The migration of human neutrophilic granulocytes in hydrated collagen lattices was studied by a combination of cinemicroscopy, and scanning and transmission electron microscopy. The basic pattern of cell migration in collagen was similar to that observed previously for these cells on inert material surfaces; i.e., a cycle of cell extension and cytoplasmic flow into the leading extension. In general, however, neutrophils in collagen were less spread than on glass or plastic surfaces. Thin lamellipodia were absent and the leading extension of the cells was often an elaborately folded pseudopodium. In addition, neutrophils migrating in collagen were never observed to have retraction fibres at the tail end of the cells, although a uropod was usually seen. In the region of the uropod, extensive blebbing of the cells often occurred, and when this happened, forward movement of the cells ceased. At the ultrastructural level, both the leading pseudopodium and the blebs at the tail of the cell were found to contain a dense cytoskeletal network from which cell organelles were excluded. Finally, the cells were found to be coated with an extensive glycocalyx, and individual collagen fibres were sometimes observed within the glycocalyx.

Several years ago, when the use of hydrated collagen lattices as a substratum for studying cell behaviour was introduced (Elsdale & Bard, 1972), it was pointed out that these lattices constitute a matrix similar to that present in connective tissue in vivo. It was found that fibroblasts in the lattices had a bipolar morphology, similar to that found in vivo, but quite different from the highly spread morphology of fibroblasts that were cultured on inert material surfaces (e.g. plastic, glass). Subsequently, in studies on fibroblast migration, it was concluded that the ruffling, leading lamelli-podium, characteristic of fibroblasts moving on material surfaces, was not typical of cell motility in vivo or in hydrated collagen lattices. Rather, cells in lattices, appeared to extend pseudopodia with fine filipodia at the leading end (Bard & Hay, 1975).

One of the important features of hydrated collagen lattices is their three-dimensional aspect. That is, cells can move within the lattices rather than being constrained to interact in two dimensions on top of a planar surface. In support of this idea, fibroblasts can align themselves along the fibrils of a hydrated collagen lattice but cannot do so if the lattice is collapsed by air-drying (Dunn & Ebendal, 1978). Similarly, fibroblasts can attach to and invade a hydrated collagen lattice in the absence of fibronectin, but fibroblast adhesion to a collapsed collagen lattice requires fibronectin (Grinnell & Bennett, 1981).

Another cell type whose migration in hydrated collagen lattices would be of particular interest is the neutrophilic granulocyte. These cells move extensively through connective tissue matrices in vivo. Although many features of neutrophil migration (Ramsey, 1972; Senda et al. 1975; Armstrong & Lackie, 1975) and chemotaxis (Zigmund, 1978) have been described in detail, almost all the studies have been carried out with cells cultured on material surfaces. Only in some of the very early literature are there descriptions of neutrophil migration in a matrix (fibrin) (e.g. see DeBruyn, 1946).

Using a combination of cinemicroscopy, and scanning and transmission electron microscopy, I have recently studied the basic features of human neutrophilic granulocyte migration in hydrated collagen lattices. The results of these studies are described in this paper.

Cells

Freshly isolated human neutrophils were routinely obtained from Dr Harry Malech, who used the following isolation protocol. Heparinized (40 i.u./ml) whole blood from normal human donors was fractionated by Ficoll-Hypaque centrifugation and the erythrocyte/neutrophil fraction was further purified by dextran sedimentation (Zakhireh & Malech, 1980). Neutrophils comprised greater than 95 % of the final dextran-separated cell population. Cell viability for all preparations was greater than 95 % as determined by Trypan blue exclusion.

Collagen gels

Hydrated collagen gels were prepared using the method of Elsdale & Bard (1972) as described previously (Grinnell & Bennett, 1982). Briefly, samples (0·05 ml) of collagen solutions (1·8 mg/ml) were spread in the centre of Permanox tissue-culture dishes and gelation was carried out for 1-2 h at 37 °C in a humidified incubator.

Incubation with cells

Isolated human neutrophils (2 × 105) were suspended in 4·0 ml of Gey’s solution containing 25 mw-Hepes buffer and foetal calf serum or bovine serum albumin, as indicated. Cell suspensions were incubated on collagen substrata, glass coverslips, or Permanox dishes at 37 °C for the times designated.

