Respiratory Medicine (1996) 90, 153-157

Detection of Pneumocystis carinii sequences in serum by polymerase chain reaction

H. MIYAWAKI*, Ji FUJITA*$, S. HOJO *, M. HARADA-~, Y. YAMAJI*, S. SUGURI~ AND J. TAKAHARA*

*First Department of Internal Medicine, and TDepartment of Parasitology, Kagawa Medical School, Kagawa, Japan

The clinical significance of the detection of Pneumocystis carinii DNA was evaluated, as well as the detection of circulating P. carinii antigen from serum using previously collected samples. Fourteen serum samples from 13 patients were diagnosed positively for P. carinii DNA by polymerase chain reaction (PCR). Ten of 14 episodes (71.4%) of pulmonary complications were compatible with P. carinii pneumonia. Two patients were definitely diagnosed as having had P. carinii pneumonia at autopsy. All patients positive for circulating antigens were also positive for P. carinii DNA, suggesting that the detection of P. carinii DNA by PCR is more sensitive compared to the detection of circulating antigens by the Ouchterlony method. It is concluded that the detection of P. carinii DNA in serum by PCR provides useful information for identifying P. carinii pneumonia.

Introduction

Pneumocystis carinii (P. carinii) is an opportunis- tic pathogen which often causes severe and fatal pneumonia in immunocompromised patients. How- ever, in the early stages of infection, patients can be treated successfully with effective agents against P. carinii, for example, sulphamethoxazole- trimethoprim and/or pentamidine. Therefore, it is very important to establish a highly sensitive, highly specific, widely applicable, and non-invasive diagnos- tic method to detect P. carinii infection.

Bronchoscopy with bronchoalveolar lavage or biopsy is currently the method employed frequently to establish the diagnosis (1). In addition, examination of induced sputum is potentially a rapid, non-invasive method of diagnosing P. carinii pneumonia (2,3).

DNA amplification by the polymerase chain reac- tion (PCR) offers a sensitive and specific means of identifying diverse organisms which may be present in biological samples. Amplification of ribosomal RNA genes has been shown to be useful in increasing the sensitivity of detection of P. carinii DNA from sputum and bronchoalveolar lavage fluid (4-7).

However, reports of the diagnosis of P. carinii infection using serum samples have been limited

Received 25 January 1995 and accepted in revised form 4 July 1995.

fAuthor to whom correspondence should be addressed at: First Department of Internal Medicine, Kagawa Medical School, 1750- 1, Miki-cho, Kita-gun, Kagawa, 761-07, Japan.

0954-611 l/96/030153+05 $12.0010

(6,8,9). This study attempted to evaluate the clinical significance of the detection of P. carinii DNA as well as the detection of circulating P. carinii antigens from serum using previously collected samples.

Materials and Methods PATIENTS

From January 1991 to December 1994, serum samples were collected prospectively from patients with pulmonary complications. During this period, 470 patients who had diseases associated with the risk of P. carinii infection were admitted to the authors’ department. Fifty-six sera (from 55 patients, 11.7%) were collected from patients with acute leukaemia, lymphoma, collagen vascular diseases, multiple myeloma, lung cancer, congenital immunodeficiency, haematological disorders, and from immature infants. All had pulmonary infiltrations such as inter- stitial, alveolar or mixed pattern as demonstrated on chest X-rays. This study attempted to detect P. carinii DNA in these samples using PCR in addition to screening to detect circulating P. carinii antigen using the Ouchterlony technique.

DNA PREPARATION, AMPLIFICATION AND DETECTION (8)

Serum (10~1) was incubated at 60°C for 6 h in a solution (20~1) containing 10 mM Tris-HCL (pH 8.3) 50 mM KCI, 1.5 mM MgCl,, 0.5 mg gelatin, 0.25% Nodidet P-40, 0.25% Tween-20, and 10 pug of proteinase K. The samples were boiled for 10 min

0 1996 W. B. Saunders Company Ltd

154 H. Miyawaki et al.

0 12 3 4 5 67 8 91011

Plate I Detection of Pneumocystis carinii DNA by polymerase chain reaction and Southern blot analysis. The amplified products were confirmed as P. carinii DNA by Southern blot analysis using a labelled fragment of the ferret P. carinii gene as a probe. Lane Nos 0 and 11: control DNA (120 base pairs); Lanes l-10: serum samples. Lane Nos 1 and 6 were diagnosed as negative.

