BASIC REVIEW
How Do Phagocytes Eat? THOMAS P. STOSSEL,
M.D.; Boston,
Massachusetts
A New York University Honors Program Lecture
Phagocytosis is a cellular function relevant for host defense against infection, tissue turnover, and other aspects of human physiology. Phagocytosis is also representative of functions wherein external stimuli activate motile events in the cell. Recognition of suitable objects by the plasma membrane of the phagocyte initiates phagocytosis. Knowledge of serum proteins that coat objects rendering them recognizable is considerable, but understanding of the chemical basis of recognition is meager. The signals activated by recognition are also not known. The work of phagocytosis that causes pseudopodia to enclose objects in vacuoles is ascribable to metabolic energy-dependent interactions between actin filaments and other contractile proteins in the peripheral cytoplasm. These interactions may also regulate the fusion of lysosomes with phagocytic vacuoles, an event important for the processing of ingested objects after phagocytosis.
PHAGOCYTOSIS is the process whereby cells internalize objects that are approximately micron-sized. Before addressing the question "how?'* let us review briefly some answers to the question, "Why is phagocytosis important?" The most familiar answers are that protists feed with it and that wandering and fixed hematopoietic cells, the polymorphonuclear leukocytes and mononuclear phagocytes, sequester, kill, or contain microorganisms that would otherwise devour higher organisms. These phagocytes also perform an important sanitation function in cleaning up damaged or aged host cells and other debris (1). Phagocytosis is also relevant for other functions. The thyroid follicle cell ingests stored thyroglobulin and then degrades it to form thyroid hormones (2). Retinal pigment epithelial cells interiorize and degrade outer rod segments to maintain the turnover of pigment vital for normal vision (3). Phagocytosis is part of the mechanism by which melanocytes transfer pigment to epidermal cells (4). Possibly the evolution of animal cells was influenced by development of a capacity to ingest microorganisms, which they subsequently captivated as their own subcellular organelles (5). The act of phagocytosis is visible under the light microscope, and the most naive observer can see what needs to be known to "explain" phagocytosis (Figure 1). The • From the Medical Oncology Unit, Massachusetts General Hospital; and the Department of Medicine, Harvard Medical School: Boston, Massachusetts. 398
Annals of Internal Medicine 89:398-402, 1978
Downloaded from https://annals.org by Tulane University user on 01/18/2019
observer notes that phagocytosis, like life, can be divided into seven stages. First, something about the surface of some objects and not others elicits recognition by the phagocytes. Second, the message, "Eat me" must be received by the cell. Third, the receipt must somehow signal the cell to do something about it, that something being the activation of effector mechanisms that are: four, an increase in stickiness of the membrane where the object has contacted the cell; five, assembly; six, movement and seven, fusion of pseudopods that surround and engulf totally or in part the recognized object, enclosing it in a phagocytic vacuole. A more sophisticated description of phagocytosis obtainable by observation with phase contrast, transmission, and scanning electron microscopy shows that the pseudopodia are pleats of peripheral cytoplasm (6, 7). This cortical material excludes the nucleus and other organelles, and therefore appears clear in phase contrast. The organelle exclusion ceases at the base of newly forming phagocytic vacuoles. At this region, lysosomes gain access to the plasma membrane and fuse with it, a process called degranulation (8). Degranulation results in the translocation of hydrolytic enzymes to the phagocytic vacuole, which is very important for postphagocytic digestive events (9). During phagocytosis some fraction of the lysosomal enzymes leaks out into the extracellular fluid (10). This lecture concerns recent studies pertaining to the mechanism of pseudopod assembly and movement, but will for completeness contain comments on the other steps involved in phagocytosis. Recognition
How do phagocytes respond to some objects and not others? The answer could theoretically emerge from an exhaustive study of surfaces that do not promote recognition, such as those of many native animal cells and of encapsulated microbes or of certain related artificial particles. These bland objects tend to be sugar coated like the phagocytes themselves, namely, they are encased in a mantle of slimy polysaccharide. Conversely, the answer could come out of an analysis of surfaces that do evoke congestion by phagocytes, for example, damaged animal cells or cells coated with phagocytosis-promoting serum proteins called "opsonins." Presently, IgG and a fragment of the third component of complement are the best defined opsonins, but there are probably others (11). Unfortunately, no simple common denominator currently ©1978 American College of Physicians
explains the different responses of phagocytes to varied surfaces. The difficulties in finding such a theme is that it is now impossible to map the precise location and configuration of recognition determinants on surfaces in anything but a gross way. The fact that one can measure the overall physical or chemical composition of a surface tells little about what may occur in the domain of the object that is interacting critically with the phagocyte membrane (Figure 2). Reception
Part of the answer to the question of phagocytic discrimination might arise from study of the external surface of phagocytes. Such an exercise could identify externally disposed molecules that bind critical surface determinants of objects. In fact, we currently think precisely in these terms of phagocytic "receptors." The receptors are actually defined by some of the rather specific effectors that activate phagocytosis (Table 1). Immunology tends to use the term "receptor" rather loosely ("Use can change the stamp of nature"), and it must be emphasized that there is little molecular definition of these phagocytic receptors. Clearly, the mere binding of an object to the surface of a phagocyte does not indicate a receptor mechanism any more than the adherence of mud to the sole of a shoe defines a mud receptor. On the other hand, the permeation of that same sole by a spike elicits a response. That response theoretically could indicate receptor involvement, but we know that the spike and not the sole confers specificity in this case. Responses evoked in the phagocyte by active recognition include the activation of certain oxidative enzymatic reactions as well as of the effector mechanisms of phagocytosis indicated above. This activation theoretically could as easily result from the capacity of recognition effector entities on objects to bypass surface determinants on the phagocyte and react nonspecifically with the plasma membrane of the phago-
Figure 2. How steric factors might influence and override overall charge effects in activation of phagocytosis. Net negative surface charge on Object 1 prevents access of effector molecule (v) to "trigger" (7) on negatively charged cell surface. However in another case, negatively charged Object 2 orients effector molecule right at trigger because of different location of negatively charged densities on cell. Opposite charges on cell and Object 3 prevent effectortrigger interaction, but in another case Object 4 promotes this interaction.
cyte, as it could from the binding of these recognition effector determinants to specific receptors. These riddles will yield to open-minded scrutiny. Signaling
How does recognition activate phagocytosis? We have no definite candidate for the signal. The main reason for this is that we do not know for sure what the signal wants to do. The identification of the signal mechanism requires a better understanding of the molecular basis of the effector mechanisms. Adhesion
Figure 1 . Visual recording by a "naive observer" of the morphology of phagocytosis. (Rendering by author's daughter, aged 6.)
