Erythrocyte membrane structural features that are critical for the lytic reaction of Spirographis spallanzani coelomic fluid hemolysin

Cony. Biochem. Physiol. Vol. 105C. No. 3, pp. 401407, Printed in Great Britain

1993 c

0306-4492/93 $6.00 + 0.00 1993Pergamon Press Ltd

ERYTHROCYTE MEMBRANE STRUCTURAL FEATURES THAT ARE CRITICAL FOR THE LYTIC REACTION OF SPIRUGRAPHIS SPALLANZANI COELOMIC FLUID HEMOLYSIN CALOGERO CANICATT~**

and

PHILIPPE

RoCHt

*Department

of Biology, University of Lecce, 73100 Lecce, Italy and tURA CNRS de Neuroendocrinologie, Universitk de Bordeaux I, Avenue des Facultis, 33405 Talence, France (Received 23 December

1992; accepted for publication 12 March 1993)

Hemolytic activity of Spirographis spallanzani coelomic fluid depends on factor(s) strongly influenced by calcium but not by sulfhydril or disulfide reagents. 2. The lytic reaction was suppressed by low zinc ion concentrations but it was not influenced by the presence of proteinase inhibitors. 3. These data indicate that S. spallunzani hemolysin is a non-enzymatic, calcium-dependent, zincinhibitable factor that occurs naturally in the coelomic fluid. 4. In the absence of calcium, enzymatic desialization converted sheep erythrocytes into susceptible targets, suggesting the involvement of erythrocyte surface sialic acid. 5. However, the inhibitory effect of the sugar on anti-rabbit lysis was partially removed by addition of calcium. 6. Attempts to characterize membrane components that are critical for hemolysis were performed by inhibition experiments. 7. We found that saccharides, glycoproteins, mucosubstances as well as rabbit erythrocyte soluble tryptic fragments were ineffective in inhibiting hemolysis. 8. Sonicated dispersion of phosphatidyl choline, phosphatidyl glycerol, phosphatidyl ethanol, sphingomyelin and cholesterol did not influence the hemolytic reaction. 9. Rabbit erythrocyte extracted from membrane lipids (chloroform phase) did not modify the lytic activity against rabbit red blood cells. 10. Conversely, the methanol phase consistently reduced the lytic capacity of the fluid. 11. The heat-stable, trypsin-resistant inhibitory factor was most probably a small molecule, since dialysis removed the inhibitory effect. Abstract-l.

INTRODUCTION

Occurrence

of hemolytic

proteins

in invertebrates

is

levels (Bernheimer and Rudy, 1986; Ken, 1988; Canicatti, 1990). In the polychaete annelids, both internal fluids (Anderson, 1980; Parrinello and Rindone, 1981; Roth et al., 1990) and external mucous secretions (Canicatti et al., 1992) possess powerful lytic activity against a variety of erythrocyte types. In Spirographisspallunzani(Polychyaeta; Sabellidae), the activity depends on a thermo-labile, calcium-dependent factor (Parrinello and Rindone, 1981; Canicatti et al., 1992) which is apparently not related to the 40 and 45 kDa Eisenia fetida (Oligochaeta) lytic system (Roth et al., 1981). In this species, although it is always difficult to safely infer the properties of an impure hemolysin, evidence for the proteinaceous nature of the hemolytic factor(s) has been obtained (Parrinello and Rindone, 1981; Canicatti et aI., 1992). However, in contrast with other invertebrate lytic systems (Bernheimer and Rudy, 1986; Canicatti et al., 1987; Roth et af., 1989; well

documented

at

all

phylogenetic

401

Canicatti, 1991) no information is available on membrane components which can interact with the S. spallanzani hemolysin. The influence of target membrane components upon cytolysis caused by lysins may give a clue regarding the mode of action of lytic molecules. For this reason, we studied the effect of various membrane components upon lytic activity produced on erythrocytes by S. spalIanzani coelomic fluid.

Collection

MATERIALS

AND METHODS

of worms and

coelomicfluid

Spirographis spallanzani (Polychaeta; Sabellidae) were collected from the harbor of Taranto (Gulf of Taranto, Italy). The coelomic fluid was obtained by incision of the body wall and immediately centrifuged at 400g for 30 min at 4°C to remove the coelomocytes. The green pigmented supernatant was pooled then dialyzed in 0.5 M Tris-HCl, 0.15 M NaCl, pH 8.0 (referred to as Tris-NaCl), divided into 2 ml aliquots and stored at -20°C until use.

