Molecular and Cellular Probes (1992) 6, 51 -58
Detection of human cytomegalovirus in peripheral mononuclear cells and urine samples using PCR P. Stanier,'* A . D . Kitchen, D. L. Taylor,3 and A . S. Tyms3 'Molecular Biology Unit, Institute of Obstetrics and Gynaecology, RPMS, Queen Charlottes and Chelsea Hospital, Goldhawk Road, London W6 OXG, UK, 'Department of Microbiology, Regional Blood Transfusion Centre, Brentwood, Essex CM15 8DP, UK and 3 Medical Research Council Collaborative Centre, Buttonhole Lane, Mill Hill, London NW7 7AD, UK (Received 17 April 1991, Accepted 21 June 1991)
Samples of peripheral blood lymphocytes from 105 different blood donors were investigated for the presence of human cytomegalovirus (HCMV) DNA using the polymerase chain reaction (PCR) with primers specific for the Pst I w fragment (IE region) . Viral DNA sequences were detected in 53 samples, a fifth of which had been previously serotyped as HCMV negative . In the latter cases, Western blot analysis re-determined two out of three individuals that were resampled as seropositive. PCR could therefore be used to extend existing methods employed for the identification of HCMV infected blood samples prior to transfusion to individuals in high risk groups . In addition, the value of PCR as a diagnostic test was evaluated in a small pilot study by comparing the results obtained with urine samples from babies suffering congenital infection and from other high risk patients, with data obtained by isolation of infectious virus or through the detection of immediate early antigens in infected cultures . Data from this study indicated that PCR is at least as sensitive as the other methods used in HCMV diagnosis. KEYWORDS :
Cytomegalovirus, polymerase chain reaction .
INTRODUCTION The ubiquitous nature of HCMV along with its potential pathogenicity in certain patient groups demands a means for improved diagnosis and the availability of antiviral chemotherapy . Disease due to HCMV may be wide-ranging in the immunocompromised with significant morbidity in AIDS patients' and organ transplant recipients .' The frequency of infection in the normal population has been shown to increase with age,' the natural route of transmission being by close or intimate contact . However, iatrogenic infection can occur through blood transfusion or organ transplantation .' In transplant recipients, HCMV infection can cause serious disease after primary infection or as a result of reactivation of endogenous virus . To avoid post transfusion disease it has become policy that HCMV seronegative blood products
should be used for transfusion to allogeneic transplant patients, most importantly when the recipient is seronegative for HCMV .5 Likewise, the use of virusfree blood products administered to newborn who are seronegative for HCMV seriously reduces the chances of post-transfusion disease .' In the blood transfusion service evidence of previous infection with HCMV is currently determined by antibody status . A number of other procedures have been developed to detect HCMV including cell culture, the detection of early viral antigens' or by DNA hybridization procedures .' These methods are often limited by a lack of sensitivity and the time to reporting . There is good evidence that HCMV persists in leucocytes after primary infection,' probably in the monocyte/lymphocyte fraction, although apart
Author to whom correspondence should be addressed . 0890-8508/92/010051 + 08 $03 .00/0
51
C~ 1992 Academic Press Limited
52
P. Stanier et al.
from one report, 10 attempts to culture virus from asymptomatic, seropositive individuals have been unsuccessful . The PCR technique allows amplification of both DNA and RNA sequences in low copy number with the potential for rapid and sensitive detection of the viral genome." In an earlier paper we have reported preliminary findings suggesting that PCR can be used to detect HCMV in the peripheral blood mononuclear cells from seropositive and from some seronegative asymptomatic blood donors ." Here we report the results of an extended study of over 100 blood donors along with preliminary findings of the application of PCR in the analysis of urine samples from transplant patients . The results indicate that PCR may have an important application in the detection and prevention of disease due to HCMV .
MATERIALS AND METHODS Blood donor mononuclear cells and clinical urine samples Mononuclear cells were separated from blood samples of 105 donors attending the Regional Blood Transfusion Centre, Brentwood, Essex, using FicollTriosil gradients . All donor samples were serotyped for HCMV using an enzyme immunoassay kit detecting total antibody (Abbott Diagnostics) . The samples were coded so that they could be studied further in a blind form . Urine samples used in this study were from three babies with severe post-natal HCMV infection (St . Mary's Hospital Medical School, Paddington) and from 16 coded normal, pregnant or transplant patients typed for cell HCMV by cell culture and/or immediate early antigens (St . George's Hospital Medical School, Tooting). Detection of immediate early antigens was as described by Steel et al .'
Hybridization Human embryonic fibroblasts (HEF) were cultured and maintained as previously reported ." HEF monolayers infected with 52 different clinical isolates of HCMV were harvested at appropriate times postinfection, total DNA was extracted, deproteinised and ethanol precipitated . For controls, DNA was also prepared from uninfected HEF cells, cells infected with adenoviruses (26 serotypes), herpes simplex viruses (two serotypes), simian-like CMV (Colburn stain) E. coli, and salmon sperm . DNA were digested with
either Eco RI or Sma I and the restriction fragments separated on 1 % agarose gels stained with ethidium bromide . DNA fragments were transferred to nylon membranes (Hybond-N, Amersham Int.) and hybridized either with the Eco RI J fragment or its Pst I w fragment subclone from the HCMV strain AD169 after random prime labelling with [a-32P]dCTP . 14
Polymerase chain reaction Digestion of the Eco RI J Fraction using the restriction endonuclease Pst I, yielded fragments which were subcloned into the single stranded phage vector M13mp 18. The DNA sequence of resulting fragments was obtained using the method of Sanger et al." PCR primers (20 bases long) spanning a region of 240 by and an internal 17 base oligonucleotide probe were selected from the region corresponding to the Pst I w fragment16 which mapped to the immediate early gene region 2 of HCMV strain Towne . 17 HCMV1 : 5' CCCGACTTTACCATCCAGTA 3' HCMV2 : 5' AAGACGAAGAGGAACTATCT 3' Probe : 5' GGGTGAAGAAGTCGAAA 3' DNA from blood was prepared for PCR by first isolating the mononuclear cell fraction . Lymphocytes were separated from 5 ml EDTA treated whole blood using ficoll density gradients . After washing in phosphate buffered saline, the lymphocyte pellets were resuspended in 215 pl of water, 25 pl of 10 x PCR buffer containing 0.45% NP40, 0 . 45% Tween 20, and 10 pl of Pronase E (20 mg ml -1 ) . After digestion at 55 ° C for 1-2 h, the DNA was phenol-chloroform extracted and ethanol precipitated . After resuspension, approximately one fifth of the DNA was used for each PCR assay . This was carried out according to the recommended protocol issued with AmpliTaq DNA polymerase (Perkin Elmer Cetus) with a specific annealing temperature of 58 ° C and 30-35 cycles of amplification . Resulting PCR products were analysed by electrophoresis on 2% Nusieve (FMC), 1 % agarose gels stained with ethidium bromide. To test for the specificity of the PCR products obtained, all gels were also Southern blotted and hybridized to an internal oligonucleotide probe, end-labelled with [a- 32 P]dATP . Viral DNA was prepared from urine samples using a standard ultracentrifugation method . Urine samples (6 ml) were clarified by centrifugation at 1000 x g for 10 min, and virus was pelleted by ultracentrifugation at 80,000 g for 1 h . Pellets were resuspended in 200 pl of buffer containing 0. 1 M NaCl, 10mM Tris-HCI and 1mM EDTA (pH 8 . 0). The resuspended DNA was incubated at 37 ° C for 3 h with 0-1 0/6 SDS, 20 pg
Detection of HCMV using PCR
RNAse and 40 µg proteinase K . After phenol-chloroform extraction, the DNA was ethanol precipitated and resuspended in the PCR assay mixture.