In some cases, the binding of objects to the surface of phagocytes is nonspecific and cannot be viewed as part of the phagocytic process (11, 12). Nearly all factors (temperature, ions, metabolic poisons, other drugs) that influence the sticking of phagocytes to various objects have identical stimulatory or inhibitory effects on the overall act of phagocytosis (12). Therefore, it is not clear whether any of these factors influences something unique to the act of adhesion per se. Moreover, there is no knowledge of the molecules involved in the sticking process. Stosse/ • Phagocytosis
Downloaded from https://annals.org by Tulane University user on 01/18/2019
399
Table 1. Operationally Defined Phagocytic "Receptors"
Receptor
Characteristics
Fc C3 "Nonspecific"
Not impaired by trypsinization of the cell; not expressed on fibroblasts Impaired by trypsinization of the cell; not expressed on fibroblasts Not impaired by trypsinization of the cell
Pseudopod Assembly and Movement
The internalization of extracellular objects by cells requires mechanical work. Work is needed for many transport processes, but the necessity for such work is intuitively more obvious in the case of phagocytosis. To account for this work, early theories of phagocytosis invoked surface forces. Cells and other objects in suspension have interfaces with their suspending solvents, and the sums of the interfacial energies hypothetically could promote or retard incorporation of objects within cells if the latter were sufficiently elastic (13). It has been impossible to test this hypothesis critically, but experiments purporting to change the surface tension of objects or cells have been both consistent (14, 15) or at variance (16) with it. The definite demonstration that phagocytosis requires active cellular metabolism (17) put physical theories into the background, but still did not explain what work the metabolic energy was being used for. Recent evidence that muscle proteins may do work in nonmuscle cells (18) has revived another old notion: that a contractile skeleton in cells would generate the work for events such as phagocytosis (19). Actin is a major protein of mammalian hematopoietic phagocytes. Actins purified from phagocytes, like muscle and nonmuscle (protoplasmic) actins, exist under physiologic conditions as extended polymers (F-actin) composed of globular monomers (G-actin). Similar actin polymers are identifiable by electron microscopy in the periphery of whole phagocytes and are the principal structures in pseudopodia surrounding objects during phagocytosis (20, 21). These filaments do not lie in any particular orientation but form a rather tangled meshwork. (The "disorganization" of cytoplasmic filaments in mammalian phagocytes contrasts sharply with the crystalline filament assays in skeletal muscles [22] or even of certain phagocytic ciliates [23].) Such a mesh work has structural rigidity by definition, especially if the strands of the meshwork are cross-linked so that there is resistance to externally applied force (24). At least two proteins exist in mammalian phagocytes that cross-link actin, and both are of very high molecular weight. One is myosin, very similar structurally and functionally to myosin of muscle, and the other is called actin-binding protein (25-27). The latter is especially potent as an actincross-linking factor and at low concentrations markedly increases the intrinsic rigidity of F-actin (28). The existence of a rigid cortical gel has long been postulated to explain the organelle exclusion from the peripheral cytoplasm (2). Controlled sequential rigidification of the cell cortex around an object could conceivably
400
September
1978
effect pseudopod extension around that object. The "signal" evoked by recognition might cause actin-binding protein to cross-link actin. But how? Studies with both immunofluorescent (30) and biochemical (31, 32) techniques do suggest that some fraction of actin-binding protein in mammalian phagocytes is concentrated in the cell periphery and is associated with membranes, possibly plasma membranes. Phagocytosis decreases the association between actin-binding protein and the membranes of macrophages, causing more actin-binding protein to be extractable during subsequent disruption of the cells (28). But what phagocytosis actually does to actin-binding protein is unknown. Even if these questions were answerable, the process of phagocytosis is probably not so simple as the consequence of sequential actin gelation. In addition to rigidity forming in the cortex adherent to the object, it would be helpful to bring more cortical material into the area, thereby providing the external force mentioned above. Moreover, an explanation is needed for the attenuation of the cortical gel at the base of the phagocytic vacuole where lysosomes fuse with the phagocytic vacuoles membrane. There is indirect evidence that the protein myosin may be involved in these functions. Myosin, in the presence of magnesium and ATP, compresses gels composed of F-actin and actin-binding protein (27, 28). Both biochemical and immunofluorescent methods show that myosin is concentrated in the cortical cytoplasm of macrophages during pseudopod formation and that myosin moves away from the base of phagocytic vacuoles as they form (32, 33). The active compression of actin filaments by myosin in pseudopodia could have several effects: [1.] More actin-rich cortical matter is brought into the pseudopods; [2.] the rigid actin gel lattice is destroyed at the vacuole's base—it is literally torn apart by myosin pulling from the lateral pseudopods (Figure 3); [3.] the pseudopod is flattened into a pleat; [4.] the controlled disruption of the cortical actin gel by myosin or other means reduces the stability of the plasma membrane. If we envision the plasma membrane as a lipid bilayer supported by an actin cytoskeleton, like cellophane stretched over chicken wire, loss of cytoskeletal support could permit the bilayer to revert to a lower energy state by assuming a spherical
Figure 3. Schematic rendition of phagocytosis by a macrophage in suspension, a. Signal transmitted after recognition activates actinbinding protein and myosin ATPase in organelle-excluding peripheral cytoplasm; b. arrows depict flow of peripheral cytoplasmic gel, which c. drives pseudopodia around object and permits lysosomes access to base of phagocytic vacuole {arrow).
