J. Cell Sci. 69, 167-178 (1984) 167Printed in Great Britain © The Company of Biologists Limited 1984

QUICK-FREEZE, DEEP-ETCH ROTARY REPLICATIONOF TRYPANOSOMA CRUZI AND HERPETOMONAS

MEGASELIAE

THAIS SOUTO-PADR6N, WANDERLEY DE SOUZAInstitute de Biofisica, Universidade Federal do Rio de Janeiro, Ilha do Fundao, 21941,Rio de Janeiro, Brasil

AND JOHN E. HEUSERDepartment of Physiology and Biophysics, Washington University School of Medicine,St Louis, Missouri 63110, U.SA.

SUMMARY

The fine structure of epimastigotes of Trypanosoma cruzi and promastigotes of Herpetomonasmegaseliae was analysed in replicas of quick-frozen, freeze-fractured, deeply etched and rotary-replicated cells. Using control cells and cells treated with Triton X-100 before glutaraldehydefixation, images were obtained that showed connections of the sub-pellicular microtubules with eachother, with the plasma membrane, and with the endoplasmic reticulum. Images were also obtainedthat showed the DNA network in the kinetoplast. Filamentous structures were found to connect thekinetoplast to the basal body, and to connect the main basal body to the accessory one. In addition,deep-etch images of detergent-extracted flagella display dynein arm substructure and the filamen-tous architecture of the paraxial structures.

INTRODUCTIONThe ultrastructure of trypanosomes, in particular Trypanosoma cruzi, has

previously been analysed by transmission electron microscopy of thin-sectioned andfreeze-fractured organisms (Martinez-Palomo, de Souza & Gonzales-Robles, 1976;de Souza, Martinez-Palomo & Gonzales-Robles, 1978). In the present paper weextend these reports, by describing the structure of epimastogotes of T. cruzi andpromastigotes of Herpetomonas magaseliae as seen in replicas of quick-frozen,freeze-fractured and deeply etched organisms. With this new preparative technique,cells can be frozen directly from life or after brief aldehyde fixation, thus avoidingpotentially harmful influences of OsCU fixation, chemical dehydration and plasticembedding. In addition, previous studies with this new technique have demonstratedthat it provides novel three-dimensional images of membrane surfaces as well as ofintracellular filaments and organelles (Heuser & Salpeter, 1979; Heuser & Kirschner,1980; Hirokawa & Heuser, 1981; Roof & Heuser, 1982). In the present report, wehave used this new preparative technique to demonstrate membrane—microtubuleinteractions in whole cells, and axonemal and kinetoplast architecture in detergent-extracted cells.

168 T. Souto-Padron, W. de Souza andj. E. Heuser

MATERIALS AND METHODS

Microorganisms

Epimastigotes of Trypanosoma cruzi, Y strain, were cultivated for 3-4 days at 28 °C in Warren's(1960) medium. Promastigotes of Herpetomonas megaseliae were cultivated in a complex medium(Roitman, Roitman & Azevedo, 1972) for 48 h at 28°C.

Fixation

The cells were collected by centrifugation and fixed in 2'5 % glutaraldehyde in (H M-cacodylatebuffer (pH7-2) for 1 h. In some experiments Triton X-100 (final concentration, O'05-O'l %) orsaponin (0-l %) was added to the fixative solution.

Quick-freezing, freeze-fracture and deep-etching

Small drops of the pellet containing the parasites were supported on 3 mmX3 mmX0-8mmcushions of rat lung and washed with distilled water containing 15 % (v/v) methanol. Then theywere quick-frozen with a liquid helium-cooled copper-block machine (Heuser et al. 1979), andfreeze-etched in a Balzer's apparatus as previously described (Heuser & Salpeter, 1979). To ensurethat the exposed areas would be optimally frozen, the depth of fracturing was reduced to aminimum: the plane of cleavage entered the sample only a few micrometres beneath its surface andlarge regions of the surface were not cleaved at all. To limit the depth of exposure around theorganelles to (H lim, etching was minimized: the specimen temperature was kept at —105°C andthe etching time was about 2-3 min. Replicas were made by rotary shadowing with a mixture ofplatinum and carbon, delivered from an electron beam gun mounted at an angle of 24 °. To minimizethe thickness of the carbon backing-film, evaporation of carbon was reduced to 3-5 s. Replicas werecleaned in full-strength chromate cleaning solution (Fisher Scientific Co.), washed in distilled waterand transferred to 75 mesh grids coated with Formvar plus carbon. Examination was carried outwith a Jeol 200CX electron microscope operated at 100 kV. The high accelerating voltage was usedto reduce heating of the replica (and hence to minimize recrystallization of platinum), and also toreduce contrast in the final photographic negatives. Stereo pairs were obtained by tilting the samplethrough ± 10 °. Micrographs were examined in negative contrast by photographically reversing thembefore printing, to make the platinum deposits look white and the background look dark; aspreviously shown (Heuser et al. 1979), this contrast-reversal enhances the three-dimensionalappearance of the images.

