Clostridium perfringens toxins (type A, B, C, D, E)

Pharmac.Ther.VoL 10, pp. 617 to 655 © PergamonPress Ltd. Printedin Great Britain

0163-7258/80/0701-0617505.00/0

Specialist Subject Editors: FRIEDRICH DORNER and JURGEN DREWS

C L O S T R I D I U M P E R F R I N G E N S TOXINS (TYPE A, B, C, D, E) JAMES L. MCDONEL

Department of Microbiology, Cell Biology, Biochemistry, and Biophysics, The Pennsylvania State University, University Park, PA 16802, U.S.A. Abstraet--Clostridum perfringens types A,B,C,D, and E produce at least 12 different antigens, referred to as toxins, that may be involved in pathogenesis. These antigens have been given the names alpha, beta, epsilon, and iota toxin ('major' toxins), and delta, theta, kappa (collagenase), lambda (protease), mu (hyaluronidase), nu (deoxyribonuclease), gamma and eta toxin ('minor' toxins). An enterotoxin and neuraminidase are also produced by certain strains. Significant progress has been made in characterizing alpha and theta toxin, and the enterotoxin, but relatively little is known about the other toxins. A variety of disease syndromes are believed to be caused by one or more of each of these toxins though their exact role in disease is not clear in most cases.

1. INTRODUCTION CIostridium perfringens produces at least 12 different soluble antigens, referred to as toxins, that may be involved in pathogenesis. Whether or not all of these antigens are directly involved in production of lesions or contribute to pathogenesis in animals or man is unclear. However, the term 'toxin' is commonly used to describe these antigens. The activities of the different toxins are given in Table 1. At least eight of the toxins are believed to be lethal. The roie in pathogenesis of eta and gamma toxins, and the enzymatic kappa, lambda, mu and nu toxins is uncertain. It is not even certain that eta and gamma toxins exist. Four of the lethal toxins, alpha, beta, epsilon, and iota, are considered to be the 'major' toxins and are used to group the species into five toxigenic types of A, B, C, D, and E. Type A strains produce predominantly alpha toxin, type B, beta and epsilon toxins, type C, beta and delta toxins, type D, epsilon toxin, and type E, iota toxin. A strain once classed as type F (Zeissler and Rassfeld-Sternberg, 1949) which has different colony appearance, elongated cell forms, heat resistant spores, and minor antigens, has since been more accurately classified as a subtype of C (Stern and Warrack, 1964). Typing of the organism is accomplished with type-specific antisera. The type neutralized by each specific antisera is given in Table 2. It is important to note that a clear-cut distinction does not always exist between different types of C. perfringens, and TABLE 1. General Description ofClostridium perfringens Toxins and Their Activities Toxin

Activity

Alpha (lecithinase) Beta Epsilon Iota Delta Theta Kappa (collagenase) Lambda (protease) Mu (hyaluronidase) Nu (deoxyribonuclease)

lethal; necrotizing; hemolytic lethal; necrotizing lethal; necrotizing lethal; necrotizing lethal, hemolytic lethal, hemolytic lethal; necrotizing; gelatinase disintegrates azocoll and hide powder, gelatinase release glucosamine from hyaluronic acid may be leucocidic in gas gangrene and post-uterine infection existence doubtful; may be lethal existence doubtful; may be lethal diarrheagenic; cytotoxic, causes glut damage; lethal inhibits cell receptor function

Gamma Eta Enterotoxin Neuraminidase

617

618

JAMES L. McDONEL

TABLE 2. Types of C. perfringens Neutralized by Various Type-Specific Antisera Type-Specific Antiserum A B C D E

Type Neutralized A A,B,C,D A,C A,D A,E

erroneous typing results can cause problems (Serrano and Schneider, 1978). Nontoxigenic variants are found within the types and 'degraded' types (Dalling and Ross, 1938; Buddle, 1954) are those which have lost their ability to produce a certain toxin. The toxins produced by the different types are given in Table 3. There is not total agreement in the literature about which strains produce which toxins. It should also be noted that a wide variety exists in amounts of toxins produced by different strains. Excellent reviews already have been written that include this kind of information about C. perfringens toxins (Willis, 1969; Ispolatovskaya, 1971; Hauschild, 1971; Smith, 1975a). C. perfringens is widely distributed in nature and normally is present in large numbers in soil, sewage, and the intestinal tract of animals and man (Willis, 1969). However, the 'organism also is found in air, dust, water, and numerous foods. In fact, C. perfringens can be considered to be ubiquitous due to its distribution throughout the ecosystems of the world. Type A strains are by far the most commonly encountered under noninfectious conditions (Collee, 1974), and they are the only ones associated with the microflora of both soil and the intestinal tract. The 'classical' (hemolytic type A) strains are more numerous in the intestinal tract than are nonhemolytic strains. Except under disease conditions, C. perfringens is not usually found in the stomach or upper and middle small intestine, while it may be isolated from the terminal ileum (Vince et al., 1972). Distribution in the large intestine is believed to be the most widespread. Types B, C, D, and E are invariably restricted to the intestinal tract, primarily of animals, and occasionally of man. Types B, C, and D have been isolated from soil but in areas where enteritis was affecting significant numbers of animals or humans. Even when inoculated directly into soil, strains of C. perfringens of intestinal origin die out within a few months at the most. It appears that they are unable to compete with the better adapted normal soil inhabitants (Smith, 1975a). Table 4 gives the general disease conditions in man and animals known to be caused by the different types of C. perfringens. It is not certain which toxins produced by each strain are directly or indirectly responsible for each of these conditions. A detailed description of each toxin's known activity under experimental conditions is given below. 2. ALPHA TOXIN (LECITHINASE, PHOSPHOLIPASE C) While strains of all five types of C. perfringens produce this toxin, type A strains usually produce the greatest quantities. On occasion a strain is isolated that is typical type A except that it does not produce detectable amounts of alpha toxin. Alpha toxin production also varies considerably with growth medium and amount of growth is not necessarily correlated with toxin production (Smith, 1975a). According to a study by MiSllby et al. (1976) the yields of alpha toxin produced by type A strains isolated from cases of gas gangrene and abdominal wounds were indistinguishable from those isolated from fecal samples from healthy persons. Not much is known about genetic control of alpha toxin production but one reported attempt to link its production to a plasmid was unsuccessful (Kr~imer and Schallehn, 1978). Suitable media and conditions for alpha toxin production have been reviewed by Ispolatovskaya (1971) and Smith (1975a).

-

.

-

-(+)

(+)

.

-

--

(+)

.

iota

-

+ -

+

delta

+

+ +

+

+

theta

+

+ +

+

+

kappa

+

-+

+

-

-

+ +

+

+

Minor Toxins lambda mu

+

+ +

+

+

nu

-

--

-

~

eta

Toxins Produced by the Five Different Types of C. p e r f r i n g e n s

+ = p r o d u c e d by s o m e strains of the type given. Q u a n t i t i e s of toxin p r o d u c e d b y different strains c a n v a r y . - = n o t k n o w n to be p r o d u c e d b y a n y s t r a i n s of the type given. ( + ) = p r o t o t o x i n , a c t i v a t i o n requires e n z y m e s . = existence d o u b t f u l 0 = not studied.

+

E

+ -

+

+

+ +

B

.

Major Toxins beta epsilon

+

C D

TYPE A

alpha

TABLE 3.

-

~ -

~

-

gamma

o

+ +

©

+

+

+ +

+

+

Other Toxins enterotoxin neuraminidase

O~

D

?

%

JAMES L. MCDONEL

620

TABLE4. Disease Conditions Caused by Different Types of C. perfringens Type

Disease

A

gas gangrene of man and animals; food poisoning; equine grass sickness; necrotizing colitis and enterotoxemia of horses lamb dysentery; enterotoxemia of foals, sheep, goats enterotoxemia of sheep (struck), calves, lambs, piglets, necrotic enteritis of man (pig-bel, Darmbrand), fowl enterotoxemia of sheep, lambs (pulpy kidney or overeating disease), goats, cattle, possibly man role in pathogenicity unclear; found in sheep, cattle; possibly responsible for colitis in rabbits

B C D E

Alpha toxin, by far the most widely studied of the perfringens toxins, has been demonstrated to have phospholipase C, lethal, necrotizing, hemolytic, and cytolytic activities (Willis, 1977). It was the first bacterial toxin to be identified as an enzyme (a phospholipase C that catalyzes the hydrolysis of the phosphodiester bond in position 3 of the lecithin molecule to produce phosphorylcholine and water-insoluble 1,2-diacylglyceride). It has been realized since the end of the 19th century that C. perfringens (then referred to as Bacillus aerogenes capsulatus and later as Fraenkel's gas bacillus and C. welchii) is the etiological agent for several disease conditions. During World War I C. perfringens came into great prominence as the most important causal organism of gas gangrene in man (Willis, 1969). Alpha toxin's enzymatic nature was first observed by Nagler (1939) and Seiffert (1939). Its mode of action as a specific phospholipase C (phosphatidylcholine: cholinephosphohydrolase E.C. 3.1.4.3) was first defined by MacFarlane and Knight (1941). Since then, much work has been done to elucidate the action of this toxin at various levels. There also is considerable interest today in using the toxin as a tool for studying biological membranes and lipids in relation to structure and function of the membranes. Extensive reviews of alpha toxin (Willis, 1969; Ispolatovskaya, 1971; Hauschild, 1971) and its application to the study of cell membranes (Avigad, 1976; Freer and Arbuthnott, 1976; Rosenberg, 1976; M~511by, 1978; Alouf, 1977; Zwaal et al., 1973), have been written. Physical Properties. As shown in Table 5, numerous authors have reported molecular weights for alpha toxin, determined by different methods, ranging from 30,000 to 106,000. The most reproducible value appears to be about 30,000 obtained by gel filtration chromatography, while SDS-polyacrylamide gel electrophoresis has yielded higher values (M611by, 1978). It is interesting to note that alpha toxin has been reported to move with beta toxin in G-100 columns and therefore it has been suggested (Akama et al., 1968; Sakurai and Duncan, 1977) that the molecular weights of both toxins are similar (beta toxin molecular weight, 30,000). Less variation has been reported for the isolectric point of alpha toxin, with a value of about 5.6 for the main component. The higher molecular weight values obtained with SDS gel electrophoresis may represent a polymer. The data from isoelectric focusing generally reveal two forms of the toxin of which the true enzyme is probably represented by the material with a pI of 5.5 (Smith, 1975a). Some reports suggest the existence of 2 4 isozymes (Sugahara et al., 1976; Litvinko et al., 1977). Studies done with SDS gels (Park et al., 1977) showed two bands for phospholipase C but only one of the bands had enzymatic activity. Both hemolytic and phospholipolytic activity were found to reside with the same (one protein) band. This is unlike Bacillus cereus toxin for which hemolytic, phospholipolytic, and lethal activities are each believed to be catalyzed by separated proteins (Ivers and Potter 1977). The physical properties of C. perfrinoens phospholipase C have been reviewed by Smyth and Arbuthnott (1974). Phospholipase C is capable of hydrolyzing both choline-containing phospholipids and glycerophosphatides. The substrate range of the enzyme among the phospholipids commonly occurring in membranes is given in Table 6. There is some uncertainty about hydrolysis of some of these phospholipids due to their reported differences in susceptibility to the enzyme. This depends upon whether they are studied as components of

621

Clostridium perfrinoens toxins (Type A, B, C, D, E)

TABLE 5. Molecular Weights and lsoelectric Points Reported for Phospholipase C of Clostridium perfringens a Authors

Year

Meduski and Volkova Bangham and Dawson Bernheimer and Grushoff Shemonova et al. Bernheimer

Gel Filtration

SDS Acrylamide

1958

Other Isoelectric Methods Point b 106,000 ~

1961 1967 1968

5.0, 5.8 31,000 51,000~

1968

et al.

Glushkova et al.

5.2, 5.5

1969

35,00042,000

1970

35,000

Teodorescu et al.

Sugahara and Ohsaka Casu et al. Isopolatovskaya Mitsui et al. M611by et al. M611by and Wadstr6m Stahl Bird et al. Smyth and Arbuthnott Takahashi et al.

Smyth and Wadstr6m Yamakawa and Ohsaka

1970 1971 1971 1973~ 1973 1973 1973 1974

5.3, 5.6 90,000 5.2, 5.5 49,000 4.6, 5.7 30,000

55,000 46,500 48,000

4.75, 5.5

1974

53,800

5.2, 5.5

1974

43,000

5.2, 5.3, 5.5

1975 1977

5.7

5.3, 5.3, 5.5 31,000

43,000

aModified from M611by (1978), Alouf (1977), and Smith (1975a). bDetermined by isoelectric focusing. CGamma irradiation inactivation. dUltracentrifugation.

intact membranes, or in their purified form. M611by (1978) has reviewed some of the problems in relation to water soluble enzymes acting upon water insoluble or partially soluble substrates. Another problem in interpreting published data on this enzyme is the fact that many authors have used preparations of the enzyme from commercial or other sources with insufficient or no regard for the fact that these preparations, unless produced under very rigorous purification standards, may contain as many as 10 contaminating active substances, including theta hemolysin, mu and nu toxins, protease, ~-glucosidase, N-acetyl-glucosamine, sulphatase, and fl-glucuronidase (M611by et al., 1973). Several of these contaminating substances have marked activity with respect to biological membranes (Smyth et al., 1975; Freer and Arbuthnott, 1976). Phospholipase C is a TABLE 6. Substrate Specific of C. perfringens Phospholipase C a Hydrolyzed Sphingomyelin Phosphatidylcholine (lecithin) Phosphatidylethanolamine Phosphatidylserine Lysophosphatidylcholine Lysophosphatidylethanolamine

Not Hydrolyzed Phosphatidylinositol Phosphatidylglycerol Diphosphatidylglycerol (cardiolipin) O-lysylphosphatidylglyc~rol

aSee van Deenen (1964), Hill and Lands (1970), Dawson (1973), Freer and Arbuthnott (1976), and M611by (1978).

622

JAMESL. McDONEL

comparatively thermostable enzyme which can retain 45 per cent of its activity when heated in a sealed ampule for 10-15 rain at 100°C (Smith and Gardner, 1950). It is also believed to be a zinc metalloenzyme (Ispolatovskaya, 1971) even though the presence of zinc ions has not been demonstrated in purified preparations of the enzyme (Smyth and Arbuthnott, 1974). However, zinc appears to be important to the stability of the enzyme and its resistance to proteases (Sato et al., 1978). The enzyme's activity is calciumdependent (Bangham and Dawson, 1961; Kurioka and Matsuda, 1976), though the role of calcium appears to be limited to its effect on the substrate, not the enzyme (Kurioka and Matsuda, 1976; Avigad, 1976; Dawson et al., 1976). Divalent cations and positively charged detergents are believed to activate the phospholipase action by giving the substrate a positive charge which optimizes attachment of the negatively charged enzyme. But, trace amounts of calcium ions have been found to be an absolute requirement besides the optimizing effects created just by charge differences (Klein, 1975; Dawson et al., 1976) because hydrolysis may occur without the positive charges when rotational freedom of individual phospholipid molecules is increased or when a monolayer of the substrate is expanded (Dawson et al., 1976). However, under normal conditions, the phospholipase C activity in vitro is inhibited by phosphate, citrate, fluoride, and other substances that tightly bind calcium (Smith, 1975a). It can also be inhibited by phosphonate and phosphinate analogues of lecithin [i.e. phosphatidylcholine with one or two direct carbon-phosphorus bonds (Rosenthal and Pousada, 1968; Rosenthal et al., 1969; MSllby, 1978)], chelating agents (in vitro and in vivo) such as diethylene triamine pentacetate (Senff and Moskowitz, 1969), and specific antitoxin. Biological Activity. Hemolysis due to alpha toxin occurs in most erythrocytes except those of horses and goats. Erythrocytes of cattle and mice seem most susceptible while those of rabbits, sheep, and humans are moderate in susceptibility (Smith, 1975a). The hydrolysis of phospholipid in the cell membrane precedes hemolysis, as can be seen by the increasing concentration of acid-soluble phosphate in the suspending medium when red cells are exposed to low concentrations of alpha toxin (Smith, 1975a). This damage to the cell membrane may result directly in hemoglobin release but other factors may be involved. An example of this is the 'hot-cold' lysis phenomenon. Erythrocytes of many species do not lyse when exposed to toxin at 37°C. However, when cooled to 4°C, rapid lysis occurs. At the same time, MSllby (1978) reports that only a comparatively slight hot-cold hemolytic phenomenon is detectable using highly purified phospholipase C from C. perfringens. The hemolytic activity of phospholipase C has been directly correlated with sphingomyelinase C activity. With Sphingomyelinase C the hot-cold phenomenon is striking because when used in its purified form lysis does not occur at all at 37°C. But, lysis of sphingomyelin-depleted red cells can be induced by several things including decreasing the temperature below the lipid phase transition temperature, or synergistic lysis with very small quantities of alpha toxin (MSllby, 1978; MSllby, 1976). Ethylenediamine tetraacetic acid (EDTA) also causes lysis of sphingomyelin-depleted cells in a manner that compares to hot-cold lysis (Smyth et al., 1975). It has been postulated that the divalent cation (mainly magnesium)-lipid interactions which stabilize the sphingomyelin-depleted membrane may be weakened by lowered temperature or addition of chelating agents (Smyth et al., 1975; MSllby, 1976). The lowered temperature may weaken hydrophobic forces (Low et al., 1974; Low and Freer, 1977) and EDTA leads to removal of ionic and hydrogen bonds between the cholesterol molecules and the polar groups of the sphingomyelin, either of which increases permeability and results in eventual lysis (MSllby, 1978). Avigad and Bernheimer (1976) showed that zinc ions (150/~M) conferred resistance to hemolysis of sheep erythrocytes by alpha toxin. They concluded that the zinc does not protect by preventing binding of the alpha toxin to the membrane. This phenomenon may support the idea that cations stabilize the membrane. However, the exact mechanism by which zinc ions protect the red cells is not known. Other studies have indicated that alpha toxin is able only to degrade membrane phospholipids and cause hemolysis of ATP-depleted erythrocytes (Frish et al., 1973; MSllby et al., 1973; MSllby et al., 1974). Taguchi and Ikezawa (1976a; 1976b) concluded

Clostridium perfringens toxins (TypeA, B, C, D, E)

623

that the hemolytic activity of phospholipase C depends upon asymmetric distribution of phospholipids in the erythrocyte membrane and accessibility of the enzyme to the phospholipids on the membrane surface (MSllby, 1978). And, several workers have postulated that accessibility of the enzyme to substrate is related to lateral pressure due to lipid packing in the outer leaflet of the phospholipid bilayer found in erythrocytes (Zwaal et al., 1975; Demel et al., 1975; van Deenen et al., 1976). Another example of the importance of accessibility is a phenomenon noted in aged red cells (Shukla et al., 1978). Aged red cells undergo microvessiculation and eventually become spherocytes which alters the accessibility of membrane phosphatidylethanolamine. This causes the older cells to become more resistant to alpha toxin activity. Further question of the direct involvement of alpha toxin phospholipase activity in hemolysis has arisen from the work of Sabban et al. (1972). They found that chicken erythrocytes are lysed by alpha toxin unable to hydrolyze phospholipids because of treatment with EDTA which binds calcium ions necessary for phospholipase C activity (Bangham and Dawson, 1961; Kurioka and Matsuda, 1976). At the same time, alpha toxin, unable to hemolyse red cells due to heating at 56°C in the presence of calcium, is capable of hydrolyzing membrane phospholipids. As mentioned before, caution must be exercised in interpreting data generated from membrane studies using alpha toxin preparations of unknown or unproven purity. MSllby et al. (1974) showed that the major portion of membrane damage in mammalian cells treated with phospholipase C was due to a contamination with theta toxin. Membrane damage could be induced by highly purified alpha toxin but lysis of cells was not noted. Further effects of phospholipase upon red cell membranes and the use of phospholipase to study lipids in relation to structure and function of membranes has been reviewed by MSllby (1978). Systems utilizing synergistic lysis of human, guinea pig and sheep erythrocytes with alpha toxin and streptococcal CAMP (Christie Atkins Munch Peterson) factor (Gubash, 1978) or other organisms (Choudhury, 1978) have recently been developed that may generate data that will aid in understanding toxin-induced hemolysis as well as membrane structure and function. Aside from its effects upon red cells, alpha toxin also lyses platelets and leucocytes (Mihancea et al., 1970), stimulates histamine release from cells in vivo (Habermann, 1960) and in vitro (Strandberg et al., 1974), and damages membranes in fibroblasts (MSllby et al., 1974; M611by, 1973) and intact muscle cells (Boethius et al., 1973). Furthermore, it causes aggregation of platelets in vitro (Sugahara et al., 1976) and in vivo (Ohsaka et al., 1978a, b). Ohsaka et al. (1978b) demonstrated that within 1-1.5 min of topical application of alpha toxin to rat mesentery, rolling leucocytes along vessel walls occurred in venules but not arterioles. Thrombi then formed in venules, capillaries, and eventually, in arterioles. They have concluded that thrombosis must be involved as an early important step in the pathogenesis of necrosis caused by alpha toxin and that these alpha-toxin induced thrombi (which result in hemostasis) may be one of the factors involved in the causation of toxemia often manifested in the late stage of gas gangrene (Ohsaka et al., 1978b). Otnaess and Holm (1976) showed that just phospholipid degradation in platelet membranes is not sufficient to cause the release reaction or aggregation of platelets. This was concluded because phospholipase C from Bacillus cereus (which did not act on sphingomyelin under their reaction conditions) degraded phospholipids but did not cause the aggregation noted to occur in response to Clostridium perfringens phospholipase C. This implies that sphingomyelin must be degraded in order for the release reaction or aggregation to occur. It has also been shown (Sugahara et al., 1977) that alpha toxin causes an increase in vascular permeability in guinea-pig skin due possibly to release of some mediator(s) from aggregating platelets (Sugahara et al., 1977; Ohsaka et al., 1978a). Wilkinson (1975) has demonstrated that the locomotor response of human monocytes is inhibited by the toxin as well. Szmigielski et al. (1978) showed that five minutes exposure to alpha toxin decreased uptake and increased leakage of a6Rb from a rabbit kidney cell line. At a later time, binding of ouabain, and uptake of 3H-uridine and 3H-glycine were inhibited, apparently due to membrane damage induced by subcytotoxic doses of the toxin. They suggested J.r'.T. 10/3--M

