Cell Research 72 (1972) 489-501
W. S. RAMSEY Department
SUMMARY The locomotion of human polymorphonuclear leucocytes (PMN) cultured on glass slides in the presence and absence of chemotactic attractant was studied using high magnification cinemicrography. Views from above and the side indicate that PMN move by extending onto the substratum a lamellipodium into which the cellular contents flow. These observations suggest that lamellipodia are extended by the flattening of a fold of membrane which moves out from the cell body. The tail and retraction fibers invariably observed at the rear of moving PMN are derived from lamellipodia which had been extended but into which the cellular contents did not flow. Thus the original adhesions of lamellipodia to the substratum are maintained until movement of the cell eventually breaks the retraction fibers. There is no evidence for gel-sol transformations in these cells. Cytoplasmic flow is continuous in moving cells and occurs throughout the entire width of the cell. Lamellinodia are randomly produced on all sides of a cell. Chemotactic movement is characterized by preferential flow -of the cellular contents into lamellipodia on the side nearest to the attractant.
Our present knowledge of mechanisms of leucocyte locomotion stems primarily from the work of two laboratories. De Bruyn [S, 6, 7, 161, in broad agreement with earlier workers, described motility of leucocytes covered by a clot as essentially similar to that of large free-living amoebae as described by Mast . This conclusion was based on the similarity of shape of moving leucocytes and amoebae and the appearance of gel-sol transformations in leucocytes. More specifically, De Bruyn  observed that granules which were adjacent to the cell membrane in lateral protuberances sometimes did not flow but remained stationary with respect to the outside of the cell. However, these granules could subsequently begin to participate in cytoplasmic streaming, suggesting a gel-sol transformation. It was also found that “sometimes granules located along the non-protuberant lateral outline of the cell also re32-721806
Although this work is main stationary”. widely quoted as demonstrating the amoeboid nature of leucocyte movement, it was noted by De Bruyn that observations on leucocytes could not be made with the same precision as those on amoebae, that it was not possible to follow the movement of any one granule for a sufficient length of time, and that it was not possible to correlate protoplasmic flow and cell motility. French workers have made extensive use of cinemicrography in studying motility of leucocytes in tissue culture, generally in spleen fragments cultured under dialysis membranes [9, 21-251. In general, this group agrees with the description of leucocyte motility as amoeboid; but they emphasize the role of the flattened hyaloplasmic membrane or veil found on the anterior end of a moving cell (I propose to call this flattened extension a lamellipodium [4, 281). These Exptl
Cell Res 72
W. S. Ramsey
workers observed that displacement of a cell occurs when a lamellipodium is extended over the substratum and the cellular contents flow into the lamellipodium. The production of true amoeboid pseudopods was also described . Robineaux later reached the opinion, however, that “It is doubtful whether the streaming plays a role in the locomotion of the cell”, presumably because streaming was not continuous . At this time it was suggested that the lamellipodium might cause motility by pulling the cell across the substratum by some mechanism involving the propagation of ruffles, as has been suggested for the action of lamellipodia in cultured motile fibroblasts. The appearance of a tail which culminated in a number of long filaments at the rear of a moving leucocyte has long been mentioned in the literature, although its significance has been in dispute. Robineaux  found that the tail of neutrophils was very adhesive for the substratum and regarded this area as a zone of reversible adhesiveness. Greenwood [I 11, while studying motility in eosinophils, found that the tail is adhesive for bacteria, platelets, and other particles. Studies on the interactions of lymphocytes and other cells [ 17, 181 led to the suggestion that the lymphocyte tail is a permanent organelle which enables the cell to adhere to and interact with other cells. These uncertainties are obviously in need of resolution. It was reported by Dixon & McCutcheon  and recently confirmed  that the presence of a chemotactic attractant affects only the direction of movement of human polymorphonuclear leucocytes (PMN) and is without effect on the cell speed. Detailed observations of locomotion during chemotaxis, however, have not been presented. It is clear that detailed observation of the locomotion of individual leucocytes is necessary if we are to approach an underExptl Cell Res 72
standing of their mechanism of locomotion. The purpose of this paper is to describe such a study, using motile cultured PMN in the presence and absence of attractant, and high magnification cinemicrography.
