j. Comp. Path. 1995 Vol. 112, 283 298

Pathological Changes in the Renal Interstitial Capillaries of Pigs Inoculated with Two Different

Strains of African Swine Fever Virus

j . c . G6mez-Villamandos, J. Herv/ts, A. M6ndez, L. Carrasco, C.J. Villeda*, P.J. Wilkinsont and M. A. Sierra

Departamento de Anatomla Patoldgica, Facultad de Veterinaria, Universidad de Cdrdoba, Avda. Medina Azahara 7, 14005 C6rdoba, Spain, *Centro de Biologla Molecular (CSIC-UAM),

Facultad de Ciencias, Universidad Autdnoma de Madrid, Spain and "~AFRC Institute for Animal Health, Pirbright Laboratory , Pirbright, Surrey, UK

Summary African swine fever is a viral disease of pigs characterized predominantly by haemorrhagic lesions. This paper reports the lesions observed in the renal interstitial capillaries of pigs inoculated with African swine fever virus strains of differing virulence: the Malawi'83 strain (haemadsorbent and highly vir- ulent) and the Dominican Republic'78 strain (haemadsorbent and moderately virulent).

In pigs infected with the Malawi'83 strain, petechial haemorrhages and microhaemorrhages were observed 5 days after inoculation and lesions were evident in the renal capillaries. Signs of phagocyte activation were noticeable in endothelial cells, with enlarged fenestrations and even loss of endothelium, leaving the basement membrane of the vessels exposed. Platelet plugs and microthrombi were also observed in these vessels. At 7 days after inoculation these lesions had intensified, and were accompanied by virus replication in the endothelial cells. In pigs infected with the Dominican Republic'78 strain, haemorrhages were more abundant and more extensive, and although no endothelial cell lesions were observed, there was intense vasodilation with diapedesis of erythrocytes.

Introduction Much of the interest expressed in viral haemorrhagic fevers (VHFs) has been directed at the origin and pathogenesis of the haemorrhages that these produce. African swine fever (ASF), which is caused by a DNA virus, has not yet been classified (Madeley, 1990). All VHFs are characterized by haemorrhages and primary replication of the virus in the mononuclear phagocyte system (MPS) (Hess, 1971; Pan, 1987) and all, including ASF, have similar pathophysiological mechanisms (Anderson et al., 1987). A study of the lesions of ASF and their functional implications might therefore yield information relevant to other VHFs.

Renal haemorrhages are a consistent result of the inoculation of pigs with virulent or moderately virulent strains of the ASF virus (ASFV) (Dardiri and Hess, 1970) and have been described as pathognomonic (Edwards et al., 1984; Mebus, 1987). This lesion, together with the distinctive nature of the renal

0021-9975/95/030283 + 16 $08.00/0 �9 1995 Academic Press Limited

284 J . c . G6mez-Villamandos et al.

vascular complex (Zollinger and Mihatsch, 1978), has focused much attention on the morphopathogenesis of vascular lesions produced by ASF in the kidney, even though this organ is not a site of primary or secondary ASFV replication (Plowright et al., 1968).

Biochemical data on the functioning of endothelial cells during experimental infection of pigs by virulent strains of ASFV have pointed to a change in their function from the 4th day after inoculation (Edwards et al., 1984; Anderson et al., 1987), which may well be associated with morphological changes in these cells, although this has not been proved. On the other hand, infection with moderately virulent strains is not associated with major variations in endothelial cell function (Anderson et al., 1987), though these strains produce serious bleeding in the kidney and in other organs (Moulton and Coggins, 1968; Wilkinson, 1989).

The present article describes the sub-cellular pathology of renal interstitial capillaries of pigs inoculated with highly and moderately virulent ASFV strains, in order to establish the morphopathogenesis of haemorrhages produced by these strains.

