JOURNAL OF PATHOLOGY, VOL.
164 175-1 84 (1 99 1)
PNEUMOCYSTIS CARINII PNEUMONIA IN SIMIAN IMMUNODEFICIENCY VIRUS INFECTION: IMMUNOHISTOLOGICAL AND SCANNING AND TRANSMISSION ELECTRON MICROSCOPICAL STUDIES A. BASKERVILLE, A. B. DOWSETT, R. W. COOK, M. J . DENNIS, M. P. CRANAGE AND P. J . GREENAWAY
Division of Pathology, Public Health Laboratory Service, Centre.fhr Applied Microbiology and Research, Porton Down, Salisburj, Wiltshire. U.K. Received 19 December 1990 Accepted 14 Junuary 1991
SUMMARY Pneurnocystis carinii pneumonia occurred in 6 o f 17 rhesus monkeys infected with simian immunodeficiency virus and was studied by immunohistochemistry and by scanning and transmission electron microscopy. A monoclonal antibody/streptavidin-biotin-peroxidase staining method was highly sensitive for detecting the organisms in small, early lesions and was much more sensitive and specific than traditional silver impregnation methods. Reprocessing of paraffin wax-embedded lung tissue for scanning electron microscopy and use o f a video printer to produce a photographic montage of light microscopic lesions allowed the same areas of tissue to be examined and compared by both methods. The ultrastructural morphology o f P. carinii in the rhesus monkey was identical to that in man, as were the histological and electron microscopic lesions, including pulmonary fibrosis. Trophozoites were seen attached to alveolar type I epithelium mainly by intimate apposition to the plasma membrane, but scanning electron microscopy also showed attachment by elongated filopodia. Few macrophages were present in infected alveoli, and though phagocytosis followed by digestion of P. carinii trophozoites was observed, it appeared to occur at a very low level. KEY
WORDS-Non-human primates, immunodeficiency, Pneumoy3stis pneumonia, immunohistochemistry, electron microscopy.
INTRODUCTION Infection of non-human primates with simian immunodeficiency virus (SIV) produces immunodeficiency, weight loss, lymphadenopathy, and neuropathological lesions,''2 a complex of disease similar to that in human acquired immune deficiency syndrome (AIDS) caused by the related human immunodeficiency lentivirus HIV- 1. SIV infection of macaque monkeys, in particular rhesus, is considered a highly relevant experimental model for the Addressee for correspondence: Dr A. Baskerville, Division of Pathology, Public Health Laboratory Service, Centre for Applied Microbiology and Research, Porton Down, Salisbury, Wiltshire,SP4 OJG, U.K.
0022-341 7/91/050175-10 $05.00 0 1991 by John Wiley & Sons, Ltd
study of AIDS3 because the animals also develop many ofthe secondary bacterial, viral, and protozoal infections which occur in AIDS patients. One of the commonest infections in human AIDS, and a frequent cause of death, is pneumonia due to Pneumocystis carinii and surveys have indicated a prevalence of up to 85 per cent. It was noted in both natural and experimental infection of macaques with SIV that a proportion also suffered from P.cariniipneumonia (PCP),'.5 though detailed pathological and ultrastructural accounts have not been published. This paper presents the findings of immunohistochemical and transmission and scanning electron microscopical studies of different stages of naturally acquired PCP in SIV-infected rhesus monkeys.
176
A BASKERVILLE E T A L .
Table I-Simian
Monkey No. 32H 27H 19H K33 K8 I67 85H 44H 68H 14H 17H 62H 67H 61H 2H 53H 29 H
immunodeficiency viruses used and disease course
Sex
9 P 9 3 d d
9
Y
9 3 P 0 d d
6
s 0
Time to death (months)
Virus SIV,,,25 1 SIVm,,32H* SIVm,,32H/CBLt SIVm,,32H SIVm,,32H SIVm,,32H SIVm,,32H SIVm,,32H SIVm,,32H SIVmd,32H SIVmd,32H SIVm,,32H SIV,,,clone 32H bloodf SIVSM SIVC," 705B blood
Died (D) or killed (K)
18 11 11;
G
14
1@ 5:
4 9 weeks
1st
5 7 15; 27 28 30 21
K K K K K K K D D K K K D K K K K
*SIV,4L32H refers to a virus stock isolated from monkey 32H and adapted to a continuous human T-cell line. tSIVm,,32H/CBL refers to the primary isolate from monkey 32H grown on human cord blood lymphocytes. :Monkey 61H was infected with SIV by direct inoculation of blood from monkey 32H.
MATERIALS AND METHODS Animals Seventeen rhesus monkeys (Macuca mulutta), 8 male, 9 female, bred in the United Kingdom and aged between 2 and 3 years at the time ofinfection, were used. The animals were fed a commercial pelleted diet supplemented with fruit and were given water ad libitum. They were caged singly and housed under containment category P3 conditions.
Viruses The viruses were given intravenously and details are summarized in Table I. Full details of the virus preparation, growth, and methods of isolation from tissues of the monkeys have been presented
Procedures For infection and all other procedures, animals were anaesthetized by intramuscular injection of ketamine hydrochloride ('Vetalar', Parke Davis).
