Phagocyte—Pathogenic Microbe Interactions

Phagocyte—Pathogenic Microbe Interactions


4MB Sizes 6 Downloads 53 Views


Phagocyte-Pathogenic Microbe Interactions ANTOINETTE RYTERAND CHANTAL DE CHASTELLIER Uniti de Microscopie Electronique, Dipartement de Biologie Moliculaire, Institut Pasteur, Paris Cedex, France I. Introduction .....................



.............................. ...............


111. Ingestion. .............................

VII. Conclusions ..................................... References . . . . . . . ...........................


I. Introduction Endocytosis is a widespread cellular function which allows the cell to ingest exogenous material. Pinocytosis, which occurs in most eukaryotic cells, corresponds to the uptake of very small particles such as ferritin, macromolecules, and low-molecular-weight solutes. Although it is difficult to visualize the initial step of pinocytosis, it probably arises by different mechanisms: fusion of membrane folds, or membrane invagination leading to the formation of depressions; in large amoebae the plasma membrane forms long channels (Chapman-Andresen, 1977; Stockem, 1970, 1977). Phagocytosis corresponds to the uptake of large particles such as microorganisms, latex beads, oil droplets, etc. Soon after particle adhesion, the cell membrane extends pseudopods around the particle, thus forming a sort of large depression termed the phagocytic cup. The latter gives rise to a phagosome upon membrane closure. As already described in excellent reviews (Holter, 1959; Jacques, 1969; North, 1970; Silverstein et al., 1977), pinosomes and phagosomes fuse with lysosomes and constitute secondary lysosomes in which the internalized material is digested by acid hydrolases. Phagocytosis is particularly well developed in Protozoa for which it represents the main feeding mechanism. The study of primitive pluricellular organisms in which digestive functions are ensured by a group of ameboid cells led Metchnikoff, 100 years ago, to parallel the behavior of these ameboid cells with that of 287 Copyright 0 1983 by Academic Press, Inc. All rights of reproduction in any form reserved. ISBN 0-12-364485-2



cells found in inflammatory lesions of vertebrates. He discovered that specialized cells endowed with a high phagocytic activity play a crucial role in the host defense against microbe invasion. Although the concept of cellular immunity has been ackowledged for over a century, there has been a real explosion of inforrnation during the last decade on the origin, physiology, and properties of the phagocytes termed “professional phagocytes. The study of microbe phagocytosis showed that in most cases, microbes are killed and degraded either by blood phagocytes [polymorphonuclear (PMN) leukocytes], or by monocytes or macrophages. Some microbes (bacteria, fungi, protozoa), however, can either avoid phagocytosis and so invade the tissues, or survive and multiply inside macrophages. The ability of intracellular parasites to escape from the microbicidal activity of professional phagocytes confers to them pathogenic properties which in many cases create severe problems of public health. The study of pathogen-host cell interplay has given rise to a huge and complex amount of data. It is therefore out of the question to give a complete survey of this field in the present article and to mention all the literature. In the different sections, we shall first describe what is known of the sequential stages constituting microbe ingestion, killing, and digestion. We shall examine at the end of each section how pathogenic microbes can avoid or inhibit these different events, by giving the most characteristic examples. ”

11. Adhesion

The initial and obligatory event involved in the phagocytic process is the contact and adhesion of molecules or particles to the cell surface. The adhesion process was especially well studied for particles by using either latex beads or isolated cells such as bacteria or erythrocytes. Adhesion depends upon the surface properties of both the particle and the host cell. One of the factors which could play a role in cell adhesion is the surface electrostatic charges. Because vertebrate and more primitive cells such as amoebae bear a net negative charge given by glycoproteins and mucopolysaccharides of their glycocalyx (Braatz-Schade and Stockem, 1973; Brandt and Freeman, 1967a; Sherbert, 1978; Stockem, 1977; Weiss, 1969), one could expect that adhesion of positively charged molecules or particles would be easier and stronger than that of negatively charged ones on account of repulsive forces. This problem has been investigated either by using molecules or particles with different electric charges or by modifying the host cell surface charges. Despite the large amount of work on this topic, no clear conclusions can be drawn for several reasons: 1. According to Gingell and Vince (1982), the correlation between cell adhesion and cell surface electrostatic properties depends upon salt concentration.



2. A wide variety of particles (latex beads, oil droplets, bacteria, erythrocytes, chlorella, yeast) and phagocytes (ameboid cells, digestive cells of hydrae, slime mold, polymorphonuclear leukocytes, macrophages) have been used. 3. Experimental conditions differ from one study to another: modifications of surface charges were performed with different concentrations of polyanions or polycations. These charged molecules were applied either to particles or to phagocytes, before or while they were in contact. 4. Removal of negative charges from the host cell surface by neuraminidase could alter other cell properties. 5 . Finally, in most cases, the authors measured particle ingestion instead of adhesion. This may introduce some distortion because adhesion and ingestion are two distinct events that depend upon different parameters (Ito et af., 1981; Rabinovitch, 1967a). For instance, we observed that latex beads could saturate the cell surface of Dicryosfelium discoideum ameboid cells although few beads were ingested (unpublished results). Despite this confusing situation, it seems possible to conclude that cationic substances generally enhance adhesion (Beukers et af., 1980; Capo et af., 1981; Davier et af., 1981 ; Deierkauf et al., 1977; Kooistra et af., 1980; McNeil et al., 1981; Pruzanski and Saito, 1978; Westwood and Longstaff, 1976). As observed in electron microscopy (Capo et af.,1981), cationic substances favor the formation of large contact areas between the particle and the phagocyte. Similar results were obtained upon removal of negative charges mostly due to sialic acid residues from the erythrocyte surface by neuraminida: 2, prior to adhesion (Capo et af., 1981). However, electrostatic forces do not play the main role in particle adhesion because, in most cases, the prey and the phagocyte are both negatively charged (Beveridge, 1980; Brandt and Freeman, 1967a; Sherbert, 1978; Weiss, 1969). This is the case for primitive cells (Protozoa, slime molds) that feed upon bacteria, and for professional phagocytes (polymorphonuclear leukocytes, macrophages) that ingest prokaryotic and eukaryotic cells. The unspecific and specific receptor sites that have been discovered in recent years certainly play a more crucial role. The term unspecific binding sites is used for adhesion of latex, oil droplets, glutaraldehyde-fixed cells, and any misunderstood attachment (Benoliel et al., 1980; Rabinovitch, 1967b; Stossel, 1975; Vogel er al., 1980). In many cases, unspecific binding seems to be due to the hydrophobicity of particles or molecules that can promote physical forces leading to adhesion between particle and cell surface. From van Oss (1978) and Mudd et af. (1934) high interfacial tension between the particle and the surrounding medium but low interfacial tension against the phagocytic cells favor binding and also engulfment. Unfortunately, at the present time no theory can satisfactorily predict how alterations of any one of these properties will affect particle adhesion.



In addition to hydrophobic forces, more specific binding sites have been found. Some of them behave as lectin receptors and recognize a specific sugar (glucose, mannose, galactose, etc.). They are located on the phagocyte cell surface (Glass et al., 1981; Stahl et al., 1978; Sung et al., 1983; Vogel et al., 1980; Warr, 1980; Weir and Ogmundsdottir, 1977). This was shown by the fact that uptake of bacteria or glycoproteins was inhibited when macrophages were previously treated by one of these sugars. Lectin-like receptors also exist on bacteria and yeast which no longer adhere to macrophages when they have been pretreated with mannose (Bar-Shavit et al., 1977; Sandin et al., 1982), or with N-acetyl-glucosamine (Levy, 1979). A similar situation was observed for the binding of yeast cells to Dictyostelium vegetative phase cells. When N-acetylglucosamine residues are blocked by wheat germ agglutinin or are missing from the cell surface of the phagocyte, yeast cells do not attach (Hellio and Ryter, 1980; Ryter and Hellio, 1980). Other binding sites seem to exist on pili (bacterial surface extensions). These proteinous thin filiaments have been shown to favor adhesion to varied kinds of cells (Beveridge, 1980; Pearce and Buchanan, 1980; Smith, 1977). Mono- and oligosaccharides containing a-D-mannose residues inhibit pili adhesion suggesting that these polysaccharidescorrespond to surface receptors for pili (Pearce and Buchanan , 1980). At last truly specific receptors were discovered some years ago on macrophages and leukocytes. Two receptors react specifically with the Fc portion of IgG molecules (Silverstein et af.,1977; Unkeless et al., 1981). One Fc receptor is trypsin, chymotrypsin, and pronase resistant and mediates the efficient binding and ingestion of IgG-antigen complexes (Arend and Mannik, 1972) and 1gGcoated particles (Griffin et al., 1975a; Mantovani et al., 1972). The other Fc receptor is trypsin-sensitive and binds to subclasses of IgG. Immunoglobulins of these subclasses are called cytophilic antibodies because they bind with high affinity to macrophage Fc receptors in the absence of antigens (Arend and Mannik, 1972). The other type of specific receptor found in professional phagocytes binds to C3, one of the complement proteins present in the serum. C3 is an inactive precursor molecule consisting of a heavy chain and a light chain joined by disulfide bridges (Silverstein et al., 1977). Specific proteases cleave a small fragment of the heavy chain, converting this inactive form into a molecule called C3b that specifically binds to C3 receptors (Reid and Porter, 1981). Both kinds of specific receptors are quite different and independent from nonspecific receptors (Griffin and Silverstein, 1974; Michl et al., 1976; Rabinovitch, 1967b). The respective role of Fc and C3b receptors in attachment or ingestion has been controvertial for many years. Some studies show that C3b receptors of both neutrophils and mononuclear phagocytes mediate attachment only whereas Fc receptors induce particle engulfment (Griffin, 1982; Hed, 1981; Hed and Stendahl, 1982). Others indicate that C3b also promotes ingestion (Muschel et al., 1977; Segerling et al., 1982; Shaw and Griffin, 1981). In a quite recent review,


29 1

Griffin (1982) discussed the difficulties linked to these studies and the possible reasons for these contradictory results. He finally proposed, as Ehlenberger and Nussenzweig (1977), that “the chief role of C3b and its receptors is to promote particle binding; the chief role of IgG and its receptors is to promote particle ingestion; the enhanced particle binding by C3b and its receptors serves to facilitate engagement of the cell’s phagocytic signal-generating Fc receptors by IgG.” Particle adhesion to cells bearing ligaad surface receptors is generally temperature and energy-independent (Benoliel et a l . , 1980; Griffin et a f . , 1975b) whereas unspecific binding may or may not depend upon these factors (Benoliel et al., 1980; Glynn, 1981; Michl et a f . , 1976; Rabinovitch, 1967b). Several authors have shown by electron microscopy that IgG-coated erythrocytes form tight and continuous contact areas with the host cell surface (Benoliel et al., 1980; Griffin et a f . , 1975b; Kaplan, 1977; Munthe-Kaas et al., 1976). This is not the case for C3 receptors which establish discontinuous contacts (Kaplan, 1977; Munthe-Kaas et al.. 1976). For unspecific or more specific binding sites, contacts differ according to the particle and the cell. In macrophages, glutaraldehyde-fixed erythrocytes and latex beads promote discontinuous binding areas whereas in Dictyostelium discoideum ameboid cells latex beads are tightly bound. In the latter cells, fixed yeast or bacteria are both loosely bound despite the presence of lectin-like receptors for yeast (Ryter and Hellio, 1980). This shows that no general rules can be established on the mode of attachment to specific or unspecific receptors. After describing the different factors intervening in particle adhesion, let us now see what pathways pathogens use to survive. Two kinds of pathogens must be distinguished: those able to multiply outside phagocytes or other cell types (bacteria and fungi), and the facultative or obligatory parasites that multiply inside phagocytes only. It is obvious that the first kind of pathogens must avoid adhesion and phagocytosis to survive whereas, for the second type, adhesion is a prerequisite for survival. Cell surface properties and environmental conditions favorable for their survival are thus quite different. We have seen that cell surface electrostatic charges do not seem to play the main role in adhesion because microbes and phagocytes are both negatively charged. Despite the repulsive forces thus generated, adhesion and ingestion generally take place very quickly. It is thus obvious that this factor does not intervene in the fate of both kinds of pathogens. In contrast, hydrophilic or hydrophobic properties of the microbe surface are much more important. The inhibitory effect of the hydrophilic bacterial surface was especially well studied in the case of smooth and rough strains of Escherichia cofi and Salmonella ophirnurium. The rough variants which are more hydrophobic are much better adsorbed and phagocytosed than the smooth hydrophilic variants that are more pathogenic (Nakano and Saito, 1970; Stendahl et al., 1973, 1981; Stjernstrom et al., 1977). A relationship between pathogenicity and the presence of hydrophilic