and 10~1 of the supematant was used directly for PCR reaction. The 5S rRNA sequence of P. carinii was amplified by PCR in a reaction mixture (100 ~1) containing 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl,, 1 mg gelatin, 200~~ each deoxy- ribonucleotide triphosphate, 2.5 U of Taq DNA polymerase (Perkin Elmer Cetus), 50pmol each of SS-sense and SS-antisense primers and 10 ~1 of DNA solution. The primers were SS-sense (5’-AGTT ACGGCCATACCTCAGA-3’) and SS-antisense (5’- AAAGCTACAGCACGTCGTAT-3’), generating a 120-bp (base pairs) product by PCR. Amplification was performed in a thermal cycler (Perkin Elmer Cetus) with a three-step cycling programme: 1 min of denaturation at 94°C followed by 2 min of annealing at 55°C and 2 min of elongation at 72°C for a total of 40 cycles. The amplified products were subjected to electrophoresis through 10/20% polyacrylamide gels and the specific P. carinii sequence (120-bp) was identified by visualization with ultraviolet light after ethidium bromide staining. The results were con- firmed for P. carinii DNA by Southern blot analysis using a labelled fragment of the ferret P. carinii gene as a probe (8).

DETECTION OF CIRCULATING ANTIGENS

Pure antigens were first prepared from the lung tissue of rates infected with P. carinii, and a rabbit was sensitized with the antigen to induce anti- P. carinii antibody production. Rat sera containing

antigenic substances then reacted with this antibody when examined by the Ouchterlony technique or immunoelectrophoresis as reported previously (10,ll).

Results

Fifty-six samples were examined out of 55 patients who had respiratory complications. Fourteen serum samples (14/56, 25%) from 13 patients (13/55, 23.6%) [one patient (Patient No. 11) provided two samples from two episodes of respiratory complications during this period] were positive for P. carinii DNA.

Plate 1 shows the results of PCR and Southern blot in selected samples. Table 1 shows patient character- istics of those positive for P. carinii DNA as shown by PCR and/or positive for circulating antigens. Nine of 14 samples were also positive for circulating P. carinii antigens.

Ten of 14 episodes (71.4%) of pulmonary compli- cations were compatible with P. carinii pneumonia. Two patients were definitely diagnosed as having had P. carinii pneumonia at autopsy (Patient Nos. 10 and 13). Probable P. carinii pneumonia (Patient Nos. 1,3, 4, 6, 7, 9, 11 and 12) was determined by chest X-ray findings (interstitial shadow) and a good response to sulphamethoxazole-trimethoprim. In four epi- sodes (Patient Nos. 2, 5, 8 and 1 l), although P. carinii infection was suggested clinically, a definite diagnosis was not obtained because an autopsy was not

Table

I

Patie

nt

chara

cteris

tics

Patie

nt

Age

and

numb

er

sex

Unde

rlying

dis

ease

Ci

rculat

ing

antig

en

Leuc

ocyte

co

unt

(id)

Lymp

hocy

te co

unt

($1)

Resp

onse

to

ST

Prog

nosis

Cl

inica

l dia

gnos

is

1 56

M

Aplas

tic

anem

ia 2

15 F

Th

ombo

cytop

enia

3 44

F Ac

ute

leuka

emia

4 54

M

Acute

leu

kaem

ia 5

48

M My

elody

splas

tic

synd

rome

6

76 M

Lu

ng

canc

er I

56 M

Lu

ng

canc

er 8

80 M

Ma

ligna

nt

lymph

oma

9 OM

Im

matur

e inf

ant

10

OM

Immu

node

ficien

cy

11

64M

Acute

leu

kaem

ia 11

’ 64

M Ac

ute

leuka

emia

12

OF

Mille

r-Diek

en

synd

rome

13

32

F

Derm

atomy

ositis

+ + - + + + - - + + - + -

1600

12

8 23

00

46

5900

53

1 93

00

1395

43

00

1032

16

900

61

6 19

00

798

2000

32

0 58

00

1450

44

00

308

5100

40

8 20

10

0 10

0 20

500

36

90

3100

62

+ - + + NA + + - + - + NA + NA

Impr

oved

Pr

obab

le PC

P*

Dead

Un

defin

ite

Impr

oved

Pr

obab

le PC

P Im

prov

ed

Prob

able

PCP

Dead

Un

defin

ite

Impr

oved

Pr

obab

le PC

P Im

prov

ed

Prob

able

PCP

Dead

Un

defin

ite

Impr

oved

Pr

obab

le PC

P De

ad

PCP

Impr

oved

Pr

obab

le PC

P De

ad

Unde

finite

Im

prov

ed

Prob

able

PCP

Dead

PC

P

ST,

Sulph

ameth

oxaz

ole-tr

imeth

oprim

; NA

, no

t an

alyse

d; PC

P,

Pneu

mocy

stis

carin

ii pn

eumo

nia.

