Journal of Clinical LaboratoryAnalysis 6:171-175 (1992)

Laboratory Diagnosis of Parvovirus B19 Infection J. Sanders Sevall, Judith Ritenhous, and J.B. Peter Specialty Laboratories, Inc., Santa Monica, California The sensitivity and application of the polymerase chain reaction (PCR) for the diagnosis of parvovirus B19 (B19) infection was investigated by simultaneously assaying a collection of 279 consecutively receivedSamples for presence of anti-B19 IgM and IgG antibodies by Western blot and for B19 DNA by PCR and dot-blot hybridization (dot-blot); samples were sera from patients with suspected B19 infection. PCR and dot-blot detected B19 DNA in 9% (16/179) and 1% Key words:

polymerase chain reaction, serology by Western blot, PCR, B19

INTRODUCTION Human parvovirus B19 infection, discovered in 1975 ( l ) , has a number of distinct clinical presentations. Generally, human parvovirus infection is a benign self-limiting illness (2-4). However, under certain circumstances the infection can be severe and life threatening. Since the virus infects and can lyse red-cell progenitors interrupting the production of red cells, patients with chronic hemolytic anemias or other red stem cell diseases are susceptible to life-threatening parvovirus infection (5,6). In pregnant women, infection can lead to fetal (7) or congenital (8) infection which sometimes causes severe anemia, congestive heart failure, generalized edema (fetal hydrops), and death (9). Immunocompromised persons, having no immune response to the parvovirus antigens, can experience persistent anemia as a result of chronic infection (10,ll). Clinical suspicion of B 19 infection is usually evaluated by serological assay for B 19-specific antibodies including IgM for recent infection and IgG for past infection (1 2). In aplastic crisis, which is the manifestation of B19 infection of patients with chronic hemolytic anemia, virus can be detected during the first days of illness by DNA detection (13) with dot-blot DNA hybridization assays usin cloned probes of B 19 (14,15). Serological diagnosis can be difficult in immunocompromised individuals with no immune response to parvovirus B 19 antigens. These individuals have persistent levels of DNA in the serum which can be detectable by a sensitive DNA assay ( 16). It is demonstrated that the polymerase chain reaction (PCR) is more sensitive for the detection of human parvovirus B 19 DNA in Ab-positive and Ab-negative (by Western blot[WB]) samples than traditional dot-blot hybridization (dot-blot) for the detection of B 19 DNA. Analysis of IgM - , IgG - , and 0 1992 Wiley-Liss, Inc.

(2/179), respectively of Ab-positive samples (IgM +/IgG - ,IgM + IgG + ,lgM - IgG +),and in 28% (15/54) and 2% (1/54), respectively, of IgM + samples. PCR also detected B19 DNA in 2% (2/100) of IgM - /IgG - samples, both of which had normal total IgG and IgM levels. PCR is of unique value because it permits diagnosis of B19 infection even in the absence of specific acute phase (IgM) and in the presence or absence of convalescentphase (IgG) Ab. o 1992Wiley-Liss, Inc.

IgM&IgG-positive clinical samples indicates that the polymerase chain reaction detects parvovirus DNA in a large number of samples which are negative by traditional dot-blot hybridizations. The current work supports the use of PCR as a more sensitive means for routine diagnosis of B19 infection and its application in patients with immune deficiency is expected to be even more effective. PCR cannot, however, replace immunological tests for laboratory evaluation of human parvovirus B 19 infection.

MATERIALS AND METHODS Materials A collection of 279 consecutively received sera from patients requesting B 19 diagnostic tests were obtained from Specialty Laboratories, Inc (Santa Monica, CA). Two sites in the B19 genome can be consistently amplified by PCR (16). The two extension primers and their B19 genome location (17) are as follows:

172 bp amplification target

5 ’-GTACGCCCATCCCCGGGACCAGTTCAGG3‘ (2060-2087)

5’-CAGGTAAACCCCTTACACCGTCCCACAC3‘ (2230-2203) 104 bp amplification target 5 ’-TGTCAAAAGCATGTGGAGTGAGGG-3’ (3183-3207) 5’-AACACCTTATAATGGTGCTCTGGG-3’ (327 1-3294). Received October 30, 1991; accepted February 18, 1992. Address reprint requests to J.S. Sevall, Specialty Laboratories, Inc., 221 1 Michigan Ave., Santa Monica, CA 90409.