• Annals of Internal Medicine • Volume 89 • Number
Downloaded from https://annals.org by Tulane University user on 01/18/2019
3
shape. Hence actin gel disruption as well as formation could promote the formation of phagocytic vacuoles. This entire picture, even if true, still leaves many questions. For example, what is the precise location of myosin molecules: Are they bound directly to the plasma membrane? Myosin molecules have two domains: a globular head that interacts with actin and a helical tail. In muscle, the tails form the backbone of myosin filaments (34). In nonmuscle cells such as macrophages, myosin filaments are not usually seen, so it is possible that the tails insert into the membrane. Second, what controls the interaction between myosin and actin? This control is well understood for muscle (22), but not for mammalian phagocytes or other nonmuscle cells. In macrophages a protein "cofactor" activates the contraction of actin by myosin (26-28); this cofactor may be a kinase enzyme that acts by phosphorylating the myosin molecule (35). Despite lack of knowledge about details, the basic idea that contractile proteins do the work of phagocytosis is becoming increasingly well substantiated. First, the requirement of ATP for actin filament meshwork contraction by myosin provides a locus for energy phagocytosis. Second, at least two examples are described wherein factors that impair contractile protein structure or function inhibit phagocytosis. Cytochalasin B
Cytochalasin B, a fungal product, produces a dose-dependent (10- 7 — 10-5 M) inhibition of phagocytosis by polymorphonuclear leukocytes (36) and macrophages (37), although binding of objects to the phagocyte membrane is relatively unaltered. Paradoxically the agent causes an increase in secretion of lysosomal enzymes into the extracellular fluid by phagocytes exposed to ingestible objects (10). Recent work indicates that cytochalasin B dissolves gels of actin cross-linked by macrophage actinbinding protein (38). It has no effect on the interaction between actin and myosin. Cytochalasin B is therefore a specific inhibitor of the gel state of actin, and it follows that solvation of the cortical actin gel could prevent pseudopod formation. Cytochalasin B could also allow myosin to compress actin meshworks in the cell periphery with great ease because myosin exerts its force on a less rigid structure than in the untreated actin gel; therefore, the rate of compression increases. This formulation explains why lysosomal granule access to the plasma membrane and subsequent secretion can be enhanced by cytochalasin B. Neutrophil Actin Dysfunction
In 1973, a male infant was encountered who, from birth on, suffered recurrent and severe pyogenic infections caused by low-grade pathogens. The infections involved predominantly the skin and gastrointestinal tracts, responded poorly to antibiotic therapy, and were characterized by absence of pus. The failure of pus formation paradoxically was in the setting of a profound neutrophilic leukocytosis. Leukocytes transfused from normal sera into the infant's circulation appeared unlike the patient's own leukocytes in inflammatory foci. Whereas normal
polymorphonuclear leukocytes, spread on glass slides, crawl about and ingest opsonized objects, the infant's polymorphonuclear leukocytes were virtually unable to perform any of these functions. However, the patient's cells responded to opsonized objects by binding to them, activating oxygen metabolism, and secreting lysosomal enzymes into the extracellular fluid. Therefore, recognition, signaling, and effector mechanisms not involving pseudopod formation were relatively unimpaired. Very little of the G-actin in homogenates of this child's polymorphonuclear leukocytes polymerized into filaments, although the amount of G-actin was normal. Under similar conditions, two thirds to three fourths of G-actin in homogenates of normal polymorphonuclear leukocytes polymerized (39). Because of continued clinical deterioration, the patient was successfully engrafted with marrow from a histocompatible sibling after chemical immunosuppression and total body irradiation. After bone marrow transplantation the patient's polymorphonuclear leukocytes had normal function. Unfortunately, the child died of pulmonary failure, probably secondary to a viral infection (40). The parents and siblings of the patient were not unusually susceptible to infection, and their polymorphonuclear leukocytes had normal function. However, the amount of G-actin that polymerized in homogenates of their polymorphonuclear leukocytes was subnormal, although more polymerized than was observed in the patient's leukocytes (Movesesian M, Stossel TP: Unpublished data). These findings and the response to homotransplantation indicate that neutrophil actin dysfunction is a genetic disease, probably of autosomal recessive inheritance. Actin and other contractile proteins from various tissues, despite great functional and structural resemblances, are now known to be products of different genes (18). Therefore it is not surprising that, as best as could be ascertained, the patient's cardiovascular, skeletal, visceral, and neurologic functions were normal. However, erythrocytes, platelets, lymphoid cells, and mononuclear phagocytes arise from the same pluripotent hematopoietic stem cell, and only polymorphonuclear leukocyte function was clearly impaired. Possibly, the relevant mutation affected only the stem-cell progeny committed to the polymorphonuclear leukocyte lineage. A more appealing explanation is that the defect causes synthesis of an actin molecule that becomes unstable with aging. Because the polymorphonuclear leukocyte is a cell with little capacity to make new structural proteins, including actin (Boxer LA, Stossel TP: Unpublished data), it cannot renew its actin. On the other hand, mononuclear phagocytes can synthesize structural macromolecules (11). A third possibility is that neutrophil actin is more susceptible than other blood cell actins to inhibition of polymerization by a cytoplasmic factor. This inhibition might be accentuated in neutrophil actin dysfunction. Unfortunately, not enough information exists concerning the turnover, structure, and function of actin in these cells or in erythrocytes or platelets to permit further speculation. However, neutrophil actin dysfunction highlights the role of F-actin in motile effector mechanisms of polymorphonuStosse/ • Phagocytosis
Downloaded from https://annals.org by Tulane University user on 01/18/2019
4 0 1
clear leukocytes and suggests that we will find more diseases ascribable to abnormalities of protoplasmic contractile proteins.
bodies. / Gen Physiol 4:373-385, 1921 ^ 14. M U D D S, M C C U T C H E O N M, L U C R E B: Phagocytosis. Physiol
15.
Pseudopod Fusion
As in the case for signaling and adhesion, we have little information with which to explain the mechanism of this function.
16. 17. 18.
Conclusion
19.
Does it matter how phagocytes eat? Phagocytes will certainly eat without our understanding how they do it. Aside from the customary justification of scientific curiosity as essential for our culture, we can note that phagocytosis is a prototype of broad class cellular functions that we now understand poorly. These functions involve a cell's responding to external (or internal) signals by initiating some active motile response. Examples are the secretion of salivary, pancreatic, and leukocyte enzymes by exocytosis; the aggregation of platelets; the control of cell locomotion that is relevant for fetal development and wound-healing; and the lack of such control seen in the invasive behavior of cancer cells. Thinking optimistically, complete knowledge of the molecular basis of phagocytosis might provide ideas for pharmacologic manipulation of related events in many normal and disease states.
20.
21.
22.
23.
24. 25.
26.
ACKNOWLEDGMENTS: Grant support: U.S. Public Health Service Grant H L 19429 and a grant from the Edwin S. Webster Foundation.
27.
• Requests for reprints should be addressed to Thomas P. Stossel, M.D.; Massachusetts General Hospital; Boston, MA 02114.
28.
Received 10 April 1978; accepted 17 May 1978.
29.
