhibitors of transglutaminase activity, cadaverine and putrescine, prevented the Ca²⁺-mediated setting of catch (Diab and Gilly, 1984). However, it should be noted that both these agents are polyamines and could interact with matrix polyanions such as glycosaminoglycans as well as transglutaminase. While transglutaminase crosslinks have never been isolated from echinoderm collagens, ε-(γ-glutamy)lysine is present in the microfibrils of C. frondosa dermis (Thurmond et al., 1997) and transglutaminase activity has been detected in the dermis (J. M. Bowness, T. J. Koob and J. A. Trotter, unpublished data). Proteases have been implicated in autotomy of the dermis in the sea cucumber Stichopus badionotus (Junqueira et al., 1980), but whether such enzymes act during reversible alterations in dermis mechanical properties is not known. The model favored by most reports implicates Ca²⁺ as the principal regulator of compliance (Wilkie, 1996).

Previous studies on sea urchin spine ligaments (Hidaka, 1983; Diab and Gilly, 1984; Shadwick and Pollock, 1988), holothurian dermis (Motokawa, 1984, 1987, 1988, 1994; Trotter and Koob, 1995; Trotter and Chino, 1997) and ophiuroid intervertebral ligaments (Wilkie, 1984, 1988) have demonstrated that experimental alteration of extracellular Ca²⁺ levels modulates the stiffness of living tissues. When specimens are challenged with Ca²⁺-free sea water, they become compliant. Following restoration of normal Ca²⁺ concentrations, specimens become stiff. This response has generally been interpreted as an indication that Ca²⁺ directly affects the extracellular matrix and, moreover, that the means by which cells normally regulate tissue compliance is via secretion and sequestration of Ca²⁺ into and out of the extracellular matrix. It has been inferred that resident cells containing secretory granules are somehow involved in regulating extracellular Ca²⁺ concentration, thereby mediating alterations in tissue compliance (Motokawa, 1984, 1987; Wilkie, 1988, 1996; Matsuno and Motokawa, 1992).

Recent experiments on both C. frondosa and Actinopyga agassizi, however, are inconsistent with this direct Ca²⁺ hypothesis (Trotter and Koob, 1995; Trotter et al., 1997; Trotter and Chino, 1997). Like all mutable connective tissues so far investigated, the inner dermis of C. frondosa and A. agassizi becomes compliant in the absence of Ca²⁺ and becomes stiff in normal Ca²⁺ levels. The stiffening response to Ca²⁺ is indirect and is probably due to a secretory event that depends on the influx of Ca²⁺ through plasmalemmal channels. Indirect evidence has suggested that one or more protein stiffening factors are contained in at least some of the granules of the resident cells and are released either by exocytosis from living cells or by membrane lysis (Trotter and Koob, 1995; Trotter and Chino, 1997). The present report describes the partial purification of the stiffening agent from C. frondosa. Further evidence is presented for a cell-derived organic plasticizing molecule in C. frondosa dermis, which also has been isolated and partially purified. Portions of this work have been published previously in abstract form (Koob-Emunds et al., 1996, 1998).

Cucumaria frondosa (Gunner, 1767) were trawled from the near-shore waters of Frenchman Bay between the months of September and November 1995 and 1996 and maintained in flow-through sea water tanks at the Mount Desert Island Biological Laboratory, Salsbury Cove, Maine, USA. All specimens were prepared from the dermis of the two ventral interambulacra, which lack podia. The dermis of C. frondosa is composed of an outer pigmented layer, accounting for approximately one-third of the dermal thickness, and a dense, white inner layer. Although the relative proportions of these two layers vary among animals, the demarcation is easily discerned. The white inner dermis was used for isolation of the stiffening factor. The pigmented outer layer was used for isolation of the plasticizing factor. Bending tests were carried out exclusively on the white inner dermis.