M a t e r i a l s a n d M e t h o d s

Animals, Hruses and Experimental Design

Large White x Landrace cross-bred pigs (n = 22) of both sexes were used for this study. The animals weighed approximately 30 kg at the beginning of the experiment and were free from infectious and parasitic diseases. They were fed with Prime Grower (Scats Ltd, Winchester, UK) and housed in environmentally controlled isolation rooms. Two of the pigs were used as uninfected controls. Eight pigs were inoculated by the intramuscular route with 105 50% haemadsorbing doses (HADs0) of the Malawi'83 (MW'83)ASFV strain (Haresnape, 1984) and 12 animals with 10 ~ HAD50 of the Dominican Republic'78 (DR'78) ASFV strain (Mebus et aI., 1978). The animals showed no changes in behaviour caused by the experimental conditions and the clinical signs observed were characteristic of ASF (Colgrove et al., 1969; Wilkinson, 1989). The inoculated pigs were killed in pairs at 1, 3, 5 or 7 days post-inoculation (dpi) with the MW'83 ASFV strain, and at 3, 5, 7, 9, 11, 13 (one animal) or 15 (one animal) dpi with the DR'78 ASFV strain. This experiment was carried out in the Institute for Animal Health, Pirbright (UK) in accordance with the Code of Practice for the Housing and Care of Animals used in Scientific Procedures.

Processing of Specimens for Light and Electron Microscopy (EM)

Tissues were fixed by vascular perfusion with glutaraldehyde 2"5% in 0" 1 M phosphate buffer (pH 7"4) at a pressure of 120 mmHg. Before perfusion, the animals were sedated with azaperone (Stresnil| (Jannsen Animal Health, Belgium) and anaesthetized with thiopental-sodium (Thiovet | (Vet Limited, England). Samples from perfused animals were embedded in paraffin wax or Epon 812 (Fluka, Buchs, Switzerland). Paraffm- wax sections were stained with haematoxylin and eosin (HE). Thick sections (1 gm) were cut and stained with toluidine blue (1% aqueous solution). For transmission electron microscopy, 50-nm sections of kidney were stained with uranyl acetate and lead citrate and examined with a Philips CM-10 transmission electron microscope. For scanning electron microscopy, glutaraldehyde-fixed samples of kidney were dehydrated through a graded series of acetone to iso-amyl acetate, critical point dried, coated with gold and examined with a Jeol JSM-6300 scanning electron microscope.

Renal Capi l lar ies in African Swine Fever 285

Morphometrical Study Several measurements of the capillary lumina were ~erformed on HE-stained sections from each of 45 medullary renal areas (0"25mm ; magnification, x250) of each control and inoculated pig. For this purpose a video-based computer linked automatic image-analysis system was used. The images were captured with a TV camera (model VK-C 150ED; Hitachi, Japan) with a 50 mm objective, projected on a monitor screen and measured by means of a Morphometry Program IMAGO developed by the Research Team SIVA (University of C6rdoba, Spain).

Statistical analysis of the cross-sectional areas of the capillary lumina was carried out with the statistical packet Statistical Analysis System (SAS). Means (x) and standard deviations (SD) were calculated. Differences were tested for significance by Student's t-test.

R e s u l t s

Pigs Inoculated with Strain MW'83

Macroscopicalfindings. The kidneys showed a few petechial haemorrhages at the cortico-medullary junction and hyperaemia of the medullary blood vessels at 5 dpi. The haemorrhages had increased in number at 7 dpi, spreading towards the inner medulla and pelvis.

Structural and ultrastructuralfindings. At 5 days, various lesions were observed in the interstitial capillaries. The lumina of some capillaries were empty as a consequence of vascular perfusion, but others contained a few circulating cells (erythrocytes and leucocytes), and the lumina of a small group of capillaries were completely occupied by cells, cellular debris and fibrin, forming microthrombi.

Cells observed in the capillary lumina were mainly monocytes, lymphoid cells and neutrophils, in addition to a limited number of platelets and erythrocytes; their morphology was normal, except that some erythrocytes had invaginations containing virions. A few monocytes displayed the cy- topathogenic effects of ASFV replication, with rounding of the nucleus, margination of chromatin and vacuolization of the cytoplasmic vacuolar system. Semi-thin sections of these cells revealed intracytoplasmic ASFV inclusion bodies, identifiable by pale-staining juxtanuclear cytoplasmic areas. Electron microscopy showed these bodies to be ASFV replication sites, char- acterized by organelle-free cytoplasmic areas with elongated membrane struc- tures and hexagonal viral particles of 175-190 nm diameter; mature, but not immature, particles had an electron-dense nucleoid (Fig. l a). Virions were released from infected cells by budding. At 7 dpi, circulating cells were more numerous, particularly lymphocytes and monocytes, which displayed necrosis associated with ASFV replication. In some interstitial capillaries, whose lumina showed no signs of cellular debris and few circulating cells, there were monocytes showing intense cytopathogenic effects and replicating virus, joined to the cytoplasmic membrane of the endothelial cells by membrane con- densation similar to that of cell junctions (Fig. lb). Some neutrophils were degranulated and contained phagosomes. Examination of the cytoplasm of these cells revealed ASFV particles, either free or within phagosomes at 7 dpi.