marrow transplant patients and neonates who suffered congenital or post-natal infection was used to determine the specificity of the Eco RI J fragment or its Pst I w fragment subclone by Southern blot analysis . With the majority of samples, hybridization to characteristic bands was observed . For example, the Eco RI J fragment hybridises to a 10 . 6 kb Eco RI band and the Pst I w fragment hybridises to a 6 . 5 kb Sma I band (Fig. 1) . Hybridization to alternative restriction fragments of DNA from some isolates denoted the loss or gain of restriction sites, presumably reflecting heterogeneity between HCMV isolates .
RESULTS Hybridization Analysis of DNA from 52 clinical isolates of HCMV recovered from AIDS patients, 13 heart or bone(a)
Ln
N
N
to
53
CO
r•
o
r-
I
c r-
F- 10 . 6 k b
10 . 6 kb
a,
a, c Y
M
3 0 H
v0 Y
Conservation of DNA sequences in different human cytomegalovirus strains. (a) Hybridization of the Eco RI J fragment to HCMV strains digested with the restriction enzyme Eco RI . (b) Hybridization of the Pst I w fragment to HCMV strains digested with the restriction enzyme Sma I . Fig . 1 .
54
P. Stanier et al.
Specificity and sensitivity of PCR
Identification of HCMV in urine samples
A 240 by fragment, defined by a pair of primers specific for the Pst I w fragment, which contains the immediate early gene region, was generated by PCR . PCR analysis was carried out on DNA from 23 different HCMV isolates (Fig . 2) . In all cases, a positive signal was obtained after hybridization with the internal probe . In addition, all non-HCMV control
Consecutive urine samples from two babies suffering post-natally acquired infection and one with a congenital infection, all with infectivity titres in excess of 5 x 10" pfu ml - ' of urine, were highly positive by PCR (data not shown) . HCMV is shed less freely in the urine of organ transplant patients and such samples therefore provided a more appropriate target for PCR analysis . Data were obtained from 16 coded urine samples from transplant patients and from control urine samples from normal individuals and pregnant women . The results are shown in Fig . 3 and the
DNA including DNA from the most related of the human herpesviruses, HHV-6 were negative by PCR . Using the conditions specified, no non-specific bands have been observed with these PCR primers . The sensitivity of the PCR test was established using dilutions of the plasmid containing the Eco RI J fragment. Between 1-10 fg of DNA was routinely amplified so that a band was visible on ethidium stained gels . This corresponded to between 300 and 3,000 copies of the HCMV genome in 10 5 cells .
'0
2
~
N
M
W LO U, = N
OD
r
O F
decoded data presented in Table 1 . When the samples were processed by ultracentrifugation, no falsenegatives were detected although two asymptomatic patients which were negative for cell-culture and IEA determination, were positive by PCR . When the urine samples were used directly after alkali denaturation
J
0 Q
U U
S
M
Y
-
0
Fl-
r- w
(D
W
U)
2
N-
240 by
240 by
Fig. 2 . PCR amplification of 23 different strains of human cytomegalovirus . All of the amplification products were shown to be specific following hybridization with an oligonucleotide probe internal to the PCR primers .
16
Fig . 3.
15
14
13
12
II
10
9
8
7
6
5
4
2
I
B
T
M
PCR amplification of human cytomegalovirus in urine samples . In the example shown, DNA was prepared from the urine samples using the ultracentrifugation procedure . Positive amplification products were confirmed following hybridization with an oligonucleotide probe internal to the PCR primers .
Detection of HCMV using PCR
55
Table 1 . Detection of HCMV in urine samples using PCR . PCR data from 16 samples, prepared either by ultracentrifugation or by direct use were compared to virological analysis by cell culture (CC) or through the detection of immediate early antigens (IEA) . *Samples that were weakly hybridization-positive by hybridization despite bands not being visible on the ethidium stained gel . Days needed to culture virus are detailed where possible as an indication of the amount of virus present in those samples (e.g . low titre =longer time to culture) . ND, no data; +, positive ; -, negative. Sample no . 1 2
3 4 5 6 7 8 9 10 11 12
13 14 15 16
(see Methods)
Details Pregnant BMT (adenovirus) Renal Tr Leukemia Pregnant Normal Heart Tr BMT Normal Infant BMT Infant BMT Pregnant Renal Tr Renal Tr
PCR was less successful which
appeared to be as a result of inhibition of the Taq polymerase .