The dermis was excised rapidly from the animal using a razor blade, and the attached musculature was stripped from the inner wall, which was then scraped with a spatula, to remove all traces of muscle, and briefly blotted to eliminate surface fluids. Specimens were held on ice during subsequent handling. Dermis segments were dissected in one of two ways depending on the intended use of the specimen. To extract macromolecules, segments were minced using razor blades to produce approximately 1 mm³ pieces. For bending tests, uniform test specimens were cut using single-edged razor blades the safety shields of which were screwed together, either directly or with an intervening metal shim. This cutting apparatus was used to produce equivalent specimens from the white inner dermis that were 3 cm long, 0.85 mm thick and 1.7 mm wide. The long axis of each specimen was parallel to that of the animal. The 0.85 mm thick side was in its radial dimension, and the 1.7 mm side was in its circumferential dimension. For all bending tests, specimens from at least three animals were assigned randomly to the treatment groups.

Gravity bending tests were performed on the white inner dermis, essentially as described previously, in order to measure viscous compliance (Trotter and Koob, 1995). Specimens were clamped loosely on the upper surface of a Lucite shelf with 1.5 cm (50 %) of the specimen held in place in the horizontal position. The free 1.5 cm of the specimen was held 2 mm above the horizontal plane by placing the end on top of a movable pin. The test was started by rapidly removing the pin, thereby allowing the unsupported 1.5 cm of the specimen to act as a cantilever beam and to bend under the force of gravity. A stopwatch was used to measure the time it took for the free end of the specimen to deflect a vertical distance of 4 mm, starting at the time when the free end of the specimen reached the horizontal position of the top of the shelf supporting the clamped end. Times were rounded up to the nearest second. Bending times of less than 1 s were routinely assigned a bending time of 1 s. Some very plastic specimens could not be mounted on the apparatus because they flowed when lifted with forceps; nevertheless, they were also assigned bending times of 1 s. Since mean values presented for compliant specimens with bending times of less than 1 s were arbitrarily assigned a value of 1 s, parametric statistics could not be used to compare means. Means ± standard deviations (S.D.) are presented for results from those experiments in which variations are relevant. The theoretical basis for employing bending tests to measure tissue viscosity and for interpreting the results from these tests are explained in detail in Trotter and Koob (1995).

The principal test solutions used in these studies were Mops-buffered artificial sea water (ASW) (0.5 mol l⁻¹ NaCl, 0.05 mol l⁻¹ MgCl₂, 0.01 mol l⁻¹ CaCl₂, 0.01 mol l⁻¹ KCl and 0.01 mol l⁻¹ Mops, pH 8.0) and EGTA-ASW, in which the CaCl₂ was replaced by 7.2 mmol l⁻¹ EGTA. To assess the effects of isolated molecules on tissue viscosity, standard specimens were initially incubated in one of the above solutions for 90 min and then in the test solutions containing purified macromolecules for an additional 90 min. The incubations were carried out at seawater ambient temperature (12–15 °C) with intermittent gentle agitation.

For the preparation of freeze–thaw (FT) extracts, the body wall was cut into approximately 2 mm×4 mm×6 mm pieces, and the dark pigmented outer dermis, including the epidermis, was separated from the white inner dermis using a razor blade. For preparation of FT extracts of inner dermis, 20 g of tissue was finely minced and extracted at seawater temperature in 5 tissue volumes of EGTA-ASW for 5 h with intermittent gentle agitation. The fluid was removed and replaced with the same volume of fresh EGTA-ASW. The tissue in EGTA-ASW was then frozen for at least 2 h at −70 °C, followed by incubation at seawater temperature until completely thawed. These freeze–thaw steps were repeated for a total of five cycles. To prepare FT extracts of the outer dermis, the epidermis was scraped off using a razor blade, and the remaining pigmented portion of the outer dermis was minced and extracted with ASW for 5 h at ambient sea water temperature. The subsequent steps were carried out as described above for the inner dermis, except that ASW was used in place of EGTA-ASW.