Blood sampling and intravenous inoculation of virus suspension were carried out using the femoral vein. Animals were killed by intravenous injection Of pentobarbitone sodium. Autopsy and histopathology Autopsy was carried out immediately after death except in the cases of the three animals which died, where there was a delay of not more than 12 h. The trachea and lungs were removed in toto. Slices of lung were taken from each lobe and fixed in 10 per cent buffered neutral formalin. After fixation, the tissue was processed by standard methods and embedded in paraffin wax. Sections were cut at 5 pm and stained by haematoxylin and eosin (H&E), periodic acid-Schiff (PAS), Gordon and Sweet's method for reticulin, Grocott's silver methenamine technique, and with anti-Pneumocystis carinii monoclonal antibody (M778, Dako Ltd., High Wycombe, U.K.) using the streptavidin-biotinperoxidase method and diaminobenzidine as indicator. Control tissue was obtained from eight uninfected normal rhesus monkeys of the same age.
PNEUMOCYSTIS CARINII PNEUMONIA IN SIV
Fig. ]-Early infection of monkey lung by p . Stained by monoclonal antibody to the organism and the streptavidinbiotin-peroxidase method. Aleveoli contain dark-staining organisms (arrows)
Transmission electron microscopy One mm3 portions of tissue were taken from consolidated areas of the lungs of two monkeys. These were immediately fixed in cold 4 per cent phosphatebuffered glutaraldehyde at pH 7.2 and later postfixed in 2 per cent buffered osmium tetroxide at 4°C. After processing through a graded series of ethanol and propylene oxide, the tissue was embedded in Araldite. Sections 1 pm in thickness were cut and stained with toluidene blue for location of suitable areas for ultrastructural examination. Ultrathin sections were stained with uranyl acetate and lead citrate, and stabilized by carbon coating in a vacuum coating unit prior to examination in a Philips EM400T transmission electron microscope.
177
Fig. 2-Heavy infection of monkey lung by P. carinii. Alveoli are packed with organisms and interalveolar septa are thickened by fibrous tissue. Monoclonal antibody to p , and
streptavidin-biotin-peroxidase
Scanning electron microscopy Specimens were prepared by the following regimens: (a) Slices of lung, approximately 1 cm2 and 3 4 mm in thickness, were taken from the formalin-fixed tissue. These were post-fixed overnight in 1 per cent buffered osmium tetroxide at 4"C, dehydrated through a graded acetone series, and critical point-dried from liquid carbon dioxide prior to mounting on SEM specimen stubs using silver dag. (b) Cubes of formalin-fixed lung tissue, approximately 4 mm in size, were plunged into liquid nitrogen slush at -210°C and cleaved with a razor blade. Tissue fragments were then postfixed, dried, and mounted as before, care being
178
A BASKERVILLE ET A L .
Fig. 3-Electron micrograph of alveolus containing trophozoites, membranes. and fibrin. Filopodia are visible in section on some surfaces. x 30 000
taken to expose the fractured tissue surface for examination.
as before, and oriented on the specimen stub so as to present for SEM examination the tissue face previously sectioned for light microscopy.
Comparative exurnination ofthe same tissue areas bj, light and scanning electron microscopy
Photographic montages of sections of waxembedded tissue which showed the presence of P. carinii by immunohistochemistry were constructed from micrographs recorded by a Sony UP 5000P video printer from a Leitz Ortholux 2 microscope fitted with a JVC CCD camera. The corresponding tissue blocks were then dewaxed with three 1 h changes of xylene at 60°C and taken from 100 to 30 per cent ethanol, through a graded series, to phosphate buffer. Tissue was then post-fixed and dried
Conductive coating Specimens for SEM were sputter-coated with a 20 nm layer of gold prior to examination in either a Cambridge Stereoscan S2A or a Hitachi S-800 scanning electron microscope.
RESULTS Macroscopic Jindings Of the 17 SIV-infected rhesus monkeys in the study, four had extensive lung lesions of PCP at
PNEUMOCYSTIS CARINII PNEUMONIA IN SIV
Fig. &Infected
179
alveolus with Pneurnocvstis cyst containing sporozoites. x 16 000
necropsy (numbers 14H, 17H, 62H, and K8). The lesions consisted of a characteristic, uniformly grey consolidation on the surface of the lungs and in cut slices, and involved virtually the whole of each lobe. The lungs of all other SIV-infected monkeys and negative controls showed no macroscopic abnormalities.
Histopathological and immunocytochemicalfindings Six animals had PCP lesions: the four with extensive pneumonia recognized at autopsy and a further two (K33 and 61H) in which the lesions were microscopic. In sections stained by H&E, the smallest lesions appeared as a few scattered foci of infiltration of a group of interalveolar septa or even single alveoli by polymorphonuclear leucocytes (PMN) and
macrophages. These foci frequently did not have any floccular material in the alveoli and P. carinii infection would not have been suspected from H&E sections alone. However, immunostaining using a monoclonal antibody to human P. carinii revealed small numbers of oval or circular organisms in these foci (Fig. 1). They were also found in alveolar walls or on alveolar epithelium with no accompanying cellular response or abnormality of alveolar cells. There was noconsistent distribution ofthe organisms in the lobe. In the four cases with massive lung involvement by PCP, the majority of alveoli were distended by foamy eosinophilic PAS-positive material, and staining with monoclonal antibody demonstrated these to be densely packed by P. carinii cysts and trophozoites (Fig. 2). Staining of consecutive sections by Grocott's method for comparison with the
180
Fig. 5-Alveolus
A BASKERVILLE ET A L .
showing intimate association between two trophozoites (T) and type I epithelial cell. Cap: capillary. x 27 300
monoclonal antibody revealed only a small proportion of the organisms as argyrophilic. In animals with heavy infestation, the P. carinii were distributed widely in the lung tissue. By monoclonal antibody staining most were detected in alveoli, but some were also present in alveolar walls, on the epithelial surface, and in the lumen of bronchi and at all levels of the smaller airways. Occasionally, a few P. carinii were found in lymphatics and veins, both free and within phagocytic cells. In the established cases, the local cellular alveolar response was minimal, though alveolar walls were thickened by fibroblasts, macrophages, and lymphocytes (Fig. 2). The alveolar epithelium in some areas had changed to a basophilic, cuboidal, or even columnar type.