RESULTS

The main components of trypanosomes can be identified in deep-etch images such

as Fig. 1, which shows a cross-fracture through the central portion of an epimastigote

form of T. cruzi. Immediately beneath the plasma membrane are found sub-pellicular

microtubules, distributed in the same pattern as is seen in thin sections. What is not

apparent in thin sections, however, is that the microtubules lie in a narrow peripheral

zone of cytoplasm, which appears to be devoid of the fine granular material that

otherwise fills the rest of the cell. This uniformly dispersed granular material, a

characteristic feature of all deep-etched cells, is thought to be soluble cytoplasmic

protein (Hirokawa & Heuser, 1981; Hirokawa, Tilney, Fujiwara & Heuser, 1982).

With the current technique, images are also obtained of the more traditional frac-

ture faces of the cell membrane; these differ only slightly from our previously

published freeze-fracture images (Martinez-PalomoeJ al. 1976; de Souzaet al. 1978),

due primarily to the rotary replication used here. This highlights the difference

between the high density of P-face particles in the membrane of the cell body and their

Ultrastructure o/Trypanosoma and Herpetomonas 169

Fig. 1. Survey view of an epimastigote form of T. cruzi. Structures such as the nucleus(/i), nucleolus (nu) and the sub-pellicular microtubules (arrows) can be seen. X31 500.

virtual absence from the membrane of the flagella (Fig. 2 versus 3). Larger expansesof fractured membrane such as are shown Fig. 3 illustrate the fact that the particlesin the plasma membrane tend to be partitioned into longitudinal clusters that corres-pond roughly to the spacing of the underlying microtubules. Also the 'cytostome' isreadily apparent in fields such as are shown in Fig. 2, where it appears as a region closeto the flagellar pocket that contains few particles but is surrounded by a single stringof intramembrane particles. Finally, the flagellar membrane itself displays a typicalciliary necklace, plus the occasional group of intramembrane particles, which we have

T. Souto-Padron, W. de Souza andjf. E. Heuser

Figs2r-3

Ultrastructure o/Trypanosoma and Herpetomonas 171

previously proposed to be involved in the attachment of the flagellum to the cell body(Martinez-Palomo et al. 1976; de Souza et al. 1978).

Returning to the deep-etch images of cell interiors, these typically reveal physicalconnections between the sub-pellicular microtubules and the plasma membrane(Figs 4, 5). The connections appear to be ~6nm diameter fibrils that are spacedalong the microtubules at intervals of ~20 nm. Similar ~6 nm fibrils are occasionallyfound to connect the sub-pellicular microtubules to subjacent elements of theendoplasmic reticulum. In addition, images of cells extracted with Triton X-100, inwhich portions of the plasma membrane are removed, reveal the underlying arrayof sub-pellicular microtubules particularly clearly (Fig. 6). In such images, similar5-6 nm fibrils are also found to connect the microtubules to each other (at intervalsof 12 to 30 nm).

Triton-extracted cells also provide numerous images of flagellar axonemes thatdisplay their outer dynein arms quite distinctly (Fig. 7). These images display sub-structure in the individual dynein arm (a bulky head attached to one A-microtubuleand a slender 'stalk' reaching over to the B-microtubule of the adjacent doublet) thatdiffers only in minor details from the substructure of outer dynein arms in otherprotozoa, as described by Goodenough & Heuser (1982). As in other protozoa, adja-cent dynein arms are spaced ~24 nm apart and appear to be linked to each other bydiagonal components termed 'linkers' by Goodenough & Heuser (1982). (SeeGoodenough & Heuser (1984) for an explanation for the piggyback-overlap of dyneinmolecules that gives rise to this image.)