*PCP

wa

s de

term

ined

by

ches

t X-

ray

findin

gs

(inter

stitia

l sh

adow

) an

d go

od

respo

nse

to ST

.

156 H. Miyawaki et al.

performed. Interestingly, all patients who were posi- tive for circulating antigens were also positive for P. carinii DNA, suggesting that the detection of P. carinii DNA by PCR is a more sensitive indicator than the detection of circulating antigen by the Ouchterlony method. In addition, one patient (Patient No. 13) who was proved to have P. carinii infection at autopsy, was negative for circulating P. carinii antigens, also suggesting the possibility that the PCR method is more sensitive than the circulat- ing antigen assay for the diagnosis of P. carinii pneumonia. Furthermore, in Patient No. 3, although circulating antigens were not detected (circulating antigens became positive 1 week later), P. carinii DNA was detected even at the onset of the pul- monary complication, suggesting that the detection of P. carinii DNA by PCR is useful for the early diagnosis of P. carinii pneumonia.

Patients who had negative results of P. carinii DNA by PCR were not clinically diagnosed as having P. carinii pneumonia.

Discussion

The present study analysed the significance of the detection of serum P. carinii DNA as well as the detection of circulating antigen using previously col- lected serum samples at the onset of pulmonary complications. Fourteen serum samples from 13 patients were diagnosed positively for P. carinii DNA by polymerase chain reaction (PCR). Ten of 14 episodes (71.4%) of pulmonary complications were compatible with P. carinii pneumonia. All patients positive for circulating antigens were also positive for P. carinii DNA, suggesting that the detection of P. carinii DNA by PCR is more sensitive compared to the detection of circulating antigens by the Ouchterlony method.

It has been shown that P. carinii antigens are present in serum from patients with P. carinii pneu- monia and are potentially useful for the early diag- nosis of P. carinii pneumonia (10,ll). Pifer et al. reported that positive circulating antigen results were obtained from 95% of fresh serum samples and 62% of frozen serum samples of the patients proven to have P. carinii pneumonia by lung aspiration (10). However, detection of circulating P. carinii antigens in serum requires complicated techniques, and is difficult to apply widely.

Conventional diagnosis of P. carinii pneumonia has usually been based on pathologic examination of appropriately stained tissue obtained via techniques of transbronchial biopsy, percutaneous lung biopsy, or open lung biopsy. The diagnostic yield of the

procedures is variable, and complications are inher- ent, especially in the setting of profound hypoxaemia and bleeding diatheses (1). Bronchoalveolar lavage is a low morbidity procedure that has been used to diagnose P. carinii pneumonia in patients with lym- phoproliferative disorders and acquired immuno- deficiency syndrome (1). In addition, examination of induced sputum is potentially a rapid, non-invasive method of diagnosing P. carinii pneumonia in patients with acquired immunodeficiency syndrome (AIDS) (2,3).

The PCR method can be used for the diagnosis of infectious diseases because it enables the detection of pathogen DNA in very small quantities of specimens in a short time. Recently, it has been reported that the diagnosis of P. carinii pneumonia by the PCR method using bronchoalveolar lavage (BAL) and sputum samples is highly sensitive and specific com- pared to conventional diagnostic methods (46). However, BAL is not a trivial examination in severe respiratory failure and the cough in P. carinii pneu- monia is almost always non-productive. Thus, this study attempted to detect P. carinii DNA by the PCR method in serum, which is easily obtained and widely applicable, to establish a non-invasive diagnostic method for the identification of P. carinii pneumonia.

It has been reported that P. carinii DNA was present in the serum of P. carinii-infected rodents and in the serum of AIDS patients with P. carinii pneumonia (9). Sepkowitz et al. reported that PCR assay for the P. carinii dihydrofolate reductase gene is equally applicable to human pneumocytosis and capable of detecting P. carinii both in BAL fluid and in the serum of AIDS patients with active P. carinii pneumonia (12). They also examined the relationship between P. carinii DNA in serum and P. carinii infection using the corticosteroid-treated rat model of pneumocytosis, and demonstrated that P. carinii DNA was detected in rat lungs by 4 weeks of immunosuppression, and in their serum by 6 weeks (12). However, it has also been reported that PCR did not detect P. carinii in blood, or sera (13).

In the present study, there appears to be a high degree of association between active P. carinii pneu- monia and the presence of P. carinii DNA in the serum. These results suggest that serum PCR for P. carinii may be useful as a diagnostic procedure in some patients with suspected P. carinii infection. This would be of particular value because current diagnosis of P. carinii pneumonia frequently requires an invasive pro- cedure such as bronchoscopy to obtain respiratory specimens for direct microscopic examination.

