INTRODUCTION. Although much is known today about the mechanisms by which virulent Staphylococcus aureus induce tissue lesions and cause clinical manifestations in mammals, our knowledge of the role played by “professional” phagocytes in host and parasite interrelationships in staphylococcal infections, is not fully understood. The extensive literature on staphylococci and their role in human disease has been the subject of several excellent comprehensive reviews (Whipple, 1965; Cohen, 1972; Jeljaszewicz, 1976). It is accepted that the interception of staphylococci by granulocytes (PMNs) takes place soon after the penetration of the cocci into the tissues. This involves the release of chemotactic factors, by the bacteria themselves (Pusell et al., 1975) or the activation, by staphylococcal factors, of chemotactic agents from complement (Ginsburg and Quie, 1980). Subsequent phagocytosis is markedly enhanced by opsonins (Koenig, 1972; Ekstadt, 1974), and in most cases the engulfed bacteria may be killed intracellularly by a variety of bactericidal agents generated by activated PMNs (Klebanoff, 1972). It is also suggested that certain lysosomal enzymes, which are released into the phagolysosome, may digest the staphylococcal cells (Cohn, 1963a; De Voe et al., 1973; Ginsburg and Sela 1976; Ginsburg, 1979) (see Sections II.B and VI). Several reports have, however, shown that intracellular staphylococci may sometimes survive within PMNs, where they multiply and eventually kill the cell (Koenig, 1972; Cohen, 1972; Pearce et al., 1976). Surprisingly, very little is known about the fate and mechanism of biodegradation of staphylococcal cell constituents, once they have been sequestered within the phagolysosomes of leucocytes. The importance of this field of research stems from the findings that non-biodegradable cell wall components of a variety of microbial species may persist within macrophges for prolonged periods, to trigger chronic infiammatory sequelae (Dannenberg, 1968; Kanai and Kondo, 1974; Ginsburg et al., 1975a; 1975b; 1976a; Ginsburg and Sela, 1976; Adams, 1976; Page et al., 1978; Ginsburg, 1979). The inability of “professional” phagocytes to degrade intracellular bacteria may be due (1) to the presence, on the surface of certain microorganisms, of shielding capsular material (Dossett et al., 1969; Smith, 1977; Wilkinson, et al., 1979; Densen and Mandell, 1980), (2) to the lack of fusion between lysosomes and phagosomes (Goren et al., 1976; Densen and Mandell, 1980), (3) to the production of leucocidins (Gladstone and Van Heyningen, 1957; Woodin, 1960; Ginsburg, 1970; Van Heyningen, 1970; Bernheimer. 1970; Ginsburg, 1972), (4) to the lack of adequate lysosomal enzymes capable of cleaving bacterial peptidoglycans (Dannenberg, 1968; Ginsburg, 1972; Ginsburg and Sela, 1976; Page et al., 1978; Ginsburg, 1979), or (5) to the presence, in serum and in inflammatory exudates, of agents which interfere with the interaction of bactericidal and bacteriolytic age with engulfed bacteria. These fields were comprehensively reviewed EW’ Jeljaszewicz, 1976; Smith, 1977; Ginsburg, 1979; and by Densen and Mandell, 1980. During the last 8 years our laboratory has been studying the host- and-parasite interrelationships in streptococcal and staphylococcal infections using biochemical, electron microscopical and tissue culture techniques. In particular we studied the mechanisms by which leu- cocytes and their isolated lysosomal agents bring about the degrada- tion of staphylococcal cell wall components, and the role which may be played by such degradation products in the initiation of chronic inflammation. The present report is an updated overview of these studies.