Growth of attractant bacteria Staphylococcus albus wereused as the attractant. They were grown at 37°C streaked on brain heart infusion agar (BBL).
Leucocyte cultures In all experiments, blood used was from one healthy adult man. PMN were collected and isolated as described by Harris  with certain modifications . In brief, a drop of blood was placed on a coverslip and allowed to clot while incubated [email protected]
min at 37°C in a moist chamber. The clot was then washed away with a buffered saline solution leaving PMN clinging to the glass surface. Wright stains of these preparations indicated at least 95 % of the cells remaining were PMN. In chemotactic preparations, a streak of S. albus was dried on a slide and pieces of a broken No. 1 coverslip were placed on either side of the bacterial streak and the area flooded with normal rabbit serum. A coverslip with adhering PMN was then inverted over the supports, forming a chamber, which was then sealed with melted paraffin. In some experiments, the procedure was modified by using Gev solution 1101 at DH 7.2 ulus 2% bovine serum albbmin both fo; washing off-the clot and as the incubation medium . The movement of cells appeared similar using either method. In all cases, cultures were incubated and photographed in a 37°C constant temperature room.
Side views To examine moving PMN from the side, the method of Harris [12, 131 for obtaining side views of fibroblasts was used. Blood was allowed to clot on 10 pm diameter glass fibers stretched across a microscope slide. After incubation the clot was washed from the fibers using Gey solution. A chamber was formed using a 2 mm thick aluminum slide in which a 20 mm diameter hole had been drilled. One side of the hole was covered by a coverslip attached with melted paraffin. The chamber was then filled with Gey solution and the glass fibers with adhering PMN were streched across the top of the chamber. A coverslip was placed over the top of the chamber, sealed with silicon stopcock grease, and the glass fibers brought up against the upper coverslip by pulling on the fiber ends which extended beyond the chamber. When chemotaxis was desired, this coverslip was streaked with S. albus. Using this arrangement and a 6 mm
Leucocyte locomotion condenser lens, leucocytes clinging to the sides of the glass fibers could be observed using an oil-immersion objective.
Photography AZeissphasecontrastmicroscope(Standard Universal) was used with a x 100, N.A. 1.30 oil-immersion objective. Films were made using a x 5 eyepiece, Sage time-lapse apparatus, 16 mm Bolex camera, and Kodak plus X Reversal film. The frame rate was 120 frames/mm with 0.20 set exuosure. Films were examined ‘with a Kodak Data Analyzer projector (L-W Photo Inc., Van Nuys, Calif.). Streak photographs were made using a Wild 35 mm still camera, Kodak Panatomic-X film, and 1 set exposure.
Three-dimensionphotography In order to demonstrate the appearance of cells in culture, a set of stereoscopic micrographs was made. PMN cultures were prepared with cells on a coverslip in the absence of chemotactic effector. Two photographs were made of the same field using a method suggested by Dr T. Betchaku. In the first photograph, the left side of the slide was raised so that the slide was tilted 6” with reference to the horizontal. This side was then returned to the stage, the right side similarly elevated, and a second photograph made. A Zeiss phase contrast microscope (Standard GFL) was used with a x 5 eyepiece and, in order to provide adequate working distance between objective lens and tilted culture chamber, a x 40, N.A. 0.75 waterimmersion phase objective. A 6 mm condenser lens was used. Photography was as above. Preparations were incubated at room temperature for 30 min after collection of the cells to ensure normal motility, and then photographed in a constant temperature room at 4°C. This low temperature prevented celluiar movement during the approx. 1.5 min required between exposures to advance the film, tilt the slide, find the correct field, and focus. The appearance of PMN at 4°C was like that at 37°C except that the cells did not move.