Transmission electron microscopy ( T E M ) Lung tissue was taken from two monkeys with extensive lesions of PCP for TEM studies and the findings were identical in both. Large numbers of thin-walled trophozoites of P. curinii completely filled many alveoli and in others formed deep layers over the alveolar type I epithelium. The trophozoites varied considerably in shape as a result of compression and interlocking with adjacent parasites (Fig. 3). Their internal structure consisted of
finely granular cytoplasm, with irregularly scattered larger electron-dense granules and a nuclear region with an ill-defined membrane and a prominent nucleolus. Other organelles were not identified and many trophozoites appeared empty. Thick-walled cysts were also present, some of which contained sporozoites (Fig. 4) and a zone of thickening of the inner wall. Collapsed cysts were also observed, appearing as crescents (Fig. 3), but with the thickened zone still intact. Short, tubular projections were present on most of the parasites. These processes have been termed filopodia7~* and were all approximately 85 nm in diameter. While filopodia at some sites were closely applied to the surface of adjacent trophozoites, alveolar epithelium, or macrophages, many contacts took place in the absence of filopodia. Close examination of the sites of attachment of trophozoites to the alveolar wall showed intimate contact between the plasma membrane of the type I epithelial cell and the membrane of the parasite, which appeared to adapt its shape to the contours of the cell (Fig. 5). In some cases, contact was over a very small area, but in others much of the involved surface of the trophozoite was applied to the epithelium. The same close contact occurred between the outer membrane of adjacent trophozoites and
PNEUMOCYSTIS CARINII PNEUMONIA IN SIV
Fig. &Alveolar macrophage containing disintegrating phagocytosed trophozoites. x 19 000
cysts, so that a composite colony of parasites was built up which appeared to be firmly anchored onto the alveolar surfaces. The cellular responses which occurred in affected alveoli were minor. Fibrin clumps, cell debris, and surfactant membranes were present in a few alveoli. Macrophages had phagocytosed trophozoites, which were identifiable within phagosomes (Fig. 6), many undergoing lysis. Occasional macrophages were seen which were surrounded by densely packed trophozoites but were not phagocytosing them, and the numerous pseudopodia appeared compressed against the cell surface. The cuboidal metaplasia of alveoli observed by light microscopy wasshown to be due to replacement of the type I epithelial cells by large numbers of electron-dense type I1 cells and immature type IT cells with few lamellated inclusions
but numerous microvilli. There was also evidence of increased fibroblastic activity and collagen fibre production in interalveolar septa in these areas. Scanning electron microscopy ( S E M ) SEM examination of normal monkey lung tissue showed that the airways, blood vessels, and alveoli had a similar structure to that described for man and other animals.’,’0 Lung consolidated by PCP revealed alveoli packed with irregular, ovoid bodies with one or more indentations and small filopodial projections (Fig. 7). Since the alveolar surfaces were obscured by the organisms, it was not possible to gain information about attachment to the epithelial cells from this type of specimen. However, by observing selected organisms near the cut edge of an alveolus
182
A BASKERVILLE E T A L .
Fig. 7-Scanning electron micrograph (SEM) of alveolus packed with Pneun.rocytir organisms. x 2820
Fig. 8-SEM of single Pneurnocy.uk trophozoite on the alveolar surface showing attachment by elongated filopodia (arrows). x 21 600
and by applying an optimum tilt angle to the specimen, several filopodia could occasionally be seen extending from the lower surface of the parasite onto the surface of the epithelial cells (Fig. 8). On the surface which had previously been sectioned, comparison of light microscopy and SEM showed clumps of P. carinii as characteristic honeycomb or spongy structures (Fig. 9) with a few interspersed macrophages. Alveolar epithelium, connective tissue, and capillaries were also identifiable. Higher magnification of the sectioned, dewaxed surface of parasite masses showed numerous hollow cup-shaped bodies, often linked by membranes and amorphous material. A different aspect of the lung tissue was obtained by freeze cleavage of the specimens. The fracture plane occasionally ran through many hollow organisms and allowed three-dimensional detail to be seen between the parasites, which was not revealed after dewaxing the sectioned block. Long thin strands were demonstrated extending from the outer organisms to the alveolar surface, apparently serving as attachment structures (Fig. lo). Thecloseapposition of trophozoites to form a honeycomb-like aggregate was also well displayed.
