Quantitative changes and ultrastructural alterations of the cornea in response to ultraviolet light II. Effects on amphibia; elucidation of desmosomal structure and basement membrane synthesis

TISSUE & CELL 1976 8 (4) 603-614 Published by Longman Group Ltd. Printed in Great Britain

LUKAS

H. MARGARITIS”,

THEMISTOCLES K. POLlTOFt and JOHN X. KOLIOPOULOSS

QUANTITATIVE CHANGES AND ULTRASTRUCTURAL ALTERATIONS OF THE CORNEA IN RESPONSE TO ULTRAVIOLET LIGHT II. EFFECTS 0:N AMPHIBIA; ELUCIDATION OF DESMOSOMAL STRUCTURE AND BASEMENT MEMBRANE SYNTHESIS ABSTRACT. The effects of far ultraviolet light irradiation upon an amphibian cornea were studied to compare the effects observed both quantitatively and ultrastructurally with data obtained after UV irradiation of mammalian corneas. The ultimate goal of this series of investigations is the elucidation of the alterations and the regeneration mechanisms, which might reflect existing morphological diversities among the species, observed in vertebrate corneas following exposure to UV light. It was found that while the epithelial cells undergo oedema after low dose exposures and are gradually damaged after high doses of UV light, 2-4 days later a new epithelium has been formed. Intercellular permeability is increased by low dose exposure as was detected by the penetration of Ruthenium Red into the intercellular clefts. Under these conditions desmosomal structure revealed a Zl-laminar configuration. The basement membrane of the amphibian, unlike that of the mammal, does not disolve away upon exposure but shows localized disruptions which are thought to accommodate the passage of leucocytes from stroma to epithelium. That a new basement membrane is subsequently formed is evident by the existence of extracellular and intracellular secretion granules. In comparison to irradiated rabbit corneas, this stroma remains remarkably at the same thickness following a high dose exposure although a noticeable disorganization of collagen arrangement is apparent. Finally, as in the case of the rabbit corneas, a secondary degeneration of endothelium was observed 4 days after a moderate dose exposure.

Drotthis-Draper law of absorption only absorbed light causes photochemical or photobiological reactions (Giese, 1964). Thus it is expected that such reactions will involve nucleic acids as well as proteins since these molecules readily absorb UV light (Smith, 1966; Setlow, 1967). However, other molecules could be indirectly involved in the overall effect of UV upon living systems. In the case of nucleic acids it has been reported that the maximum sensitivity of the cells during UV exposure occurs at the early S phase (Domon and Rauth, 1969) due to inhibition of DNA synthesis possibly caused by breakage of double helix hydrogen bonds

Introduction ULTRAVIOLET light has been used extensively as an experimental tool to investigate the function and structure of organelles, cells, tissues and organisms. On the basis of the From the Dept. of Biology, University of Athens, Greece, and the Biological Laboratories, Harvard University, Cambridge, Mass., U.S.A. Present address: * Dept. of Biology, University of Athens, Panepistimiopolis-Kouponia, Athens-621, Greece. t Dept. of Physics, Cornell University, U.S.A. $ University Eye Clinic, Ithaca, N.Y., National Ophthalmological Center, Mesogion-170, Holargos, Athens, Greece. Received 5 April 1976. Revised 16 August 1976. 603

MARGARITIS.

604

(Painter, 1970; Smith, 1966). Moreover the formation of cytoplasmic RNA has been found to be 66% reduced following UV exposure in some cells (Painter, 1970). There is evidence that proteins are sensitive to UV light in proportion to the number of disulfide bonds in their structure-these bonds break upon exposure (Augenstine et al., 1960). Enzymes are inactivated either as a result of the above effect or due to the absorption of UV by aromatic amino acids (Setlow, 1967). Unsaturated fatty acids do absorb UV energy (Nikolaides, 1966), whereas it has been shown that lipid composition changes in E. coli after UV exposure (Jacobson and Yotrin, 1975) and membrane lipid peroxidation is induced in vivo and in vitro (Johnson et al., 1968; Roshchupkin et al., 1975). Carbohydrates are considered to be indirectly affected (Smith, 1966). Thus, glycogen content increases following UV irradiation (Johnson et al., 1968) and hyaluronic acid is polymerized in vivo and in vitro.