In other interstitial capillaries, the lumina were occluded by platelet plugs

286 J.C. G6mez-Vi l l amandos e t all.

Fig. 1. (a) MW'83 strain, 5 dpi. Medullary capillary with erythrocytes, lymphocytes (L) and monocytes ~M). A monocyte shows ASFu replication site (arrowheads). EM. Bar, 2 gm. Inset: Immature virions and membranous structures in the viral replication site. Bar, 200 nm. (b) MW'83 strain, 7 dpi. Monocyte with ASFV replication site (*) attached to normal endothelial cell of a cortical capillary by membrane condensations (arrows). There is an ASFV particle in the subendothelium (arrowhead). EM. Bar, 1 gm.

composed of enlarged, degranulated platelets with extended pseudopodia (Fig. 2a) and by microthrombi composed of cells, cellular debris and virions surrounded by fibrinous networks (Fig. 2b).

In some areas, the endothelial cells of a proportion of the interstitial capillaries had long cytoplasmic projections which surrounded cellular debris and erythrocytes, thus considerably increasing the size of endothelial cells. There was also an increase in primary and secondary lysosomes in these cells, as well as a notable increase in micropinocytotic vesicles compared with the endothelial cells from control swine and those killed at 5 dpi. Despite this enlargement, most such cells retained their junctions with neighbouring endothelial cells; clearly formed intercellular junctions were observed between the cytoplasmic projections (Fig. 3). However, in other capillaries almost all of the endothelial lining showed increased fenestration and a loss of junctions between endothelial cells, with erythrocytes, cellular debris and the occasional

R e n a l C a p i l l a r i e s i n A f r i c a n S w i n e Fever 287

Fig. 2. (a) MW'83 strain, 5 dpi. Plug of degranulated platelet and cellular debris in a capillary with loss of endothelial ceils and limited by a basement membrane (arrows). EM. Bar, 2 I-tm. (b) MW'83 strain, 7 dpi. Networks of fibrin (arrows) which are associated with cellular debris and an ASFV particle (arrowhead) in a medullary capillary. The endothelial ceils of this capillary are ultrastructurally normal. EM. Bar, 1 pm.

viral particle between the endothelial cells and the basal lamina of the capillary (Fig. 4a).

The most severe lesion noted in the interstitial capillaries at 5 dpi was loss of endothelial cells (Figs 2a, 4a) and disorganization of the basement membrane; erythrocytes and cell debris had moved towards the interstitium, producing haemorrhages (Fig. 4b). In these and the previously ment ioned capillaries, large platelets showing various degrees of degranulation adhered to the subendothelial connective tissue by membrane condensations (Fig. 5). At- tachment of platelets to ultrastructurally normal endothelial cells was also observed. Animals killed at 7 dpi showed similar lesions, but with a notable increase in the intensity and extent of interstitial capillary damage. At this stage of the infection, the main observation was that numerous endothelial cells in the cortical and medullary interstitial capillaries revealed cyto- pathogenic effects and virus replication sites. The presence of viral particles at different stages of maturi ty was also noted; these showed evidence of budding, towards both the vascular lumen and the surrounding interstitial tissue (Fig. 6).

288 J . C . Gbmez-Vi l lamandos e t al.

Fig. 3. (a) MW'83 strain, 5 dpi. Cortical capillary with swollen and activated endothelial cells. These cells are characterized by cell junctions with other endothelial cells (arrows); intracytoplasmic cellular debris and proliferation of lysosomes and pseudopods. EM. Bar, 2 gm.