Microscopic analyses

Light microscopic observations were made using a Zeiss microscope and 40 × water-immersion objective. Time-lapse films were made with a Bolex Camera and Sage Intervalometer using 83s interval. The film used was Kodak 16 mm Plus × Reversal Movie Film and was developed commercially. Films were analysed with the aid of a model 224 Photo-Optical Data Analyzer. Prints of the time-lapse films were made from 35 mm negatives of individual frames photographed with a Honeywell Repronar slide duplicator.

Samples for scanning electron microscopic (SEM) analyses were fixed and processed as described previously (Grinnell & Bennett, 1982). Critical-point-dried specimens were coated with approx. 20 nm of gold/palladium using a rotary shadower (Ladd Research Industries) and observed and photographed in an ETEC scanning electron microscope at 10 kV (8 mm working distance to the specimens).

Samples for transmission electron microscopic (TEM) analyses were fixed and processed similarly as described previously (Grinnell & Bennett, 1982), with the following modifications.

The primary fixative was I % glutal, o·2 % tannic acid. Secondary fixation was carried out with i % OsO4 in 01 M-Na phosphate buffer (pH 6) for 20 min at 4 °C. The Epon mixture used was: 13 ml of Epon, 7 ml of NMA, 8 ml of DDSA, and 0-56 ml of DMP-30 (Ladd Research Ind.). Thin sections stained with uranyl acetate and lead citrate (Grinnell & Bennett, 1982) were observed and photographed using a Philips 300 electron microscope at 80 kV.

Materials

Calf skin collagen was a product of the Collagen Corporation. Thermanox culture dishes were obtained from Lux Scientific Corp. Gey’s solution (10 x) and foetal calf serum were obtained from GIBCo. Bovine serum albumin and Hepes buffer were obtained from Sigtna Chem. Co. Glutaraldehyde and OsO4 were obtained from Polysciences, Inc.

Neutrophil morphology and invasion of collagen lattices

In Fig. 1, neutrophils are shown attached and spread on plastic (A), glass (B) and hydrated collagen lattices (c). The extent of cell spreading was much greater on plastic and glass than on hydrated collagen. This is shown by the absence from cells on collagen of the thin lamellar regions of spread cytoplasm typical of cells on plastic or glass (A, B, open arrows). Another interesting feature was that cells on collagen lacked retraction fibres, that are characteristic of neutrophils moving on plastic or glass (A, B, closed arrows). There were, however, thin extensions of cells on collagen (c, open arrows), which in cinemicroscopic studies were found to be at the leading edge of the cells (Fig. 3, below). The surface morphology of neutrophils on glass and hydrated collagen lattices as observed by scanning electron microscopy will be shown later (Figs. 5, 6).

Fig. 1.

Appearance of neutrophils on different substrata. Leucocytes were incubated on plastic (A), glass (B), or hydrated collagen lattices (c) for 90 min in Gey’s solution supplemented with 10% foetal calf serum. Cells on plastic and glass substrata exhibited numerous lammelipodia (open arrows) and retraction fibres (closed arrows). These structures were absent on cells on collagen lattices, which were observed to be mote rounded and often had branched extensions at the leading edge (open arrows). Other details are given in Methods and Materials, ×375

Fig. 1.

Appearance of neutrophils on different substrata. Leucocytes were incubated on plastic (A), glass (B), or hydrated collagen lattices (c) for 90 min in Gey’s solution supplemented with 10% foetal calf serum. Cells on plastic and glass substrata exhibited numerous lammelipodia (open arrows) and retraction fibres (closed arrows). These structures were absent on cells on collagen lattices, which were observed to be mote rounded and often had branched extensions at the leading edge (open arrows). Other details are given in Methods and Materials, ×375

Fig. 2.

Invasion of neutrophils into hydrated collagen lattices. Leucocytes were incubated on hydrated collagen lattices for 90 min in Gey’s solution supplemented with 2 % BSA. Extensive invasion of cells into the lattices was observed as indicated in three different planes of focus (A, B and c) of the same field. In A, cell 1 was in focus on top of the lattice; in B, cell 2 was in focus at a lower plane; and in c, cell 3 was in focus at the lowest plane. Other details are given in Methods and Materials. ×525.

Fig. 2.