capsules has also been shown for Bacteroides fragilis (Simon et al., 1982) and Streptococcus pneumoniae (Roberts, 1979). At last, slime produced in large amounts by Pseudomonas aeruginosa also seems to prevent adhesion and ingestion (Schwarzmann and Boring, 1971). In this case, however, it is not clear whether inhibition is related to the hydrophilic properties of the slime or (and) the presence of a glycoprotein which was shown to inhibit phagocytosis (Bishop et al., 1982). In conclusion, hydrophilic surface properties constitute one of the factors allowing several pathogenic bacteria to escape from adhesion. In contrast, it is not known whether adhesion and ingestion of facultative or obligatory parasites is favored by their surface hydrophobicity. The only clue suggesting that this factor can intervene is the presence of glycolipids on the cell wall of Mycobacterium species (Goren et al., 1980) which could confer strong hydrophobic properties. Obviously, the crucial factor in microbe adhesion is the presence of receptors on phagocytes or microbes. Lectin-like receptors are very efficient for adhesion, and their interaction with the polysaccharidic glycocalix of many bacteria and fungi has been well demonstrated for plant and animal cells (Costerton et al., 1981; Pistol, 1981; Smith, 1977). Although this type of linkage seems to play a role in pathogenicity (Costerton et al., 1981; Smith, 1977) its role in microbe uptake by professional phagocytes is still poorly documented. It was shown that lectin-like receptors implicated in bacterial pili adhesion do not necessarily enhance uptake of piliated bacteria by professional phagocytes. The relationship between the presence of pili and the rate of ingestion, particularly well investigated for Neisseria and E. coli, showed that pili generally enhance uptake of piliated E. coli strains (Blumenstock and Jann, 1982; Silveblatt et al., 1979) but reduce that of Neisseria gonorrhoeae (King and Swanson, 1978; Thomas et al., 1972; Witt et al., 1976). Attachment of intracellular parasites such as Trypanosoma cruzi (Alcantara and Brener, 1980; Nogueira and Cohn, 1976), Chlamydiapsittaci, and C . trachomatis (Byme and Moulder, 1978; Levy, 1979) seems to be mediated by a protease-sensitive component of the host cell surface, which has not yet been identified as a lectin. Fc and C3 receptors found on the professional phagocyte surface are of great importance in defense mechanisms. The numerous studies on the effects of opsonization illustrate the importance of these receptors for microbe adhesion and ingestion. The increased phagocytosis of opsonized microbes was demonstrated for a wide variety of bacteria, fungi, and Protozoa. We will mention only one conclusive example concerning Mycoplasma: these microbes can grow on the surface of cultured fibroblasts or macrophages and are immediately ingested upon addition of serum (Jones and Hirsch, 1971). This means that antibodies present in the serum of immune animals or man considerably decrease the probability that pathogenic bacteria escape from adhesion. However, the presence of a



capsule can suppress the opsonization effect. Capsules are usually less antigenic than cell wall components and can mask antigen-antibody complexes located in a deeper layer, thus impairing binding between IgG and Fc receptors (Horwitz, 1982; Simon er al., 1982; Stinson and van Oss, 1971; Wilkinson et al., 1979). This shows that capsulated bacteria have a greater chance of avoiding adhesion because of their hydrophilic surface and this masking phenomenon. Opsonization of obligatory parasites favors their adhesion and engulfment but may also induce their killing after phagocytosis because, as described in Section IV, the presence of IgG on the prey’s surface triggers the microbicidal mechanism of professional phagocytes. Opsonization therefore decreases the chance of survival of all kinds of pathogens. It is not excluded, in contrast, that C3 receptors are implicated in Leishmania attachment to macrophages (Mauel, 1980) without inducing its killing in the absence of IgG, and in Babesia adhesion to the erythrocyte surface (Chapman and Ward, 1977). Another way of escaping from adhesion for nonparasitic pathogens is the production of vaned toxic substances that impair professional phagocytes or inhibit chemotaxis and phagocytosis. These substances are produced by many bacterial species (Alouf, 1976; Ofek er al., 1972; Schwab, 1975). Some are located on the cell wall, for example, protein M of Streptococcus (Fox, 1976; Jones and Schwab, 1970) and protein A of Staphylococcus (Dossett er al., 1969; Iwata and Uchida, 1980; Musher et al., 1981). Other toxins are excreted by bacteria such as Streptococcus (Alouf, 1980; Bernheimer, 1974; Roberts, 1979; Weissmann er al., 1963). Pseudomonas aeruginosa (Nonoyama et al., 1979), Borderella perrusis (Utsumi er al., 1978), Serratia miscotorum (Saeki et al., 1974), and Corynebacrerium diphtheria (Pappenheimer, 1977). Even viruses and tumor cells (Fauve and Hevin, 1980) can produce substances that inhibit PMN leukocyte chemotaxis and suppress the normal inflammatory response.

111. Ingestion Ingestion corresponds to particle capture by membrane extensions that surround and finally fully enclose the particle in a phagocytic vacuole. The shape of these cell processes and their connection with the particle surface differ according to the nature of the particle and the kind of phagocyte. In some cases, the host plasma membrane forms thin filopodia that establish discrete contacts with particles (McNeil et al., 1981). In other cases, it forms thin lamellipodia that remain tightly apposed to the particle surface throughout the engulfment process (Silverstein et al., 1977; Stossel, 1976; Stossel and Hartwig, 1976). This mode of ingestion was observed for opsonized particles phagocytosed by macrophages



and is due to their interaction with Fc receptors (Griffin et al., 1975b). Griffin et al. (1975b, 1976) proposed that the initial interaction of immune ligands with the particle generates a process that requires the continuous apposition of receptors to ligands until the particle is fully enclosed within the phagocytic vacuole. This mechanism, called the “zipper mechanism,” was not shown for all receptors (Kaplan, 1977; MacRae et a f . , 1980; Munthe-Kaas et a f . , 1976; Ryter and Hellio, 1980). Inversely, during latex uptake that is not related to specific receptors but rather to bead hydrophobicity tight contacts are established (personal observation). Ingestion is quite rapid; in favorable conditions it occurs within 10-30 seconds after the first contact between phagocyte and prey (Bowers, 1980; MacRae ti?al., 1980). It is temperature dependent since it does not take place below 18’C in professional phagocytes (Silverstein et a f . , 1977) and 16°C in Acanthamoeba castelfanii for which the optimal growth temperature is 30°C (Bowers, 1977). This threshold is probably lower for primitive phagocytes growing around 20°C in their natural environment. Ingestion also requires metabolic energy (Stossel, 1975) but the nature of the cellular organelle consuming this chemical energy is unknown. One likely candidate is the contractile apparatus. On thin section electron micrographs one observes immediately below the membrane of the phagocytic cup a thick zone from which all cytoplasmic organelles are excluded by an anastomosing meshwork of microfilaments. Immunofluorescence (Stendahl et al., 1980; Valerius et a f . , 1981) and the heavy meromyosin labeling technique applied to glycerinated cells demonstrated that this meshwork contains actin, actin-binding proteins, and myosin (Taylor and Condeelis, 1979). It is clear that the appearance of the network closely follows particle adhesion but the mechanism which triggers contractile protein mobilization is not yet elucidated. Griffin et af.(1976) proposed that in the case of IgG-coated particles, the ligand-receptor interaction generates a signal (maybe the release of actinbinding protein from the plasma membrane) that would initiate polymerization and aggregation of contractile proteins and lead to the extension of phagocytic pseudopods. Pseudopod extension would bring about further receptor-ligand interaction and this in turn would generate further contractile protein association. This hypothesis is certainly very attractive but cannot account for the many cases in which contacts between particle and phagocytic membrane are loose and discontinuous. From scanning electron microscopic observations made during the engulfment process, it appeared that contact areas were randomly located along pseudopods or filopods (MacRae et a f . , 1980; Saint-Guillain e? a f . , 1980; personal observation). This means that the signal triggering contractile protein mobilization can propagate even in the absence of a progressive and permanent contact between particle and phagocyte. The factors triggering the dissociation of the filament network once a particle



has been internalized also remain unknown. Experiments in which ingestion was slowed down by particle coating with Con A showed that the zone of organelle exclusion corresponding to the microfilament network disappeared from the bottom of the phagocytic cup before phagosome closure (de Chastellier and Ryter, 1982). This suggests that phagosome closure is not a prerequisite for local microfilament network dissociation. Different electron microscope techniques have revealed changes in the phagocytic membrane during the ingestion process. One of these modifications, that seems to be directly related to the extension of phagocytic pseudopods and the mobilization of contractile proteins, was observed in Dictyostelium discoideum. The use of Oschman and Wall’s technique (glutaraldehyde fixation in the presence of calcium) (Oschman and Wall, 1972) showed that calcium deposits were especially abundant along the inner face of the filopod and phagocytic cup plasma membrane (de Chastellier and Ryter, 1981, 1982), that is to say where the membrane was underlayered with a filament network. Slackened phagocytosis and chemotaxis experiments indicated that the formation of calcium deposits was related to the mobilization of contractile proteins (de Chastellier and Ryter, 1982) and not to their dissociation. These calcium deposits seem to result from a phosphatase activity but their exact meaning is not yet established. The most attractive hypothesis is that they correspond to calcium channels, more especially as many authors have shown that calcium promotes phagocytosis (Hartwig et al., 1980; Stossel, 1975) and is more abundant in actin filament-rich cellular regions (Taylor et al., 1980). Changes in membrane coat polysaccharides were also found during endocytosis in different cells. Sialic acid residues were no longer detectable in pinocytic pits in an epithelial cell (de Bruyn er al., 1978), and wheat germ agglutinin receptors (mainly N-acetyl-glycosamine residues) disappeared from the phagocytic membrane of Dictyostelium discoideum during yeast ingestion (Ryter and Hellio, 1980). Several years ago, Brandt and Freeman (1967b) observed that in the amoeba Chaos chaos the plasma membrane electric resistence considerably decreased prior to the formation of endocytic channels; this was accompanied by a doubling of the electron transparent inner side of the membrane. A spin label study of macrophage plasma membrane also suggested that a nonrandom clustering of lipids and surface proteins occurred during phagocytosis (Horvath et al., 1981). Another modification concerning membrane lipids was reported by Karnovsky and Wallach (1961) and Sastry and Hokin (1966). These authors observed an increased phosphorus incorporation into phosphatidylinosito1 and phosphatidic acid during phagocytosis, that seemed to be related to phospholipid degradation. It is not known whether this phenomenon is related to changes in membrane viscosity (Berlin and Fera, 1977) or membrane depolarization (Horvath et al., 1981), also associated with phagocytosis. Frustrated phagocytosis performed by spreading eosinophils on large nonphagocytosable sur-



faces coated with anti-antibody complexes showed that two new proteins became accessible to iodination in the early attachment phase (Thorne et al., 1980) as if vertical movements of proteins had taken place during this process. Very similar observations have been made in macrophages during phagocytosis (Howard et a f . , 1982). The observation of freeze-fractured preparations during phagocytosis showed, however, no modifications in the number, size, or distribution of intramembrane particles (Favard-SCrCno et al., 1981). This suggests that the changes in lipids, cell coat polysaccharides, or proteins observed during endocytosis do not correspond to a major redistribution of proteins in phagocytic membranes. All these data indicate that the plasma membrane undergoes an important remodeling during endocytosis. The reason for these changes remains obscure. In particular, it is not known to what extent they actually deal with plasma membrane distortion occurring during the engulfment process and its interaction with contractile proteins. Interaction between pathogens and phagocyte during ingestion has not yet been studied. It is obvious that ingestion of bacteria is ensured by the phagocyte and not by the prey, because inert particles and dead microbes are phagocytosed as well as living bacteria. Some bacteria may, however, inhibit their own phagocytosis by excreting toxins. By changing the phagocyte membrane permeability, the latter could prevent or block contractile protein mobilization as suggested by Manjula and Fischetti (1980) for Streptococcus M protein. Penetration of obligatory parasites for which ingestion is an obligatory event does not seem to correspond to an active mechanism either (Baker and Liston, 1978; Mauel, 1980). It is only in the case of blood parasites, such as Pfasmodiunrand Babesia, that penetration into erythrocytes is apparently ensured by the parasite itself (Aikawa et al., 1978). This situation is however very peculiar because red blood cells are devoid of phagocytic activity.

IV. Microbicidal Activity Before describing the process leading to microbe killing, it is important to point out that microbicidal properties vary considerably with the kind of phagocyte. Polymorphonuclear (PMN) leukocytes, short living cells that constitute the first manifestation of host cellular defense against microbe invasion by the blood stream, exhibit strong and well-defined cytotoxic properties. Those of macrophages that are long living cells vary according to the kind of macrophage (peritoneal, alveolar, monocyte-derived macrophages differentiating in vitro, etc.), the animal species (Nguyen et al., 1982), and mostly upon their physiological state (Cohn, 1978; Karnovsky and Lazdins, 1978; North, 1978). The latter depends upon the stimulations macrophages have been submitted to. Macrophages from unstimulated animals are termed “resident,” those stimulated by



immunological mechanisms in vivo are called “activated,” and those which come from animals having received an injection of various substances (caseinate, glycogen, peptone, thioglycolate, etc.) (North, 1978) are termed “elicited.” Both activated and elicited macrophages display a pronounced ruffling of the plasma membrane, an increased capacity for adhering to and spreading onto a substratum, an enhanced capacity for phagocytosis, and a high number of lysosomes. They exhibit physiological properties such as high acid hydrolase activity, excretion of neutral proteinases, specific elastase and collagenase, plasminogen activator. They can also produce cytotoxic compounds that will be described below. All these biochemical properties are expressed to varied extents in PMN leukocytes and elicited or activated macrophages (Cohn, 1978; Karnovsky and Lazdins, 1978; Nathan and Root, 1977). They also vary according to the substances used for elicitation (Briles et al., 1981). It is thus not surprising that the large amount of data accumulated for the past 10 years on the professional phagocyte microbicidal response to microbe infection often appear contradictory or at least confusing. This is also due to the fact that microbicidal activity corresponds to a quite complex phenomenon which is not totally elucidated. Roughly, cytotoxic properties of PMN leukocytes and macrophages are of two orders: the production of toxic oxidative compounds and that of toxic cationic proteins. A. OXIDATIVE KILLING Many years ago, Baldridge and Gerard (1933) observed that phagocytosis in leukocytes and granulocytes was accompanied by a marked oxygen consumption that was shown to be insensitive to cyanide (Klebanoff, 1982). In the past 20 years there have been major advances in our understanding of this phenomenon and its role in the microbicidal and cytotoxic properties of phagocytes (Allen, 1979; Badwey and Karnovsky, 1980; Klebanoff, 1982; Klebanoff and Clark, 1978; Weiss and Lobuglio, 1982). It is now well established that the enhanced oxygen consumption is used for production of hydrogen peroxide and other toxic products such as superoxide anions (0,- ), hydroxyl radicals (-OH), and singlet This phenomenon has been termed “respiratory burst” or “oxoxygen (‘0,). idative metabolism. Superoxide anion is the primary product of the respiratory burst. It results from the activity of a plasma membrane NADH or NADPH oxidase: ”


+ 2 0 2 + 2 02- + NAD(P) + H +

Hydrogen peroxide is formed from two superoxide anions by a dismutation reaction, in which one radical is oxidized and the other reduced: 02-

+ 0 2 - + 2 H + + 0 2 + H202

This dismutation can occur spontaneously or be catalyzed by superoxide dismutase (Fridovich, 1975).