Removal of the tissues and clarification of the extracts was accomplished by centrifugation at 27 000 gfor 30 min. The resulting supernatants were dialyzed against 0.8 mol l⁻¹ NaCl, 20 mmol l⁻¹ Tris-HCl, pH 8.0, in preparation for anion-exchange chromatography. The dialyzed extracts were first applied to a 5 ml HiTrapQ column (Pharmacia Biotech, Sweden) in the same buffer to eliminate highly charged macromolecules. After the sample had been applied, the column was washed with 10 ml of the same buffer, and the flow-through and column wash were combined as the unbound fraction. The bound material was then eluted with 25 ml of 3.0 mol l⁻¹ NaCl in 20 mmol l⁻¹ Tris-HCl, pH 8.0. The flow-through and bound fractions were dialyzed against either EGTA-ASW or ASW in preparation for bending tests. Both stiffening and plasticizing activities were found exclusively in the unbound fractions.

A second HiTrapQ column was used for further purification of the stiffening and plasticizing factors. The unbound fractions from the first column were dialyzed against 0.05 mol l⁻¹ NaCl, 20 mmol l⁻¹ Tris-HCl, pH 8.0, then applied to a 5 ml HiTrapQ column pre-equilibrated with the same buffer. After washing the column with 10 ml of the starting buffer, the column was eluted at 2 ml min⁻¹ with a 30 ml linear gradient from 0.05 mol l⁻¹ to 0.8 mol l⁻¹ NaCl in the Tris-HCl buffer. Fractions (2 ml) were collected and monitored for protein by absorbance at 280 nm using a Beckman DU40 spectrophotometer. NaCl concentration in each fraction was assessed by measuring osmolality using a vapor pressure osmometer (Wescor, Logan, UT, USA). For bending tests, 250 μl of each protein-containing fraction was added to 2.25 ml of EGTA-ASW or ASW and dialyzed against EGTA-ASW or ASW for 5 h. Five standard specimens were incubated in these fractions for at least 90 min prior to testing. The remaining portions of the biologically active fractions were combined, dialyzed against 0.8 mol l⁻¹ NaCl, 20 mmol l⁻¹ Tris-HCl, pH 8.0, and chromatographed on either a 1.6 cm×60 cm Sephacryl S-100 or a S-200 HR gel-filtration column (Pharmacia Biotech, Sweden) eluted at 0.8 ml min⁻¹. Fractions (4 ml) were collected and monitored at 280 nm. Samples (2 ml) of each protein-containing fraction were dialyzed against EGTA-ASW or ASW. Five specimens were incubated in the dialyzed fractions for at least 90 min prior to testing.

All purification steps were monitored using sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS–PAGE). Samples (100–200 μl) of each fraction from the chromatographic steps were added to 1.8 ml of 95 % ice-cold acetone for precipitation of macromolecules. Following incubation at −20 °C for 30 min, the precipitates were collected by centrifugation at 14 000 gfor 20 min and then resuspended in 95 % ice-cold acetone. The precipitates were again collected by centrifugation as above, dried under vacuum, then redissolved in SDS–PAGE sample buffer containing the reducing agent β-mercaptoethanol. The samples were electrophoresed on linear-gradient (4 % to 20 %) SDS–PAGE gels, which were then stained with Coomassie Brilliant Blue R 250 for visualization of proteins and Alcian Blue to detect proteoglycans and glycosaminoglycans.

Specimens were prepared and examined using an electron microscope as described previously (Trotter and Koob, 1995). Briefly, freshly dissected full-thickness specimens of the body wall were fixed for 36 h at 20 °C in 2.5 % glutaraldehyde, 0.1 mol l⁻¹ Mops, 0.41 mol l⁻¹ NaCl, 0.05 mol l⁻¹ MgCl₂, 0.01 mol l⁻¹ CaCl₂, 0.01 mol l⁻¹ KCl, pH 7.9. They were then rinsed for 24 h in several changes of the same solution lacking glutaraldehyde, post-fixed for 2 h in 1 % OsO₄, 0.1 mol l⁻¹ sodium cacodylate, pH 7.3, stained for 2 h in the dark in 0.5 % uranyl acetate in water, dehydrated in increasing concentrations of ethanol and embedded in Spurr’s resin. Ultrathin sections were stained using uranyl acetate and lead citrate and examined with Hitachi H-600 electron microscopes.