Identification of HCMV in blood donors DNA was prepared from the mononuclear fraction from PBLs of 105 blood donor samples and examined for HCMV sequences by PCR (Fig . 4). Approximately 50% were positive by PCR with confirmation by hybridization with the internal probe . A summary of results after decoding is presented in Table 2 . On no occasion was a PCR product detected in negative control DNA . After decoding, it was determined that three of the PCR-negative samples were from seropositive individuals and 11 of the PCR-positive samples were from individuals that were seronegative . Three of the latter patients were resampled as described in Stanier et al. 12
Ultra
Direct
+ +• +' + + + + + +
+ + + + -
+
+
CC
IEA
6 days 6 days -
+ ND
+ 8
days
+ 10
days
+ 7
days
ND ND ND ND +
ruled out the possibility that false-positive results were due to contamination . 18 The reproducibility of the data was confirmed when individuals who were seronegative but PCR positive, gave the same results when resampled . Western blotting, regarded as a useful additional antibody test, 19 was weakly positive for two of the three individuals, 12 suggesting that antibody to the virus was present, but at very low levels . Whether these seronegative but HCMV genome positive individuals are infectious was not determined . Thiers et al ." demonstrated that PCRpositive serum samples from patients negative for all serological markers for hepatitis-B virus did carry infectious virus . It must be considered likely therefore that blood from the individuals described here who were PCR positive, although seronegative, must be a potential cause of infection when administered to patients at risk of HCMV infection . It has been reported that antigenic heterogeneity between different HCMV strains can cause falsenegative results with antibody tests 21 and indeed this
DISCUSSION AND CONCLUSIONS
may explain the difference between the PCR and antibody tests . The primers used in this study map to
It is clear from this study that a number of individuals considered to be negative for HCMV, on the basis of their antibody status, were apparently harbouring virus in their peripheral blood mononuclear cells . Careful handling of samples and rigorous controls
a region of the HCMV genome encoding an immediate early gene product, 17 thought to be highly conserved . We have cross-hybridized a probe spanning this gene product in 52 different HCMV isolates and also specifically amplified sequences in this region in 23 different isolates without failure . Recent reports
56
P. Stanier et al .
M +
Antibody typing + + + -
I- 240 by
WE
240 by
Fig. 4. PCR amplification of human cytomegalovirus from peripheral blood mononuclear cells in asymptomatic blood donors . Serological typings determined by a total antibody detection kit are shown . Positive amplification products were confirmed following hybridization with an oligonucleotide probe internal to the PCR primers .
PCR results from peripheral blood mononuclear cells obtained from blood donors compared with their status for HCMV positively determined by antibody detection Table 2 .
PCR
+
42
3
11
49
Ab
suggest serological cross-reactivity22 and homology at the DNA level23 between HCMV and human herpesvirus-6 (HHV-6) . The IE region of the HCMV genome that we amplified did not cross-hybridize with HHV-6 or with any of the other viruses tested . A number of other recent reports have described the PCR for the successful detection of HCMV in
blood", " although in the main, viral DNA was detected only in patients with HCMV viremia . Cassol et al .," however, reported two cases where HCMV was detected in seropositive asymptomatic patients and the findings of Morris et al ." were similar to those reported here . In earlier reports, Shrier et a! .9 were able to detect HCMV RNA transcripts in the blood of seropositive asymptomatic patients and also one of 12 seronegative patients by in situ hybridization whilst Nelson et al." detected HCMV immediate early RNA in approximately 0.01% of PBMCs from 20 asymptomatic HCMV positive individuals using the same technique . Interestingly, PCR has recently been used to detect murine cytomegalovirus DNA in latently infected mice, 29 supporting the view for the sensitivity of this test in any HCMV screening programme . The potential of PCR as a diagnostic tool was further illustrated by the ability to detect DNA in the urine samples of patients who were positive by
Detection of HCMV using PCR culture or by the detection of immediate early antigens . It was possible to detect reproducibly HCMV from a series of urine samples from two babies that were known to be shedding large quantities of the virus (in excess of 5 x 104 pfu ml -1 ) . This was not unexpected and agreed with a larger study reported by Demmler et al.' who detected HCMV sequences in 100% (n=46) of culture-positive urine samples from congenitally infected babies . In their study, they were able to detect HCMV sequences by adding a small quantity of the urine sample direct to the PCR mixture. Analysis of urine samples from transplant patients in the present study indicated that this method was unreliable . Good sensitivity (100%) and specificity (83%) when compared to cell culture and/ or by the detection of immediate early antigens was only obtained when the urine samples were prepared by ultracentrifugation . The unreliability of the direct method is apparently due to the presence of an inhibitor of the Taq polymerase found in the urine, but is probably also related to the lower titres of virus encountered in these types of patients . The recent PCR technology has been used to facilitate the detection of DNA sequences in a variety of diagnostic applications . We have shown that HCMV sequences can be detected both in the mononuclear cells of asymptomatic individuals and in urine samples from at risk individuals using PCR . The first of these findings may prove to be helpful in the screening of blood donations before administering to specific patient groups . The second enables rapid screening of newborns, which may act as an early alert for special development follow-up procedures to be implemented .
ACKNOWLEDGEMENTS HCMV specific DNA probes were kindly provided by Dr Fleckenstein, Erlangen, Gemany ; Epstein-Barr virus DNA was provided by Dr Paul Farrall, Ludwig Institute, St . Mary's Hospital Medical School, London; human herpesvirus type 6 was provided by the late Dr Bob Hoeness, National Institute for Medical Research, London . We are also much in debt to Yvonne Tryhorn (St . George's Hospital Medical School, London) for supplying coded urine samples from both normal and transplant patients .
REFERENCES 1 . Quinnan, G. V., Mazur, H ., Rook, A . H . et al. (1984) . Herpes infections in the acquired immune deficiency syndrome. The Journal of the American Medical Association 252, 72-7 . 2 . Glen, J . (1981) . Cytomegalovirus infections following renal transplantation . Review of Infectious Diseases 3, 1151-78 .