Terminology The term lamellipodia, introduced by Abercrombie et al.  to name flattened extensions of cells used in locomotion, will be used for the flattened structures found on the sides of motile PMN which have been variously called veils, hyaloplasmic membranes, or undulating membranes. The side of the PMN adjacent to the substratum will be called the lower sihe, and that exposed to the medium the upper. In a moving PMfi the advancing end will be-called the front end, the other end the rear.
RESULTS Observations of PMN locomotion
When PMN were observed moving on a glass surface without being covered with a
clot it became readily apparent that motility under these conditions involves alternation between being highly spread over the glass surface and protruding up into the medium with only a relatively small area against the glass. Figs l-10 show single frames at 8 set intervals of a film of a cell moving toward the right in response to chemotactic attractam. In figs l-5 the cell is relatively flat against the substratum. The edge of the cell advances while the cell is widely spread. In figs 6-10 the rear end of the cell begins to move forward, and in the process the cell begins to protrude away from the substratum. Thus, two distinguishable elements are involved in PMN motility: spreading of the front of the cell over the substratum, followed by forward movement of the rear end of the cell accompanied by protrusion of the cell into the medium. The phase halos in figs 5-10 were caused by the thickening optical section of the cell as it protruded into the medium. This and other protuberances are readily seen by focusing up and down. Further demonstration of this point may be found in fig. 11, which is a stereoscopic pair of photographs of PMN on a glass surface. Some cells are widely spread over the surface, whereas others protrude into the medium. Examination of PMN immediately after isolation revealed all cells to be widely spread over the glass surface. It was only after the cells began to move actively about and actually change position, within 10 min, that the typical appearance described above developed. Obviously, the best optical conditions for observing cellular features occur when the cell is widely spread over the substratum. Figs 12-23 show single frames printed 2 set between frames from a film of a highly spread cell which was just in the process of beginning active motility. The cell was moving toward chemotactic attractant at Exptl
Ceil Res 72
W. S. Ramsey
Figs I-10. Actively moving PMN in culture 3 h. Frames taken from a 16 mm film, original rate 0.5 set between exposures, here printed at 8 set between frames. Mark represents 10 ,um. Arrow 1 indicates a lamellipodium which was extended and then developed into a retraction fiber as the cell moved. Arrow 2 indicates a lamellipodium into which the cellular contents flowed as the cell move.
the bottom of the field. In fig. 12, an extended lamellipodium may be seen at one end of the cell (arrow 1). In figs 13-15, cytoplasm has flowed into the lamellipodium as indicated by the movement of granules. A new Exptl Cell Res 72
lamellipodium is then extended on the front of the cell in figs 15-18 and in figs 19-23 cytoplasm again flows into this structure. In fig. 23, the phase halo around the front of the cell suggests increased thickness in this
region. There is also a net movement of cytoplasm from the rear to the front of the cell. The effect is to cause displacement of the cell toward the bottom of the field. In order to get a better understanding of the mechanism of lamellipodium extension, enlargements were made of the area of one emerging lamellipodium, starting one frame (0.5 set) before fig. 14 and proceeding at 0.5 set intervals to two frames before fig. 17. These enlargements are reproduced in figs 24-35. In figs 28-32, an optically dense, horseshoe-shaped curve appears and moves out into the extending lamellipodium. This curve was formed by a combination of an optically dense region moving along the edge of the lamellipodium with another similar region moving out from the cell body. In figs 31-35, the horseshoe-shaped curve disappears while at the same time the lamellipodium extends. This may be interpreted as the movement of a fold of membrane which moves over the upper surface of the cell out over the lamellipodium toward its edge where it is flattened, thus causing extension of the lamellipodium over the substratum. In this series from figs 24-35, the actual extension covers a distance of 3 pm. Thus the height of the fold would be expected to be about 3 pm. Extensions of the cell surface which wave about in the medium and form phagocytotic vacuoles were often observed in cells which were not actively displacing. These protrusions appear similar to those: found in macrophages . In contrast, lamellipodia extended against the substratum by actively displacing cells were not observed to lift from the substratum or to ruffle. Folds extending from the cell body to the edge of the lamellipodium were sometimes observed (fig. 5). Such folds extended perpendicular to the edge of the lamellipodium and did not appear to move.