The toxicity of H,O, is enhanced many fold by its transformation into a variety of highly toxic substances. As an example, the presence of peroxidase, a halide, and H,O, generates HOCI, OC1- , and CI,, of which concentration is pH dependent (Klebanoff, 1982). This catalysis is of special interest because peroxidases are present in high concentrations in certain phagocytes (Daems et af., 1979; van Furth et al., 1970; Nichols et af., 1971; Root and Stossel, 1974; Simmons and Karnovsky, 1973). The release of PMN leukocyte myeloperoxidase into phagosomes has been detected by cytochemistry (Briggs et af., 1975b). In addition, the microbicidal effect of this enzyme was demonstrated in a very convincing way by Locksley el af. (1982). Living toxoplasms, uncoated or coated with eosinophil peroxidase, were given to resident peritoneal macrophages which lack peroxidase granules. The peroxidase-coated toxoplasms were killed whereas uncoated ones survived. Recent observations made in patients with myeloperoxidase deficiency indicate however that this enzymatic reaction is not a major microbicidal mechanism of phagocytic cells (Thong, 1982). In the presence of superoxide anions and traces of metal, hydrogen peroxide can also lead to the production of singlet radicals and hydroxide radicals which are both very toxic because they react with a wide variety of cellular compounds. The interest for singlet oxygen involvement in phagocytosis began with the finding that in PMN leukocytes phagocytosis is associated with the emission of light (Allen, 1979). Subsequent studies indicated that chemiluminescence is a general property of stimulated phagocytes and is due to two general mechanisms, one involving peroxidase and the other the production of superoxide anions (Klebanoff, 1982). This property is now widely used to estimate quantitatively the phagocyte oxidative response to the endocytic process. Stimulation of oxidative metabolism is a very rapid process (20 to 50 seconds) (Root and Stossel, 1974) triggered by molecule or particle adhesion. This event corresponds to the induction of a plasma membrane enzyme, the NADH or NADPH oxidase. Cytochemical staining for this enzyme applied to PMN leukocytes showed that the final reaction product was located on the outer face of the plasma membrane (Briggs et af., 1975a) (Fig. 1). This does not prove however that NAD(P)H oxidase is exposed to this side of the membrane (Badwey et af., 1980; Karnovsky et al., 1981). After phagocytosis, a cerium precipitate resulting from the cytochemical reaction was found along the phagosomal membrane (Fig. 3). This indicates that the killing process takes place during and after phagocytosis. Oxidative products are also released into the extracellular medium and can impair nearby microbes. The generation of oxygen metabolites was first studied in PMN leukocytes. It was more difficult to detect in macrophages than in PMN leukocytes, first, because the reaction is generally weaker, and second, because it varies according to the physiological state of macrophages (Badwey and Karnovsky, 1980; Cohn,

Fic. I . Cytochemical demonstration of H202formation by the technique of Briggs er al. (1975a) in a PMN leukocyte incubated for 20 minutes in the presence of phorbolymyristate acetate. The dense deposit is visible along the plasma membrane and in some pinosomes (observations made by Ryter et al.. 1983b). FIG. 2. The same cytochemical reaction as in Fig. I , performed in elicited alveolar macrophages. In this phagocyte, the oxidative activity is much weaker than in PMN leukocytes. The dense precipitate is visible in pinosomes but is absent from the plasma membrane (arrows) (Ryter et al., 1983b).

FIG. 3. Cytochemical demonstration of H202 production during phagocytosis of zymosan (2)by PMN leukocytes. The dense precipitate is visible along the plasma membrane and phagosome membrane (courtesy of Karnovsky er al.. 1981).



1978; Johnston et al., 1978; Kamovsky and Lazdins, 1978; Nathan and Root, 1977). NADH oxidase activity can, in fact, be visualized by cytochemistry only in activated or elicited macrophages (personal observation) (Fig. 2).

Although NAD(P)H oxidase seems to be the triggering site of the oxidative burst, the mechanism of induction is not yet understood. From transmembrane potential measurements, Miles et al. (1981) proposed that superoxide anion production is related to the membrane depolarization observed during phagocytosis. This potential change would precede superoxide production and act as a signal for the phagocytic response. On the other hand, Badwey and Karnovsky (1980) showed that the signal that triggers the oxidative burst is Ca or Mg dependent. However, the role of these cations remains obscure because calcium at certain concentrations can also inhibit NAD(P)H oxidase (Lew and Stossel, 1980). Only certain kinds of molecules stimulate NAD(P)H oxidase. Nonopsonized microbes generally have no effect whereas IgG and some other substances [phorbolmyristateacetate (PMA), ionophores] induce this activity (Figs. 1 and 2) (Badwey and Kamovsky, 1980). The stimulation of the oxygen burst by immunoglobulins suggests that Fc receptors could be implicated in this mechanism (Henrichs et al., 1982). Their role in microbe killing was especially well demonstrated for opsonized Toxoplasma (Wilson et al., 1980), Trypanosoma cruzi (Nathan el al., 1979), and Trypanosoma dionisii (Thorne et al., 1978). In many other studies, however, the role of IgG as a trigger of the killing process was less conclusive because the stimulating effect of IgG on phagocytosis, killing, or digestion was not studied independently. The possible role of sialic acid in the initiation of the oxidative metabolism was also studied but the results are contradictory. One study suggested that it is required for the stimulation of this process (Tsan and McIntyre, 1976) whereas others showed that it is its removal by neuraminidase that stimulates superoxide production (Henrichs et al., 1982; Mills et al., 1981). Despite the lack of knowledge on the triggering mechanism of the oxidative burst, it is now admitted that the production of oxidative compounds plays a crucial role in the microbicidal properties of professional phagocytes. This was demonstrated by the relationship between 0, and H,O, production and killing properties, and especially by the study of phagocytes isolated from patients with chronic granulomatous disease that have poor oxidative properties and low bactericidal activity (Klebanoff and Clark, 1978; Quie, 1972). It is obvious that microbes that can survive inside professional phagocytes use defense mechanisms against the oxidative burst. Three possibilities can be envisaged: (1) parasites inhibit or do not trigger NAD(P)H oxidase activity; (2) they are insensitive to oxidative products; or (3) their enzymatic equipment allows them to rapidly destroy toxic substances or prevent their transformation into more toxic ones.


30 1

As said above, triggering of the oxidative burst may be prevented by the absence of opsonins on the microbe surface. In addition, the chemical surface properties of the microbe certainly play a role in this process but there are no data on this topic. The second mode of resistance concerns the microbe insensitivity to toxic substances. Resistance to hydrogen peroxide and 0,- was reported for Mycobacteria (Jackett er al., 1978; Mitchison et al., 1963). These authors showed that virulent strains are more resistant to H,O, than avirulent ones. Jackett et al. (1978) concluded, however, that the loss of virulence in attenuated strains is not necessarily linked to a loss of resistance; other factors are probably involved in the mechanism of attenuation. Toxoplasma and Trypanosoma were also shown to be very resistant to hydrogen peroxide (Murray and Cohn, 1979; Nathan et al., 1979). The enzymatic content of microbes may also influence their survival. We have already mentioned that the presence of peroxidase inside phagosomes leads to microbe killing because peroxidase in the presence of hydrogen peroxide and halide gives rise to highly toxic products. Microbes rich in this enzyme are thus more easily killed. Conversely, catalase that degrades H,O, and thereby prevents production of more toxic substances seems to be responsible for the resistance of many pathogenic bacteria (Leijh ef al., 1980). Finally, several bacterial species (Escherichia coli, Proteus, P seudomonas, Salmonella, Klebsiella) are quite rich in superoxide dismutase (Britton et al., 1978). These authors suggested that this enzymatic activity could increase bacterial resistance to superoxide radicals by rapidly degrading them. However, this resistance mechanism does not appear to be very efficient becuase it is generally admitted that 0,- radicals are much less toxic than products resulting from their dismutation.

B. CATIONIC PROTEINS Another antimicrobial agent, ‘‘phagocytin,” was discovered by Hirsch ( 1956) in rabbit granulocytes. Later studies demonstrated that it was localized in lysosomes of PMN leukocytes and macrophages of several animals and man (Avila, 1979). This agent is composed of several proteins that are highly cationic because they are rich in basic amino acids, mainly arginine (Zeya and Spitznagel, 1966, 1968). Their bactericidal activities differ with the kind of microbe (bacteria or fungi), the bacterial species, and even the bacterial strain (Elsbach et al., 1979; Patterson et al., 1980; Weiss et al., 1982; Zeya and Spitznagel, 1966, 1968). Some are especially active against gram-positive bacteria (Odeberg and Olsson, 1975) whereas others specifically kill gram-negative bacteria (Weiss et al., 1978). Their visualization under the light microscope with the fast green stain (Spitznagel and Chi, 1963)or under the electron microscope (Weiss et al., 1976)



showed that they bind to the bacterial surface. This attachment is certainly mediated by the outer membrane lipopolysaccharides of gram-negative bacteria or the gram-positive cell wall carbohydrates. Their specificity probably resides in the different nature of these cell surfaces. The main effect of cationic proteins is to change the membrane permeability (Beckerditeet al., 1974; Elsbach et al., 1979; Zeya and Spitznagel, 1966). In the case of the protein isolated from rabbit PMN, that specifically kills gram-negative bacteria, Elsbach et af. (1979) showed that it is associated with phospholipase A. Its bactericidal properties are conserved even after dissociation from this enzyme, thus indicating that the killing property belongs to the cationic protein only. Cationic proteins display microbicidal activity only after they have been released into phagosomes and therefore depend upon phagosome lysosome fusion. Their mode of action is therefore different from that of the oxidative mechanism which is triggered before engulfment and takes place in all phagosomes whatever the time at which lysosome fusion occurs. This also suggests that the resistance of intracellular parasites to cationic proteins or respiratory burst is due to quite different microbe properties or behavior. To our knowledge, such killing mechanisms have not been described in primitive phagocytes for which microbe phagocytosis represents the main feeding process. These phagocytes ingest large amounts of bacteria and probably display efficient microbicidal properties because ingested bacteria are quickly impaired and digested. The first clue to the existence of similar killing properties in these cells resides in the quite recent discovery of a protein isolated from Entanioeba histolyrica that alters the membrane permeability of E . coli (Young et af., 1982). Since its mode of action is analogous to that of cationic proteins it could deal with microbe killing.

V. Phagosome-Lysosome Fusion The fist evidence of fusion between phagosomes and lysosomes was obtained several years ago by light microscopic observations of PMN leukocytes and activated macrophages (Hirsch, 1962; Hirsch et a f . , 1968; Zucker-Franklin and Hirsch, 1964). Dark granules, identified as lysosomes, were no longer visible after phagocytosis or induced pinocytosis. Their disappearance was due to fusions with newly formed endosomes and was named “degranulation.” The term of phagosome-lysosome fusion corresponds to phagosome fusion with primary or secondary lysosomes. In fact, reported observations generally concern fusions with secondary lysosomes because their Occurrence is based on the presence inside phagosomes of fluorescent or electron-dense markers pinocytosed before particle ingestion.



Fusions generally occur soon after phagosome formation. Quantitative measurements of phagosome-lysosome fusion in Dictyosrelium discoideum show a “burst” of fusions within 15 minutes after particle addition (Favard-SCrCno et al., 1981). Fusions are less frequent afterward although phagosomes and lysosomes are still available. It is interesting to note that fusions observed in virro among endosomes and lysosomes isolated from Acanrhamoeba casrellanii also occur mostly during the first 10-15 minutes of incubation (Oates and Touster, 1976, 1978). Absence of fusion is therefore not due to exhaustion of lysosomes but to another factor. In professional phagocytes, fusions are also very rapid and may even take place during particle ingestion. In this case, the lysosomal content can be released into the extracellular medium (Nichols, 1982; Pryzwansky er al., 1979).