172

Sevall et al.

oxycytidine, thymidine, and deoxyguanosine with a total volume of 50 pl. Thermal cycling was 94°C for 30 seconds and 64°C for 30 seconds for a total of 60 cycles. After amplification, 25 pl of the reaction was used for further analysis by agarose gel electrophoresis in 3% Nusieve, 1 % agarose with 1 direct visualization by ethidium bromide fluorescence or 2 12.5% polyacrylamide gel electrophoresis of PCR products amplified from 32P-end-labeled extension primers and direct 3 detection by autoradiography. A modified parvovirus B 19 DNA oligonucleotide dot-blot 4 hybridization is as follows (16). Two hundred microliters of 5 1 N NaOH was added to aliquotes (23 and 46 p,1) of serum and boiled for 5 minutes (100°C). The samples were dotted 6 ontoZetaprobenylon membraneprewetted with2 x SSC, neu7 tralized with 2 X SSC, and baked under vacuum (1 5 inches of Hg) at 80°C for 1 hour. Prehybridization was at 65°C for 4-24 8 hours in 3 x SSC, 3% SDS, 5 x Denhardt’s (1 %bovine serum albumin, 1% ficoll, and 1% polyprollidine), and 120 p,g single9 stranded salmon sperm DNA/ml in sealed KapaWScotchpak 10 heat-sealable pouches (Minneapolis, MN). One to 10 x lo6 cpm of a randomly labeled isolated B19 DNA fragment of 11 2,500 bp [sequence position 2430-4920 electrophoretically 12 isolated from a pGemini plasmid carrying the entire B19 sequence (17)] was added in fresh hybridization solution with 5% dextran sulfate and incubated 24 hours. After hybridization, filters were washed in 1 X SSC containing 0.1% SDS for 15 minutes at room temperature with two buffer changes. Then A B C D E F G H the filters were washed in 0.1 X SSC containing 0.1% SDS at 65°C for 45 minutes with two buffer changes. The washed Fig. 1. Dot-blot hybridization between B19 DNA standards, positive and filters were blotted dry and exposed to X-Omat S film with negative clinical parvovirus controls, and the 5‘-end-labeled consensus oliintensifying screens for 48 to 72 hours at - 70°C. gonucleotide probe. In column A positions 1-12, a serial twofold dilution IgM or IgG antibodies to parvovirus B 19 were detected by (position 12, 100-fold dilution and position 1, 204.800-fold dilution) of a known DNA positive clinical sample is applied to the nylon membrane. The WB. The antigen (Ag) was derived from Escherichia coli cells quantitative standard B 19 bacterial plasmid was dotted in columns C-H at expressing the entire B19 structural gene encoding the 84 kd 25 ng, 5 ng, 500 pg. 250 pg, 125 pg, and 50 pg, respectively. The standard capsid protein as a 196 kd fusion protein with P-galactosidase. was dotted in TE (lane 12) or in serum (position 11). In columns C-H, lane Partially purified Ag was subjected to PAGE followed by 10, negative clinical serums are dotted on the nylon membrane. The remainelectroelution transfer to nitrocellulose (NC) sheets, residual ing spots (columns B-H, lanes 9 through 1) are available for dilution of clinical semm samples. The sample application and hybridization proceprotein binding sites were blocked with nonfat dry milk dure are in Materials and Methods. (casein), and the NC sheet was divided into lanes. Lanes were incubated with serum at 1:25 dilution for 1 hour, washed, Primers were synthesized based on the gene sequence of B 19 and incubated with alkaline phosphatase-labeled goat antion an Applied Biosystems DNA Synthesizer (Applied Biohuman IgG or IgM Ab. Bound Ab were then visualized with systems, Foster City, CA) by Dr. John Tomich, Department nitroblue tetrazolium in the presence of 5-bromo-4-chloro-3of Medical Genetics, Children’s Hospital (Los Angeles, CA). indolyl phosphate. During assays for specific IgM, Gullsorb Due to amplification inhibitors in the serum, 200 p1 of the (Gull Laboratories, Inc) was routinely added to each patient’s clinical specimens was boiled 10 minutes, cleared by censerum before incubation with Ag to block IgM rheumatoid trifugation, and 5-10 pl of the cleared supernate was amplifactor(l8,19). fied. DNA was amplified by PCR with 50 pmoles of each extension primer. When end-labeled extension primers were RESULTS added, 5 x lo6 cpm of one extension primer labeled at the 5’-end with [y-”PIATP (ICN, Irvine, CA; 3,500 mCi/mM) B19 DNA in serum was detected quantitatively by dot-blot was added to the unlabeled dNTP pool. Reaction conditions (16) and qualitatively by PCR. The hybridization probe is were 25 mM Tris-HC1, pH 8.3, 5 mM MgC12, 50 mM NaCI, isolated from a plasmid (pGemini) containing the entire B 19 and 0.5 mM each of triphosphates of deoxyadenosine, de- DNA sequence and did not cross-hybridize under conditions