Fig. II a, 6. Stereoscopic view of PMN cultured 35 min on a glass substratum. Mark represents 50 pm. Note some cells flattened against the substratum and some protruding into the medium.
Effect of chemotaxis on the locomotory mechanism
Low power cinemicrography of cells indicates that movement in the presence of chemotactic attractant differs from that in the absence of attractant only in that chemotactic movement is directional . When high-magnification cinemicrographs were examined, it became apparent that the mechanism of directional movement is related to the process of cytoplasmic flow rather than to lamellipodium extension. In both the presence and absence of chemotactic attractant, lamellipodia extend apparently at random, on all sides of the cell. In cells responding to chemotactic attractant, however, the cell cytoplasm tends to flow preferentially into lamellipodia which are on the side facing the attractant. In the absence of attractant, cytoplasm flows into the various lamellipodia, apparently at random, although the previous polarity of the cell is often maintained for long periods of time.
Source of retraction fibers and tail
It has long been noted that the moving PMN has a tail which terminates in long fibers. These appear like retraction fibers observed in fibroblasts . Cinemicrography of motile Exptl Cell Res 72
W. S. Ramsey
Cell Res 72
Leucocyte locomotion 495 PMN indicates that the retraction fibers are formed by lamellipodia which had been extended but into which the cellular contents did not flow. The formation of retraction fibers indicates that such lamellipodia are adhesive to the substratum. Subsequent movement of the cell causes the trailing lamellipodium to stretch out from the adhesive point and form a retraction fiber, one end of which is attached to the substratum and the other end to the moving cell at the region of the tail. A sequence showing this process is illustrated in figs l-10. Arrow 1 indicates a lamellipodium which was extended and abandoned as the cell contents flowed into a lamellipodium at a slightly different location. The location of the edge of the lamellipodium remains static, however, presumably because of its adhesion to the substratum. As expected, a retraction fiber forms at this site. Intimations of a similar occurrence may be seen in figs 12-23, arrow 2. In this case a lamellipodium is extended and, although granules slowly flow into this structure, it is unused in the sense that the cell does not move in that direction; the lamellipodium is consequently slowly displaced toward the rear of the cell as cytoplasmic flow into another lamellipodium causes the cell to move. In figs 12-23 one should also note that displacement of the front of the cell is not accompanied by displacement of the rear margin of the cell. Rather the cell contents (as indicated by granules and vacuoles) at the rear move forward leaving an apparently empty cell envelope. It appears that once a lamellipodium becomes adherent to the substratum the adhesion is not reversible and is lost only when retraction fibers break or,
after being greatly extended, pull loose from the substratum. The so-called tail appears to be only a collection of the proximal ends of retraction fibers. Side views Moving PMN were photographed from the side in order to test the observations that cells periodically protrude from the substratum and that the mechanism of cell motility involves extension of lamellipodia into which the cell contents flow. Since PMN were photographed moving on the side of glass fibers, they were not in the same configuration with respect to gravity as cells as usually observed, although one would not imagine this to be a serious problem as cell movement appears similar whether the cells are on the top of a slide or on the bottom of a coverslip. Single frames from a film of such cells, printed at an interval of 4 set between frames, are shown in figs 36-48. The cell originally protruded away from the fiber (fig. 36). In figs 37-40 a thin structure is extended from the base of the cell out along the substratum toward the right. In figs 41-45 the cell contents flow into this extension and the cell is thereby displaced toward the right. It seems reasonable to conclude that the structure extended from the side of a moving cell in figs 36-48 corresponds to the lamellipodia shown in figs 12-23 and 24-35. Cells viewed from the side were found to change their shapes continually, even when not actually displacing. Indeed, waves were even observed to be propagated at the upper surface and to move toward the base of the cell (figs 3843). Cells were also observed to bend and oscillate without actual displacement. In
Figs 12-23. Spread PMN beginning active movement toward chemotactic attractant at bottom of field, cells in culture 18 min. Framestakenfrom a 16 mm film, originalrate 0.5 set betweenframes,hereprinted at 2 set between frames. Mark represents 10 pm. Arrow 1, extensionof lamellipodiumand eventual flowing of cytoplasmic contents into lamellipodium with movement of front of cell. Arrow 2, lamellipodium which was extended but not used in cell displacement. Exptl
W. S. RamseJj
Figs 24-35. Extension of lamellipodium. Fig. 24 is an enlargement of the frame just prior to the frame in fig. 14. Figs 25-35 are subsequent frames, 0.5 set between frames. Mark represents IO pm. Arrow, (figs 28-31) indicates formation of horseshoe-shaped region.
