Quantitative polymerase chain reaction for ... - Clinical Chemistry

Clinical Chemistry 43:10 1843–1849 (1997)

Molecular Pathology and Genetics

Quantitative polymerase chain reaction for human herpesvirus diagnosis and measurement of Epstein–Barr virus burden in posttransplant lymphoproliferative disorder Xin Bai, Gregory Hosler, Beverly Barton Rogers, D. Brian Dawson, and Richard H. Scheuermann*

Human herpesviruses can cause acute diseases such as chicken pox or mononucleosis, but also may reactivate during immunosuppression and result in severe or life-threatening illnesses such as shingles or lymphoproliferative disorders. We report the development and validation of a quantitative PCR method to measure viral burden for all eight human herpesviruses (HSV1, HSV2, VZV, EBV, CMV, HHV6, HHV7, and KSHV) in patients’ samples. The method uses an internal standard that is coamplified with the viral target, allowing quantification of viral genomes in absolute terms (e.g., viral targets/mL of blood) and ruling out false-negative results. We demonstrate that transplant patients with lymphoproliferative disorder carry an EBV viral burden 3 logs higher than nontransplant patients. EBV titers in transplant patients without a lymphoproliferative disorder are between these values. This quantitative PCR method may aid in differentiating clinically significant vs latent viral burden in immunosuppressed patients. Eight different herpesviruses that infect humans have been identified to date— herpes simplex viruses 1 and 2 (HSV1 and HSV2),1 varicella zoster virus (VZV), Epstein– Barr virus (EBV), cytomegalovirus (CMV), human herpesviruses 6 and 7 (HHV6 and HHV7), and Kaposi sarcomaassociated herpesvirus (KSHV)—also known as human

Department of Pathology and Laboratory of Molecular Pathology, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX 75235-9072. * Author for correspondence. Fax 214-648-4070; e-mail scheuerm@ utsw.swmed.edu. 1 Nonstandard abbreviations: HSV, herpes simplex virus; VZV, varicella zoster virus; EBV, Epstein–Barr virus; HHV, human herpesvirus; KSHV, Kaposi sarcoma-associated herpesvirus; ICS, internal calibration standard; PTLD, posttransplant lymphoproliferative disorder. Received May 15, 1997; revision accepted June 23, 1997.

herpesviruses 1– 8 (HHV1– 8), respectively. As a family these viruses typically cause self-limiting illnesses in immunocompetent patients. However, in the immunosuppressed, such as patients on chemotherapy, those with HIV infection, or patients on immunosuppressive therapy after transplant, acute infection or reactivation from a previous infection can be devastating. Whereas primary CMV infection in apparently healthy individuals usually goes unnoticed, CMV-induced pneumonitis in bone marrow transplant patients is often fatal [1]; EBV causes mononucleosis in apparently healthy individuals but can cause a life-threatening lymphoproliferative disorder in the immunosuppressed [2]. PCR amplification has recently been used as a primary modality to diagnose CMV, EBV, and KSHV infection, particularly in the immunosuppressed. Direct viral detection by PCR has advantages over other diagnostic methods. Viral culture for certain herpesviruses, such as EBV and HHV6, is technically difficult and not done in routine diagnostic laboratories. Serologic responses, frequently used to diagnose viral infection in immunocompetent patients, are not reliable in the immunosuppressed. Whereas qualitative PCR can detect viral genomes, it is sometimes difficult to interpret the significance of a positive result for herpesviruses because of the presence of virus in latently infected cells. As an example, 85% of apparently healthy adults have serologic evidence of prior EBV infection and carry ;1 viral genome/105 B cells [3]. For direct detection to be clinically useful, the diagnostic method must be able to differentiate latency from disease. We have developed a PCR technique to detect and quantify viral burden for all eight HHV family members. The technique relies on the coamplification of a DNA internal calibration standard (ICS) included in each PCR reaction with the same primers that recognize the viral

1843

1844

Bai et al.: PCR for HHV diagnosis and EBV burden in PTLD

target. This approach controls for amplification efficiency differences between samples [4, 5], provides quantification, and controls for false-negative results. The assay shows low detection limits, specificity, and precision, essential qualities for use in clinical diagnostic laboratories.

