954

The nuclei were normal, and budding from the cell surface which is described for many other viruses, but not for coronaviruses, was not seen. Control cells showed none of these changes. Fsecal emulsion from patient X was also inoculated into primary human embryo-kidney monolayers. Both these cells and sections of inoculated organ cultures showed immunofluorescence with convalescent serum from patient X and fluorescein-conjugated antihuman globulin. In the monolayers, after 2 days’ incubation, small particulate areas of fluorescence were seen in the cytoplasm of some of the cells (fig. 3a),

which progressed after a further 4 days’ incubation to large inclusion-like masses (fig. 3b). This appearance has been described in coronavirus-infected cells.ll No fluorescence was seen in uninoculated cultures, In the organ cultures 12 hours after inoculation, fluorescence was seen only in the epithelium of the villi and not in other cells-similar to the appearance of pig intestine infected with TGE.11 The increase in the amount of fluorescence between the 2nd and 4th days in human embryo-kidney cells suggests multiand not phagocytosis. More work is required to elucidate the relationship of this virus to gastroenteritis, and to the animal intestinal coronaviruses. Sharpel and Mebusl2 have found antibodies to calf coronavirus in human sera. These particles cannot be identified as coronaviruses on morphology alone, but the typical electron microscopic appearance in cell sections is very suggestive. Definite characterisation of the particles as coronaviruses must await further work.

plication

REFERENCES

McIntosh, K. Curr. Topics Microbiol. Immun. 1974, 63, 85. Stair, E. L., Rhodes, M. B., White, R. G., Mebus, C. A. Am. J. vet. Res. 1972, 33, 1147. 3. Lee, K. M. Ann. N. Y. Acad. Sci. 1956, 66, 191. 4. Mebus, C. A., Stair, E. L., Rhodes, M. B., Twiehaus, M. J. Am. J. vet. Res. 1973, 34, 145. 5. Mathan, M., Mathan, V. I., Swaminathan, S. P., Yesudoss, S., Baker, S. J. Lancet, 1975, i, 1068. 6. Caul, E. O., Paver, W. K., Clarke, S. K. R. ibid. 1192. 7. Berry, D. M., Cruikshank, J. G., Chu, H. P., Wells, R. J. H. Virology, 1. 2.

1964, 23, 403. Almeida, J D., Berry, D. M., Cunningham, C. H. Hamre, D., Hofstad, M. S., Mallucci, L., McIntosh, K., Tyrrell, D. A. J. Nature, 1968, 220, 650. 9. Becker, W. B., McIntosh, K., Dees, J. H., Chanock, R. M. J. Virol. 1967, 1, 1019. 10. Oshiro, L. S., Schieble, J. H., Lennette, E. H. J. gen. Virol. 1971, 12, 161. 11. Pensaert, M. B., Burnstein, T., Haelterman, E. O. Am. J. vet. Res. 1970, 31, 771. 12. Sharpel, R., Mebus, C. A. Lancet, 1975, i, 639. 8.

CLINICAL SIGN OF OBSTRUCTED AXOPLASMIC TRANSPORT Fig. 2-Thin section of columnar epithelial cell from human embryo in. testinal-organ culture inoculated with faeces from patient X: (a) show ing crescents of thickened membrane lining the cystoplasmic vesicle (virus particles are in the vesicle); (b) inclusion containing tubulat structures.

(x 66 000. Bar=nm.)

DAVID MCLEOD

Moorfields Eye Hospital, London EC1V 2PD and retrograde axoplasmic transport in retinal ganglion-cell axons can be interrupted by axonal ischæmia. This report is believed to be the first to illustrate how this phenomenon can be observed clinically in man in cases of retinal vascular disease. The intense retinal "whiteness" of small cottonwool spots and at the periphery of larger areas of retinal ischæmia represents gross localised axonal distension secondary to the cessation axoplasmic flow.

Summary

Orthograde

INTRODUCTION

3-Human embryo-kidney monolayer inoculated with faeces from patient X (indirect immunofluorescence with convalescent serum from patient X); (a) 2 days after inoculation; (b) 6 days after inoculation.

Fig.