The pathogenetic significance of the detection of P. carinii DNA or circulating antigens has not yet

Pneumocystis carinii sequences in serum 157

been established. However, the reported occurrence of extrapulmonary pneumocytosis (14) suggests that in some patients, P. carinii may have a blood-borne phase. Evidence by DNA amplification that P. carinii may be blood-borne offers important information concerning the natural history of this infection and would readily explain the pathogenesis of extra- pulmonary pneumocytosis.

The results of the present study indicate that the detection of P. carinii DNA from serum by PCR can be used for diagnosis of P. carinii pneumonia. Although false-positive results by PCR should be considered, because of the repeated PCR positivity of these samples and the presence of several negative controls at each step of digestion and amplification, it is not believed that these specimens represent false- positives due to cross-contamination. To monitor P. carinii DNA in serum, especially in immunocom- promised patients, the PCR method seems to be useful for confirming the existence of P. carinii infection.

It is concluded that the detection of P. carirzii DNA from serum by PCR amplification provides a valu- able tool for rapid diagnosis and early treatment of P. carinii pneumonia, and may obviate the need for invasive examination.

Acknowledgements

We would like to thank Dr. Moriyasu Tsuji for his technical support in detecting P. carinii circulating antigens.

References

1. Ognibene FP, Shelhamer J, Gill V er al. The diagnosis of Pneumocystis car&ii pneumonia in patients with the acquired immunodeficiency syndrome using sub- segmental bronchoalveolar lavage. Am Rev Respir Dis 1994; 129: 929-932.

2. Bigby TD, Margolskee D, Curtis JL et al. The useful- ness of induced sputum in the diagnosis of Pneumo- cystis carinii pneumonia in patients with the acquired

3.

4.

5.

6.

I.

8.

9.

10.

11.

12.

13.

14.

immunodeficiency syndrome. Am Rev Respir Dis 1986; 133: 515-518. Kovacs JA, Ng VL, Masur H et a/. Diagnosis of Pneumocystis carinii pneumonia: improved detection in sputum with use of monoclonal antibodies. N Engl J Med 1988; 318: 589-593. Wakefield AE, Guiver L, Miller RF, Hopkin JM. DNA amplification on induced sputum samples for diagnosis of Pneumocystis carinii pneumonia. Lancet 1991; 337: 1378-1379. Wakefield AE, Pixley FJ, Banerji S et al. Detection of Pneumocystis carinii with DNA amplification. Lancet 1990; 336: 451453. Lipschik GY, Gill VJ, Lundgren JD et al. Improved diagnosis of Pneumocystis carinii infection by polymer- ase-chain reaction on induced sputum and dlood. Lancet 1992: 340: 203-206. Leigh TR, Kangro HO, Gazzard BG, Jeffries DJ, Collins JV. DNA amplification by the polymerase chain reaction to detect sub-clinical Pneumocystis carinii colonization in HIV-positive and HIV-negative male homosexuals with and without respiratory symptoms. Respir Med 1993; 87: 525-529. Kitada K, Oka S, Kimura S et al. Detection of Pneumocystis carinii sequences by polymerase chain reaction: animal models and clinical application to noninvasive specimens. J Clin Microbial 1991; 29: 1985-1990. Schluger N, Godwin T, Sepkowitz K et al. Application of DNA amplification to pneumocystosis: presence of serum Pneumocystis carinii DNA during human and experimentally induced Pneumocystis carinii pneumo- nia. J Exp Med 1992; 176: 1327-1333. Pifer LL, Hughes WT, Stagno S, Woods D. Pneumo- cystis carinii infection: evidence for high prevalence in normal and immunosuppressed children. Pediatr 1978; 61: 3541. Tsuji M. Diagnosis for Pneumocystis carinii pneumonia. Asian Med 1991; 34: 572-577. Sepkowitz K, Schluger N, Godwin T, Armstrong D, Cerami A, Bucal R. DNA amplification in experimental pneumocystosis: characterization of serum Pneumo- cystis carinii DNA and potential P. carinii carrier states. J Infect Dis 1993: 168: 421426. Ro;x P, Lavrard I, Poirot JL et al. Usefulness of PCR for detection of Pneumocystis carinii DNA. J Clin Microbial 1994; 32: 232&2326. Telzak EE, Cote RJ, Gold JWM, Campbell SW, Armstrong D. Extrapulmonary Pneumocystis carinii infections. Rev Infect Dis 1990; 12: 380.