Fresh inner dermis specimens, containing living cells, were plastic when incubated in EGTA-ASW. These conditions were therefore employed to assay stiffening activity. Five freeze–thaw cycles of inner dermis in EGTA-ASW were found to release activity that caused a marked increase in bending times (Fig. 1). Mean bending times in nine experiments ranged from 17.8 to 105.5 s for tissues incubated in the EGTA-ASW freeze–thaw extract compared with less than 1 s for control tissues incubated in EGTA-ASW alone. The freeze–thaw extract produced bending times comparable with those of the control tissues incubated in ASW. The extract of inner dermis collected prior to the freeze–thaw cycles failed to stiffen live specimens, indicating that the freeze–thaw cycles were required to release the stiffening agent (Table 1).

In contrast to the above results, FT extracts of outer dermis in EGTA-ASW did not increase bending times of inner dermis specimens (Table 1). Pre-washing of the tissue prior to beginning the freeze–thaw cycles was not necessary to extract stiffening agent, indicating that release of stiffening activity upon freezing and thawing did not rely on the prior extraction of extracellular components (Table 1).

Fresh inner dermis specimens were stiff when incubated in ASW containing normal Ca²⁺ levels (ASW in Figs 1 and 2). These conditions were therefore used to assay plasticizing activity in outer dermis extracts. Five freeze–thaw cycles of the outer dermis in ASW released activity that caused a marked decrease in bending times (Fig. 2). In six separate experiments, mean bending times of specimens incubated in the freeze–thaw extract in ASW were consistently below 1 s compared with 22.5–73.1 s for control tissues incubated in ASW alone. In most experiments, inner dermis specimens treated with the extract of the outer dermis could not be mounted on the apparatus because the tissue flowed when lifted with forceps. The extract was as effective on frozen and thawed inner dermis specimens as it was on fresh specimens (data not shown), indicating that the plasticizing activity acted directly on the extracellular matrix and not via cell-mediated processes.

In contrast to the effects of outer dermis extracts, FT extracts of inner dermis in ASW did not reduce the bending times of specimens incubated in ASW, nor was plasticizing activity detected in the ASW in which the outer dermis specimens were initially incubated (i.e. without FT cycles) (Table 2). As was found for the inner dermis, the first wash of the outer dermis was not necessary for the subsequent extraction and detection of plasticizing activity in freeze–thaw extracts (Table 2).

The stiffening agent eluted from a strong anion exchanger (at pH 8.0) at NaCl concentrations between 0.2 and 0.3 mol l⁻¹ (Fig. 3). SDS–PAGE analysis indicated that the biologically active fractions consisted of proteins ranging in apparent molecular mass between 8 and 200 kDa. Fractions 11–14 contained a previously identified glycoprotein, stiparin, which aggregates collagen fibrils in vitro (Trotter et al., 1996) but was ineffective at stiffening live specimens. Negatively charged, Alcian-Blue-staining material that eluted near the end of the gradient similarly had no effect on bending times.

The stiffening agent eluted from a Sephacryl S-200 HR gel filtration column in 0.8 mol l⁻¹ NaCl, 20 mmol l⁻¹ Tris-HCl, pH 8.0, with a K_av between 0.35 and 0.45 [K_av=(V_e−V₀)/(V_t−V₀), where V_e is the elution volume of the molecule, V₀ is the void volume and V_t is the total volume] (Fig. 4). SDS–PAGE analysis of these fractions showed several proteins ranging in apparent molecular mass between 15 and 50 kDa. Only one major protein with apparent molecular mass of 38 kDa (arrow, Fig. 4B) was common to the active fractions.

