Parasitology (1975), 71, 483-491 With 4 plates and 1 figure in the text
Structure and invasive behaviour of Plasmodium knowlesi merozoites in vitro L. H. BANNISTER,* G. A. BUTCHER,t E. D. DENNISf and G. H. MITCHELLf * Departments of Biology and Anatomy, and fo/ Chemical Pathology, Guy's Hospital Medical School, London SE1 9RT (Received 14 May 1975) SUMMARY
The structure and invasive behaviour of extracellular erythrocytic merozoites prepared by a cell sieving method have been studied with the electron microscope. Free merozoites contain organelles similar to those described in late schizonts of Plasmodium knowlesi. Their surface is lined by a coat of short filaments. On mixing with fresh red cells, merozoites at first adhere, then cause the red cell surface to invaginate rapidly, often with the formation of narrow membranous channels in the red cell interior. As the merozoite enters the invagination it forms an attachment by its cell coat to the rim of the pit, and finally leaves this coat behind as it is enclosed in a red cell vacuole. Dense, rounded intracellular bodies then move to the merozoite periphery, and apparently rupture to cause further localized invagination of the red cell vacuole. The merozoite finally loses its rhoptries, the pellicle is reduced to a single membrane and the parasite becomes a trophozoite. Invasion is complete by 1 min after adhesion, and the trophozoite is formed by 10 min.
In recent years electron microscopical methods have provided much information regarding the structure and behaviour of different species of Plasmodium (see the reviews of this subject by Rudzinska & Vickerman, 1968; Rudzinska, 1969; Aikawa, 1971). In particular, knowledge of the merozoite stages has increased greatly and their cytology has been described in some detail for several species, including P. knowlesi (Aikawa, Cook, Sakoda & Sprinz, 1969). Because of the difficulty of obtaining sufficient numbers of extracellular forms, attention has centred mainly on parasites fixed within the schizont during the final stages of merogony, although free merozoites have also been studied (see Aikawa, 1967; Garnham, Bird, Baker & Killick-Kendrick, 1969). In 1969, Ladda, Aikawa & Sprinz gave a detailed fine structural account of the process of erythrocyte invasion by merozoites of P. berghei yoelii and of P. gallinaceum as seen in random samples of heavily infected blood (see also Ladda, 1969). It is known from direct observation of in vitro material that after merozoites have been released from schizonts, their entry into erythrocytes is extremely 33
L. H. BANNISTER AND OTHERS
rapid (Trager, 1956; Dvorak, Miller, Whitehouse & Shiroishi, 1975); thus the chances of finding appropriate stages of penetration in electron microscopic sections of randomly sampled material are low, requiring the examination of large numbers of specimens. However, Ladda et al. (1969) clearly demonstrated that merozoites attach to erythrocyte membranes by their anterior prominences, then cause the host cell surface to invaginate deeply, and eventually to enclose the parasite within a vacuole which it continues to inhabit throughout the erythrocyte phase of its life. Dvorak et al. (1975) have also shown, in in vitro studies of P. knowlesi, that the behaviour of merozoites is even more complex than this, since invasion involves various sequential changes in the shapes of both host and parasite cells. Several problems concerning the structure and behaviour of merozoites remain to be settled, including the nature of their cell surface and of the various dense bodies within the cytoplasm, and the details of erythrocyte invasion. In a previous communication the structure and invasive behaviour of merozoites prepared by a new sieving technique have been reported briefly (Bannister, Butcher, Dennis & Mitchell, 1975). This method, which is described in the previous paper (Dennis, Mitchell, Butcher & Cohen, 1975), has allowed the examination of large quantities of extracellular merozoites and facilitated studies of their early invasive stages in red cells. A detailed account of these investigations is now presented. MATERIALS AND METHODS