References
30. B O X E R LA, R I C H A R D S O N S, F L O Y D A: Identification of actin-binding
1. METSCHNIKOFF E: Sur la lutte des cellules de l'organisme contre l'invasion des microbes. Ann Inst Pasteur 1:321-336, 1887 2. RODESCH FR, N E V E P, D U M O N T JE: Phagocytosis of latex beads by isolated thyroid cells. Exp Cell Res 60:354-360, 1970 3. Y O U N G RW, BOK D: Participation of the retinal pigment epithelium in the rod outer segment renewal process. / Cell Biol 42:392-403, 1969 4. CRUICKSHANK CND, HARCOURT SA: Pigment donation in vitro. J Invest Dermatol 42:183-184, 1964 5. CAVALIER-SMITH T: The origin of nuclei and of eukaryotic cells. Nature 256:463-468, 1975 6. BESSIS M: Living Blood Cells and their infrastructure. Springer Verlag, Berlin, Heidelberg, New York, 1973
protein in membrane of polymorphonuclear leukocytes. Nature 263:259,261 1976 31. D A V I E S WA, STOSSELL TP: Peripheral hyaline blebs (podosomes) of macrophages. / Cell Biol 75:941-955, 1977 32. H A R T W I G JH, DAVIES WA, STOSSEL TP: Evidence for contractile protein translocation in macrophage spreading, phagocytosis and phagolysosome formation. / Cell Biol 75:956-967, 1977 33. BROTSCHI EA, L A B A T E A, J A N T Z E N E, STOSSEL TP: Immunofluores-
cent localization of myosin in resting and phagocytosing macrophages (abstract). Clin Res 26:502A, 1978 34. L O W E Y S, SLAYTER HS, W E E D S AG, B A K E R H: Substructure of the
myosin molecule. I. Subfragments "of myosin by enzymatic dedgradaticm. J Mol Biol 42:1-29, 1969 TROTTER J A, ADELSTEIN RS: Evidence for myosin light chain kinase and phosphatase in macrophages (abstract). Biophys J 21:13a, 1978 ZIGMOND SH, HIRSCH JG: Effects of cytochalasin B on polymorphonuclear leukocyte locomotion, phagocytosis and glycolysis. Exp Cell Res 73:383-393, 1972 ALLISON AC, D A V I E S P, D E P E T R I S S: Role of contractile microfilaments in macrophage movement and endocytosis. Nature [New Biol] 232:153-155, 1971 H A R T W I G JH, STOSSEL TP: Interactions of actin myosin and an actinbinding protein of rabbit pulmonary macrophages. III. Effects of cytochalasin B.J Cell Biol 71:295-303, 1976 BOXER LA, H E D L E Y - W H Y T E ET, STOSSEL TP: Neutrophil actin dysfunction and abnormal neutrophil behavior. N Engl J Med 291:10931099, 1974
7. M O O R E PL, B A N K HL, BRISSIE NT, SPICER SS: Phagocytosis of bacte-
ria by polymorphonuclear leukocytes. A freeze-fracture, scanning electron microscope, and thin-section investigation of membrane structure. / Cell #70/76:158-174, 1978 8. ROBINEAUX J, FREDERIC J: Contribution a 1 etude des granulations neutrophiles des polynucleaires par la microcinematographie en contraste de phase. CR Soc Biol (Paris) 149:486-492, 1955 9. G O R E N MB: Phagocyte lysosomes: interactions with infectious agents, phagosomes, and experimental perturbations in function. Ann Rev Microbiol 31:507-533, 1977 10. W E I S S M A N N G, G O L D S T E I N I, H O F F S T E I N S, C H A U V E T G,
11. SILVERSTEIN SC, S T E I N M A N R M , C O H N ZA: Endocytosis. Ann
36.
37.
38.
39.
Rev
Biochem 46:669-722, 1977 12. STOSSEL TP: Phagocytosis: recognition and ingestion. Semin Hematol 14:83-116, 1975 13. F E N N WO: The theoretical response of living cells to contact with solid
September
35.
ROBI-
NEAUX R: Yin/Yang modulation of lysosomal enzyme release from polymorphonuclear leukocytes by cyclic nucleotides. Ann NY Acad Sci 256:222-232, 1975
402
Rev
14:210-275, 1934 VAN OSS CJ, G I L L M A N CF: Phagocytosis as a surface phenomenon. II. Contact angles and phagocytosis of encapsulated bacteria before and after opsonization by specific antiserum and complement. / ReticuloendothelSoc 12:497-502, 1972 W R I G H T CS, D O D D MC: Phagocytosis. Ann NY Acad Sci 59:945, 950, 1954 KARNOVSKY ML: Metabolic basis of phagocytic activity. Physiol Rev 42:143-168, 1962 K O R N ED: Biochemistry of actomyosin-dependent cell motility (a review). Proc Natl Acad Sci USA 75:588-599, 1978 L E W I S WH: The role of a superficial plasmagel layer in changes of form, locomotion, and division of cells in tissue cultures. Z Exp Zellforsch 23:1-7, 1939 KEYSERLINGK DG: Elektronenmikroskopische Untersuchung uber die Differenzierungsvorgange im Cytoplasma von segmentierten neutrophilen leukozyten wahrend der Zellbewegung Exp Cell Res 51:79-91, 1968 R E A V E N EP, A X L I N E SG: Subplasmalemmal microfilaments and microtubules in resting and phagocytizing cultivated macrophages / Cell Biol 59:12-27, 1973 H U X L E Y HE: Electron microscope studies on the structure of natural and synthetic protein filaments from striated muscle J Mol Biol 3:281308 1963 TUCKER JB: Endocytosis and streaming of highly gelated cytoplasm alongside rows of arm-bearing microtubules in the ciliate Nassula J Cell Sci 29:213-232, 1978 FERRY JD: Protein gels. Adv Protein Chem 4:2-78, 1948 H A R T W I G JH, STOSSEL TP: Isolation and properties of actin, myosin and a new actin-binding protein in rabbit alveolar marcrophages. / Biol Chem 250:5699-5705, 1975 STOSSELL TP, H A R T W I G JH: Interactions between actin, myosin and a new actin-binding protein of rabbit alveolar marcrophages. Macrophage myosin Mg + + -adenosine triphosphatase requires a cofactor for activation by actin. J Biol Chem 250:5706-5712, 1975 BOXER LA, STOSSEL TP: Interactions of actin, myosin and an actinbinding protein of chronic myelogenous leukemia leukocytes. / Clin Invest 57:964-976, 1976 STOSSEL TP, H A R T W I G JH: Interactions of actin, myosin and a new actin-binding protein of rabbit pulmonary macrophages. II. Role in cytoplasmic movement and phagocytosis. J Cell Biol 68:602-619, 1976 HEILBRUNN LV: The Dynamics of Living Protoplasm. New York, Academic Press, Inc. Publishers, 1956
40. C A M I T T A BM, Q U E S E N B E R R Y PJ, P A R K M A N R, B O X E R LA, STOSSEL
1978 • Annals of Internal Medicine • Volume 89 • Number
Downloaded from https://annals.org by Tulane University user on 01/18/2019
TP, CASSIDY JR, RAPPEPORT JN, N A T H A N DG: Bone marrow transplantation for a syndrome of neutrophil dysfunction. Exp Hematol 5:109-116, 1977
3
How Do Phagocytes Eat? THOMAS P. STOSSEL,
M.D.; Boston,
Massachusetts
A New York University Honors Program Lecture
Phagocytosis is a cellular function relevant for host defense against infection, tissue turnover, and other aspects of human physiology. Phagocytosis is also representative of functions wherein external stimuli activate motile events in the cell. Recognition of suitable objects by the plasma membrane of the phagocyte initiates phagocytosis. Knowledge of serum proteins that coat objects rendering them recognizable is considerable, but understanding of the chemical basis of recognition is meager. The signals activated by recognition are also not known. The work of phagocytosis that causes pseudopodia to enclose objects in vacuoles is ascribable to metabolic energy-dependent interactions between actin filaments and other contractile proteins in the peripheral cytoplasm. These interactions may also regulate the fusion of lysosomes with phagocytic vacuoles, an event important for the processing of ingested objects after phagocytosis.