Introduction. Although a wealth of knowledge exists today on the biochemical pathways of biosynthesis, turnover and autolysis of bacterial cell wall components in vitro (1, 2), surprisingly very little is actually known about the mechanisms of biodegradation of microbial constituents in_vivo. One should differentiate between bactericidal and bacteriolytic processes induced by leukocytes since killed, but non-degraded, microbial cells may persist within macrophages to trigger chronic inflammation (3, 4). The present communication further supports our contention (5, 6) that the degradation of microbial cell wall components by leukocytes may be due to activation, by leukocytic cationic proteins, of autolytic wall enzymes rather than to the direct cleavage of the cells by lysosomal hydrolases. The modulation of bacteriolysis by anionic polyelectrolytes will be described and discussed in relation to the pathogenesis of chronic inflammation and afequelae.

Introduction. In previous studies it could be shown that autolytic wall enzymes of bacteria can be activated by some cationic proteins (1). Recently we determination of firming our data lytic enzyme but obtained results from chemical and end group determination of the cleavage products from peptidoglycan confirming our data that even lysozyme acted not only as a muralytic enzyme but also as a cationic protein (2). In order to elucidate the mechanisms of the activation of autolytic wall processes by cationic proteins and of the direct respectively indirect muralytic actions of lysozyme we performed investigations on heated cells.

Although a voluminous literature exists today on the mechanisms by which "professional" leukocytes (granulocytes and maerophages) intercept with, engulf and eventually kill phagocytosed microorganisms ~ (1, 2), surprisingly very little is known about the mechanisms of degradation and elimination of bacteria from tissues. It is well established that phagocytic cells are endowed with numerous hydrolyric enzymes, including the key cell wall splitting enzyme-lysozyme, which can theoretically cleave, surface, eeU wall and cytoplasmic constituents of bacteria. Also, fresh mammalian sera are known to contain a complex mixture of heat-labile complement components (3) heat-stable lysozyme and platelet-derived cationic proteins (~-lysins) (4) which have been shown to kill and partially lyse certain microbial constituents. Surprisingly, however, the majority of virulent microorganisms are highly refractory to both leukocyte and serum lyric agents. Throughout this communication we shall use the general term bacteriolysis to denote the degradation of the cell walls, the outer membranes and the cytoplasmic constituents of bacteria. The term cell wall lysis will be used to describe the specific biochemical degradation of the bacterial peptidoglycan. This may also be accompanied by the rupture of the protoplast membrane and the release of cytoplasmic constituents. The inability of leukocyte and serum factors to induce bacteriolysis is linked to the presence, upon most bacterial surfaces, of'lipopolysaccharides, polysaccharides-teichoic complexes and certain lipids and waxes, which hinder the accessibility of the major cell wall splitting enzyme-lysozyme to the peptidoglycan (1). Once however this obstacle is overcome, the peptidoglycan ist degraded, and the protoplasts burst due to their high osmotic pressure, releasing degradation products of both cell wall and cytoplasmic constituents into the surrounding medium. The very extensive literature on these subjects has been recently summarized and reviewed (5-7). It has als0 been suggested that the process of killing and biochemical degradation of microbial constituents, either following phagocytosis or following treatment with fresh complement and lysozyme-sufficient serum are probably mediated by different mechanisms (6-8). While extensive loss of wall material and cytoplasmic entities is usually accompanied by a bactericidal reaction, the killing of bacteria either by the oxygen-dependent (9) or by the non-oxygen dependent bactericidal systems of leukocytes (5, 10) and serum (3), is not necessarily accompanied by a substantial bacteriolysis. The distinction between a bactericidal and a bacteriolyric process is important, in view of the observations that poorly degraded non-viable microbial constituents may persist for long periods both extracellularly and within phagocytic cells, to trigger and perpetuate chronic inflammatory sequellae (6, 7, 11-13). Furthermore, while degradation products of grampositive and acid-fast bacteria have been shown to be endowed with distinct pharmacological properties (14), to modulate the immune responses (15), to activate the complement cascade (16) and to be cytopathic for mammalian cells (14), the in vivo release of lipopolysaccharides of the outer membrane of gramnegative bacteria may result in severe coagulation defects (Shwartzman phenomenon) and in endotoxic shock (17). Poorly degraded cell wall components of bacteria have also been shown to be translocated within macrophages from one tissue site to another, thus contributing perhaps to the dissemination of granulomatosis (18-20). Although the nature of the biochemical pathways involved in bacterial biodegradation in tissues has not been fully elucidated, it has been recently suggested that a cooperation among leukocyte factors (21, 22), serum components (23), the bacterial own autolytic wall enzymes (21, 22) and certain antibiotics (24), may act in accord to induce a massive breakdown of cell wall and cytoplasmic constituents of bacteria. It is well established that the autolytic systems present in every bacterial cell, control cell division, the deposition of new cell wall material and the regeneration after treatment with certain antibiotics (25, 36). Autolytic enzymes have been isolated from many bacterial species and were found to possess muramidase, Nacetyl glucosaminidase, amidase and peptidase activities (25). It is also known that certain antibiotics, mainly of the penicillin and cephalosporin series, are capable of killing microorganisms, presumably by activating their autolytic wall enzymes (27). These intraceUular enzymes are thought to be controlled by endogenous lipid material (e.g.- phospholipids, lipoteichoic acid) (27). Thus, any agent present in leukocytes or in tissue fluids, which will disrupt the balance between autolytic enzymes and their naturally occurring inhibitors, may lead to the activation of autolysins, and concomitantly to the release of toxic bacterial agents. In view of the complex interrelationship which exist between bacteria and host factors in infectious and imflammatory sites, it was of interest to clarify some of the mechanisms and the factors involved in the biodegradation and persistence of microbial constituents in tissues. The following is an overview of our studies on this subject, employing staphylococci and streptococci as model systems, and using biochemical and electron microscopical techniques (28). The peptidoglycan of Staphylococcus aureus was labelled during the logarithmic phase of growth with 14C-N-acetylglucosamine. When such labelled cells were incubated for several hours at 37 ~ in acetate buffer pH 5.0, with small amounts of crude human leukocyte extracts or with more purified lysosomal extracts, a substantial amount of the radioactivity, associated with the cell walls was solubilized. Electron microscopical analysis of these reaction mixtures revealed the accumulation of cell wall fragments, and both amorphous and intact cytoplasmic constituents still retaining their typical morphologies (6, 7, 21, 22, 28, 29). Since the pH optimum for this reaction process was found to be on the acid side and compatible with the pH optima of many of the acid hydrolases known to be present in the leukocyte preparation (1) we postulated that the breakdown of the bacterial ceils was mediated by acid hydrolases. The findings, however, that heat treatment did not destroy the capacity of the leukocyte extracts to induce cell wall degradation, and that purified radiolabelled staphylococcal cell walls became completely refractory to the lytic effect of the extracts (21, 22), suggested that the wall degradation observed was probably not caused by the leukocyte hydrolases. Since leukocyte lysosomes are known to be rich in heat-stable argininerich bactericidal cationic proteins (LCP) (30) and in myeloperoxidase (MPO) (1) (also a cationic protein) it was reasonable to try and employ them, instead of the total leukocyte mixture, to lyse the staphylococci. Indeed we found that as little as 0.5-1/zg/ml of nuclear histone, poly-L-lysine, poly-L-arginine or MPO and 10-50/ag/ml of crystalline pancreatic ribonuclease or cytochrome C (all cationic in nature) were sufficient to induce massive loss of cell wall material from 108 log-phase staphylococci. Furthermore, small amounts of the membrane.damaging agents phospholipase A2, and polymyxins B and E were also capable of inducing cell wall lysis, as determined by the release of N-acetyl-glucosamine (8, 31-33). Since purified staphylococcal cell walls (devoid of cytoplasmic structures) were extremely refractory to any of the cationic polyelectrolytes or to the membrane-damaging agents, we postulated that perturbation of the staphylococcal membrane by these agents might have resulted in the activation of endogenous enzymes