ParvovirusB19 Infection

173

A A

B

C

D

E

F

G

H

base pairs:

172 104 -

primers-

B A B C

D E W S

b 8 C

D E W

S A B C

D E W

base pairs:

172104-

primers-

Fig. 2. Polymerase chain reaction products analyzed by gel electrophoresis with direct autoradiography on 12.5% polyacrylarnide gels (A) or with direct ethidium bromide fluorescence on 3% Nusieve, 1% agarose gels (B). Two amplification targets were run simultaneously and immediately applied to the gels for product analysis as in Materials and Methods. Lane S con-

tains a 123 base pair standard ladder (Bethesda Research Labs) and separate DNA targets amplified for 20 (left), 40 (middle), and 60 (left) thermal cycles. Lanes W and H are the amplification of a water control. Lanes A-G contain a ten-fold serial dilution of the linearized pGemini B 19 DNA. Lane A has I ng per assay and lane G 1 .O fg, respectively.

described with the control pGemini plasmid. Sensitivity was determined by probe hybridization with dilutions of linearized plasmid DNA containing the B 19 genome (Fig. 1 , lanes 1 1 and 12, columns C-H) and was monitored with every hybridization by a two-fold dilution of a reference-positive serum (lanes 1-12, column A). The probe was capable of detecting 50 pg B 19 DNA, equivalent to lo6viral particles, whereas 8 negative control serums at 23 or 46 p-1 (example: lane 10, columns C-H) showed no hybridization under the washing conditions described. Because serum in the standard curve reduces the sensitivity of the dot-blot hybridization (lane 1 1, columns B-H), 25 p1 ofnegative-control serum was added to the linearized, viral DNA plasmid prior to sample preparation for dotting the DNA sample. The intensity of the autoradiograph was reduced when compared to the standard DNA

dotted in 10 mM Tris-HC1 (pH = 8.0), 1 mM EDTA. Calibration of B 19-positive clinical serum samples requires the viral DNA standard curve to be dotted with an equivalent amount of serum. The lack of a well-characterized culture system for titered human parvovirus B 19 standards requires the bacterial plasmid containing B 19 DNA in serum to be the calibration standard. It is assumed the purified plasmid B 19 DNA is equivalent to encapsidated viral DNA. The sensitivity of the polymerase chain reaction was quantitated by the minimum amount of target DNA required for amplification of specific target sites. Target sizes of 100-200 bp were reproducibly amplified with thermal cycling between 95°C for 30 seconds and 65°C for 30 seconds. Two targets could be amplified in the same amplification mix. The highest sensitivity is obtained with increasing amplification cycles

-

174

Sevall et al.

M A BC DE

- +

S

M + l 2 3 4 5 6 7

Fig. 3. Polymerase chain amplificationof clinical specimens. Typical amplification in the clinical laboratory has a serial dilution (10 pg, lane A to 0.001 pg, lane E) of the bacterial plasma carrying the entire genome of parvovirus B19 (Materials and Methods) in TE. A heat cleared negative serum specimen (lane -), a heat cleared serum B 19 positive specimen (lane ) and 100

pg of the B 19 bacterial plasma in a heat cleared plasma sample are the negative and positive controls. Lane M is a molecular weight DNA marker (BoehringerMannheim DNA V marker). Lanes 1-7 are clinical serum specimens being tested for the presence of the B 19 virus. Lanes 1-4 are negative and lanes 5-7 are positive serum specimens.