Although the mechanisms controlling fusion are unknown, in virro studies performed with natural or artificial membranes have given some information on the conditions in which membrane fusion takes place (Papahadjopoulos er al., 1977; Poste and Allison, 1973). It is now well established that fusion occurs when the phospholipid layers of both membranes are in contact, which implies that membrane proteins must be mobile and pushed apart (Volsky and Loyter, 1978). Protein motility and segregation were demonstrated by freeze-fracture preparations in which phagosome-lysosome fusion areas were smooth and devoid of intramembrane particles (Amherdt er al., 1978; Batz and Wunderlich, 1976; Bowers, 1980; Favard-Strho er al., 1981; Orci et al., 1977). It is thus not surprising that membrane protein immobilization by cross-linking agents such as concanavalin A (Con A) and polyanionic substances inhibits membrane fusion. As shown by the following examples contradictory results were obtained by different authors, indicating that inhibition of fusion also depends upon other factors. Edelson and Cohn (1974) and Storrie (1979) reported that Con A-containing vacuoles do not fuse with secondary lysosomes whereas Goldman and Raz ( 1975) concluded their fusion because these vacuoles contain acid phosphatase. The same discrepancy was found between morphological observations and cytochemistry in Dictyostelium discoideum having phagocytosed Con A-coated yeast. The phagosome membrane remained tightly apposed to the Con A-coated yeast surface for at least 1 hour whereas it immediately loosened around untreated yeast, thus suggesting that fusions with lysosomes were inhibited in the presence of Con A. Cytochemistry showed, however, that acid phosphatase had been discharged into both kinds of phagosomes (Figs. 4, 5, and 6). A likely explanation would be that Con A inhibits fusion with large secondary lysosomes but does not prevent fusion with tiny primary lysosomes. The importance of vesicle size in the fusion process is in good accord with previous in virro experiments (Poste and Allison, 1973; Wilschert et al., 1981).



FIGS. 4 A N D 5 Con A-coated yeast-containing phagosome. Despite the tight apposition of the phagosome membrane to the yeast particle surface, suggesting that no lysosome fusions have occurred (Fig. 4). acid phosphatase activity is cytochemically revealed inside the phagosome. The dense deposit remains located in discrete areas at the periphery of the yeast particle (Fig. 5 ) (personal observation). FIG. 6. Untreated yeast-containing phagosome showing a large space between the yeast surface and phagosome membrane. Acid phosphatase activity is visible inside the yeast particle (personal observation).

Polyanionic substances also seem to inhibit lysosome fusion (Alexander, 1981; Geisow et al., 1980; Goren et al., 1976; Hart and Young, 1978; Kielian and Cohn, 1982). As for Con A, however, the lysosome size plays a role because Kielian et al. (1982) also observed that dextran sulfate vacuoles did not fuse with large phagosomes but fused with small pinosomes. Cyclic AMP has also been suspected of inhibiting lysosome fusion. This assumption comes from the fact that a relation could be established between the low frequency of fusions between Mycobacterium tuberculosis-containing phagosomes and lysosomes and the increased cellular CAMPconcentration induced



by phagocytosis of this bacillus (Lowrie et al., 1975). The absence of such a cAMP increase after phagocytosis of bacteria unable to inhibit lysosome fusion strengthened this hypothesis (Lowrie et al., 1975; Carrol et al., 1979). A direct effect of cAMP on lysosome fusion was however not found in the course of in vitro experiments. Rather, this nucleotide seemed to stimulate fusion between isolated secondary lysosomes and pinosomes (Oates and Touster, 1980). The discrepancy between in vivo and in vitro systems could be due to the fact that cAMP affects several cellular functions in vivo (Confer and Eaton, 1982). Inhibition of fusion observed after Mycobacrerium uptake could be produced by other metabolic perturbations. Finally amines and NH,Cl inhibit phagosome lysosome fusion (Gordon et al., 1980; Hart and Young, 1975; Young er al., 1981). However, the mechanism of this inhibition remains obscure because other amines such as chloroquine seem to stimulate fusion (Hart and Young, 1978, 1979). In conclusion, studies on inhibition of lysosome fusion are still confusing partly for technical reasons. As discussed by Pesanti (1978) and Goren et al. (1980), fluorescent markers must reach a high concentration inside phagosomes in order to be detected. This means that phagosome fusion with a small number of tiny fluorescent lysosomes will not be seen. The electron microscope seems to be a more sensitive tool although a given thin section represents only a small fraction of phagosome and the marker is not necessarily present in this phagosome portion. The comparison of the number of phagosomes containing either an electron-dense marker or a fluorescent marker indeed showed that the number of fusions observed under the electron microscope was higher than that found by fluorescence (Goren et al., 1980). The fact that phagocytes were often loaded with fusion inhibitors for 3-5 days before particle ingestion could also distort the results because the accumulation of these substances could impair the digestive functions of the cells. The use of matrix-bound indicator dyes or fluorescent molecules has shown that the lysosome pH is around 4.5 to 6 (Heiple and Taylor, 1982; Jacques and Bainton, 1978; Kielian and Cohn, 1982; Mandell, 1970; Reijngoud and Tager, 1977). Geisow et al. (1981) and Segal et al. (1981) observed a slight pH increase 2-3 minutes before its lowering. They proposed that this short period of alkaline pH favors cationic protein binding to microbes and thereby accelerates their killing. It is not yet known, however, whether lysosome fusion and acidification are two functionally independent events or not. The dissection of these processes will require techniques with an increased spatial and temporal resolution. Two mechanisms could be responsible for the acidification of lysosome pH. Either a membrane bound H+ trans-ATPase driving H+ into lysosomes (Reijngoud and Tager, 1977) or a Donnan equilibrium generated by the presence of glycoproteins with low isoelectric point (Goldman and Rottenberg, 1973). At



present, none of these hypotheses has been definitively proved (Reijngoud, 1978) but recent observations made in different laboratories are in favor of the first one. The acidic environment constitutes optimal conditions for lysosomal enzyme activity. The latter comprises several hydrolases active against lipids, nucleic acids, complex lipids, polysaccharides, glycoproteins, or proteins. In theory, many of these enzymes should degrade the lysosomal membrane. Although several authors have tried to determine whether the plasma membrane was modified after its internalization and contact with lysosomal enzymes, this problem is still a matter of debate. This question is difficult to study first, because the phagosome membrane fraction is more or less contaminated by other cell membranes, and second, because it is difficult to distinguish the modifications that occur during the engulfment process from those resulting from lysosomal enzyme degradation. The large amount of data obtained up to now seem to indicate however that the phagosome membrane is very resistant and that its degradation is neglectible during its life span (Steinman et a f . , 1983). Let us now see how parasites escape from cationic protein killing and lysosomal hydrolase degradation. Several ways are used: (1) inhibition of phagosome-lysosome fusion; (2) resistance to lysosomal enzymes; and (3) escape from phagosome. 1. Inhibition of phagosome-lysosome fusion. This mode of resistance seems to be the most commonly used by bacteria (Goren, 1977). It was especially well studied in the case of Mycobacterium tuberculosis and related strains (Armstrong and d'Arcy Hart, 1975; Hart et a f . , 1972). At least four substances produced by M . tuberculosis under appropriate conditions are susceptible to inhibit phagosome-lysosome fusion: sulfatide, polyglutamic acid, CAMP,and ammonia. Sulfatide (2,3,6,6'-tetraacetyl-trehalose-2'-sulfate) is a strong anionic substance which corresponds to the major representative of a group of unusual sulfated glycolipids produced by this bacillus (Middlebrook et a f . , 1959). A correlation between the production of sulfatides and virulence was established. Goren et a f . (1974, 1976) showed that yeast-containing phagosomes did not fuse with secondary lysosomes loaded with fluorescent or electron-dense markers when macrophages had previously pinocytosed sulfatides. It still remains to be shown that the amount of sulfatides naturally produced by M . tuberculosis is sufficient to inhibit lysosome fusion. In addition, its precise location inside the bacterial cell wall has not been established. It is thus not known whether it is in contact with the phagosome membrane or not (Goren et al., 1980). Electron micrographs show that the virulent strains M . tuberculosis (Armstrong and d'Arcy Hart, 1975) and M . avium (personal observation) both inhibit lysosome fusion and are surrounded in phagosomes by an electron-transparent capsule-like zone (Fig. 7) that does not seem to exist in in vitro growing bacteria. This material, also found in M . fepraemurium(Hart et al., 1972), was isolated

FIGS.7 AND 8. Macrophage phagosomes containing bacteria of the pathogenic strain M. aviurn (Fig. 7) or the nonpathogenic strain M. aurum (Fig. 8) (DL dense lysosomal material). In Fig. 7 , bacteria are surrounded by an electron-transparent zone (arrows) which is not visible in Fig. 8 (Frihel e t a / . , 1983).



and shown to correspond to “mycoside,” a complex peptidoglycolipid usually located in the Mycobacterium cell wall (Draper and Rees, 1973). It is interesting that the avirulent strain M. aurum is generally not surrounded by this “capsule” and is quickly digested (personal observation; Fig. 8). However, its role in fusion inhibition remains questionable because M. lepraemurium-containing phagosomes fuse with lysosomes. Polyglutamic acid, the accumulation of which in secondary lysosomes was shown to inhibit fusion with phagosomes (Hart and Young, 1978), is naturally covalently linked to the cell wall of virulent strains of M. tuberculosis (Vilkas and Markavits, 1972). The role of this substance is, however, subject to controversy concerning the kind of isomer present in the Mycobacrerium cell wall and its amount, which is not related to bacterial virulence (Draper, 1981). In addition, polyglutamic acid is probably associated with peptidoglycan and therefore is deeply buried in the cell wall. It is thus certainly not in contact with the phagosome membrane. As mentioned above, cyclic AMP concentration was shown to increase inside macrophages after phagocytosis of M. tuberculosis and M. microri (Lowrie er af., 1975). Such an increase was not found after infection by M. fepraemirrium and Salmonella ryphimurium which do not inhibit lysosome fusion (Carrol er af., 1979). The assumption of an inhibitory effect of cAMP was strengthened by the fact that attenuated strains of M. tuberculosis, which inhibit phagosome-lysosome fusion for a limited time, produced only a temporary rise in cAMP (Lowrie et af., 1975). The participation of this cyclic nucleotide is not definitively established, however, because cAMP inhibits other cell processes such as phagocytosis, the oxidative process, the release of lysosomal enzymes, and the killing of intracellular microorganisms (Confer and Eaton, 1982; Draper, 1981). Ammonia is formed in excess in unbalanced physiological conditions by M. tuberculosis and, as other amines, seems to inhibit lysosome fusion (Hart and Young, 1978; Seglen and Reith, 1976). That such a production occurs in phagosomes where nutritional conditions are presumably quite different from those of in virro growth remains to be proved. Lysosome fusion inhibition was also clearly demonstrated several years ago for Toxopfusma gondii (Jones and Hirsch, 1972). The reason for this inhibition is unknown but these authors observed that phagosomes containing intact parasites were surrounded by several mitochondria and endoplasmic reticulum cisternae. These organelles could protect phagosomes against oncoming lysosomes. A relationship between virulence, viability, and lysosome fusion was observed for several fungi and bacteria. This is the case for Coccidioides immitis (Beaman and Holmberg, 1980), Nocazdia arreroides (Davis-Scibienski and Beaman, 1980), and Brucelfa suis (Oberti er af., 1981). The resistance of these pathogens has been much less intensively studied than that of Mycobacteria, and the factors responsible for lysosome inhibition are still unknown.



Rickettsiae and Chlamydiae also escape from lysosome fusion. The way in which Chlamydiae avoid lysosome fusion is still poorly documented (Todd and Storz, 1975; Wyrick and Brownridge, 1978) although the multiplication and development of these parasites have been intensively studied (Becker, 1978; Storz and Spears, 1977). Eissenberg and Wyrick (1981) observed that after double infection with E. coli and Chlamydiae, E. coli was quickly digested whereas Chlamydiae remained intact. They concluded that the factor inhibiting lysosome fusion does not act on the general phagocyte behavior but intervenes locally at the level of Chlamydia-containingphagosomes. It was also shown that the number of phagocytosed microbes plays a crucial role in their survival. When the multiplicity of infection is one Chlamydia per macrophage, phagosomelysosome fusion appears to be inhibited and microbe survival is very high. In contrast, a high multiplicity (100 bacteria per phagocyte) leads to a low survival and phagosome-lysosome fusions were frequently observed (Wyrick and Brownridge, 1978). The comparison of Chlamydia survival in normal PMN leukocytes or in those isolated from patients with myeloperoxidasedeficiency or granulomatousdisease showed that oxygen-dependent antimicrobicidal activity is not essential for Chlamydia killing (Yong et al., 1982). This confirms that phagosome-lysosome fusion is the crucial stage upon which survival or death depends. The mechanism of lysosome fusion inhibition developed by Rickettsia rsutsugamushi seems to depend upon the cytoplasmic environment of their phagosomes. Rikihisa and Ito (1979) observed that phagosomes in which microbes were intact were surrounded by a high number of glycogen particles (Fig. 13). This would isolate phagosomes from other cytoplasmic organelles, and more especially lysosomes, and allow Rickettsiae to dissolve the phagosome membrane and become free in the cytoplasm before phagosome-lysosome fusion. It must be pointed out, however, that inhibition of fusion is never total whatever the pathogen. One always finds a portion of phagosomes that has fused and contains degraded microbes. The number of fusions and microbe survival seem to depend upon many parameters dealing with the number of infecting microbes, and the state of activation of phagocytes. One of the factors which seems to promote phagosome-lysosome fusion is microbe opsonization. The presence of IgG on the microbe surface could mask surface exposed substances that normally inhibit lysosome fusion. In most studies, however, IgG stimulation of lysosome fusion could not be distinguished from IgG stimulation of the oxidative burst and microbe killing. It is therefore quite possible that the increased number of fusions also results from the increased amount of killed microbes that can no longer inhibit lysosome fusion. 2. Resistance to lysosomal hydrolases. This second mechanism of survival inside phagosomes seems to be less commonly used by pathogens. Only three microbes are known to display this property.