In the interstitial tissue at 5 dpi, the changes seen were moderate-to-intense oedema, microhaemorrhages and cellular infiltrates, mainly composed of macrophages, neutrophils and a few lymphocytes and eosinophils. At 5 dpi, and more frequently at 7 dpi, some macrophages displayed cytopathogenic effects and ASFV replication sites (Fig. 7). Signs of necrosis in cells infiltrating the interstitium were present from 5 dpi, but such necrosis was particularly intense at 7 dpi. Cellular debris was observed both extracellularly and within macrophages.

Pigs Inoculated with Strain DR'78

Macroscopicalfindings. The first lesions noted (7 dpi) consisted of a small number of petechial haemorrhages; these were more numerous in the medulla and in the deeper part of the cortex, and became more widespread as the disease ran its course, with petechiae all over the kidney and suffusion in the medulla and pelvis at 9 dpi. Haemorrhage intensified as the disease progressed.

Structural and ultrastructural findings. From 7 dpi, microhaemorrhages made their appearance, but there were no morphological changes suggestive of

Renal Capillaries in African Swine Fever 289

Fig. 4. (a) MW'83 strain, 5 dpi. Two medullary capillaries. In one capillary the endothelial cells are slightly swollen and contain intracytoplasmic debris (E), and there is an erythrocyte and cell debris between the endothelial cells and basement membrane (arrows). The other vessel shows loss of endothelial cells and is limited by basement membrane only (arrowheads). EM. Bar, 2 gm. (b) MW'83 strain, 5 dpi. Capillary with loss of endothelial cells and disorganization of basement membrane (arrows); erythrocytes and cell debris are present in the interstitium. EM. Bar, 2 gm.

disruption of endothelial cells in the interstitial capillaries; nor were there lesions indicating direct action of the virus on these cells (Figs 8 and 9). However, there were erythrocytes with extended cytoplasmic projections, producing signs of diapedesis and causing cytoplasmic invagination. The area of the cytoplasm of the endothelial cells next to the erythrocyte was free of organelles and with abundant micropinocytotic vesicles (Fig. 10). Separation was greater between endothelial cell nuclei than in the MW'83 experiment and in the control animals. In some interstitial capillaries, lymphocytes, monocytes and neutrophils were attached to the endothel ium, with cytoplasmic projections extending through the endothelial lining (Fig. 8); these cells, and also erythrocytes, were occasionally seen inside the cytoplasm of the endothelial cells, as well as between the endothelial cell and the capillary basement membrane . Diapedesis and extravascular blood cells were not observed in control pigs and pigs killed at 3 and 5 dpi.

Monocytes contained large amounts of cellular debris, which caused them

290 J.C. G 6 m e z - V i i l a m a n d o s e t al.

Fig. 5. MW'83 strain, 5 dpi. A degranulated pIatelet with a pseudopodium attached to a capillary basement membrane by junction-like structures (arrowheads). EM. Bar, 500 nm.

to swell considerably (Fig. 8). Nevertheless, examination of all the fields (700 fields at 90 x 90 gm) revealed only one viral particle, which was associated with an erythrocyte, and no virus replication site was observed.

Vasodilation, with increased distance between endothelial cell nuclei, and the at tachment of circulating cells to the endothelium were confirmed by scanning electron microscopy, as was a "stretching" of the vascular bed.

In the interstitial tissue, accompanied by ever-increasing perivascular haem- orrhages and oedema which at 11 dpi were capable of causing the destruction of the renal interstitium, there was an abundant infiltrate of mononuclear cells around the capillaries (Figs 8 and 9). This infiltrate included numerous lymphocytes, plasma cells and large cells with rounded nuclei and prominent nucleoli; the latter cells had a few cytoplasmic organelles, notable among which were rough endoplasmic reticulum cisternae, and were identifiable as stimulated lymphoid cells. There were also enlarged macrophages showing phagocytosis of cellular debris and erythrocytes, but with no signs of ASFV replication (Fig. 9).

As the disease progressed, vasodilation, perivascular cell infiltration and

R e n a l C a p i l l a x i e s i n A f r i c a n S w i n e Fever 291

Fig. 6. MW'83 strain, 7 dpi. Endothelial cells with ASFV replication site and virus budding towards the interstitium and vascular lumen. EM. Bar, 500 nm.

oedema, haemorrhage, a t tachment of mononuclear cells to the endothel ium and the ensuing diapedesis all increased, as did erythrocyte diapedesis.