57
3 . Griffiths, P . D . & Baboonian, C . (1984) . A prospective study of primary cytomegalovirus infection during pregnancy (final report) . British Journal of Medicine 91, 307-315 . 4 . Wilhelm, J . A ., Matter, L . & Schopfer, K . (1986) . The risk of transmitting cytomegalovirus to patients receiving blood transfusions . The Journal of Infectious Diseases 154, 169-171 . 5 . Ho, M . (1982) . Cytomegalovirus : Biology and Infection . pp. 95-104 . New York : Plenum Medical . 6 . Yeager, A . S ., Crumet, F . C., Hafleigh, E . B ., Arvin, A . M ., Bradley, J . S . & Prober, C . C . (1981) . Prevention of transfusion-acquired cytomegalovirus infections in newborn infants. Journal of Paediatrics 98, 287--91 . 7 . Steel, H . M., Booth, J . C ., Tryhorn, Y . S . & Stern H . (1988) . A simple immunoalkaline phosphatase method for the rapid diagnosis of cytomegalovirus (CMV) infection . Serodiagnosis and Immunotherapy of Infectious Diseases 2, 193-200 . 8 . Spector, S . A ., Rua, J . A ., Spector, D. L . & McMillan, R . (1984) . Detection of human cytomegalovirus in clinical specimens by DNA-DNA hybridisation . The Journal of Infectious Diseases 150, 121-6 . 9 . Schrier, R . D ., Nelson, J . & Aldstone, M . D . A . (1985) . Detection of human cytomegalovirus in peripheral blood lymphocytes in a natural infection . Science 230, 1048-51 . 10 . Diosi, P ., Moldovan, E . & Tomescu, N . (1969) . Latent cytomegalovirus infection in blood donors . British Medical journal 4, 660-2 . 11 . Saiki, R . K ., Gelfond, D . H ., Stoffel, S . et al. (1988) . Primerdirected enzymatic amplification of DNA with a thermostable DNA polymerase . Science 239, 487-94 . 12 . Stanier, P., Taylor, D . L ., Kitchen, A. D ., Wales, N ., Tryhorn, Y . & Tyms, A . S . (1989) . Persistence of cytomegalovirus in mononuclear cells in peripheral blood from blood donors . British Medical journal 299, 897-8. 13 . Taylor, D . L ., Taylor-Robinson, D ., Jefferys, D . J . & Tyms, A . S . (1988) . Characterisation of cytomegalovirus isolates from patients with AIDS by DNA restriction analysis . Epidemiology and Infection 101, 483-94. 14 . Feinberg, A . P. & Vogelstein, B . (1984) . A technique for radiolabelling DNA fragments to a high specific activity . Analytical Biochemistry 137, 266-7 . 15 . Sanger, F ., Nicklen, S . & Coulsen, A . R . (1977) . DNA sequencing with chain temination inhibitors . Proceedings of the National Academy of Science, USA 74, 5463-6 . 16 . Greenaway, P. J ., Oram, J. D ., Downing, R . G . & Patel, K . (1982) . Human Cytomegalovirus DNA : BamH1, EcoRl, and Pstl restriction endonuclease cleavage maps . Gene 18, 355-60 . 17. Sternberg, R . M., Witte, P . R. & Stinski M . F . (1985) . Multiple spliced and unspliced transcripts from human cytomegalovirus immediate-early region 2 and evidence for a common initiator site within immediateearly gene region 1 . Journal of Virology 56, 665-75 . 18. Kwok, S . & Higuchi, R . (1989) . Avoiding false positives with PCR . Nature 339, 237-8 . 19. Towbin, H ., Staehelin, T . & Gordon, J . (1979) . Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets : Procedure and some applications. Proceedings of the National Academy of Science, USA 76,4350-4 . 20. Thiers, V ., Nakajima, E., Kremsdorf, D . et al. (1988) . Transmission of hepatitis B from hepatitis-B-seronegative subjects . Lancet ii, 1273-6 .
58
P. Stanier et al.
21 . Faix, R . C. (1985) . Cytomegalovirus antigenic heterogeneity can cause false-negative results in indirect hemagglutination and complement fixation antibody assays. Journal of Clinical Microbiology 22, 768-71 . 22 . Larcher, C ., Huemer, H . P ., Margreiter, R . & Dierich, M . P . (1988) . Serological crossreaction of human herpesvirus-6 with cytomegalovirus . Lancet ii, 963-4 . 23 . Efstathiou, S ., Compels, U . A ., Craxton, M. A . et al . (1988). DNA homology between a novel human herpes virus (HHV-6) and human cytomegalovirus . Lancet i, 575-6 . 24 . Shibata, D ., Martin, W . J., Appleman, M. D ., Causey, D . M., Leedom, J. M . & Arnheim, N . (1988) . Detection of cytomegalovirus DNA in peripheral blood of patients infected with human immunodeficiency virus . The Journal of Infectious Diseases 158, 1185-92 . 25 . Jiwa, N . M ., Van Cemert, G . W ., Raap, A . K . et al. (1989) . Rapid detection of human cytomegalovirus DNA in peripheral blood leukocytes of viremic transplant recipients by the polymerase chain reaction . Transplantation 48, 72-6 . 26 . Cassol, S . A ., Poon, M-C ., Pal, R . et al. (1989) . Primer-
mediated enzymatic amplification of cytomegalovirus (CMV) DNA : application to the early diagnosis of CMV infection in marrow transplant recipients . Journal of Clinical Investigation 83, 1109-15 . 27 . Morris, D . J ., Kimpton, C . P . & Corbitt, G . (1989) . Persistance of cytomegalovirus in peripheral blood from blood donors. British Medical journal 2", 1164-5 . 28 . Nelson, J . A ., Cnann Jr ., J . W . & Ghazal, P. (1990) . Regulation and tissue specific expression of human cytomegalovirus. Current Topics in Microbiology and Immunology 154, 75-100 . 29 . Klotman, M . E ., Henry, S . C ., Greene, R . C ., Brazy, P . C ., Klotman, P. E . & Hamilton, J . D . (1990) . Detection of mouse cytomegalovirus nucleic acid in latently infected mice by in vitro enzymatic amplification . The Journal of Infectious Diseases 161, 220-5 . 30 . Demmler, C. J ., Buffone, G . J ., Schimbor, C. M. & May, A . M . (1988) . Detection of cytomegalovirus in urine from newborns by using polymerase chain reaction DNA amplification . The Journal of Infectious Diseases 158, 1177-84 .