Living merozoites were obtained by separation from schizonts in the manner set out in the previous paper. For electron microscopy of merozoites alone, 5 ml of cell suspension from the sieve apparatus (containing 5 x 107 to 108 merozoites per ml) were allowed to run into an equal volume of 6 % glutaraldehyde fixative (buffered with 0-2 M phosphate at pH 7-2). In studies of penetration, 20 ml aliquots of merozoite suspension collected from the apparatus without cooling were added to 0-2 ml (approximately 2-5 x 109) of packed normal red cells and inoculated at 37 °C. Samples (4 ml) of this material were taken 1, 5 and 10 min later and added to equal volumes of the 6 % glutaraldehyde fixative. Samples were fixed for 1 h at room temperature. The material was then spun at 1500g (for merozoite-red cell suspensions) or 3000g (for merozoites alone) for 10 min, then the fixative was removed and the pellet resuspended in 0-2 M phosphate buffer (pH 7-4) for initial washing. Further centrifugation into pellets was carried out in a Beckman 'Microfuge'. Pellets were washed several times over 8 h in 0-2 M phosphate buffer at pH 7-4, treated with 1 % buffered osmium tetroxide for 2 h, dehydrated in an ethanol series and embedded in ' TAAB' epoxy resin. Sections cut on a diamond knife were stained with uranyl acetate and lead citrate, and were viewed in RCA EMU-4A and Hitachi HU 12A electron microscopes.
Red cell invasion by P. knowlesi
Structure of extracellular merozoites
Erythrocytic merozoites harvested with the cell sieve apparatus closely resemble those within mature schizonts (see Aikawa et al. 1969). Merozoites are pyriform, fusiform or ellipsoidal in shape with an anterior apical prominence; they measure about 2 by 1-5 /tm (PI. 1A). The nucleus is situated at the posterior end and is usually semicircular in longitudinal sections of merozoites; the nuclear contents are typically finely fibrillar and closely packed with occasional larger dense granules. A typical nuclear envelope, perforated by pore complexes, is present. The cytoplasm is largely occupied by uniformly spaced ribosomes set in a matrix of very fine (2-3 nm) filaments and granules. Profiles of a double membrane bounded body (mitochondrion: PI. IE) are often arranged in a series of two or three suggesting that a single coiled structure has been sectioned longitudinally. The inner membrane of this organelle is coated with granular or filamentous material. Associated with the mitochondrion, and often lying between it and the nucleus or forming an S-shaped profile in the centre of the cell, is the multilamellar body formed of two or more parallel cisternae of smooth membranes (PI. 1A). Various membrane bounded bodies with dense interiors occupy the anterior half of the merozoite. These include the rhoptries (paired organelles), the smaller fusiform 'micronemes' (Scholtyseck & Mehlhorn, 1970) many of which, like the rhoptries, converge on the apical prominence, and clusters of larger (50-80 nm) bodies of a more circular profile (PI. 1 A, B, C); the latter will be referred to as ' microspheres' in the present work to avoid confusion with the other types of dense membrane bounded body present in the cell. Groups of interconnected membranous channels and vacuoles, some with dense contents, were seen in some merozoites, but in this study none of these appeared to be associated specifically with the rhoptries, unlike the similar structures reported in P. brasilianum by Sterling, Aikawa & Nussenzweig (1972). The outer surface of the whole merozoite is covered with a cell coat consisting of short (20 nm) filaments projecting at right angles from the external membrane (PI. 1C). These filaments are about 4nm in diameter although somewhat wider at their distal end (PI. 1C). The outer membrane of the merozoite has the usual unit structure (Aikawa, 1967) and beneath this lies a thin layer of pellicular cytoplasm (PI. 1C). Intermediate and inner pellicular membranes, which are sometimes separated from each other by a variable gap (PI. 1C), probably artifactual, underlie this cytoplasm. Discontinuities of the intermediate and inner membranes occur beneath the apical prominence, at the cytostome (PI. ID) and occasionally in other parts of the cell. A few longitudinal microtubules are sometimes visible immediately beneath the inner membrane of the pellicle. The structures within the apical prominence are similar to those observed in other species of Plasmodium (see e.g. P. lophurae: Aikawa, 1967) and consist of three polar rings (PI. IB), the most basal of which forms a conical collar. The 33-2
L. H. BANNISTER AND OTHERS
exact shape of this complex is variable; in the more spherical merozoites, which may be undergoing degenerative changes, the apical prominence is flattened and the rings and collars form a series of concentric annuli. Narrow canals from the paired organelles contact the flat surface of the apical prominence, and small dense masses of cytoplasm are associated with the tips of these canals. Invasion of erythrocytes Samples fixed 1 min after the addition of red cells to merozoite suspensions contained parasites at all stages of invasion, and complete differentiation into ring forms had occurred by 10 min. Whilst initial adhesion may be between any part of the merozoite and the red cell surface (PI. 2 A), it is known from in vitro observations that penetration commences only when the apical prominence makes contact with the red cell (see Introduction). In the electron microscope such early stages show the apposition