The plasticizing agent eluted from a strong anion exchanger at pH 8.0 at a NaCl concentration of approximately 0.1 mol l⁻¹ (Fig. 5). SDS–PAGE analysis of the single active fraction showed the presence of proteins ranging in molecular mass from 6 to 300 kDa (Fig. 5). Plasticizing activity eluted from a Sephacryl S-100 HR gel filtration column in 0.8 mol l⁻¹ NaCl, 20 mmol l⁻¹ Tris-HCl, pH 8.0, with a K_av of 0.5 (Fig. 6). SDS–PAGE analysis of this single active fraction showed three Coomassie-Blue-stained bands. The predominant protein migrated as a doublet with an apparent molecular mass of approximately 9 kDa (lower arrow, Fig. 6). This protein was also present, albeit in smaller amounts, in the following fraction. A Coomassie-Blue-staining doublet migrated with an apparent molecular mass of approximately 15 kDa and was apparently present in the two flanking fractions (upper arrow Fig. 6). A disperse Coomassie-Blue-staining material migrated between 9 and 4 kDa and was not detected in the neighboring fractions.

To determine whether the stiffener and plasticizer are spatially segregated in the dermis, as suggested by these results, we searched for stiffener in extracts of the outer dermis and for plasticizer in extracts of the inner dermis, employing extraction and purification methods identical to those described above. The results of bending tests on the Sephacryl S-100 fractions of the partially purified inner dermis extracts are shown in Fig. 7. As expected, stiffening activity was detected in fractions 13–15, with the greatest activity appearing in fraction 14. Detection of stiffening activity verified that the correct fractions had been collected from the anion-exchange column and that the extract retained biological activity through the purification procedure. Fractions 17–20 were tested for plasticizing activity. Plasticizing activity, which routinely reduces bending times to less than 1 s, was not detected in these fractions, indicating that the inner dermis contains relatively low levels, if any, of the plasticizer.

The results of bending tests on fractions from the Sephacryl S-100 chromatography of the partially purified outer dermis extract are shown in Fig. 8. Plasticizing activity was detected in fractions 18–21, verifying that the outer dermis extract retained biological activity through the purification procedure. Fractions 12–16 contained no detectable stiffening activity, indicating that the outer dermis contains little, if any, stiffener comparable with that found in the inner dermis.

Electron microscopy revealed granular cells in both the inner and outer dermis (Fig. 9). As described previously (Trotter and Koob, 1995), the granular cells of the inner dermis (Fig. 9A) were surrounded, either individually or in groups, by a continuous external lamina and contained, in separate cells, three different populations of membrane-bound granules, distinguished by size, morphology and electron density. The granular cells of the outer dermis were also surrounded by an external lamina and contained three different classes of granule (Fig. 9B). The least abundant granule was round and small (approximately 200 nm in diameter) and very electron-dense. The most abundant granule was larger (approximately 400–700 nm in diameter), round and less electron-dense. The third type of granule was very large (greater than 1 μm in diameter) and pleomorphic. Many images suggested that two or more granules had fused to create a larger granule. The contents were highly variable in electron density and, in the larger granules, had a flocculent appearance. With respect to the granular cells, two features distinguished the inner from the outer dermis. The cell density was notably higher in the outer dermis, and the cells with the largest granules constituted a greater proportion of the total granular cell population in the outer dermis. Neither of these variables was evaluated quantitatively. Both observations were based on the cells encountered in a series of contiguous electron microscopic fields through the thickness of randomly selected regions of dermis.

We have previously extracted an organic stiffening agent from the inner dermis of C. frondosa (Trotter and Koob, 1995). On the basis of the observation that this agent was non-dialyzable and destroyed by boiling, we speculated that it was macromolecular and possibly a protein. The present results confirm that the stiffening factor released by freezing and thawing the inner dermis is a protein, which has now been partially purified by anion-exchange and gel-filtration chromatography. At pH 8.0, it carries a slight negative charge, as shown by its elution from a strong anion-exchange column at low counter-ion concentration. The active fractions from a gel filtration column contained proteins ranging in apparent molecular mass from 15 to 50 kDa. However, only one protein, with an apparent molecular mass of 38 kDa, was common to all three fractions. Moreover, the relative abundance of this protein corresponded with the extent of stiffening activity in the three fractions, further suggesting that it is the stiffener. Purification of the stiffening molecule to homogeneity coupled with bioassays will be necessary to determine whether this protein is the cell-derived stiffening factor.

