JOURNAL OF
PATHOLOGY, VOL. 166: 195-1 98 (1 992)
A SEARCH FOR PNEUMOCYSTIS CARINII IN POST-MORTEM LUNGS BY DNA AMPLIFICATION SARAH E.
PETERS* ANN
E.
WAKEFIELD*, KATHRYN SINCLAIR*, PETER
R. MILLARD? AND
JULIAN M. HOPKIN$
*Molecular Infectious Diseuses Group, Department of Paediatrics, Institute of Molecular Medicine and ?Department of Pathology, John Rudclife Hospital, Oxford; SOsler Chest Unit, Churchill Hospital, Oxford, U.K. Received 14 May 1991 Accepted 1 July 1991
SUMMARY DNA amplification of specific sequences and subsequent oligonucleotide hybridization were used to search for Pneumocystzs curinii in post-mortem lung samplings from non-immunosuppressed individuals ranging from 15 to 70 years of age. No P. carinii-specific DNA was detected in 45 DNA amplification reactions from 15 lungs. KEY
WORDS-Pneumocystis carinii post-mortem lungs, DNA amplification.
INTRODUCTION
METHODS
Serological surveys suggest that asymptomatic infection with P. carinii (the main cause of pneumonia in the immunosuppressed’) occurs in the great majority of healthy Post-mortem studies on lung tissue using Grocott’s methenamine silver stain for P. carinii have suggested that small numbers of the parasite can be observed in the lungs of non-immunosuppressed individual^.^ It has been widely assumed that opportunistic, characteristically diffuse P. carinii pneumonia in the immunosuppressed results from reactivation of a focus of organisms carried within the lungs which may, by analogy with tuberculosis, be the residuum of this asymptomatic childhood infection. Recent studies usingan efficient technique ofmonoclonal antibody binding to screen post-mortem lungs for P. carinii have challenged the concept of parasite carriage in normal 1ungs.j To test this further, we have conducted a survey of post-mortem lungs using DNA amplification by the polymerase chain reaction, a technique that offers very high s ecificity and sensitivity in identifying the parasite.69
P
Addressee for correspondence: Dr S. E. Peters, Molecular Infectious Diseases Group, Department of Pdediatrics, Institute of Molecular Medicine, John Radcliffe Hospital, Oxford
OX3 9DU, U.K.
0022-341 7/92/02019544$05.00 0 1992 by John Wiley & Sons, Ltd.
A complete lobe of lung tissue (right and left upper and lower lobes in rotation) was obtained at post-mortem examination from each of 15 individuals aged 15-75 years. Causes of death included cardiovascular disease (myocardial infarction, cerebral vascular accident, congestive cardiac failure) or trauma; in none did the clinical records or post-mortem examination suggest the possibilities of immunosuppression or primary lung disease. Each lobe was homogenized in a Waring blender and DNA extracted from a sample of lg by proteinase K digestion (500 pg ml-’) at 50°C in the presence of 10 mM EDTA and 0.25 per cent SDS, followed by phenolshloroform extraction and ethanol precipitation. After each lung sample, a buffer negative control was processed in an identical way to test for cross-contamination. The blender was also acid-cleaned (1 M HCl) between samples. For DNA amplification, we used our previously described oligonucleotide primers and probes6 Oligonucleotide primers pAZ102-E and pAZl02-H were used in a 50 pl amplification reaction mixture final concentrations 15 mM KCl, 10 mM Tris (pH 8.0), 0.01 per cent (w/v) gelatine, 3 m MgCl,, ~ 400 p~ dNTPs (Boehringer Mannheim), 1 p~ oligonucleotide primer, and 0.025 units/ pl of ‘Amplitaq’ (Perkin Elmer Cetus)]. Denaturation was at 94°C
196
S . E. PETERS ET AL.