Detection of human cytomegalovirus in peripheral mononuclear cells and urine samples using PCR P. Stanier,'* A . D . Kitchen, D. L. Taylor,3 and A . S. Tyms3 'Molecular Biology Unit, Institute of Obstetrics and Gynaecology, RPMS, Queen Charlottes and Chelsea Hospital, Goldhawk Road, London W6 OXG, UK, 'Department of Microbiology, Regional Blood Transfusion Centre, Brentwood, Essex CM15 8DP, UK and 3 Medical Research Council Collaborative Centre, Buttonhole Lane, Mill Hill, London NW7 7AD, UK (Received 17 April 1991, Accepted 21 June 1991)
Samples of peripheral blood lymphocytes from 105 different blood donors were investigated for the presence of human cytomegalovirus (HCMV) DNA using the polymerase chain reaction (PCR) with primers specific for the Pst I w fragment (IE region) . Viral DNA sequences were detected in 53 samples, a fifth of which had been previously serotyped as HCMV negative . In the latter cases, Western blot analysis re-determined two out of three individuals that were resampled as seropositive. PCR could therefore be used to extend existing methods employed for the identification of HCMV infected blood samples prior to transfusion to individuals in high risk groups . In addition, the value of PCR as a diagnostic test was evaluated in a small pilot study by comparing the results obtained with urine samples from babies suffering congenital infection and from other high risk patients, with data obtained by isolation of infectious virus or through the detection of immediate early antigens in infected cultures . Data from this study indicated that PCR is at least as sensitive as the other methods used in HCMV diagnosis. KEYWORDS :
Cytomegalovirus, polymerase chain reaction .
INTRODUCTION The ubiquitous nature of HCMV along with its potential pathogenicity in certain patient groups demands a means for improved diagnosis and the availability of antiviral chemotherapy . Disease due to HCMV may be wide-ranging in the immunocompromised with significant morbidity in AIDS patients' and organ transplant recipients .' The frequency of infection in the normal population has been shown to increase with age,' the natural route of transmission being by close or intimate contact . However, iatrogenic infection can occur through blood transfusion or organ transplantation .' In transplant recipients, HCMV infection can cause serious disease after primary infection or as a result of reactivation of endogenous virus . To avoid post transfusion disease it has become policy that HCMV seronegative blood products
should be used for transfusion to allogeneic transplant patients, most importantly when the recipient is seronegative for HCMV .5 Likewise, the use of virusfree blood products administered to newborn who are seronegative for HCMV seriously reduces the chances of post-transfusion disease .' In the blood transfusion service evidence of previous infection with HCMV is currently determined by antibody status . A number of other procedures have been developed to detect HCMV including cell culture, the detection of early viral antigens' or by DNA hybridization procedures .' These methods are often limited by a lack of sensitivity and the time to reporting . There is good evidence that HCMV persists in leucocytes after primary infection,' probably in the monocyte/lymphocyte fraction, although apart
Author to whom correspondence should be addressed . 0890-8508/92/010051 + 08 $03 .00/0
51
C~ 1992 Academic Press Limited
52
P. Stanier et al.
from one report, 10 attempts to culture virus from asymptomatic, seropositive individuals have been unsuccessful . The PCR technique allows amplification of both DNA and RNA sequences in low copy number with the potential for rapid and sensitive detection of the viral genome." In an earlier paper we have reported preliminary findings suggesting that PCR can be used to detect HCMV in the peripheral blood mononuclear cells from seropositive and from some seronegative asymptomatic blood donors ." Here we report the results of an extended study of over 100 blood donors along with preliminary findings of the application of PCR in the analysis of urine samples from transplant patients . The results indicate that PCR may have an important application in the detection and prevention of disease due to HCMV .
MATERIALS AND METHODS Blood donor mononuclear cells and clinical urine samples Mononuclear cells were separated from blood samples of 105 donors attending the Regional Blood Transfusion Centre, Brentwood, Essex, using FicollTriosil gradients . All donor samples were serotyped for HCMV using an enzyme immunoassay kit detecting total antibody (Abbott Diagnostics) . The samples were coded so that they could be studied further in a blind form . Urine samples used in this study were from three babies with severe post-natal HCMV infection (St . Mary's Hospital Medical School, Paddington) and from 16 coded normal, pregnant or transplant patients typed for cell HCMV by cell culture and/or immediate early antigens (St . George's Hospital Medical School, Tooting). Detection of immediate early antigens was as described by Steel et al .'
Hybridization Human embryonic fibroblasts (HEF) were cultured and maintained as previously reported ." HEF monolayers infected with 52 different clinical isolates of HCMV were harvested at appropriate times postinfection, total DNA was extracted, deproteinised and ethanol precipitated . For controls, DNA was also prepared from uninfected HEF cells, cells infected with adenoviruses (26 serotypes), herpes simplex viruses (two serotypes), simian-like CMV (Colburn stain) E. coli, and salmon sperm . DNA were digested with
either Eco RI or Sma I and the restriction fragments separated on 1 % agarose gels stained with ethidium bromide . DNA fragments were transferred to nylon membranes (Hybond-N, Amersham Int.) and hybridized either with the Eco RI J fragment or its Pst I w fragment subclone from the HCMV strain AD169 after random prime labelling with [a-32P]dCTP . 14
Polymerase chain reaction Digestion of the Eco RI J Fraction using the restriction endonuclease Pst I, yielded fragments which were subcloned into the single stranded phage vector M13mp 18. The DNA sequence of resulting fragments was obtained using the method of Sanger et al." PCR primers (20 bases long) spanning a region of 240 by and an internal 17 base oligonucleotide probe were selected from the region corresponding to the Pst I w fragment16 which mapped to the immediate early gene region 2 of HCMV strain Towne . 17 HCMV1 : 5' CCCGACTTTACCATCCAGTA 3' HCMV2 : 5' AAGACGAAGAGGAACTATCT 3' Probe : 5' GGGTGAAGAAGTCGAAA 3' DNA from blood was prepared for PCR by first isolating the mononuclear cell fraction . Lymphocytes were separated from 5 ml EDTA treated whole blood using ficoll density gradients . After washing in phosphate buffered saline, the lymphocyte pellets were resuspended in 215 pl of water, 25 pl of 10 x PCR buffer containing 0.45% NP40, 0 . 45% Tween 20, and 10 pl of Pronase E (20 mg ml -1 ) . After digestion at 55 ° C for 1-2 h, the DNA was phenol-chloroform extracted and ethanol precipitated . After resuspension, approximately one fifth of the DNA was used for each PCR assay . This was carried out according to the recommended protocol issued with AmpliTaq DNA polymerase (Perkin Elmer Cetus) with a specific annealing temperature of 58 ° C and 30-35 cycles of amplification . Resulting PCR products were analysed by electrophoresis on 2% Nusieve (FMC), 1 % agarose gels stained with ethidium bromide. To test for the specificity of the PCR products obtained, all gels were also Southern blotted and hybridized to an internal oligonucleotide probe, end-labelled with [a- 32 P]dATP . Viral DNA was prepared from urine samples using a standard ultracentrifugation method . Urine samples (6 ml) were clarified by centrifugation at 1000 x g for 10 min, and virus was pelleted by ultracentrifugation at 80,000 g for 1 h . Pellets were resuspended in 200 pl of buffer containing 0. 1 M NaCl, 10mM Tris-HCI and 1mM EDTA (pH 8 . 0). The resuspended DNA was incubated at 37 ° C for 3 h with 0-1 0/6 SDS, 20 pg
Detection of HCMV using PCR
RNAse and 40 µg proteinase K . After phenol-chloroform extraction, the DNA was ethanol precipitated and resuspended in the PCR assay mixture.