624

JAI~IESL. McDONEL

that the toxin causes disturbances in potassium transport through the cellular membrane with accompanying changes in Na+-K+-ATP-ase activity. The phospholipid content of alpha toxin treated versus untreated cultured embryonic chick muscle cells was reported (Kent, 1978) to be similar but cellular degradation and synthesis of phosphatidylcholine and sphingomyelin was reported to be 3-4 times greater in the treated than untreated cells. There is considerable difficulty in interpreting much of the data published over past years dealing with alpha toxin's role in disease because of lack of purity of the toxin preparations used. With this problem in mind, some attempt will be made to summarize numerous disease syndromes thought to be induced by the toxin. The systemic effects of alpha toxin are affected by the route of administration (Smith, 1975a). When injected intravenously, a 5 to 20 per cent drop in the red cell count occurs within a few hours followed shortly thereafter by death. Other effects are destruction of platelets, a drop in clotting time, and widespread capillary damage. Alpha toxin has been recovered from the serum of a patient suffering from C. perfringens septicemia (Moore et al., 1976). Under experimental conditions ninety per cent of the toxin injected disappears from circulating blood within 5 min of injection (Ellner, 1961). The toxin that disappears from the blood appears in the following organs: liver, 72 per cent, lungs, 15 per cent, kidneys, 8 per cent, and spleen, 5 per cent. Toxin does not appear to bind to skeletal muscle. In fact, when injected intramuscularly, apparent leakage into blood or lymph, vascular damage, and effects on platelets are absent. The muscle cells' plasma membranes are altered which probably causes the apparent degeneration of mitochondria, sarcoplasmic reticulum, and nuclei (Strunk et al., 1967). The disease gas gangrene (anaerobic myonecrosis) can be caused by six different species of Clostridium, including perfringens. Individual lesions often yield more than one species, though C. perfringens is recovered from 80~o of gas gangrene cases (Weinstein and Barza, 1973). It is believed that between 10,000 and 12,000 cases of this disease occur per year in the United States today (Hitchcock et al., 1975). Mortality varies from 0 to 100 per cent. Once started, the pathological process of gas gangrene can often result in death within 30-48 hrs. The exact involvement of alpha toxin in the disease is not clear. Kameyama et al. (1975) demonstrated that guinea pigs immunized with purified alpha toxin did not develop gas gangrene when challenged simultaneously with the organism, alpha toxin, and kappa toxin. Animals immunized with kappa toxin did develop gas gangrene when challenged with the organism and alpha toxin. It is likely that other toxins of C. perfringens type A mentioned in this review play a role in the development of the disease. Theta toxin has lethal, necrotic, and hemolyzing capabilities similar to those of alpha toxin. Kappa toxin is a potent protease, while nu toxin acts as a 'spreading factor'. Mu toxin affects DNA, fibrinolysin is produced, neuraminidase destroys the immunological receptors on erythrocytes, hemagglutinin inactivates the group factor A on erythrocytes, and the 'circulating factor' inhibits phagocytosis (Hitchcock et al., 1975). Intense and 'woody hard' edema develops in the area of the infection which frequently causes occlusion of the microcirculation. This gives rise to hypoxia which enhances growth of the organism. Thrombosis of local vessels occurs followed by production of hydrogen sulfide and carbon dioxide gases. Within as little as 12-20 hr the defense mechanisms of the host can be completely overwhelmed. All organs of the body are likely to be involved at this point. The treatment of choice is crystalline sodium penicillin G (recommended doses vary from 10 million to 32 million units per day) given intravenously, and possibly cephalothin (Keflin), coupled with hyperbaric oxygen (3 ATA) for 2 hr periods, and surgical debridement. These procedures, when used in concert, give consistent survival rates of 95 per cent assuming sufficiently early diagnosis and treatment of the disease. None of the treatments alone is nearly as effective. There has been some controversy over the use of hyperbaric oxygen as there has been over the effectiveness of antitoxin in treatment of the disease. Petrov et al. (1978) recently reported that polymerization of alpha toxoid can triple the immune response to the monomeric form. The development of highly immunogenic toxoid preparations could give toxoid prophy-

Clostridium perfringens toxins (Type A, B, C, D, E)

625

laxis a more central role in combating this disease. Much work needs yet to be done to distinguish the role played in pathogenesis of each of the toxins (in their highly purified state) produced by C. perfringens. Necrotic enteritis of chickens is thought (AI-Sheikhly and Truscott, 1977a; AI-Sheikhly and Truscott, 1977b; AI-Sheikhly and Truscott, 1977c) to be caused in large part by alpha toxin from C. perfringens type A. There is a similar disease, ulcerative enteritis, of quail caused by type C (Bickford, 1975) and, in fact, it has been reported that necrotic enteritis of chickens is due to type C organisms (Smith, 1975b). The disease symptoms described parallel closely those characteristic for type C enteritis in piglets. Typical lesions, which develop as early as 1 hour after infusion of organisms, include edema in the lamina propria, desquamation of epithelial cells, especially at villus tips, coagulation necrosis of villus tips, and eventually (8-12 hrs), necrosis of the entire villus structures. Other organs are not overtly affected but abnormal erythrocytes are present in them at 12 hrs. The histological damage at villus tips has striking similarities to the histopathology described for C. perfringens enterotoxin (see below). Suggestive evidence has been presented (A1-Sheikhly and Truscott, 1977c) that alpha toxin is the primary cause of necrosis in this disease, though more work needs to be done with purified alpha and theta toxins (both potent necrotizing agents) before concrete conclusions can be drawn about their contribution to disease symptoms. 3. BETA TOXIN Beta toxin, which may be coded for by a plasmid (Duncan et al., 1978), is produced exclusively by C. perfringens types B and C. These are the organisms most likely responsible for necrotic enteritis in man and animals. Very little is known about this toxin and interpretation of the few mode-of-action studies that have been done must be made with caution because only recently has the toxin been successfully prepared in a highly purified form. It is particularly likely that toxin preparations used in earlier studies have been contaminated with alpha toxin (and very possibly several other toxins produced by these organisms). The known physical properties of beta toxin are summarized in Table 7. The earliest report of purification was that of Akama et al. (1968). When they performed gel chromatography of crude material it was found that alpha toxin and beta toxin moved together and appeared inseparable by contemporary techniques. To overcome this problem, antiserum to alpha toxin was used to adsorb the toxin component prior to further separation on Sephadex G-100 of the immunoreacted material. The purified beta toxin showed none of the activities associated with toxins produced by type A organisms. It appeared to retain serological activity (guinea pigs were successfully immunized) and biological activity (measured by lethality after IV injection into mice) was retained. Worthington and Miilders (1975b) reported partial purification of beta toxin using chromatography on Sephadex G-50, G-100, and DEAE cellulose. The resultant material was not homogeneous on polyacrylamide gels. The toxin had 800,000 MLDso per mg nitrogen, a typical protein UV absorption spectrum, pI values of 5 and 6, and a molecular weight of TABLE 7. Molecular Weights and lsoelectric Points Reported for Beta Toxin of C. perfringens Authors Worthington and Miilders Sakurai and Duncan AI-Saadi and Lauerman

Year

Gel Filtration

1975 1978 1979"

SDS Acrylamide 42,000 + 2,000

30,000

Isoelectric Point 5,6 5.53, 5.06

20,000

"Personal communication, unpublished data.

626

JAMESL. MCDONEL

42,000 + 2,000 as determined on SDS gel electrophoresis. Sakurai and Duncan (1977; 1978) reported a 340 fold purification of beta toxin from type C organisms by ammonium sulfate fractionation, gel filtration through Sephadex G-100, isoelectric focusing on a pH 3 to 6 gradient, and immunoaffinity chromatography. The resultant purified toxin gave a single band with polyacrylamide gel electrophoresis. They reported that the toxin appears to be a single polypeptide chain protein with a molecular weight of approximately 30,000 and a major component with a pI value of 5.53 and a minor peak with a pI value of 5.06. The toxin was found to be heat labile (75 per cent loss of biological activity after 5 min at 50°C) and destroyed by trypsin after 30 min at 37°C. The 50 per cent lethal dose for a mouse was calculated to be 1.87/~g of the purified beta toxin. Unpublished data by A1-Saadi and Lauerman (personal communication) indicate that beta toxin has a molecular weight of 20,000, one disulfide bond, and is trypsin sensitive (Table 7). They found that the purified beta toxin has a MLDso of 4 #g. In a recent report Sakurai and Duncan (1979) demonstrated that beta toxin production by cultures of type C organisms could be enhanced by pH-controlled medium containing 1 per cent glucose, starch or sucrose, but that toxin synthesis was not related to growth yield of the organism. They also described a guinea pig skin test that allowed for distinguishing beta and alpha toxin activities, which the mouse lethality test does not necessarily do. Beta toxin is probably of most widespread concern to veterinary medicine. The most common types of C. perfringens to cause problems in animals are B, C, and D. Type B distribution is not thought to be world wide (not found in North America, Australia, or New Zealand) but is found in Europe, South Africa, and the Middle East. Necrotic enteritis from type C is reportedly world wide in distribution affecting calves, lambs, pigs, and possibly feeder cattle (Kennedy et al., 1977). While numerous other toxins are produced by types B and C (see Table 3) it is generally believed that beta toxin is the one primarily responsible for the symptoms noted in necrotic enteritis of the animals listed (Willis, 1969; Hauschild, 1971). Some successs has been reported (Hogh, 1976) in immunizing piglets against type C infection using anti-beta toxin serum. Some slight differences in symptoms between enteritis caused by type B versus type C may be attributed to the production by type B of epsilon toxin which is thought to increase intestinal permeability (Bullen and Batty, 1956), and mu toxin (hyaluronidase) which enable the necrotoxins to spread more widely in the intestinal tissues. This same effect is noted in the development of the characteristic purplish necrotic area produced when beta toxin is injected intradermally in guinea pig skin. Beta reactions produced by type B filtrates are very irregular in shape due to the presence of mu toxin while those produced by type C filtrates (most of which do not contain mu toxin) are smaller and more nearly circular (Willis, 1969). It is also believed that beta toxin is responsible for necrotizing jejunitis (enteritis necroticans) in man as well. This disease, also known as 'pig bel' (a similar syndrome had the name 'Darmbrand' in Germany and Scandinavia from 1944-1949) is endemic in the highlands of Papua New Guinea and has been reported in Africa, South East Asia, and even America (Lawrence and Walker, 1976). Sporadic cases have been reported in other western countries. Lawrence and Walker (1976) have proposed an interesting hypothesis concerning the development of enteritis necroticans. While the disease is endemic and relatively common in the highlands of New Guinea, it is rare in the lowlands, where poor nutrition and a predominance of type C organisms in the soil are equivalent to the conditions in the highlands. These workers postulated that the sweet potato, a staple food in the highlands, is the indirect cause of the prevalence of the disease. Sweet potatoes have been shown to contain heat-stable inhibitors of trypsin (Sugiura et al., 1973), a protease that readily destroys beta toxin activity. The contributory factors to development of disease are a low protein diet, low pancreatic tryptic activity, plus continuous consumption of trypsin inhibitors which results in inadequate proteolysis of beta toxin. In fact, this proposal has been supported experimentally with guinea pigs. Enteritis developed in guinea pigs given type C toxin filtrates and soybean material (which natur-

Clostridium perfrin#ens toxins (Type A, B, C, D, E)

627

ally contains protease inhibitors), while pigs given the same amount of beta toxin preparation but with heat-inactivated soybean material did not develop symptoms. Early studies on the pharmacological activity of filtrates from types B and C indicate that beta toxin can cause contraction of smooth muscle, liberate pigments from perfused liver, raise pulmonary arterial pressure and venous pressure, and decrease systemic pressure which usually is accompanied by heart block. Type B filtrates can cause liberation of histamine from perfused cat lung, and both types liberate a slowly-acting muscle stimulant (Kellaway and Trethewie, 1941). As mentioned before, however, the presence of unaccounted toxins, and use of 'specific' antiserum prepared against beta toxin that may not have been highly purified necessitates cautious interpretation of these data. More recently Parnas (1976) demonstrated that beta toxin administered into the rabbit jejunum and ileum shows paralysis of the motor activity of the intestine which he believed to be similar to the paralyzing activity in the intestine noted in piglet dysentery. 4. EPSILON TOXIN C. perfringens types B and D produce ~in essentially inactive prototoxin that, when digested with trypsin becomes a highly potent toxin called epsilon toxin. This toxin is of particular interest to veterinary medicine in that it causes a rapidly fatal enterotoxemia (also called pulpy kidney disease, and overeating disease) of sheep and other animals (Hauschild, 1971). Some conditions for maximal toxin production in culture have been described by Kulshrestha (Kulshrestha, 1974a; Kulshrestha, 1974b) and Smith (1975a). Radioimmunoassays for detection of toxin and antibodies (as little as 0.004 IU/ml) have been developed that are useful in laboratory studies and in screening large herds of vaccinated animals (Bernath, 1975; Lieberman, 1975). The purified toxin may also be assayed by a precipitin test, a microcomplement fixation test (Marucci and Fuller, 1971), a single radial diffusion test (Mancini et al., 1965), and reverse phase passive hemagglutination (detects as little as 0.6 ng/ml) (Beh and Buttery, 1978). Early attempts at purification of the prototoxin by column chromatography were made by Orlans et al. (1960) which resulted in a calculated molecular weight of 38,000 + 5,000 (see Table 8) and a sedimentation coefficient of 2.8 S. Thompson (1962; 1963) reported a molecular weight of 40,500 and a sedimentation coefficient of 2.48 S. TABLE 8. Molecular Weights and Sedimentation Coefficients of Epsilon Prototoxin of Clostridium perfringens Type D

Authors

Year

Orlans Thompson Habeeb Habeeb

1960 1962 1964 1972

Habeeb

1975

Worthington and Miilders AI-Saadi and Lauerman

1977 1979

Molecular Weight 38,000 + 5,000 40,500 23,200-25,000 34,250 ~ 33,200 b + 1,300 24,000 ¢ 34,250" 33,100 b 25,000 d 27,500 ~ 32,700 f 31,200 r,g

Sedimentation Coefficient 2.8 S 2.48 S 2.85 S 2.15 S 2.15 S

25,000 f'h

aDetermined by ultracentrifugation. bDetermined by sedimentation equilibration. CDetermined by gel filtration (Sephadex G-100). ~Determined by gel filtration (Sephadex G-100 with borate buffer). eDetermined by gel filtration (Sephadex G-100 with phosphate buffer). fActivated epsilon toxin. gDetermined by SDS gel electrophoresis. hpersonal communication, unpublished data.

628

JAMESL. McDONEL

Habeeb (1969) reported the presence in purified preparations of multiple electrophoretic forms of the prototoxin that were immunologically indistinguishable. Habeeb (1964) first reported a molecular weight of 23,500-25,000 and a sedimentation coefficient of 2.85 S then later, values ranging from 24,00(O34,250 (Habeeb, 1972) and 25,000-34,250 (Habeeb, 1975) with a stokes radius Of 1.96nm. Worthington and M~ilders (1977) reported a molecular weight of 32,700 for the prototoxin and 31,200 for the activated epsilon toxin. The isoelectric point for the major peak of prototoxin was 8.02. The activated toxin had two major peaks with pI values of 5.36 and 5.74. Both peaks contained material toxic for mice though the acidic fraction was more toxic on the average than the basic fraction. The significant change in isoelectric point and the small change in molecular weight that resulted upon activation of the prototoxin suggested that a small highly basic peptide(s) was being removed. This confirmed a similar observation made earlier by Habeeb (1969). A1-Saadi and Lauerman (personal communication, unpublished data) have used SDS gels to estimate a molecular weight of 25,000 for epsilon prototoxin. Epsilon prototoxin consists of one polypeptide chain of 311 amino acids with aspartate, threonine, serine, glutamate, valine, and lysine accounting for over 60 per cent of the total amino acid residues (Habeeb, 1975). The molecule has no free cysteine but contains one blocked cysteine. The N- and C- terminal residues are lysine. While activation is usually affected by the enzymatic activity of trypsin, some activation can occur spontaneously in growing cultures, presumably due to the activities of proteolytic enzymes (such as kappa and lambda toxins) produced by the bacteria (Smith, 1975a). It has been shown (Bhown and Habeeb, 1977) that activation is caused by a scission of the peptide bond between lys14-ala15 of the prototoxin molecule resulting in a small peptide (14 amino acids) split from the N-terminal to give active epsilon toxin. The epsilon prototoxin and toxin contain a high amount of fl-conformation (Habeeb et al., 1973). The activation of prototoxin results in some conformation change. But, there is a tendency of the cleavage product to bind to the remaining active toxin molecule which gives the toxin a conformation and serological identity similar to that of the prototoxin. The rebinding of the cleavage product can be prevented under the proper conditions. It is not clear whether the conformation changes resulting from activation are a necessary prerequisite for toxicity (Bhown and Habeeb, 1977). While rare or non-existent as a cause for human disease, epsilon toxin is responsible for acute enterotoxemia in animals, most commonly sheep, but found also in goats and cattle. A contributing factor in enterotoxemia is a sudden supply of diet (hence 'overeating disease') which results in passage of undigested food and C. perfringens type D organisms from the rumen to the intestines (Hauschild, 1971). This is followed by multiplication and epsilon prototoxin production by the organism in the intestine where proteolytic enzymes in the intestinal fluid (Niilo, 1965) activate the epsilon toxin. The highly toxic epsilon toxin, which increases intestinal permeability (Bullen and Batty, 1956; Bullen, 1970), then enters the blood (Jansen, 1967a) in substantial doses and causes typically swollen hypermic kidneys, lung edema, and excess pericardial fluid. The kidneys become pulpy (hence 'pulpy kidney disease') in sheep, but not in cattle, within a few hours of death (Jansen, 1960). Clinical symptoms (prostration and convulsions preceding death) are rarely seen (Hauschild, 1971), though affected sheep do exhibit a typical nervousness (Buxton and Morgan, 1976; Hartley, 1956). Immunity to the disease can be induced by injection of toxin, or toxoid (Jansen, 1967a; Jansen, 1967b; Oxer et al., 1971; Hyder, 1973; Worthington et al., 1973; Kennedy et al., 1977). A considerable amount of work has been done to study the effects of epsilon toxin on the brains of experimental animals because nervousness is a principle clinical sign in sheep with acute cases of enterotoxemia. Hartley (1956) described a focal symmetrical encephalomalacia (FSE) in brain stems of sheep affected by enterotoxemia caused by C. perfringens type D. Griner (1961) and Griner and Carlson (1961) produced brain lesions in sheep and mice by intravenous injection of epsilon toxin, and Gardner (1974) stated that the primary morphological change in the mouse brain was severe endothelial damage. He reported that the brain edema revealed a quantitative increase in water

Clostridium perfrin#ens toxins (TypeA, B, C, D, E)

629

content of the brain tissue and swelling of protoplasm in astrocyte and astrocyte processes around blood vessels. A more detailed study was reported in 1974 by Morgan and Kelley in which the light and electron microscopes were used to detect morphological changes in periaxonal and intramyelinic edema in the white matter of the cerebellar corpus medulare. There also was noted a swelling of axon terminals and dendrites in the grey matter adjacent to the left ventricles while endothelial degeneration appeared to be a feature of older lesions. In support of Griner and Carlson (1961) and Gardner (1974), Morgan et al. (1975) concluded that the primary lesion of this intoxication in the brain is in the vascular endothelium. They reported the extravasation of horseradish peroxidase (HRP) from the brain vasculature of clinically normal mice given lethal or sublethal doses of epsilon toxin. Within 20 min of intravenous administration HRP was found throughout the brain. Since detectable morphological damage usually takes 6 hrs to develop (Gardner, 1974; Morgan and Kelly, 1974), vascular leakage occurs several hrs before development of observable brain lesions. The minimal dose to produce leakage of HRP in mouse brains was 100 #g. Doses below 50/~g had no effect while 150-200 #g caused extensive leakage. Some differences in description by different investigators in pathology caused by this toxin may be due to the fixation or toxin preparation used. Worthington and MiJlders (1975a) demonstrated vascular leakage in mouse brain using epsilon toxin and 125I-polyvinylpyrrolidone and 125I-human serum albumin. With larger doses (4,000 MLDso) death, accompanied by extravasation of the albumin, occurred within 2-3 minutes. With smaller doses the effects took a longer period of time. Buxton and Morgan (1976) further characterized the pathology induced by epsilon toxin in lambs brains (FSE) as severe endothelial damage with swelling of perivascular astrocyte end feet, malacic changes, capillary hemorrhages, and choroidal edema. The extravasation of H R P occurred within 50 min of intravenous injection. It is also believed that epsilon toxin acts by binding to receptors in the brain vascular walls because administration of a formalinized toxoid prior to exposure to active toxin prevented extravasation of H R P and other typical vascular alterations (Buxton, 1976). The sequence of events in epsilon toxin intoxication has been described as follows (Buxton and Morgan, 1976). Epsilon toxin reacts with specific receptor sites causing vascular endothelial cells to degenerate. This results in tight junctions becoming patent, which alters fluid dynamics. The effect is that astrocyte end feet begin to swell. As degenerative changes progress, serum proteins, and eventually red cells, leak out, and the end feet rupture which causes edema. The edema then produces the clinical signs of nervousness seen in cases of acute lesions observed in actual cases of enterotoxemia of sheep. A detailed series of studies has recently been reported by Buxton in which binding and activity of epsilon toxin in mice (1978a), guinea pigs (1978b), cells from guineapigs, mice, rabbits, and sheep (1978b) were studied. In the mouse model it was found that epsilon toxin bound to the luminal surface of the endothelial lining of blood vessels in the brain, liver, lungs (sparse), and heart (slight) and to cells lining the loops of Henle, the distal convoluted tubules in kidney, and to the hepatic sinusoids. No binding was detected in the large or small intestines, or smooth (intestinal) and skeletal muscles in the body. The toxin has also been shown (Buxton, 1978b) to cause an increase in capillary permeability in guinea pig skin, and an elevation of serum levels of cyclic 3'-5' adenosine monophosphate in mice. A scheme has been proposed to explain the dramatic increase in blood glucose reported (Gardner, 1973a; Gardner, 1973b) to occur in the preclinical phase of type D enterotoxemia whereby hepatocyte-bound epsilon toxin (Buxton, 1978b) stimulates cyclic 3'-5' adenosine monophosphate production which in turn causes breakdown of glycogen to glucose. Buxton's use of the term 'enterotoxin' for type D epsilon toxin may be premature in that the intestine has not been demonstrated as a target organ. In fact, the evidence available to date suggests that the intestine does not have receptor sites for the toxin (Buxton, 1978a). Studies done with isolated cells (Buxton, 1978c) showed that guinea pig peritoneal