A 1
2
3
4
5
6
7
8
9
10
11
12 13
344-
B
Fig. 1-DNA amplification reactions from post-mortem lungs using P.carinii primers. (A) Agarose gel electrophoresis, in the presence of ethidium bromide; (B) autoradiography of oligonucleotide hybridization. Lanes 1 and 13: 1 k b molecular weight markers (sizes shown in bp). Lane 2: lung ‘X’ DNA ( x 5 dilution). Lane 3: lung ‘X’ DNA ( x 25 dilution). Lane 4: no template DNA negativecontrol for lung ‘X’. Lane 5: buffer negative control ( x 5 dilution). Lane 6: buffer negative control ( x 25 dilution). Lane 7: no template DNA negative control for buffer control. Lane 8: lung ‘Y’ DNA ( x 5 dilution). Lane 9: lung ‘Y’ DNA ( x 25 dilution). Lane 10: no template DNA negative control for lung ‘Y’. Lane 11: no template D N A negative control. Lane 12: P. carinii DNA positive control. The arrow marks the position of specific amplification product
for 1.5min, annealing at 55°C for 1.5min, and extension at 72°C for 2min for 40 cycles. Three DNA amplification reactions were performed on the extracted DNA from each of the 15 lung samples (equivalent to 7.5 mg tissue in total or the screening by microscopy of 400 cm2 of lung). Negative controls with no added template DNA were included after each lung and buffer sample (again to monitor for cross-contamination) and also a positive control of bronchial lavage from a case of proven P. carinii pneumonia. To test that the conditions for DNA
amplification and the quality of each DNA template from the lung samplings were adequate, polymerase chain reaction (PCR) amplification was also performed on the co-extracted human DNA, using primers for the human antithrombin gene.6 Amplification products were electrophoresed in 1.5 per cent agarose gel and examined by (i) direct visualization with ultraviolet light after ethidium bromide staining and (ii) oligonucleotide hybridization after Southern transfer to ‘Hybond N’ filters (Amersham), with a 32Pend-labelled internal primer
DNA AMPLIFICATION FOR PNEUMOCYSTIS
197
samples despite the power of the DNA amplification and subsequent oligonucleotide hybridization technique. In a parallel study, we have derived estimates of the sensitivity of our methods by seeding homogenized human lung with P. carinii organisms derived from the rat model. A band of amplified DNA visible on ethidium bromide staining indicates the presence of 100organisms or more in any particular sample; a band visible only on oligonucleotide hybridization denotes smaller numbers of organisms and indeed as few as 1-2 per sample (manuscript submitted for publication). In our present postmortem lung study, positive DNA amplification with the control human antithrombin primers demonstrates the satisfactory state of the extracted DNA and functioning of the DNA amplification -516 conditions. 394 These results accord with our previous observations on failure to detect P. carinii by DNA amplification in bronchoscopic alveolar lavage samples from non-immunosuppressed subjects’~* and the data on post-mortem lungs screening for the parasite using monoclonal a n t i b ~ d y Our . ~ results contrast Fig. 2-Ethidium bromide-stained agarose gel electrophoresis of DNA amplification reactions from post-mortem lungs using with the reports of visualization of small numbers of human anti-thrombin primers. Lane 1: lung ‘X’ DNA ( x 5 P. carinii in the lungs of non-immunosuppressed dilution). Lane 2: no template DNA negative control for lung ‘X’. individuals by methenamine silver ~ t a i n i n g ;the ~ Lane 3: lung ‘Y’ DNA ( x 5 dilution). Lane 4: no template nega- specificity and sensitivity of the DNA amplification tive control for lung ‘Y’. Lane 5: human DNA positive control (the arrow marks the position of specific amplification product). reaction imply that the silver stain and microscopy Lane 6: no template DNA negative control for human DNA. findings may have been spurious, and it is acknowlLane 7: 1 kb molecular weight markers (sizes shown in bp) edged that certain fungal spores can be misidentified as P. carinii using such techniques.’ Our results are ~ A Z 1 0 2 - L 2Filters .