The first one is Mycobacferium lepraemurium for which phagosome-lysosome fusion was observed in cultured macrophages and fibroblasts (Brown and Draper, 1970, 1976; Hart et al., 1972) and in vivo (Brown and Draper, 1976). This microbe seems to be especially resistant to cationic proteins and lysosomal enzymes. The presence of “mycoside” which surrounds microbes inside phagosomes (Draper, 1981) could be responsible for this property. A rather similar compound has been found in other Mycobacferium species. Cytochemical staining for acid phosphatase showed that lead precipitate remained located at the periphery of this transparent material in phagosomes containing the pathogenic strain M . avium; in the case of M. aurum, a nonpathogenic strain that does not form this capsule, the precipitate was in direct contact with bacteria (Frkhel et al., 1983; Figs. 9 and 10). This suggests that this material could act as a barrier to lysosomal enzyme diffusion. As in the case of fusion inhibition, resistance to lysosomal enzymes is never total. Generally, many M . leprae are killed during the first days of infection. The survivers can later multiply and invade the host cell.

FIGS.9 AND 10. Cytochemical demonstration of acid phosphatase in phagosomes containing M.

aviurn (Fig. 9) or M. aurum (Fig. 10). While the dense precipitate is in contact with M.aurum and

tends to penetrate inside bacteria, it remains located at the periphery of the transparent zone surrounding M. nurium in Fig. 9, as if this material inhibited lysosomal enzyme diffusion (Frkhel er al., 1983).


31 1

The two other pathogens presenting a high resistance to hydrolases are Protozoa: Leishmaniae and Trypanosomas. Phagosome-lysosome fusion was detected after Leishmania phagocytosis by using fluorescent or electron-dense markers (Alexander and Vickerman, 1975; Chang and Dwyer, 1976). The electron-dense lysosomal content was also detected inside Leishmania-containing phagosomes of bone marrow-derived macrophages (Ryter e? al., 1983a) (Fig. 11) and staining for acid phosphatase confirmed the presence of this hydrolase in phagosomes (Alexander and Vickerman, 1975; Lewis and Peters, 1977; Ryter et al., 1983). In addition the membrane impermeability of this pathogen was illustrated by the fact that lead precipitate was strictly located in the lysosomal material surrounding the parasites but not in the parasites. After modification of the membrane permeability by killing of Leishmania inside phagosomes (Rabinovitch et af., 1982; Ryter et al., 1983a) lead precipitate was found inside the parasites. The reason for such a great impermeability is not understood. Contrary to Mycobacterium species that have complex cell walls, Leishmania is limited by

FIG. I I . Cytochemical staining for acid phosphatase in macrophages infected with Leishmania mexicana amazoniensis (L). The dense deposit is visible inside phagosomes but remains located in the dense lysosomal material (DL) surrounding the parasites. The latter remain intact and devoid of precipitate except in a lysosome of a Leishmania (L) (Ryter er al.. 1983a).



its plasma membrane only and is not protected by a thick plysaccharide coat (personal observation). Trypanosoma cruzi trypomastigotes also momentarily resist hydrolases. ‘Their phagosomes fuse with lysosomes (Kress et al., 1977; Milder and Kloetzel, 1980; Nogueira and Cohn, 1976) and although part of them are destroyed (Nogueira and Cohn, 1976) some resist and escape later into the host cytoplasm. The resistance of these parasites varies according to the species (Liston and Baker, 1978; Thorne et al., 1979) and their development stage (Nogueira and Cohn, 1976; Thorne et al., 1979). It is not established whether these variations reflect varied hydrolase resistances or varied sensitivities to microbicidal products. 3. Escape from phagosomes. Early evidence that intracellular parasites can escape from macrophage phagosomes and multiply in direct contact with the host cytoplasm was given with Trypanosoma cruzi (Nogueira and Cohn, 1976). This phenomenon seems to occur by the progressive degradation of the phagosome membrane (Fig. 12). Forty-eight hours after infection, most parasites are in the

FIG. 12. Transverse section through an intracellular trypomastigote (T) of Trypanosomacruzi 90 minutes after macrophage infection. The parasite membrane (short arrow) has a typical unit membrane structure with underlying subpellicular microtubules. Another modified membrane is also seen around the parasite. It is thinner and has only a single leaflet (dashed arrow). This membrane corresponds to the phagosome membrane in the course of its degradation (see inset at higher magnification) (courtesy of Nogueira and Cohn, 1976).

FIG.13. Rickettsia tsutsugamushi escaping from a phagosome. Intact Rickettsia-containing phagosomes are generally located in glycogen-richcytoplasmic (upper right inset) areas. Some parasites have "dissolved" the phagosome membrane and are free in the cytoplasm (lower left inset) (courtesy of Rikihisa and Ito, 1979).



cytoplasm and start to multiply. A rather similar process was observed with Rickettsia species. As already mentioned, Rickettsia tsutsugamushi, after avoiding lysosome fusion because of its location in glycogen regions, seems to dissolve the phagosome membrane (Rikihisa and Ito, 1979; Fig. 13). R . rickettsii can also become free in the cytoplasm and is later found inside the endoplasmic reticulum (ER) (Silverman and Wisseman, 1979) (Fig. 14). The way in which it reaches the ER remains obscure but it could occur by budding. Its penetration is accompanied by a considerable swelling of the cisternae leading after 120 hours to the appearance of a huge and unique vacuole. The host cell finally lyses,

FIG. 14. Chicken embryo fibroblast infected with R. rickcrtsii. The rickettsia are located in dilated rough endoplasmic reticulum (short arrows) and are membrane-bound (long arrows) (courtesy of Silverman and Wisseman, 1979).



liberating the microbes which invade neighbor cells. The development of Rickettsia orientalis (Higashi, 1962) and Scrub Typhus Rickettsia (Erwing et a l . , 1978) is slightly different. Parasites free in the cytoplasm extrude directly from the host cell surface by a budding process. It must be pointed out, however, that Rickettsia infection and multiplication have been studied in their normal environment, that is to say in nonprofessional phagocytes that do not possess developed microbicidal properties and a digestive apparatus. They probably survive less easily in PMN leukocytes or macrophages. A recent morphological study of Legionella pneumophila infection showed that these bacteria were killed by PMN leukocytes but survived in macrophages when the number of bacteria was low (Katz and Hashemi, 1982). Some hours after infection, bacteria were found in ER cisternae. As no pictures of phagocytosis could be obtained at low multiplicity, it is not yet known whether bacteria were first located in phagosomes and then reached the ER. Another defense mechanism used from inside the phagosomes of phagocytes by nonobligatory parasites is the production of toxins that rapidly impair and sometimes lyse the host cell. As already mentioned in Sections I1 and 111, many bacteria contain or excrete toxins. In some cases, they may immediately prevent microbe ingestion but some of them can also act later. For example, it has been shown that surface lipids of Corynebacterium ovis induced the swelling of ER, Golgi apparatus, and phagosomes leading finally to macrophage lysis (Hard, 1973). Although no toxin has yet been identified for Legionella pneumophila, it is possible that macrophage destruction observed at high bacterial multiplicity (Katz and Hashemi, 1982) is due to toxin production. Bordetella pertusis was shown to contain and excrete large amounts of adenyl-cyclase (Confer and Eaton, 1982). This enzyme is internalized by phagocytosis and catalyzes the formation of cyclic AMP, thereby disrupting normal phagocyte functions and probably inducing their further killing. This could explain why humans infected by this pathogenic bacterium are quite vulnerable to secondary infection (Confer and Eaton, 1982). The mechanism of membrane penetration used by diphtheria toxin is very interesting because it is pH dependent (van Heyningen, 1981; Sandvig and Olsnes, 1980). The toxin molecule is composed of two peptides: one binds to a membrane receptor and, at low pH, promotes membrane crossing by the other peptide. After phagocytosis of Corynebacterium diphtheria, the toxin could cross the phagosome membrane after pH lowering thereby destroying the phagocyte even after bacterium killing and digestion. Another interesting pH-dependent escape mechanism has been discovered for certain viruses (Helenius et a l . , 1980; Genchault er a l . , 1981). Virus particles that have been phagocytosed can escape from phagosomes after pH lowering. The viral envelope fuses with the phagosome membrane thus liberating viral nucleic acid into the cytoplasm.



VI. Membrane Recycling During Endocytosis Endocytosis results in an extensive interiorization of the plasma membrane varying between 1 and 20 times the total cell surface per hour, depending upon cell type and culture conditions (Bowers and Olszewski, 1972; Bowers et al., 1981; Githens and Karnovsky, 1973;Hubbard and Cohn, 1975; Ryter and de Chastellier, 1977; Steinman et al., 1976, 1983; Stockem, 1973; Weisman and Korn, 1967). This raises two main problems: (1) what is the fate of the interiorized membrane and (2) how is the plasma membrane renewed? As described in the preceding section, the main fate of endocytic vesicles is to fuse with lysosomes and deliver their content to them. Although the surface area of incoming vesicles is much larger than the dimensions of the preexisting secondary lysosomes, the membrane area of the total intracellular compartment preserves a constant value throughout endocytosis in both macrophages and L cells (Steinman er al., 1976). Similar observations were made in ameboid cells (Bowers er al., 1981; Ryter and de Chastellier, 1977). The maintenance of the vacuolar compartment at a constant surface area implies a rapid reduction in vesicle size. Degradation of membrane components by lysosomal enzymes is, however, too slow to ensure this equilibrium (Hubbard and Cohn, 1975). The second problem concerns plasma membrane renewal. As shown in macrophages and ameboid cells, phagocytosis or pinocytosis does not induce a reduction of the cell surface area (Bowers er al., 198I ; Ryter and de Chastellier, 1977; Steinman ef al., 1976). This means that the plasma membrane must be replaced at a corresponding rate. De novo synthesis of membrane constituents is too slow to account for the rapid replacement of internalized membrane (Silverstein er al., 1977; Werb and Cohn, 1972). In addition, all studies on the degradation rate of plasma membrane constituents point to low values (for review see Hubbard, 1978). All these considerations led several authors to propose that plasma membrane internalized during pinocytosis or phagocytosis was recycled back to the cell surface. Indirect evidence for such a recycling came from varied morphological studies in amoebae, professional phagocytes, or other cell types, using endocytic markers, such as HRP, yeast particles, latex beads (Bowers er al., 1981; Ryter and de Chastellier, 1977; Steinman et al., 1976), or noncovalently linked markers to unspecified membrane components, such as cationized ferritin (Farquhar, 1978). Direct evidence for membrane recycling has only recently become available by studying the fate of plasma membrane antigens during pinocytosis (Schneider et af., 1979a,b; Tulkens et al., 1980), or that of radioactive markers, covalently bound to specific plasma membrane constituents, during pinocytosis (Burgert and Thilo, 1983; Thilo and Vogel, 1980) or phagocytosis (de Chastellier er al., 1983; Muller et al., 1980; Storrie et al., 1981). After internalization in the vacuolar compartment, the antigen or label reappears at the cell surface



(Burgert and Thilo, 1983; Muller et al., 1980, 1983; Schneider et al., 1979b; Storrie et al., 1981; Thilo and Vogel, 1980; Tulkens et al., 1980). As shown by the autoradiographic study of Muller et al. (1983), this phenomenon occurs rapidly since labeled phagolysosome membrane constituents reappeared at the cell surface within 5-10 minutes. The concept of recycling raises the question of the organization of this membrane flow. Most likely, the incoming pathway is constituted by the endocytic vesicles, but the cytological nature of the outgoing pathway is unknown. During pinocytosis the interiorized membrane could be returned to the cell surface in the form of small vacuoles, thus establishing a direct shuttle between cell surface and endosomal membranes (Muller er af., 1983; Steinman et al., 1976). These small vesicles could be issued by endosomal membranes that have not yet fused with lysosomes or by secondary lysosomes. The hypothesis of Duncan and Pratten (1977) is that the bulk of the endocytic vesicle membrane is recycled before it has even come into contact with lysosomes. Recent work from Burgert and Thilo (1983) in the macrophage cell line P388D, is in accord with this hypothesis. A similar membrane shuttle was observed during phagocytosis between the phagosome membrane and plasma membrane. In macrophages, the autoradiographic study of Muller et af. (1980) showed that the radioiodinated phagolysosome membrane proteins returned rapidly to the cell surface, most likely via small vesicles that bud from the phagolysosome and subsequently fuse with the plasma membrane. In ameboid cells, in which the endosomal compartment area is at least equal to that of the cell surface (Bowers et af., 1981; Ryter and de Chastellier, 1977), the mixing of membrane between incoming endocytic vesicles and preexisting ones seems predominant with respect to the direct shuttle between phagosome and plasma membrane (de Chastellier er af., 1983). An alternative pathway would include the Golgi elements, on the basis that certain pinocytosed substances appear in the Golgi complex (Farquhar, 1978; Thyberg, 1980; Thyberg et al., 1980). The exact pathway taken by the interiorized membrane between the cell surface and Golgi complex is, however, not clearly established. The most likely explanation is that endocytic vesicles fuse first with lysosomes and then, for membrane retrieval, with the Golgi complex, as suggested by Tulkens et al. (1980) and the recent work of Schwarz and Thilo (1983). This intense membrane traffic between plasma membrane and endosomal membrane shows that the phagosomal membrane is not a stable structure. A continous communication exists between the phagosome content and the extracellular medium: pinocytic vesicles bring extracellular substances to phagosomes and, inversely, the phagosome content is transported outside the cell. The latter process could play a role in the transport of microbe antigens and their exposure to the macrophage surface. At the present time, no data have been obtained on this membrane flow in



professional phagocytes infected by intracellular pathogens. It is quite possible that the presence of living microbes inside phagosomes disturbs membrane flow especially with microbes that inhibit lysosome fusion. Although the loose or tight contact of the phagosome membrane with the internalized particle does not necessarily reflect its capacity to fuse with lysosomes (de Chastellier et al., 1983, see also Section V), it is striking that the membrane of M.avium-containing phagosomes remains in tight apposition with the electron-transparent zone throughout microbe multiplication (personal observation), as if no exchanges took place with other membrane compartments. Inversely, Leishmania mexicana amazonensis-containing phagosomes that fuse with lysosomes are transformed after 24-48 hours into huge parasitophorous vacuoles (Chang, 1980; Rabinovitch et a!., 1982). The size increase which is partly due to fusions with pinocytic vesicles (M. Rabinovitch, personal communication) could reflect an unbalanced membrane flow between phagosome and plasma membrane that could favor parasite multiplication. In conclusion, membrane flow could be an important phenomenon in host-invader interplay and represents an interesting new field of research.