C. CANtcArTi and P.

402 Chemicals

All enzymes, carbohydrates, mucosubstances, lipids, protease inhibitors and proteins were purchased from Sigma Chemical Co., St. Louis. Salts were purchased from Fluka (Paris). Hemolytic

assay

Hemolytic activity was tested against a 5% suspension of rabbit erythrocytes (Bio-Merieux, France). The degree of hemolysis was calculated according to the method of Parrinello and Rindone (1981) after 10min incubation at room temperature. Proteolytic

assay

Maskel and Di Capua’s (1988) method of in vitro gelatin lysis assay was employed. Briefly, a solution of 0.5% gelatin and 1.5% agarose was prepared in 0.1 M Tris-HCI, pH 7.0. Ten millilitres were poured into a Petri dish. Wells of 6 mm in diameter were cut in the solid agarose and filled with 30 ~1 of solution to be tested. After 6 hr incubation at 37°C undigested gelatin was precipitated with 5 ml of 15 g HgCI,, 20 ml of 12 N HCI in 80 ml distilled water. The diameter of the clear circle around each well was measured. Protease

inhibition

Inhibition of both hemolytic and proteolytic activities were performed by using 1 mM (final concentration) of the following proteinase inhibitors: pepstatin, chimostatin, puromicin, bestatin, antitosyl-L-lysine chloro-methyl pain, benzamidine, ketone (TLCK) and phenyl -methyl - sulfonyl -fluoride (PMSF). Soybean trypsin inhibitor (STI) at 1 mg/ml final concentration was also employed. Erythrocyte

enzyme

treatments

Rabbit and sheep erythrocyte (Bio-Merieux, France) surfaces were enzymatically modified. Rabbit cell suspension (10 ml) was incubated while shaking at 37°C with 10 ml of trypsin solution (2.5 mg/ml) in 0.5 M Tris-HCl, 0.15 M NaCl, pH 7.5. After 1 hr, the erythrocytes were centrifuged at 2000g for 5 min and the red colored supernatant removed by suction. Sheep erythrocyte suspension (10 ml) was incubated with 0.4 unit of sialidaze in 1 ml of 0.1 M sodium acetate buffer pH 5.5. After 1 hr incubation at 37”C, the cell suspension was centrifuged at 2000g for 5 min at 4°C. Both pelleted trypsinized and desialized erythrocytes were resuspended in Tris-NaCl at 5% concentration before use. Preparation

of tryptic fragments

Soluble tryptic fragments from rabbit erythrocytes were prepared according to the method reported by Giga ef al. (1985). Washed erythrocytes were suspended in 100 ml of 0.15 M Tris-HCI, pH 7.5, containing 25 mg of trypsin. Erythrocyte suspensions were incubated while shaking for I hr at 37°C then

ROCH

centrifuged at 2000g for 20 min at 4°C. Trichloracetic acid (50% final concentration) was added to the red colored supernatant. The resulting precipitate was removed by centrifugation for 20 min at 50,OOOg. The supernatant containing the tryptic fragments was neutralized with NaOH then dialyzed against distilled water at 4°C and lyophilized. Erythrocyte

membrane

lipid extraction

Lipid extracts were prepared from packed rabbit erythrocyte ghosts according to the method of De Gier and Van Deener (196 1). Two millilitres of ghosts were stirred with 10 ml of a 3: 1 mixture of chloroform-methanol for 10 min at room temperature. The 2 distinct phases were separated by centrifugation at 2000g for IO min at 4°C. Both the chloroform and the methanol phases were evaporated under vacuum and resuspended in 2 ml Tris-NaCl and designated Chp and Mp respectively. Inhibitory

experiments

Various substances, such as sugars, glycoproteins, mucous substances and lipids were tested by mixing v/v appropriate concentrations of inhibitors with Tri-NaCl dialyzed coelomic fluid. The inhibitory capacity was evaluated by substracting the percent value of hemolysis obtained after inhibition treatment from that calculated for reaction mixtures of untreated coelomic fluid. To take into account the dilution effect of the inhibitory solutions, 1: 2 diluted coelomic fluid was used as control. RESULTS Hemolytic