Since the UV absorbing macromolecules are forming supramolecular structures, the last are equally affected directly or indirectly. Thus in irradiated skin, a swelling followed by fragmentation and subsequent dissolution of the basement membrane was observed (Epstein, 1970) and an increase in total epithelium thickness due to oedema was detected. Leucocyte migration reportedly takes place in some cases as well (Johnson et al., 1968). Physiologically, UV induces pinocytosis in Amoeba and changes membrane permeability (Giese, 1964). This work investigates the possible effects

POLITOF

AND

KOLIOPOULOS

of UV light on the amphibian cornea which has been shown to differ both quantitatively and qualitatively from the mammalian cornea (Margaritis et al., 1976), and compares the results with previously reported data concerning UV exposure of mammalian corneas (Margaritis et al., 1975). Moreover, we decided to investigate the effect of UV irradiation on membrane permeability by using the polyvalent basic dye Ruthenium Red as an intercellular tracer (Luft, 1971a). Materials and Methods The right eyes of eight salamanders, Triturus c. cristatus, were irradiated with UV light

emitted by a UVS-12 mercury lamp (Ultraviolet Products Inc., San Gabriel, California). The lamp intensity was calibrated with a J-225 intensity gauge from the same company. Emission lines at 2600 (major intensity), 3300 and 4000 A were detected after a spectroscopic analysis. The center of the lamp’s filter was placed 5 cm away from the surface of the cornea under irradiation. The UV intensity at that point was 400 pW/cm2. Prior to UV exposure of the eyes (the rest of the animal was suitably protected) the salamanders were slightly anesthetized with a 1:2500 solution of MS-222 (Sandoz). Following irradiation the animals were left to recover in tap water and were decapitated after a certain period of time as indicated in Table 1. The corneas were subsequently dissected and fixed for 90 min at 22°C with 2.5 % glutaraldehyde (Polysciences Inc.) in 0.1 M sodium cacodylate buffer pH, 7.4, containing 800 ppm Ruthenium Red (BDH Ltd). Washing was carried out in the same

Table 1.Protocol of the U V irradiation experiments _ ~~ Total dose (erg/cm? 105 8x105 2x 106 5 x lo6 (a)* 5 x IO6 (b) 20 x lo6 (aj 20 x lo6 (b) 40 x 106

Low dose Low dose Moderate dose Moderate dose Moderate dose High dose High dose High dose

Rate (erg/cm2 day) 105 8x 105 2x 10” 5x 10” 5x10” 20x 10” 5x 10” 5x10”

* (a) and (b) are used throughout the text for convenience.

Number of exposures 1 1 1 1 1 1 4 8

Dissection (days after the last exposure) 1 1 1 1 4 2 1

I

ALTERATIONS

OF

CORNEA

IN

UV

LIGHT

buffer containing 6% sucrose and 400 ppm RR for 40 min at 4°C. PostfIxation with 2% aqueous 0~04 included 600 ppm RR and lasted 90 min at 4°C. After a second wash as before the tissues were dehydrated in ethanol and embedded in Epon-Araldite (Margaritis, 1974). Sectioning and subsequent treatment for quantitative analysis and electron microscopy were performed as described elsewhere (Margaritis et al., 1975, 1976). The unexposed (left) eyes served as controls throughout the quantitative and ultrastructural processing. Only cross-sections of the center of the cornea were studied for the effects of irradiation.

605

sure, stays normal with a moderate dose, and decreases with a high-dose exposure (Table 2, Fig. 1).

Results I. Quantitative data (a) Epitheliunz. The total thickness of cornea1 epithelium increases with a low-dose expo-

Fig. 1. Histogram showing the variations in epithelium thickness as a function of UV dose exposure. , irradiated (right) eyes; 0, unirradiated (left) eyes. (a) and (b) arc explained in Table 1.