~ were washed at high stringency relevant to the use of DNA amplification as a diag(48°C) and exposed to radiographic film at - 80°C nostic method for P. carinii pneumonia, and comwith intensifying screens for 16 h. Homogenization, bined with the data from bronchoscopic lavage DNA extraction and PCR were performed in a studies show that P. carinii is not detectable in the laminar flow cabinet, and disposable tips, tubes, and lungs of non-immunosuppressed individuals. Low reagent aliquots were used to avoid contamination. levels of amplified DNA (detectable only by oligonucleotide hybridization and due to small numbers of organisms) are recognised in a small but signifiRESULTS cant proportion (20 per cent) of immunosuppressed No P.carinii-specific DNA amplification product HIV positive individuals without P. carinii pneu(345 bp) was detected by ethidium bromide staining monia, but strong amplified DNA signals (visible nor by subsequent oligonucleotide hybridization on simple ethidium bromide staining and equivalent in any of the lung samplings or buffer controls. to 100 organisms per sample) are found almost (Figure 1 shows the results from two lungs, X and exclusively (98 per cent) in those with P. carinii Y . ) The antithrombin primers showed efficient p n e ~ m o n i a . ~ The data are also relevant to the uncertain epiamplification of antithrombin-specific sequence from the human DNA in each of the samples (Fig. 2 ) . demiology and biology of P. carinii. The absence of detectable P. carinii in the lung samples tested does not favour the concept that there is regular pulmonDISCUSSION ary carriage of the organism which acts as a source No amplified P. carinii DNA was detected in any for expansion of parasite numbers during immunoof the three reactions taken from each of the 15 lung suppression. Unless some other tissue is a repository
1
2
3
4
5
6
7
c
198
S. E. PETERS ET A L .
for dormant organisms (and there are no data to support such a possibility), fresh infection of the immunosuppressed by the parasite is the more likely event. Reported case clusters of P. carinii pneumonia'' suggest that horizontal transmission between immunosuppressed individuals may be significant, whilst the recent demonstration, by comparative DNA studies, of the fungal origins of P. cariniil'-'' emphasizes the possibility of an, as yet unidentified, environmental source of infectious particles. ACKNOWLEDGEMENTS
This work was funded by the Medical Research Council and the Wellcome Trust. AEW is supported by The Royal Society. REFERENCES 1. Hopkin JM. Pneumocyslis carinii. Oxford: Oxford University Press.
1991.
2 . Meuwissen JHET, Tauber I, Leevwenberg A D E M , Beckers PJA, Sieben M . Parasitologic and serologic observations of infection with pneumocystis in humans. Jfnferr Dis 1977; 1 3 6 43-49. 3. Wakefield AE, Stewart TJ, Moxon ER, Marsh K, Hopkin JM. Infection with PneurnocyJlis curinii is prevalent in healthy Gambian children. Trans R Soc Trop Med 1990; 84: 800-802. 4. Sheldon WH, Subclinical pneumocystis pneumonitis. Am J Dis Child 1959; 97: 287-297. 5. Millard PR, Heryet AR. Observations favouring Pneumocysti.y carinii pneumonia as a primary infection: a monoclonal antibody study o n paraffin sections. J Pathol1988; 154 365-370. 6. Wakefield AE, Pixley FJ, Banerji S , el ul. Amplification of niitochondrial ribosomal RNA sequences from Pneumocystis carinii D N A of rat and human origin. Moi Biochem Parusitol 1991; 43: 69-76. 7. Wakefield AE, Pixley FJ, Banerji S , ei ul. Detection of Pneumoeysri.7 currmi with D N A amplification. Lance1 1990; 336 45 1-453. 8 Wakefield AE. Guiver L, Miller RF, Hopkin J M . D N A amplification on sputum samples for diagnosis of PneumocJstrs carinu pneumonia. Lancer 1991;331: 1378-1379. 9. Reinhardt DJ, Kaplan W, Chandler F W , Morphologic resemblance of rygomycete spores to P. carinii in tissue. Am Re11 Resp Dis 1977; 115: 170-172. 10. Ruebush TK, Weinstein RA, Baehner RL. An outbreak of pneumocystis pneumonia in children with acute lymphocytic leukaemia. Am J Dis Child 1978; 132: 143-148. I I . Edman JC, Kovacs JA, Masur H, et a / . Ribosomal R N A sequences show Pneumocystis carinii to be a member of the fungi. Nature 1988; 334: 519-522. 12 Pixley FJ, Wakefield AE, Banerji S , Hopkin JM. Mitochondria1 gene sequences show fungal homology for Pneumocwtis carinii. Mol Microhiol1991: 5: 1347-1351
PATHOLOGY, VOL. 166: 195-1 98 (1 992)
A SEARCH FOR PNEUMOCYSTIS CARINII IN POST-MORTEM LUNGS BY DNA AMPLIFICATION SARAH E.