Invasion of neutrophils into hydrated collagen lattices. Leucocytes were incubated on hydrated collagen lattices for 90 min in Gey’s solution supplemented with 2 % BSA. Extensive invasion of cells into the lattices was observed as indicated in three different planes of focus (A, B and c) of the same field. In A, cell 1 was in focus on top of the lattice; in B, cell 2 was in focus at a lower plane; and in c, cell 3 was in focus at the lowest plane. Other details are given in Methods and Materials. ×525.

Fig. 3.

Formation of branched extensions at the leading edge of the cells. Leucocytes were incubated on hydrated collagen lattices for 60 min in medium supplemented with 2 % BSA. Subsequently, the cells were observed by time-lapse microscopy. The cell shown here was tracked for 12-2 min. A series of frames (times indicate seconds) from the middle of the sequence is shown. Initially, a branched extension (arrow) and blunt extension (arrowhead) appeared. The extensions increased in length up to 12 s and then cytoplasmic flow into both extensions occurred. At 27 s, another extension (arrow) occurred at the tip of the leading edge, which can clearly be seen to be branched at 30 s (arrow). Subsequently, cytoplasmic forward flow occurred until only a uropod was visible at 48 s (arrow). Other details are given in Methods and Materials, ×500.

Fig. 3.

Formation of branched extensions at the leading edge of the cells. Leucocytes were incubated on hydrated collagen lattices for 60 min in medium supplemented with 2 % BSA. Subsequently, the cells were observed by time-lapse microscopy. The cell shown here was tracked for 12-2 min. A series of frames (times indicate seconds) from the middle of the sequence is shown. Initially, a branched extension (arrow) and blunt extension (arrowhead) appeared. The extensions increased in length up to 12 s and then cytoplasmic flow into both extensions occurred. At 27 s, another extension (arrow) occurred at the tip of the leading edge, which can clearly be seen to be branched at 30 s (arrow). Subsequently, cytoplasmic forward flow occurred until only a uropod was visible at 48 s (arrow). Other details are given in Methods and Materials, ×500.

During a 90 min incubation, extensive invasion of neutrophils into hydrated collagen lattices occurred. The same field of cells is shown at different planes of focus in Fig. 2. On top of the lattice, cell no. 1 was in focus. Moving lower into the lattice, cell no. 2 was in focus. At a deeper level, cell no. 3, which was not visible at all at the top of the lattice, was in focus. It should be noted that many of the cells had very elongated cylindrical shapes. This is not typical of neutrophils migrating on material surfaces, but has been reported for these cells moving through a fibrin matrix (DeBruyn, 1946).

Analysis of neutrophil migration by cinemicroscopy

In order to characterize the invasion of individual neutrophils into the collagen matrix, cinemicroscopic studies were carried out. It should be pointed out that cine studies performed under these circumstances are more difficult than with cells migrating on material surfaces because the cells can move in three dimensions instead of two. Many cells moved in and out of the plane of focus during filming and could be observed only for a short time. Also, cell protuberances that were extended from above and below cells were not as visible as extensions from the cell margins. Cells that were tracked in the place of focus for at least 5 min were found to have migration rates in the range from 5 to 8 μm/min. When analysed under similar incubation conditions on plastic surfaces, the cell migration rates were found to be slightly higher, usually around 10-13 /tm/min. The rate of cell migration in collagen gels should be taken as an underestimate of the actual rate of migration because movements into the third dimension are not counted. Therefore, the differences in the rates of cell movement in collagen compared to plastic surfaces may not be significant.

The cell shown in Fig. 3 was followed for 12 ·2 min and moved 74 μm during that time. The forward movement of the cell shown in Fig. 3 cannot be observed in the photomicrographs shown because there are no fixed points of reference. That movement was occurring in the direction indicated, and the distances over which the cell migrated, were established by making tracings of the cell migration pathway from the time-lapse film. At the initial time shown (00 s), the cell was just beginning to protrude a thin branched extension at the leading edge (arrow). A blunt extension on the side of the cell (arrowhead) appeared at the same time. The thin extension reached its maximal length 12 s later, after which the cytoplasmic flow from the trailing body into the extension was observed (15 and 18 s). Before the cell had completed its forward flow, another fine extension emerged from the leading edge (27 s, arrow). The branched appearance of this extension can be seen readily at 30 s (arrow). Subsequently, over the next 18 s, the cell cytoplasm moved forward into the leading pseudopodium until only the uropod was visible (48 s, arrow).