these casesthe upper end was never observed to touch the substratum. Cytoplasmic flow The most casual observation of a living PMN gives a striking impression of incessant cytoExptl
Cell Res 72
plasmic flow, as indicated by the movement of granules, vacuoles, and nucleus. In a locomotive PMN the direction of cytoplasmic flow is always into the lamellipodium when the cell is actually moving although this is always preceded and followed by apparently
Figs 36-48. from a 16 represents propagated
Side view of a cell moving toward chemotactic attractant. Cells in culture 2 h. Frames taken mm film, original rate 0.5 set between exposures, here printed at 4 set between frames. Mark 10 pm. Arrow 1 indicates lamellipodium extended along glass fiber. Arrow 2 indicates wave from upper portion of cell. Exptl
W. S. Ramsey
Figs 49-52. Streak photographs of PMN. Marks represent 10 Pm. Figs 49-51, actively displacing cells in culture 30-60 min. Fig. 52, cell 5 min after isolation with little cytoplasmic flow.
random flow. The flow in PMN which are not actually displacing is in any direction and frequent changes of direction occur. In all observations of displacing PMN there was no evidence of an area containing flowing granules adjacent to an area in which granules did not flow, as would be expected if there were a gel-sol transformation, as in amoebae. There was no evidence therefore, for a cylinder of gel containing flowing sol. On the contrary, in moving PMN, granules Exptl Cell Res 72
were invariably observed to move throughout the entire width of the cell. The only obstructions to granule movement appear to be the nucleus and vacuoles which, while moving themselves, sometimes cause diversion of the flow of granules. Long exposure photographs have been used to demonstrate the pattern of granule flow in amoebae [l]. In these organisms it is easy to demonstrate channels of moving granules adjacent to regions of immobile particles. When similar
photographs were made of moving PMN (figs 49-51) no regions of immobile granules adjacent to a flowing region were observed. Every frame of the portion of film in figs l-10 was enlarged, printed, and carefully examined on this point (45 frames, 0.5 set between frames). No evidence which would suggest a gel-sol transformation was found. Similar conclusions may be reached after examination of the flow of granules in figs 24-35. Fig. 52 is a streak photograph of a cell which had just been isolated and which was greatly spread with little cytoplasmic flow apparent. In such cells individual granules sometimes make saltatory motions and regions of flow appear adjacent to immobile regions. When such a cell begins normal displacement, however, cytoplasmic flow occurs throughout the width of the cell. DISCUSSION The formation of a lamellipodium on the side of a leucocyte and subsequent flow of cytoplasmic contents into this structure appears to correspond to the side views reported here of extension of a flattened protrusion along the substratum followed by cytoplasmic flow into it. The extension of such a lamellipodium appears to occur through the flattening of waves which proceed from the cell body out toward the margin. This conclusion is supported by views from the side, which also reveal waves traveling down the cell surface from the uppermost region to that nearest the substratum. Lamellipodia are produced on all sides of a moving cell. When the cell is travelling in a relatively straight line (as during chemotaxis), however, the cytoplasmic contents tend to flow preferentially into lamellipodia on one side. The lamellipodium appears therefore to be an integral aprt of the locomotory system of the PMN.