630

JAMES L. MCDONEL

macrophages were altered in morphology and eventually killed. The cells became swollen, the nuclear and cytoplasmic membranes became 'blistered' and noncontinuous, and the cytoplasm appeared structureless. It is believed that the outer surface of these cells represents the location of the receptors for epsilon toxin. Other cell types studied did not respond to the toxin (pulmonary alveolar macrophages from guinea pigs and rabbits, lymphocytes from guinea-pigs and sheep, and peritoneal macrophages and Lan Schultz ascites-tumor cells from mice). 5. THETA TOXIN Theta toxin (also known as perfringolysin O), which is produced by C. perfringens types A-E, is a member of the group of bacterial protein toxins known as 'thiol or SH-activated cytolysins'. These toxins have often been referred to as 'oxygen-labile hemolysins' (Van Heyningen, 1950; Smith, 1975a) in the past but the variability of oxygen lability and dependence of lability upon degree of purity have caused the nomenclature to be improved (Bernheimer, 1974). The term 'cytolysins' more accurately describes their ability to affect many cell types other than erythrocytes (Smyth and Duncan, 1978; Bernheimer, 1976; Alouf, 1977; McCartney and Arbuthnott, 1978). The cytolysins, which are produced by gram positive members of the families Bacillaceae, Streptococcaceae, and Lactobacillaceae, are all related in the following ways: (1) addition of cholesterol causes irreversible loss of biological activity; (2) hyperimmune horse serum causes crossneutralization; (3) they are lethal and cardiotoxic; (4) they have similar pH and temperature optima for cytolytic activity; (5) they lyse numerous erythrocyte species though mouse erythrocytes are characteristically resistant (Smyth and Duncan, 1978). Molecular weights and isoelectric points reported for theta toxin are given in Table 9. Smyth et al. (1972) reported a molecular weight of 61,500 and Smyth and Arbuthnott (1972) reported a pI value of 6.5. Mitsui et al. (1973b) reported two species of theta toxin, 0A and 0B with molecular weights of 53,000 and 50,000, and pI values of 7.05 and 6.65, respectively. Smyth (1975) reported four electrophoretically distinguishable components of theta toxin which he designated 0x, 02, 03, and 04. Most recently, Yamakawa et al. (1977) reported a molecular weight of 51,000 for theta toxin which was in close agreement with the molecular weight calculated from the amino acid composition. The protein was found to be composed of approximately 465 amino acid residues of which over 51 per cent were lysine, aspartic acid, serine, glutamic acid, and valine. There was found to TABLE 9. Physical Characteristics of Theta Toxin from Clostridium perfringens Authors

Year

Smyth et al. Smyth and Arbuthnott Mitsui et al.

1972 1972 1973 b

Mtillby and Wadstr6m Hauschild et al. Smyth

1973 1973 1975

Yamakawa et al. Hauschild et al. Smyth

1977 1973 1974

Molecular Weight

Isoelectric Point

61,500 b 0g 53,000 ~ 0a 50,000 a 60,000-65,000" 74,000 b 0~ 60,800-62,000 b 02 61,500 b 03 59,500 b 04 59,000 b 01-04 200,000-250,000 d 51,000 b 35,000~0,000 a'e 46,000 a'e

6.5 7.05 6.65 6.8 6.8-6.9 6.5-6.6 6.1-6.3 5.7-5.9

"Determined by gel filtration. bDetermined by SDS gel electrophoresis. 'Determined by ultracentrifugation. dDetermined by pore gradient polyacrylamide gel electrophoresis. 'These may represent cleavage products of the theta toxin that have reduced or no activity.

Clostridium perfringens toxins (Type A, B, C, D, E)

631

be one cysteic acid residue per molecule. Mitsui et al. (1973b) also did amino acid analysis of 0n and 0a fractions. The molar ratios they reported (converted to number of residues by Smyth and Duncan, 1978) are in agreement with Yamakawa et al. (1977) for some amino acids but not others. Lysine was determined to be the N-terminal residue which confirms the earlier observation of Hauschild et al. (1973). The purified theta toxin showed a single band on SDS gel electrophoresis. Theta toxin is responsible for the zone of clear hemolysis that surrounds colonies on blood agar plates. It is also noted for its lethal and necrotizing activities (Smith, 1975a). A report by Soda et al. (1976) suggests that the hemolytic and lethal activities of the toxin may be separate. It is reversibly inactivated by mild exposure to oxygen or oxidizing agents and is reactivated by thiol-reducing agents (but not other reducing agents, Mitsui and Hase, 1979) such as cysteine, 2-mercaptoethanol, dithiothreitol and sodium thioglycolate (Smyth and Duncan, 1978). It seems likely that theta toxin contains - S H groups (which are subject to reversible oxidation to S-S linkages) that are essential to its hemolytic activity (Smith, 1975a). Mitsui and Hase (1979) have recently supported this concept with data suggesting that the active form of the toxin has free thiol groups in its active site. These are changed to disulfide bonds with loss of activity by the exchange reaction between disulfide reagents having high Eo (oxidation-reduction potential) value. In addition, an unknown sulfide bond (possibly extra-molecular) in the toxin's inactive form may be reductively cleaved by thiols to form free thiol groups, thereby affecting recovery of the activity. Theta toxin is thermolabile (Duncan, 1975) and markedly temperaturedependent in that its activity rises sharply as temperature increases to 37°C but drops off sharply as the temperature is raised above 37°C. The hemolytic activity also is markedly dependent upon pH with the optimum activity at pH 6.7-6.8. Hydrogen ion concentrations above or below these values cause rapid loss of activity (Smith, 1975a). It has been known for many years that cholesterol irreversibly inactivates reduced theta toxin (Howard et al., 1953; Smyth and Duncan, 1978). The stereochemical specificity of sterol inhibition of theta toxin has been studied by Howard et al., 1953) and Hase et al. (1976). Sterols which contain alcohol functional groups, are an important subclass of the compounds called steroids. Steroids are compounds containing the basic perhydrocyclopentanophenanthrene carbon skeleton (see Masoro, 1968, for a lucid description of these substances). The fl-OH group on C-3 was found to be essential on cholesterol and related sterol molecules for inhibitory activity. The presence of an ~-OH group, an ester bond, or a keto group on C-3 caused the sterols to be incapable of inactivating the toxin. In addition, a hydrophobic group on C-17 (D ring) appears to be a prerequisite for inhibitory activity (Hase et al., 1976). Unpublished data of Watson and Smyth have been reported (1978) to support, for theta toxin, studies done with streptolysin O (Watson and Kerr, 1974; Prigent and Alouf, 1976) which showed that various other substitutions on the sterol molecules that interfere with the fl-OH of C-3 also alter inactivation potential. There are numerous problems associated with determining the number of molecules of a particular sterol that are necessary to inactivate a molecule of theta toxin (reviewed by Smyth and Duncan, 1978). In view of these problems, Smyth (1975) has reported that 1,500 cholesterol molecules are required per theta toxin molecule to affect inactivation. The exact nature of interaction between cholesterol and theta toxin is not known. Some hypotheses have been presented for other cytolysins that are similar to theta toxin, and more highly studied (Smyth and Duncan, 1978). Cells that do not contain cholesterol are insensitive to the SH-activated cytolysins which have been studied. The hemolytic activity of theta toxin is abolished if the toxin is preincubated with erythrocyte membranes (Smyth, 1975). It has been shown indirectly that theta toxin appears to bind to erythrocyte membranes and that the inactivating substance in membranes is lipid and must be a fraction containing cholesterol (or ergosterol). Phospholipids do not inactivate the toxin (Hase et al., 1975). Binding of theta toxin to cells appears, unlike hemolysis, to be independent of temperature. Binding occurs at 0°C while lysis does not. But if cells bound with theta toxin are raised in temperature to

632

JAMESL. MCDONEL

37°C they lyse (Hase et al., 1975). It is believed that hemolysis due to cytolysins is a multi-hit phenomenon and theta toxin has been demonstrated to require 2 and sometimes 3 hits for hemolysis to occur (Inoue et al., 1976). Mitsui et al. (1979) have suggested that 7 molecules of theta toxin per human erythrocyte are sufficient to cause complete hemolysis. They reported that the toxin causes altered bilayer portions ('cavities') in the membrane structure of the erythrocytes. They also noted a random aggregation of intramembranous particles on the protoplasmic face of membranes treated with low concentrations of theta toxin. Theta toxin has been shown to have a membrane effect upon several cell types aside from erythrocytes. Wilkinson (1975) demonstrated that the locomotor response of human blood neutrophils is inhibited by theta toxin. Freholm et al. (1978) have shown that theta toxin acts quickly, and, prior to development of clear-cut morphological alterations, antagonizes noradrenaline-stimulated cyclic 3'-5' adenosine monophosphate (cyclic AMP) accumulation and lipolysis in isolated rat adipocytes. They concluded that the onset of affect is so rapid that an enzymatic mechanism of action is unlikely. They hypothesize that lipolysis inhibition results from a decrease in cellular concentration of cyclic AMP and possibly ions such as calcium. Strandberg et al. (1974) demonstrated that theta toxin causes a zinc-inhibitable release of histamine from rat mast cells. Gill and King (1975) reported an interesting phenomenon in that membrane ghosts, prepared by lysing erythrocytes with theta toxin, did not reseal or vessicularize. Thelestam and MSllby (1975a; 1975b) demonstrated that functional pores were formed by theta toxin in human diploid fibroblasts that allowed for passage of nucleotides but not larger molecules. Unlike other membrane-active agents they studied, theta toxin did not cause an increase in the size of the functional pores with time. The size of pores and the functional alterations noted by Thelestam and M/511by led them to support the postulation that theta toxin interacts with membrane cholesterol in these cells. The absence of morphological alterations in fibroblasts may be due to the integrity of the rapid membrane repair mechanisms which are lacking in erythrocyte ghosts. The activity of theta toxin is particularly important, aside from pathological considerations, in that many studies on mechanisms of action reported in the past for other substances such as C. perfringens alpha toxin (M/511by et al., 1976; Smyth et al., 1975) and neuraminidase (Chien et al., 1975; Den et al., 1975; Rood and Wilkinson, 1974) have been shown to be contaminated with theta toxin. Smyth et al. (1975) showed that commercial preparations of alpha toxin caused arcshaped structures to form in the membranes of horse erythrocyte ghosts. These are similar to the 'pits' or 'holes' found during immune lysis of erythrocytes by antibody and complement (Dourmashkin and Rosse, 1966; Iles et al., 1973). These same structures were found when the ghosts were treated with highly purified theta toxin, while they were absent when treated with highly purified alpha toxin. When large doses of purified theta toxin were used the ghosts were actually fragmented. Similar arc-shaped structures have been shown to be produced by other cytolysins such as cerolysin (Cowell et al., 1978) and tetanolysin (Rotten et al., 1976). By using lipid dispersions and liposomes, Smyth et al. (1975) were able to show that cholesterol was the membrane constituent necessary for ring formation. The relationship of structural alterations to the lytic mechanism is unknown at this time (Smyth and Duncan, 1978). An interesting description of the course of events for hemolysis by a cytolysin such as theta toxin has been postulated by Smyth and Duncan (1978) in which the toxin sequesters cholesterol in the membrane thereby permitting a lipid phase-transition to occur. The result is an increase in membrane fluidity and permeability. Much more work must be done before the exact mechanism whereby the observed structural and functional alterations occur as a result of the action of theta and related thio-activated cytolysins. The role played by theta toxin in disease is unknown at this time. It has been shown to be lethal upon intravenous injection into mice but its lethal effect is not generally believed to be additive when co-injected with alpha toxin. However, according to Soda et

Clostridium perfringens toxins (Type A, B, C, D, E)

633

al. (1976), theta toxin, when co-injected with alpha toxin into mice, enhances the speed of

death even though the dose of toxin necessary to kill the animal is not lowered. Muscle tissue is not thought to adsorb theta toxin and may even suppress its production. For this reason there is question about the toxin's role in clostridial myonecrosis (Duncan, 1975). 6. ENTEROTOXIN C. perfringens enterotoxin is responsible for one of the most common types of food poisoning in the United States (Genigeorgis, 1975; Duncan, 1970; Chakrabarty et al., 1977). It is most often associated with meat, poultry and gravy-containing dishes (Smart et al., 1979; Duncan, 1975; Genigeorgis, 1975; Duncan, 1970) which are served or stored in large quantities. An extensive treatment of all aspects of the disease and the causative organisms has been presented by Walker (1975). While food poisoning due to C. perfringens was described as early as 1945 (McClung, 1945) it was not until the classic report by Hobbs et al. (1953) that the organism began to acquire its deserved attention as one of major public importance. Most if not all cases of classical C. perfringens food poisoning are caused by strains from type A. Strains isolated from human food poisoning cases do not appear to be limited to any particular group of type A, that is to say, they may be (a) beta-hemolytic, heat sensitive, like gas gangrene strains, (b) partially hemolytic, heatresistant, (c) partially hemolytic, or heat-sensitive (Hauschild, 1973; Hall et al., 1963; Hauschild and Thatcher, 1967; Hauschild and Thatcher, 1968; Sutton and Hobbs, 1968). Pinegar and Stringer (1977) recently reported the isolation of lecithinase-negative type A strains believed to be responsible for food poisoning outbreaks. However, caution must be exercised in that many enterotoxin positive strains are very weak producers of alpha toxin, and very sensitive assays are necessary to detect the toxin. The standard egg yolk plate test is likely to be of insufficient sensitivity to detect the small amounts of alpha toxin from a weak producer. Enterotoxin production appears not to be restricted to any particular agglutinating serotype of type A, though it it may be common to the entire group of Hobb's serotypes (Niilo, 1973c; Chakrabarty and Narayan, 1979). Strains of Hobb's serotypes 4, 5 and 7 have not been shown to produce enterotoxin however. It has been proposed that these strains, isolated from food poisoning cases, may have lost their ability to produce enterotoxin. Enterotoxin production by strains of type C (Skjelkval6 and Duncan, 1975a; Skjelkval6 and Duncan, 1975b) and type D (Uemura and Skjelkval6, 1976) have been described but the role in intestinal disease of entertoxin produced by these strains is not clear at this time. The disease is thought to be confined to humans though there is a report that may link type A enterotoxin to equine grass sickness (Ochoa and Velandia, 1978). Onset of primary symptoms of the disease (diarrhea and abdominal cramps) is usually 6-12 hrs and duration is usually 12-24 hrs. Some cases involve nausea and headache, but emesis and fever are rare (Hobbs, 1969). Fatal cases are rare but have been associated with elderly or debilitated persons. Early studies done with whole cells were performed by introduction of viable cultures of C. perfringens into the intestinal tracts of lambs (Hauschild et al., 1967), sheep (Niilo et al., 1971) and rabbits (Duncan et al., 1968; Duncan and Strong, 1969b; Duncan and Strong, 1971; Strong et al., 1971). The discovery that an enteropathogenic factor (Duncan and Strong, 1969a; Hauschild et al., 1970b; Niilo, 1971; Hauschild et al., 1970a) is produced during sporulation (Duncan and Strong, 1969a; Duncan et al., 1972) led to the isolation and purification (Hauschild and Hilsheimer, 1971; Stark and Duncan, 1972; Sakaguchi et al., 1973; Hauschild et al., 1973; Scott and Duncan, 1975; Uemura and Skjelkval6, 1976; Enders and Duncan, 1977) of a protein toxin. This toxin has been shown to be the factor primarily responsible for experimental disease induced in lambs (Hauschild et al., 1967; Hauschild et al., 1970a; Niilo, 1971), sheep (Niilo, 1971; Niilo, 1972; Niilo et al., 1971), calves (Niilo, 1973a; Niilo, 1973b), chickens (Niilo, 1974; Niilo, 1976), dogs (Bartlett et al., 1972), rats (McDonel, 1974; McDonel and Asano, 1975;

634

JAMES L. McDONEL

McDonel et al., 1978), rabbits (Duncan and Strong, 1971; Strong et al., 1971; Hauschild et al., 1971; Niilo, 1971; McDonel et al., 1978; McDonel and Duncan, 1975c; McDonel and Duncan, 1977), guinea pigs (Niilo, 1971), mice (Miwantani et al., 1978; Yamamoto et al., 1979), monkeys (Uemura et al., 1975; Duncan and Strong, 1971; Hauschild et al., 1971) and humans (Strong et al., 1971; Skjelkval6 and Uemura, 1977b). The physical characteristics of C. perfringens enterotoxin are outlined in Table 10. Aside from values obtained with SDS gels, the average molecular weight determined by most workers has been about 35,000. Anomalous aggregation of the enterotoxin into high molecular weight forms when subjected to detergent gel electrophoresis has been reported by Skjelkval6 and Duncan (1975b) and Enders and Duncan (1976). They reported that the high molecular weight values corresponded to multiples of a theoretical subunit of molecular weight of 17,500. However, in our experience, SDS gel electrophoresis of enterotoxin has consistently given a single band with a position equivalent to a molecular weight of about 35,000 (McDonel and Smith, unpublished observation). Yotis and Catsimpoolas (1975) have reported isoelectric focusing and isotachophoretic data that support the concept of the enterotoxin being a two-component entity. However, Enders and Duncan (1976) and Granum and Skjelkval6 (1977) also have demonstrated that the protein is a single polypeptide chain with a molecular weight of about 34,000. McDonel and Smith (1979) recently reported an in vitro system to synthesize C. perfringens enterotoxin. They found that in vitro synthesized material that was precipitated with anti-enterotoxin serum had formed three bands on SDS gels that represented TABLE 10. Physical Characteristics of Enterotoxin from Clostridium perfringens Type A

Authors

Year

Hauschild and Hilsheimer Stark and Duncan Sakaguchi

1971 1972

et al.

36,000 a + 4,000 34,000 a +_ 3,000 35,000 c

et al.

1973

Skjelkval6 and D u n c a n Skjelkval6 and D u n c a n

1975a 1975 b

Yotis and Catsimpoolas Freiben and Duncan Enders and Duncan

G r a n u m and Skjelkval6 McDonel and Smith

1976

1977 1979

by by by by by by

Sedimentation Coefficient

4.3 4.3

40,000 d 36,000 c 33,000 e 37,000f-Type C 36,000 a 33,400 c 68,000 b 85,000 b 105,000 ° 140,000 b

3.08

2,79

4.3

2.92

4.5 major 4.6 minor

1975 1975

Isoelectric Point

4.3

1973

Hauschild

~Determined bDetermined ~Determined dDetermined °Determined fDetermined

Molecular Weight

38,000 f 33,500 b 52,000 b 69,000b-most c o m m o n 82,000 b 108,000 b 140,000 b 34,000 c 17,500b 35,000 b 52,000 b

filtration chromatography. SDS gel electrophoresis. sedimentation equilibration. ultracentrifugation (Archibald method). amino acid analysis. polyacrylamide gel electrophoresis.