2. Quantitative light microscopical data of the variations in epithelium and stroma thicknesses as a function of UV dose exposure

Table

Epithelium

Thickness (P)

Dose (ers/cm2)

Ratio UVjC

Stroma -

% total

Thickness (P)

% total

uv

36.3k2.2 26.1k2.1

1.39+0.14

35.4f3.2 31.4k2.8

66.3k6.7 61.2k2.4

64.6k7.9 68.6k3.0

uv

33.1&1.0 24.9* 1.4

1.33+0.08

32.6k3.0 30.5t1.9

68.6k2.0 57.0* 1.6

67.3k2.5 69.6k2.7

uv

20.2& 1.1 21.1kO.5

0.95kO.06

26.6& 32.0+

1.5 1.2

55.7k1.4 45.1k1.8

73.4k2.4 68.Ok3.4

5 x lo6 (a)

uv

26.8k2.1 25.9kO.8

1.03&0.09

33.8k3.2 29.7+ 1.3

52.7k1.4 61.2k2.4

66.3k2.6 70.3k3.4

5 x 106 (b)

uv

21.OkO.9 24.Ok2.3

0.88+0.09

28.1k1.3 32.1 k3.3

53.5kO.9 50.7k1.3

71.9+ 1.7

37.6z1.2

37.8kO.9 38.0+1.0

62.122.1

105

C 8x 105

C

2x106

C

C

C

20 x lo6 (a) 20 x 10R(b)

uv

-

22.9kO.6

uv

18.1k1.7 21.2cO.5

0.85+0.08

31.1k3.1 32.0+ 1.2

4O.lk1.6 45.1k1.8

68.9k3.9 68.Ok3.4

uv

9.2k1.2 21.2kO.5

0.43kO.06

17.5k2.9 32.0+ 1.2

43.4k5.4 45.1k1.8

82.5k3.4 68.Ok3.4

C

(a), (b): see explanation in Table 1. UV: Irradiated (right) eyes. C: Unirradiated (left) eyes of the same animal.

39

-

C

C

40x 106

-

67.9k3.0

MARGARITIS,

606

(b) Stroma. No significant changes in the overall stromal thickness have been detected (Table 2). II. Ultrastructural

data

(a) Epithelium. Light microscopical observations indicated that dramatic swelling of the cells occurs 24 hr after a low-dose exposure whereas higher doses result in thinner epithelium consisting of irregularly shaped cells. Two to four days after exposure the overall appearance of the epithelium resembles that of the unirradiated corneas with the exception of some cell shape irregularities. A detailed ultrastructural analysis showed that the swelling of epithelium resulting from a low-dose exposure is followed by the appearance of large intercellular spaces (Fig. 4). The most superficial intercellular clefts are

Fig. 2. The intercellular density (arrow) presumably occludentes (10s erg/cma).

POLITOF

AND

KOLIOPOULOS

often found to be heavily labeled with RR (Fig. 2) giving rise to increase in the electron density of desmosomes (Fig 5) and appearance of dense intercellular granules (Fig. 3). Whenever the desmosomes are slightly RR labeled they reveal a 21-laminar structure (Figs. 6, 18). Under these irradiation conditions the tight junctions located at the outermost region of the superficial cells (Margaritis et al., 1976) were observed to be intact. Epithelium damage occurs after a highdose exposure. Cytoplasm seems degenerated in some areas, the nuclear material becomes condensed (Fig. 7) and leucocytes appear between the basement membrane and the basal cells (Fig. 14). However, even after a low-dose exposure, the nuclei of some basal cells are found to contain large intranuclear granules (Fig. 8).

spaces of some superficial cells show increased electron due to penetration of Ruthenium Red through the zonulae x 27,000.

Fig. 3. The intercellular penetration of RR following a low-dose exposure reveals granules attached at the external surface of the plasma membranes (arrow) (10s erg/ cma). x 54,000. Fig. 4. Occasionally the large intercellular spaces formed after irradiation flocculent material (8 x 10s erg/cma). x 14,000.

contain

a

Fig. 5. Some desmosomes are heavily labeled with RR due to intercellular permeability change and show intracellular plaques extending up to 2200 A (arrows) (I O5 erg/cma). x 65,000. Fig. 6. Desmosomes which are very slightly labeled by RR reveal a 2 I-laminar structure (see also Fig. 18) of about 2200 8, total diameter (arrows). Compare with Fig. 5 (105 erg/cma). x 64,000. Fig. 7. A high-dose exposure results in damaged cytoplasm of epithelial cells, whereas the less affected cells contain nuclei with clumped chromatin (20 x IOBerg/cm2) (b) (see Table I). x 2800. Fig. 8. Some epithelial cells are found to contain nuclei with highly condensed chromatin and large granules (arrow) after a low-dose exposure (10s erg/cm”). x 27,000. Figs. 9-10. The basement membrane undergoes disruptions after moderate dose exposure (asterisk) (5 x lO”erg/cma) (b) (see Table I). Fig. 9, x 8600; Fig. IO, x 23,000. Fig. 11. Two days after a high-dose exposure some basal epithelial cells contain large pools of granular material resembling ,&glycogen particles (20 x IO” erg/cm2) (a) (see Table I). x 23,000. Inset: x 55,000. Figs. 12-13. The granular material observed in cytoplasmic pools (see Fig. I I) is frequently associated with the basement membrane (asterisk), either in a limited fashion following a low-dose exposure (Fig. 13) or in an extended form 4 days after a moderate dose (Fig. 12). Fig. 12, 5 x 1Oa erg/cm2 (b) (see Table I). x 32,000; Fig. 13, 10s erg/cma. x 44,000.