Also visible in deep-etch images of detergent-extracted flagella is the characteristic'paraxial structure'. As was apparent in thin sections, this structure is composed of acomplex fabric of ~8 nm filaments woven together by narrower oblique elements.Particularly distinct in deep-etch replicas are the numerous short projections thatconnect the paraxial structure with the B-microtubules of the adjacent axoneme.

The kinetoplast can also be easily recognized in deep-etch images. In non-extractedcells, in which the mitochondrial membrane is intact, its core of kinetoplast DNAlooks like a dense rod that is opposed to the portion of the mitochondrial membranelocated near the nucleus. This rod is connected by thin filamentous structures to theopposite side of the mitochondrion, where its membrane approaches the basal body(data not shown). In Triton X-100-extracted cells, where membranes are largelyextracted and the granular material of the cytoplasm is removed so that structures likethe kinetoplast can be seen to advantage, it is apparent that the kinetoplast DNAfilaments are woven together to form radially oriented bundles. These are connectedto each other laterally by thinner filaments that appear to branch and anastomose tocreate polygonal 'facets'. In the periphery of the kinetoplast, these filaments appearto extend between the thicker bundles of DNA and remnants of the mitochondrial

Fig. 2. Epimastigote form of T. cruzi. While few particles are seen on the flagellar mem-brane {/) many appear on the membrane that encloses the cell body (cb). The cytostome(cy) can be clearly seen, x28 000.Fig. 3. General view of the plasma membrane of T. cruzi. X42000.

172 T. Souto-Padron, W. de Souza andj. E. Heuser

Figs 4—6

Ultrastructure o/Trypanosoma and Herpetomonas 173

Fig. 7. Longitudinal view of the axoneme (a) of H. megaseliae showing dynein arms inwhich heads (arrowhead) and stalks (arrow) can be seen. Connections of the filamentsforming the paraxial structure (p) to the axoneme can also be seen. X225 000.

membrane (which usually persist only on the basal body side). In addition, stouterfilaments (11 nm in diameter) connect the remains of the mitochondrial membrane tothe two basal bodies, one of which is always associated with the flagellum (Figs 8—9).Filamentous structures also connect the basal body at the base of the flagellum to asecond, lateral or accessory basal body (Fig. 9).

Figs 4—S. Cross-sections of epimastigotes of T. cruzi. The arrows in Fig. 4 indicate fibrilsconnecting the sub-pellicular microtubules to the endoplasmic reticulum (er, indicated bya dense line). The arrows in Fig. 5 indicate connections of the sub-pellicular microtubulesto the plasma membrane, x 130000.

Fig. 6. Promastigotes of H. megaseliae treated with Triton X-100 immediately beforeglutaraldehyde fixation. The plasma membrane removed, revealing the underlying layerof sub-pellicular microtubules (mt) arranged in a parallel array. Fibrils can be seen con-necting the sub-pellicular microtubules to each other (arrow). X 170000.

174 T. Souto-Padron, W. de Souza andjf. E. Heuser

Figs 8-9

Ultrastructure o/Trypanosoma and Herpetomonas 175

DISCUSSION

Previous electron microscopy of T. cruzi has depended primarily upon preparationof thin sections of chemically fixed cells. This traditional approach entails somepotential problems. Actin filaments, if not associated with other proteins, can breakdown during osmium tetroxide fixation (Maupin-Zamier & Pollard, 1978) or duringchemical dehydration (Small, 1982). Also, plastic embedding can create a back-ground electron density that obscures the finer filamentous structures of thecytoplasm (Wolosewick & Porter, 1976). Trypanosomes have also been prepared forelectron microscopy by freeze-fracture, a method that yields particular informationon the internal organization of their membranes. Freeze-fracture, however, cannotprovide additional information on internal structures such as the kinetoplast or theaxoneme-paraxial assembly inside flagella. In contrast, the deep-etch technique usedhere offers the advantage that it allows simultaneous examination of the fracturefaces and the true surfaces of various membranes, as well as the topography ofintracellular structures that have not been fixed with osmium tetroxide or dehydratedwith chemicals.