Morphometrical and Statistical Results with the DR'78 and MW'83 ASFV Strains

The vascular renal lesions resulting from inoculation of the DR'78 strain suggested the existence of interstitial capillary vasodilation. For that reason, statistical analysis was performed on the morphometr ical results for the medullary capillary area of pigs inoculated with the DR'78 strain. This analysis confirmed that there had been a striking increase in the luminal size of these vessels from 7 dpi (P<0"001) (Fig. 11).

A similar study was performed on the less damaged interstitial capillaries (without microthrombi, plugs of platelets or severe endothelial lesions) of the pigs inoculated with MW'83 strain, but no significant values were obtained.

D i s c u s s i o n

The present study revealed renal haemorrhages in the pigs inoculated with the virulent strain MW'83 from 5 dpi. At that time no replication of the virus

292 J . C . G 6 m e z - V i l l a m a n d o s et aL

Fig. 7. MW'83 strain, 5 dpi. Cortical interstitium shows macrophages with ASFV replication sites (*). EM. Bar, I gin.

in the endothelial cells of the interstitial capillaries was observed. ASFV replication in these cells was therefore rejected as the cause of haemorrhage in ASF, a cause by some authors (Maurer et al., 1958; Colgrove et al., 1969; Sierra et al., 1989). However, coinciding with the presence of haemorrhages, there was endothelial damage, consisting of phagocytic activation of the endothelial cells, increased capillary fenestration and even necrosis and loss of endothelial cells; the latter lesions resulted in exposure of the capillary basement membrane, to which activated platelets were attached. This may be one of the causes of the disseminated intravascular coagulation (DIC) characteristic of ASF (Pan, 1987; Villeda et al., 1993). The endothelial lesions described constitute the morphological manifestation of the biochemical values indicative of endothelial dysfunction in acute ASF, such as high levels of factor VIII (Edwards et al., 1984) and a reduction in prostacyclin (Anderson et al., 1987).

The absence of a resident macrophage population in renal interstitial capillaries may influence the phagocytic activity of endothelial cells in these vessels. Previous studies have shown an absence of phagocytic activity in endothelial cells of the liver (Sierra et al., 1987), spleen (Konno et al., 1972)

Renal C a p i l l a r i e s i n African Swine Fever 293

Fig. 8. DR'78 strain, 7 dpi. Dilated capillaries with activated monocytes (M), a lymphocyte (L) and a neutrophil (N) adherent to endothelial cells. Peripheral interstitium is oedematous and haem- orrhagic. EM. Bar, I0 btm.

and lung (Sierra et al., 1990) in ASF, as these organs have resident macrophage populations which clear the vascular lumen of cellular debris. The cellular debris, which undergoes phagocytosis by macrophages, arises from the intense necrotic activity occurring in the liver (Sierra et al., 1987), spleen (Konno et al., 1972; Minguez et al., 1988) and lymph nodes (Minguez et al., 1988), starting at 3-4 dpi with virulent strains; but necrosis in the kidney is a rare phenomenon and only present at 7 dpi (G6mez-Villamandos et al., 1995).

Cytokines, released by macrophages and monocytes, may play an important part in the establishment of DIC (Bevilacqua et al., 1986; Halstead, 1989). Macrophages and monocytes showing ASFV replication were noted in the present experiment from 5 dpi, and it has been suggested that cytokines released by these cells occur in ASF (Villeda et al., 1993). Two important cytokines are interleukin-1 and tumour necrosis factor-a, to which a major role has been attributed in the pathogenesis of the DIC present in the VHFs, such as Argentine haemorrhagic fever, Rift Valley fever and dengue haemorrhagic fever. In these diseases the virus replicates in mononuclear phagocytic system (MPS) cells and not in endothelial cells (Halstead, 1989;

294 J .c . G6mez-ViUamandos et al.

Fig. 9. DR'78 strain, 9 dpi. Cortical interstitium with haemorrhage and cellular infiltrate of lymphocytes, plasma cells and macrophages (M) with phagocytic cell debris. The interstitial capillary (C) is empty and the endothelial ceils are uhrastructarally normal. EM. Bar, 5 p,m.