(Fig. 2). The products of the PCR targets were detected by two methods. The most rapid detection method is to visualize the PCR products on 1% agarose, 3% nusieve gels by ethidium bromide staining (Fig. 2B). Maximum sensitivity at 45 cycles is approximately 0.1 pg, or two orders of magnitude more sensitive than dot-blot hybridization. Detection by autoradiography of 32P-end-labeled PCR products after electrophoresis on 12.5% polyacrylamide gels is shown in Figure 2A. The PCR products were labeled by concurrent amplification in the presence of 32P-end-labeled extension primers. After 30 thermal cycles, the sensitivity of both PCR products (172 bp and 104 bp) by autoradiography is 0.1 pg of B19 DNA (Fig. 2, lane E). Compared to direct ethidium bromide fluorescence of PCR products after 45 cycles, autoradiography of PCR products after 30 thermal cycles is equivalent. For rapid diagnosis and high sensitivity, amplification was routinely done for 45 cycles and the products were detected by ethidium bromide stain. A typical clinical amplification is shown in Figure 3 where positive and negative serum controls are amplified alongside the amplified clinical specimens. The sensitivity and application of PCR for the diagnosis of B19 infection was investigated by assaying a collection of clinical samples that were simultaneously studied for the presence of B 19 viral IgM & IgG antibodies and for B 19 DNA by dot-blot hybridization. Eighty-six antibody negative con-

trols were B 19 DNA negative by both dot-blot hybridization and PCR analysis. In the group of 267 consecutively received samples, IgM and/or IgG Ab to B19 were detected in 64% (179/279) by WB using a recombinant antigen (Table 1). PCR and dotblot detected B19 DNA in 9% (16/179) and I % (2/179), respectively, of the Ab-positive samples (Table 1). In the early phase of the Ab response (IgM /IgG - or IgM /IgG ), PCR detected B19 DNA in 28% (1354) of samples vs. 2% (1/54) by dot-blot. In 1% (1/125) of samples with a convalescent Ab pattern (IgM-/IgG+), PCR and dot-blot each detected B 19 DNA. PCR and dot-blot also detected B 19 DNA in 2% (2/100) and 1% (UlOO), respectively, of IgM -/IgG samples, both of which had normal total IgG and IgM levels.

+

+

+

+

DISCUSSION PCR is more sensitive than dot-blot hybridization in detecting parvovirus B19 DNA in sera positive for anti-B19 IgM and/or IgG Ab by Western blot (9% vs. l%), and especially (28% vs. 2%) in early stages of the Ab response (IgM+). Dot-blot hybridization has a sensitivity of 50 pg, whereas PCR can detect 100 fg target with 40 cycles of amplification by ethidium bromide stain or 30 cycles by autoradiography with end-labeled primers. In general, PCR is 100-fold more sen-

TABLE 1. ParvovirusDNA Analysis Antibody status IgM No. of samples DNA PCR positive, % DNA dot blot positive, %

+ /IgG4 I (25) 0 (0)

IgM

+ iIgG-

50 14 (28) I(2)

IgM

-

/IgG-

125 1(1) 1(1)

IgM - iIgG I00 2 (2) 1(1)

Totals 279 18 ( 6 ) 3 (1)

Parvovirus B19 Infection

sitive for DNA detection than standard dot-blot hybridization for B 19 parvovirus. PCR and dot-blot hybridization are comparable (2% vs. 1%) for detecting B19 DNA in sera from individuals with suspected B19 infection during the viremic state prior to the Abresponse (IgM - /IgG - by Western blot). The immunological detection of B 19 infection is not, however, supplanted by the use of the DNA detection assay. However, PCR can be used to confirm suspected B19 infection in early stages (IgM ) in individuals exposed to B 19 who are at risk (e.g., fetus, patients with chronic hemolytic anemias) for serious complications (e.g . ,hydrops fetalis, fetal death, transient aplastic crisis). PCR should be used to c o n f m suspected B 19 infection in immunocompromised individuals who will not develop an Ab response and thus, if untreated are at risk for developing severe, life-threatening anemias due to bone marrow erythropoietic failure.