4.43

4.38 major 4.50 minor

Clostridium perfrinoens toxins (Type A, B, C, D, E)

635

molecular weights of approximately 17,000, 35,000, and 52,000. They have proposed that the enterotoxin may be a precursor (molecular weight > 35,000) of a spore coat component (see below). The spore coat component is envisioned to have a molecular weight of approximately 17,000, while the precursor (enterotoxin) is a single polypeptide chain that has one or more components contained within. This may explain the 'subunits' reported by Enders and Duncan (1976) and Skjelkval6 and Duncan (1977b) and the homology between enterotoxin and spore coat components (Freiben and Duncan, 1973; 1975). Frieben and Duncan (1975) reported the existence of enterotoxin-like proteins, that were serologically reactive with anti-enterotoxin serum, which could be extracted from spores of enterotoxin-positive as well as enterotoxin-negative strains. These proteins were labeled Type I (mol. wt. = 36,500; pI = 4.43), Type II (mol. wt. = 23,000; pI = 4.36), and Type III (mol. wt. = 14,500; pI = 4.52). The enterotoxin reported by Skjelkval6 and Duncan (1975a; 1975b) to have a molecular weight of 36,000-37,000 was purified from type C strains and is apparently indistinguishable from enterotoxin produced by type A strains. Sakaguchi et al. (1973) reported that the molecular weight of enterotoxin purified by ammonium sulfate and gel filtration through Sephadex G-200 was similar to that reported elsewhere, but they did not report the exact values they obtained. Amino acid analysis of the enterotoxin has been done which indicates that the enterotoxin molecule, with approximately 300 amino acid residues, is composed predominantly of aspartic acid, serine, leucine, glutamic acid, isoleucine, glycine, and threonine (Hauschild et al., 1973). Granum and Skjelkval6 (1977) reported results that agreed closely with those of Hauschild et al. (1973). However, Hauschild et al. reported the presence of 2 cysteic acid residues per molecule of enterotoxin while Granum and Skjelkval6 reported 1 cysteic acid residue. Enders and Duncan (1976) alluded to having found an amino acid composition comparable to that reported by Hauschild et al., but values were not given. Other characteristics of the enterotoxin molecule that have been reported are as follows. It is heat labile, i.e., 60°C for 5 min destroys biological activity though serological activity can be retained (10 per cent of original serological activity) even after 80 min at 60°C (Naik and Duncan, 1978a). Loss of serological activity is more rapid at 60°C if the pH is 6.0 or less. If heat-inactivated in food samples, about 12 per cent of the serological activity destroyed by heating at 60°C can be restored by treatment with urea for 1 hr at room temperature (Naik and Duncan, 1978a). Granum and Skjelkval6 (1977) reported 90 per cent loss of biological activity at 60°C after 1 min. They reported that enterotoxin heat-inactivated at 55°C for 1 min (70 per cent loss of activity) regained 50 per cent of original activity after 4 days. Enterotoxin heated for 15 minutes or more did not regain activity. The enterotoxin loses its activity during pronase and subtilisin treatment, but is insensitive to trypsin, chymotrypsin, bacterial lipase, steapsin, s-amylase, papain, and neuraminidase (Duncan and Strong, 1969a; Hauschild and Hilsheimer, 1971). Loss of activity occurs at pH 1.0, 3.0, 5.0, and 12.0 (Duncan and Strong, 1969a). It is essentially free of nucleic acids, fatty acids, lipid phosphorus, and reducing sugars, and has a u.v. absorption spectrum with a maximum of 278-280 nm and a minimum of 250 nm. It has a stokes' radius of 2.6 nm (Hauschild and Hilsheimer, 1971; Sakaguchi et al., 1973). Enders and Duncan (1976) have suggested that the acidic protein may have hydrophobic regions created by clustered hydrophobic amino acids which comprise about 40 per cent of the residues in the molecule. They also were unable to find in the enterotoxin the presence of ligands, free fatty acids, protease activity, or phospholipase activity. They reported the presence of less than 0.1 per cent lipid, 0.2 per cent nucleic acid, and no protein-bound carbohydrate. As mentioned above, Duncan and Strong (1969a) and Duncan et al. (1972) have demonstrated that enterotoxin is produced only by sporulating cells which distinguishes this toxin from other toxins produced by C. perfringens. Sporulation and enterotoxin production have been shown to be dependent upon pH, temperature, and carbohydrate availability in culture (Labbe and Duncan, 1974; Labbe and Duncan, 1975). Mutants blocked in sporulation earlier than stage III do not produce enterotoxin while ones

636

JAMESL. MCDONEL

blocked at stage III produce small amounts, and ones blocked at late stage IV and stage V are able to produce enterotoxin (Duncan et al., 1972). They concluded that the enterotoxin is a sporulation-specific gene product. Duncan (1973) and Labbe and Duncan (1977a; 1977b) have suggested that enterotoxin formation, which may involve a stable messenger RNA, appears to begin approximately 2.5-3 hrs after inoculation into sporulation medium. This corresponds to late state II or early stage III of sporulation. McDonel and Smith (1979) have isolated polysomes from cultures 1 hr after inoculation that produce enterotoxin in vitro. Duncan et al. (1973) described a paracrystaUine inclusion that is found only in sporulating cells that are enterotoxin producers, which may be an aggregate of enterotoxin or enterotoxin-like spore coat protein material. When sporulation mutants were studied the inclusions appeared only in cells blocked no earlier than stage IV of sporulation. The relationship of enterotoxin t o spore coat proteins was further detailed by Frieben and Duncan (1973; 1975). They found that anti-enterotoxin serum precipitated protein material extracted from spores of enterotoxin-positive as well as negative strains. Furthermore, the material was associated with spore 'coat' fractions but not spore 'core' fractions. They concluded that the enterotoxin is a spore coat structural component. Further characterization of this spore coat protein revealed that three distinct enterotoxin-like proteins can be extracted which Frieben and Duncan (1975) named types I, II, and III (mol. wts., 14,500, 23,000 and 36,500, respectively; see Table 10). All three of these proteins retained biological activity characteristic of purified enterotoxin (mol. wt., 35,000). The enterotoxin-like proteins (types I and II) particularly support the concept of a subunit(s) within the enterotoxin molecule that has typical biological and serological activity. McDonel and Smith (1979) have demonstrated that proteins synthesized in vitro by polysomes extracted from vegetative cells do not produce protein(s) precipitable by anti-enterotoxin serum. However, polysomes from cells grown in Duncan-Strong sporulation medium (1968) produced anti-enterotoxin serum precipitable proteins after one hour (early stage of sporulation) of incubation. At six hrs, 12 per cent of the total protein produced by the polysomes was precipitable with the antienterotoxin serum. The precipitated material (see Table 10) had molecular weights of 17,000, 35,000, and 52,000 as determined by SDS gel electrophoresis. These data support the earlier observation that enterotoxin is a sporulation-specific product (Duncan and Strong, 1969a; Duncan et al., 1972) and has given rise to the suggestion (McDonel and Smith, 1979) that enterotoxin may be a precursor molecule for the spore coat component described by Frieben and Duncan (1973; 1975). The presence of plasmids has been described for C. perfringens (Sebald et al., 1975; Sebald and Br~fort, 1975; Rood et al., 1978a; Rood et al., 1978b) but aside from one identified plasmid that may control beta toxin production (Duncan et al., 1978), plasmid control of other toxins including enterotoxin has not yet been demonstrated. Several biological and serological assay systems have been developed for the detection and quantitation of enterotoxin. It appears that the biological activity of the enterotoxin is more labile than the serological activity and the specific (biological/serological) activity tends to vary considerably from one preparation to another. While biologically active material will always be serologically active, the converse is not necessarily true. Conversion factors (Niilo, 1975) from assays of one activity to the other should be avoided. For mode of action studies, enterotoxin doses that are presented only in terms of mass (i.e. 10 #g/injection) have very limited meaning. It also is important to note that the biological activity of enterotoxin solutions, even when stored frozen at - 2 0 ° C begin to lose activity within weeks of storage. While reference will be made in the discussions to follow of quantities of enterotoxin (mass units) it is to be understood that biological activity/ unit mass varies considerably from one enterotoxin preparation to another. The earliest assay of the enterotoxin's biological activity was the rabbit ligated ileal loop test (Duncan and Strong, 1969a). This assay is based upon the response of the intestine to the enterotoxin, which is a net secretion of fluid and electrolytes into the lumen. The result is a noticeable expansion of the loops due to the excess fluid. The assay is inconvenient, relatively insensitive, and expensive. Stark and Duncan (1972) reported

Clostridium perfringens toxins(TypeA, B, C, D, E)

637

that the minimal dose detectable by this assay is 140-200 EU (erythemal units, see definition below) or 29 #g (Genigeorgis et al., 1973). Hauschild et al. (1971) described a rapid and more sensitive ligated loop technique in rabbits that could detect as little as 2.5 EU (6.25 #g) within 90 min of intraluminal challenge with enterotoxin. Niilo (1974) has described a chicken ligated ileal loops assay system that will detect 20-30 #g per loop (biological activity unknown). Probably the most reliable assay of biological activity reported to date is one based on the observations by Hauschild (1970) and Stark and Duncan (1971) that the enterotoxin, when injected intradermally, causes erythema in guinea pig or rabbit skin. The erythemal unit has been defined as that amount of enterotoxin that causes a zone of erythema 0.8 cm in diameter on the skin of a depilated guinea-pig which has received intradermal injections of the enterotoxin (Stark and Duncan, 1971). Results are usually read at 18-24 hrs after injection. This assay is more sensitive, less expensive, and easier than the ileal loop test. It can detect, with reasonable accuracy, quantities of enterotoxin as low as 0.25-0.5 EU. Detection has been reported of quantities as small as 0.064).125 #g (Genigeorgis et al., 1973). However, a linear dose-response curve can be seen only over a range of about 0.5-2.0 EU, at best. Specific activities of highly purified enterotoxin range from as high as 5,000 to as low as 2,000 EU/mg protein. It is very unusual to encounter specific activities over 3,600 EU/mg and the most commonly encountered activity is around 2,000 to 2,500 EU/mg. Another assay that has received some use is the mouse lethality assay (Hauschild and Hilsheimer, 1971; Stark and Duncan, 1971; Genigeorgis et al., 1973; Niilo, 1975). It has been proposed that mouse lethality results from the enterotoxin's activity on the portion of the brain controlling respiration (Miwantani et al., 1978). No effect upon the heart, embryonic heart cells grown in culture, or neuromuscular junctions was noted. Niilo (1975) reported that one LOso for the average 18 g mouse was about 5 EU (1.46 #g) when injected intravenously. A highly sensitive biological assay has recently been developed by McDonel, McClane and Marks (unpublished observation) that utilizes inhibition by the enterotoxin of plating efficiency of Vero (African green monkey kidney) cells grown in tissue culture (McClane and McDonel, 1979). This effect has been used to detect as little as 2.5 x 10 -4 EU (0.1 ng of protein). This corresponds to a concentration of 2.5 x 10 -3 EU/ml (1 ng/ml). A linear dose-response curve is obtained with dilutions containing between 1.25 × 10 -3 and 1.25 x 10 -2 EU (0.5-5 ng). Doses over 0.25 EU (100ng) cause inhibition of plating efficiency to approach 100 per cent. They have described a new unit of biological activity, called the plating efficiency unit (PEU), which is defined as that amount of enterotoxin causing inhibition of plating of 25 per cent of 200 cells inoculated into test wells (100 #1). Enterotoxin with a specific activity of 2,000 EU/mg has 400,000 PEU/mg (1 PEU = 2.5 ng, or 25 ng/ml). Erythemal activity and plating efficiency inhibiting activity correlate very closely with one another. Serological assays have been developed with excellent sensitivity and reproducibility. The Ouchterlony double-immunodiffusion method performed on standard microscope slides (Stark and Duncan, 1971) allows identification of enterotoxin (Genigeorgis et al., 1973). However, this method is not nearly as sensitive as electroimmunodiffusion, single gel diffusion tube, or reverse passive hemagglutinin (RPHA) techniques. Duncan and Somers (1972) reported that electroimmunodiffusion could detect as little as 0.01 #g (1 #g/ml) of enterotoxin while single gel diffusion could detect 3 #g/ml. Genigeorgis et al. (1973) reported that microslide diffusion could detect 0.013#g (0.50#g/ml) while the single gel diffusion method could detect 0.3/~g (0.9 ug/ml). Counterimmunoelectrophoresis (CIEP), a rapid, sensitive, and specific assay, has been used to detect as little as 0.002 #g (0.2/~g/ml) of enterotoxin (Naik and Duncan, 1977b). RPHA is reported to be sensitivie to quantities of enterotoxin as small as 0.05 ng (0.5 ng/ml) (Genigeorgis et al., 1973; Uemura et al., 1973). Fluorescent antibody has been used to detect intracellular enterotoxin in small numbers of sporulating cells (Niilo, 1977). The efficacy of numerous of these serological tests for laboratory detection of enterotoxigenic

638

JAMESL. McDONEL

C. perfringens has been reviewed by Niilo (1978; 1979). Skjelkval6 and Uemura (1977a) have utilized RPHA and CIEP to detect enterotoxin in feces of food poisoning victims. Naik and Duncan (1977a; 1978b) have used CIEP to detect enterotoxin in foods and human feces. At one time it was believed that fluid accumulation in the gut resulting from C. perfringens enterititis was due to alpha toxin (Nygren, 1962), but alpha toxin now is known not to be responsible for the diarrhea in this disease (Hauschild et al., 1967; Hauschild et al., 1971). The enterotoxin's ability to cause fluid accumulation in intestinal loops was first demonstrated in rabbits (Duncan and Strong, 1969a) and lambs (Hauschild et al., 1970b). The erythemal reaction to the enterotoxin in guinea pig and rabbit skin (Hauschild, 1970) was used by Stark and Duncan (1971) to develop an assay for its biological activity. They also demonstrated that the erythema is associated with an increase in capillary permeability (Stark and Duncan, 1972). The increase in capillary permeability was reported by Niilo (1971) to occur within 15-20 min of intradermal injection. In the same report Niilo described the systemic effects of the enterotoxin when injected intravenously into rabbits, lambs, and guinea-pigs. The responses of all three species are outlined in Table 11. The enterotoxin has been shown to be capable of causing depression of blood pressure in sheep as well (Niilo, 1972). Marked differences were noted in a similar study done with calves (Niilo, 1973b) in that they did not develop lacrimation, salivation, nasal discharge, or diarrhea, but intestinal hyperemia and mucosal sloughing did occur. It is believed that the enterotoxin has a strong predilection for the intestinal capillary bed. Niilo also reported that fluid accumulation in bovine Thiry fistulas can occur within 30 min exposure to the enterotoxin (1973a). All biological activities of C. perfringens enterotoxin so far described have been found to be neutralizable with specific antiserum prepared against the enterotoxin. Immunization of sheep (Niilo et al., 1971) against the enterotoxin prevented development of characteristic systemic reactions upon intravenous injection of the enterotoxin. However, the immunized sheep were as susceptible as control sheep to intraluminal enterotoxin challenge. At the same time Uemura et al. (1975) showed that intraluminal enterotoxin challenge in monkeys confers no detectable immunity (serum anti-enterotoxin titer) in the animals. It appears that with intraluminal challenge (the challenge route of C. perfringens food poisoning) enterotoxin does not penetrate to the immune system either to induce formation of specific antibodies or to be neutralized by them. But, Uemura et al. (1974) found that 85 per cent of Brazilians and 65 per cent of Americans sampled had substantial anti-enterotoxin serum titers. While the results of the study by Niilo et al.,

TABLE11. Systemic Effects of Enterotoxin When Injected Intravenously into Lambs, Rabbits, and Guinea-Pios ~ I. ParasympathomimeticProperties capillary permeability vasodilationin intestine increased intestinalmotility

ii. Other Effects diarrhea lacrimation salivation nasal discharge lassitude dyspnea hyperemicintestinalmucosa congestionof: liver lungs spleen kidneys ~FromNiilo (1971).

Clostridium perfringens toxins (Type A, B, C, D, E)

639

1971) would imply that these titers in humans are unable to prevent C. perfringens food poisoning, accessibility of the enterotoxin to the blood stream must be possible by a mechanism absent in simple intraluminal challenge with enterotoxin in experimental models. The frequency of detectable titers in humans studied gives good indication of the high exposure and carrier rate of C. perfringens type A food poisoning organisms. Skjelkval6 and Uemura (1977a; 1977b) reported an increase in serum anti-enterotoxin titer over a 60 day period in cases of human food poisoning due to C. perfringens but enterotoxin was not detected in the serum. Progress has been made during the last five years in defining precisely the physiological and pathological changes caused in the gut by C. perfringens enterotoxin (McDonel, 1979). It is important to review the literature during these years in considerable detail since current reports and reviews continue to be published that appear to be without awareness of the considerable progress that has been made (Yamamoto et al., 1979; Plotkin et al., 1979). The rat ileum was used as the first model system to describe transport alterations due to the enterotoxin (McDonel, 1974). In the rat ileum, when treated with the enterotoxin, there was a reversal of transport from absorption in controis, to secretion of fluid, sodium, and chloride. Glucose absorption was inhibited while transport of potassium and bicarbonate was unaffected. Fluid secretion was found to increase with enterotoxin dose up to a point beyond which excess enterotoxin had no effect. This saturation phenomenon was confirmed in another report in which rabbit ileal loops were challenged with a range of 50 (minimum dose to cause fluid accumulation) to 1,000 erythemal units (EU) of enterotoxin (McDonel and Duncan, 1977). The amount of fluid accumulation was indistinguishable in loops treated with each dose. Protein levels (an indication of tissue damage) in loop contents increased with dose in loops treated with 50 to 250 EU but did not change with challenges from 250 to 1,000 EU. Desquamation of intestinal epithelial cells was noted to occur at villus tips in enterotoxin-treated loops. Niilo (1971) has reported epithelial desquamation after intravenous injection of enterotoxin. Intestinal epithelial desquamation from intraluminal challenge with the enterotoxin has since been confirmed in rats (McDonel and Asano, 1975; McDonel and Duncan, 1975b) and rabbits (McDonel and Duncan, 1977; Hauschild et al., 1973; McDonel et al., 1978). When net transport of sodium in the rat ileum was resolved into unidirectional fluxes (McDonel and Asano, 1975) it was found that sodium uptake (influx) was identical in control and enterotoxin-treated animals. Net loss of sodium in enterotoxin-treated animals was shown to be due to nearly a two-fold increase in sodium efflux to the lumen. Studies done with rat everted ileal sacs (McDonel and Duncan, 1975a) revealed that the enterotoxin causes a significant reduction in oxygen consumption by the sacs. It was concluded that oxidative metabolism is inhibited while glycolysis is unaffected because glucose was absorbed and lactate produced in equivalent quantities by enterotoxintreated and control sacs. A similar (30 per cent) enterotoxin-induced decrease in oxygen consumption by mitochondria from rat liver and beef heart was reported (McDonel and Duncan, 1976) though the rate of phosphate esterification (ATP production) coupled to substrate oxidation (P/O ratios) was unaffected. Removal of enterotoxin from the rat gut after as little as 10 rain exposure will not prevent typical responses from developing at their normal rate, and overt tissue damage can occur within 15 min of exposure to the enterotoxin (McDonel and Duncan, 1975a; McDonel, 1974). Yamamoto et al. (1979) reported that removal of enterotoxin from mouse ligated loops caused a reduction in fluid accumulation, even if the loop was exposed to enterotoxin for 30 min prior to removal. They concluded that the enterotoxin does not bind freely to the intestinal cells. However, much more sophisticated binding techniques must be used before conclusions can be made about cell membraneenterotoxin binding interactions (see below). Studies with the rabbit have revealed that the enterotoxin is most active in the ileum, mildly active in the jejunum, and nearly inactive in the duodenum (Hauschild et al., 1973). Activities measured included tissue damage and transport of fluid, sodium, chloride, and glucose. Net sodium fluxes in all J.P.T. I0/3--N

640

.lAMES L. MCDONEL

three sections of the small intestine were resolved into unidirectional fluxes (McDonel, unpublished data). The results in the ileum confirmed those already reported for the rat (McDonel and Asano, 1975), but sodium transport appeared unaffected in the jejunum and duodenum. However, slight tissue damage was evident in jejunal sections so the relationship between tissue damage and electrolyte transport in this model system is not yet clear (McDonel, 1980). It is becoming increasingly evident that C. perfringens enterotoxin acts by direct cell membrane damage (McDonel, 1979). McDonel and Duncan (1975b) have shown that cycloheximide, an inhibitor of protein synthesis, does not prevent tissue damage and fluid and glucose transport alterations induced by the enterotoxin. This evidence indicates that induction of de novo synthesis of protein does not play a role in the mode of action. Furthermore, levels of cyclic 3'-5' adenosine monophosphate were not different in control and enterotoxin-treated intestinal mucosa. It appears that, rather than to activate novel metabolic pathways, synthesis, or existing enzymes, the enterotoxin acts through interference with energy production and cellular maintenance of structure and function. Support for the concept of direct membrane effect is found in data from an electron microscopic analysis of rat and rabbit intestinal tissue (McDonel et al., 1978). Scanning electron micrographs revealed that the enterotoxin causes the formation of blebs and morphologically altered epithelial cells localized at villus tips. With transmission electron micrographs it could be seen that blebs were forming in the apical membrane of epithelial cells. Brush borders lost their characteristic folded configuration and large quantities of membrane and cytoplasm from the terminal web region and below were being lost to the lumen. A sequential process was described whereby the highly structured brush borders were losing their characteristic folded configuration with the concomitant loss of significant quantities of membrane and cytoplasm to the intestinal lumen. Prior to lysis of these cells, internal organelles and membranes appeared essentially normal. This suggested that later changes in organelles were due to cell destruction rather than cell destruction resulting from organelle deterioration. No effect was noted upon epithelial cell basolateral membranes until cellular lysis caused a general breakdown of the cytoplasmic membrane. The chronology of these events implicates the brush border membrane of villus tip epithelial cells as being the enterotoxin's primary site of action (McDonel, 1980). Gyobu and Kodama (1978) have reported that C. perfringens enterotoxin is cytotoxic for Vero (African green monkey kidney) cells. An in vitro model system for studying the mechanism of action of the enterotoxin has recently been established (McClane and McDonel, 1979) which utilized Vero cells. The cells have proven to be very sensitive in that enterotoxin alters their morphology, viability, and macromolecular synthesis. Gross morphological damage was observed within 30 rain exposure of monolayers to enterotoxin, and approximately 75 per cent of the cell had detached within 40 rain. The affected cells had large blebs on their cytoplasmic membranes that were reminiscent of those described for intestinal cells treated with the enterotoxin (McDonel et al., 1978). Matsudo and Sugimoto (1979) have since confirmed this phenomenon in Hela and Vero cells. Vero cell viability dropped to nearly 50 per cent within 35-40 rain. Doses of enterotoxin as low as 0.1 ng (1 ng/ml) caused small but detectable inhibition of plating efficiency while more than 100ng (1 ug/ml) caused the inhibition to approach 100 per cent. It has been demonstrated (McClane and McDonel, 1980) that functional 'holes' (Thelestam and MiSllby, 1976) are being formed in vero cell membranes by the enterotoxin. The holes allow nearly 100 per cent of l*C-aminoisobutyric acid (AIB, mol. wt. 103), 50-60 per cent of 3H-uridine (mol. wt. 244), 25-30 per cent of 51Cr (complexed with polypeptides, tool. wt. 3,000-3,500), and 0 per cent of 3H-RNA (tool. wt. > 25,000) to pass out of treated cells within 30 rain. The size of these holes does not appear to increase with time since length of exposure to enterotoxin does not increase the minimum size for molecules to pass through the membrane. The degree of leakage is, however, dose dependent. They have also noted that growing cells are more sensitive to the enterotoxin than