marrow transplant patients and neonates who suffered congenital or post-natal infection was used to determine the specificity of the Eco RI J fragment or its Pst I w fragment subclone by Southern blot analysis . With the majority of samples, hybridization to characteristic bands was observed . For example, the Eco RI J fragment hybridises to a 10 . 6 kb Eco RI band and the Pst I w fragment hybridises to a 6 . 5 kb Sma I band (Fig. 1) . Hybridization to alternative restriction fragments of DNA from some isolates denoted the loss or gain of restriction sites, presumably reflecting heterogeneity between HCMV isolates .
RESULTS Hybridization Analysis of DNA from 52 clinical isolates of HCMV recovered from AIDS patients, 13 heart or bone(a)
Ln
N
N
to
53
CO
r•
o
r-
I
c r-
F- 10 . 6 k b
10 . 6 kb
a,
a, c Y
M
3 0 H
v0 Y
Conservation of DNA sequences in different human cytomegalovirus strains. (a) Hybridization of the Eco RI J fragment to HCMV strains digested with the restriction enzyme Eco RI . (b) Hybridization of the Pst I w fragment to HCMV strains digested with the restriction enzyme Sma I . Fig . 1 .
54
P. Stanier et al.
Specificity and sensitivity of PCR
Identification of HCMV in urine samples
A 240 by fragment, defined by a pair of primers specific for the Pst I w fragment, which contains the immediate early gene region, was generated by PCR . PCR analysis was carried out on DNA from 23 different HCMV isolates (Fig . 2) . In all cases, a positive signal was obtained after hybridization with the internal probe . In addition, all non-HCMV control
Consecutive urine samples from two babies suffering post-natally acquired infection and one with a congenital infection, all with infectivity titres in excess of 5 x 10" pfu ml - ' of urine, were highly positive by PCR (data not shown) . HCMV is shed less freely in the urine of organ transplant patients and such samples therefore provided a more appropriate target for PCR analysis . Data were obtained from 16 coded urine samples from transplant patients and from control urine samples from normal individuals and pregnant women . The results are shown in Fig . 3 and the
DNA including DNA from the most related of the human herpesviruses, HHV-6 were negative by PCR . Using the conditions specified, no non-specific bands have been observed with these PCR primers . The sensitivity of the PCR test was established using dilutions of the plasmid containing the Eco RI J fragment. Between 1-10 fg of DNA was routinely amplified so that a band was visible on ethidium stained gels . This corresponded to between 300 and 3,000 copies of the HCMV genome in 10 5 cells .
'0
2
~
N
M
W LO U, = N
OD
r
O F
decoded data presented in Table 1 . When the samples were processed by ultracentrifugation, no falsenegatives were detected although two asymptomatic patients which were negative for cell-culture and IEA determination, were positive by PCR . When the urine samples were used directly after alkali denaturation
J
0 Q
U U
S
M
Y
-
0
Fl-
r- w
(D
W
U)
2
N-
240 by
240 by
Fig. 2 . PCR amplification of 23 different strains of human cytomegalovirus . All of the amplification products were shown to be specific following hybridization with an oligonucleotide probe internal to the PCR primers .
16
Fig . 3.
15
14
13
12
II
10
9
8
7
6
5
4
2
I
B
T
M
PCR amplification of human cytomegalovirus in urine samples . In the example shown, DNA was prepared from the urine samples using the ultracentrifugation procedure . Positive amplification products were confirmed following hybridization with an oligonucleotide probe internal to the PCR primers .
Detection of HCMV using PCR
55
Table 1 . Detection of HCMV in urine samples using PCR . PCR data from 16 samples, prepared either by ultracentrifugation or by direct use were compared to virological analysis by cell culture (CC) or through the detection of immediate early antigens (IEA) . *Samples that were weakly hybridization-positive by hybridization despite bands not being visible on the ethidium stained gel . Days needed to culture virus are detailed where possible as an indication of the amount of virus present in those samples (e.g . low titre =longer time to culture) . ND, no data; +, positive ; -, negative. Sample no . 1 2
3 4 5 6 7 8 9 10 11 12
13 14 15 16
(see Methods)
Details Pregnant BMT (adenovirus) Renal Tr Leukemia Pregnant Normal Heart Tr BMT Normal Infant BMT Infant BMT Pregnant Renal Tr Renal Tr
PCR was less successful which
appeared to be as a result of inhibition of the Taq polymerase .