PHAGOCYTOSIS is the process whereby cells internalize objects that are approximately micron-sized. Before addressing the question "how?'* let us review briefly some answers to the question, "Why is phagocytosis important?" The most familiar answers are that protists feed with it and that wandering and fixed hematopoietic cells, the polymorphonuclear leukocytes and mononuclear phagocytes, sequester, kill, or contain microorganisms that would otherwise devour higher organisms. These phagocytes also perform an important sanitation function in cleaning up damaged or aged host cells and other debris (1). Phagocytosis is also relevant for other functions. The thyroid follicle cell ingests stored thyroglobulin and then degrades it to form thyroid hormones (2). Retinal pigment epithelial cells interiorize and degrade outer rod segments to maintain the turnover of pigment vital for normal vision (3). Phagocytosis is part of the mechanism by which melanocytes transfer pigment to epidermal cells (4). Possibly the evolution of animal cells was influenced by development of a capacity to ingest microorganisms, which they subsequently captivated as their own subcellular organelles (5). The act of phagocytosis is visible under the light microscope, and the most naive observer can see what needs to be known to "explain" phagocytosis (Figure 1). The • From the Medical Oncology Unit, Massachusetts General Hospital; and the Department of Medicine, Harvard Medical School: Boston, Massachusetts. 398
Annals of Internal Medicine 89:398-402, 1978
Downloaded from https://annals.org by Tulane University user on 01/18/2019
observer notes that phagocytosis, like life, can be divided into seven stages. First, something about the surface of some objects and not others elicits recognition by the phagocytes. Second, the message, "Eat me" must be received by the cell. Third, the receipt must somehow signal the cell to do something about it, that something being the activation of effector mechanisms that are: four, an increase in stickiness of the membrane where the object has contacted the cell; five, assembly; six, movement and seven, fusion of pseudopods that surround and engulf totally or in part the recognized object, enclosing it in a phagocytic vacuole. A more sophisticated description of phagocytosis obtainable by observation with phase contrast, transmission, and scanning electron microscopy shows that the pseudopodia are pleats of peripheral cytoplasm (6, 7). This cortical material excludes the nucleus and other organelles, and therefore appears clear in phase contrast. The organelle exclusion ceases at the base of newly forming phagocytic vacuoles. At this region, lysosomes gain access to the plasma membrane and fuse with it, a process called degranulation (8). Degranulation results in the translocation of hydrolytic enzymes to the phagocytic vacuole, which is very important for postphagocytic digestive events (9). During phagocytosis some fraction of the lysosomal enzymes leaks out into the extracellular fluid (10). This lecture concerns recent studies pertaining to the mechanism of pseudopod assembly and movement, but will for completeness contain comments on the other steps involved in phagocytosis. Recognition
How do phagocytes respond to some objects and not others? The answer could theoretically emerge from an exhaustive study of surfaces that do not promote recognition, such as those of many native animal cells and of encapsulated microbes or of certain related artificial particles. These bland objects tend to be sugar coated like the phagocytes themselves, namely, they are encased in a mantle of slimy polysaccharide. Conversely, the answer could come out of an analysis of surfaces that do evoke congestion by phagocytes, for example, damaged animal cells or cells coated with phagocytosis-promoting serum proteins called "opsonins." Presently, IgG and a fragment of the third component of complement are the best defined opsonins, but there are probably others (11). Unfortunately, no simple common denominator currently ©1978 American College of Physicians
explains the different responses of phagocytes to varied surfaces. The difficulties in finding such a theme is that it is now impossible to map the precise location and configuration of recognition determinants on surfaces in anything but a gross way. The fact that one can measure the overall physical or chemical composition of a surface tells little about what may occur in the domain of the object that is interacting critically with the phagocyte membrane (Figure 2). Reception
Part of the answer to the question of phagocytic discrimination might arise from study of the external surface of phagocytes. Such an exercise could identify externally disposed molecules that bind critical surface determinants of objects. In fact, we currently think precisely in these terms of phagocytic "receptors." The receptors are actually defined by some of the rather specific effectors that activate phagocytosis (Table 1). Immunology tends to use the term "receptor" rather loosely ("Use can change the stamp of nature"), and it must be emphasized that there is little molecular definition of these phagocytic receptors. Clearly, the mere binding of an object to the surface of a phagocyte does not indicate a receptor mechanism any more than the adherence of mud to the sole of a shoe defines a mud receptor. On the other hand, the permeation of that same sole by a spike elicits a response. That response theoretically could indicate receptor involvement, but we know that the spike and not the sole confers specificity in this case. Responses evoked in the phagocyte by active recognition include the activation of certain oxidative enzymatic reactions as well as of the effector mechanisms of phagocytosis indicated above. This activation theoretically could as easily result from the capacity of recognition effector entities on objects to bypass surface determinants on the phagocyte and react nonspecifically with the plasma membrane of the phago-
Figure 2. How steric factors might influence and override overall charge effects in activation of phagocytosis. Net negative surface charge on Object 1 prevents access of effector molecule (v) to "trigger" (7) on negatively charged cell surface. However in another case, negatively charged Object 2 orients effector molecule right at trigger because of different location of negatively charged densities on cell. Opposite charges on cell and Object 3 prevent effectortrigger interaction, but in another case Object 4 promotes this interaction.
cyte, as it could from the binding of these recognition effector determinants to specific receptors. These riddles will yield to open-minded scrutiny. Signaling
How does recognition activate phagocytosis? We have no definite candidate for the signal. The main reason for this is that we do not know for sure what the signal wants to do. The identification of the signal mechanism requires a better understanding of the molecular basis of the effector mechanisms. Adhesion
Figure 1 . Visual recording by a "naive observer" of the morphology of phagocytosis. (Rendering by author's daughter, aged 6.)