activity

Undiluted Tris-NaCl dialyzed S. spaflanzani coelomic fluid exhibits a powerful lytic activity against rabbit erythrocytes. Measured after 10 min at room temperature, the degree of hemolysis was 70.2 + 1.3%. Calcium (5 mM) supplemented coelomic fluid produced an almost complete hemolysis (98.0 f 0.1%) of target erythrocytes, confirming the previously reported enhancing effects of Ca*+ on the hemolytic reaction induced by Spirographis coelomic fluid (Parrinello and Rindone,l981). We found here that calcium was the only ion capable of increasing the degree of hemolysis. The other metal ions of the first transitional periodic table tested here were ineffective or inhibitors (Fig. 1). A concentration of 5 mM of Zn2+ completely abrogated hemolysis whereas Cu2+ produced about 62% inhibition. Mn2+ and Cd’+ were less inhibitory, producing 36 and 26% inhibition, respectively. Fe*+ and Mg2+ were, instead, ineffective. The ion inhibitory effects depend upon concentration. At least for Zn*+ very low concentrations (0.05 mM) were found to be critical to abrogate the lytic activity (Fig. 1). Thiol and disulfide covalent modification reagents did not influence the hemolytic activity against rabbit erythrocytes. Preincubation of coelomic fluid

Lytic reaction of Spirographiscoelomic fluid Dogno

100,

of homolyala tr)

403

However, adding S-10 mM CaCl, to the inhibition mixture partially removed the inhibitory effect of the sialic acids (Fig. 4). Inhibitory experiments with carbohydrates, proteins and mucosubstances

--*-__;_, 0 .Ol.OZ

; , , ; 1 2.6

catkin -c...

; , , : /

.06 con;c3entratl)ons

+m Z”..

*

cu..

-

Y”..

6

&I *I%..

4

840..

Fig. 1. Effect of various concentrations of metal ions on hemolytic activity against rabbit erythrocytes. Results are expressed as arithmetic mean, n = 3. SEM always < 0. I and not illustrated.

sulfhidryl reagents N-ethylmaleimide and P-hydroxymercuribenzoate did not reduce the hemolytic activity at any concentration tested (Table 1). The disulfide reagent dithiotreitol was also ineffective. Hemolytic activity was not influenced by adding proteinase inhibitors which affected, instead, the proteolytic capacity of the fluid (Table 2). Measured after 4 hr incubation at room temperature on gelatin plates at pH 7.0, 30~1 of S. spallunzani coelomic fluid produced a diameter of proteolysis of 1.2 + 0.3 mm. Adding the proteinase inhibitors chymostatin, puromicin, bestatin, antipain, benzamidine and TLCK consistently reduced the proteinase activity. A strong inhibition of about 67% was observed with 1 mM PMSF, whereas concentrations of 1 mg/ml ST1 completely abrogated the fluid’s proteolytic activity (Table 2).

with

Target surface enzymatic modification

Rabbit erythrocytes constituted highly susceptible targets for S. spallunzani hemolytic factor(s), whereas sheep erythrocytes were lysed in a medium that only contained Ca*+ ions (Fig. 2). These results are consistent with the influence of N-acetyl-neuraminic acid (sialic acid) residues during hemolysis. For this reason, we analyzed the effect of preincubating coelomic fluid with increasing concentrations of commercially pure N-acetyl-muraminic acid. As shown in Fig. 3, in the absence of calcium, only high N-acetylmuraminic acid concentration (10 mg/ml) inhibited hemolytic reactions against rabbit erythrocytes.

Table

I. Effect

of sulphydryl

and disulfide

as arithmetical

Saccharides did not influence the hemolytic reaction vs rabbit erythrocytes. As listed in Table 3, the monosaccharides D-ghCOSe, D-galactose, D-fucose, D-xylose, D-arabinose, D-ribose, D-fructose and Lsorbose did not have any important inhibitory effect. Also, the modified sugars D-galactosamine, D-mannosamine, N-acetyl-D-giuD-glucosamine, N-acetyl-D-galactosamine, N-acetyl-Dcosamine, c(-methyl glucopyranosade and mannosamine, m-methyl mannosamine were ineffective in producing inhibition. The dissaccharides a-lactose and saccharose were also not inhibitory. Soluble tryptic fragments from rabbit erythrocytes, prepared as described in Materials and Methods, were also not inhibitory. As shown in Fig. 5, none of the concentrations tested induced significant decreases in the hemolytic power of coelomic fluid. The glycoprotein fetuin tested herein was also ineffective (Table 4). Both mucin type I II, as well as chondroi’tin sulfate A, did not produce inhibition at any concentration tested (Table IV). Inhibitory experiments with lipids