Clostridium perfringens toxins(TypeA, B, C, D, E)

641

cells in confluent monolayer. More than 50 per cent protection can be confirred on the cells by pretreating them with 0.3 M sucrose, 5 per cent albumin, or 20-40 per cent polyethylene glycol. Membrane damage and resulting permeability changes have been used to explain the depression by enterotoxin of glucagon-stimulated AIB transport in cultured rat hepatocytes (Giger and Pariza, 1978). Further evidence of a membrane effect comes from binding studies (McDonel and McClane, 1979) in which it has been demonstrated that 125I-enterotoxin binds in a specific (binding inhibitable by pretreatment with native enterotoxin) manner to the vero cells. After the isolation of a strain of vero cells resistant to the enterotoxin from the highly sensitive cells, the apparent existence of high and low affinity binding sites were described in both cell types. The differei]ce was that the site density on the resistant cells (1.35 x 105 sites/cell) was ten-fold less than that on the sensitive cells (1.30 x 106 sites/cell). The association constants for high and low affinity sites were similar for both cell types. This indicates that the primary difference between the two cell types is in the density of sites. The biological effect of the enterotoxin in resistant cells (as determined by analysis of DNA synthesis, 3H-uridine release, and plating efficiency inhibition) was found also to be ten-fold less than in sensitive cells. It has been concluded that binding to the membrane is necessary for biological activity. Work is currently in progress to identify and isolate a specific compound (receptor) within the cell membrane with which the enterotoxin interacts. It is believed (McDonel, 1979; McDonel, 1980) that the enterotoxin is altering the configuration or structure of some component of the membrane which results in structural and functional changes in the membrane. Other binding studies have been performed with cells isolated from rabbit intestine, liver, kidney, and brain (McDonel, unpublished data). The 125I-enterotoxin was found to bind equally to cells from the intestine, liver, and kidney, but not the brain. Detailed studies with the intestinal cells revealed two classes of saturable binding sites. The rate and amount of binding to cells appeared to be temperature dependent in that lesser amounts of enterotoxin bound as the temperature was lowered from 37°C to 4°C. The binding site density on intestinal cells was 1.96 x 10 6 sites/cell. A problem exists in that reversibility of the binding has not yet been demonstrated. Either the enterotoxin interacts with a compound in the membrane in such a way as to become trapped, or it is being absorbed into the membrane or the cell in such a way that escape is impossible. C. perfringens enterotoxin is similar in ways to cholera, Escherichia coli, staphylococcal and shigella enterotoxins in that they all cause fluid and electrolyte accumulation in treated ileal loops. Staphylococcal and shigella enterotoxins have been shown to cause gut tissue damage under the proper conditions although the mechanisms appear dissimilar to that of perfringens (McDonel et al., 1978). However, in contrast to the other enterotoxins, perfringens enterotoxin inhibits glucose uptake, energy production, and macromolecular synthesis. It also causes breakdown of membrane structure and function by an apparently direct interaction with the outer cell membrane. Therefore, perfringens enterotoxin probably falls into a separate category from the other enterotoxins with regard to mechanism of action. 7. IOTA TOXIN Iota prototoxin, produced only by type E strains, was first described in 1943 (Bosworth, 1943). It is similar to epsilon prototoxin in that it must be activated to a toxic form by proteolytic enzymes. The enzymes can be those produced by the growing organisms such as lambda toxin (proteinase) (Craig and Miles, 1961) or by those added to the prototoxin such as trypsin (Ross et al., 1949). Very little is known about the iota toxin molecule. Purification has not yet been reported and therefore the few studies that have been done to define its activity must be interpreted with caution. Orcutt et al. (1978) have reported that iota toxin has an estimated molecular weight of about 70,000 and that toxoid (formalinized toxin) can be used

642

JAMESL. MCDONEL

to produce antisera which protects against experimental challenge with active iota toxin. The toxin has been shown to be heat labile (it is inactivated with 15 min treatment at 53°C) and it is rapidly inactivated by pH values between 5,2 and 4.2 (Craig and Miles, 1961). Craig and Miles (1961) have shown that iota toxin causes an increase in capillary permeability when injected intradermally into a guinea-pig. Larger doses injected intradermally cause necrosis. When injected intravenously, the toxin is lethal. Some controversy has existed over the role of type E strains in disease. Generally speaking, most sources conclude that these strains do not appear to play a role in pathogenicity (Cowie-Whitney, 1970; Flatt et al., 1974; Sterne and Batty, 1975). However, indirect evidence has recently been published by Patton et al. (1978) and Orcutt et al. (1978) that type E strains producing iota toxin may be responsible for enterotoxemia, or 'hemorrhagic typhlitis' in rabbits that have experienced diet-related stress. Patton et al. (1978) described cecal hemorrhage and edema in rabbits with the enterotoxin attributed to type E. However, there were difficulties reported in isolating type E organisms from many of the affected animals, and iota toxin was not detected in the isolated cultures grown on artificial medium. Until iota toxin is purified, characterized, and shown to induce characteristic lesions in experimental models, no firm conclusions can be made about the role of iota toxin in disease in animals or man. A recent report by LaMont et al. (1979) has described a toxin from cecal contents of rabbits with clindamycin-associated colitis that is neutralized by C. perfringens type E antiserum. The toxin has an apparent molecular weight (as determined by column chromatography) of 45,000, is heat labile (loss of activity with 60°C for 10 min), and pronase-sensitive. Its activity includes mouse lethality, HeLa cell toxicity, and severe epithelial cell necrosis in rabbit rectal explants maintained in organ culture. These authors suggest that the toxic material may be iota toxin but an organism was not isolated and the toxic material was not generated in culture. Further work will be necessary before the nature and origin of this toxic material can be identified. 8. DELTA TOXIN Delta toxin is produced only by young cultures of certain strains of types B and C. There appears not to be any correlation between delta toxin production and isolation of organisms from diseased animals. For example, strains of type B associated with lamb dysentery have been shown to produce delta toxin while strains from enterotoxemia of sheep and goats in Iran reportedly have not (Smith, 1975a). Type C strains which have been isolated from sheep with 'struck' in England, and cattle in Japan are producers of delta toxin (Yamagishi et al., 1971) while type C strains isolated from calves and lambs in the United States, and man in Germany and New Guinea do not appear to be active producers (Smith, 1975a). Very little is known about delta toxin's role, if any, in disease. It has been shown to be hemolytic only for erythrocytes of sheep, goats, pigs and cows (Oakley and Warrack, 1953; Brooks et al., 1957), and it is lethal but not necrotizing upon intradermal injection. Tixier and Alouf (1976) have described a purification scheme for delta toxin which involves ammonium sulfate precipitation and gel chromatography with Sephadex G-75. They reported that the purified toxin is a basic protein occurring as two forms with pI values of 8.8 and 9.4, and a molecular weight, as estimated by SDS gel electrophoresis, of about 42,000. Surprisingly, no studies have been reported since that deal with the mode of action of this purified delta toxin. 9. KAPPA TOXIN Kappa toxin attacks collagen and gelatin and therefore is categorized as (Oakley et al., 1946) and a gelatinase (Oakley et al., 1948). The enzyme hydrolysis of nonpolar regions of the collagen molecule characterized by (Gly-Pro-R)n, where R is an amino acid (Duncan, 1975). It is produced

a collagenase catalyzes the the sequence primarily by

Clostridium perfringens toxins (Type A, B, C, D, E)

643

strains of types A, D, and E, though strains of types B and C have been reported to produce the toxin. Testing for kappa toxin should be done with native collagen since other proteolytic enzymes which are unable to act on native collagen will attack gelatin and hide powder (Smith, 1975a). Azocoll (hide powder coupled with an azo dye) can be used as an indicator of kappa toxin with some reliability and specificity if pretreated with papain and polyvinylpyrrolidone (PVP). The resultant azocoll is called azocoll EP (Kameyama and Akama, 1970). Relatively little is known about the kappa toxin molecule or its role in disease. Kameyama and Akama (1971) have purified the molecule by gel filtration (Sephadex G-100) and DEAE-cellulose. They have concluded from their studies that kappa toxin apparently is protein in nature, free of carbohydrate and phosphorus, heat labile (enzymatic activity destroyed by 60°C for 10 min), and sensitive to low pH (activity destroyed by pH 4.5). The molecular weight, as determined by gel filtration (Sephadex G-150, -200) was estimated to be about 80,000. Kappa toxin has been shown to be lethal (approximately 30/~g is lethal to a mouse within one hr) and necrotic when injected intravenously into mice and intracutaneously into rabbits, respectively (Kameyama and Akama, 1971). Hemorrhage in the rabbits appeared as early as 5 rain after injection of as little as 0.7 #g of the toxin. Necrosis developed within a few days. Extensive destruction of connective tissue in the absence of visible changes in the muscle layer was noted in guinea-pigs injected subcutaneously with the toxin. Extensive hemorrhage in lungs but not other organs was noted after subcutaneous and intramuscular injection in guinea pigs. From these observations it is possible to envision kappa toxin playing a role in the development of gas gangrene by softening muscle tissue and contributing to the spread of alpha toxin and the organisms producing these toxins. Some muscle damage has been reported to occur from purified alpha toxin as well (Strunk et al., 1967; Kameyama and Akama, 1971) though the absolute purity question is still a problem in the interpretation of these observations. Kameyama et al. (1975) have reported that guinea pigs immunized with kappa toxin were resistant to challenge with the organisms (type A) and kappa toxin but not the organisms and alpha toxin. On the other hand, guinea pigs immunized against alpha toxin were resistant to challenge with the organisms and either alpha or kappa toxin. They concluded that alpha toxin is the main contributor to development of local infection and gas gangrene. However, gas gangrene is difficult to quantitate in degree as a response, and the purity of their toxin preparations for immunization and challenge is not that well documented. Much further study is needed to ascertain the roles and interdependence of these agents in development of disease (see lambda toxin, also). 10. LAMBDA TOXIN Lambda toxin is a protease that is produced by strains of types B, E, and some strains of D. Very little is known about this toxin and little if any meaningful work has been done with it in the past 20 years. Lambda toxin degrades numerous proteins such as hide powder, azocoll, gelatin, casein, hemoglobin, and seracin. It can be distinguished from kappa toxin (collagenase) in that it does not degrade native collagen. Early studies indicate that it is inhibited by normal nonimmune serum, is destroyed by pH below 5 or above 9, is heat labile, and is inhibited by lambda 'antitoxin', egg albumin, cysteine, cyanide, and citrate (Smith, 1975a; Willis, 1969). The purity and exact identity of preparations used in these studies are unknown and therefore these observations should not be regarded as firm. The role of lambda toxin in disease is unclear though its likely contribution to breakdown of muscle and connective tissue is bound to be of some significance (see kappa toxin). Lambda and kappa toxin both would be expected to contribute to spread of infection (supply of amino acids and peptides for organism growth, and breakdown of physical barriers) and other toxins (such as alpha toxin). Some amounts of these toxins are likely absorbed and distributed systemically and therefore able to contribute to

644

JAMESU McDONEL

generalized toxemia. The possibility has also been suggested that conversion of epsilon prototoxin to active toxin may be enhanced in some instances by lambda toxin (Willis, 1969). 11. MU TOXIN Mu toxin, which is a hyaluronidase, is produced in largest amounts by strains of type B, but is also produced by strains of types A and D, and, according to some sources, by type C (Smith, 1975a; Willis, 1969). It acts by releasing glucosamine from hyaluronic acid, a polysaccharide universally present in connective tissue that probably binds water to hold cells together in a jellylike matrix. Like most of the other minor toxins, very little of real substance is known about this toxin because it has not been studied in detail in recent years. Studies with material of undefined purity have indicated that mu toxin is heat labile (activity is destroyed rapidly by temperatures above 50°C) (Robertson et al., 1940), is relatively resistant to high and low pH values (activity remains at pH values from 3.9-8.5) (Rogers, 1948) though it is inactivated by pH 9 (Hale, 1944). Mu toxin has been demonstrated to have hemolytic, necrotic, and lethal activities. It has been shown to increase skin permeability (Willis, 1969; Smith, 1975a) and therefore it may contribute to more rapid spread of infection in gas gangrene cases. However, no relationship has been found between mu toxin production and organism virulence. The same problem exists for this toxin as exists for the others that have not been studied in depth in a purified state. The importance of mu toxin in disease is unknown, though certainly it does not appear to be essential for development of gas gangrene. It may contribute to severity or speed of spread of the disease.

12. NU TOXIN Nu toxin, which is a deoxyribonuclease, is produced by strains from all five types of C. perfringens, though most frequently by types A and C. It is believed to have lethal, hemolytic, or necrotic activities. It is, however, a leucocidin in that its ribonuclease activity results in destruction of nuclei of polymorphonuclear leucocytes and muscle cells when injected intramuscularly in rabbits (Robb-Smith, 1945). Nu toxin's involvement in gas gangrene appears to be reflected by the absence of a leucocyte response during the disease. 13. GAMMA AND ETA TOXINS The existence of gamma toxin (strains of types B and C) and eta toxin (one strain of type A) is doubtful. Suggestive evidence of their existence has arisen from discrepancies in neutralizing values of specific antisera used with these strains of organisms. However, until specific active substances are isolated and purified from suspect cultures, the existence of these toxins must be considered speculative at best.

14. NEURAMINIDASE Neuraminidase (sialidase) has been shown to be produced by strains of types A, B, C, D, and E (McCrea, 1947; Collee, 1965; Fraser and Collee, 1975; Fraser, 1978). A neuraminidase is an enzyme which cleaves acylneuraminic acid residues from oligosaccharides, glycoproteins, glycolipids or colonimic acid (Cabexas, 1978). Rood and Wilkinson (1976) have described three neuraminidase enzymes produced by C. perfringens with molecular weights of 310,000, 105,000, and 64,000. The role of neuraminidase in disease is unclear at this time. Various activities of neuraminidase in numerous systems have been reviewed by Rosenberg and Schengrund (1976). Neuraminidase may contribute to the

Clostridium perfringens toxins (Type A, B, C, D, E)

645

virulence and invasiveness of an organism but certain neuraminidase-negative organisms such as C. novyi type A are invasive and cause gas gangrene (Fraser, 1978). And there are neuraminidase-positive organisms such as C. tertium that are not thought to be pathogenic. At this time it is only possible to suggest that neuraminidase may contribute to a large number of other factors responsible for the development of disease caused by certain organisms that produce the enzyme. 15. OTHER SUBSTANCES Certain strains of all types of C. perfringens produce a hemagglutinin. Numerous other factors have been described and referred to as fibrinolysin, thyroid stimulating factor (Macchia et al., 1967), bursting, factor, hemolytic (non-alpha, theta, or delta) factor, and circulating factor. The activities (and in some instances the existence) of these factors are either unknown, in question, or in doubt. Generally speaking, the studies that have been done with these substances were carried out many years ago in the absence of sophisticated modern techniques and standards. There may be considerable value to reinvestigation of these components in contemporary laboratories. Numerous clinical conditions other than those already mentioned, have been attributed to C. perfrinoens though it seems that little attention has been paid by clinicians to identification of the type or strain of the organisms causing the diseases. Clostridial septicemia has been described in association with postpartum and postabortal infection (Bornstein et al., 1964; MacLennon, 1962), traumatic wound infections (MacLennon, 1962), and neoplastic disease (Wynne and Armstrong, 1972). Less common clinical syndromes are septic arthritis (Schiller et al., 1979), pneumonia and empyema (Bayer et al., 1975), meningitis (MacKay et al., 1971), corneal ulcer (Stern et al., 1979), myonecrosis (Mohr et al., 1978), necrotising enterocolitis in infants (Kosloske et al., 1978), and cystitis (Maliwan, 1979). Animal diseases such as focal abscess and intestinal disease in horses (MacKay et al., 1979; Wierup, 1977) and myositis in marine mammals (Greenwood and Taylor, 1978) have been reported as well. C. perfringens is a ubiquitous organism that emerges from time to time as a confirmed or suspected etiologic agent in a variety of diseases in animals and man. Surprisingly little is known about ecological and physiological factors that cause the organism to move from being a passive resident to an active pathogen. Furthermore, relatively little is known about the numerous toxins produced by the organism and their role in pathogenicity. Clinicians traditionally have paid minimal attention to C. perfringens in the stools of patients with intestinal disorders. Reports abound where 'clostridia' have been isolated without regard for identifying the species, and C. perfringens has been identified without regard for type, or toxins produced. In the late 1960's and early 1970's the number of reported outbreaks of C. perfringens food poisoning rose to unprecedented levels because public health officials spent more time looking for this organism as a cause of disease previously categorized as 'etiology unknown'. It is likely that a similar change in incidence of other intestinal disorders of unconfirmed etiology would occur if C. perfringens were given more serious consideration as the responsible agent. Basic science has equally ignored the toxins' physical and chemical properties, and modes of action. Hopefully more attention will be paid in the future to the possible if not probable role of C. perfringens toxins in intestinal disease ranging from simple diarrhea to necrotizing enterocolitis. Acknowledgements--This work was supported in part by Public Health Service Grant AI 14161-02 from the National Institute of Allergy and Infectious Diseases and by funds from the Agricultural Experiment Station, the Pennsylvania State University.

REFERENCES AKAMA, K., OTANI,S. and KAMEYAMA,S. (1968) Purification of beta toxin of Clostridium perfringens type C. Jap. J. Med. Sci. Biol. 21: 423-426.