PETERS* ANN
E.
WAKEFIELD*, KATHRYN SINCLAIR*, PETER
R. MILLARD? AND
JULIAN M. HOPKIN$
*Molecular Infectious Diseuses Group, Department of Paediatrics, Institute of Molecular Medicine and ?Department of Pathology, John Rudclife Hospital, Oxford; SOsler Chest Unit, Churchill Hospital, Oxford, U.K. Received 14 May 1991 Accepted 1 July 1991
SUMMARY DNA amplification of specific sequences and subsequent oligonucleotide hybridization were used to search for Pneumocystzs curinii in post-mortem lung samplings from non-immunosuppressed individuals ranging from 15 to 70 years of age. No P. carinii-specific DNA was detected in 45 DNA amplification reactions from 15 lungs. KEY
WORDS-Pneumocystis carinii post-mortem lungs, DNA amplification.
INTRODUCTION
METHODS
Serological surveys suggest that asymptomatic infection with P. carinii (the main cause of pneumonia in the immunosuppressed’) occurs in the great majority of healthy Post-mortem studies on lung tissue using Grocott’s methenamine silver stain for P. carinii have suggested that small numbers of the parasite can be observed in the lungs of non-immunosuppressed individual^.^ It has been widely assumed that opportunistic, characteristically diffuse P. carinii pneumonia in the immunosuppressed results from reactivation of a focus of organisms carried within the lungs which may, by analogy with tuberculosis, be the residuum of this asymptomatic childhood infection. Recent studies usingan efficient technique ofmonoclonal antibody binding to screen post-mortem lungs for P. carinii have challenged the concept of parasite carriage in normal 1ungs.j To test this further, we have conducted a survey of post-mortem lungs using DNA amplification by the polymerase chain reaction, a technique that offers very high s ecificity and sensitivity in identifying the parasite.69
P
Addressee for correspondence: Dr S. E. Peters, Molecular Infectious Diseases Group, Department of Pdediatrics, Institute of Molecular Medicine, John Radcliffe Hospital, Oxford
OX3 9DU, U.K.
0022-341 7/92/02019544$05.00 0 1992 by John Wiley & Sons, Ltd.
A complete lobe of lung tissue (right and left upper and lower lobes in rotation) was obtained at post-mortem examination from each of 15 individuals aged 15-75 years. Causes of death included cardiovascular disease (myocardial infarction, cerebral vascular accident, congestive cardiac failure) or trauma; in none did the clinical records or post-mortem examination suggest the possibilities of immunosuppression or primary lung disease. Each lobe was homogenized in a Waring blender and DNA extracted from a sample of lg by proteinase K digestion (500 pg ml-’) at 50°C in the presence of 10 mM EDTA and 0.25 per cent SDS, followed by phenolshloroform extraction and ethanol precipitation. After each lung sample, a buffer negative control was processed in an identical way to test for cross-contamination. The blender was also acid-cleaned (1 M HCl) between samples. For DNA amplification, we used our previously described oligonucleotide primers and probes6 Oligonucleotide primers pAZ102-E and pAZl02-H were used in a 50 pl amplification reaction mixture final concentrations 15 mM KCl, 10 mM Tris (pH 8.0), 0.01 per cent (w/v) gelatine, 3 m MgCl,, ~ 400 p~ dNTPs (Boehringer Mannheim), 1 p~ oligonucleotide primer, and 0.025 units/ pl of ‘Amplitaq’ (Perkin Elmer Cetus)]. Denaturation was at 94°C
196
S . E. PETERS ET AL.