The migration of another cell at higher magnification is shown in Fig. 4. This cell was observed for 6 ·75 min and moved 46·1 μm. At the initial time shown (000 s), blunt extensions from the cell were observed (arrowheads). The flow of cytoplasm into the extension toward the bottom of the field resulted in the formation of a broad leading edge (·021 s, arrowhead; 0·27 s). Often, collagen fibrils observed in the collagen gel seemed more phase-dense in the region of the moving neutrophils as if the fibrils were under stress (069 s, arrowheads). Studies with scanning electron microscopy, however, did not indicate any change in the collagen matrix in the vicinity of the cells (see Fig. 5, below).

Fig. 4.

Blebbing at the tail end of the cells. Experimental details are the same as for Fig. 3. The cell shown here was tracked for 6·75 min. At the initial time shown, several blunt extensions from the leading edge of the cell were observed (arrowheads). Cytoplasmic flow into the extensions resulted in the formation of a broad leading edge (021, arrowhead; 027). Collagen fibrils in the region of the cells often appeared more conspicuous (069, arrowheads). Beginning at 99 s, forward movement of the tail end of the cell ceased, and extensive blebbing of the cells began and continued for the next 18 s (arrowheads). Subsequently, the tail end of the cell continued to move forward. Other details are given in Methods and Materials, × 1500.

Fig. 4.

Blebbing at the tail end of the cells. Experimental details are the same as for Fig. 3. The cell shown here was tracked for 6·75 min. At the initial time shown, several blunt extensions from the leading edge of the cell were observed (arrowheads). Cytoplasmic flow into the extensions resulted in the formation of a broad leading edge (021, arrowhead; 027). Collagen fibrils in the region of the cells often appeared more conspicuous (069, arrowheads). Beginning at 99 s, forward movement of the tail end of the cell ceased, and extensive blebbing of the cells began and continued for the next 18 s (arrowheads). Subsequently, the tail end of the cell continued to move forward. Other details are given in Methods and Materials, × 1500.

The forward migration of the neutrophil ceased for a short time beginning at 099 s. At this time, prominent blebs were seen in the trailing portion of the cell (arrowhead). Blebbing continued for the next 18 s. When blebbing stopped, the cytoplasm from the tail portion continued to move forward (120 and 123 s). The observation of blebbing in the tail ends of these cells is a novel finding that has not been reported for neutrophils migrating on material surfaces. The appearance of the blebs was also observed by scanning and transmission electron microscopy as will be described (Figs. 5 and 7).

Fig. 5.

Appearance of neutrophils in hydrated collagen lattices as observed by SEM. Leucocytes were incubated on hydrated collagen lattices for 30 min in medium supplemented with 20 % foetal calf serum. Cells on top of or just under the top of the collagen lattice are shown. No destruction or distortion of the collagen fibrils was observed in either case. Many cells were found to have elaborately folded extensions (A, arrow), and sometimes folded extensions at one end of the cell and blebs at the other were observed (c, small arrow and large arrow, respectively). Cells with broad extensions were also evident (B, arrow). Cell blebs were observed even when the tail portions of the cells were above the collagen gels (D, arrow). Very elongated cells were also seen, both within (E) and on top of the lattices (F). Other details are given in Methods and Materials, A, D: ×6IOO;B, c: ×44OO;E: ×35OO;F: × 5250.

Fig. 5.

Appearance of neutrophils in hydrated collagen lattices as observed by SEM. Leucocytes were incubated on hydrated collagen lattices for 30 min in medium supplemented with 20 % foetal calf serum. Cells on top of or just under the top of the collagen lattice are shown. No destruction or distortion of the collagen fibrils was observed in either case. Many cells were found to have elaborately folded extensions (A, arrow), and sometimes folded extensions at one end of the cell and blebs at the other were observed (c, small arrow and large arrow, respectively). Cells with broad extensions were also evident (B, arrow). Cell blebs were observed even when the tail portions of the cells were above the collagen gels (D, arrow). Very elongated cells were also seen, both within (E) and on top of the lattices (F). Other details are given in Methods and Materials, A, D: ×6IOO;B, c: ×44OO;E: ×35OO;F: × 5250.

Fig. 6.