VII. Conclusions This brief review shows that host-invader interplay is a complex phenomenon, that deals with different stages of phagocytosis or of the digestion process according to the pathogen. The best understood mechanism of pathogen escape is that used by bacteria able to grow outside cells. They succeed in avoiding adhesion and phagocytosis because of their peculiar surface properties. In contrast, mechanisms of resistance developed by intracellular pathogens are poorly understood. This is due first to the complexity of cellular processes implicated in engulfment, microbicidal properties, and lysosome fusion. Although great advances have been made in the past 10 years in our knowledge of these events, their exact mechanism and control are still far from being elucidated. Second, pathogen resistance is multifactorial and only exceptionally the property of a single determinant. Microbes must first inhibit or resist the oxidative burst and then neutralize the cationic proteins and lysosomal enzymes either by preventing lysosome fusion or by resisting lysosomal content. The study of these different processes is difficult because they form a cascade of interdependent events. It is obvious that the crucial stage of resistance concerns the oxidative burst because when microbes have been impaired by oxidative products, they can no longer inhibit phagosome-lysosome fusion and lose their resistance properties to lysosomal components. In addition, pathogen resistance to these different processes is seldom an all or



none phenomenon and varies according to microbe multiplicity, which complicates the kinetic study of their behavior. Finally, their resistance also depends upon the defense properties of the phagocyte which vary, for macrophages, with their state of activation and their origin. It is thus difficult to draw general conclusions from results obtained under different experimental conditions. We have focused our attention mostly on pathogen resistance toward professional phagocytes because these cells represent the first cellular defense of animals and they are the best equipped cells for microbe killing. It is obvious that when microbes have escaped from phagocyte vigilance, they can invade cells of other tissues with a greater chance of surviving and multiplying. Although in many cases, intracellular parasite multiplication finally leads to host cell destruction, the real goal of the pathogen is not to kill the host cell but to use its machinery to grow. This situation is similar to that of prokaryotic cells living in association with Protozoa, ameboid cells, algae, aquatic invertebrates, and insects. In several cases this cohabitation corresponds to a real symbosis in which the parasite confers specific functions to the host cell. This phenomenon is observed in phagocytic cells, for example, in a strain of Amoeba proreus that survives only when it is colonized by bacteria (Jeon and Jeon, 1976). The latter multiply inside phagosomes that do not fuse with lysosomes. It is interesting to note that the behavior of invading parasites varies with the host cell. As an example, when Chlorella (a Cyanophycea) are phagocytosed by the A. proteus strain cited above, phagosome-lysosome fusions occur and they are digested; when they are ingested by Paramecium or Hydra, phagosomes do not fuse with lysosomes and these organisms become symbionts (Karakashia and Karakashia, 1973; Karakashia and Rudzinska, 1981; Muscatine er al., 1975). These two examples suggest that symbiont-host cell interplay probably depends upon resistance or inhibition mechanisms similar to those observed in pathogen-professional phagocyte interaction. One of the most studied symbioses is that of Rhizobium which confers its nitrogen fixation properties to the host plant cell. This situation is, however, different from the previous ones, because plant cells do not display a phagocytic activity and bacterial penetration corresponds to a complex process (Dart, 1975). In any case, and whatever the cell type, parasitism is of special interest because it could represent a preliminary stage of intracellular adaptation that would have led to the total state of symbiosis displayed by chloroplasts and mitochondria, according to a very attractive hypothesis (Stanier, 1970). REFERENCES Aikawa, M . , Miller, L. H . , Johnson, J . , and Rabbege, J. (1978). J . Cell B i d . 77, 72-82 Alcantara, A . , and Brener, Z. (1980). Exp. Parusirol. 50, 1-6.



Alexander, J. (1981). Exp. Parasirol. 52, 261-270. Alexander, J., and Vickerman, K. (1975). J . Prororool. 22, 502-508. Allen, R. C. (1979).In “Lysosomes in Biology and Pathology” (J. T. Dingle, J. Jacques, and I. H. Shaw, eds.), Vol. 6, pp. 197-233. North-Holland Publ., Amsterdam. Alouf, J. E. (1976). In “The Specificity and Action of Animal, Bacterial and Plant Toxins” (P. Cuatrecasas, ed.), Series B, Vol I, pp. 221-270. Chapman & Hall, London. Alouf, J. E. (1980). Pharmacol. Ther. 11, 661-717. Amherdt, M., Baggiolini, M., Perrelet, A., and Orci, L. (1978). Lab. Invest. 39, 398-404. Arend, W. P., and Mannik, M. (1972). J. Exp. Med. 136, 514-531. Armstrong, J. A., and d’Arcy Hart, P. D. (1975). J. Exp. Med. 142, 1-16. Avila, J. L. (1979). In “Lysosomes in Biology and Pathology” (J. T. Dingle, J. Jacques, and 1. H. Shaw, eds.), Vol. 6, pp. 235-266. North-Holland Publ., Amsterdam. Badwey. J. A., and Karnovsky, M. L. (1980). Annu. Rev. Biochem. 49, 695-726. Badwey, J. A., Cumutte, J. T., Robinson, J. M., Lazdins, J. K., Briggs, R. T., Karnovsky, M. J., and Karnovsky, M. L. (1980). J. Cell. Physiol. 105, 541-551. Baker, J. R., and Liston, A. J. (1978). J . Gen. Microbiol. 104, 79-89. Baldridge, C. W., and Gerard, R. W. (1933). Am. J. Physiol. 103, 235-236. Bar-Shavit, Z., Ofek, I., Goldman, R., Mirelman, D.. and Sharon, N. (1977). Biochem. Biophys. Res. Commun. 78,455-460. Batz, W., and Wunderlich, F. (1976). Arch. Microbiol. 109, 215-220. Beaman, L., and Holmberg, C. A. (1980). Infect. Immun. 28, 594-600. Becker, Y. (1978). Bacreriol. Rev. 42, 274-306. Beckerdite, S., Mooney, C., Weiss, L., Franson, R., and Elsbach, P. (1974). J. Exp. Med. 140, 396-409. Benoliel, A. M., Capo, C., Bongrand, P., Ryter, A,, and Depieds, R. (1980). Immunology 41, 547-560. Berlin, R. D., and Fera, J. P. (1977). Proc. Narl. Acud. Sci. U.S.A. 74, 1072-1076. Bemheimer. A. W. (1974). Biochim. Biophys. Acra 344, 27-50. Beukers, H., Deierkauf, F. A. Blom. C. P., Deierkauf, M., Scheffers, C. C., and Rienezsma, J. C. (1980). Chem. Biol. Interact. 33, 91-100. Beveridge, T. J. (1980). Can. J . Microbiol. 26, 643-653. Bishop, 0..Orr, T.. and Bartell, P. F. (1982). Infecr. Immun. 37, 378-381. Blumenstock, E., and Jann, K. (1982). Infect. Immun. 35, 264-269. Bowers, B. (1977). Exp. Cell Res. 110, 409-417. Bowers, B. (1980). J . Cell Biol. 84, 246-260. Bowers, B., and Olszewski, T. E. (1972). J . Cell Biol. 53, 681-694. Bowers, B., Olszewski, T. E., and Hyde, J. (1981). J . Cell Biol. 88, 509-515. Braatz-Schade, K., and Stockem, W. (1973). Z. Zellforsch. 137, 97-109. Brandt, P. W., and Freeman, A. R. (1967a). J. Colloid Interface Sci. 25, 47-56. Brandt, P. W., and Freeman, A. R. (1967b). Science 155, 582-585. Briggs, R. T., Drath, D. B., Kamovsky, M. L., and Kamovsky, M. J. (1975a). J. Cell Biol. 67, 566-586. Briggs, R. T., Karnovsky, M. L., and Karnovsky, M. J. (1975b). J. Cell Biol. 64, 254-260. Briles, D. E., Lehmeyer, J., and Forman, C. (1981). Infect. Immun. 33, 380-388. Britton, L., Malinowski, D. P., and Fridovich, I. (1978). J. Bucreriol. 134, 229-236. Brown, C. A., and Draper, P. (1970). J . Parhol. 102, 21-26. Brown, I. N., and Draper, P. (1976). Infecr. Immun. 13, 1199-1204. Bruyn, de, P. P. H., Michelson, S., and Becker, R. P. (1978). J . Cell Biol. 78, 379-389. Burgert, H. G., and Thilo, L. (1983). Exp. Cell Res. 144, 127-142. Byme. G. I., and Moulder, J. W. (1978). Infect. Immun. 19, 598-606.


32 1

Capo, C., Bongrand, P., Benoliel, A. M., Ryter, A,, and Depieds, R. (1981). Ann. Immunol. 132 D, 165-173. Carrol, M. E. W., Jackett, P. S., Aber, V. R.. and Lowrie, D. B. (1979). J. Gen. Microbiol. 110, 42 1-429. Chang, K. G. (1980). Science 209, 1240-1242. Chang, K. P., and Dwyer, D. M. (1976). Science 193, 678-680. Chapman-Andersen, C. (1977). Physiol. Rev. 57, 371-385. Chapman, W. E., and Ward, P. A. (1977). Science 196, 67-70. Chastellier, de, C., and Ryter, A. (1981). Biol. Cell. 40, 109-118. Chastellier, de, C., and Ryter, A. (1982). Biol. Cell. 43, 121-128. Chastellier, de, C., Ryter, A., and Thilo, L. (1983). Eur. J . Cell Biol. 30, 233-243. Cohn, Z. A. (1978). J. Immunol. 121, 813-816. Confer, D. L., and Eaton, J. W. (1982). Science 217, 948-950. Costerton, J. W., Irvin, T., and Cheng, K. 1. (1981). Annu. Rev. Microbiol. 35, 299-324. Daems, W. T., Roos, D., van Berkel, Th. J. C., and van der Rhee, H. J. (1979). I n “Lysosomes in Biology and Pathology” (J. T. Dingle, P. J. Jacques, and I. H. Shaw, eds.). Vol. 6, pp. 463-514. North-Holland Publ., Amsterdam. Dart, P. J. (1975). In “The Development and Functions of Roots” (J. G. Torrey and D. T. Clarkson, eds.), pp. 467-506. Academic Press, New York. Davier, P. F., Rennke, H. G., and Cotran, R. S. (1981). J. Cell Sci. 49, 69-86. Davis-Scibienski, C., and Beaman, B. L. (1980). Infecr. Immun.28, 610-619. Deierkauf, F. A., Beukers, H., Deierkauf, M., and Riemzsma, J. C. (1977). J. Cell. Physiol. 92, 169- 175. Dossett, J. H., Kronvall, G., Williams, R. G., Jr., and Quie, P. G. (1969). J . Immunol. 103, 1405- 1410. Draper, P. (1981). In “Microbial Perturbation of Host Defense” (F. O’Grady and H. Smith, eds.), pp. 143-157. Academic Press, New York. Draper, P., and Rees, R. J. W. (1973). J. Gen. Microbiol. 77, 79-87. Duncan, R., and Pratten, M. K. (1977). J. Theor. Biol. 66, 727-735. Edelson, P. J., and Cohn, Z. A. (1974). J. Exp. Med. 140, 1364-1386. Ehlenberger, A. G., and Nussenzweig, V. (1977). J . Exp. Med. 145, 357-371. Eissenberg, L. G., and Wyrick, P. B. (1981). Infecr. Immun.32, 889-896. Elsbach, P., Weiss, J., Franson, R. C., Beckerdite-Quagliata, S., Schneider, A., and Harris, L. (1979). J. Biol. Chem. 254, 11OOO-11009. Erving, E. P., Takeuchi, A., Shirai, A., and Osterman, J. V. (1978). Infecr. Immun. 19, 1068- 1075. Farquhar, M. G. (1978). J . Cell Biol. 77, R35-R42. Fauve, R., and Hevin. B. (1980). In “The Host Invader Interplay” (H. van Bossche, ed.), pp. 285-292. Janssen Research Foundation. Elsevier, Amsterdam. Favard-Skkno, C., Ludosky, M. A., and Ryter, A. (1981). J . Cell Sci. 51, 63-84. Fox, E. N. (1976). Eacreriol. Rev. 38, 57-86. Frkhel, C., Ryter, A,, Rastogi, N., and David, H. L. (1983). In preparation. Fridovich, 1. (1975). Annu. Rev. Biochem. 44, 147-159. Furth, van, P., Hirsch. J. G.. and Fedorko, M. E. (1970). J . Exp. Med. 132, 794-812. Geisow, M. I., Beaven, G. H., d’Arcy Hart,P., and Young, M. R. (1980). Exp. Cell Res. 126, 159-165. Geisow, M. J., d’Arcy Hart,P., and Young, M. R. (1981). J . CellEiol. 89, 645-652. Genchault, J. L., Steffan, A. M., Bingen, A., and Kim, A. (1981). Virology 112, 375-384. Gingell, D., and Vince, S . (1982). J. Cell Sci. 54, 255-285. Githens, S., 111, and Karnovsky, M. L. (1973). J. Cell Biol. 58, 536-548.