610

MARGARITIS,

Two days after a high dose some basal cells contain extended pools of granular material (Fig. 11). (b) Basement membrane (basal lamina). This structure undergoes dramatic morphological alterations after moderate dose exposures. Disorganization and partial disruption have been observed (Figs. 9, 10, 12), while at the same time there is evidence for the appearance of granular material either in the basal cells-basement membrane extracellular region (Figs. 12, 13) or inside the basal cells (Fig. 11) as already mentioned. (c) Substantia propria (stroma). Leucocytes are very often present deep in the stroma 1 day after a moderate dose or in the vicinity of the basement membrane 2 days after a high-dose exposure (Fig. 15). Similar cells are found in the basal cells region of the epithelium under the same conditions (Fig. 14). Finally, a disorganization of the familiar hexagonal arrangement of collagen fibers has been detected by exposing the corneas to a four times repeated moderate dose (Fig. 16). (d) Endothelium. Degeneration of the endothelial cells seems to occur either 4 days after a moderate dose exposure or even after exposure to the same dose four times (Fig. 17). Discussion (a) Epithelium. Ultraviolet light induces lipid peroxide formation (Johnson et al., 1968; Meffert and Lohrisch, 1971) which in turn causes swelling of organelles and cells (Hunter et al., 1964). To the overall increase in thickness an increase in mitotic rate contributes

POLITOF

AND

KOLIOPOLJLOS

as in the case of UV irradiated skin (Johnson et al., 1968). Presumably a low-dose exposure results in epithelium swelling for the above reasons as was also reported for irradiated rabbit corneas (Margaritis rt al., 1975). Higher doses lead to a thin epithelium because the superficial cells, lethally affected by UV, drop out. The appearance of RR in the intercellular spaces of the outermost epithelial cells may be explained as follows: RR is unable to penetrate tight junctions (Luft, 1971b) as was demonstrated also by the absence of electron density from the intercellular epithelial spaces of unirradiated salamander cornea (Margaritis et al., 1976). Thus, we assume that UV has affected the intercellular permeability of epithelial cells in either or both of the following ways: (i) UV has been shown to increase the level of lipid peroxidation in membranes (Giese, 1964; Hunter et al., 1964; Meffert and Lohrisch, 1971; Meffert et al., 1972; Roshchupkin et al., 1975) which in turn are believed to play a role in the control of membrane permeability (Hunter et a/., 1964; Robinson, 1965, 1966) although Leibowitz and Johnsson (1971) doubt whether there is a direct relationship between these two effects. (ii) UV results in an increase of SH groups (Johnson et al., 1968) which are thought also to affect membrane permeability (Tomlinson and Rich, 1969). We therefore consider that such an increase in intercellular permeability is responsible for the observed leakage of RR through the zonulae occludentes, although a tracer when added to fixatives may be incapable of testing in viva or ifz vitro permeability. A similar

Figs. 14-15. Micrographs illustrating the similarity in morphology between cells found in either side of the basement membrane and considered to be leucocytes, 2 days after a high-dose exposure (20 x lo6 erg/cm2) (a) (see Table 1). x 13,000. Fig. 16. A high-dose distribution of collagen

exposure results in the disorganization of the hexagonal-like fibers (10 x loo erg/cm2) (b) (see Table 1). x 72,000.

Fig. 17. The endothelial cells were found to be degenerated 4 days after exposure of the cornea to a moderate dose (20 x 106 erg/cm2 (b) and 5 x lo6 erg/cm2 (b)) (see Table 1). x 15,000.