DISCUSSION This is the first published account of SEM findings and histopathological and ultrastructural changes in PCP in a SIV-infected non-human primate. The ultrastructural morphology of the various stages of P. carinii was identical to that described in man and the This suggests that the immunodeficient rhesus monkey is a good model for the study of PCP in man. A technique for parallel investigation of paraffin-blocked lung tissue by immunohistochemistry and SEM using a rapid photographic montage to aid location proved to be helpful. These combined techniques made possible the study of two early cases of simian PCP, in which only a small number of organisms was present. The lesions were small and non-specific and would have been overlooked on H&E staining alone, but the P. carinii were detected by monoclonal antibody staining. These early cases gave an important insight into the pathogenesis of the infection. The size of all forms of the parasite permits inhalation directly into the smallest airways and alveoli. The organism then attaches to the alveolar epithelial surface, either by close apposition of the membranes or by using
PNEUMOCYSTIS CARINII PNEUMONIA IN SIV
183
Fig. 9-SEM of alveolus of dewaxed lung, showing spongy appearance of Pnrumocysris masses. A few macrophages are also present. x 1410
Fig. 10-SEM of freeze-cleaved lung, illustrating hollow organisms and attachment to the alveolar surface by long strands. x 940
elongated filopodia to act as anchors. We have clearly demonstrated this by SEM and therefore confirmed in vivo an attachment process previously described only in tissue culture'' and not in vivo.I2 Attachments involving extensions of trophozoite cytoplasm which Long et interpreted in the rat lung as hooks were also seen, but we regard these as a fortuitous appearance of the intimate contact of the irregular surfaces of parasite and epithelium. In addition, in our primate lungs, long strands connecting the masses of P . carinii to the alveolar surfaces were also demonstrated by SEM after freeze cleaving. In the early stages of PCP, there was little or no cellular response to the parasites in an alveolus and few P. cariniiwere phagocytosed. Consequently, proliferation of trophozoites, their conversion to cysts, sporozoite formation, and completion of the life cycle continued until most alveoli became packed with the parasites. Examination of heavy P. carinii infection in the monkeys also showed that few macrophages were present in alveoli, and though phagocytosis did take place, it was at a very low level. Where parasites had been engulfed, several trophozoites were present in separate phagosomes in a macrophage. In some
cases, trophozoite membranes had disintegrated, showing that killing can occur. We did not observe phagocytosis of cyst forms, though this has been d e ~ c r i b e d .Interalveolar '~ septa were greatly thickened by fibrosis and cellular infiltration in monkeys with advanced PCP, to an extent which would cause impairment of gaseous exchange. Since PCP in the monkey probably reflects accurately the disease process in man, this supports earlier suggestions15 that P. carinii itself causes pulmonary fibrosis. The reason for the establishment of P. carinii infection in immunosuppressed individuals is not known, though the mode of dealing with the initial stages of infection is probably the crucial difference between the immunodeficient and competent states. P. carinii is considered to be ubiquitous in the environment and not uncommonly inhaled by normal subjects, the organisms being removed by nonspecific clearance mechanisms; in the case of the alveoli, predominantly by phagocytosis. The failure of this clearance process appears to be due to inadequate mobilization of sufficient numbers of macrophages and perhaps also to a low efficiency of phagocytosis. Whether the total number of pulmonary macrophages is reduced in SIV- or HIV-infected
184
A BASKERVILLE ET A L.
subjectsis unknown, but the failuremay be indirectly linked to the deficiencies in cell-mediated immunity characteristic of these infections. ACKNOWLEDGEMENTS
We are grateful for the skilled assistance of Mr R. T. Raymond, Mr R. T. Moore, Miss P. Freemantle, and Mrs E. Elphick, and to Mr C. Inman for providing the micrograph for Fig. 8. REFERENCES I . Chalifoux LV, Ringler DJ, King NW e f al. Lymphadenopathy in macaques experimentally infected with the simian immunodeficiency virus. Am JParhul 1987; 1 2 8 1 0 4 1 10. 2. Baskin GB, Murphey-Corb M, Watson EA, er ol. Necropsy findings in rhesus monkeys experimentally infected with cultured simian immunodeficiency virus (SIV) Delta. Vet Pathol1988; 2 5 456 467. 3. King NW. Simian models of acquired immunodeficiency syndrome (AIDS): a review. Vet Purhol 1986; 23: 345-353. 4. Marchevsky A, Rosen MJ, Chrysial G . Pulmonary complications of AIDS: a clinicopathologic study of 70 cases. Hum Parhol 1985; 16: 659-670.
5. Baskerville A, Ramsay AD, Cranage MP, of 01. Histopathological changes in simian immunodeficiency virus infection. J Parhol 1990; 162: 67-75. 6. Cranage MP, Cook N, Johnstone P, et a/. SIV infection of rhesus macaques: in vivo titration ofinfectivity and development of an experimental vaccine. In: Schellekens H, Horzinek M, eds. Animal Models in AIDS. Amsterdam: Elsevier, 1990; 103-1 13. 7. Barton EG, Campbel WG. Further observations on the ultrastructure of Pnromocysris. Arcl7 Parhol 1967; 83: 537-534. 8. Barton EG, Campbell WG. Pneumocptis carinii in lungs of rats treated with cortisone acetate. Am J Puthol 1969; 54: 209-236. 9. Kuhn, C, Finke EH. The topography of the pulmonary alveolus: scanning electron microscopy using different fixations. J Ulrrasrrucr Res 1972;38: 161-172. 10. Nowell JA, Tyler WS. Scanning electron microscopy of the surface morphology of mammalian lungs. Am Rev Resp Dis 1971; 103: 313-328. 11. Murphy MJ, Pifer LL, Hughes WT. Pneumocysfis carinii in vifro: a study by scanning electron microscopy. Am J Parhol 1977; 86 387402. 12. Henshaw NG. Carson JL, Collier AM. Ultrastructural observations of Pnrumocwtrs cariniiattachment to rat lung. JInfecr Dis 1985; 151: 181-1 86. 13. Long EG, Smith JS, Meier JL. Attachment of Pneumocy~riscurinii to rat pneumocytes. Lab Invesl 1986; 54: 609-615. 14. Hasleton PS, Curry A, Rankin EM. Pneumocystis carinif pneumonia: a light microscopical and ultrastructural study. JCIin Purhol 1981; 34: 1138-1146. 15. Hasleton PS, Curry A. Pneumocystis corinii: the continuing enigma. Thorax 1982; 37: 481485.