A CONSTANT flow of subcellular particles and molecules takes place within the axons of nerve-cells. This flow is called "axoplasmic transport", and the visual pathway provides physiologists with a convenient experimental system for its investigation. Using enzyme markers and autoradiography, slow and rapid phases of

955 flow have been demonstrated in retinal ganglion-cell axons, and the same techniques have indicated that retrograde axonal flow also takes place between synaptic terminals in the lateral geniculate body and the perikaryon situated in the inner retina.’-3 Although it depends on oxidative metabolism, axoplasmic flow does not require physical continuity between the axon and the perikaryon; axonal transportation is controlled locally within the neurite, and is responsible for the aggregation of organelles and proteins on both sides of local interruptions of nerve fibres, due to crushing or ischxmia.4 After carotid embolisation in pigs,5 Shakib and Ashton described an accumulation of organelles in terminal axonal swellings within areas of focal retinal ischaemia;6 the ultrastructural features were similar to those seen in a patient with systemic hypertension and retinal cottonwool spots.’ Mitochondrial aggregation was ascribed to "reactive" organelle proliferation in injured axons whose cell-bodies were situated outside the ischaemic

orthograde axoplasmic

area.

Although

this

interpretation

is

no

longer tenable,

Shakib and Ashton made the crucial observation that localised mitochondrial accumulation within distended axon-bulbs accounts for the intense fluffy-white appearance of a small cottonwool spot, while diffuse axonal swelling and intracellular oedema, seen within larger areas of ischaemia, produces a greyish-white discolouration of the retina. By unifying this clinicopathological correlation with the physiological concept of axonal flow, it is possible to illustrate the interruption of orthograde and retrograde axoplasmic transport in the human retina by detailed clinical observation of the ophthalmoscopic signs resulting from retinal vascular occlusions. MATERIAL AND METHODS

Fundus photographs and intravenous fluorescein angiograms from over a hundred patients with retinal arterial occlusions have been examined. These included cases of central retinal-

macula

right photograph: Fig.TheI-Fundus fovea is central; the optic disc is

eye. on

the

right.

Fig. 2-Fundus photograph: macula left eye. The fovea is central; the optic disc is on the left.

artery occlusion, central arterial occlusion with sparing of the cilioretinal arteriolar territory, isolated cilioretinal arteriolar occlusion, central retinal-vein occlusion with cilioretinal infarction, ischaemic optic neuropathy with cilioretinal arteriolar occlusion, branch retinal-artery occlusion, and multiple cottonwool spots.

FINDINGS

In all cases, the ophthalmoscopic signs of retinal ischaemia conformed to a recognisable pattern. The distribution of grey discolouration and dense white opacification depended on the topographical relationship between the boundary of the ischxmic area and the orientation of retinal nerve-fibre bundles. Dense white opacification was only seen where large numbers of nerve fibres entered or emerged from the edge of the ischaemic area. The zone of opacity was either at or close to this boundary, but the remainder of the ischasmic area showed only diffuse grey pallor accentuated around the fovea. No dense opacity was found at the periphery of the infarct either where the nerve-fibre bundles ran parallel to its edge, or along the central horizontal retinal meridian, since no nerve fibres cross the horizontal raphe. Fig. 1 is a fundus photograph from a patient with incomplete obstruction of the central retinal vein in the right eye. Because of pre-existing partial occlusion of the posterior ciliary arterial circulation, the venous occlusion resulted in retinal infarction in the parapapillary cilioretinal arteriolar territory. A zone of dense white opacification is visible along the temporal margin of the infarct, where the orientation of the papillomacular bundle lies perpendicular to the border of the ischaemic area. This zone is less marked along the superior edge of the cilioretinal territory, and is absent along those parts of the inferior edge of the infarct where nerve fibre bundles run parallel to this border. The remaining juxtapapillary cilioretinal territory showed a’ greyish white discolouration.

956 This

of cilioretinal infarction illustrates the interof ruption orthograde axoplasmic flow in axons derived from ganglion-cells in the central arterial territory. The accumulation of organelles in distended axons produces an opaque swollen margin, while juxtapapillary retinal translucency represents intracellular axonal oedema without mitochondrial aggregation. Cessation of retrograde axoplasmic transport due to retinal ischaemia can also be demonstrated. Fig. 2 is a fundus photograph from a patient with an embolic occlusion of an inferotemporal branch retinal artery in the left eye; the ischsemic retina occupies the lower temporal quadrant of the photograph. The retinal infarct has a greyish-white discolouration, though the parafoveal region is more opaque reflecting the increased thickness of the inner retinal layers in this region. However, the appearance of the horizontal and vertical limits of the infarct should be contrasted. A narrow zone of dense white opacity marks the nasal border of the ischaemic area, while at the superior edge of the infarct (along the horizontal raphe), grey retinal translucency gives way almost imperceptibly to normal retinal transparency. The opaque nasal margin of the ischaemic area represents localised axonal distension and organelle aggregation due to interrupted retrograde axoplasmic flow following axonal ischaemia; the ganglion cell bodies lying within the infarct are destroyed. case