Appearance of neutrophils on glass substrata as observed by SEM. Experimental details are the same as for Fig. 5 except for the substratum. Cells with elaborately folded extensions into the medium above the cell were observed (A, arrow). In many cases, broad flat lammelipodia were observed (B) and retraction fibres were also seen (c, arrow). Some cells had regions of extensively spread cytoplasm but the cell bodies were round (D). Other details are given in Methods and Materials. A, B, D: × 3500; c: × 1850.

Fig. 6.

Appearance of neutrophils on glass substrata as observed by SEM. Experimental details are the same as for Fig. 5 except for the substratum. Cells with elaborately folded extensions into the medium above the cell were observed (A, arrow). In many cases, broad flat lammelipodia were observed (B) and retraction fibres were also seen (c, arrow). Some cells had regions of extensively spread cytoplasm but the cell bodies were round (D). Other details are given in Methods and Materials. A, B, D: × 3500; c: × 1850.

Fig. 7.

Appearance of neutrophils in hydrated collagen lattices as observed by TEM. Leucocytes were incubated on hydrated collagen gels for 60 min in medium supplemented with 2 % BSA. The cells shown in A and B appeared to be similar to cells observed by SEM in Fig. 5B and 5 A, respectively. At the TEM level, neutrophils were observed to have a glycocalyx around their surfaces (open arrows) and individual collagen fibrils were found often within the glycocalyx (thin arrows). Blebs were observed to contain dense cytoplasm (B, arrowheads) that was continuous with the cortical meshwork of the cells from which organelles were excluded. The dense cortical network was also readily seen at the leading end of the cells (large closed arrows). Other details are given in Methods and Materials, A: × 14600; B: × 10000.

Fig. 7.

Appearance of neutrophils in hydrated collagen lattices as observed by TEM. Leucocytes were incubated on hydrated collagen gels for 60 min in medium supplemented with 2 % BSA. The cells shown in A and B appeared to be similar to cells observed by SEM in Fig. 5B and 5 A, respectively. At the TEM level, neutrophils were observed to have a glycocalyx around their surfaces (open arrows) and individual collagen fibrils were found often within the glycocalyx (thin arrows). Blebs were observed to contain dense cytoplasm (B, arrowheads) that was continuous with the cortical meshwork of the cells from which organelles were excluded. The dense cortical network was also readily seen at the leading end of the cells (large closed arrows). Other details are given in Methods and Materials, A: × 14600; B: × 10000.

Neutrophils observed by scanning and transmission electron microscopy

In order to clarify further some of the features of neutrophil migration observed in the cinemicroscopic studies, scanning and transmission electron microscopic observations were made. Consistent with the light microscopic observations, neutrophils in collagen matrices were found to be tapered (Fig. 5A-D) or elongated (Fig. 5E, F), and generally lacking thin lamellar regions of cell spreading observed with cells on glass (Fig. 6B-D). It should be mentioned that the elongated cell morphology was not a result of cells squeezing through the collagen matrix, because it was seen for cells on top of the matrix (Fig. 5F) as well as within (Fig. 5E).

The presence of blebs was observed on many cells in the collagen matrix (Fig. 5 c, D, large arrows). These structures contained cortical cytoplasm that was continuous with the cytoplasm in the main body of the cell (Fig. 7B, arrowheads) and are probably composed of a dense cytoskeletal network, from which cell organelles are excluded (Senda et al. 1975; Boyles & Bainton, 1979). It should be pointed out that blebs on cells in collagen and retraction fibres found with cells on glass (Fig. 6 c, arrow) both occurred at the tail end of the cells. It is likely that retraction fibres are residual adhesions of the neutrophils to the substratum (Ramsey, 1972). The blebs, however, are probably not involved in cell-collagen adhesion, since they were observed even when the tail end of the cell was above the collagen matrix (Fig. 5 D, arrow).

Structural correlates of the leading extensions of the cell surface observed in cine-microscopy were also observed by SEM. In most cases, the extensions were elaborate folds of membrane (Fig. 5 A, arrow) and would appear either branched or blunt depending upon the angle from which they were viewed in light microscopy. Similar membrane folds were also extended into the medium from cells on glass substrata (Fig. 6 A, arrow) and have been described previously (Ramsey, 1972; Armstrong & Lackie, 1975). The cell in Fig. 5 c demonstrated both folds at the leading edge and blebs at the trailing end. The broader extension of the leading edge of cells in collagen (Fig. 5B) is analogous to the lamellar structure observed with cells on glass (Fig. 6 B). As observed by TEM, the folds and broad extensions at the leading edge of the cells in collagen were observed to contain a dense cortical cytoskeletal network from which cell organelles were excluded for the most part (Fig. 7 A, B, large solid arrows).