164 175-1 84 (1 99 1)
PNEUMOCYSTIS CARINII PNEUMONIA IN SIMIAN IMMUNODEFICIENCY VIRUS INFECTION: IMMUNOHISTOLOGICAL AND SCANNING AND TRANSMISSION ELECTRON MICROSCOPICAL STUDIES A. BASKERVILLE, A. B. DOWSETT, R. W. COOK, M. J . DENNIS, M. P. CRANAGE AND P. J . GREENAWAY
Division of Pathology, Public Health Laboratory Service, Centre.fhr Applied Microbiology and Research, Porton Down, Salisburj, Wiltshire. U.K. Received 19 December 1990 Accepted 14 Junuary 1991
SUMMARY Pneurnocystis carinii pneumonia occurred in 6 o f 17 rhesus monkeys infected with simian immunodeficiency virus and was studied by immunohistochemistry and by scanning and transmission electron microscopy. A monoclonal antibody/streptavidin-biotin-peroxidase staining method was highly sensitive for detecting the organisms in small, early lesions and was much more sensitive and specific than traditional silver impregnation methods. Reprocessing of paraffin wax-embedded lung tissue for scanning electron microscopy and use o f a video printer to produce a photographic montage of light microscopic lesions allowed the same areas of tissue to be examined and compared by both methods. The ultrastructural morphology o f P. carinii in the rhesus monkey was identical to that in man, as were the histological and electron microscopic lesions, including pulmonary fibrosis. Trophozoites were seen attached to alveolar type I epithelium mainly by intimate apposition to the plasma membrane, but scanning electron microscopy also showed attachment by elongated filopodia. Few macrophages were present in infected alveoli, and though phagocytosis followed by digestion of P. carinii trophozoites was observed, it appeared to occur at a very low level. KEY
WORDS-Non-human primates, immunodeficiency, Pneumoy3stis pneumonia, immunohistochemistry, electron microscopy.
INTRODUCTION Infection of non-human primates with simian immunodeficiency virus (SIV) produces immunodeficiency, weight loss, lymphadenopathy, and neuropathological lesions,''2 a complex of disease similar to that in human acquired immune deficiency syndrome (AIDS) caused by the related human immunodeficiency lentivirus HIV- 1. SIV infection of macaque monkeys, in particular rhesus, is considered a highly relevant experimental model for the Addressee for correspondence: Dr A. Baskerville, Division of Pathology, Public Health Laboratory Service, Centre for Applied Microbiology and Research, Porton Down, Salisbury, Wiltshire,SP4 OJG, U.K.
0022-341 7/91/050175-10 $05.00 0 1991 by John Wiley & Sons, Ltd
study of AIDS3 because the animals also develop many ofthe secondary bacterial, viral, and protozoal infections which occur in AIDS patients. One of the commonest infections in human AIDS, and a frequent cause of death, is pneumonia due to Pneumocystis carinii and surveys have indicated a prevalence of up to 85 per cent. It was noted in both natural and experimental infection of macaques with SIV that a proportion also suffered from P.cariniipneumonia (PCP),'.5 though detailed pathological and ultrastructural accounts have not been published. This paper presents the findings of immunohistochemical and transmission and scanning electron microscopical studies of different stages of naturally acquired PCP in SIV-infected rhesus monkeys.
176
A BASKERVILLE E T A L .
Table I-Simian
Monkey No. 32H 27H 19H K33 K8 I67 85H 44H 68H 14H 17H 62H 67H 61H 2H 53H 29 H
immunodeficiency viruses used and disease course
Sex
9 P 9 3 d d
9
Y
9 3 P 0 d d
6
s 0
Time to death (months)
Virus SIV,,,25 1 SIVm,,32H* SIVm,,32H/CBLt SIVm,,32H SIVm,,32H SIVm,,32H SIVm,,32H SIVm,,32H SIVm,,32H SIVmd,32H SIVmd,32H SIVm,,32H SIV,,,clone 32H bloodf SIVSM SIVC," 705B blood
Died (D) or killed (K)
18 11 11;
G
14
1@ 5:
4 9 weeks
1st
5 7 15; 27 28 30 21
K K K K K K K D D K K K D K K K K
*SIV,4L32H refers to a virus stock isolated from monkey 32H and adapted to a continuous human T-cell line. tSIVm,,32H/CBL refers to the primary isolate from monkey 32H grown on human cord blood lymphocytes. :Monkey 61H was infected with SIV by direct inoculation of blood from monkey 32H.
MATERIALS AND METHODS Animals Seventeen rhesus monkeys (Macuca mulutta), 8 male, 9 female, bred in the United Kingdom and aged between 2 and 3 years at the time ofinfection, were used. The animals were fed a commercial pelleted diet supplemented with fruit and were given water ad libitum. They were caged singly and housed under containment category P3 conditions.
Viruses The viruses were given intravenously and details are summarized in Table I. Full details of the virus preparation, growth, and methods of isolation from tissues of the monkeys have been presented
Procedures For infection and all other procedures, animals were anaesthetized by intramuscular injection of ketamine hydrochloride ('Vetalar', Parke Davis).