of the flat end of the apical prominence and the red cell surface, with a narrow gap of about 10 nm lying between the two. Often part of the red cell surface is partially engulfed within the inverted conical end of the parasite and narrow channel systems run deep into the red cell interior (PI. 2B). A little later on invagination of the red cell has occurred in the region of contact to form a wide pit with a few subsidiary channels or vacuoles extending from its deepest point into the red cell interior (PI. 2C). As the pit continues to deepen and the merozoite to enter further into the red cell the cell coat of the parasite is removed from the invading end, but a line of attachment is maintained between the outer rim of the pit and the cell coat. Little of this coat is therefore carried into the interior of the pit but at first it accumulates externally, then, as the merozoite is finally enclosed within the red cell, the coat is shed completely and disperses into the surrounding medium (P1.3B.C). The shape of the merozoite becomes distorted as it passes through the opening of the pit (Pis. 2D; 3 A), but once inside the erythrocyte it resumes a rounded profile (PI. 3C, D). At a slightly later stage its outline becomes irregular as the microspheres move to the periphery of the merozoite causing its surface to bulge (PI. 4A); later these bodies are seen to have ruptured, releasing their finely granular contents into the erythrocyte vacuole to cause further invaginations of its limiting membrane (PL 4B, C). The parasite subsequently has an elliptical or irregular outline and in many cases only a single pellicular membrane is visible (PL 4D). Micrographs often show membrane debris in the vacuole containing the parasite, or passing out through a narrow external aperture to the exterior; such material could have come from the erythrocyte rather than the merozoite exterior, and how the pellicular complex is reduced to a single thickness of membrane is not clear. With the formation of vacuoles containing red cell cytoplasm the parasite becomes a typical trophozoite with a more extensive surface area. It is noteworthy that the rhoptries and microspheres remain visible up to the time of transition into the trophozoite stage (e.g. PL 4 A), although the density of
Red cell invasion by P. knowlesi
one or both rhoptries was sometimes seen to be lower after red cell entry. The process of rhoptry discharge itself was not elucidated in this study. Micronemes appear to be much reduced in number after the initial invasion sequence, but again, the precise fate of these organelles remains uncertain. DISCUSSION
The well-preserved fine structure of merozoites prepared by the sieving method shows that although some of them may be damaged by this technique of separation, many merozoites are little affected, as also shown by the viability studies described in the previous paper (Dennis et al. 1975). A number of structural features require comment. The consistent appearance and regular arrangement of short filaments forming the outer coat of the merozoite show that the coat is similar to the glycocalyces of other types of cell, and does not merely consist of aggregated material derived from the vacuole of the schizont. The coat appears to be responsible for the initial adhesion to the erythrocyte membrane and for maintaining a seal between the merozoite and the rim of the invaginating red cell surface during invasion. The parasite coat may assist in invasion by stabilizing the red cell membrane at the rim of the invagination, thus limiting its expansion to the area within this ring. The properties of this coat are unusual. It is known that it differs chemically from the glycocalyces of non-parasitic cells in its relative lack of sialyl groups (Seed, Aikawa, Sterling & Rabbege, 1974); electrophoretic mobility measurements of merozoites indicate that they possess a much lower surface charge than do erythrocytes (Gregson, Bannister & Butcher, unpublished results), a feature which is likely to assist in adhesion to red cells. Since the coat is sloughed into the extracellular medium, it may be important in stimulating antibody production ; it is therefore interesting that antibodies to soluble serum antigens which may correspond with such material have been detected in malaria infections (Wilson, 1974). The membrane bounded dense bodies of various kinds appear to fall into three distinct categories: the rhoptries, micronemes and the rounded 'microspheres'. It has been suggested (see Aikawa, 1971) that the various types of smaller dense body are precursors of the paired organelles, and that the channel systems sometimes seen in association with the paired organelles enable the smaller bodies to coalesce into the larger ones. The present results indicate that while the micronemes may be involved in the