646

JAMES L. MCDONEL

ALOUF, J. E. (1977) Cell membranes and cytolytic bacterial toxins. In: The specificity and action of animal, bacterial and plant toxins. Receptors and Recognition. pp. 219-270. CUATKECASAS,P. (Ed.) Series B, Vol. 1. John Wiley and Sons, N.Y. AL-SrlEXKHLY,F. and TRUSCOTT, R. B. (1977a) The pathology of necrotic enteritis of chickens following infusion of broth cultures of Clostridium perfringens into the duodenum. Arian Dis. 21 : 230-240. AL-SHEIKHLY, F. and TRUSCOTT, R. B. (1977b) The pathology of necrotic enteritis of chickens following infusion of crude toxins of Clostridium perfringens into the duodenum. Avian Dis. 21: 241-255. AL-SHEIKHLY, F. and TRUSCOTT, R. B. (1977c) The interaction of Clostridium perfringens and its toxins in the production of necrotic enteritis of chickens. Avian Dis. 21: 256-263. AVIGAD, G. (1976) Microbial Phospholipases. In: Mechanisms in bacterial toxinology, pp. 99-167. BERNHEIMER, A. W. (Ed.). John Wiley and Sons, N.Y. AVIGAD, L. S. and BERNHEIMER, A. W. (1976) Inhibition by zinc of hemolysis induced by bacterial and other cytolytic agents. Infect. lmmun. 13: 1378-1381. BANGnAM, A. D. and DAWSON, R. M. C. (1961) Electrokinetic requirements for the reaction between Cl. perfringens ~-toxin (phospholipase C) and phospholipid substrates. Biochim. Biophys. Acta 59: 103-115. BARTLETT, M. L., WALKER, H. W. and ZIPRIN, R. (1972) Use of dogs as an assay for Clostridium perfringens enterotoxin. Appl. Microbiol. 23: 196-197. BAYER, A. S., NELSON, S. C., GALPIN, J. E., CHOW, A. W. and GUZE, L. B. (1975) Necrotizing pneumonia and empyema due to Clostridium perfringens. Report and review of the literature. Am. J. Med. 59: 851-856. BEn, K. J. and BUTTERY, S. H. (1978) Reverse phase passive haemagglutination and single radial immunodiffusion to detect epsilon antigen of Clostridium perfringens type D. Austr. Vet. J. 54: 541-544. BERNATH, S. (1975) Solid phase radioimmunoassays for quantitative antibody determination of Clostridium perfringens type D epsilon toxin. Appl. Microbiol. 30: 499-502. BERNHEIMER, A. W. (1974) Interactions between membranes and cytolytic bacterial toxins. Biochim. Biophys. Acta 344: 27-50. BERNEIEIMER, A. W. (1976) Sulfhydryl activated toxins. In: Mechanisms in bacterial toxinology, pp. 85-97. BERNHEIMER, A.W. (Ed.). John Wiley and Sons, N.Y. BERNHEIMER, A. W. and GRUSHOFF, P. (1967) Cereolysin: Production, purification and partial characterization. J. Gen. Microbiol. 46: 143-150. BERNHEIMER, A. W., GRUSHOFF, P. and AVIGAD,U S. (1968) Isoelectric analysis of cytolytic bacterial proteins. J. Bacteriol. 95: 2439-2441. BHOWN, A. S. and HABEES, A. F. S. A. (1977) Structural studies on epsilon prototoxin of Clostridium perfringens type D localization of the site of tryptic scission necessary for activation to epsilon toxin. Biochem. Biophys. Res. Commun. 78: 889-896. BICKEORD, A. A. (1975) Comments on ulcerative enteritis. Am. d. Vet. Res. 36: 586. BIRD, R. A., LOW, M. G. and STEPHEN, J. (1974) Immunopurification of phospholipase C (or-toxin) from CIostridium perfringens. FEBS Lett. 44: 279-281~ BOETHIUS, J., RYDQUIST, B., M6LLBY, R. and WADSTR6M, T. (1973) Effect of a highly purified phospholipase C on some electrophysiological properties of the frog muscle fiber membrane. Life Sci. 13: 171-176. BORNSTEIN, D. L., WEINBERG, A. N., SWARTZ, M. N. and KUNZ, L. J. (1964) Anaerobic infections. Review of current experience. Medicine 43: 207-232. BOSWORTH, T. J. (1943) On a new type of toxin produced by Clostridium welchii. J. Comp. Pathol. 53: 245-255. BROOKS, M. E., STERNE, M. and WARRACK, G. H. (1957) A reassessment of the criteria used for type differentiation of Clostridium perfringens. J. Pathol. Bacteriol. 74: 185-195. BUDDLE, M. B. (1954) 'Degraded' strains of CIostridium welchii type C isolated from sheep in New Zealand. J. Comp. Path. TheE. 64: 217-224. BULLEN, J. J. (1970) Role of toxins in host-parasite relationships. In: Microbial Toxins. p. 233. SJL, S. J., KADIS, S. and MONTIE, T. C. (Eds,) Vol. I. Academic Press, N.Y. BULLEN, J. J. and BATTY,I. (1956) The effect of Clostridium welchii type D culture filtrates on the permeability of the mouse intestine. J. Path. Bact. 71: 311-323. BUXTON, D. (1976) Use of horseradish peroxidase to study the antagonism of CIostridium welchii (Clostridium perfringens) type D epsilon toxin in mice by the formalinized epsilon prototoxin. J. Comp. Pathol. 86: 67-72. BUXTON, D. (1978a) The use of an immunoperoxidase technique to investigate by light and electron microscopy the sites of binding of Clostridium welchil type D epsilon toxin in mice. J. Med. Microbiol. 11: 289-292. BUXTON, D. (1978b) Further studies on the mode of action of Clostridium welchii type D epsilon toxin. J. Med. Microbiol. 11 : 293-298. BUXTON, D. (1978c) In vitro effects of CIostridium welchii type D epsilon toxin on guinea pig, mouse, rabbit and sheep cells. J. Med. Microbiol. 11 : 299-302. BUXTON, T. and MORGAN, K. T. 1976. Studies of lesions produced in the brains of colostrum deprived lambs by Clostridium welchii (CIostridium perfringens) type D toxin. J. Comp. Pathol. 86: 435-447. CAaEXAS, M. (1978) Determination of the inhibitory effect of several compounds on neuraminidases from virus influenza, V.. cholerae and Cl. perfringens. Int. J. Biochem. 9: 47-49. CASU, A., PALA, V., MONACELLI, R. and NANNI, G. (1971) Phospholipase C from Clostridium perfringens: purification by electrophoresis on acrylamide-agarose gel. ltal. J. Biochem. 20: 166-178. CHAKRABARTY,A. K. and NARAYAN, K. G. (1979) Pathogenesis of Hobbs' heat-sensitive spore forming Clostridum perfringens type A strain. Microbiol. Immunol. 23: 213-221. CHAKRABARTY, A. K., NARAYAN, K. G. and CHADIRAMANI,N. K. (1977) Association of CIostridium perfringens type A with human diarrhoeal cases. Ind. J. Med. Res. 65: 495-499. CBIEN, S.-F., YEVlCn, S. J., L1, S.-C. and LI, Y.-T. (1975) Presence of endo fl-N-actylglucosaminidase and protease activities in the commercial neuraminidase preparations isolated from Clostridium perfringens. Biochem. Biophys. Res. Commun. 65: 683-691.

Clostridium perfringens toxins (Type A, B, C, D, E)

647

CHOUDHURY,T. K. (1978) Synergistic lysis of erythrocytes by Propionibacterium acnes. J. Clin. Microbiol. 8: 238-241. COLLEE, J. G. (1965) A myxovirus receptor-inactivating agent occurring in cultures of Clostridium welchii. J. Pathol. Bacteriol. 90: 1-11. COLLEE, J. G. (1974) Clostridium perfringens (CI. welchii) in the human gastrointestinal tract. In: The Normal Microbial Flora of Man, pp. 205-219. SKINNER,F. A. and CARR, J. G. (Eds.). Academic Press, N.Y. COWELL,J. L,, KIM, K. S. and BERNHEIMER,A. W. (1978) Alteration by cereolysin of the structure of cholesterolcontaining membranes. Biochem. Biophys. Acta 507: 230-241. COWlE-WHITNEY, J. (1970) Some aspects of the enteritis complex of rabbits. In: Nutrition and disease in experimental animals, pp. 122-131. TAVERNOR,W. D. (Ed.). Williams and Wilkins Co., Baltimore. CRAIG, J. P. and MILES,A. A, (1961) Some properties of the iota-toxin of CIostridium welchii, including its action on capillary permeability. J. Path. Bact. 81: 481-493. DALLING,T. and Ross, H. E. (1938) Clostridium welchii: Notes on the relationship between the types of cultures and the production of toxin. J. Comp. Path. Ther. 51 : 235-249. DAWSON, R. M. C. (1973) Specificity of enzymes involved in the metabolism of phospholipids. In: Form and function ofphospholipids, pp. 97-116. ANSELL,G, B., HAWTHORNE,J. N. and DAWSON, R. M. C. (Eds.). Elsevier, Amsterdam. DAWSON, R. M. C., HEMINGTON,N. L., MILLER, N. G. A. and BANGHAM,A. D. (1976) On the question of electrokinetic requirement for phospholipase C action. J. Membrane Biol. 29: 179-184. DEMEL, R. A., GUERTSVANKESSEL,W. S. M., ZWALL, R. F. A., ROELOFSEN,B. and VAN DEENEN,L. L, M. (1975) Relation between various phospholipase actions on human red cell membranes and the interfacial phospholipid pressure in monolayers. Biochim. Biophys. Acta 406: 9%107. DEN, H., MALINZAK, D. A. and ROSENBERG,A. (1975) Cytotoxin contaminants in commercial Clostridium perfringens neuraminidase preparations purified by ali~nity chromatography. J. Chromat. 111: 217-222. DOURMASHK1N,R. R. and ROSSE, W. F. (1966) Morphologic changes in the membranes of red blood cells undergoing hemolysis. Am. J. Med. 41: 699-709. DUNCAN, C. L. (1970) Clostridum perfringens food poisoning. J. Milk Food Technol. 33: 35-41. DUNCAN, C. L. (1973) Time of enterotoxin formation and release during sporulation of Clostridium perfringens type A. J. Bacteriol. 113: 932-936. DUNCAN,C. L. (1975) Role of clostridial toxins in pathogenesis. In: Microbiology~1975. pp. 283-291. SCHLEssINGER, D. (Ed.). American Society for Microbiology, Washington, D.C. DUNCAN,C. L., KING, G. J. and FR1EBEN,W. R. (1973) A paracrystalline inclusion formed during sporulation of enterotoxin-producing strains of Clostridium perfringens type A. J. Bacteriol. 114: 845-859. DUNCAN, C. L., ROKOS, E. A., CI~RISTENSON,C, M. and ROOD, J. I. (1978) Multiple plasmids in different toxigenic types of Clostridium perfringens possible control of beta toxin production. In: Microbiology 1978. pp. 246-248. SCHLESSINGER,D. (Ed.). American Society for Microbiology, Washington, D.C. DUNCAN, C. L. and SO~ERS, E. B. (1972) Quantitation of Clostridium perfringens type A enterotoxin by electroimmunodiffusion. Appl. Microbiol. 24: 801-804. DUNCAN, C. L. and STRONG,D. H. (1968) Improved medium for sporulation of Clostridium perfrin#ens. Appl. Microbiol. 16: 82-89. DUNCAN,C. L. and STRONG,D. H. (1969a) Ileal loop fluid accumulation and production of diarrhea in rabbits by cell-free products of Clostridium perfrinyens. J. Bacteriol. 100: 86-94. DUNCAN, C. L. and STRONG,D. H. (1969b) Experimental production of diarrhea in rabbits with Clostridium perfringens. Can. J. Microbiol. 15: 765-770. DUNCAN, C. L. and STRONG,D. H. (1971) Clostridium perfringens type A food poisoning. 1. Response of the rabbit ileum as an indication of enteropathogenicity of strains of Clostridium perfringens in monkeys. Infect. Immun. 3: 167-170. DUNCAN, C. L., STRONG,D. H. and SEaALD, M. (1972) Sporulation and exterotoxin production by mutants of Clostridium perfringens. J. Bacteriol. 110: 378-391. DUNCAN, C. L., SUGIYAMA,H. and STRONG, D. H. (1968) Rabbit ileal loop response to strains of Clostridum perfringens. J. Bacteriol. 95: 1560-1566. ELLNER, P. D. (1961) Fate of partially purified C14-1abelled toxin of Clostridium perfrinoens. J. Bacteriol. 82: 275-283. ENDERS, G. L., JR. and DUNCAN, C. L. (1976) Anomalous aggregation of Clostridium perfringens enterotoxin under dissociating conditions. Can. J. Microbiol. 22: 1410-1414. ENDERS, G. L., JR. and DUNCAN, C. L. (1977) Preparative polyacrylamide gel electrophoresis purification of Clostridium perfringens enterotoxin. Infect. and lmmun. 17: 425-429. FLATT, R. E., WEISBROTla,S. H. and KRAUS, A. L. (1974) Metabolic, traumatic, mycotic, and m{scellaneous diseases of rabbits. In: The biology of the laboratory rabbit, pp. 437-440. WEISaROTn, S. H., FLATT, R. E. and KRAUS,A. L. (eds.). Academic Press, N.Y. FRASER, A. G. (1978) Neuraminidase production by Clostridia. J. Med. Microbiol. 11: 269-280. FRASER,A. G. and COLLEE,J. G. (1975) The production of neuraminidase by food-poisoning strains of Clostridium wellchii (C. perfringens). J. Med. Microbiol. 8: 251-263. FREER,J. H. and ARaUTHNOTr,J. P. (1976) Biochemical and morphologic alterations of membranes by bacterial toxins. In: Mechanisms of bacterial toxinology, pp. 169-193. BERNrlEIMER,A. W. (Ed.). John Wiley and Sons, N.Y. FREHOLM, B. B., Mt)LLBY, R., ANDERSON, P., MALblQUIST, T. and SMYTH, C. J. (1978) Inhibition of noradrenaline-stimulated lipolysis and cyclic AMP accumulation in isolated rat adipocytes by purified phospholipase C and theta-toxin from Clostridium perfringens. Acta Pharmacol. Toxicol. 42: 23-34. FRIEBEN,W. R. and DUNCAN,C. L. (1973) Homology between enterotoxin protein and spore structural protein in Clostridium perfringens type A. Eur. J. Biochem. 39: 393-3401. FRIEBEN, W. R. and DUNCAN, C. L. (1975) Heterogeneity of enterotoxin-like protein extracted from spores of Clostridium perfringens type A. Eur. J. Biochem. 55: 455-463.

648

JAMESL. McDONEL

FRISH, A., GAZITT, Y. and LOYTER, A. (1973) Metabolically controlled hemolysis of chicken erythrocytes. Biochim. Biophys. Acta 291: 690-700. GARDNER, D. E. (1973a) Pathology of Clostridium welchii type D enterotoxaemia. I. Biochemical and haematological alterations in lambs. J. Comp. Pathol. 83: 499-507. GARDNER, D. E. (1973b) Pathology of Clostridium welchii type D enterotoxaemia. III. Basis of the hyperglycaemic response. J. Comp. Pathol. 83: 525-529. GARDNER, D. E. (1974) Brain oedema: an experimental model. Br. J. Expt. Pathol. 55: 453-457. GENIGEORGlS, C. (1975) Public health importance of Clostridium perfringens. J. Am. Vet. Med. Assoc. 167: 821-827. GENIGEORGIS, C., SAKAGUCrlI, G. and RIEMANN, H. (1973) Assay methods for Clostridium perfringens type A enterotoxin. Appl. Microbiol. 26: 111-115. GIGER, O. and PARIZA, M. (1978) Depression of amino acid transport in cultured rat hepatocytes by purified enterotoxin from Clostridium perfringens. Biochem. Biophys. Res. Commun. 82: 378-383. GILL, D. M. and KING, C. A. (1975) The mechanism of action of cholera toxin in pigeon erythrocyte lysates. J. Biol. Chem. 250: 6425--6432. GLUSnKOVA, A. I., NENASHEAV,V. P. and GOLDSHMID, V. K. (1969) A study of nutrient medium and the toxin of Cl. perfringens type A by means of gel filtration through Sephadex gel. In: Proceedinos of the Inter-Institute Conference, Stavropol, 1966. In Russian, cited by Smyth and Arbuthnott, 1974. GRANUM, P. E. and SK.IELKVALI~,R. (1977) Chemical modification and characterization of enterotoxin from Clostridium perfringens type A. Acta Pathol. Microbiol. Scand. Sect. B 85: 89-94. GREENWOOD, A. G. and TAYLOR, D. C. (1978) Clostridial myositis in marine mammals. Vet. Record. 103: 54-55. GRXNER, L. A. (1961) Enterotoxemia of sheep. I. Effect of Clostridium perfringens type D toxin on the brains of sheep and mice. Am. J. Vet. Res. 22: 429-442. GRINER, L. A. and CARLSON,W. D. (1961) Enterotoxemia of sheep. II. Distribution of 131I-radioiodinated serum albumin in brains of Clostridium perfringens type D intoxicated lambs. Am. J. Vet. Res. 22: 443-446. GUBASH, S. M. (1978) Synergistic hemolysis phenomenon shown by an alphatoxin-producing CIostridium perfringens and streptococcal CAMP factor in presumptive streptococcal grouping. J. Clin. Microbiol. 8: 480-488. GYOBU, Y. and KODAMA, H. (1978) Vero cell microcell culture technique for the detection of neutralization antibody against Clostridium perfrinc,lens enterotoxin and its application to sero-epidemilogy. J. Food Hyg. Soc. Jap. 19: 294-298. HAREEB, A. F. S. A. (1964) Studies on ~-prototoxin of Clostridium perfrinoens type A. Can. J. Biochem. 42: 545-554. HABEEB, A. F. S. A. (1969) Studies on E-prototoxin of Clostridium perfringens type D. I. Purification methods: evidence for multiple forms of E-prototoxin. Arch. Biochem. Biophys. 130: 430-440. HABEEB, A. F. S. A. (1972) Chemical and physicochemical properties of E-prototoxin of Clostridium perfrinfens type D. Fed. Proc. 31: 932. HABEEB, A. F. S. A. (1975) Studies on ~-prototoxin of Clostridium perfringens type D physicochemical and chemical properties of E-prototoxin. Bioehim. Biophys. Acta 412: 62-69. H~EEB, A. F. S. A., LEE, C. L. and ATASSl, M. Z. (1973) Conformational studies on modified proteins and peptides, Part 7. Conformation of epsilon prototoxin and epsilon toxin from Clostridium perfringens, conformational changes associated with toxicity. Biochim. Biophys. Acta 322: 245-250. H~ERMANN, E. (1960) Zur Toxikologie und Pharmakologie des Gasbrandgiftes (Clostridium welchii type A) und seiner Komponenten. Naunyn-Schmiedeberg's Arch. Exp. Path. Pharmak. 238: 502-524. HALE, C. W. (1944) Studies on diffusing factors 5. The influence of some environemtnal conditions on the activity of hyaluronidase. Biochem. J. 38: 368-370. HALL, H. E., ANGELOTTi, R., LEWIS, K. H. and FOTER, M. J. (1963) Characteristics of Clostridium perfringens strains associated with food and food-borne disease. J. Bacteriol. 85: 1094-1103. HARTLEY, W. J. (1956) A focal symmetrical encephalomalacia of lambs. N.Z. Vet. J. 4: 129-135. HASE, J., MITSUI, K. and SHONAKA, E. (1975) Clostridium perfringens exotoxins. III. Binding of 0-toxin to erythrocyte membrane, Japan. J. Exp. Med. 45: 433-438. HASE, J., MITSUI, K. and SnONAKA, E. (1976) Clostridium perfringens exotoxins. IV. Inhibition of the 0-toxin induced hemolysis by steroids and related compounds. Japan. J. Exp. Med. 46: 45-50. HAUSCHILD, A. H. W. (1970) Erythemal activity of the cellular enteropathogenic factor of Clostridium perfringens type A. Can. J. Microbiol. 16: 651-654. HAUSCHILD, A. H. W. (1971) Clostridium perfringens toxins types B, C, D, and E. In: Microbial Toxins. Vol. IIA. Bacterial Protein Toxins. pp. 159-188. KADIS, S., MONTIE, T. C. and AJL, S. J. (Eds.). Academic Press, N.Y. HAUSCHILD, A. H. W. (1973) Food poisoning by Clostridium perfringens. Can. Inst. Food Sci. 7?echnol. J. 6: 106-110. HAUSCH1LD, A. H. W. and HILSHE1MER, R. (1971) Purification and characteristics of the enterotoxin of Clostridium perfringens type A. Can. J. Microbiol. 17: 1425-1433. HAUSCHILD, A. H. W., H1LSHEIMER,R. and MARTIN, W. G. (1973) Improved purification and futher characterization of Clostridium perfringens type A enterotoxin. Can. J. Microbiol. 19: 1379-1382. HAUSCHILD, A. H. W., HILSHEIMER, R. and ROGERS, C. G. (1971) Rapid detection of Clostridium perfringens enterotoxin in a modified ligated intestinal loop technique in rabbits. Can. J. Microbiol. 17: 1475-1476. HAUSCHILD, A. H. W., LECROISEY,A. and ALOUE, J. E. (1973) Purification of Clostridium perfringens type C theta toxin. Can. J. Microbiol. 19: 881-885. HAUSCHILD, A. H. W., NnLo, L. and DORWARD, W. J. (1967) Experimental enteritis with food poisoning and classical strains of CIostridium perfringens type A in lambs. J. Infect. Dis. 117: 379-386. HAUSCHILD, A. H. W., NIILO, L. and DORWARD, W. J. (1970a) Enteropathogenic factors of food-poisoning Clostridium perfringens type A. Can. J. Microbiol. 16: 331-338. HAUSCHILD, A. H. W., NIlLO, L. and DORWARD, W. J. (1970b) Response of ligated intestinal loops in lambs to an enteropathogenic factor of Clostridium perfringens type A. Can. J. Microbiol. 16: 339-343.