Blood sampling and intravenous inoculation of virus suspension were carried out using the femoral vein. Animals were killed by intravenous injection Of pentobarbitone sodium. Autopsy and histopathology Autopsy was carried out immediately after death except in the cases of the three animals which died, where there was a delay of not more than 12 h. The trachea and lungs were removed in toto. Slices of lung were taken from each lobe and fixed in 10 per cent buffered neutral formalin. After fixation, the tissue was processed by standard methods and embedded in paraffin wax. Sections were cut at 5 pm and stained by haematoxylin and eosin (H&E), periodic acid-Schiff (PAS), Gordon and Sweet's method for reticulin, Grocott's silver methenamine technique, and with anti-Pneumocystis carinii monoclonal antibody (M778, Dako Ltd., High Wycombe, U.K.) using the streptavidin-biotinperoxidase method and diaminobenzidine as indicator. Control tissue was obtained from eight uninfected normal rhesus monkeys of the same age.
PNEUMOCYSTIS CARINII PNEUMONIA IN SIV
Fig. ]-Early infection of monkey lung by p . Stained by monoclonal antibody to the organism and the streptavidinbiotin-peroxidase method. Aleveoli contain dark-staining organisms (arrows)
Transmission electron microscopy One mm3 portions of tissue were taken from consolidated areas of the lungs of two monkeys. These were immediately fixed in cold 4 per cent phosphatebuffered glutaraldehyde at pH 7.2 and later postfixed in 2 per cent buffered osmium tetroxide at 4°C. After processing through a graded series of ethanol and propylene oxide, the tissue was embedded in Araldite. Sections 1 pm in thickness were cut and stained with toluidene blue for location of suitable areas for ultrastructural examination. Ultrathin sections were stained with uranyl acetate and lead citrate, and stabilized by carbon coating in a vacuum coating unit prior to examination in a Philips EM400T transmission electron microscope.
177
Fig. 2-Heavy infection of monkey lung by P. carinii. Alveoli are packed with organisms and interalveolar septa are thickened by fibrous tissue. Monoclonal antibody to p , and
streptavidin-biotin-peroxidase
Scanning electron microscopy Specimens were prepared by the following regimens: (a) Slices of lung, approximately 1 cm2 and 3 4 mm in thickness, were taken from the formalin-fixed tissue. These were post-fixed overnight in 1 per cent buffered osmium tetroxide at 4"C, dehydrated through a graded acetone series, and critical point-dried from liquid carbon dioxide prior to mounting on SEM specimen stubs using silver dag. (b) Cubes of formalin-fixed lung tissue, approximately 4 mm in size, were plunged into liquid nitrogen slush at -210°C and cleaved with a razor blade. Tissue fragments were then postfixed, dried, and mounted as before, care being
178
A BASKERVILLE ET A L .
Fig. 3-Electron micrograph of alveolus containing trophozoites, membranes. and fibrin. Filopodia are visible in section on some surfaces. x 30 000
taken to expose the fractured tissue surface for examination.
as before, and oriented on the specimen stub so as to present for SEM examination the tissue face previously sectioned for light microscopy.
Comparative exurnination ofthe same tissue areas bj, light and scanning electron microscopy
Photographic montages of sections of waxembedded tissue which showed the presence of P. carinii by immunohistochemistry were constructed from micrographs recorded by a Sony UP 5000P video printer from a Leitz Ortholux 2 microscope fitted with a JVC CCD camera. The corresponding tissue blocks were then dewaxed with three 1 h changes of xylene at 60°C and taken from 100 to 30 per cent ethanol, through a graded series, to phosphate buffer. Tissue was then post-fixed and dried
Conductive coating Specimens for SEM were sputter-coated with a 20 nm layer of gold prior to examination in either a Cambridge Stereoscan S2A or a Hitachi S-800 scanning electron microscope.
RESULTS Macroscopic Jindings Of the 17 SIV-infected rhesus monkeys in the study, four had extensive lung lesions of PCP at
PNEUMOCYSTIS CARINII PNEUMONIA IN SIV
Fig. &Infected
179
alveolus with Pneurnocvstis cyst containing sporozoites. x 16 000
necropsy (numbers 14H, 17H, 62H, and K8). The lesions consisted of a characteristic, uniformly grey consolidation on the surface of the lungs and in cut slices, and involved virtually the whole of each lobe. The lungs of all other SIV-infected monkeys and negative controls showed no macroscopic abnormalities.
Histopathological and immunocytochemicalfindings Six animals had PCP lesions: the four with extensive pneumonia recognized at autopsy and a further two (K33 and 61H) in which the lesions were microscopic. In sections stained by H&E, the smallest lesions appeared as a few scattered foci of infiltration of a group of interalveolar septa or even single alveoli by polymorphonuclear leucocytes (PMN) and
macrophages. These foci frequently did not have any floccular material in the alveoli and P. carinii infection would not have been suspected from H&E sections alone. However, immunostaining using a monoclonal antibody to human P. carinii revealed small numbers of oval or circular organisms in these foci (Fig. 1). They were also found in alveolar walls or on alveolar epithelium with no accompanying cellular response or abnormality of alveolar cells. There was noconsistent distribution ofthe organisms in the lobe. In the four cases with massive lung involvement by PCP, the majority of alveoli were distended by foamy eosinophilic PAS-positive material, and staining with monoclonal antibody demonstrated these to be densely packed by P. carinii cysts and trophozoites (Fig. 2). Staining of consecutive sections by Grocott's method for comparison with the
180
Fig. 5-Alveolus
A BASKERVILLE ET A L .
showing intimate association between two trophozoites (T) and type I epithelial cell. Cap: capillary. x 27 300
monoclonal antibody revealed only a small proportion of the organisms as argyrophilic. In animals with heavy infestation, the P. carinii were distributed widely in the lung tissue. By monoclonal antibody staining most were detected in alveoli, but some were also present in alveolar walls, on the epithelial surface, and in the lumen of bronchi and at all levels of the smaller airways. Occasionally, a few P. carinii were found in lymphatics and veins, both free and within phagocytic cells. In the established cases, the local cellular alveolar response was minimal, though alveolar walls were thickened by fibroblasts, macrophages, and lymphocytes (Fig. 2). The alveolar epithelium in some areas had changed to a basophilic, cuboidal, or even columnar type.