The PMN lamellipodium differs from that observed in time-lapse films of moving fibroblasts in several important respects [2, 3, 13, 15, 281. In PMN, the lamellipodia are relatively temporary structures which are extended, adhere to the substratum, and are then filled with cellular contents and cease to be lamellipodia. In fibroblasts, the lamellipodia are much less temporary entities which adhere to and move over the substratum while pulling the cell body behind them, all the time remaining flattened. There is no cytoplasmic flow in fibroblasts. PMN lamellipodia are extended at an average speed of about 20 pm/min; the analogous figure for fibroblast lamellipodia displacement would be about 0.60 pm/min, although fluctuations of the margin move at 5 pm/min . The ruffle in a fibroblast lamellipodium is formed by the uplift of the margin and is generally propagated back toward the cell body away from the margin. Ruffling does not occur in PMN. The margins of PMN lamellipodia were not observed to lift from the substratum and when folds do occur they are perpendicular to the lamellipodium margin or are generated well back of the margin and move from the cell body out toward the margin. In non-displacing cells, surface extensions are often seen to wave about in the medium and to form phagocytic vacuoles. The relationship between these extensions and the lamellipodia used in cell locomotion is not clear. The present study suggests that the tail and retraction fibers should be regarded as somewhat incidental features of PMN motility, results rather than causes of cell displacement. The retraction fibers observed to extend from the tail to substratum are apparently greatly stretched cylinders of cell envelope extending from the moving cell to the original sites of adhesion of lamellipodia to Expti
W. S. Ramsey
the substratum. The tail is an accumulation of proximal ends of retraction fibers which collect at the most distal portion of the moving cell. PMN lamellipodia resemble those of fibroblasts to the extent that both provide points of adhesion to the substratum and therefore may serve as the source of retraction fibers. Previously reported experiments  showed that a PMN moving toward chemotactic attractant held in a micromanipulator reversed its direction of movement almost immediately after the location of attractant was shifted to behind the cell. This change in direction of movement was accompanied by appearance of a lamellipodium at the region of the original tail, followed by the eventual appearance of a tail in the region of the original lamellipodium. It seems likely that the material in the original tail was converted into lamellipodium. The experiments were performed at too low a magnification, however, to provide resolution sufficient to distinguish between this interpretation and the movement of the original lamellipodium over the cell to the region of the original tail. In any case, it would appear no longer necessary to postulate an area of reversible adhesions which would alternately adhere and deadhere to the substratum during PMN locomotion . Rather, one need postulate only areas of nonreversible adhesion of the lower sides of lamellipodia. The inhibiting effect of these adhesions on motility would be overcome as movement of the cell created sufficient tension to break the retraction fibers or the adhesions. Such a mechanism implies the continual synthesis of new lamellipodial material in moving cells as material would be left behind as the cell displaced. It remains to be seen if this analysis of the origin of retraction fibers and their role in PMN locomotion is of general application in leucocyte motility. Careful examination of films of moving Expil
Cell Res 72
PMN does not support the suggestion that gel-sol transformations are involved. Cytoplasmic flow appeared to occur throughout the entire width of the cell. The direction of flow was always toward the lamellipodium being filled as the cell displaced. At other times, however, the flow was in various directions. Cytoplasmic flow occurred continuously in all living PMN whether the cell was displacing or not. The only exception to this statement are cells which have just been collected and are greatly spread against the glass surface. Under these conditions one sometimes observes movements of granules in streams adjacent to immobile granules when cytoplasmic flow is just beginning. It seems possible that earlier observations suggesting a gel-sol structure [5, 6, 7, 161 were based on similar cells, or on cells under less desirable optical conditions, such as under a clot. With phase microscopy the image depends on the thickness and refractive index of the structure being observed. A gel-sol interface would probably involve a change in refractive index. On the other hand, the appearance of a structure does not necessarily indicate a gel-sol interface. A better way to obtain evidence for the gel or sol state is to observe the movement of particles. These considerations would lead one to question the contention [22, 231 that gel-sol structures have been observed in cells and lamellipodia in the absence of observations on particle movements. Of the modes of PMN movement discussed on p. 485-486, only the type in which the extension of a lamellipodium is followed by cytoplasmic flow into this organelle has been found to occur. This process is accompanied by alternate flattening of the cell against the substratum and protrusion into the medium. No amoeboid-type pseudopods were observed, nor was evidence obtained for a gel-sol structure. PMN lamellipodia were
not found blasts.