Glass, E., Stewart, J., and Weir, D. M. (1981). Immunology 44, 529-534. Glynn, P. J. (1981). Cyrobios 30, 153-166. Goldman, R.. and Raz, A. (1975). Exp. Cell Res. 96, 393-405. Goldman, R., and Rottenberg, H. (1973). FEES Len. 33, 233-238. Gordon, A. H., d’Arcy Hart, P., and Young, M. R. (1980). Narure (London) 286, 79-80. Goren, M. B. (1977). Annu. Rev. Microbiol. 31, 507-533. Goren, M. B., Brokl, 0.. and Schaefer, W . B. (1974). Infecr. Immun. 9, 142-149. Goren, M. B., d’Arcy Hart, P., Young, M. R.,and Armstrong, J. A. (1976). Proc. Narl. Acad. Sci. U.S.A. 73, 2510-2514. Goren, M. B., Swendsen, C. L., and Henson, J. (1980). I n “Mononuclear Phagocytes: Functional Aspects” (R. van Furth, ed.), pp. 999-1038. Nijhoff, The Hague. Griffin, F., Jr. (1982). In “Advances in Host Defense Mechanisms” (J. F. Gallin and A. S. Fauci, eds.), Vol. I, pp. 31-55. Raven, New York. Griffin, F. M., Jr., and Silverstein, S. C. (1974). J. Exp. Med. 139, 323-336. Griffin, F. M., Jr., Bianco, C., and Silverstein, S. C. (1975a). J . Exp. Med. 141, 1269-1277. Griffin, F. M., Jr., Griffin, J . A., bider, J. E., and Silverstein, S. C. (1975b). J. Exp. Med. 142, 1263- 1282.

Griffin, F. M., Jr., Griffin, J. A., and Silverstein, S. C. (1976). J . Exp. Med. 144, 788-4309, Hard, G. C. (1973) Infecr. Immun. 12, 1439-1449. Hart, d’Arcy, P.,and Young, M. R. (1975). Narure (London) 256, 47-49. Hart, d’Arcy, P., and Young, M. R. (1978). Exp. Cell Res. 114, 486-490. Hart, d’Arcy, P., and Young, M. R. (1979). Exp. CellRes. 118, 365-375. Hart, d Arcy, P., Armstrong, J. A., Brown, C. A,, and Draper, P. (1972). Infecr. Immun. 5, 803-807.

Hartwig, 1. H., Yin, H. L., and Stossel, T. P. (1980). In “Mononuclear Phagocytes: Functional Aspects” (R.van Furth, ed.), Part 11, pp. 971-996. Nijhoff, The Hague. Hed, J. (1981). Monogr. Allergy 17, 92-111. Hed, J., and Stendahl, 0. (1982). Immunology 45, 727-736. Heiple, J. M., and Taylor, D. L. (1982). J. Cell Biol. 94, 143-149. Helenius, A., Kartenbeck, J., Simons, K., and Fries, E. (1980). J. Cell Biol. 84, 404-420. Hellio, R., and Ryter, A. (1980). J. Cell Sci. 41, 89-104. Henrichs, P. A. J., van Erne-van der Tol, E., and Verhoef, J. (1982); Anronie Leeuwenhoek 48, 191- 192.

Heyningen, van, S. (1981). Nature (London) 292, 293-294. Higashi, N. (1962). I n “Recent Advances in Studies of Tsutsugamushi Disease in Japan” (T. Tamija, ed.), pp. 145-156. Medical Culture. Hirsch, 1. G. (1956). J . Exp. Med. 103, 589-611. Hirsch, J. G. (1962). J . Exp. Med. 116, 827-834. Hirsch, J . G., Fedorko, M. E., and Cohn, Z. A. (1968). J . Cell Biol. 38, 629-632. Holter, H. (1959). Inr. Rev. Cyrol. 8, 481-504. Horvath, L. I., Bagyinka, C. S., and Sandor, M. (1981). Period. Biol. 83, 143-145. Horwitz, M. A. (1982). Rev. Infect .Dis. 4, 104-123. Howard, F. D., Petty, H. R., and Mc Connell, H. M. (1982). J. Cell Biol. 92, 283-288. Hubbard, A. L. (1978). In “Transport of Macromolecules in Cellular Systems,” pp. 363-398. Dahlem-Konferenzen. Abakon, Berlin. Hubbard, A. L., and Cohn, Z. A. (1975). J . Cell Biol. 64, 461-479. Ito, T., Ueda, M. J., Okada, I. S., and Ohnishi, S. I. (1981). J . Cell Sci. 51, 189-201. Iwata, K., and Uchida, K. (1980). I n “The Host Invader Interplay” (H. van Bossche, ed.), pp. 319-325. lanssen Research Foundation. Elsevier, Amsterdam. Jackett, P. S., Aber, V. R., and Lowrie, D. B. (1978). J. Gen. Microbiol. 104, 37-45.



Jacques, P. J. (1969). In “Lysosomes in Biology and Pathology” (J. T. Dingle and H. B. Fell, eds.), Vol 2, pp. 395-420. North-Holland Publ., Amsterdam. Jacques, Y. V., and Bainton, D. F. (1978). Lab. Invest. 39, 179-185. Jeon, K. W., and Jeon, M. S. (1976). J . Cell. Physiol. 89, 337-344. Johnston, R. B., Godzik, C. A., and Cohn, 2 . A. (1978). J. Exp. Med. 148, 115-127. Jones, T. C.. and Hirsch, J. G . (1971). J . Exp. Med. 133, 231-259. Jones, T. C.. and Hirsch, J. G. (1972). J. Exp. Med. 136, 1173-1194. Jones, J. M., and Schwab. J. H. (1970). Infect. Immun. 1, 232-242. Kaplan, G. (1977). Scund. J. Immunol. 6, 797-807. Karakashian, M. W., and Karakashian, S. J. (1973). Exp. Cell Res. 81, 1 1 1-119. Karakashian, S. J., and Rudzinska, M. A. (1981). Exp. Cell Res. 131, 387-393. Kamovsky, M. L., and Lazdins, J. K. (1978). J . Immunol. 121, 809-813. Kamovsky, M. L., and Wallach, P. F. H. (1961). J. Biol. Chem. 236, 1895-1901. Karnovsky, M. J., Robinson, J. M., Briggs, R. T . , and Karnovsky, M. L. (1981).Histochem. J. 13, 1-22. Katz, S. M., and Hashemi, S. (1982). Lab. Invest. 46, 24-38. Kielian, M. C., and Cohn, Z. A. (1982). J . Cell Biol. 93, 875-882. Kielian, M. C.. Steinman, R. M., and Cohn, 2 . A. (1982). J . Cell Biol. 93, 866-874. King, G. J., and Swanson, J. (1978). Inject. Immun. 21, 575-584. Klebanoff, S. J. (1982). In “Advances in Host Defense Mechanisms” (J. 1. Gallin and A. S. Fauci, eds.), Vol. I, pp. 1 1 I- 162. Raven, New York. Klebanoff, S. J., and Clark, R. A. (1978). “The Neutrophil: Function and Clinical Disorders.” North-Holland h b l . , Amsterdam. Kooistra, T., Duursma, A. M., Bouma, J . H. W.. and Gurber, M. (1980). Biochim. Biophys. Acru 631, 439-450. Kress, Y., Tanowitz, H., Bloom, B., and Wittner, M. (1977). Exp. Parasitol. 41, 385-396. Leijh, P. C. J., van Zwet, Th.L., and van Furth, R. (1980). Infect. Immun. 30,421-426. Levy, N. J . (1979). Infect. Immun. 25, 946-953. Lew, P. D., and Stossel, T. P. (1980). J. Clin. Invest. 67, 1-9. Lewis, D. H., and Peters, W. (1977). Ann. Trop. Med. Parusitol. 71, 295-312. Liston, A. J . , and Baker, J. R. (1978). J. Gen. Microbiol. 107, 253-262. Locksley, R., Wilson, C. B., and Klebanoff, S. J. (1982). J. Clin. Invest. 69, 1099- I I 1 I . Lowrie, D. B., Jackett, P. S . , and Ratcliff. N. A. (1975). Nature (London) 254, 600-602. MacRae, E. K., Pryzwansky, K. B., Cooney, M. H., and Spitznagel, J. K. (1980). Cell Tissue Res. 209, 65-70. McNeil, P. L. (1981). J. CellSci. 49, 311-340. McNeil, P. L., Hohman, T. C., and Muscatine, L. (1981). J. Cell Sci. 52, 243-269. Mandell, G. L. (1970). Proc. Soc. Exp. Eiol. Med. 134, 447-449. Manjula. B. W., and Fischetti, V. A. (1980). J . Exp. Med. 151, 695-708. Mantovani, B. Rabinovitch, M., and Nussenzweig, V. (1972). J . Exp. Med. 135, 780-792. Mauel. J. (1980). In “The Host Invader Interplay” (H. van Bossche. ed.), pp. 165-177. Janssen Research Foundation. Elsevier, Amsterdam. Michl, J., Ohlbaum, D. J., and Silverstein, S. C. (1976). J. Exp. Med. 144, 1465-1483. Middlebrook, G., Coleman, C. M., and Schaefer. W. B. (1959). Proc. Nut/. Acad. Sci. U.S.A. 45, 1801- 1804. Milder, R., and Kloetzel, J. (1980). Parasitology 80, 139-145. Miles, P. R., Bowman, L., and Costranova, V. (1981). J . Cell. Phvsiol. 106, 109-117. Mills, E. L., Debets-Ossekopp, Y., Verbrugh. H. A , , and Verhoef, J. (1981). Infect. Immun. 32, 1200- 1205. Mitchison, D. A., Selkon, J. B., and Llyod, J. (1963). J. Pathol. Bacteriol. 86, 377-386.