The present observations also establish that the sea cucumber dermis contains a plasticizing factor. This factor was only detected in extracts of the outer dermis. That the plasticizer originates in the outer dermis itself was verified by carefully removing the epidermis before the pre-incubation and freeze–thaw extraction. Partial purification of the plasticizing factor established that it is a small molecule (<15 kDa under denaturing conditions) carrying a slight negative charge at pH 8.0. Gel filtration chromatography indicated that the plasticizer is one or more of three low-molecular-mass proteins. Which of these three proteins, alone or in combination, is responsible for the plasticizing activity remains an open question and the focus of current research efforts.

Although future analyses will be necessary positively to identify and characterize the stiffening and plasticizing factors, the present results rule out several previously identified macromolecules from the dermis of C. frondosa (Trotter et al., 1995, 1996). Neither factor is a proteoglycan, as indicated both by the low NaCl concentration needed to elute the molecules from a strong anion-exchange column and by their staining characteristics after SDS–PAGE. Further, neither factor is stiparin, a recently identified glycoprotein from inner dermis that has been shown to aggregate isolated native collagen fibrils from C. frondosa (Trotter et al., 1996). The present results also exclude the possibility that either factor is a neurotransmitter or structurally related small molecule (Birenheide et al., 1998).

Both the stiffening and plasticizing factors appear to be stored inside cells of the dermis rather than in compartments of the extracellular matrix. Neither was extracted at detectable levels from live tissue under the conditions employed to pre-incubate the tissues before freezing, whereas matrix macromolecules such as proteoglycans and stiparin freely leached from the specimens under these conditions. We previously established by electron micrographic analysis that five freeze–thaw cycles effectively lysed the plasmalemma of inner dermis granular cells as well as lysing a significant proportion of the granules themselves, thereby releasing cellular/granular contents into the extracellular space (Trotter and Koob, 1995). On the basis of these considerations, one or more of the granule types found in the inner dermis is thought to be the source of stiffener. A similar logic applies to the granular cells of the outer dermis, which are the most probable source of plasticizer. The present results do not indicate which cell or granule types contain stiffener or plasticizer. It should be emphasized that the stiffener and plasticizer may not have been derived directly from lysed cells. Cell lysis may have caused the release of unidentified factors that allowed stiffener or plasticizer to diffuse out of the matrix. Possible factors include proteases or protease activators, the actions of which could have released either whole molecules or proteolytic fragments of parent molecules, which we have identified as ‘stiffener’ and ‘plasticizer.’ The release of proteases has been postulated to be involved in autotomy in sea cucumber dermis (Junqueira et al., 1980). Other enzymes, such as transglutaminase (Diab and Gilly, 1984), or enzyme activators may also have been released. Immunocytochemical experiments localizing stiffener and plasticizer in tissue will be needed to address these uncertainties. In addition, experiments similar to those reported previously (Trotter et al., 1996) using isolated, native dermis collagen fibrils will prove crucial for determining whether these factors promote or inhibit fibril–fibril interactions or act through some other mechanisms. The conclusion that the contents of granular cells in C. frondosa dermis are important regulators of tissue stiffness is consistent with previously stated hypotheses implicating secretory granules in the regulation of tissue compliance in sea cucumber dermis and other mutable echinoderm connective tissues (Wilkie, 1979, 1984, 1988, 1996; Holland and Grimmer, 1981; Smith et al., 1981; Motokawa, 1982a,b, 1984, 1988; Hidaka and Takahashi, 1983). Moreover, our results confirm and extend recent experimental evidence showing that cell lysis in sea urchin spine ligaments (Szulgit and Shadwick, 1994) and sea cucumber dermis (Motokawa, 1994) releases factors that stiffen the extracellular matrix. However, the present results differ from the earlier interpretations of this effect of cell lysis in that they implicate proteins, rather than Ca²⁺, as the effector molecules stored in the granules responsible for modulating tissue mutability (Wilkie, 1979, 1996; Matsuno and Motokawa, 1992).