The influence of lipids upon erythrocyte lysis caused by S. spallanzani hemolysin was studied using both sonicated dispersion of commercially pure lipids and lipids extracted from rabbit erythrocyte membranes. As listed in Table 5, sonicated dispersions of five purified lipids including cholesterol, did not interfere at 1 mg/ml final concentration. Higher concentrations up to 5 mg/ml, were also not inhibitory (not shown). The degree of hemolysis caused by S. spallunzani hemolysis upon rabbit erythrocytes did not vary significantly in the presence of increasing volumes of the chloroform fraction obtained from rabbit erythrocyte membranes (Fig. 6). Interestingly, adding increasing volumes of the methanol fraction to reaction mixtures progressively reduced the hemolytic activity. The inhibitory factor(s) was apparently not removed by trypsin treatment of the methanol fraction (Fig. 7). Conversely, boiling at 100°C for 5 min partially removed the inhibitory effect of the methanol phase. Dialysis with a molecular cut-off of

reagents

N-ethylamleimide

I

on lysis of rabbit

means k SEM, Degree

Concentrations

glyco-

erythrocytes.

Results

n = 3

of hemolysis

(%)

P-hydroxymercuribenzoale

Dilhiotreitol

0

80.2 * 0.

ImM

76.7 i_ 0.1

76.7 k 0.

76.7 + 0.1

5mM

80.2 * 0.1

82.5 i

80.2 k 0.1

IOmM

79.6 f 0.1

76.8 f 0. I

82.5 k 0.1

50 mM

73.8 k 0.1

8l.o+o.l

75.8 * 0.2

80.7 * 0.1

I 0. I

82.7 + 0.3

C. CANICATT~and

404

Table 2. Effect of different protease inhibitors on hemolytic and protease activities of S. spallanzani coelomic fluid. Results as arithmetical means + SEM, n = 3 Degree of homolysis (%) NOW DMSO Ethanol Pepstatin A (1 mM) Chymostatin (I mM) Puromycin (I mM) Bestatin (I mM) Antipain (I mM) ST1 (1 mg/ml) Benzamidine (I mM) TLCK (I mM) PMSF (1 mM)

79.1 83.5 82.9 85.1 80.2 85.1 83.5 79.7 82.4 84.0 85.1 83.5

f0.1 f 0.1 + 0.1 +O.l + 0. I kO.1 k 0.1 + 0. I f 0.1 + 0.1 20.1 + 0. I

68000 Da almost completely effect of the methanol phase

Diameter of proteolysis (mm) I.2 1.0 1.2 1.2 0.7 0.9 0.9 0.7

-F_0.3 * 0.5 f 0.3 f 0.4 * 0.2 * 0.2 * 0.5 * 0.5 0 0.9 + 0.6 0.9 + 0.6 0.4 f 0. I

removed the inhibitory (Fig. 7).

DISCUSSION

In accordance with Parrinello and Rindone (198 l), we found that the cytolytic activity of Spirographs spallanzani is induced by a calcium-dependent factor(s) that occurs naturally in the coelomic fluid. This activity is not significantly reduced by thiol or disulfide covalent modifying reagents. This suggests that sulfhydryl and disulfide bound groups are not essential in modulating activity of the native lytic factor. This finding differentiates the S. spallunzani hemolysin from many thiol-activated hemolytic toxins that occur in some species of Cnidarians (Bernheimer and Rudy, 1986) and bacteria (Freer, 1982; Harris, 1986). Zinc ions were a powerful inhibitor of the hemolytic activity of S. spahmzani coelomic fluid. This characteristic is shared with many other invertebrate lytic proteins (Avigard and Bernheimer, 1987; Canicatti and Grasso, 1988; Roth ef al., 1989; Leonard et al., 1990; Canicatti, 1991; Stabili et al., 1992) and most probably represents a common character-