Mudd, S., McCutcheon, M., and Lucke, B. (1934). Physiol. Rev. 14, 210-275. Muller. W. A., Steinman, R. M., and Cohn, Z. A. (1980). J . Cell Biol. 86, 304-314. Muller, W. A., Steinman, R. M., and Cohn, Z. A. (1983). J. Cell Biol. %, 29-36. Munthe-Kaas. A. C., Kaplan, G., and Seljelid, R. (1976). Exp. Cell Res. 103, 201-212. Murray, H. W . , and Cohn, Z. A. (1979). J. Exp. Med. 150, 938-949. Muscatine, L., Cook, C. B., Pardy, R. L., and Pool, R. R. (1975). Symp. SOC.Exp. Biol. Med. 29, 175-203. Muschel, R. J., Rosen, N., and Bloom, B. R. (1977). J . Exp. Med. 145, 175-186. Muscher, D. M., Verbrugh, H. A., and Verhoef, J. (1981). J . Immunol. 127, 84-88. Nakano. M., and Saito, K. (1970). Jpn. J . Microbiol. 14, 451-461. Nathan, C. F., andRoot, R. K. (1977). J . Exp. Med. 146, 1648-1662. Nathan, C. E., Nogueira, N., Juangbhanich, C., Ellis, J., and Cohn, Z. A. (1979). J . Exp. Med. 149, 1056-1068. Nguyen, 9 . Y.,Peterson, P. K., Verbrugh, H. A., Quie, P. G . , and Hoidal, J. R. (1982). Infect. Immun. 36, 504-509. Nichols, B. A. (1982). Lab. Invest. 47, 235-246. Nichols, 9. A.. Bainton, D. F., and Farquhar, M. G. (1971). J. Cell Biol. 50, 498-515. Nogueira, N., and Cohn, Z. (1976). J . Exp. Med. 143, 1402-1420. Nonoyama, S., Kojo, H., Mine, Y., Nishida, M., Goto, S., and Kuwahara, S. (1979). Infect. Immun. 24, 399-403. North, R. J. (1970). Semin. Hemarol. 7, 162-171. North, R. J. (1978). J. Immunol. 121, 806-816. Oates, P. J., and Touster, 0. (1976). J . Cell Biol. 68, 319-338. Oates, P. J., and Touster, 0. (1978). J. Cell Biol. 79, 217-234. Oates, P. J., and Touster, 0. (1980).J . Cell Biol. 85, 804-810. Oberti, J., Caravano, R., and Roux, J. (1981). Ann. Insr. Pasreur Immunol. 132 D, 201-206. Odeberg. H.,and Olsson, I. (1975). J. Clin. Invest. 56, I 1 18-1 124. Ofek, I., Bergner-Rabinowitz, S., and Ginsberg, I. (1972). Infect. Immun. 6, 459-464. Orci, L., Perrelet, A., and Friend, D. C. (1977). J. Cell Biol. 75, 23-30. Oschman, J. L., and Wall, B. J. (1972). J . Cell Biol. 55, 58-73. Oss, van, C. J. (1978). Annu. Rev. Microbiol. 32, 19-39. Papahadjopoulos. W . , Vail, J., Newton, C., Nir, S., Jacobson, K.,Poste, G., and Lazo, R. (1977). Biochim. Biophys. Acra 465, 579-598. Pappenheimer, A. M. (1977). Annu. Rev. Biochem. 46, 69-94. Patterson-Delafield, J., Martinez, R. J., and Lehrer, R. I. (1980). Infect. Immun. 30, 180-192. Pearce, W. A., and Buchanan, T. M. (1980). I n “Bacterial Adherence” (E. H. Beachey, ed.), Series B, Vol. 6, pp. 291-340. Chapman & Hall, London. Pesanti, E. I. (1978). Infect. Immun. 20, 503-511. Pistel, T. G. (1981). Annu. Rev. Microbiol. 35, 85-1 12. Poste, G., and Allison, A. C. (1973). Biochim. Biophys. Acra 300, 421-465. Pruzanski, W., and Saito, S. (1978). Exp. CeNRes. 117, 1-13. Pryzwansky, K. B., MacRae, E. K.,Spitznagel, J. K., and Cooney, M. H. (1979). Cell 18, 1025- 1032. Quie, P. G. (1972). In “Phagocyte Mechanisms in Health and Disease” (R. C. William and H. D. Fudenberg, eds.), pp. 139-150. Intercontinental Medical, New York. Rabinovitch, M. (1967a). Exp. Cell Res. 46, 19-28. Rabinovitch, M. (1967b). Proc. SOC. Exp. Biol. Med. 124, 396-399. Rabinovitch, M., Dedet, J.-P., Ryter, A., Robineaux, R., Topper, G., and Brunel, E. (1982). J. Exp. Med. 155, 415-431. Reid, K. 9 . M., and Porter, R. R. (1981). Annu. Rev. Biochem. 50, 433-464.



Reijngoud, D. J. (1978). Trends Biochem. Sci. 3, 178-180. Reijngoud, D. J., and Tager, J. M. (1977). Biochim. Biophys. Acta 472, 419-449. Rikihisa, Y ., and Ito, S. (1979). J. Exp. Med. 150, 703-708. Roberts, R. B. (1979). In “Principles and Practice of Infectious Diseases” (G. L. Mandell, R. G. Douglas, and J. E. Bennett, eds.), pp. 1589-1616. Wiley, New York. Root, R. K., and Stossel, T. P. (1974). J. Clin. Invest. 53, 1207-1215. Ryter, A., and de Chastellier, C. (1977). J . Cell Biol. 75, 200-217. Ryter, A., and Hellio, R. (1980). J. CeN Sci. 41, 75-88. Ryter, A., Dedet, J.-P., and Rabinovitch, M. (1983a). Exp. Parasirol. 55, 233-242. Ryter, A., Lepoivre, M., and Petit, I. F. (1983b). In preparation. Saeki, K . , Endo, K., Tasaka, K., and Yamasuki, H. (1974). Jpn. J . Phurmacol. 24, 109-118. Saint-Gullain, M. L., Vray, B., Hoebeke, J., and Leloup, R. (1980). In “Scanning Electron Microscopy,” pp. 205-212. I1 Semin. Inc., AMF O’Hare, Chicago, Illinois. Sandin, R. L., Rogers, A. L., Patterson, R. J., and Benke, E. (1982). Infect. Immun. 35, 79-85. Sandvig, K., and Olsnes, S. (1980). J . Cell Biol. 87, 828-832. Sastry, P. S.,and Hokin, L. E. (1966). J . Biol. Chem. 241, 3354-3361. Schneider, Y. J., Tulkens, P., de Duve, C., and Trouet, A. (1979a). J. Cell Biol. 82, 449-465. Schneider, Y. J., Tulkens, P., de Duve, C., and Trouet, A. (1979b). J . Cell Biol. 82, 466-474. Schwab, J. H. (1975). Bacreriol. Rev. 39, 121-143. Schwarz. H., and Thilo, L. (1983). Eur. 1. Cell Biol. (in press). Schwarzmann, S., and Boring, J. R., 111. (1971). Infect Immun. 3, 762-767. Segal, A. W.. Geisow, M., Garcia, R., Harper, A., and Miller, R. (1981). Nature (London) 290, 406-409. Segerling, M., Hahn, H., and Opperkuch, W. (1982). Immunobiology 162, 39-45. Seglen, P. O., and Ruth, A. (1976). Exp. Cell Res. 100, 276-280. Shaw, D. R., and Griffin, F. M. (1981). Nature (London) 289, 409-41 I . Sherbert, G. V. (1978). In “The Biophysical Characterization of the Cell Surface,” pp. 36-145. Academic Press, New York. Silveblatt, F. J., Dreyer, J., and Schauer, S. (1979). Infect Immun. 24, 218-223. Silverman, D. I., and Wisseman, C. L. (1979). Infect. Immkn. 26, 714-727. Silverstein, S. C., Steinman, R. M., and Cohn, 2. A. (1977). Annu. Rev. Biochem. 46, 669-722. Simmons, S. R.,and Kamovsky, M. L. (1973). J . Exp. Med. 138, 44-63. Simon, G. L., Klempner, M. S., Kasper, D. L., and Gorbach, S. L. (1982). J. Infecr. Dis. 145, 72-77. Smith, H. (1977). Bacreriol. Rev. 41, 475-500. Spitznagel, J. K., and Chi, H. C. (1963). Am. J . Parhol. 43, 697-708. Stahl, P. D., Rodman, J. S., Miller, M., and Schlessinger, P. H. (1978). Proc. Narl. Acad. Sci. U.S.A. 75, 1399-1403. Stanier, R. Y. (1970). Symp. SOC. Gen. Microbiol.. 20th pp. 1-38. Steinman, R. M., Brodie, S. E., and Cohn, Z. A. (1976). J. Cell Biol. 68, 665-687. Steinman, R. M., Mellman, 1. S., Muller, W. A., and Cohn, Z. A. (1983). J . Cell Biol. 96, 1-28. Stendahl, 0.. Tagesson, C., and Edebo, M. (1973). Infect. Irnmun. 8, 36-41. Stendahl, 0. I . , Hartwig, J. H., Brotschi, E. A,, and Stossel, T. P. (1980). J . Cell Biol. 84, 215-224. Stendahl, 0. I., Dahlgren, C., Edebo, M., and Ohman, L. (1981). Monogr. Aflergy 17, 12-27. Stinson, M. W., and van Oss, C. J. (1971). J. Rericuloendothel. SOC. 9, 503-512. Stjernstrom, I . , Magnusson, K. E., Stendahl, 0..and Tagesson, C. (1977). Infect. Immun. 18, 26 1-265. Stockem, W. (1970). Mikroskopie 24, 332-344. Stockem, W. (1973). Z. Zellforsch. 136, 433-446.



Stockem, W. (1977). In “Mammalian Cell Membranes” (G. A. Jamieson and D. M. Robinson, eds.), Vol 5, pp. 151-195. Butterworths, London. Storrie, B. A. (1979). J . Cell Biol. 83, LP1312. Stome, B. A., Dreesen, T. D., and Maurey, K. M. (1981). Mol. Cell. Biol. 1, 261-268. %OR, I . , and Spears, P. (1977). Curr. Top. Microbiol. Immunol. 76, 167-214. Stossel, T. P. (1975). Semin. Hemarol. 12, 83-116. Stossel, T. P. (1976). J. Rericuloendorhel. SOC. 19, 237-245. Stossel, T. P., and Hartwig, J. H. (1976). J . Cell Biol. 68, 602-619. Sung, S. S . I.. Nelson, R. S . , and Silverstein, S . C. (1983). J. Cell Biol. W, 160-166. Taylor, D. L., and Condeelis, I. (1979). Inr. Rev. Cyrol. 56, 57-142. Taylor, D. L., Blinks, J. R., and Reynolds, G. (1980). J. Cell Biol. 86, 599-607. Thilo, L., and Vogel, G. (1980). Proc. Natl. Acad. Sci. U.S.A. 77, 1015-1019. Thomas, D. W., Hill, J. C., and Tyeryar, F. J. (1972). Infect. Immun. 8, 98-104. Thong, Y. H. (1982). Med. Hypotheses 8, 249-254. Thorne, K. I. I., Svvennsen, R. J., and Franks, D. (1978). Infecr. Immun. 21, 798-805. Thorne, K. J. I., Glauert, A. M., Svvennsen, R. J., and Franks, D. (1979). Parasitology 79, 367-379. Thome, K. J. I., Oliver, R. C., and Glauert, A. M. (1980). J. Cell Sci. 42, 367-378. Thyberg, J. (1980). Eur. J. Cell Biol. 23, 95-103. Thyberg, J., Nilsson, J., and Hellgren, D. (1980). Eur. J. Cell Biol. 23, 85-94. Todd, W. J., and Storz, J. (1975). Infect. Immun. 12, 638-646. Tsan, M. F., and McIntyre, P. A. (1976). J . Exp. Med. 143, 1308-1316. Tulkens, P., Schneider, Y. J., and Trouet, A. (1980). In “Mononuclear Phagocytes: Functional Aspects” (R. van Furth. ed.), Part I, pp. 613-648. Nijhoff, The Hague. Unkeless, J. C., Fleit, H., and Mellman, I. S . (1981). Adv. Immunol. 31, 247-270. Utsumi, S., Sonoda, S., Imagawa, T., and Kanoh, M. (1978). Biken J . 21, 121- 135. Valerius, N. H., Stendahl, 0.. Hartwig, J. H., and Stossel, T. P. (1981). Cell 24, 195-202. Vilkas, E., and Markavits, J. (1972). C. R. Acad. Sci. Paris 275 C , 913-915. Vogel, G., Thilo, L., Schwarz, H., and Steinhart, R. (1980). J . Cell Biol. 86, 456-465. Volsky, D. J., and Loyter, A. (1978). Biochim. Biophys. Acra 514, 213-224. Warr, G. A. (1980). Biochem. Biophys. Res. Commun. 93, 737-745. Weir, D. M., and Ogmundsdottir, H. (1977). Clin. Exp. Immunol. 30, 323-329. Weisman, R. A., and Korn, E. D. (1967). Biochemistry 6, 485-497. Weiss, J., Franson, R. C., Schmeidler, K., and Elsbach, P. (1976). Biochim. Biophys. Acra 436, 154- 169. Weiss, J., Elsbach, P., Olsson, I . , and Odeberg, H. (1978). J . Biol. Chem. 253, 2664-2672. Weiss, J., Victor, M.. Stendahl, O., and Elsbach, P. (1982). J. Clin. Invest. 69, 959-970. Weiss, L. (1969). Inr. Rev. Cyrol. 26, 63-105. Weiss, S. I., and Lobuglio, A. T. (1982). Lab. Invest. 47, 5-18. Weissmann, G., Keiser, H., and Bernheimer, A. W. (1963). J . Exp. Med. 118, 205-222. Werb, Z., and Cohn, Z. A. (1972). J. Biol. Chem. 247, 2439-2446. Westwood, F. R., and Longstaff. E. (1976). Br. J . Cancer 33, 392-399. Wilkinson, B. J., Peterson, P. K.,and Quie, P. G.(1979). Infect. Immun. 23, 502-508. Wilschert, J., Diizgiines, N., and Papahadjopoulos, D. (1981). Biochemistry 20, 3126-3133. Wilson, C. B., van Tsai, V., and Remington, I. S. (1980). J . Exp. Med. 151, 328-334. Win, K., Veale, D. R., and Smith, H. (1976). J. Med. Microbiol. 9, 1-12. Wyrick, P. B., and Brownridge, E. A. (1978). Infecr. Immun. 19, 1054-1060. Yong, E. C., Klebanoff, S. J., and Kuo, C. C. (1982). Infecr. Immun. 37, 422-426. Young, I. D. E., Young, T. M., Lu, L. P., Unkeless, J. C., and Cohn,Z. A. (1982). J . Exp. Med. 156, 1677-1690.

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS Young, M . R . , d'Arcy Hart, P., and Geisow, M. J . (1981). Exp. Cell Res. 135, 442-445. Zeya, H. I . , and Spitznagel, J . K . (1966). J . Bacterial. 91, 755-762. Zeya, H. I . , and Spitznagel, J. K . (1968). J . Exp. Med. 127, 927-941. Zucker-Franklin, D., and Hirsch, J . G. (1964). J . Exp. Med. 120, 569-576.