612

MARGARITIS.

leakage has been induced in vitro withacetone treatment which similarly leads to undetectable morphological changes (Goodenough and Revel, 1970). The above permeability change has been shown in rabbit corneas as well (Margaritis et al., 1975) and facilitated the penetration of RR into the desmosomes. Thus, these structures revealed a 21-laminar configuration (Fig. 18) strikingly different than the one reported for amphibian epidermis (Kelly, 1966) or other cells (Luft, 1971b). Even though the structure of desmosomes is dynamic and variable we feel confident that RR labeling is a useful procedure for the investigation of desmosomes or other junctions(Leik and Kelly, 1970) throughout living systems. The intercellular granules observed after penetration of RR might represent mucopolysaccharides and have been observed in other cell systems (Luft, 1971 b) where there are no tight junctions to block the passage of the tracer. It is known that the nuclear material of cells which undergo DNA synthesis is sensitive to UV irradiation (Painter, 1970). Since a small percentage of cornea1 epithelial cells

,L

II

:.::.:.:.:.:.:.:.:.:.:.:

POLITOF

AND

KOLIOPOULOS

is expected to be in that phase during irradiation, then the observed abnormal intranuclear granules, after a low-dose exposure, may represent the product of UV action, possibly due to crosslinking of DNA to proteins (Smith, 1966). Under higher doses the epithelial cells are damaged primarily as a result of enzyme inactivation (Wills and Willkinson, 1967). The origin of the cytoplasmic granules observed in the basal cells 2 days after a high dose remains unclear because it is impossible to discriminate morphologically between ribosomes and /3-glycogen. However it seems that the granules represent glycogen for the following reasons: (i) There is evidence that ribosomal RNA synthesis is inhibited in cells exposed to UV (Nozu, 1972). (ii) The granules are morphologically identica to those observed extracellularly and thought to be secreted for basement membrane synthesis. (iii) It has been reported that UV irradiated skin shows an increase in glycogen content 15 hr after exposure (Johnsson et al., 1968). (iv) Jt has been found that the same (basal) cells in another (human) cornea contain glycogen (Smelter and Gzanics, 1965). Thus we believe that these granules are playing a role in the repair of PAS positive basement membrane components.

UM Et_ :.:.:.:*:.:.:.:.:.:.:.:.!A

,L

IL ’

B C DE F G H

. .

‘.‘.:.‘.:.‘.‘.:.:.‘.‘.‘.

I. I.

.

.

.* .

.

.

,-:

IJ

:A

((

.

1

, I

Fig. 18. Schematic representation of the desmosomal structure as revealed after labeling with Ruthenium Red (illustrated in Figs. 5, 6). Whenever heavily labeled, the intercellular layers plus two half unit membranes either side do not show any substructure (refer to Fig. 5). However, all the intracellular layers are apparent in both lightly and heavily labeled desmosomes. UM, unit membrane of adjacent cells; EL, extra(inter)cellular layers; extremely electron dense in heavily RR-labeled desmosomes as in Fig. 5; IL, intracellular layers showing the same electron density in either heavily or lightly stained desmosomes. A, 310-330 A; B, 380400 A; C, 45tS-470 A; D, 550-600 A; E, 66&700 A; F, 900-950 A; G, 1200-1250 A; H, 1400-1450 A; I, 1700-1800 A: J, 2200-2300 A.

(b) Basement membrune (basal lam&a). This structure turned out to be a very active system following UV exposure, partly dissolving to accommodate the passage of leucocytes to epithelium. However, it is not clear whether this membrane is affected directly by UV-this might be possible by breakage of disulfide bonds (Augenstine et a/., 1960) known to be crosslinking agents in basement membranes in general (Kefalides, 1973)-or whether it is enzymatically digested We have occasionally observed such a disruption of basement membrane accompanied by a proximal appearance of leucocytes in unirradiated corneas (unpublished results). This observation suggests that UV exposure amplifies this physiological response.* At the * In fact we consider the secretion process observed after a low-dose exposure to be due to a physiological repair function and not due to irradiation.