DISCUSSION

Although physiologists have been studying axoplasmic transport for many years, the implications of these investigations for ocular disease are only just emerging.9 10 These cases illustrate how the failure of axoplasmic transport due to retinal ischaemia can be observed in the fundus of the eye by simple ophthalmoscopy. Although retinal cottonwool spots commonly result from focal retinal ischaemia, their dense whiteness contrasts with the appearance of the retina after occlusion of the central retinal artery. The complete failure of retinal axoplasmic flow after central arterial occlusion precludes any accumulation of organelles, and is reflected in a grey translucency due to diffuse intracellular axonal oedema." Clinical observations also indicate that the pale swelling of the optic disc in ischaemic optic neuropathy is due to the cessation of orthograde axoplasmic flow in ganglion-cell axons.12 I thank Dr J. F. Cullen and Mr A. C. Bird for permission their patients, and Mr K. Sehmi for the illustrations.

to

describe

REFERENCES

McEwan, B. S., Grafstein, B. J. cell. Biol. 1968, 38, 494 Chou, S. M. Neurology, 1970, 20, 607. 3. La Vail, J. H., Winston, K. R., Tish, A. Brain Res. 1973, 58, 470. 4. Lubinska, L. Progr. Brain Res. 1964, 13, 1. 5. Dollery, C. T., Henkind, P., Paterson, J. W., Ramalho, P. S., Hill, D. W. Br. J. Ophthal. 1966, 50, 285. 6. Shakib, M., Ashton, N. ibid. 1966, 50, 325. 7. Ashton, N., Harry, J. Trans. Ophthal. Soc. U.K. 1963, 83, 91. 8. McLeod, D. Br. J. Ophthal. 1975, 59, 486. 9. Levy, N. S. Invest. Ophthal. 1974, 13, 691. 10. Anderson, D. R., Hendrickson, A. ibid. 1974, 13, 771. 11. Kroll, A. J. Archs. Ophthal. 1968, 79, 453. 12. McLeod, D. Unpublished. 1. 2.

Hypothesis CŒLIAC DISEASE: A CAUSE OF VARIOUS ASSOCIATED DISEASES? BRIAN B. SCOTT

M. S. LOSOWSKY

University Department of Medicine, St. James’s Hospital, Leeds LS9 7TF

Deposition in other organs of immune complexes originating from the small-intestinal mucosa is suggested as a possible reason, in some patients, for the described association between cœliac

Summary

disease and

a

range of "autoimmune" diseases. INTRODUCTION

IN coeliac disease (c.D.) gluten causes small-intestinal mucosal damage. Although the underlying mechanism is not known, there is much evidence that it has an immunological basis, and immune complexes have been demonstrated in the mucosa both after gluten challengel2 and in untreated coeliac disease.3 There are reported associations between c.D. and a number of diseases of unknown aetiology in which immunological mechanisms are thought to be involved (so-called "autoimmune" diseases). The mechanism of these associations has not been explained; it seems unlikely that they are all fortuitous, and a common underlying immunological defect has been considered.4This paper suggests, partly by analogy with the relation between c.D. and dermatitis herpetiformis (D.H.), that small-intestinal mucosal damage may sometimes be the basis of disease in other organs as a result of deposition there of circulating immune complexes originating from immunological reactions in the small-intestinal mucosa. This possibility has

important therapeutic implications. Caeliac Disease and Dermatitis Herpetiformis The association between these two conditions is close in that most (probably at least 4 out of 55) patients with D.H. have c.D. In the peripheral blood of patients with D.H., immune complexes are common and tend to disappear when gluten is removed from the diet.6 The skin lesions contain deposits of immunoglobulin, particularly IgA, and C3 complement,suggesting the deposition of immune complexes. The skin lesions also improve on a gluten-free diet.89 These findings are compatible with the hypothesis that circulating immune complexes, originating from an immunological response in the smallintestinal mucosa, are deposited in the skin under certain conditions, causing the skin lesion of D.H. Ccp/!’ac Disease and other Associated Diseases The diseases (usually regarded as autoimmune in origin) shown in the accompanying table have been reported in association with either definite or possible c.D., although the frequencies of the associations are un-

known. The confidence with which c.D. can be postulated as a related disease varies greatly in these reports. Some of the reports are of patients with small-intestinal villous atrophy in whom either the diagnosis of c.D. was