initial steps of invasion, these, the rhoptries and the microspheres appear to remain separate from each other throughout the process; the microspheres, indeed, play a distinctive part in the later stages of invasion. There is no evidence that, as suggested by Scholtyseck & Mehlhorn (1970) the various dense bodies coalesce at any point in their life span, although some degree of fusion cannot be excluded. The process of invasion in Plasmodium knowlesi is in many respects similar to that described in P. berghei and P. gallinaceum by Ladda et al. (1969). The present results confirm that the parasite does not actually penetrate the red cell membrane but comes to lie within an erythrocyte vacuole. The means by which the parasite
L. H. BANNISTER AND OTHERS
Fig. 1. Diagrammatic summary of the chief events of red cell invasion by a merozoite. Note that although the final stage of rhoptry emptying is shown in the intracellular form, this is an inferred rather than an observed event.
induces vacuole formation is not entirely clear but ultrastructural evidence points to the induction of a rapid, localized increase in red cell surface area, sometimes also resulting in the formation of deep narrow channels within the erythrocyte (PI. 2B, C). These channels appear to originate at or near the point of contact of the red cell membrane with the apical prominence of the merozoite and may therefore derive from the activity of the micronemes or rhoptries. Ladda et al. (1969) have suggested that these bodies may contain a surfactant material which could cause rapid stretching of the red cell membrane. Expansion of the latter is not normally associated with the formation of holes large enough to cause haemolysis (see also Dvorak et al. 1975), nor is there any noticeable change in the trilaminar structure of the red cell membrane. It is therefore possible that substances released from the various dense bodies of the merozoite are incorporated into the red cell membrane to cause its ex-
Red cell invasion by P. knowlesi
pansion. It is known that various amphiphilic compounds such as some classes of anaesthetics can, in low concentrations, cause such membrane effects by disordering the arrangement of the phospholipid bilayers; recently it has been suggested that such materials may produce surface invaginations in red cells by causing the inner bilayer to expand more than the outer one (Sheetz & Singer, 1974), and it is possible that such a mechanism might be responsible for the changes observed in merozoite invasion; the various membranous whorls and clefts known to occur in red cells infected with Plasmodium (see Aikawa, 1971) and the changes in antigenicity and permeability of the host red cell (Homewood & Neame, 1974; Sherman & Tanigoshi, 1974) may also be consequences of such an alteration and expansion of the red cell vacuole membrane during and after invasion. Cytochemical tests are in progress to determine the chemical nature of the dense bodies and to examine their effects on the red cell membrane. The mechanism of invasion initially involves the adhesion of the apical prominence to the rapidly invaginating cell surface, the zone of membrane expansion being limited to the region between the point of adhesion and the rim of the invagination where the coat of the merozoite contacts the red cell. The merozoite would then seem to be pulled by its apical attachment out of its cell coat by the invaginating red cell membrane or perhaps sucked into the rapidly expanding vacuole; no other force than that generated by the expanding red cell membrane need therefore be invoked to explain its entry into the vacuole. In any case, there is little evidence from observations of live merozoites of P. knowlesi that they have any motility of their own (Butcher & Cohen, 1970). A diagram summarizing the invasion sequence is shown in Pig. 1. The sequence of structural changes also complements the in vitro observations of Dvorak et al. (1975) on cell invasion. Two major events occur in the hostparasite relationship during invasion, namely the initial entry by the parasite and then the change in shape of the merozoite associated with the release of microsphere material. Both of these also involve a sudden expansion of red cell membrane, and probably explain the observed in vitro changes in red cell shape at the beginning and at the end of invasion. However, much remains to be discovered about the precise mechanism of invasion and the many adaptations which the merozoite undoubtedly shows to the intracellular mode of parasitism. The authors wish to thank the Medical Research Council (U.K.) and the World Health Organization for financial support. They are also much indebted to Professor S. Cohen for his advice and encouragement, and to Miss Patricia Wilmot, Mrs Christina Cunliffe and Mr D. Lovell for technical assistance.