ALTERATIONS

OF

CORNEA

IN

UV LIGHT

phase of recovery, however, it is demonstrated that the basal cells are capable of secreting material for the formation of a new basement membrane where necessary, as is the case in unirradiated corneas during minor repairs (Margaritis et al., 1976) and in other corneas during cornea1 disease (Kenyon, 1969). Moreover, a similar fragmentation of basement membranes occurs in UV irradiated skin (Epstein, 1970). Compared to rabbit cornea, this basal lamina is highly UV resistant, since the former totally disappears under the same high-dose exposure (Margaritis et al., 1975). (c) Substantia propriu (stroma). Although this cornea’s epithelium is thinner (less UV absorbance) compared to rabbit cornea (Margaritis et al., 1975), the overall effect of UV upon stromal organization and total thickness is less dramatic. The total thickness does not change significantly even after a high-dose exposure (40 x 106 erg/cm2), whereas half dose has been shown to cause a duplication of stromal thickness in the rabbit (Margaritis et ul, 1975). Such a UV resistance could be compatible with the existence of sutural fibers as those observed in the shark (Goldman and Benedek, 1971), although we did not observe such structures in this cornea. The only noticed effect of UV upon stroma, namely that of the hexagonal

613

pattern disorganization, might be explained by a breakage of mucopolysaccharide molecules known to connect adjacent collagen fibers and shown to depolymerize after UV exposure (Johnson et al., 1968). (d) Endothelium. A very surprising effect which was similarly reported to occur in irradiated rabbit corneas under exactly the same conditions (Margaritis et al., 1975) concerns the secondary degeneration of endothelial cells 4 days after a moderate dose exposure. The interpretation of this effect remains obscure because the fate of the cells at later intervals has not been determined. However, a reasonable explanation might involve the diffusion of toxic molecules produced in the degenerated epithelial cells during the first or second day of irradiation. It is clear, therefore, that ultraviolet radiation is very useful for the investigation of structures or functions which are otherwise undetected. Acknowledgements

This work was supported by a grant from the Empirikion Foundation of Athens, Greece. Completion of the work was performed in the laboratory of Dr D. Branton (The Biological Laboratories, Harvard University), to whom we are deeply indebted.

References AUGENSTINE,L. G., CARTER, J. G., NELSON, D. R. and YOCKEY, H. P. 1960. Radiation effects at the macromolecular level. Radiar. Res., Suppl. 2, 19-48. DOMON, M. and RAUTH, A. M. 1969. Ultraviolet-light irradiation of mouse L cells: effects on cells in the DNA synthesis phase. Radiat. Res., 40, 414-429. EPSTEIN, J. H. 1970. Ultraviolet carcinogenesis. In Photophysiology (ed. A. C. Giese), Vol. V, pp. 235-273. Academic Press, New York and London. GIESE, A. C. 1964. Studies on ultraviolet radiation action upon animal cells. In Photophysiology (ed. A. C. Giese), Vol. V, pp. 203-245. Academic Press, New York and London. GOLDMAN, J. N. and BENEDEK, G. B. 1971. The relationship between morphology and transparency in the unswelling cornea1 stroma of the shark. Invest.Ophth., 6, 574-600. GOODENOUGH, D. A. and REVEL, J. P. 1970. A fine structural analysis of intercellular junctions in the mouse liver. J. Cell Biol., 45, 273-290. HUNTER, F. E., JR., SCOTT, A., HOFFSTEN, P. E., GUERRA, F., WEINSTEIN,J., SCHNEIDER, A., SCHUTZ, B., FINK, J., FORD, L. and SMITH, E. 1964. Studies on the mechanism of ascorbate-induced swelling and lysis of isolated liver mitochondria. J. biol. Chem., 239, 604-613. JACOBSON, A. F. and YOTRIN, M. B. 1975. Changes in lipid composition of E. coli following y- and UVirradiation. 19th Ann. Meet. Bioph. Sot., Philadelphia, Penn., 18-21 Feb. 1975. Abstract in Bioph. J., 15 (2), 199a.