Transmission electron microscopy ( T E M ) Lung tissue was taken from two monkeys with extensive lesions of PCP for TEM studies and the findings were identical in both. Large numbers of thin-walled trophozoites of P. curinii completely filled many alveoli and in others formed deep layers over the alveolar type I epithelium. The trophozoites varied considerably in shape as a result of compression and interlocking with adjacent parasites (Fig. 3). Their internal structure consisted of
finely granular cytoplasm, with irregularly scattered larger electron-dense granules and a nuclear region with an ill-defined membrane and a prominent nucleolus. Other organelles were not identified and many trophozoites appeared empty. Thick-walled cysts were also present, some of which contained sporozoites (Fig. 4) and a zone of thickening of the inner wall. Collapsed cysts were also observed, appearing as crescents (Fig. 3), but with the thickened zone still intact. Short, tubular projections were present on most of the parasites. These processes have been termed filopodia7~* and were all approximately 85 nm in diameter. While filopodia at some sites were closely applied to the surface of adjacent trophozoites, alveolar epithelium, or macrophages, many contacts took place in the absence of filopodia. Close examination of the sites of attachment of trophozoites to the alveolar wall showed intimate contact between the plasma membrane of the type I epithelial cell and the membrane of the parasite, which appeared to adapt its shape to the contours of the cell (Fig. 5). In some cases, contact was over a very small area, but in others much of the involved surface of the trophozoite was applied to the epithelium. The same close contact occurred between the outer membrane of adjacent trophozoites and
PNEUMOCYSTIS CARINII PNEUMONIA IN SIV
Fig. &Alveolar macrophage containing disintegrating phagocytosed trophozoites. x 19 000
cysts, so that a composite colony of parasites was built up which appeared to be firmly anchored onto the alveolar surfaces. The cellular responses which occurred in affected alveoli were minor. Fibrin clumps, cell debris, and surfactant membranes were present in a few alveoli. Macrophages had phagocytosed trophozoites, which were identifiable within phagosomes (Fig. 6), many undergoing lysis. Occasional macrophages were seen which were surrounded by densely packed trophozoites but were not phagocytosing them, and the numerous pseudopodia appeared compressed against the cell surface. The cuboidal metaplasia of alveoli observed by light microscopy wasshown to be due to replacement of the type I epithelial cells by large numbers of electron-dense type I1 cells and immature type IT cells with few lamellated inclusions
but numerous microvilli. There was also evidence of increased fibroblastic activity and collagen fibre production in interalveolar septa in these areas. Scanning electron microscopy ( S E M ) SEM examination of normal monkey lung tissue showed that the airways, blood vessels, and alveoli had a similar structure to that described for man and other animals.’,’0 Lung consolidated by PCP revealed alveoli packed with irregular, ovoid bodies with one or more indentations and small filopodial projections (Fig. 7). Since the alveolar surfaces were obscured by the organisms, it was not possible to gain information about attachment to the epithelial cells from this type of specimen. However, by observing selected organisms near the cut edge of an alveolus
182
A BASKERVILLE E T A L .
Fig. 7-Scanning electron micrograph (SEM) of alveolus packed with Pneun.rocytir organisms. x 2820
Fig. 8-SEM of single Pneurnocy.uk trophozoite on the alveolar surface showing attachment by elongated filopodia (arrows). x 21 600
and by applying an optimum tilt angle to the specimen, several filopodia could occasionally be seen extending from the lower surface of the parasite onto the surface of the epithelial cells (Fig. 8). On the surface which had previously been sectioned, comparison of light microscopy and SEM showed clumps of P. carinii as characteristic honeycomb or spongy structures (Fig. 9) with a few interspersed macrophages. Alveolar epithelium, connective tissue, and capillaries were also identifiable. Higher magnification of the sectioned, dewaxed surface of parasite masses showed numerous hollow cup-shaped bodies, often linked by membranes and amorphous material. A different aspect of the lung tissue was obtained by freeze cleavage of the specimens. The fracture plane occasionally ran through many hollow organisms and allowed three-dimensional detail to be seen between the parasites, which was not revealed after dewaxing the sectioned block. Long thin strands were demonstrated extending from the outer organisms to the alveolar surface, apparently serving as attachment structures (Fig. lo). Thecloseapposition of trophozoites to form a honeycomb-like aggregate was also well displayed.