Identification of HCMV in blood donors DNA was prepared from the mononuclear fraction from PBLs of 105 blood donor samples and examined for HCMV sequences by PCR (Fig . 4). Approximately 50% were positive by PCR with confirmation by hybridization with the internal probe . A summary of results after decoding is presented in Table 2 . On no occasion was a PCR product detected in negative control DNA . After decoding, it was determined that three of the PCR-negative samples were from seropositive individuals and 11 of the PCR-positive samples were from individuals that were seronegative . Three of the latter patients were resampled as described in Stanier et al. 12
Ultra
Direct
+ +• +' + + + + + +
+ + + + -
+
+
CC
IEA
6 days 6 days -
+ ND
+ 8
days
+ 10
days
+ 7
days
ND ND ND ND +
ruled out the possibility that false-positive results were due to contamination . 18 The reproducibility of the data was confirmed when individuals who were seronegative but PCR positive, gave the same results when resampled . Western blotting, regarded as a useful additional antibody test, 19 was weakly positive for two of the three individuals, 12 suggesting that antibody to the virus was present, but at very low levels . Whether these seronegative but HCMV genome positive individuals are infectious was not determined . Thiers et al ." demonstrated that PCRpositive serum samples from patients negative for all serological markers for hepatitis-B virus did carry infectious virus . It must be considered likely therefore that blood from the individuals described here who were PCR positive, although seronegative, must be a potential cause of infection when administered to patients at risk of HCMV infection . It has been reported that antigenic heterogeneity between different HCMV strains can cause falsenegative results with antibody tests 21 and indeed this
DISCUSSION AND CONCLUSIONS
may explain the difference between the PCR and antibody tests . The primers used in this study map to
It is clear from this study that a number of individuals considered to be negative for HCMV, on the basis of their antibody status, were apparently harbouring virus in their peripheral blood mononuclear cells . Careful handling of samples and rigorous controls
a region of the HCMV genome encoding an immediate early gene product, 17 thought to be highly conserved . We have cross-hybridized a probe spanning this gene product in 52 different HCMV isolates and also specifically amplified sequences in this region in 23 different isolates without failure . Recent reports
56
P. Stanier et al .
M +
Antibody typing + + + -
I- 240 by
WE
240 by
Fig. 4. PCR amplification of human cytomegalovirus from peripheral blood mononuclear cells in asymptomatic blood donors . Serological typings determined by a total antibody detection kit are shown . Positive amplification products were confirmed following hybridization with an oligonucleotide probe internal to the PCR primers .
PCR results from peripheral blood mononuclear cells obtained from blood donors compared with their status for HCMV positively determined by antibody detection Table 2 .
PCR
+
42
3
11
49
Ab
suggest serological cross-reactivity22 and homology at the DNA level23 between HCMV and human herpesvirus-6 (HHV-6) . The IE region of the HCMV genome that we amplified did not cross-hybridize with HHV-6 or with any of the other viruses tested . A number of other recent reports have described the PCR for the successful detection of HCMV in
blood", " although in the main, viral DNA was detected only in patients with HCMV viremia . Cassol et al .," however, reported two cases where HCMV was detected in seropositive asymptomatic patients and the findings of Morris et al ." were similar to those reported here . In earlier reports, Shrier et a! .9 were able to detect HCMV RNA transcripts in the blood of seropositive asymptomatic patients and also one of 12 seronegative patients by in situ hybridization whilst Nelson et al." detected HCMV immediate early RNA in approximately 0.01% of PBMCs from 20 asymptomatic HCMV positive individuals using the same technique . Interestingly, PCR has recently been used to detect murine cytomegalovirus DNA in latently infected mice, 29 supporting the view for the sensitivity of this test in any HCMV screening programme . The potential of PCR as a diagnostic tool was further illustrated by the ability to detect DNA in the urine samples of patients who were positive by
Detection of HCMV using PCR culture or by the detection of immediate early antigens . It was possible to detect reproducibly HCMV from a series of urine samples from two babies that were known to be shedding large quantities of the virus (in excess of 5 x 104 pfu ml -1 ) . This was not unexpected and agreed with a larger study reported by Demmler et al.' who detected HCMV sequences in 100% (n=46) of culture-positive urine samples from congenitally infected babies . In their study, they were able to detect HCMV sequences by adding a small quantity of the urine sample direct to the PCR mixture. Analysis of urine samples from transplant patients in the present study indicated that this method was unreliable . Good sensitivity (100%) and specificity (83%) when compared to cell culture and/ or by the detection of immediate early antigens was only obtained when the urine samples were prepared by ultracentrifugation . The unreliability of the direct method is apparently due to the presence of an inhibitor of the Taq polymerase found in the urine, but is probably also related to the lower titres of virus encountered in these types of patients . The recent PCR technology has been used to facilitate the detection of DNA sequences in a variety of diagnostic applications . We have shown that HCMV sequences can be detected both in the mononuclear cells of asymptomatic individuals and in urine samples from at risk individuals using PCR . The first of these findings may prove to be helpful in the screening of blood donations before administering to specific patient groups . The second enables rapid screening of newborns, which may act as an early alert for special development follow-up procedures to be implemented .
ACKNOWLEDGEMENTS HCMV specific DNA probes were kindly provided by Dr Fleckenstein, Erlangen, Gemany ; Epstein-Barr virus DNA was provided by Dr Paul Farrall, Ludwig Institute, St . Mary's Hospital Medical School, London; human herpesvirus type 6 was provided by the late Dr Bob Hoeness, National Institute for Medical Research, London . We are also much in debt to Yvonne Tryhorn (St . George's Hospital Medical School, London) for supplying coded urine samples from both normal and transplant patients .
REFERENCES 1 . Quinnan, G. V., Mazur, H ., Rook, A . H . et al. (1984) . Herpes infections in the acquired immune deficiency syndrome. The Journal of the American Medical Association 252, 72-7 . 2 . Glen, J . (1981) . Cytomegalovirus infections following renal transplantation . Review of Infectious Diseases 3, 1151-78 .