In some cases, the binding of objects to the surface of phagocytes is nonspecific and cannot be viewed as part of the phagocytic process (11, 12). Nearly all factors (temperature, ions, metabolic poisons, other drugs) that influence the sticking of phagocytes to various objects have identical stimulatory or inhibitory effects on the overall act of phagocytosis (12). Therefore, it is not clear whether any of these factors influences something unique to the act of adhesion per se. Moreover, there is no knowledge of the molecules involved in the sticking process. Stosse/ • Phagocytosis
Downloaded from https://annals.org by Tulane University user on 01/18/2019
399
Table 1. Operationally Defined Phagocytic "Receptors"
Receptor
Characteristics
Fc C3 "Nonspecific"
Not impaired by trypsinization of the cell; not expressed on fibroblasts Impaired by trypsinization of the cell; not expressed on fibroblasts Not impaired by trypsinization of the cell
Pseudopod Assembly and Movement
The internalization of extracellular objects by cells requires mechanical work. Work is needed for many transport processes, but the necessity for such work is intuitively more obvious in the case of phagocytosis. To account for this work, early theories of phagocytosis invoked surface forces. Cells and other objects in suspension have interfaces with their suspending solvents, and the sums of the interfacial energies hypothetically could promote or retard incorporation of objects within cells if the latter were sufficiently elastic (13). It has been impossible to test this hypothesis critically, but experiments purporting to change the surface tension of objects or cells have been both consistent (14, 15) or at variance (16) with it. The definite demonstration that phagocytosis requires active cellular metabolism (17) put physical theories into the background, but still did not explain what work the metabolic energy was being used for. Recent evidence that muscle proteins may do work in nonmuscle cells (18) has revived another old notion: that a contractile skeleton in cells would generate the work for events such as phagocytosis (19). Actin is a major protein of mammalian hematopoietic phagocytes. Actins purified from phagocytes, like muscle and nonmuscle (protoplasmic) actins, exist under physiologic conditions as extended polymers (F-actin) composed of globular monomers (G-actin). Similar actin polymers are identifiable by electron microscopy in the periphery of whole phagocytes and are the principal structures in pseudopodia surrounding objects during phagocytosis (20, 21). These filaments do not lie in any particular orientation but form a rather tangled meshwork. (The "disorganization" of cytoplasmic filaments in mammalian phagocytes contrasts sharply with the crystalline filament assays in skeletal muscles [22] or even of certain phagocytic ciliates [23].) Such a mesh work has structural rigidity by definition, especially if the strands of the meshwork are cross-linked so that there is resistance to externally applied force (24). At least two proteins exist in mammalian phagocytes that cross-link actin, and both are of very high molecular weight. One is myosin, very similar structurally and functionally to myosin of muscle, and the other is called actin-binding protein (25-27). The latter is especially potent as an actincross-linking factor and at low concentrations markedly increases the intrinsic rigidity of F-actin (28). The existence of a rigid cortical gel has long been postulated to explain the organelle exclusion from the peripheral cytoplasm (2). Controlled sequential rigidification of the cell cortex around an object could conceivably
400
September
1978
effect pseudopod extension around that object. The "signal" evoked by recognition might cause actin-binding protein to cross-link actin. But how? Studies with both immunofluorescent (30) and biochemical (31, 32) techniques do suggest that some fraction of actin-binding protein in mammalian phagocytes is concentrated in the cell periphery and is associated with membranes, possibly plasma membranes. Phagocytosis decreases the association between actin-binding protein and the membranes of macrophages, causing more actin-binding protein to be extractable during subsequent disruption of the cells (28). But what phagocytosis actually does to actin-binding protein is unknown. Even if these questions were answerable, the process of phagocytosis is probably not so simple as the consequence of sequential actin gelation. In addition to rigidity forming in the cortex adherent to the object, it would be helpful to bring more cortical material into the area, thereby providing the external force mentioned above. Moreover, an explanation is needed for the attenuation of the cortical gel at the base of the phagocytic vacuole where lysosomes fuse with the phagocytic vacuoles membrane. There is indirect evidence that the protein myosin may be involved in these functions. Myosin, in the presence of magnesium and ATP, compresses gels composed of F-actin and actin-binding protein (27, 28). Both biochemical and immunofluorescent methods show that myosin is concentrated in the cortical cytoplasm of macrophages during pseudopod formation and that myosin moves away from the base of phagocytic vacuoles as they form (32, 33). The active compression of actin filaments by myosin in pseudopodia could have several effects: [1.] More actin-rich cortical matter is brought into the pseudopods; [2.] the rigid actin gel lattice is destroyed at the vacuole's base—it is literally torn apart by myosin pulling from the lateral pseudopods (Figure 3); [3.] the pseudopod is flattened into a pleat; [4.] the controlled disruption of the cortical actin gel by myosin or other means reduces the stability of the plasma membrane. If we envision the plasma membrane as a lipid bilayer supported by an actin cytoskeleton, like cellophane stretched over chicken wire, loss of cytoskeletal support could permit the bilayer to revert to a lower energy state by assuming a spherical
Figure 3. Schematic rendition of phagocytosis by a macrophage in suspension, a. Signal transmitted after recognition activates actinbinding protein and myosin ATPase in organelle-excluding peripheral cytoplasm; b. arrows depict flow of peripheral cytoplasmic gel, which c. drives pseudopodia around object and permits lysosomes access to base of phagocytic vacuole {arrow).