614

MARGARITIS,

POLITOF

JOHNSON,B. E., DANIELS, F., JR. and MAGNUS, 1. A. 1968. Response

AND

KOLIOPOULOS

of human skin to ultraviolet light. In Photophysiology (ed. A. C. Giese), Vol. V, pp. 139-201. Academic Press, New York and London. KEFALIDES,N. A. 1973. Structure and biosynthesis of basement membranes. Int. Rev. Conn. Tissue Res., 6, 63-104. KELLY, D. E. 1966. Fine structure of desmosomes, hemidesmosomes and an epidermal globular layer in developing newt epidermis, J. Cell Biol., 28, 51-72. KENYON, K. R. 1969. The synthesis of basement membrane by the cornea1 epithelium in bullous keratopathy. Invest. Ophth., 8, 156-168. LEIBOWITZ, M. E. and JOHNSON, M. C. 1971. Relation of lipid peroxidation to loss of cations trapped in liposomes. J. Lipid Res., 12, 662-670. LEIK, J. and KELLY, D. E. 1970. Septate junctions in the gastrodermal epithelium of Phialidium: a fine structural study utilizing Ruthenium Red. Tissue & Cc/l, 2, 435-441. LUFT, J. H. 1971a. Ruthenium Red and Violet. I. Chemistry, purification, methods of use for electron microscopy and mechanism of action. Amt. Rec., 171 (3), 347-368. LUFT, J. H. 1971b Ruthenium Red and Violet. II. Fine structural localization in animal tissues. Anaf. Rec., 171, (3), 369415. MARGARITIS, L. H. 1974. Programmed synthesis of specific proteins in cellular differentiation. A contribution to the study of chorion formation in the follicles of D. melunogaster. Ph.D. Thesis, University of Athens, Greece. MARGARITIS,L. H., POLITOF,Th. K. and KOLIOPOULOS,J. X. 1975. Quantitative changes and ultrastructural alterations of the cornea after exposure to ultraviolet light. 1. Effects on rabbits. Min. Med. Greta, 3, (3), 295-305. MARGARITIS, L. H., POLITOF, Th. K. and KOLIOPOULPS, 1. X. 197.5. Quantitative and comparative ultrastructure of the vertebrate cornea. I. Urodele ambhibia. Tissue & Cell, 8, 591-602. MEFFERT, H. and LOHRISCH, 1. 1971. Hemmung der Atmung von Epidermiszellen und Fibroblasten durch Produkte der Lipidperoxydation. Dem. Wschr., 157, 793-801. MEFFERT,H., DRESSLER,Ch. and MEFFERT,B. 1972. UV-provozierte Lipidperoxydation in Epidermis. Korium und Subkutis des Menschen in vitro. Acta biol. med. germ., 29, 667-675. NIKOLAIDES,N. 1966. Lipids, membranes and the human epidermis. In The Epidermis (eds. W. Montagna and W. C. Lobitz), pp. 51 l-538. Academic Press, New York. Nozu, K. 1972 Effects of ultraviolet irridation on biological functions of DNA and RNA. In Aspecfs of Cellular and Molecular Physiology (ed. K. Haguchi), pp. 147-165. Faculty of Osaka University Press. PAINTER, R. B. 1970. The action of UV light on mammalian cells. In Photophysiology (ed. A. C. Giese), Vol. V, pp. 169-189. Academic Press, New York and London. ROBINSON, J. D. 1965. Structural changes in microsomal suspensions. III. Formation of lipid peroxides. Archs. Biochem. Biophys., 112, 17&l 79. ROBINSON, J. D. 1966. Interaction between protein sulphydryl groups and lipid double bonds in biological membranes. Nature, Lond., 212, 199-200. ROSHCHUPKIN, D. I., PELENITSYN,A. B., POTAPENKO, A. YA., TALITSKY, V. V. and VLADIMIROV,Yu. A. 1975. Study of the effects of ultraviolet light on biomembranes. IV. The effect of oxygen on UV-induced hemolysis and lipid photoperoxidation in rat erythrocytes and liposomes. Photorh. Photobiol., 21, 63-69. SETLOW,J. K. 1967. The effects of ultraviolet radiation and photoreactivation. In Comprehensive Biochemistry (eds. M. Florkin and E. H. Stotz), pp. 157-209. Vol. 27, Elsevier Pub]. Co., Amsterdam, London, New York. SMELCER,G. K. and OZANICS, V. 1965. New concepts in anatomy and histology of the cornea. In The Cornea, World Congress (eds. J. H. King and J. W. McTigue), pp. l-20. Butterworths, Washington. SMITH, K. C. 1966. Physical and chemical changes induced in nucleic acids by ultraviolet light. Radiat. Res., Suppl. 6, 54-79. TOMLINSON,H. and RICH, S. 1969. Relating lipid content and fatty acid synthesis to ozone injury of tobacco leaves. Phytopathology, 59, 12841286. WILLS, E. D. and WILLKINSON, A. E. 1967. The effect of irradiation on lipid peroxide formation in subcellular fractions. Radiaf. Res., 31, 732-747.