As mentioned earlier, the collagen matrix around the neutrophils as observed by SEM did not appear to be reorganized or distorted and there was no evidence of proteolysis (Fig. 5). It should be pointed out, however, that the cells were found to have an extensive glycocalyx (Fig. 7 A, open arrows), and collagen fibrils were often found within the glycocalyx (Fig. 7 A, B, small arrows).

The purpose of the studies reported in this paper was to characterize the morphological features of neutrophilic granulocytes migrating in collagen gels. The basic pattern of migration of neutrophils in collagen was similar to that previously observed for these cells migrating on material surfaces (DeBruyn, 1946; Ramsey, 1973; Armstrong & Lackie, 1975). That is, there was a cycle of cell extension and cytoplasmic flow into the extensions. The observation that the overall morphology of neutrophils moving on material surfaces or through collagen was similar, despite the differences in neutrophil adhesion to these different surfaces (see below) is consistent with the view that the series of changes in shape associated with neutrophil motility are an autochthonous feature of these cells (Keller & Cottier, 1981).

Several dissimilar aspects that were observed between neutrophil migration in collagen and on material surfaces can be explained by differences in cell adhesiveness to the substrata. Recently, it has been reported that neutrophil adhesion to collagen is much lower than to glass (Brown & Lackie, 1981). The decreased cell adhesion on collagen can explain why the cells appear less well-spread and lack retraction fibres as they move through the collagen gels (cf. Keller, Barandum, Kistler & Ploem, 1979). Also, the lack of retraction fibres associated with neutrophil migration in collagen gels supports the view that retraction fibres are not essential to migration, but are remnants of strong cell adhesions that were formed initially at the leading edge of the cells (Ramsey, 1972).

One of the most interesting features of neutrophil migration through collagen gels was the blebbing at the tail end of the cells. The presence of cortical cytoplasm in the blebs indicated that they were not ‘blister artifacts’ (Hasty & Hay, 1978). Why cells bleb is unknown. It has been suggested, however, that contraction of the cell cytoskeleton parallel to the plasma membrane causes the protrusion of blebs (Meek & Puck, 1979). In the case of neutrophils, the presence of a circumferentially organized band of cortical microfilaments along the cell membrane has been described (Boyles & Bainton, 1979), and local contraction of this band could provide the necessary force generation for bleb formation. Since forward movement of cell cytoplasm ceased during periods of blebbing, it can be suggested that contraction of the cortical band of microfilaments provides the force in the rear region of the cells for the forward flow of cytoplasm. The idea that a contraction in the rear region of the cells may be important in the forward flow of cytoplasm has long been postulated (see discussion by Armstrong & Lackie, 1975). An alternative possibility is that neutrophils contract around their entire surfaces, but only the membrane in the tail region is susceptible to blebbing.

The present findings confirm the recent observation that neutrophils have a glycocalyx on their surfaces (Hoffstein, Weissman & Pearlstein, 1981). The appearance of the glycocalyx, however, was found to be continuous and uniform, not globular. This difference might have resulted from the different staining methods used by us and the other investigators. The glycocalyx may be important in neutrophil adhesion to collagen.

Finally, although studies on neutrophil chemotaxis in collagen gels have not been carried out, they should be very informative. The ability of cells to turn in response to chemotactic stimuli might be altered in the collagen gels. For cells on glass substrata, new protrusions were generally observed at the leading edge of the cells (Zigmund, Levitsky & Kreel, 1981). Asa result, reorientation of cells in a chemotactic field generally occurred by a slow series of turning movements. Neutrophils in collagen matrices, however, were able to form extensions from anywhere along the cell margins as well as at the leading edge of the cells.

The studies reported in this paper were carried out while I was on sabbatical leave visiting the laboratory of Dr J. P. Trinkaus in the Biology Department at Yale University. I am indebted to Dr Trinkaus and his students for their assistance with many aspects of this work. Also, Dr Harry Malech in the Department of Medicine provided invaluable assistance. SEM studies were carried out with the help of Dr Alan Pooley and TEM studies were carried out with the help of Douglas Keene and John Hoffpauir. This research was supported by grants from the N.I.H., CA22451 (to J. P. Trinkaus) and CA14609.