57
3 . Griffiths, P . D . & Baboonian, C . (1984) . A prospective study of primary cytomegalovirus infection during pregnancy (final report) . British Journal of Medicine 91, 307-315 . 4 . Wilhelm, J . A ., Matter, L . & Schopfer, K . (1986) . The risk of transmitting cytomegalovirus to patients receiving blood transfusions . The Journal of Infectious Diseases 154, 169-171 . 5 . Ho, M . (1982) . Cytomegalovirus : Biology and Infection . pp. 95-104 . New York : Plenum Medical . 6 . Yeager, A . S ., Crumet, F . C., Hafleigh, E . B ., Arvin, A . M ., Bradley, J . S . & Prober, C . C . (1981) . Prevention of transfusion-acquired cytomegalovirus infections in newborn infants. Journal of Paediatrics 98, 287--91 . 7 . Steel, H . M., Booth, J . C ., Tryhorn, Y . S . & Stern H . (1988) . A simple immunoalkaline phosphatase method for the rapid diagnosis of cytomegalovirus (CMV) infection . Serodiagnosis and Immunotherapy of Infectious Diseases 2, 193-200 . 8 . Spector, S . A ., Rua, J . A ., Spector, D. L . & McMillan, R . (1984) . Detection of human cytomegalovirus in clinical specimens by DNA-DNA hybridisation . The Journal of Infectious Diseases 150, 121-6 . 9 . Schrier, R . D ., Nelson, J . & Aldstone, M . D . A . (1985) . Detection of human cytomegalovirus in peripheral blood lymphocytes in a natural infection . Science 230, 1048-51 . 10 . Diosi, P ., Moldovan, E . & Tomescu, N . (1969) . Latent cytomegalovirus infection in blood donors . British Medical journal 4, 660-2 . 11 . Saiki, R . K ., Gelfond, D . H ., Stoffel, S . et al. (1988) . Primerdirected enzymatic amplification of DNA with a thermostable DNA polymerase . Science 239, 487-94 . 12 . Stanier, P., Taylor, D . L ., Kitchen, A. D ., Wales, N ., Tryhorn, Y . & Tyms, A . S . (1989) . Persistence of cytomegalovirus in mononuclear cells in peripheral blood from blood donors . British Medical journal 299, 897-8. 13 . Taylor, D . L ., Taylor-Robinson, D ., Jefferys, D . J . & Tyms, A . S . (1988) . Characterisation of cytomegalovirus isolates from patients with AIDS by DNA restriction analysis . Epidemiology and Infection 101, 483-94. 14 . Feinberg, A . P. & Vogelstein, B . (1984) . A technique for radiolabelling DNA fragments to a high specific activity . Analytical Biochemistry 137, 266-7 . 15 . Sanger, F ., Nicklen, S . & Coulsen, A . R . (1977) . DNA sequencing with chain temination inhibitors . Proceedings of the National Academy of Science, USA 74, 5463-6 . 16 . Greenaway, P. J ., Oram, J. D ., Downing, R . G . & Patel, K . (1982) . Human Cytomegalovirus DNA : BamH1, EcoRl, and Pstl restriction endonuclease cleavage maps . Gene 18, 355-60 . 17. Sternberg, R . M., Witte, P . R. & Stinski M . F . (1985) . Multiple spliced and unspliced transcripts from human cytomegalovirus immediate-early region 2 and evidence for a common initiator site within immediateearly gene region 1 . Journal of Virology 56, 665-75 . 18. Kwok, S . & Higuchi, R . (1989) . Avoiding false positives with PCR . Nature 339, 237-8 . 19. Towbin, H ., Staehelin, T . & Gordon, J . (1979) . Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets : Procedure and some applications. Proceedings of the National Academy of Science, USA 76,4350-4 . 20. Thiers, V ., Nakajima, E., Kremsdorf, D . et al. (1988) . Transmission of hepatitis B from hepatitis-B-seronegative subjects . Lancet ii, 1273-6 .
58
P. Stanier et al.
21 . Faix, R . C. (1985) . Cytomegalovirus antigenic heterogeneity can cause false-negative results in indirect hemagglutination and complement fixation antibody assays. Journal of Clinical Microbiology 22, 768-71 . 22 . Larcher, C ., Huemer, H . P ., Margreiter, R . & Dierich, M . P . (1988) . Serological crossreaction of human herpesvirus-6 with cytomegalovirus . Lancet ii, 963-4 . 23 . Efstathiou, S ., Compels, U . A ., Craxton, M. A . et al . (1988). DNA homology between a novel human herpes virus (HHV-6) and human cytomegalovirus . Lancet i, 575-6 . 24 . Shibata, D ., Martin, W . J., Appleman, M. D ., Causey, D . M., Leedom, J. M . & Arnheim, N . (1988) . Detection of cytomegalovirus DNA in peripheral blood of patients infected with human immunodeficiency virus . The Journal of Infectious Diseases 158, 1185-92 . 25 . Jiwa, N . M ., Van Cemert, G . W ., Raap, A . K . et al. (1989) . Rapid detection of human cytomegalovirus DNA in peripheral blood leukocytes of viremic transplant recipients by the polymerase chain reaction . Transplantation 48, 72-6 . 26 . Cassol, S . A ., Poon, M-C ., Pal, R . et al. (1989) . Primer-
mediated enzymatic amplification of cytomegalovirus (CMV) DNA : application to the early diagnosis of CMV infection in marrow transplant recipients . Journal of Clinical Investigation 83, 1109-15 . 27 . Morris, D . J ., Kimpton, C . P . & Corbitt, G . (1989) . Persistance of cytomegalovirus in peripheral blood from blood donors. British Medical journal 2", 1164-5 . 28 . Nelson, J . A ., Cnann Jr ., J . W . & Ghazal, P. (1990) . Regulation and tissue specific expression of human cytomegalovirus. Current Topics in Microbiology and Immunology 154, 75-100 . 29 . Klotman, M . E ., Henry, S . C ., Greene, R . C ., Brazy, P . C ., Klotman, P. E . & Hamilton, J . D . (1990) . Detection of mouse cytomegalovirus nucleic acid in latently infected mice by in vitro enzymatic amplification . The Journal of Infectious Diseases 161, 220-5 . 30 . Demmler, C. J ., Buffone, G . J ., Schimbor, C. M. & May, A . M . (1988) . Detection of cytomegalovirus in urine from newborns by using polymerase chain reaction DNA amplification . The Journal of Infectious Diseases 158, 1177-84 .