P. ROCH

istic of the invertebrate cytolysins. Due to the antagonistic role played by Ca*+ (activator) and Zn*+ (inhibitor), it has been suggested that they represent the in ho regulating ions that participate in the lytic reaction (Canicatti, 1992). However, the limited analysis provided so far and the unknown mechanism of action of the ions on the hemolytic molecules do not permit any definite answer. Although evidence for the presence of proteolytic activity in S. spnllanzani coelomic fluid is clearly indicated in our results, it appeared that these proteinases do not account for hemolytic activity. By using external mucous secretions from S. spallanzani we previously demonstrated that absorption of mucus with erythrocyte ghosts abrogated hemolysis but not proteolysis (Canicatti et al., 1992). In addition, the kinetics of hemolysis by coelomic fluid reported by Parrinello and Rindone (1981) suggest that lysis is induced by a non-catalytic process. Thus, S. spalIanzani hemolysin is not a proteolytic enzyme. This conclusion is consistent with what is known for other invertebrate, lytic systems (Roth et (II., 1989; Roth et ul., 1990; Canicatti, 1991) although some protease inhibitors, especially the serine-protease inhibitor PMSF, are able to reduce the lytic activity of certain invertebrate fluids (Leonard et al., 1990). In the absence of calcium, rabbit but not sheep erythrocytes were susceptible to lysis by both Spirographs coelomic fluid (Parrinello and Rindone. 1981) and mucus (Canicatti c’t al., 1992). In the presence of calcium, both targets were lysed. We confirmed here these findings and provided new evidence that sialidaze treatment of sheep erythrocytes converted these cells. in the absence of calcium, into sensitive targets. On the basis of similar results obtained in a sea star species, it was postulated by Leonard et al. (1990) that sialic acid is most probably involved in the binding of hemolytic molecules through an analogous effect to that produced in mammals (Fearon, 1978).

Fig. 2. Hemolytic activity of S. spullanzani coelomic fluid against sheep (A) and rabbit (B) erythrocytes. Results are expressed as arithmetic mean + SEM, n = 3. Normal = undiluted coelomic fluid. 5 mM Ca*+ = coelomic fluid supplemented with 5 mM of Ca 2+ Trypsin = activity against trypsinized erythrocytes. Desial = activity against desialized erythrocytes.

Lytic reaction of Spirographis coelomic fluid 100

DWJIW ot hem0ly.i. ____

405

Table 3. Effect of simple and modified sugars on hemolytic reaction erythrocytes. Results as arithmetical means, n = 3. SEM
t9b)

of rabbit

T

Degree of hemolysis (%)

Sugar concentrations

26

I 0

.oa

.07

.I

.*

.6

Concentrations

0.6

f.?6

6

M

(mg/ml)

Fig. 3. Influence of various concentrations of N-acetylmuraminic acid on hemolytic activity vs rabbit erythrocytes. Arithmetic means, n = 3. SEM <0.2 except for 0 where SEM = 1.0. In S. spallanzani, the inhibition induced by N-acetylneuraminic acid can be suppressed by adding calcium indicating that the inhibition is reversible. Most probably the negative charges of the sugar are the only components that interfere with hemolysin-target interactions. Once neutralized by cation supplementation, the hemolysin-sheep erythrocyte interaction can occur. Further analysis on the activity of this sugar seems desirable to clarify its action. Composition of invertebrate hemolysin binding sites has been approached by testing the inhibitory effects of various sugars (Roth et al., 1981; Bertheussen, 1983). In only a few cases an effective inhibition of the hemolytic reaction was measured. As our results indicate, apart from sialic acid, none of the simple or modified saccharides were able to prevent rabbit erythrocyte hemolysis suggesting that carbohydrates are most probably not the receptor molecules responsible for binding of lytic molecules on erythrocyte surfaces. On the other hand, the lytic effect is not prevented by glycoproteins nor by mucins nor chondroitin sulfate.

25mM

50mM

100mM

94.2 89.8 82.2 88.6 87.3 86.0 81.1 93.6 91.1 89.8 90. I 81.1 81.2 84.8 88.6 84.8 86.0 87.3 83.5 94.9

87.3 86.0 88.6 88.6 88.6 83.5 91.1 77.2 88.6 88.6 83.5 83.5 87.3 86.0 88.6 89.8 86.0 87.0 92.4

89.8 86.0 88.6 88.6 88.6 74.6 91.1 79.7 84.7 79.9 80.9 82.2 82.2 86.2 79.9 87.3 86.0 79.7 90.0