Our results with this technique essentially confirm previous freeze-fracture studieson various organisms (T. cruzi: Martinez-Palomo et al. 1976; de Souza et al. 1978;T. brucei: Smith, Njogu, Cayer & Jarlfors, 1974; Hogan & Patton, 1976; Vickerman& Tettley, 1977; Leptomonas collosoma: Linder & Staehelin, 1977; Herpetomonassamuelpessoai: de Souza, Chaves & Martinez-Palomo, 1979; Leptomonas samueli:Souto-Padr6n, Gonc.alves de Lima, Roitman & de Souza, 1980; and Leishmaniamexicana amazonensis: Benchimol & de Souza, 1980). In addition, the deep-etchimages presented here provide new information on internal structures such as the sub-pellicular microtubules, the flagellar axoneme and the kinetoplast—basal bodyassembly.

The flagellum

Previous thin-section studies have shown that axonemes in T. cruzi flagella have thetypical 9+2 pattern of microtubules, but these studies have not provided any clearimages of dynein arms or radial spokes. In contrast, the present study illustrates thefact that when an axoneme happens to lie parallel or slightly obliquely to the fractureplane, but slightly beneath it, deep-etching will reveal its surface structure very

Fig. 8. General aspect of an epimastigote form of T. cruzi that was extracted with TritonX-100 before glutaraldehyde fixation. The dense network of fine fibrils that comprise thekinetoplast (k) is evident. Stouter fibrils (arrows) are seen to connect the kinetoplast to thebasal body (b). The flagellum is at (/). X102000.

Fig. 9. Epimastigote form of T. cruzi extracted with Triton X-100, showing thekinetoplast DNA network (k) and the stout fibrils that connect the kinetoplast to the basalbody (b). At the large arrow, other fibrils are seen to connect the basal body at the baseof the flagellum to the accessory basal body (•) located laterally. A thin arrow shows a smallspace or 'facet' in the kinetoplast DNA network, while an arrowhead denotes a larger facettherein. X88000.

176 T. Souto-Padron, W. de Souza andj. E. Heuser

clearly. This has shown that the architecture of the outer dynein arms in T. cruziaxonemes conforms closely to the description given by Goodenough & Heuser (1982)for other protozoa. (That study should be consulted for details.)

The sub-pellicular microtubules

A characteristic feature of trypanosomes is a layer of microtubules beneath theirplasma membrane. These microtubules remain attached to the plasma membrane andto each other when cells are ruptured and their plasma membranes isolated (Hunt &Ellar, 1974; de Souza, 1976; Linder & Staehelin, 1977; Voorheis, Gale, Owen &Edwards, 1978; Dwyer, 1980), indicating they are indeed attached to the plasmamembrane. Previous thin-section and negative-stain studies have shown that theseattachments take the form of microtubule side-arms that recur with a periodicity of22-24 nm and extend towards the plasma membrane as either single entities or tripletsthat fan out laterally to attach to the membrane. In our deep-etch images ofmicrotubule-membrane crossfractures, we also see two or three projections fanningout laterally from each microtubule to reach the adjacent membrane. These small'fans' recur at ~20nm in our replicas. The exact mode of their attachment to theplasma membrane remains to be seen, but in this regard it should be pointed out thatthe non-random deployment of intramembrane particles and lipids seen in the plasmamembrane of these organisms (Souto-Padr6n & de Souza, 1983) most probablyreflects the patterning of the underlying lattice of microtubules.

The chemical nature of these microtubule-membrane bridges remains unknown,but the present observations tend to discount Dwyer's (1980) suggestion that theymight be dynein. This is because dynein has such a distinct substructure when seenby deep-etching (Fig. 7) while the side-arms in question look completely different(Figs 4-6). The latter are much thinner overall, and lack any of the globular expan-sions of dynein. We are more impressed with their structural similarity to deep-etchedmicrotubule-associated proteins (Heuser & Kirschner, 1980).

The kinetoplast

Previous thin-section studies of trypanosomes have demonstrated that thekinetoplast of epimastigotes consists of a complex array of delicate filaments (2-5 nmin diameter), which here and there contact the inner mitochondrial membrane(reviewed by Vickerman & Preston, 1976). Electron microscopy of the DNA in thekinetoplast, isolated after lysis of the cell, indicates that it consists of a network of20 000 to 30 000 'minicircles', each with a length of ~0-45 pirn, plus some larger circles(called 'maxicircles') each with a length of about 10 pirn (reviewed by Borst & Hoeij-makers, 1979). In the present study, deep-etching has permitted three-dimensionalexamination of the kinetoplast—DNA network in intact cells. We find that the DNAis packed into dense bundles that belie its underlying arrangement into circles.Presumably, this underlying structure becomes apparent only during isolation, whenthe DNA may lose some components involved in its in situ packing. Unfortunately,the images provided here offer no ready explanation as to why maxicircles are seenpreferentially at the periphery of spread DNA preparations (cf. fig. 2 of Hadjuk &

Ultrastructure o/Trypanosoma and Herpetomonas 177

Cosgrove, 1979; figs 1-6 of Hoijmakers & Weijers, 1980; fig. 1 of Borst & Hoeij-

makers, 1979; and fig. 3 of Leon et al. 1980).