presumably associated with the autolytic systems (21, 22, 25, 26). As autolytic wall enzymes are known to be heat-labile (25, 26), it was of in,terest to try and reactivate the lytic process by the addition of freshly-harvested viable staphylococci (as donors of autolysins) to the heat killed radiolabeled staphylococci or to the purified labelled cell walls, in the presence of several of the cationic proteins or the membrane-damaging agents. Indeed, such mixtures resulted in a substantial loss of radiolabelled wall material. This process was completely blocked by anionic polyelectrolytes. We suggested, therefore, that the activators of autolysins interacted with the viable bacteria to release the autolytic enzymes, which in turn attacked and degraded the radiolabelled substances. Similar results were recently described with Bacillus subtilis (34). It has also been suggested that pneumococci, gonococci, meningococci, Streptococcus faecalis and perhaps listeriae may also likewise be degraded in vivo following the activation of their autolysins, and not through the direct action of lysosomal enzymes (28). Further experiments showed that Staphylococcus aureus, which had been cultivated in the presence of sub-inhibitory concentrations of penicillin G, became much more susceptible to wall lysis, following treatment with leukocyte extracts (24,35) suggesting a collaboration between/3-1actam antibiotics, leukocyte factors and bacterial autolysins in bacteriolysis (see 27). Other studies from our laboratory (23) have also shown that contrary to the accepted belief, both fresh and heat-treated human serum, when properly diluted, also lysed log-phase grampositive staphylococci and Streptococcus faecalis at pH 5.0. Since anionic polyelectrolytes also inhibited cell lysis induced by serum (23), and since heat-killed bacteria became resistant to lysis by serum, we postulated that, as in the case of leukocyte extracts, lysis was induced by a heat-stable factor, presumably ~-lysin of platelet origin (4) present in serum, which activated the autolytic systems of the bacteria. To further elucidate the mechanism of bacteriolysis, we have analyzed this process by electron microscopy. In collaborative studies with Prof. P. Giesbrecht and Dr. J. Wecke of the Robert Koch Institute in Berlin, we found (36) that a few hours after the addition of either crystalline lysozyme (500/ag) or pancreatic ribonuclease (50/ag/ml) to log-phase staphylococci, the first signs of cell damage could already be seen. These consisted of the formation of small, periodically-arranged lytic sites between the cell wall and the cytoplasmic membrane of the cross wall. This was followed by the formation of a distinct gap between the cell wall proper and the cytoplasmic membrane. The degradation of the peripheral cell wall continued to the opposite side of the cell, and extended gaps underneath the wall could be detected long before the cell wall itself was peeled off as large ribbons. The cross wall often appeared to be already disintegrated during the early phase of lysozyme or ribonuclease action. The release of the wall left apparently intact protoplasts, which still retained their original shape, and even the invagination of the cross walls were conserved. At this point over 70 % of the toal radioactivity associated with the cell wall was solubilized after three to four hours, and the radioactivity could not be sedimented at 100,000 • g suggesting the formation of solubilized peptidoglycan. Since lysozyme which had been heated to destroy its enzymatic activity still retained its ability to induce cell wall lysis, we postulated that lysozyme in this system did not act as an enzyme but as a cationic protein. Further studies from our laboratory (28) and in collaboration with the Robert Koch Institute (to be published in detail) have revealed that very similar ultrastructural changes in the staphylococci also took place following phagocytosis by mouse non-elicited macrophages in culture. On the other hand, although staphylococci, which had been injected into mouse or rat tissues, and which were phagocytosed by both PMNs and macrophages, underwent rapid loss of their cytoplasmic constituents, presumably by digestion with lysosomal hydrolases (37-39), no apparent damageto the cell walls was evident for many days, suggesting that the autolytic wall enzymes might have been inhibited. Since the degradation of the staphylococcal cell walls in vitro was completely inhibited by a variety of anionic polyelectrolytes like heparin, dextran sulfate, polyanethole sulfonate (a synthetic heparin), as well as by cationic polyelectrolytes like histones, poly-L-lysine, poly-L-ar~inine, etc., when used at 10-100 /ag/ml/10 ~ staphylococci (concentrations 10- 100 fold higher than those employed to activate the autolytic wall enzymes (see above), we also postulated that a delicate balance between activators and inhibitors may determine whether or not bacterial wall material may be degraded in tissue lesions in vivo (40,41). Finally, iecent studies (42) have also shown that high-molecular-weight degradation products of staphylococcal cell walls derived following treatment of the bacteria either in buffers (spontaneous wall lysis) or by small amounts of leukocyte extracts, were found to possess very strong chemot~ctic activities for PMNs in vitro, and to induce severe inflammatory lesions when injected intraarticularly to rats. Thus, it may be concluded that the fate of bacterial peptidoglycans in leukocytes in inflamed tissues may be dependent on the one hand on the availability of agents capable of activating autolytic wall enzymes in bacteria, and on the other hand on the presence in tissues of inhibitory substances (polyelectrolytes) which are capable of blocking bacteriolysis. It is, however, not aimed t9 rule out the possibility that other still unknown mechanisms may function in the complex milieu of inflammation, which may bring about the biodegradation of bacteria. The employment of certain antibiotics and other pharmacological agents, yet to be discovered, which will be capable of changing the balance between activators and inhibitors of autolytic enzymes, may contribute to a better understanding of the mechanisms involved in the survial and persistence of microbial agents in vivo. It is also obvious that the release of large quantities of microbial constituents following incomplete biodegradation may prove to be deleterious to tissues. Finally, our studies on staphylococci do not shed light on the mechanisms of biodegradation of other microbial species of medical imprtance, and is only a reflection of one possible mechanism, which may or may not be common to all microorganisms.

This year marks the centennial anniversary of Elie Metchnikoff's discovery of the pivotal role played by "professional" phagocytes in body defenses against invading microorganisms. His cellular theory dealt with the phagocytic events and the postphagocytic killing, and also alluded to the digestion of the internalized bacteria by the "cystases," later shown to be associated with the lysosomal apparatus of leukocytes. Fo date, despite the fact that numerous studies have described in great detail the mechanisms by which serum and leukocytes kill microorganisms (1, 8, 13, 24), surprisingly little is actually known about the biochemical pathways of degradation and mechanisms of disposal of microbial constituents once they have been sequestered within phagolysosomes (3, 9, 10, 13, 24). It is usually taken for granted that the numerous hydrolytic enzymes, including the key bacteriolytic enzyme lysozyme (muramidase), present in lysosomes of "professional" phagocytic cells [granulocytes or polymorphonuclear neutrophils (PMNs), and macrophages] are capable, (at least theoretically) of stripping off bacterial coats, thus exposing the peptidoglycan to cleavage by touramidase. Yet, the majority of pathogenic microorganisms are highly refractory to lysozyme action (9, 10, 17). One should also bear in mind that, while a massive breakdown of microbial cell walls eventually may lead to a bactericidal reaction, the mere killing of a microorganisms, either by leukocyte or by serum factors, may not necessarily be followed by a bacteriolytic reaction. The importance of elucidating the mechanism of microbial biodegradation in tissues stems from the observation that in many infectious diseases there is "storage" of nonbiodegraded microbial cell wall components within macrophages for long periods, which may be responsible for the perpetuation and propagation of chronic inflammatory sequelae and tissue destruction (10, 18). We have recently postulated (11, 14, 15) that bacteriolysis, and the biochemical degradation that ensues after bacteria have been attacked by serum or by leukocytes, may involve close cooperation among heat-stable serum factors, cationic proteins and phospholipase A2 of leukocytes, and heat-labile endogenous bacterial autolyric wall enzymes. This cooperation is affected markedly by anionic polyelectrolytes, likely to accumulate in inflammatory exudates, which may shut down autolysis and, thus, contribute to unfavorable postinfectious sequelae (10, 19). The present communication is a summary of efforts from our laboratory to gain insight into the mechanisms of lysis of Staphylococcus attreus, chosen as a model, by lysosomal enzymes of human blood leukocytes (3, 8-11, 14, 15, 19).