Clostridium perfringens toxins (Type A, B, C, D, E)

649

HAUSCHILD, A. H. W. and THATCHER, F. S. (1967) Experimental food poisoning with heat-susceptible Clostridium perfringens type A. J. Food Sci. 32: 467-469. HAUSCHILD, A. H. W. and THATCHER, F. S. (1968) Experimental gas gangrene with food-poisoning heatsusceptible Clostridium perfringens type A. Can. J. Microbiol. 14: 705-709. HAUSCHILD, A. H. W., WALCROFT, M. J. and CAMPBELL,W. (1971) Emesis and diarrhea induced by enterotoxin of Clostridium perfringens type A in monkeys. Can. J. Microbiol. 17:1141-1143. HILL, E. E. and LANDS, W. E. M. (1970) Phospholipid metabolism. In: Lipid metabolism, pp. 185-277. WAKIL, S. J. (Ed.). Academic Press, N.Y. HITCHCOCK, C. R., DEMELLO,F. J. and HAGLIN, J. J. (1975) Gangrene infection, new approaches to an old disease. Surg. Clin. N. Am. 55: 1403-1410. HOBBS, B. C. (1969) Clostridium perfringens and Bacillu~ cereus infections. In: Food-borne infections and intoxications, pp. 131-173. RIEMAN, H. (Ed.). Academic Press, N.Y. HOBBS, B. C., SMITH, S. F., OAKLEY,C. T., WARRACK,G. H. and CRUICHSHANK, J. F. (1953) Clostridium welchii food poisoning. J. Hyg. 51: 74-101. HOGH, P. (1976) Experimental studies on serum treatment and vaccination against Cl. perfringens type C infection in piglets. Dev. Biol. Stand. 32: 69-76. HOWARD, J. G., WALLACE,K. R. and WRIGHT, G. P. (1953) The inhibitory effects of cholesterol and related sterol on heamolysis by streptolysin O. Br. J. Exp. Pathol. 34: 174-180. HYDER, G. (1973) Enterotoxemia or pulpy kidney disease vaccine. Indian Vet. J. 50: 972-978. ILLS, G. H., SEEMAN,P., NAYLOR, D. and CINADER, B. (1973) Membrane lesions, immune lysis surface rings, globule aggregates and transient openings. J. Cell. Biol. 56: 528-539. INOUE, K., AKIYAMA,Y., KINOSHITA, T., HIGASHI, Y. and AMANO, T. (1976) Evidence for a one-hit theory in the immune bactericidal reaction and demonstration of a multi-hit response for hemolysis by streptolysin O and Clostridium perfringens theta-toxin. Infect. Immun. 13: 337-344. ISPOLATOVSKAYA,M. V. (1971) Type A Clostridium perfringens toxin. In: Microbial Toxins. Vol. IIA. Bacterial Protein Toxins. pp. 109-158. KAD1S, S., MONTIE, T. C. and AJL, S. J. (Eds.). Academic Press, N.Y. IVERS, J. T. and POTTER, N. N. (1977) Production and stability of hemolysin, phospholipase C, and lethal toxin of Bacillus cereus in food. J. Food Prot. 40: 17-21. JANSEN, B. C. (1960) The diagnosis of pulpy kidney disease. J. S. African Vet. Med. Assoc. 31: 15-20. JANSEN, B. C. (1967a) The production of a basic immunity against pulpy kidney disease. Onderstepoort. J. Vet. Res. 34:65 80. JANSEN, B. C. (1967b) The duration of immunity to pulpy kidney disease of sheep. Onderstepoort. J. Vet. Res. 34: 333-334. KAMEYAMA,S. and AKAMA,K. (1970) Titration of kappa toxin of Clostridium perfringens and kappa antitoxin by a modified azocoll method. Japan. J. Med. Sci. Biol. 23: 31-46. KAMEYAMA, S. and AKAMA, K. (1971) Purification and some properties of kappa toxin of Clostridium perfringens. Japan. J. Med. Sci. Biol. 24: 9-23. KAMEYAMA, S., SATO, H. and MURATA, R. (1975) The role of g-toxin of Clostridium perfringens in experimental gas gangrene in guinea pigs. Japan. J. Med. Sci. Biol. 25: 200. KELLAWAY, C. H. and TRETHEWIE,E. R. (1941) The injury of tissue cells and the liberation of pharmacologically active substances by the toxins of Cl. welchii types B and C. Aust. J. Exp. Biol. Med. Sci. 19:17 27. KENNEDY, K. K., NORRIS~ S. J., BECKENHAUER,W. H. and WHITE, R. G. (1977) Clostridium perfringens type C and type D toxoid. Am. J. Vet. Res. 38: 1515-1518. KENT, C. (1978) Inhibition of muscle cell fusion by phospholipase C alterations in phospholipid metabolism. Fed. Proc. 37: 1404. KLEIN, R. (1975) The action of phospholipase C from Clostridium welchii by quinine: an absolute requirement for calcium ions. Chem. Phys. Lipids 15: 15-26. KOSLOSKE, A. M., ULRICH, J. A. and HOFFMAN, H. (1978) Fulminant necrotising enterocolitis associated with clostridia. Lancet ii 1014-1016. KR.~MER, J. and SCHALLEHN, G. (1978) Characterization of plasmid DNA in a lecithinase-positive and in a lecithinase-negative strain of Clostridium perfringens. Zbl. Bakt. Hyg. I. Abt. Orig. A 241 : 438-447. KULSHRESTHA, S. B. (1974a) Effect of carbohydrates on the production of lethal toxin by Clostridium perfringens types B and C with particular reference to beta toxin and epsilon toxin. Indian. J. Anita. Sci. 43: 6284531. KULSHRESTHA, S. B. (1974b) Effect of period of incubation and pH on the production of beta and epsilon toxins by Clostridium welchii type B and type C. Indian. J. Anim. Sci. 43: 987-990. KURIOKA, S. and MATSUDA, M. (1976) Phospholipase C assay using p-nitrophenylphosphorylcholine togeth& with sorbital and its application to studying the metal and detergent requirement of the enzyme. Anal. Biochem. 75: 281-289. LABBE, R. G. and DUNCAN, C. L. (1974) Sporulation and enterotoxin production by Clostridium perfringens type A under conditions of controlled pH and temperature. Can. J. Microbiol. 20: 1493-1501. LAaBE, R. G. and DUNCAN, C. U (1975) Influence of carbohydrates on growth and sporulation of Clostridium perfringens type A. Appl. Microbiol. 29: 345-351. LARBE, R. G. and DUNCAN, C. L. (1977a) Evidence for stable messenger ribonucleic acid during sporulation and enterotoxin synthesis by Clostridium perfringens type A. J. Bacteriol. 129: 843-849. LABBE, R. G. and DUNCAN, C. L. (1977b) Spore coat protein and entero-toxin synthesis in Clostridium perfrinoens. J. Bacteriol. 131: 713-715. LAMONT, J. T., SONNENBLICK, E. B. and ROTHMAN, S. (1979) Role of Clostridial toxin in the pathogenesis of clindamycin colitis in rabbits. Gastroenterology 76: 356-361. LAWRENCE, G. and WALKER, P. D. (1976) Pathogenesis of enteritis necroticans in Papua New Guinea. Lancet. i: 125-126. LIEBERMAN, M. (1975) A highly efficient solid phase radioimmunoassay for a clostridial toxin. Anal. Biochem. 67: 115-121.

650

JAMES L. MCDONEL

LITVINKO, N. M., KHURGIN, Y. I. and SHEMANOVA, G. F. (1977) Effect of low molecular weight substrate fragments and their analogs on activity of isoenzymes of phospholipase C E.C.-3.1.4.3 from Clostridium perfringens type A. Biochemistry (Eng. Transl. Biokhimiya) 42: 838-842. Low, D. K. R. and FREER, J. H. (1977) Biological effects of highly purified fl-lysin (sphingomyelinase C) from Staphylococcus aureus. FEMS Microbiol. Lett. 2: 133-138. Low, D. K. R., FREER, J. H., ARBUTHNOTT, J. P., MOLLBY, R. and WADSTROM, T. (1974) Consequences of sphingomyelin degradation in erythrocyte ghost membranes by staphylococcal fl-toxin (sphingomyelinase C). Toxicom. 12:279 285. MACCHIA, V., BATES,R. W. and PASTAN,I. (1967) The purification and properties of a thyroid-stimulating factor isolated from Clostridium perfringens. J. Biol. Chem. 242: 3726-3730. MACFARLANE, M. G. and KNIGHT, B. C. J. G. (1941) The biochemistry of bacterial toxins. I. The lecithinase activity of Cl. welchii toxins. Biochem. J. 35: 884-902. MACKAY, N. N. S., GRUNEBERG, R. N., HARRIES, B. J. and THOMAS, P. K. (1971) Primary Clostridium welchii meningitis. Br. Med. J. l: 591-592. MACKAY, R. J., CARLSON,G. P. and HIRSH, D. C. (1979) Clostridium perfringens associated with a focal abscess in a horse. J. Am. Vet. Med. Assoc. 175: 71-72. MACLENNON, J. D. (1962) Histotoxic clostridial infections of man. Bact. Rev. 26: 177-276. MALIWAN, N. (1979) Emphysematous cystitis associated with Clostridium perfringens bacteremia. J. Urol. 121: 819-820. MANCINI, G., CARBONARA, A. O. and HEREMANS, J. F. (1965) Immunochemical quantitation of antigens by single radial immunodiffusion. Immunochem. 2: 235-254. MARUCCI, A. A. and FULLER, T. C. 0971) Quantitative micro-compliment fixation test. Appl. Microbiol. 21: 260--264. MASORO, E. J. (1968) Physiological chemistry of lipids in mammals. Saunders Co., Philadelphia. MATSUOA, M. and SUGIMOTO, N. (1979) Calcium-independent and dependent steps in action of Clostridium perfringens enterotoxin on Hela and Vero cells. Biochem. Biophys. Res. Commun. 91 : 629-636. MCCARTNEY, C. and ARBUTrlNOTT, J. P. (1978) Mode of action of membrane-damaging toxins produced by staphylococci. In: Bacterial toxins and cell membranes, pp. 89-127. JELJASZEWICZ, J. and WADSTR6M, T. (Eds.). MCCLANE, B. A. and McDONEL, J. L. (1979) The effects of CIostridium perfringens enterotoxin on morphology, viability, and macromolecular synthesis in vero cells. J. Cell. Physiol. 99: 191-200. MCCLANE, B. A. and McDONEL, J. L (1980) Characterization of membrane permeability alterations induced in Vero cells by Clostridium perfringens enterotoxin. Biochim. Biophys. Acta. (In press). McCLUNG, L. S. (1945) Human food poisoning due to growth of Clostridium perfringens (C. welchii) in freshly cooked chicken. J. Bacteriol. 50: 229-231. MCCREA, J. F. (1947) Modification of red-cell agglutinability by Cl. welchii toxins. Aust. J. Expt. Biol. Med. Sci. 25: 127-136. MCDONEL, J. L. (1974) In vivo effects of Clostridium perfringens enteropathogenic factors on the rat ileum. Infect. Immun, 10: 1156-1162. MCDONEL, J. L. (1979) The molecular mode of action of Clostridium perfringens enterotoxin. Am. J. Clin. Nutr. 32: 210-218. M cDONEL, J. L. (1980) Mechanism of action of Clostridium perfringens enterotoxin. J. Food Technol. 34:91-95. MCDONEL, J. L. and ASANO, T. (1975) Analysis of unidirectional sodium fluxes during diarrhea induced by Clostridium perfringens enterotoxin in the rat terminal ileum. Infect. Immun. 11 : 526-529. McDONEL, J. L., CHANG, L. W., POUNDS, J. G. and DUNCAN, C. L. (1978) The effects of Clostridium perfringens enterotoxin on rat and rabbit ileum: an electron microscopic study. Lab. Invest. 39: 210-218. MCDONEL, J. L. and DUNCAN, C. L. (1975a) Effects of Clostridium perfringens enterotoxin on metabolic indexes of everted rat ileal sacs. Infect. Immun. 12: 274-280. McDONEL, J, L. and DUNCAN, C. L. (1975b) Mechanism of action of Clostridium perfringens enterotoxin in the rat ileum: protein synthesis and cycli 3'-5' adenosine monophosphate. Proceeding of the 1lth Joint Conference U.S.-Japan Cooperative Medical Science Program, Cholera Panel, pp. 297-305. McDONEL, J. L. and DUNCAN, C. L. (1975c) Histopathological effects of Clostridium perfringens enterotoxin in the rabbit ileum. Infect. Immun. 12: 1214-1218. MCDONEL, J. L. and DUNCAN, C. L. (1976) Effect of Clostridium perfringens enterotoxin on mitochondrial respiration. Infect. Immun. 15: 999-1001. MCDONEL, J. L. and DUNCAN, C. L. (1977) Regional localization of activity of Clostridium perfringens type A enterotoxin in the rabbit ileum. J. Infect. Dis. 136: 661-666. McDONEL, J. L. and MCCLANE, B. A. (1979) Binding versus biological activity of Clostridium perfringens enterotoxin in veto cells. Biochem. Biophys. Res. Commun. 87: 497-504. MCDONEL, J. L. and SMITH, W. P. (1979) An in vitro system to study enterotoxin synthesis and sporulation in Clostridium perfringens type A. Abstract. Abstr. Ann. Meeting, Am. Soc. Microbiol. p. 16. MEDUSKI, J. W. and VOLKOVA, M. S. (1958) Clostridium perfringens (welchii) inaktywowaniem promieniami gamma. Medycyna Dosw. Mikrobiol. 10: 247-254. (Biol. Abs. 1960, 35: 38506). M1HANCEA, N., BITTNER, J., TOACSEN, E., CEACURUANU,A. and TEODORESCU, G. (1970) Influence de la toxine perfringens sur la phagocytose et al viabilite leucocytarie. Arch. Roum. Pathol. Exp. Microbiol. 6: 159167. MITSUI, K., MITSUI, N, and HASE, J. (1973a) CIostridium perfringens exotoxins. I. Purification and properties of *t-toxin. Japan. J. Exp. Med, 43: 65-80. MlTSU1, K., MITSUI, N. and HASE, J. (1973b) Clostridium perfringens exotoxins. II. Purification and some properties of 0-toxin. Japan, J. Exp. Med. 43: 377-391. MITSOl, K. and HASE, J. (1979) Clostridium perfringens exotoxins. VI. Reactivity of perfringolysin O with thiol and disulfide compounds. Japan. J. Exp. Med. 49: 13-18. MiTsul, K., SEKIYA, T., NOZAWA, Y. and HASE, J. (1979) Alteration of human herythrocyte plasma membranes

Clostridium perfringens toxins (Type A, B, C, D, E)

651

by perfringolysin O as revealed by freeze-fracture electron microscopy. Studies on Clostridium perfringens exotoxins V. Biochim. Biophys. Acta 554: 68-75. MIWANTANI, T., HONDA, T., ARAI, S. and NAKAYAMA, A. (1978) Mode of lethal activity of Clostridium perffingens type A enterotoxin. Japan. J. Med. Sci. Biol. 31 : 221-223. MOHR, J. A., GRIFFXTHS,W., HOLM, R., GARCIA-MORAL, C. and FLOURNOY, D. J. (1978) Clostridial myonecrosis ('gas gangrene') during cephalosporin porphylaxis. J. Am. Med. Assoc. 239: 847-849. M6LLaY, R. (1973) Purification of two bacterial phosphotipases C and some effects on cell membranes. Ph.D Dissertation, Karolinska Institute, Stockholm. M6LLBY, R. (1976) Effect of staphylococcal beta-haemolysin (sphingo-myelinase C) on cell membranes. In: Staphylococci and Staphylococcal diseases, pp. 665-667. JELJASZEWICZ,J. (Ed.). Gustav Fischer Verlag, Stuttgart, N.Y. M6LLBY, R. (1978) Bacterial Phospholipases. In: Bacterial toxins and cell membranes, pp. 367-424. JELJASZEW1CZ, J. and WADS'rR6M, T. (Eds.). Academic Press, N.Y. M6LLBY, R,, HOLME, T., NORD, C.-E., SMYTH, C. J. and WADSTR6M, T. (1976) Production of phospholipase C (alpha toxin) haemolysis and lethal toxins by Clostridium perfringens types A to D. J. Gen. Microbiol. 96: 137 144. M6LLRY, R., NORD, C. E. and WADSTR6M, T. (1973) Biological activities contaminating preparations of phospholipase C (or-toxin) from Clostridium perfringens. Toxicon. I1: 139-147. M6LLaY, R., TrlELESTAM,M. and WADSTR6M, T. (1974) Effect of Clostridium perfringens phospholipase C (alpha toxin) on the human diploid fibroblast membrane. J. Membrane Biol. 16: 313-330. M6LLBY, R. and WADSTR6M, T. (1973) Purification of phospholipase C (a-toxin) from Clostridium perfringens. Biochim. Biophys. Acta 321: 569-584. MOORE, A., GOTTFRIED, E. L., STONE, P. H. and COLEMAN,M. (1976) Case report. Clostridium perfringens septicemia with detection of phospholipase C activity in the serum. Am. J. Med. Sci. 271: 59-63. MORGAN, K. T. and KELLY, B. G. (1974) Ultrastructural study of brain lesions produced in mice by the administration of Clostridium perfringens type D toxin. J, Comp. Pathol. 84: 181-191. MORGAN, K. T., KELLY, B. G. and BUXTON, D. (1975) Vascular leakage produced in the brains of mice by CIostridium welchii type D toxin. J. Comp. Pathol. 85: 461-466. NAGLER, F. P. O. (1939) Observations on a reaction between lethal toxins of Cl. welehii (Type A) and human serum. Br. J. Exp. Path. 20: 473-485. NAIK, H. S. and DUNCAN, C. L. (1977a) Enterotoxin formation of foods by CIostridium perfringens type A. d. Food Safety 1: 7-18. NAIK, H. S. and DUNCAN, C. L. (1977b) Rapid detection and quantitation of CIostridium perfringens enterotoxin by counterimmunoelectrophoresis. Appl. Environ. Microbiol. 34: 125-128. NAIK, H. S. and DUNCAN, C. U (1978a) Thermal inactivation of CIostridium perfringens enterotoxin. J. Food Prot. 41: 100-103. NAIK, H. A. and DUNCAN, C. L. (1978b) Detection of Clostridium perfringens enterotoxin in human fecal samples and anti-enterotoxin in sera. J. Clin. Microbiol. 7: 337-340. NIILO, L. (1965) Bovine enterot6xemia. Ill. Factors affecting the stability of the toxins of Clostridium perfringens types A, C and D. Can. Vet. J. 6: 38-42, NIILO, L. (1971) Mechanism of action of the enteropathogenic factor of Clostridium perfringens type A. Infect. Immun. 3: 100-106. NIILO, L. (1972) The effect of enterotoxin of Clostridium welchii (perfringens) on the systemic blood pressure of sheep. Res. Vet. Sci. 13: 503-505. NIILO, L. (1973a) Fluid secretory response of bovine Thiry jejunal fistula to enterotoxin of CIostridium perfringens. Infect. lmmun. 7: 1-4. NnLO, L. (1973b) Effect of calves of the intravenous injection of the enterotoxin of Clostridium perfringens type A. J, Comp. Pathol. 83: 265-269. NnLO, L. (1973c) Antigen homogeneity of enterotoxin from different agglutinating serotypes of Clostridium perfringens. Can. J. Microbiol. 19: 521-524. NIILO, L. (1974) Response of ligated intestinal loops in chickens to the enterotoxin of Clostridium perfringens. AppL Microbiol. 28: 889-891. NnLO, L. (1975) Measurement of biological activities of purified and crude enterotoxin of Clostridium perfringens. Infect. Immun. 12: 44(Y442. NnLO, L. (1976) The effect of enterotoxin of Clostridium welchii (perfringens) type A on fowls. Res. Vet. Sci. 20: 225-226. NImO, L. (1977) Enterotoxin formation by Clostridium perfringens type A studied by the use of fluorescent antibody. Can. J. Microbiol. 23: 908-915. NHLO, L. (1978) Efficacy of laboratory tests for the detection of enterotoxigenic Clostridium perfringens. Can. J. Microbiol. 24: 633-635. NnLO, L. (1979) Identification of enterotoxigenic Clostridium perfrinoens type A in mixed cultures. Can. J. Comp. Med. 43: 98-101. NnLO, L., HAUSCmLD, A. H. W. and DORWARD, W. J. (1971) Immunization of sheep against experimental Clostridium perfringens type A enteritis. Can. J. Microbiol. 17: 391-395. NYGREN, B. (1962) Phospholipase C-producing bacteria and food poisoning. Acta Pathol. Microbiol. Scand. (Suppl.) 160: 1-88. OAKLEY, C. L. (1949) The toxins of CIostridium welchii type F. Br. Med. J. 1 : 269-270. OAKLEY, C. L. and WARRACK, G. H. (1953) Routine typing of Clostridium welchii. J. Hyg. Camb. 51: 102107. OAKLEY,C. L., WARRACK,G. H. and VAN HEYNINGEN,W. E. (1946) The collagenase (~c toxin) of CI. welchii type A. J. Pathol. Bacteriol. 58: 229-235. OAKLEY, C. L., WARRACK, G. H. and WARREN, M. E. (1948) The kappa and lambda antigens of Clostridium welchii. J. Pathol. Bacteriol. 60: 495-503.