• Annals of Internal Medicine • Volume 89 • Number
Downloaded from https://annals.org by Tulane University user on 01/18/2019
3
shape. Hence actin gel disruption as well as formation could promote the formation of phagocytic vacuoles. This entire picture, even if true, still leaves many questions. For example, what is the precise location of myosin molecules: Are they bound directly to the plasma membrane? Myosin molecules have two domains: a globular head that interacts with actin and a helical tail. In muscle, the tails form the backbone of myosin filaments (34). In nonmuscle cells such as macrophages, myosin filaments are not usually seen, so it is possible that the tails insert into the membrane. Second, what controls the interaction between myosin and actin? This control is well understood for muscle (22), but not for mammalian phagocytes or other nonmuscle cells. In macrophages a protein "cofactor" activates the contraction of actin by myosin (26-28); this cofactor may be a kinase enzyme that acts by phosphorylating the myosin molecule (35). Despite lack of knowledge about details, the basic idea that contractile proteins do the work of phagocytosis is becoming increasingly well substantiated. First, the requirement of ATP for actin filament meshwork contraction by myosin provides a locus for energy phagocytosis. Second, at least two examples are described wherein factors that impair contractile protein structure or function inhibit phagocytosis. Cytochalasin B
Cytochalasin B, a fungal product, produces a dose-dependent (10- 7 — 10-5 M) inhibition of phagocytosis by polymorphonuclear leukocytes (36) and macrophages (37), although binding of objects to the phagocyte membrane is relatively unaltered. Paradoxically the agent causes an increase in secretion of lysosomal enzymes into the extracellular fluid by phagocytes exposed to ingestible objects (10). Recent work indicates that cytochalasin B dissolves gels of actin cross-linked by macrophage actinbinding protein (38). It has no effect on the interaction between actin and myosin. Cytochalasin B is therefore a specific inhibitor of the gel state of actin, and it follows that solvation of the cortical actin gel could prevent pseudopod formation. Cytochalasin B could also allow myosin to compress actin meshworks in the cell periphery with great ease because myosin exerts its force on a less rigid structure than in the untreated actin gel; therefore, the rate of compression increases. This formulation explains why lysosomal granule access to the plasma membrane and subsequent secretion can be enhanced by cytochalasin B. Neutrophil Actin Dysfunction
In 1973, a male infant was encountered who, from birth on, suffered recurrent and severe pyogenic infections caused by low-grade pathogens. The infections involved predominantly the skin and gastrointestinal tracts, responded poorly to antibiotic therapy, and were characterized by absence of pus. The failure of pus formation paradoxically was in the setting of a profound neutrophilic leukocytosis. Leukocytes transfused from normal sera into the infant's circulation appeared unlike the patient's own leukocytes in inflammatory foci. Whereas normal
polymorphonuclear leukocytes, spread on glass slides, crawl about and ingest opsonized objects, the infant's polymorphonuclear leukocytes were virtually unable to perform any of these functions. However, the patient's cells responded to opsonized objects by binding to them, activating oxygen metabolism, and secreting lysosomal enzymes into the extracellular fluid. Therefore, recognition, signaling, and effector mechanisms not involving pseudopod formation were relatively unimpaired. Very little of the G-actin in homogenates of this child's polymorphonuclear leukocytes polymerized into filaments, although the amount of G-actin was normal. Under similar conditions, two thirds to three fourths of G-actin in homogenates of normal polymorphonuclear leukocytes polymerized (39). Because of continued clinical deterioration, the patient was successfully engrafted with marrow from a histocompatible sibling after chemical immunosuppression and total body irradiation. After bone marrow transplantation the patient's polymorphonuclear leukocytes had normal function. Unfortunately, the child died of pulmonary failure, probably secondary to a viral infection (40). The parents and siblings of the patient were not unusually susceptible to infection, and their polymorphonuclear leukocytes had normal function. However, the amount of G-actin that polymerized in homogenates of their polymorphonuclear leukocytes was subnormal, although more polymerized than was observed in the patient's leukocytes (Movesesian M, Stossel TP: Unpublished data). These findings and the response to homotransplantation indicate that neutrophil actin dysfunction is a genetic disease, probably of autosomal recessive inheritance. Actin and other contractile proteins from various tissues, despite great functional and structural resemblances, are now known to be products of different genes (18). Therefore it is not surprising that, as best as could be ascertained, the patient's cardiovascular, skeletal, visceral, and neurologic functions were normal. However, erythrocytes, platelets, lymphoid cells, and mononuclear phagocytes arise from the same pluripotent hematopoietic stem cell, and only polymorphonuclear leukocyte function was clearly impaired. Possibly, the relevant mutation affected only the stem-cell progeny committed to the polymorphonuclear leukocyte lineage. A more appealing explanation is that the defect causes synthesis of an actin molecule that becomes unstable with aging. Because the polymorphonuclear leukocyte is a cell with little capacity to make new structural proteins, including actin (Boxer LA, Stossel TP: Unpublished data), it cannot renew its actin. On the other hand, mononuclear phagocytes can synthesize structural macromolecules (11). A third possibility is that neutrophil actin is more susceptible than other blood cell actins to inhibition of polymerization by a cytoplasmic factor. This inhibition might be accentuated in neutrophil actin dysfunction. Unfortunately, not enough information exists concerning the turnover, structure, and function of actin in these cells or in erythrocytes or platelets to permit further speculation. However, neutrophil actin dysfunction highlights the role of F-actin in motile effector mechanisms of polymorphonuStosse/ • Phagocytosis
Downloaded from https://annals.org by Tulane University user on 01/18/2019
4 0 1
clear leukocytes and suggests that we will find more diseases ascribable to abnormalities of protoplasmic contractile proteins.
bodies. / Gen Physiol 4:373-385, 1921 ^ 14. M U D D S, M C C U T C H E O N M, L U C R E B: Phagocytosis. Physiol
15.
Pseudopod Fusion
As in the case for signaling and adhesion, we have little information with which to explain the mechanism of this function.
16. 17. 18.
Conclusion
19.
Does it matter how phagocytes eat? Phagocytes will certainly eat without our understanding how they do it. Aside from the customary justification of scientific curiosity as essential for our culture, we can note that phagocytosis is a prototype of broad class cellular functions that we now understand poorly. These functions involve a cell's responding to external (or internal) signals by initiating some active motile response. Examples are the secretion of salivary, pancreatic, and leukocyte enzymes by exocytosis; the aggregation of platelets; the control of cell locomotion that is relevant for fetal development and wound-healing; and the lack of such control seen in the invasive behavior of cancer cells. Thinking optimistically, complete knowledge of the molecular basis of phagocytosis might provide ideas for pharmacologic manipulation of related events in many normal and disease states.
20.
21.
22.
23.
24. 25.
26.
ACKNOWLEDGMENTS: Grant support: U.S. Public Health Service Grant H L 19429 and a grant from the Edwin S. Webster Foundation.
27.
• Requests for reprints should be addressed to Thomas P. Stossel, M.D.; Massachusetts General Hospital; Boston, MA 02114.
28.
Received 10 April 1978; accepted 17 May 1978.
29.