like those of fibro-
I wish to thank Dr Trinkaus for suggesting the problem and for advice and encouragement during the research. Discussions with Dr Albert K. Harris are also aratefullv acknowledged. This research was supported by a grant from the National Science Foundation to Dr J. P. Trinkaus (GB 21240). The author was supported by USPHS training grant HD 00032 to the Department of Biology, Yale University.
REFERENCES 1. Abe, T H, Primitive motile systems in cell biology (ed R D Allen & N Kamiya) p. 221. Academic Press, New York (1964). 2. Abercrombie, M & Ambrose, E J, Exptl cell res 15 (1958) 332. 3. Abercrombie, M, Heaysman, J E M & Pegrum, S M, Exptl cell res 59 (1970) 393. 4. - Ibid 60 (1970) 437. 5. DeBruyn, P P H, Anat ret 89 (1944) 43. 6. - Ibid 93 (1945) 295. 7. -Ibid 95 (1946) 177. 8. Dixon, H M & McCutcheon, M, Proc sot exptl biol med 34 (1936) 173. 9. Frederic, J & Robineaux, R, J physiol 43 (1951)
12. Harris, A K. J cell biol 43 (1969) 165a. 13. - The role.of adhesion and the’cytoskeleton in fibroblast locomotion. PhD dissertation. Yale University (1971). 14. Harris, H, J path01 bact 66 (1953) 135. 15. Ingram, V M, Nature 222 (1969) 641. 16. Kass, L & DeBruyn, P P H, Anat ret 159 (1967) 115. 17. McFarland, W & Heilman, D H, Nature 205 (1965) 887. 18. McFarland, W, Heilman, D H & Moorhead, J F, J exptl med 124 (1966) 851. 19. Mast, S 0, J morph01 physiol 41 (1926) 347. 20. Ramsev. W S. Exntl cell res 70 (1972) 129. 21. Robin&&x, R, Rev hematol 9 (1954)‘364. 22. - Primitive motile systems in cell biology (ed R D Allen & N Kamiya) p. 351. Academic Press, New York (1964). 23. Robineaux, R & Pinet, J, Ciba Foundation symposium on cellular aspects of immunity (ed G E W Wolstenholme & M O’Connor). Little, Brown & Co, Boston (1960). 24. Robineaux, R, Pinet, J & Kourilsky, R, Compt rend sot biol 156 (1962) 1025. 25. - Nouv rev franc hematol 2 (1962) 797. 26. Snyderman, R, Gewurz, H & Mergenhagen, S E, J exptl med 128 (1968) 259. 27. Taylor, A C & Robbins, E, Dev bio17 (1963) 660. 28. Trinkaus, J P, Betchaku, T & Krulikowski, L S, Exptl cell res 64 (1971) 291. 29. Warfel, AH & Elberg, S S, Science 170 (1970) 446.
10. Gey, G 0 & Gey, K K, Am j cancer 27 (1936) 45. 11. Greenwood, B, Brit j dermatol, suppl. 3 (1969) 36.
Received August 30, 1971
Cell Res 72