DISCUSSION This is the first published account of SEM findings and histopathological and ultrastructural changes in PCP in a SIV-infected non-human primate. The ultrastructural morphology of the various stages of P. carinii was identical to that described in man and the This suggests that the immunodeficient rhesus monkey is a good model for the study of PCP in man. A technique for parallel investigation of paraffin-blocked lung tissue by immunohistochemistry and SEM using a rapid photographic montage to aid location proved to be helpful. These combined techniques made possible the study of two early cases of simian PCP, in which only a small number of organisms was present. The lesions were small and non-specific and would have been overlooked on H&E staining alone, but the P. carinii were detected by monoclonal antibody staining. These early cases gave an important insight into the pathogenesis of the infection. The size of all forms of the parasite permits inhalation directly into the smallest airways and alveoli. The organism then attaches to the alveolar epithelial surface, either by close apposition of the membranes or by using
PNEUMOCYSTIS CARINII PNEUMONIA IN SIV
183
Fig. 9-SEM of alveolus of dewaxed lung, showing spongy appearance of Pnrumocysris masses. A few macrophages are also present. x 1410
Fig. 10-SEM of freeze-cleaved lung, illustrating hollow organisms and attachment to the alveolar surface by long strands. x 940
elongated filopodia to act as anchors. We have clearly demonstrated this by SEM and therefore confirmed in vivo an attachment process previously described only in tissue culture'' and not in vivo.I2 Attachments involving extensions of trophozoite cytoplasm which Long et interpreted in the rat lung as hooks were also seen, but we regard these as a fortuitous appearance of the intimate contact of the irregular surfaces of parasite and epithelium. In addition, in our primate lungs, long strands connecting the masses of P . carinii to the alveolar surfaces were also demonstrated by SEM after freeze cleaving. In the early stages of PCP, there was little or no cellular response to the parasites in an alveolus and few P. cariniiwere phagocytosed. Consequently, proliferation of trophozoites, their conversion to cysts, sporozoite formation, and completion of the life cycle continued until most alveoli became packed with the parasites. Examination of heavy P. carinii infection in the monkeys also showed that few macrophages were present in alveoli, and though phagocytosis did take place, it was at a very low level. Where parasites had been engulfed, several trophozoites were present in separate phagosomes in a macrophage. In some
cases, trophozoite membranes had disintegrated, showing that killing can occur. We did not observe phagocytosis of cyst forms, though this has been d e ~ c r i b e d .Interalveolar '~ septa were greatly thickened by fibrosis and cellular infiltration in monkeys with advanced PCP, to an extent which would cause impairment of gaseous exchange. Since PCP in the monkey probably reflects accurately the disease process in man, this supports earlier suggestions15 that P. carinii itself causes pulmonary fibrosis. The reason for the establishment of P. carinii infection in immunosuppressed individuals is not known, though the mode of dealing with the initial stages of infection is probably the crucial difference between the immunodeficient and competent states. P. carinii is considered to be ubiquitous in the environment and not uncommonly inhaled by normal subjects, the organisms being removed by nonspecific clearance mechanisms; in the case of the alveoli, predominantly by phagocytosis. The failure of this clearance process appears to be due to inadequate mobilization of sufficient numbers of macrophages and perhaps also to a low efficiency of phagocytosis. Whether the total number of pulmonary macrophages is reduced in SIV- or HIV-infected
184
A BASKERVILLE ET A L.
subjectsis unknown, but the failuremay be indirectly linked to the deficiencies in cell-mediated immunity characteristic of these infections. ACKNOWLEDGEMENTS
We are grateful for the skilled assistance of Mr R. T. Raymond, Mr R. T. Moore, Miss P. Freemantle, and Mrs E. Elphick, and to Mr C. Inman for providing the micrograph for Fig. 8. REFERENCES I . Chalifoux LV, Ringler DJ, King NW e f al. Lymphadenopathy in macaques experimentally infected with the simian immunodeficiency virus. Am JParhul 1987; 1 2 8 1 0 4 1 10. 2. Baskin GB, Murphey-Corb M, Watson EA, er ol. Necropsy findings in rhesus monkeys experimentally infected with cultured simian immunodeficiency virus (SIV) Delta. Vet Pathol1988; 2 5 456 467. 3. King NW. Simian models of acquired immunodeficiency syndrome (AIDS): a review. Vet Purhol 1986; 23: 345-353. 4. Marchevsky A, Rosen MJ, Chrysial G . Pulmonary complications of AIDS: a clinicopathologic study of 70 cases. Hum Parhol 1985; 16: 659-670.
5. Baskerville A, Ramsay AD, Cranage MP, of 01. Histopathological changes in simian immunodeficiency virus infection. J Parhol 1990; 162: 67-75. 6. Cranage MP, Cook N, Johnstone P, et a/. SIV infection of rhesus macaques: in vivo titration ofinfectivity and development of an experimental vaccine. In: Schellekens H, Horzinek M, eds. Animal Models in AIDS. Amsterdam: Elsevier, 1990; 103-1 13. 7. Barton EG, Campbel WG. Further observations on the ultrastructure of Pnromocysris. Arcl7 Parhol 1967; 83: 537-534. 8. Barton EG, Campbell WG. Pneumocptis carinii in lungs of rats treated with cortisone acetate. Am J Puthol 1969; 54: 209-236. 9. Kuhn, C, Finke EH. The topography of the pulmonary alveolus: scanning electron microscopy using different fixations. J Ulrrasrrucr Res 1972;38: 161-172. 10. Nowell JA, Tyler WS. Scanning electron microscopy of the surface morphology of mammalian lungs. Am Rev Resp Dis 1971; 103: 313-328. 11. Murphy MJ, Pifer LL, Hughes WT. Pneumocysfis carinii in vifro: a study by scanning electron microscopy. Am J Parhol 1977; 86 387402. 12. Henshaw NG. Carson JL, Collier AM. Ultrastructural observations of Pnrumocwtrs cariniiattachment to rat lung. JInfecr Dis 1985; 151: 181-1 86. 13. Long EG, Smith JS, Meier JL. Attachment of Pneumocy~riscurinii to rat pneumocytes. Lab Invesl 1986; 54: 609-615. 14. Hasleton PS, Curry A, Rankin EM. Pneumocystis carinif pneumonia: a light microscopical and ultrastructural study. JCIin Purhol 1981; 34: 1138-1146. 15. Hasleton PS, Curry A. Pneumocystis corinii: the continuing enigma. Thorax 1982; 37: 481485.