A 1
2
3
4
5
6
7
8
9
10
11
12 13
344-
B
Fig. 1-DNA amplification reactions from post-mortem lungs using P.carinii primers. (A) Agarose gel electrophoresis, in the presence of ethidium bromide; (B) autoradiography of oligonucleotide hybridization. Lanes 1 and 13: 1 k b molecular weight markers (sizes shown in bp). Lane 2: lung ‘X’ DNA ( x 5 dilution). Lane 3: lung ‘X’ DNA ( x 25 dilution). Lane 4: no template DNA negativecontrol for lung ‘X’. Lane 5: buffer negative control ( x 5 dilution). Lane 6: buffer negative control ( x 25 dilution). Lane 7: no template DNA negative control for buffer control. Lane 8: lung ‘Y’ DNA ( x 5 dilution). Lane 9: lung ‘Y’ DNA ( x 25 dilution). Lane 10: no template DNA negative control for lung ‘Y’. Lane 11: no template D N A negative control. Lane 12: P. carinii DNA positive control. The arrow marks the position of specific amplification product
for 1.5min, annealing at 55°C for 1.5min, and extension at 72°C for 2min for 40 cycles. Three DNA amplification reactions were performed on the extracted DNA from each of the 15 lung samples (equivalent to 7.5 mg tissue in total or the screening by microscopy of 400 cm2 of lung). Negative controls with no added template DNA were included after each lung and buffer sample (again to monitor for cross-contamination) and also a positive control of bronchial lavage from a case of proven P. carinii pneumonia. To test that the conditions for DNA
amplification and the quality of each DNA template from the lung samplings were adequate, polymerase chain reaction (PCR) amplification was also performed on the co-extracted human DNA, using primers for the human antithrombin gene.6 Amplification products were electrophoresed in 1.5 per cent agarose gel and examined by (i) direct visualization with ultraviolet light after ethidium bromide staining and (ii) oligonucleotide hybridization after Southern transfer to ‘Hybond N’ filters (Amersham), with a 32Pend-labelled internal primer
DNA AMPLIFICATION FOR PNEUMOCYSTIS
197
samples despite the power of the DNA amplification and subsequent oligonucleotide hybridization technique. In a parallel study, we have derived estimates of the sensitivity of our methods by seeding homogenized human lung with P. carinii organisms derived from the rat model. A band of amplified DNA visible on ethidium bromide staining indicates the presence of 100organisms or more in any particular sample; a band visible only on oligonucleotide hybridization denotes smaller numbers of organisms and indeed as few as 1-2 per sample (manuscript submitted for publication). In our present postmortem lung study, positive DNA amplification with the control human antithrombin primers demonstrates the satisfactory state of the extracted DNA and functioning of the DNA amplification -516 conditions. 394 These results accord with our previous observations on failure to detect P. carinii by DNA amplification in bronchoscopic alveolar lavage samples from non-immunosuppressed subjects’~* and the data on post-mortem lungs screening for the parasite using monoclonal a n t i b ~ d y Our . ~ results contrast Fig. 2-Ethidium bromide-stained agarose gel electrophoresis of DNA amplification reactions from post-mortem lungs using with the reports of visualization of small numbers of human anti-thrombin primers. Lane 1: lung ‘X’ DNA ( x 5 P. carinii in the lungs of non-immunosuppressed dilution). Lane 2: no template DNA negative control for lung ‘X’. individuals by methenamine silver ~ t a i n i n g ;the ~ Lane 3: lung ‘Y’ DNA ( x 5 dilution). Lane 4: no template nega- specificity and sensitivity of the DNA amplification tive control for lung ‘Y’. Lane 5: human DNA positive control (the arrow marks the position of specific amplification product). reaction imply that the silver stain and microscopy Lane 6: no template DNA negative control for human DNA. findings may have been spurious, and it is acknowlLane 7: 1 kb molecular weight markers (sizes shown in bp) edged that certain fungal spores can be misidentified as P. carinii using such techniques.’ Our results are ~ A Z 1 0 2 - L 2Filters .~ were washed at high stringency relevant to the use of DNA amplification as a diag(48°C) and exposed to radiographic film at - 80°C nostic method for P. carinii pneumonia, and comwith intensifying screens for 16 h. Homogenization, bined with the data from bronchoscopic lavage DNA extraction and PCR were performed in a studies show that P. carinii is not detectable in the laminar flow cabinet, and disposable tips, tubes, and lungs of non-immunosuppressed individuals. Low reagent aliquots were used to avoid contamination. levels of amplified DNA (detectable only by oligonucleotide hybridization and due to small numbers of organisms) are recognised in a small but signifiRESULTS cant proportion (20 per cent) of immunosuppressed No P.carinii-specific DNA amplification product HIV positive individuals without P. carinii pneu(345 bp) was detected by ethidium bromide staining monia, but strong amplified DNA signals (visible nor by subsequent oligonucleotide hybridization on simple ethidium bromide staining and equivalent in any of the lung samplings or buffer controls. to 100 organisms per sample) are found almost (Figure 1 shows the results from two lungs, X and exclusively (98 per cent) in those with P. carinii Y . ) The antithrombin primers showed efficient p n e ~ m o n i a . ~ The data are also relevant to the uncertain epiamplification of antithrombin-specific sequence from the human DNA in each of the samples (Fig. 2 ) . demiology and biology of P. carinii. The absence of detectable P. carinii in the lung samples tested does not favour the concept that there is regular pulmonDISCUSSION ary carriage of the organism which acts as a source No amplified P. carinii DNA was detected in any for expansion of parasite numbers during immunoof the three reactions taken from each of the 15 lung suppression. Unless some other tissue is a repository