Molinas et al., 1989; Peters et al., 1989). The attachment of platelets and monocytes to ultrastructurally normal endothelial cells and the activation of endothelial cells from 5 dpi, coinciding with virus replication in nearby macrophages and monocytes, suggest that these endothelial cells undergo changes at the molecular level, perhaps due to the action of cytokines, which induce the expression of the adhesion molecules of leucocytes and platelets (Grau and Lou, 1993) and activation of the endothelial cells (Cavender et al., 1989).

The kidneys of pigs inoculated with the DR'78 strain revealed intense, widespread haemorrhage from 7 dpi, but neither endothelial lesions nor intravascular coagulation was observed. Morphometrical and ultrastructural data suggest that these haemorrhages are associated with intense vasodilation; this would be expected to give rise to increased vascular permeability, which in turn would be responsible for the intense interstitial oedema and diapedesis observed in this study. Diapedesis of erythrocytes would seem to be the cause of haemorrhage in the kidneys of pigs inoculated with moderately virulent strains; this mechanism has been suggested as the cause of the intestinal

R e n a l C a p i l l a r i e s i n A f r i c a n S w i n e F e v e r 2 9 5

Fig. 10. DR'78 strain, 7 dpi. Diapedesis of an erythrocyte in a cortical capillaly. EM. Bar, 200 nm.

haemorrhage in turkey haemorrhagic enteritis (Saunders et al., 1993). Di- apedesis of mononuclear cells gives rise to intense perivascular infiltration of macrophages, lymphocytes, lymphoblasts, plasma cells and neutrophils. In the infiltrates the macrophages appear to be activated, leading to phagocytosis of cell debris and erythrocytes. The transmigration of cells from capillaries is a process that may also be mediated by activated monocytes and macrophages or by immunological mechanisms, or both (Mantovani and Dejana, 1989). Monocytes and macrophages were activated but no virus replication was observed in these cells in the pigs inoculated with strain DR'78. On the other hand, immunological mechanisms may have played a role in DR'78-infected animals, given the lympho-plasmocytic nature of the cell infiltrate and the existence of proliferative mesangial glomerulonephritis (unpublished data) from 7 dpi. These lesions are associated with the deposition ofimmunoglobulins and complement (C3) (Muller-Peddinghaus and Trautwein, 1974; Miyanchi et al., 1992), and they may reflect the high rates of immunoglobulin detection from 5 to 6 dpi in swine inoculated with attenuated strains (S/mchez-Vizcaino et al., 1981).

The absence of intravascular coagulation in the kidneys of pigs killed after inoculation with DR'78 may have been due to the absence of endothelial

296 J.C. G 6 m e z - V i l l a m a n d o s e t aL

1600 I I [ [ I I I I

1400 / I 1200 1000 800 600

_

200 ~ / ~ - -

0 1 3 5 7 9 11 13 15

Days post- inoculat ion

Fig. 11. Measurements (mean _ SD) of the luminal areas of the capillaries of pigs killed at 0 dpi (control animals) and pigs inoculated with strain DR'78.

lesions and not to ASFV replication in MPS cells present in the kidney. To support this hypothesis, it should be noted that DIC in animals receiving strains of moderate virulence has been described in pigs that had died of the disease; before death they had developed symptoms and lesions similar to those described for pigs inoculated with virulent strains of ASFV, including a high rate of virus replication in MPS cells (Pan, 1987; FernAndez et al., 1992). This is reinforced by the results with DR'78-infected animals dying at 8-10 dpi (unpublished data); the intravascular coagulation in these animals might therefore have been attributable to the same mechanisms that the authors described for the MW'83 strain.

Morphological results obtained in the present study suggest that highly virulent strains of ASFV produce renal haemorrhages as a result of intense endothelial damage, influenced by the phagocytic activation that occurs in endothelial cells to ensure that cell debris from elsewhere is removed from the circulation, virus replication in endothelial cells playing only a secondary role. With strains of moderate virulence, haemorrhage is a consequence of an increase in vascular permeability with diapedesis of erythrocytes, in which immunological processes may play a main role.