L. H. BANNISTER AND OTHERS
REFERENCES AIKAWA, M. (1967). Ultrastructure of the pellioular complex of Plasmodium fallax. Journal of Cell Biology 35, 103-13. AIKAWA, M. (1971). Plasmodium: the fine structure of malarial parasites. Experimental Parasitology 30, 284-320. AIKAWA, M., COOK, R. T., SAKODA, J . J . & SPRINZ, H. (1969). Fine structure of the erythro-
cytic stages of Plasmodium knoivlesi: A comparison between intracellular and free forms. Zeitschrift fur Zellforschung und Mikroskopische Anatomie 100, 271-84. BANNISTER, L. H., BUTCHER, G. A. DENNIS, E. D. & MITCHELL, G. H. (1975). Studies on the
structure and invasive behaviour of merozoites of Plasmodium knowlesi. Transactions of the Royal Society of Tropical Medicine and Hygiene 69, 5. BUTCHER, G. A. & COHEN, S. (1970). Schizogony of Plasmodium knoivlesi in the presence of normal and immune sera. Transactions of the Royal Society of Tropical Medicine and Hygiene 64, 470. DENNIS, E. D., MITCHELL, G. H., BUTCHER, G. A. & COHEN, S. (1975). In vitro isolation of
Plasmodium knowlesi merozoites using polycarbonate sieves. Parasitology 71, 475-81. DVORAK, J . A., MILLER, L. H., WHITEHOUSE, W. C. & SHIROISHI, T. (1975). Invasion of
erythrocytes by malaria merozoites. Science 187, 748-50. GARNHAM, P. C. C , BIRD, R. G., BAKER, J . R. & KILLICK-KENDRICK, R. (1969). Electron
microscope studies on the motile stages of malaria parasites. Transactions of the Royal Society of Tropical Medicine and Hygiene 63, 328-32. HOMEWOOD, C. A. & NEAME, K. D. (1974). Malaria and the permeability of the host erythrocyte. Nature, London 252, 718-19. LADDA, R. L. (1969). New insights into the fine structure of rodent malarial parasites. Military Medicine 134, 825-64. LADDA, R., AIKAWA, M. & SPRINZ, H. (1969). Penetration of erythrocytes by merozoites of mammalian and avian malarial parasites. Journal of Parasitology 65, 633-44. RUDZINSKA, M. A. (1969). The fine structure of malaria parasites. International Review of Cytology 25, 161-99. RUDZINSKA, M. A. & VICKERMAN, K. (1968). The fine structure. In Infectious Blood Diseases of Man and Animals (ed. D. Weinman and M. Ristic), vol. 1, pp. 217-306. New York: Academic Press. SCHOLTYSECK, E. & MEHLHORN, H. (1970). Ultrastructural study of characteristic organelles (paired organelles, micronemes, micropores) of sporozoa and related organisms. Zeitschrift fur Parasitenkunde 34, 97-129. SEED, T. M., AIKAWA, M., STERLING, C. & RABBEGE, J . (1974). Surface properties of extra-
cellular malaria parasites: morphological and cytochemical study. Infection and Immunity 9, 750-61. SHEETZ, M. P. & SINGER, S. J . (1974). Biological membranes as bilayer couples. A molecular mechanism of drug-erythroeyte interactions. Proceedings of the National Academy of Sciences of the U.S.A. 71, 4457-61. SHERMAN, I. W. & TANIGOSHI, L. (1974). Glucose transport in the malarial (Plasmodium lophurae) infected erythrocyte. Journal of Protozoology 21, 603-7. STERLING, C. R., AIKAWA, M. & NUSSENZWEIG, R. S. (1972). Morphological divergence in a