1
2
3
4
5
6
7
c
198
S. E. PETERS ET A L .
for dormant organisms (and there are no data to support such a possibility), fresh infection of the immunosuppressed by the parasite is the more likely event. Reported case clusters of P. carinii pneumonia'' suggest that horizontal transmission between immunosuppressed individuals may be significant, whilst the recent demonstration, by comparative DNA studies, of the fungal origins of P. cariniil'-'' emphasizes the possibility of an, as yet unidentified, environmental source of infectious particles. ACKNOWLEDGEMENTS
This work was funded by the Medical Research Council and the Wellcome Trust. AEW is supported by The Royal Society. REFERENCES 1. Hopkin JM. Pneumocyslis carinii. Oxford: Oxford University Press.
1991.
2 . Meuwissen JHET, Tauber I, Leevwenberg A D E M , Beckers PJA, Sieben M . Parasitologic and serologic observations of infection with pneumocystis in humans. Jfnferr Dis 1977; 1 3 6 43-49. 3. Wakefield AE, Stewart TJ, Moxon ER, Marsh K, Hopkin JM. Infection with PneurnocyJlis curinii is prevalent in healthy Gambian children. Trans R Soc Trop Med 1990; 84: 800-802. 4. Sheldon WH, Subclinical pneumocystis pneumonitis. Am J Dis Child 1959; 97: 287-297. 5. Millard PR, Heryet AR. Observations favouring Pneumocysti.y carinii pneumonia as a primary infection: a monoclonal antibody study o n paraffin sections. J Pathol1988; 154 365-370. 6. Wakefield AE, Pixley FJ, Banerji S , el ul. Amplification of niitochondrial ribosomal RNA sequences from Pneumocystis carinii D N A of rat and human origin. Moi Biochem Parusitol 1991; 43: 69-76. 7. Wakefield AE, Pixley FJ, Banerji S , ei ul. Detection of Pneumoeysri.7 currmi with D N A amplification. Lance1 1990; 336 45 1-453. 8 Wakefield AE. Guiver L, Miller RF, Hopkin J M . D N A amplification on sputum samples for diagnosis of PneumocJstrs carinu pneumonia. Lancer 1991;331: 1378-1379. 9. Reinhardt DJ, Kaplan W, Chandler F W , Morphologic resemblance of rygomycete spores to P. carinii in tissue. Am Re11 Resp Dis 1977; 115: 170-172. 10. Ruebush TK, Weinstein RA, Baehner RL. An outbreak of pneumocystis pneumonia in children with acute lymphocytic leukaemia. Am J Dis Child 1978; 132: 143-148. I I . Edman JC, Kovacs JA, Masur H, et a / . Ribosomal R N A sequences show Pneumocystis carinii to be a member of the fungi. Nature 1988; 334: 519-522. 12 Pixley FJ, Wakefield AE, Banerji S , Hopkin JM. Mitochondria1 gene sequences show fungal homology for Pneumocwtis carinii. Mol Microhiol1991: 5: 1347-1351