Kinetoplast—basal body association

During the normal life-cycle of trypanosomes, the position of the kinetoplast

relative to the nucleus changes; in the amastigote form the kinetoplast is located in the

anterior portion of the cell, while in the trypomastigote form it is located in the

posterior region. In both cases, however, the kinetoplast is located close to the basal

body. Although previous ultrastructural studies have not shown any structural links

between the two, they also tend to hold together during isolation (Braly, Simpson &

Kretzen, 1977), except when divalent cations are removed (Simpson, 1972). The

deep-etch images presented here illustrate the fact that the structural entities main-

taining this connection are relatively stout fibrils, ~11 nm in diameter. Their chemi-

cal nature remains to be determined.

This work has been supported by CNPq, FINEP, UNDP/World Bank/WHO SpecialProgramme for Research and Training in Tropical Diseases and a grant from the U.S.P.H.S.National Institute of General Medical Science to J.E.H.

REFERENCES

BENCHIMOL, M. & DE SOUZA, W. (1980). Freeze-fracture study of the plasma membrane ofLeishmania mexicana amazonensis.J. Parasit. 66, 941-947.

BORST, P. & HOEIJMAKERS, J. H. J. (1979). Kinetoplast DNA. Plasmid 2, 20-40.BRALY, P., SIMPSON, L. & KRETZEN, F. (1974). Isolation of kinetoplast-mitochondrial complexes

from Leishmania tarentolae.J. Protozool. 21, 782-790.DE SOUZA, W. (1976). Association membrane-microtubules chez Trypanosoma cruzi.J. Microsc.

Biol. Cell. 25, 189-190.DE SOUZA, W., CHAVES, B. & MARTINEZ-PALOMO, A. (1979). Freeze-fracture study of the cell

membrane of Herpetomonas samuelpessoai. J. Parasit. 65, 109-116.DE SOUZA, W., MARTINEZ-PALOMO, A. & GONZALES-ROBLES, A. (1978). The cell surface of

Trypanosoma cruzi: cytochemistry and freeze-fracture. J. Cell Sci. 33, 285-299.DWYER, D. M. (1980). Isolation and partial characterization of surface membranes from Leish-

mania donovani.J. Protozool. 27, 176-182.GOODENOUGH, U. W. & HEUSER, J. E. (1982). Substructure of the outer dynein arm. J. Cell Biol

95, 798-815.GOODENOUGH, U. W. & HEUSER, J. E. (1984). Structural comparison of purified dynein proteins

with in situ dynein arms. J. molec. Biol. (in press).HADJUK, S. L. & COSGROVE, W. B. (1979). Kinetoplast DNA from normal and dyskinetoplastic

strains of Trypanosoma equiperdum. Biochim. biophys. Ada 561, 1-9.HEUSER, ] . E. & KIRSCHNER, M. W. (1980). Filament organization revealed in platinum replicas

of freeze-dried cytoskeletons.7. Cell Biol. 86, 212-234.HEUSER, J. E., REESE, T. S., JAN, L-Y-., DENNIS, M. J. & EVANS, L. (1979). Synaptic vesicle

exocytosis captured by quick freezing and correlated with quantal transmitter release, j . Cell Biol.81, 275-300.

HEUSER, J. E. & SALPETER, S. R. (1979). Organization of acetylcholine receptors in quick-frozen,deep-etched, and rotary-replicated Torpedo postsynaptic membrane. J . Cell Biol. 82, 150-173.

HIROKAWA, N. & HEUSER, J. E. (1981). Quick-freeze, deep-etch visualization of the cytoskeletonbeneath surface differentiations of intestinal epithelial cells. J . Cell Biol. 91, 399-409.