Acknowledgments We thank Prof. Dr A. Jover and Professor Dr E Vifiuela for their support, Professor Dr T. Moyano, Chief of Electron Microscopy Service of the UCO, F. Gracia and M.J . Gutierrez for the morphometrical study, C. Romero, G. H. Hutchings and S. M. Williams for technical assistance in the laboratory and T. Kinsella and L. Fitzpatrick for assistance in the animal unit. This work was supported by grants from

Renal Capillaries in African Swine Fever 297

CAICYT, Junta de Andalucla and Junta de Extremadura (Spain) and the Ministry of Agriculture, Food and Fisheries (UK).

R e f e r e n c e s

Anderson, E. C., Williams, S. M., Fisher-Hoch, S. P. and Wilkinson, P.J . (1987). Arachidonic acid metabolites in the pathophysiology of thrombocytopaenia and haemorrhage in acute African swine fever. Research in Veterinary Science, 42, 387-394.

Bevilacqua, M. P., Pober,J. S., Majeu, G. R., Fiers, W., Cotran, R. S. and Grimbrone, M. A. (1986). Recombinant tumor necrosis factor induces procoagulant activity in cultured human vascular endothelium. Characterization and comparison with the actions of interleukin-1. Proceedings of the National Academy of Sciences USA, 83, 4533.

Cavender, D. E., Edelbaum, D. and Ziff, M. (1989). Endothelial cell activation induced by tumor necrosis factor and lymphotoxin. American Journal of Pathology, 134, 551 560.

Colgrove, G. S., Halterman, E. O. and Coggins, L. (1969). Pathogenesis of African swine fever in young pigs. American Journal of Veterinary Research, 30, 1343-1359.

Dardiri, A. H. and Hess, W. R. (1970). Pathogenesis of African swine fever in pigs. Proceedings of the US Animal Health Association, 74, 213-224.

Edwards, J. F., Doods, W. J. and Slauson, D. O. (1984). Coagulation changes in African swine fever virus infection. American Journal of Veterinary Research, 45, 2414-2420.

Fern/mdez, A., P6rez, J., Carrasco, L., Bautista, M.J. , Sfinchez-Vizcaino, J. M. and Sierra, M. A. (1992). Distribution of ASFV antigens in pig tissues experimentally infected with two different Spanish virus isolates. Journal of Veterinary Medicine (Series B), 39, 393-402.

G6mez-Villamandos, J. C., Herv~s, J., M~ndez, A., Carrasco, L., Villeda, C., Wil- kinson, P.J. and Sierra, M. A. (1995). Ultrastructural study of the renal tubular system in acute experimental African swine fever: virus replication in glomerular mesangial cells and in the collecting ducts. Archives of Virology, in press.

Grau, G. E. and Lou, J. (1993). TNF in vascular pathology: the importance of platelet- endothelium interactions. Research in Immunology, 144, 355-363.

Halstead, S. B. (1989). Antibody, macrophages, dengue virus infection, shock and hemorrhage: a pathogenic cascade. Reviews of Infectious Diseases, 11,830-839.

Haresnape, J. M. (1984). Isolation of ASFV from ticks of the Ornithodorus moubata complex (Ixodoidea: Argasidae) collected within the ASF enzootic area of Malawi. Epidemiology and Infection, 101, 173-185.

Hess, W. R. (1971). African swine fever virus. In: Virology Monograph, Springer-Verlag, New York, pp. 1-33.

Konno, S., Taylor, W. D., Hess, W. R. and Heuschele, W. P. (1972). Spleen pathology in African swine fever. Cornell Veterinarian, 3, 486-506.

Madeley, C. R. (1990). The morphology and structure of viruses. In: Topley and I4~lson's Principles of Bacteriology, l~rology and Immunity, 8th Edit., Vol 4, M. T. Parker and L. H. Collier, General Eds. and L. H. Collier and M. C. Timbury, Vol. Eds, Edward Arnold, London, pp. 11-39.

Mantovani, A. and Dejana, E. (1989). Cytokines as a communication signal between leukocytes and endothelial cells. Immunology Today, 10, 370-375.

Maurer, F., Grisemer, R. and Jones, T. C. (1958). The pathology of African swine fever. A comparison with hog cholera. American Journal of Veterinary Research, 19, 517-539.

Mebus, C. A. (1987). Pathobiology and pathogenesis. In: African Swine Fever, Y. Becker, Ed., Martinus Nijhoff Publishing, Boston, pp. 21-30.