References
30. B O X E R LA, R I C H A R D S O N S, F L O Y D A: Identification of actin-binding
1. METSCHNIKOFF E: Sur la lutte des cellules de l'organisme contre l'invasion des microbes. Ann Inst Pasteur 1:321-336, 1887 2. RODESCH FR, N E V E P, D U M O N T JE: Phagocytosis of latex beads by isolated thyroid cells. Exp Cell Res 60:354-360, 1970 3. Y O U N G RW, BOK D: Participation of the retinal pigment epithelium in the rod outer segment renewal process. / Cell Biol 42:392-403, 1969 4. CRUICKSHANK CND, HARCOURT SA: Pigment donation in vitro. J Invest Dermatol 42:183-184, 1964 5. CAVALIER-SMITH T: The origin of nuclei and of eukaryotic cells. Nature 256:463-468, 1975 6. BESSIS M: Living Blood Cells and their infrastructure. Springer Verlag, Berlin, Heidelberg, New York, 1973
protein in membrane of polymorphonuclear leukocytes. Nature 263:259,261 1976 31. D A V I E S WA, STOSSELL TP: Peripheral hyaline blebs (podosomes) of macrophages. / Cell Biol 75:941-955, 1977 32. H A R T W I G JH, DAVIES WA, STOSSEL TP: Evidence for contractile protein translocation in macrophage spreading, phagocytosis and phagolysosome formation. / Cell Biol 75:956-967, 1977 33. BROTSCHI EA, L A B A T E A, J A N T Z E N E, STOSSEL TP: Immunofluores-
cent localization of myosin in resting and phagocytosing macrophages (abstract). Clin Res 26:502A, 1978 34. L O W E Y S, SLAYTER HS, W E E D S AG, B A K E R H: Substructure of the
myosin molecule. I. Subfragments "of myosin by enzymatic dedgradaticm. J Mol Biol 42:1-29, 1969 TROTTER J A, ADELSTEIN RS: Evidence for myosin light chain kinase and phosphatase in macrophages (abstract). Biophys J 21:13a, 1978 ZIGMOND SH, HIRSCH JG: Effects of cytochalasin B on polymorphonuclear leukocyte locomotion, phagocytosis and glycolysis. Exp Cell Res 73:383-393, 1972 ALLISON AC, D A V I E S P, D E P E T R I S S: Role of contractile microfilaments in macrophage movement and endocytosis. Nature [New Biol] 232:153-155, 1971 H A R T W I G JH, STOSSEL TP: Interactions of actin myosin and an actinbinding protein of rabbit pulmonary macrophages. III. Effects of cytochalasin B.J Cell Biol 71:295-303, 1976 BOXER LA, H E D L E Y - W H Y T E ET, STOSSEL TP: Neutrophil actin dysfunction and abnormal neutrophil behavior. N Engl J Med 291:10931099, 1974
7. M O O R E PL, B A N K HL, BRISSIE NT, SPICER SS: Phagocytosis of bacte-
ria by polymorphonuclear leukocytes. A freeze-fracture, scanning electron microscope, and thin-section investigation of membrane structure. / Cell #70/76:158-174, 1978 8. ROBINEAUX J, FREDERIC J: Contribution a 1 etude des granulations neutrophiles des polynucleaires par la microcinematographie en contraste de phase. CR Soc Biol (Paris) 149:486-492, 1955 9. G O R E N MB: Phagocyte lysosomes: interactions with infectious agents, phagosomes, and experimental perturbations in function. Ann Rev Microbiol 31:507-533, 1977 10. W E I S S M A N N G, G O L D S T E I N I, H O F F S T E I N S, C H A U V E T G,
11. SILVERSTEIN SC, S T E I N M A N R M , C O H N ZA: Endocytosis. Ann
36.
37.
38.
39.
Rev
Biochem 46:669-722, 1977 12. STOSSEL TP: Phagocytosis: recognition and ingestion. Semin Hematol 14:83-116, 1975 13. F E N N WO: The theoretical response of living cells to contact with solid
September
35.
ROBI-
NEAUX R: Yin/Yang modulation of lysosomal enzyme release from polymorphonuclear leukocytes by cyclic nucleotides. Ann NY Acad Sci 256:222-232, 1975
402
Rev
14:210-275, 1934 VAN OSS CJ, G I L L M A N CF: Phagocytosis as a surface phenomenon. II. Contact angles and phagocytosis of encapsulated bacteria before and after opsonization by specific antiserum and complement. / ReticuloendothelSoc 12:497-502, 1972 W R I G H T CS, D O D D MC: Phagocytosis. Ann NY Acad Sci 59:945, 950, 1954 KARNOVSKY ML: Metabolic basis of phagocytic activity. Physiol Rev 42:143-168, 1962 K O R N ED: Biochemistry of actomyosin-dependent cell motility (a review). Proc Natl Acad Sci USA 75:588-599, 1978 L E W I S WH: The role of a superficial plasmagel layer in changes of form, locomotion, and division of cells in tissue cultures. Z Exp Zellforsch 23:1-7, 1939 KEYSERLINGK DG: Elektronenmikroskopische Untersuchung uber die Differenzierungsvorgange im Cytoplasma von segmentierten neutrophilen leukozyten wahrend der Zellbewegung Exp Cell Res 51:79-91, 1968 R E A V E N EP, A X L I N E SG: Subplasmalemmal microfilaments and microtubules in resting and phagocytizing cultivated macrophages / Cell Biol 59:12-27, 1973 H U X L E Y HE: Electron microscope studies on the structure of natural and synthetic protein filaments from striated muscle J Mol Biol 3:281308 1963 TUCKER JB: Endocytosis and streaming of highly gelated cytoplasm alongside rows of arm-bearing microtubules in the ciliate Nassula J Cell Sci 29:213-232, 1978 FERRY JD: Protein gels. Adv Protein Chem 4:2-78, 1948 H A R T W I G JH, STOSSEL TP: Isolation and properties of actin, myosin and a new actin-binding protein in rabbit alveolar marcrophages. / Biol Chem 250:5699-5705, 1975 STOSSELL TP, H A R T W I G JH: Interactions between actin, myosin and a new actin-binding protein of rabbit alveolar marcrophages. Macrophage myosin Mg + + -adenosine triphosphatase requires a cofactor for activation by actin. J Biol Chem 250:5706-5712, 1975 BOXER LA, STOSSEL TP: Interactions of actin, myosin and an actinbinding protein of chronic myelogenous leukemia leukocytes. / Clin Invest 57:964-976, 1976 STOSSEL TP, H A R T W I G JH: Interactions of actin, myosin and a new actin-binding protein of rabbit pulmonary macrophages. II. Role in cytoplasmic movement and phagocytosis. J Cell Biol 68:602-619, 1976 HEILBRUNN LV: The Dynamics of Living Protoplasm. New York, Academic Press, Inc. Publishers, 1956
40. C A M I T T A BM, Q U E S E N B E R R Y PJ, P A R K M A N R, B O X E R LA, STOSSEL
1978 • Annals of Internal Medicine • Volume 89 • Number
Downloaded from https://annals.org by Tulane University user on 01/18/2019
TP, CASSIDY JR, RAPPEPORT JN, N A T H A N DG: Bone marrow transplantation for a syndrome of neutrophil dysfunction. Exp Hematol 5:109-116, 1977
3