+

ACKNOWLEDGMENTS We are grateful to Me1 Agopian for clinical specimens, Viphea Mam for technical expertise, and Dr. Karen L. McKeown for helpful discussions and critical review of the manuscript.

REFERENCES I . Cossart YE, Field AM, Cant B, Widdows D: Parvovirus-like particles in human sera. Lancet i:72-73, 1975. 2. Plummer FA, Hammond GW, Forward K: An erythema infectiosumlike illness caused by human parvovirus infection. N Engl J Med 313: 74-79, 1985. 3. Chaorba T, Coccia P, Holman RC: The role of parvovims B 19 in aplastic crisis and erythema infectiosum (fifth disease). J Infect Dis 154: 383-393, 1986. 4. Mayo DR, Vance DW: Parvovirus B 19 as the cause of a syndrome resembling lyme arthritis. N E n g l J M e d 324:419-420, 1991.

175

5. Pattison JR, Jones SE, Hodgson J, Davis LR, Stroud CE, Murtaza L: Parvovirus infections and hypoplastic crises in sickle cell disease. Lancer 1 :664-67 I , 198I . 6. Takahashi M, Moriyama Y, Shibata A, Takai K: Anemia caused by parvovirus in an adult patient with acute lymphoblastic leukemia in complete remission. Eur J Haemotol46:47, 199 1 . 7. Morey AL, Nicolini U, Welch CR, Economides D, Chamberlain PF, Cohen BJ: Parvovirus-B 19 infection and transient fetal hydrops. Lancet 337:496, 1991. 8. Wright IMR, Williams ML, Cohen BJ: Congenital parvovirus infection. Arch Dis Child 66:253-254, 1991. 9. Miller ME, Stiehm ER: Immunology and resistance to infection. In Infectious Diseases of the Fetus and Newborn Infant, 2nd ed, JS Remington, JO Klein, eds. W.B. Saunders, Philadelphia, 1983, pp 27-68. 10. Kurtzman GJ, Cohen B, Meyers P,Amunullah A, Young NS: Persistent B19 parvovirus infection as a cause of severe chronic anaemia in children with acute lymphocytic leukemia. Lancet 2: 1159-1 162, 1988. 11. Gahr M, Pekrun A, Eiffert H: Persistence of parvovirus B19-DNA in blood of a child with severe combined immunodeficiency associated with chronic red cell aplasia. Eur J Pediatr I50:470-472, 1991, 12. Anderson MJ, Higgens PG, Davis LR, Wilman JS, Jones SE, Kidd IM, Pattison JR, Tyrrell DAJ: Experimental parvoviral infection in humans. JInfect Dis 152:257-265, 1985. 13. Anderson MJ: Human parvovirus infections. J Virol Methods 17: 175-181, 1987. 14. Clewley JP: Detection of human parvovirus using a molecularly cloned probe. JMedVirol 15:173-181, 1985. 15. Anderson MJ, Jones SE, Minson AC: Diagnosis of human parvovirus infection by dot-blot hybridization using cloned viral DNA. JMed Virol 15: 163- 172, 1985. 16. Sevall JS: Detection of panovirus B19 by dot-blot and polymerase chain reaction. Mol Cell Probe 4:237-246, 1990. 17. Shade RO, Blundell MC, Cotmore SF, Tattersall P, Astell C: Nucleotide sequence and genome organization of human parvovirus B 19 isolated from the serum of a child during aplastic crisis. J Virol58921-936, 1986. 18. Porter DD, Porter AE, Larsen AE, Hadlow WJ: Immunoenzyme western blotting analysis of antibody specificity in Aleutian disease of mink, a parvovirus infection. J Virol52:745-749, 1984. 19. Ozawa K , Young N: Characterization of capsid and noncapsid proteins of B 19 parvovirus propagated in human erythroid bone marrow cell cultures. J Virol61:2627-2630, 1987.