652

JAMES L. MCDONEL

OcrtOA, R. and VELANDIA, S. (1978) Equine grass sickness: serologic evidence of association with Clostridium perfringens type A enterotoxin. Am. J. Vet. Res. 39: 1049-1051. OHSAKA, A., SUGAHARA,T., TAKAHASHI, T. and YAMAYA, S. (1978a) Effects of ~t-toxin (phospholipase C) of Clostridium perfringens on microcirculation. In: Toxins, animal, plant, and microbial, pp. 1031-1048. ROSENBERG, P. (Ed.) (Suppl. No. 1, Toxicon). Pergamon Press, N.Y. OHSAKA, A., TSUCHIYA, M., OSHIO, C., MIYAIRA, M., SUZUKI, K. and YAMAKAWA,Y. (1978b) Aggregation of platelets in the mesenteric microcirculation of the rat induced by or-toxin (phospholipase C) of Clostridium perfringens. Toxicon. 16: 333-341. ORCUTT, R. P., FOSTER, H. L. and JONAS,A. M. (1978) Clostridium perfringens type E enterotoxemia as the cause of acute diarrheal death or "hemorrhagic typhlitis" in rabbits. Abstract No. 100. Am. Assoc. Lab. Animal Sci. Publ. 78-4. ORLANS, E. S., RICHARDS,C. B. and JONES, V. E. (1960) Clostridium welchii epsilontoxin and antitoxin. Immunology 3: 28-44. OTNAESS, A-B. and HOLM, T, (1976) The effect of phospholipase C on human blood platelets. J. Clin. Invest. 57: 1419-1425. OXER, D. T., MINTY, D. W. and LIEFMAN, C. E. (1971) Vaccination trials in sheep with clostridial vaccines with special reference to passively acquired Cl. welchii type D antitoxin in lambs. Aust. Vet. J. 47: 134-140. PARK, Y., BABAJIMOPOULUS,M. and MIKOLAJCIK, E. M. (1977) Purification and activity of Clostridium perfringens alpha toxin. J. Food Prot. 40: 831-834. PARNAS, J. (1976) About the influence of fl toxin of Clostridium perfringens type C on the motorics of intestine segments in vitro. Part 1. Zentralbl. Bacteriol. Parasitenkd. 234: 243-246. PATTON, N. M., HOLMES,H. T., RIGGS, R. J. and CHEEKE, P. R. (1978) Enterotoxemia in rabbits. Lab. Anim. Sci. 28: 536-540. PETROV, R. V., SHEMANOVA,G. F., KAVERZNEVA,E. D., LISSEL,V. L., PANTELEEV,E. I., DMITR1EVA,L. N. and RUDNEVA, T. B. (1978) Immunogenicity and thymus-dependence of polymerized Clostridium perfringens ct-toxoid. Bull. Expt. Biol. Med. 84: 1743-1737. PLOTKIN, G. R., KLUGE, R. M. and WALDMAN,R. G. (1979) Gastroenteritis: etiology, pathophysiology and clinical manifestations. Med. 58: 95-114. P1NEGAR, J. A. and STRINGER, M. F. (1977) Outbreaks of food poisoning attributed to lecithinase-negative Clostridium welchii. J. Clin. Pathol. 30: 491-492. PRIGENT, O. and ALOUF, J~ E. (1976) Interaction of streptolysin O with sterols. Biochim. Biophys. Acta 443: 288 300. ROBB-SMITH, A. H. T. (1945) Tissue changes induced by C1. welchii type A filtrates. Lancet ii: 362-368. ROBERTSON, W. B., ROPES, M. W. and BAUER, W. (1940) Mucinase: a bacterial enzyme which hydrolyzes synovial fluid mucin and other mucins. J. Biol. Chem. 133: 261-276. ROGERS, H. J. (1948) The complexity of the hyaluronidases produced by microorganisms. Biocbem. J. 42: 633~i40. ROOD, J. I., SCOTT, V. N. and DUNCAN, C. L. (1978a) Identification of a transferable tetracycline resistance plasmid (pCW3) from Clostridium perfringens. Plasmid 1 : 563-570. ROOD, J. I., MAH~R, E, M., SOMERS,E. B., CAMPOS,E. and DUNCAN, C. L. (1978b) Isolation and characterization of multiple antibiotic-resistant Clostridium perfringens strains from porcine feces. Antimicrob. Aoents Chemother. 13: 871-880. ROOD, J. I. and WXLKINSON, R. G. (1974) Affinity chromatography of Clostridium perfringens sialidase non specific adsorption of haemagglutinin, haemolysin and phospholipase C to Sepharosyl-glycyltryosyl [N-(paminophenyl) oxamic acid]. Biochim. Biophys. Acta 334: 168-178. ROOD, J. I. and WILKINSON, R. G. (1976) Relationship between hemagglutinin and sialidase from Clostridium perfringens CN3870: chromatographic characterization of the biologically active proteins. J. Bacteriol. 126: 841-844. ROSENBERG, A. and SCHENGRUND, C. L. (1976) Sialidases. In: Biological roles of sialic acid. pp. 295-359. ROSENBERG, A. and SCHENGRUNÙ, C. L. (Eds.). Plenum Press, N.Y. ROSENBERG, P. (1976) Bacterial and snake venom phospholipases: enzymatic probes in the study of structure and function in biolectrically excitable tissues. In: Animal, plant, and microbial toxins, Vol. 2. pp. 229-262, OHSAKA, A., HAVASm, K. and SAWAI, Y. (Eds.). Plenum Press, N.Y. ROSENTHAL, A. F., CHODSKY, S. V. and HAN, S. C. H. (1969) Relationship between chemical structure and platelet-aggretation activity of prostaglandins. Biochim. Biophys. Acta 187: 385-392. ROSENTHAL, A. F. and POUSADA, M. (1968) Inhibition of phospholipase C by phosphonate analogs of glycerophosphatides. Biochim. Biophys, Acta 164: 226-237. Ross, H. E., WARREN, M. E. and BARNES,J. M. (1949) Clostridium welchii iota toxin: its activation by trypsin. J. Gen. Microbiol. 3: 148-152. ROTTEM, S., HARDEGREE, M. C., GRAROWSKI, M. W., FORNWALO, R. and BARILE, M. F. (1976) Interaction between tetanolysin and mycoplasma cell membrane. Biochim. Biophys. Acta 455: 876-888. S,~BAN, E., LASER, Y. and LoveR, A. (1972) Resolution of the hemolytic and the hydrolytic activities of phospholipase-C preparations from Clostridium perfringens. Eur. J. Biochem. 28: 373-380. SAKAGUCHI, G., UEMURA, T. and TIEMANN, H. P. (1973) Simplified method for purification of Clostridium perfringens type A enterotoxin. Appl. Microbiol. 26: 762-767. SAKURAI, J. and DUNCAN, D. L. (1977) Purification of beta toxin from Clostridium perfringens type C. Infect, lmmun. 18: 741-745. SAKURAI,J. and DUNCAN, C. L. (1978) Some properties of beta toxin produced by Clostridium perfringens type C. Infect. Immun. 21: 678-680. SAKURAI, J. and DUNCAN, C. L. (1979) Effect of carbohydrates and control of culture pH on beta toxin production by Clostridium perfringens type C. Microbiol. lmmunol, 23: 313-318. SATO, H., YAMAKAWA,Y., ITO, A. and MURATA,R. (1978) Effect of zinc and calcium ions on the production of alpha-toxin and proteases by Clostridium perfringens. Infect. Immun. 20: 325-333.

Clostridium perfringens toxins (Type A, B, C, D, E)

653

SCHILLER, M., DONNELLY, P. J., MELO, J. C. and RAFF, M. J. (1979) CIostridium perfringens septic arthritis. Clin. Orthop. 139: 92-96. SCOTT, V. N. and DUNCAN, C. L. (1975) Affinity chromatography purification of Clostridium perfringens enterotoxin. Infect. Immun. 12: 536-543. SEBALD, M., BOUANCHAUD, D. and BIETH, G. (1975) Nature plasmidique de la resistance a plusieurs antibiotiques chez C. perfringens type A, souche 659. C.R. Acad. Sci. (Paris) 280: 2401-2404. SEBALD, M. and BRI~FORT, G. (1975) Transfer du plasmide tetracycline-chloramphenicol chez Clostridium perfringens. C. R. Acad. Sci. (Paris)281: 317-319. SEIFEERT,G. (1939) A reaction in human serum with Cl. perfringens toxin. Z. ImmunForsch. 96: 515-520. SENFE,L. M. M. and MOSKOWITZ, M. (1969) Relation of in vitro inhibition by chelates of Clostridium perfringens a-toxin to their ability to protect against experimental toxemia. J. Bacteriol. 98: 29-35. SERRANO,A. M. and SCHNEIDER,I. S. (1978) New modification of Willis and Hobbs' method for identification of Clostridium perfringens. Appl. Environ. Microbiol. 35: 809-810. SHEMANOVA,G. R., VLASOVA,E. G., TSVETKOV, V. S., LOGUNOV, A. I. and LEVIN. ]7. B. (1968) Extraction and properties of Clostridium perfringens lecithinase. Biochemistry, N. Y (Eng. cd) 33:110-115. SHUKLA, S. D., COLEMAN, R., F1NEAN, J. B. and MICHELL, R. H. (1978) The use of phospholipase C to detect structural changes in the membranes of human erythrocytes aged by storage. Biochim. Biophys. Acta 512: 341-349. SKJELKVALIg,R. and DUNCAN, C. L. (1975a) Enterotoxin formation by different toxigenic types of Clostridium perfringens. Infect. lmmun, l l : 563-575. SKJELKVALI~,R. and DUNCAN, C. L. (1975b) Characterizztion of enterotoxin purified from Clostridium perfringens type C. Infect. Immun. l l: 1061-1068. SKJELKVALE,R. and UEMURA,T. (1977a) Detection of enterotoxin in faeces and anti-enterotoxin in serum after Clostridium perfringens food-poisoning. J. Appl. Bacteriol. 42: 355-363. SKJELKVALI~,R. and UEMURA,Z. (1977b) Experimental diarrhoea in human volunteers following oral administration of Clostridium perfringens enterotoxin. J. Appl. Bacteriol. 43: 281-286. SMART,J. L., ROBERTS, T. A., STRINGER,M. F. and SHAH, N. (1979) The incidence and serotypes of Clostridium perfringens on beef, pork, and lamb carcasses. J. Appl. Bacteriol. 46: 37~383. SMITH, L. US. (1975a) The pathogenic anaerobic bacteria. (2nd ltd.). Charles C. Thomas. Springfield, IL. SMITH, L. DS. (1975b) Clostridial infections. In: Isolation and identification of Avian Pathogens. pp. 95-96. HITCHNER, S. B., DOMERMUTH, C. H., PURCHASE, H. G. and WILLIAMS, J. E. (Eds.) Am. Assoc. Avian Pathologists. SMITH, L. US. and GARDNER, V. M. (1950) The anomalous heat inactivation of Clostridium perfringens lecitfiinase. Arch. Biochem. 25: 54-60. SMYTH, C. J. (1974) Multiple forms of Clostridium perfringens 0-toxin (0-haemolysin): physical properties and biological characteristics. Proc. Soc. Gen. Microbiol. 1: 56. SMYTH, C. J. (1975) The identification and purification of multiple forms of 0-haemolysin (0-toxin) of Clostridium perfringens type A. J. Gen. Microbiol. 87: 219-238. SMYTH, C. J. and ARBUTHNOTT,J. P. (1972) Characteristics of Clostridium perfringens type A a-toxin purified by isoelectric focusing. J. Gen. Microbiol. 71 : ii (abstract). SMYTH, C. J. and ARBUTHNOTT, J. P. (1974) Properties of CIostridium perfringens (welchii) type A or-toxin (Phospholipase C) purified by electrofocusing. J. Med. Microbiol. 7: 41-66. SMYTH, C. J., ARBUTHNOTT,J. P. and FREER, J. H. (1972) Purification properties and biological characteristics of Clostridium perfringens type A 0-toxin. J. Gen. Microbiol. 73: xxvi (abstract). SMYTH, C. J. and DUNCAN, C. L. (1978) Thiol-activated (oxygen-labile) cytolysins. In: Bacterial toxins and cell membranes, pp. 129-183. JELJASZEWlCZ,J. and WADSTRgM, T. (Eds.). Academic Press, N.Y. SMYTH, C. J., FREER, J. H. and ARBUTVlNOTT,J. P. (1975) Interaction of Clostridium perfringens 0-hemolysin, a contaminant of commercial phospholipase C, with erythrocyte membranes and lipid dispersions, a morphological study. Biochim. Biophys. Acta 382: 479-493. SMYTH,C. J. and WADSTROM,T. (1975) Isoelectric focusing in thin layer polyacrylamid¢ gel combined with a zymogian method for detecting enzyme microheterogeneity: sample application. Anal. Biochem. 65: 137-152. SODA,S., ITO, A. and YAMAMOTO,A. (1976) Production and properties of 0-toxin of Clostridium perfringens with special reference to lethal activity. Japan. J. Med. Sci. Biol. 29: 335-349. STAHL, W. L. (1973) Phospholipase C purification and specificity with respect to individual phospholipids and brain microsomal membrane phospholipids. Arch. Biochem. Biophys. 154: 47-55. STARK,R. L. and DUNCAN,C. L. (1971) Biological characteristics of Clostridium perfringens type A enterotoxin. Infect. Immun. 4: 89-96. STARK, R. L. and DUNCAN, C. L. (1972a) Transient increase in capillary permeability induced by Clostridium perfringens type A enterotoxin. Infect. lmmun. 5: 147-150. STARK, R. L. and DUNCAN, C. L. (1972b) Purification and biochemical properties of Clostridium perfringens type A enterotoxin. Infect. Immun. 6: 662-673. STERN, G. A., HODES, B. L. and STOCK, E. L. (1979) Clostridium perfringens corneal ulcer. Arch. Ophthalmol. 97: 661-663. STERN, M. and WARRACK, G. H. (1964) The types of Clostridium perfringens. J. Pathol. Bacteriol. 88: 279-283. STERNE, M. and BATTY, L (1975) Criteria for diagnosing clostridial infection. In: Pathogenic clostridia, pp. 79-84. Butterworth and Co., London. STRANDBERG, K., MOLLBY, R. and WADSTROM, T. (1974) Histamine release from mast cells by highly purified phospholipase C (~t-toxin) and theta toxin from Clostridium perfringens. Toxicology. 12: 199-208. STRONG, D. H., DUNCAN, C. L. and PERNA, G. (1971) Clostridium perfringens type A food poisoning. II. Response of the rabbit ileum as an indication of enteropathogenicity of strains of Clostridium perfringens in human beings. Infect. lmmun. 3: 171-178.

654

JAMES L. McDONEL

STRUNK, S. W., SMITH,C. W. and BLUMBERG,J. M. (1967) Ultrastructural studies on the lesion produced in skeletal muscle fibers by crude type A Clostridium perfrinoens toxin and its purified alpha fraction. Am. J. Pathol. 50: 89-107. SUGAHARA, T. and OHSAKA, S. (1970) Two molecular forms of Clostridium perfringens ~-toxin associated with lethal, hemolytic and enzymatic activities. Japan. J. Med. Sci. Biol. 23: 61-66. SUGAHARA, T., TAKAHASHI,T., YAMAVA,S. and OHSAKA, A. (1976) In vitro aggregation of platelets induced by alpha-toxin (phospholipase C) of Clostridium perfringens. Japan. J. Med. Sci. Biol. 29: 255-263. SUGAHARA, T., TAKAHASHI,T., YAMAYA,S. and OHSAKA, A. (1977) Vascular permeability increase by or-toxin (phospholipase C) of Clostridium perfringens. Toxicology 15: 81-87. SUGIURA, M., OGISO, T., TAKEUTL K., TAMURO, S. and ITO, A. (1973) Studies on trypsin inhibitors in sweet potato. I. Purification and some properties. Biochim. Biophys. Acta 328: 407-417. SUTTON, R. G. A. and HoaBs, B. C. (1968) Food poisoning caused by heat-sensitive Clostridium welchii. A report of five recent outbreaks. J. Hy#. Camb. 66: 135-146. SZMIGIELSK1,S., JANIAK,M., MOLLBY,R., WADSTRt~M,T. and JELJASZEWICZ,J. (1978) Metabolism of rabbit kidney cells incubated in vitro with phospholipase C from Staphylococcus aureus and Clostridium perfringens. Toxicology 16: 567-574. TAGUCHI, R. and IKEZAWA, H. (1976a) Studies on the hemolytic and hydrolytic actions of phospholipases against mammalian erythrocyte membranes. Archs. Biochem. Biophys. 173: 538-545. TAGUCHI, R. and |KEZAWA, H. (1976b) Studies on the mode of action of phospholipases on mammalian erythrocytes. In: Animal, plant, and microbial toxins. Vol. I. pp. 429-436. OHSAKA, A., HAYASHI, K. and SAUAI, Y. (Eds.). Plenum Press, N.Y. TAKAHASH1,T., SUGAHARA,T. and OHSAKA, A. (1974) Purification of Clostridium perfringens phospholipase C (or-toxin) by affinity chromatography on agarose-linked egg-yolk lipoprotein. Biochim. Biophys. Acta 351: 155-171. TEODORESCU, G. H., BITTNER, J. and CERCARCANU, A. (1970) Donn6es preliminaires sur la purification et la structure de l'alphatoxine Clostridium perfrinoens. Arch. Roum. Pathol. Exp. Microbiol. 29: 541-544. THELESTAM, M. and M6LLBV, R. (1975a) Determination of toxin-induced leakage of different size nucleotides through the plasma membrane of human diploid fibroblasts. Infect. Immun. 11: 640-648. THELESTAM, M. and M6LLBV, R. (1975b) Sensitive assay for detection of toxin-induced damage to the cytoplasmic membrane of human diploid fibroblasts. Infect. Immun. 12: 225-232. THELESTAM, M. and M6LLBY, R. (1976) Cytotoxic effects on the plasma membrane of human diploid fibroblasts, a comparative study of leakage tests. Med, Biol. 54: 39-49. THOMPSON, R. O. (1962) Crystalline e-prototoxin from Clostridium welchii. Nature 193: 69-70. THOMPSON, R. O. (1963) The fractionation of Clostridium welchii e-antigen on cellulose ion exchanges. J. Gen. Microbiol. 31: 79-90. TIXIER, G. and ALOUF, J. E. (1976) Purification and some properties of Clostridium perfringens delta toxin. Ann. Microbiol. (Paris) 127B: 509-524. UEMURA, T., GENIGEORGIS, C., RIEMANN, H. P. and FRANTI, C. E. (1974) Antibody against Clostridium perfringens type A enterotoxin in human sera. Infect. lmmun. 9: 470-471. UEMURA, T., SAKAGUCHI,G., ITOH, T., OKAZAWA,K. and SAKAI, S. (1975) Experimental diarrhea in cynomolgus monkeys by oral administration with Clostridium perfringens type A viable cells or enterotoxin. Japan. J. Med. Sci. Biol. 28: 165-177. UEMURA, T., SAKAGUCHI,G. and RIEMANN, H. P. (1973) In vitro production of Clostridium perfringens enterotoxin and its detection by reversed passive hemagglutination. Appl. Microbiol. 26: 381-385. UEMURA, T. and SKJELKVAL~,R. (1976) An enterotoxin produced by Clostridium perfringens type D. Purification by affinity chromatography. Acta Pathol. Microbiol. Scand. Sect. B 84: 414-420. VAN DEENEN, L. L. M. (1964) The specificity of phospholipases. In: Metabolism and physiological significance of lipids, pp. 155-178. DAWSON, R. M. C. and RHODES, D. N. (Eds.). John Wiley, London. VAN DEENEN,L. L. M., DEMEL, R. A., GUERTS VAN KESSEL, W. S. M., KAMP, H. H., ROELOFSEN,B., VERKLEIJ, A. J., WIRTZ, K. W. A. and ZWAAL, R. F. A. (1976) Phospholipases and monolayers as tools in studies on membrane structure. In: The structural basis of membrane function, pp. 21-38. HATEFI, Y. and DJAVADIOHANIANCE, L. (Eds.). Academic Press, N.Y. VAN HEYNINGEN, W. E. (1950) Bacterial toxins. C. J. Thomas, Springfield, Ill. VINCE, A., DYER, N. H., O'GRADY, F. W. and DAWSON, A. M. (1972) Bacteriological studies in Crohn's disease. J. Med. Microbiol. 5: 219-229. WALKER, H. W. (1975) Food borne illness from Clostridium perfringens. Crit. Rev. Food Sci. Nutr. 7: 71-104. WATSON, K. C. and KERR, E. J. C. (1974) Sterol stuctural requirements for inhibition of streptolysin O activity. Biochem. J. 140: 95-98. WEINSTEIN, L. and BARZA, M. A. (1973) Gas gangrene. N. Engl. J, Med. 289: 1129-1131. W1ERUP, M. (1977) Equine intestinal clostridiosis. An acute disease in horses associated with high intestinal count of Clostridium perfringens type A. Acta Vet. Scand. (suppl.) 62-64. WILLIS, T. A. (1969) Clostridia of Wound Infection. Butterworth, London. WILLIS, A. T. (1977) Anaerobic Bacteriology: Clinical and Laboratory Practice. (3rd Edition). Butterworths, London. WILKINSON,P. C. (1975) Inhibition of leukocyte locomotion and chemotaxis by lipid specific bacterial toxins. Nature (Lond.) 255: 485-487. WORTHINGTON, R. W. and MOLDERS, M. S. G. (1975a) Effect of Clostridium perfringens epsilon toxin on the blood brain barrier of mice. Onderstepoort. J. Vet. Res. 42: 25-28. WORTHINGTON, R. W. and MOLDERS, M. S. G. (1975b) The partial purification of Clostridium perfringens beta toxin. Onderstepoort. J. Vet. Res. 42: 91-98. WORTHINGTON, R. W. and MOLDERS, M. S. G. (1977) Physical changes in the epsilon prototoxin molecule of Clostridium perfringens during enzymatic activation. Infect. Immun. 18: 549-551. WORTHINGTON, R. W., MOLDERS, M. S. G. and VAN RENSBURG,J. J. (1973) Enzymatic activation of Clostridium

Clostridium perfringens toxins (Type A, B, C, D, E)

655

perfrinoens epsilon toxin and some biological properties of activated toxin. Onderstepoort. J. Vet. Res. 40: 151-154. WYNNE, J. W. and ARMSXRON~, D. (1972) Clostridial septicemia. Cancer 29: 215-221. YAMAGmm, T., YOSmZAWA, J., KAWAZ, M., SEO, N. and NISHIDA, S. (1971) Identification of isolates of Clostridium perfrinoens types C and D by agglutination and fluorescent-antibody methods. Appl. Microbiol. 21: 787-793. YAMAKAWA, Y., ITO, A. and SATO, H. (1977) Theta toxin of Clostridium perfrinoens. I. Purification and some properties. Biochim. Biophys. Acta 494:301 313. YAMAKAWA, Y. and OHSAKA, A. (1977) Purification and some properties of phospholipase C (a-toxin) of Clostridium perfringens. J. Biochem. 81:115 126. YAMAMOTO, K., OHxsm, I. and SAKAGUCHI,G. (1979) Fluid accumulation in mouse ligated intestine inoculated with Clostridium perfringens enterotoxin. Appl. Environ. M icrobiol. 37: 181-186. YOT~S, W. W. and CATSIMPOOLAS, N. (1975) Scanning isoelectric focusing and isotachophoresis of Clostridium perfringens type A enterotoxin. J. Appl. Bacteriol. 39: 147-156. ZEISSLER, J. and RASSVELD-STERNBEgG,L. (1949) Enteritis necroticans due to Clostridium welchii type F. Br. Med. J. 1: 267-269. ZWAAL, R. F. A., ROELOFSEN, B. and COLt,EY, C. M. (1973) Localization of red cell membrane constituents. Biochim. Biophys. Acta 300: 159-182. ZWAAL, R. F. A., ROELOFSEN, B., COMFURIUS, P. and VAN DEENEN, L. L. M. (1975) Organization of phospholipids in human red cell membranes as detected by the actions of various purified phospholipases. Biochim. Biophys. Acta 406: 83-96.

J.p.'r. 10/3---o