None D-glucose o-galactose D-fucose D-xylose o-arabinose r_-arabinose o-ribose D-fruCtOSe L-sorbose o-glucosamine o-galactosamine N-acetyl-D-mannosamine N-acetyl-o-glucosamine N-acetyl-o-galactosamine o-mannosamine or-methyl-glucopyranoside a -methyl-mannose Lactose Saccharose

For many invertebrate hemolysins, the influence of membrane lipids upon erythrocyte lysis has been well established (Bernheimer and Avigard, 1976; Canicatti et al., 1987; Roth et al., 1989; Canicatti, 1990; Stabili et al., 1992). It was demonstrated that preincubation of lytic fluid with sonicated dispersions of purified lipids completely abrogated hemolysis. Such inhibition could be the result of hemolysin trapped on lipid microvesicles since hemolysin bound to lipid microvesicles has been observed by SDS-PAGE analysis (Roth et al., 1989). The results here are not consistent with the above finding. None of the lipid dispersions that we tested were able to prevent the hemolysis of rabbit erythrocytes. Also the lipids extracted from target erythrocyte membranes by chloroform were not able to inhibit hemolysis. Somewhat surprising was the dose-dependent inhibition of hemolysis by methanol phase erythrocyte membrane extracts. The inhibitory factor(s) was apparently

100I-

0

-

”

I

*

Normal

fluid

6 mY

ca++

10 rnM

Ca++

0 .Ol

.os

.a1

.‘9

.*I

.I

Concentrations

2

4

8

(milml)

Fig. 4. Degree

of hemolysis of N-acetyl-muraminic acidtreated coelomic fluid in absence (AM acid) or in presence of 5 mM and 10 mM of Ca*+ compared with the lysis induced by untreated fluid (normal fluid). Results as arithmetic means + SEM, n = 3.

Fig. 5. Effect of different concentrations of rabbit erythrocyte tryptic fragments on hemolytic activity vs rabbit erythrocytes. Results as arithmetic means, n = 3, SEM ~0.9 not visible.

406

C. CANlCATTi

Table 4. Effect of glycoprotein, mucins and chondroi’tin-sulfate A on hemolytic reaction. Results as arithmetical means, n = 3. SEM < I .4 Degree of homolysis

(%)

0.1 mg/ml

I mg/ml

IO mg/ml

None Fetuin

88.1 84.8

87.3

87.3

Mucin type I Mucin type II Chondroi’tin-sulfate

86.0 89.8 86.0

89.8 87.3 83.5

81.0 84.3 87.3

Table

A

5. Effect on sonicated dispersion of lipids on hemolytic reaction. Results as arithmetical means f SEM, n = 3 Degree of hemolysis

None I mg/ml I mg/ml I mg/ml I mg/ml I mg/ml

not removed by boiling for 5 min but was almost completely removed by dialysis, suggesting that a heat-stable molecule smaller than 68000 Da could be implicated.

0, 0

3

6

12

-L-

Mothwwl

fraction

Fig 6. Influence of both from rabbit erythrocyte cytes by coelomic fluid. SEM

Untnated

MF

50

26

of extracted

Ul

+

100

llplds Chloroform

trffitlon

methanol and chloroform fractions ghosts on the lysis of rabbit erythroResults as arithmetic means, n = 3, ~0.1 not visible.

MF-TrypaIn

MF-100%

Further analyses are in progress to explore the chemical nature of the inhibitor factor since it may provide clues regarding the erythrocyte receptor molecules involved in the binding of Spirographis lysin. Acknowledgements-This work was supported in part by a Convention Internationale du CNRSICNR-Italic and by the Ministere de 1’Education Nationale (C.C.). We are grateful to Professor E. L. Cooper for critical appraisal of the manuscript. REFERENCES

(%)

83.0 i 0. I 82.2 + 0.1 82.1 +O.l 80.4 rt_0.1 81.2+0.1 80.1 + 0.1

phosphatidyl-choline phosphatidyl-glycerol phosphatidyl-ethanol sphingomyelin cholesterol

and P. ROCH

f3F-dlaly#la

Fig. 7. Effect of various treatments on the methanol fraction (MF) ability to inhibit the lysis of rabbit erythrocytes compared with normal lytic activity (Untreated). MFTrypsin = methanol fraction treated with 1 mg/ml trypsin. MF-100°C = methanol fraction boiled for 5 min. MFdialysis = activity of the methanol fraction after dialysis. Results as arithmetic means k SEM. n = 3.