Mebus, C. A., Dardiri, A. H., Hamdy, F. M., Ferris, D. H., Hess, W. R. and Callis, J .J . (1978). Some characteristics of African swine fever virus isolated from Brazil

298 J . c . G6mez-ViUamandos e t aL

and the Dominican Republic. Proceedings of the US Animal Health Association, 82, 232-236.

Minguez, I., Rueda, A., Dominguez,J. and S/mchez-Vizcaino. (1988). Double labelling immunohistological study of African swine fever virus infected spleen and lymph nodes. Veterinary Pathology, 25, 193-198.

Miyanchi, Y., Nakayama, H., Uchida, K., Uetsuka, K., Hasegawa, A. and Goto, N. (1992). Glomerulopathy with IgA deposition in the dog. Journal of Veterinary Medicine Sciences, 54, 969-975.

Molinas, F. C., Bracco, M. M. E. and Maiztegui, J. I. (1989). Hemostasis and the complement system in Argentine haemorrhagic fever. Reviews of Infectious Diseases, 11, 762-770.

Moulton, J. E. and Coggins, L. (1968). Comparison of lesions in acute and chronic African swine fever. CorneU Veterinarian, 58, 364-388.

Muller-Peddinghaus, R. and Trautwein, G. (1974). Spontaneous glomerulonephritis. Classification and immunopathology. Veterinary Pathology, 14, 1-13.

Pan, I. C. (1987). Spontaneously susceptible cells and culture methodologies for African swine fever virus. In: African Swine Fever, Y. Becker, Ed., Martinus Nijhoff Publishing, Boston, pp. 81-126.

Peters, C.J., Liu, C. T., Anderson, G. W., Morrill, J. C. andJahrling, P. B. (1989). Pathogenesis of viral haemorrhagic fever. Reviews of Infectious Diseases, 11, 743-749.

Plowright, W., Parker, J. and Staple, R. F. (1968). The growth of a virulent strain of swine fever virus in domestic pigs. Journal of Hygiene, Cambridge, 66, 117 134.

SAnchez-Vizcaino, J. M., Slauson, D. O., Ruiz-Gonzalvo, F. and Valero, F. (1981). Lymphocyte function and cell-mediated immunity in pigs with experimentally induced African swine fever. American Journal of Veterinary Research, 42, 1335-1341.

Saunders, G. K., Pierson, F. W. and van den Hurk, J. V. (1993). Haemorrhagic enteritis virus infection in turkeys: a comparison of virulent and avirulent virus infections, and a proposed pathogenesis. Avian Pathology, 22, 47-58.

Sierra, M. A., Bernabd, A., Mozos, E., Mendez, A. andJover, A. (1987). Ultrastructure of the liver in pigs with experimental African swine fever. Veterinary Pathology, 24, 460-462.

Sierra, M. A., Quezada, M., Fern/mdez, A., Carrasco, L., G6mez-Villamandos, J. C., Martin de las Mulas, J. and S/mchez-Vizcaino, J. M. (1989). Experimental African swine fever: evidence of the virus in interstitial tissues of the kidney. Veterinary Pathology, 26, 173 176.

Sierra, M. A., Carrasco, L., G6mez-Villamandos, J. C., Martin de las Mulas, J., Mdndez, A. andJover, A. (1990). Pulmonary intravascular macrophages in lung of pigs inoculated with African swine fever viruses of differing virulence. Journal of Comparative Pathology, 102, 323-334.

Villeda, C., Williams, S. M., Wilkinson, P.J. and Vifiuela, E. (1993). Haemostatic abnormalities in African swine fever: a comparison of two ASF virus strains of different virulence. Archives of Virology, 130, 71 83.

Wilkinson, P.J. (1989). African swine fever virus. In: Virus Infection of Vertebrates, Vol II; Virus Infections ofPorcines, N. B. Pensaert, Ed., Elsevier, Amsterdam, pp. 17-35.

Zollinger, H. U. and Mihatsch, M.J. (1978). Histology of normal kidney tissue. In: Renal Pathology in Biopsy, H. U. Zollinger and M. J. Mihatsch, Eds, Springer- Verlag, New York, pp. 21 45.

Received, September 7 th, 1994 1 Accepted, December 19th, 1994]