HIROKAWA, N., TILNEY, L. G., FUJIWARA, K. & HEUSER, J. E. (1982). Organization of actin,myosin, and intermediate filaments in the brush border of intestinal epithelial cells. J. Cell Biol.94, 425-443.

178 T. Souto-Padron, W. de Souza andj. E. Heuser

HOEIJMAKERS, J. H. J. & WEIJERS, P. J. (1980). The segregation of kinetoplast DNA networks inTrypanosoma brucei. Plasmid 4, 97-116.

HOGAN, J. C. JR & PATTON, C. L. (1976). Variation in intramembrane components ofTrypanosoma brucei from intact and X-irradiated rats. A freeze-cleave study. J. Protozool. 23,205-2115.

HUNT, R. C. & ELLAR, D. J. (1974). Isolation of the plasma membrane of a trypanosomatidflagellate: general characterization and lipjd composition. Biochim. biopkys. Ada 339, 173-189.

LEON, W., FRASCH, A. C. C , HOEIJMAKERS, J. H. G., FASE-FOWLER, F., BORST, P., BRUNEL,

F. & DAVISON, J. (1980). Maxi-circles and mini-circles in kinetoplast DNA from Trypanosomacruzi. Biochem. biophys. Ada 607, 221-231.

LINDER, J. C. & STAEHELIN, L. A. (1977). Plasma membrane specialization in a trypanosomatidflagellate. J . Ultrastrud. Res. 60, 246-262.

MARTINEZ-PALOMO, A., DE SOUZA, W. & GONZALES-ROBLES, A. (1976). Topographical dif-ferences in the distribution of surface coat components and intramembrane particles..7. Cell Biol.69, 507-513.

MAUPIN-SZAMIER, P. & POLLARD, T. D. (1978). Actin filament destruction by osmium tetroxide.J. Cell Biol. 77, 837-852.

ROITMAN, C , ROITMAN, I. & AZEVEDO, H. P. (1972). Growth of an insect trypanosomatid at 37 °Cin a defined medium. J. Protozool. 19, 346—349.

ROOF, D. J. & HEUSER, J. E. (1982). Surfaces of rod photoreceptor disk membranes: integralmembrane components. J . Cell Biol. 95, 487-500.

SIMPSON, L. (1972). The kinetoplast of the haemoflagellates. Int. Rev. Cytol. 32, 139.SMALL, J. V. (1982). Organization of actin in the leading edge of cultured cells: influence of osmium

tetroxide and dehydration on the ultrastructure of actin networks..?. Cell Biol. 91, 695-705.SMITH, D. S., NJOGU, A. R., CAYER, M. & JARLFORS, U. (1974). Observations of freeze-fractured

membranes of a trypanosome. Tissue & Cell 6, 223-241.SOUTO-PADR6N, T. & DE SOUZA, W. (1983). Freeze-fracture localization of filipin-cholesterol

complexes in the plaama membrane of Trypanosoma cruzi. J. Parasit. 69, 129-137.SOUTO-PADR6N, T., GONCALVES DE LIMA, V. M. Q., ROITMAN, I. & DE SOUZA, W. (1980). Fine

structure study of Leptomonas samueli by the freeze-fracture technique. ZentblBakt. ParasitKdeAbt. 7 62, 145-157.

VICKERMAN, K. & PRESTON, T. M. (1976). Comparative cell biology of the kinetoplastidflagellates. In Biology ofKinetoplastids (td. W'. H. R. Lunsden&D. A. Evan), vol. 1, pp. 35-130.New York, London: Academic Press.

VICKERMAN, K. & TETLEY, H. (1977). Recent ultrastructure studies on trypanosomes. Annls Soc.beige Mid. trop. 57, 441-455.

VOORHEIS, H. P., GALE, J. S., OWEN, M. J. & EDWARDS, W. (1979). The isolation and partialcharacterization of the plasma membrane from Trypanosoma brucei. Biochem. J. 180, 11-24.

WARREN, L. (1960). Metabolism of Schizotrypanum cruzi, Chagas. I. Effect of culture age andsubstrate concentration on respiratory rate. J. Parasit. 46, 529-539.

WOLOSEWICK, J. J. & PORTER, K. R. (1976). Stereo high-voltage electron microscopy of whole cellsof the human diploid line, WI-38. Am. J. Anat. 147, 303-324.

{Received 9 November 1983-Accepted, in revised form, 29 February 1984)