Anderson R. S. (1980) Hemolysin and hemagglutinins in the coelomic fluid of a polychaete annelid Gl.vcera dibranchiata. Biol. Bull. 159, 259-268. Avigard L. S. and Bernheimer A. W. (1987) Inhibition of hemolysins by zinc and its reversal by histidine. f&r. Immunol. 19, 110~115. Bernheimer A. W. and Avigard L. S. (1976) Properties of a toxin from the sea anemone Stoichactis helianfhus, including specific binding to sphingomyelin. Proc. Nat/ Acad. Sci. (U.S.A.) 73, 467471. Bernheimer A. W. and Rudy B. (1986) Interaction between membranes and cytolytic peptides. Biochim. hiophys. Acru 864, 1233141. Bertheussen K. (1983) Complement-like activity in sea urchin coelomic fluid. Dev. Comp. Immunol. 7, 21-31. Canicatti C., Parrinello N. and Arizza V. (1987) Inhibitory activity of sphingomyelin on hemolytic activity of coelomic fluid of Holothuria polii (Echinodermata). Dev. Camp. Immunol. 11, 29-35. Canicatti C. and Grass0 M. (1988) Biodepressive effect of zinc on humoral effector of Holothuria polii immune response. Marine Biol. 99, 393-397. Canicatti C. (1990) Hemolysins: pore-forming proteins in invertebrates. Experientia 46, 239-244. Canicatti C. (1991) Binding of Paracenrrofus lividus (Echinoidea) hemolysin. Camp. Biochem. Physiol. 98A. 463468. Canicatti C. (1992) The echinoderm lytic system. Bull. Zool. 59, 159-166. Canicatti C., Ville P., Pagliara P. and Roth Ph. (1992). Hemolysin from mucus of Spirographis spallanzani (Polychaeta: Sabellidae). Marine Biol. 114, 453.-458. De Gier J. and Vandeemer L. L. M. (1961) Some lipid characteristics of red cell membranes of various animal species, Biochim. biophys. Acts 49, 286-296. Fearon D. T. (1987) Regulation by membrane sialic acid of /I lH-dependent decay-dissociation of amplification C, convertase of the alternative pathway. Proc. Natl Acad. Sci. (U.S.A.) 75, 1971-1975. Freer J. H. (1982) Cytolytic toxins and surface acttvity. Toxicon 20, 217-225. Giga Y., Suton K. and Ikai A. (1985) A new multimeric hemagglutinin from the coelomic fluid of the sea urchin Anthocidaris crassispina. Biochemistry 24, 446 l-4467. Harris J. B. (1986) Natural Toxins in Animals, Planrs and Microbia. Oxford University Press, Oxford. Ken W. R. (1988) Sea anemone toxin structure and action. In The Biology of Hematocysts (Edited by Hessinger D. and Lenhoff H.), pp. 375405. Academic Press, New York. Leonard L. A., Strandberg J. D. and Winkelstein J. A. (1990) Complement-like activity in the sea star, Asrerias forbesi. Dev. Camp. Immunol. 14, 19-30. Maskel S. M. and Di Capua R. A. (1988) Qualitative assays for the protease activity of Lymunrria disparunclear polyhedrosis virus. J. invert. Path. 51, 139-144. Parrinello N. and Rindone D. (1981) Studies on the natural hemolytic system of the annelid worm Spirographis spal lanzani Viviani (Pol_vchaeta). Dev. Comp. Immunol. 5, 3342.

Lytic reaction of Spirogruphis coelomic fluid Roth Ph., Valembois P., Davant N. and Lassegues M. (1981) Protein analysis of earthworm coelomic fluid. II. Isolation and biochemical characterization of the Eisenia foerida andrei factor (EFAF). Comp. Biochem. Physiol. 69B, 829436.

Roth Ph., Canicatti C. and Valembois P. (1989) Interactions between earthworm hemolysins and sheep red blood cell membranes. Biochim. biophys. Acta 983, 193-198.

407

Roth Ph., Giangrande A. and Canicatti C. 1990. Comparison of hemolytic activity in eight species of polychaetes. Marine Biol. 107, 199-203. Stabili L., Pagliara P., Metrangolo M. and Canicatti C. (1992) Comparative aspect of echinoidea cytolysins: the cytolytic activity of Spherechinus granularis (Echinoidea) coelomic fluid. Comp. Biochem. Physiol. lOlA, 553-556.