mammalian malarial parasite: the fine structure of Plasmodium brasilianum. Proceedings of the Helminthological Society of Washington 39, 109-29. TRAGER, W. (1956). The intracellular position of malarial parasites. Transactions of the Royal Society of Tropical Medicine and Hygiene 50, 419-20. WILSON, R. J . M. (1974). The production of antigens by Plasmodium falciparum in vitro. International Journal for Parasitology 4, 537-47.
KEY TO THE LETTERING OF THE PLATES ap I mn ms
apical prominence multilamellar body microneme microsphere
n ps r
nucleus pseudopodium rhoptry
Parasitology, Vol. 71, Part 3
L. Jl. BANXISTlvR AND OTHERS
(Facing p. 490)
axv aaxsixxva 'H
8 W»d '\L 'PA
SHHHJ.0 axv HHXSIXXYH 'H '1
Parasitology, Vol. 71, Part 3 i
c; L. H. BAXXISTER AXD OTHERS
Red cell invasion by P. knowlesi
EXPLANATION OF PLATES PLATE 1
A. Longitudinal section of an extracellular merozoite prepared by the sieving method, showing apical prominence (above) rhoptries and nucleus. B. Parasagittal section through the anterior half of a merozoite, showing three polar rings, part of a rhoptry, micronemes and microspheres. C. Section through the merozoite pellicle, showing the cell coat composed of bristle-like filaments, and also demonstrating the relationships of the three sets of pellicular membranes, the inner two being separated in this instance by a space. D. Section through the cytostome (arrow), showing two dense bands in the cytoplasm on either side. E. Section through the double membrane-bounded body (mitochondrion). PLATE 2
A. A merozoite adhering by its cell coat to the surface of a red cell. B. A slightly later state of adhesion, showing the formation of an intimate contact between the apical end of the merozoite and the red cell surface which has become partially drawn into the apical end. Note the presence of a long narrow channel in the red cell (arrow). C. Early invagination of the red cell surface, showing attachment between the rim of the pit and the merozoite cell coat (arrows), and a subsidiary vacuole at the base of the pit. D. Later stage showing distortion of the merozoite as it passes into the vacuole. PLATE 3
A. Further invasion, showing greater merozoite distortion as it passes through the orifice of the invagination. Note the rounded shape of the forming vacuole. B. Details of the attachment zone between parasite cell coat and red cell surface. C. Newly entered merozoite, with cell coat debris visible at red cell surface. Note the persistence of a rhoptry and of microspheres. D. A merozoite inside red cell vacuole, demonstrating the continuing presence of a triplelayered pellicle. PLATE 4
A. Section showing dense microspheres which have moved to the merozoite periphery, causing its pellicle to bulge and the red cell vacuole membrane to distort. B. The microspheres appear to have ruptured, releasing granular material and initiating local invaginations of the membrane lining the red cell vacuole. C. Another example of a transforming merozoite, showing the protrusion of a pseudopodium from its base, and the presence of the apical prominence. Note the membranous debris (left). D. Early trophozoite with two vacuoles or surface depressions containing red cell cytoplasm. The nucleus is visible below.
Printed in Great Britain