As we have pointed out previously (Trotter and Koob, 1995), Ca²⁺ in the bathing medium appears to be unimportant as a direct modulator of tissue compliance in C. frondosa dermis. The results of the experiments presented here confirm and extend this conclusion. The stiffening agent at all stages of purification caused an increase in tissue stiffness in the presence of a Ca²⁺ chelator and in the virtual absence of Ca²⁺ in the bathing medium (for levels of Ca²⁺ in specimens treated with EGTA/ASW, see Trotter and Koob, 1995). Moreover, the plasticizing factor caused the specimens to become compliant in sea water containing normal amounts of Ca²⁺ (10 mmol l⁻¹). These observations suggest that the stiffening and plasticizing of the extracellular matrix brought about by these protein factors are independent of the bulk Ca²⁺ concentration in the tissue. The experiments presented here do not rule out the possibility that small amounts of Ca²⁺, perhaps the Ca²⁺ contained in certain granules, participate in the cellular mechanisms responsible for the stiffening or plasticizing of the matrix. For example, intracellular Ca²⁺ may be critically involved in secretory activity responsible for releasing the regulating factors into the extracellular matrix (see Trotter and Chino, 1997). Nonetheless, on the basis of the isolation of the two protein factors that can effectively modulate tissue stiffness independent of the Ca²⁺ concentration, it seems highly unlikely that Ca²⁺ itself is the stiffening agent.

It can be inferred that the relative abundance of the stiffening factor is markedly greater in the inner dermis than in the outer dermis because stiffening activity was not detected in freeze–thaw extracts of the latter. Conversely, it can be inferred that the outer dermis contains relatively more plasticizer than the inner dermis. These observations suggest that the two factors might be spatially segregated in the dermis, with the plasticizer being stored in cells of the outer dermis while the stiffener is stored in cells of the inner dermis. However, it has also been found that living specimens of inner dermis alone are reversibly plasticized by Ca²⁺ chelation, suggesting that both plasticizing and stiffening factors are located in the cells of the inner dermis (Trotter and Koob, 1995). It is possible that the plasticizing agent in the inner dermis differs from that isolated from the outer dermis. Further experiments will be needed to answer this question.

The abundance of the two factors in the dermis is relatively low (approximately five orders of magnitude less than collagen), especially when compared with the amount of collagen, proteoglycans and stiparin, suggesting that both factors have limited and very specific interactions in the matrix. This, in turn, implies that alterations of local rather than global interactions regulate tissue compliance. The constitutive state of the extracellular matrix may be such that slight alterations in the relative proportions of stiffener and plasticizer would have dramatic effects on overall tissue compliance. Alternatively, the effects of these factors might be amplified by other matrix macromolecules. For example, they could act as reversible activators of other matrix macromolecules or enzymes that, once acted upon by one or the other factor, set into motion a cascade of interactions among the macromolecules responsible for mediating collagen fibril interactions. We have shown that stiparin binds to and aggregates isolated collagen fibrils (Trotter et al., 1996), and our recent experiments (J. A. Trotter, K. Chino and G. Lyons-Levy, unpublished results) have indicated that at least two matrix glycoconjugates inhibit the binding of stiparin to fibrils. Amplification of the effects of the stiffener and plasticizer could be mediated via these matrix macromolecules.

Treatment of the inner dermis with the plasticizer preparations resulted in a dramatic loss in stiffness: treated specimens flowed under the force of gravity. This behavior is similar to the rapid and irreversible loss of tensility in intervertebral ligaments during autotomy in brittlestar arms (Wilkie, 1988). The plasticizing factor from the sea cucumber dermis may be related to the brittlestar autotomy factor. Unfortunately, this factor has yet to be isolated and identified. However, another autotomy-promoting factor that has been purified from fluids released from autotomizing sea stars, when injected into the coelom of unstressed conspecific individuals, caused a complex behavioral response ending with multiple arm autotomy and body wall softening (Mladenov et al., 1989). Whether this effect was a direct action of the factor on the autotomizing tissues or was mediated via another pathway has not been resolved.

These experiments were supported by grants to J.A.T. from the National Science Foundation and the Office of Naval Research and to T.J.K from the Shriners of North America.