Armstrong
,
P. B.
&
Lackie
,
J. M.
(
1975
).
Studies on intercellular invasion in vitro using rabbit peritoneal neutrophil granulocytes (PMNs)
.
J. Cell Biol
.
65
,
439
462
.
Bard
,
J. B. L.
&
Hay
,
E. D.
(
1975
).
The behavior of fibroblasts from the developing avian cornea: Morphology and movement in situ and in vivo
.
Boyles
,
J.
&
Bainton
,
D. F.
(
1979
).
Changing patterns of plasma membrane-associated filaments during the initial phases of polymorphonuclear leukocyte adherence
.
J. Cell Biol
.
82
,
347
368
.
Brown
,
A. F.
&
Lackie
,
J. M.
(
1981
).
Fibronectin and collagen inhibit cell-substratum adhesion of neutrophil granulocytes
.
Expl Cell Res
.
136
,
225
231
.
Debruyn
,
P. P. H.
(
1946
).
The amoeboid movement of the mammalian leukocyte in tissue culture
.
Anat. Rec
.
95
,
177
191
.
Dunn
,
G. A.
&
Ebendal
,
T.
(
1978
).
Contact guidance on oriented collagen gels
.
Expl Cell Res
.
111
.
475
479
.
Elsdale
,
T.
&
Bard
,
J.
(
1972
).
Collagen substrata for studies on cell behavior
,
J. Cell Biol
.
54
,
626
637
.
Grinnell
,
F.
&
Bennett
,
M. H.
(
1981
).
Fibroblast adhesion on collagen substrata in the presence and absence of plasma fibronectin
.
J. Cell Sci
.
48
,
19
34
.
Grinnell
,
F.
&
Bennett
,
M. H.
(
1982
).
Ultrastructural studies of cell-collagen interactions
.
Meth. Enzym
.
82
,
535
544
.
Hasty
,
D. L.
&
Hay
,
E. D.
(
1978
).
Freeze-fracture studies of the developing cell surface
.
J. Cell Biol
.
78
,
756
768
.
Hoffstein
,
S. T.
,
Weissman
,
G.
&
Pearlstein
,
E.
(
1981
).
Fibronectin is a component of the surface coat of human neutrophils
.
J. Cell Sci
.
50
,
315
327
.
Keller
,
H. U.
,
Barandun
,
S.
,
Kistler
,
P.
&
Ploem
,
J. S.
(
1979
).
Locomotion and adhesion of neutrophil granulocytes. Effects of albumin, fibrinogen, and gamma globulins studied by reflection contrast microscopy
.
Expl Cell Res
.
122
,
351
362
.
Keller
,
H. U.
&
Cottier
,
H.
(
1981
).
Crawling-like movementsand polarization in non-adherent leukocytes
.
Cell Biol. Int. Rep
.
5
,
3
7
.
Meek
,
W. D.
&
Puck
,
T. T.
(
1979
).
Role of the microfibrillar system in knob action of transformed cells
.
J. supramolec. Struct
.
12
,
335
354
.
Ramsey
,
W. S.
(
1972
).
Locomotion of human polymorphonuclear leukocytes
.
Expl Cell Res
.
72
,
489
501
.
Senda
,
N.
,
Tamura
,
H.
,
Shibata
,
N.
,
Yoshitake
,
J.
,
Kondo
,
K.
&
Tanaka
,
K.
(
1975
).
The mechanism of the movement of leukocytes
.
Expl Cell Res
.
91
,
393
407
.
Zakhireh
,
B.
&
Malech
,
H. L.
(
1980
).
The effect of colchicine and vinblastin on the chemotactic response of human monocytes
.
J. Immun
.
125
,
2143
2153
.
Zigmund
,
S. H.
(
1978
).
Chemotaxis by polymorphonuclear leukocytes
.
J. Cell Biol
.
77
,
269
287
.
Zigmund
,
S. H.
,
Levitsky
,
H. I.
&
Kreel
,
B. J.
(
1981
).
Cell polarity: An examination of its behavioural expression and its consequences for polymorphonuclear leukocyte chemotaxis
.
J. Cell Biol
.
89
,
585
592
.