Materials and Methods Viral DNAs. Template DNAs for the different viruses were isolated from either purified virus (HSV2, VZV, EBV, CMV, HHV6, and HHV7 from Advanced Biotechnologies), viral culture supernatant (HSV1), the BC-3 cell line (KSHV [6]), or a patient’s sample (parvovirus). Viral strains used for HSV2, VZV, EBV, CMV, and HHV6 were G, Rod, B95-8, AD169, and Z-29, respectively. Although the analysis of EBV amplification presented uses the Type I or A strain of EBV isolated from B95-8 cells, EBV amplification was also observed with the Jijoye Burkitt lymphoma line that carries the Type II or B strain of EBV [7]. Indeed, the sequences recognized by all four EBVspecific PCR primers are identical between the two strain types [7]. Amplification conditions. All PCR reactions were performed with 2.5 units of AmpliTaq DNA polymerase (Perkin-Elmer) in 13 PCR buffer with 1.5 mmol/L MgCl2, 20 pmol of each primer, and 100 mmol/L each of dATP, dCTP, dGTP, and dUTP in 50 mL final volume. Amplification reactions were initially heated to 95 °C for 2 min and then subjected to cycles of 94 °C for 0.5 min, 60 °C for 0.5 min, and 72 °C for 1.0 min, followed by a final extension at 72 °C for 9.0 min in a GeneAmp 9600 thermocycler (Perkin-Elmer). PCR primers were specifically selected for performance under these amplification conditions. Synthetic ICS construction. The synthetic ICS was constructed essentially as described previously [5]. Briefly, oligonucleotides of 100 residues were designed with 20 complementary residues at the 39 ends. These were annealed and extended in five PCR cycles with Expand Long Template PCR enzyme (Boehringer Mannheim) to generate a 180-bp linked product. In the second step, a new 100-mer oligonucleotide that overlaps the 180-bp linked product by 20 nucleotides was added, together with the initiating 100-mer oligonucleotide. Amplification for an additional five cycles generated a linked product of 260 bp. This process was continued for nine steps to generate the full-length insert. In the final step, amplification for 30 cycles was used to generate sufficient material for cloning. This 812-bp linked sequence was cloned into the polylinker region of the pSP64polyA plasmid. Subclones were sequenced to verify structure. Because a total of 75 PCR cycles were needed to complete the insert synthesis, the six subclones sequenced all contained mutations introduced during PCR linkage. Mutations in one of these subclones were corrected by site-directed mutagenesis with the Transformer Site-directed Mutagenesis

Kit (Clontech), and the resulting plasmid, HHVQ-1, was resequenced to verify structure.

Results and Discussion Primer evaluation. Identification of optimal primers is the most important aspect of the development of a PCR procedure. These primers should combine specificity, low detection limits, and ease of use. Although certain criteria can be used to select primers on the basis of the sequence of the DNA target to be amplified, such as G/C content, length, or Tm, we have found that optimal primers still have to be evaluated empirically. Several primer pairs were selected for each HHV family member, either from the literature or from the DNA sequence. In the latter case, primers chosen were usually 20 nucleotides long, with a G/C content of 40 – 60% evenly distributed over the length of the oligonucleotide, and usually with a G or C 39 terminus, such that a PCR product between 200 and 600 bp would be generated. Primer pairs were first evaluated for amplification specificity with viral DNA mixed with human genomic DNA under routine PCR conditions. At high concentrations of viral DNA and low concentrations of genomic DNA (1 ng), most primer pairs gave single amplification products of the size predicted from the viral sequence (e.g., Fig. 1A, lanes 4, 7, 10, 13, 16, and 19). As the amount of genomic DNA was increased to 100 ng, some primer pairs generated additional bands (primer sets 1A, 1C, 1D, and 1E). This was even more evident when the amount of viral DNA was reduced 10-fold (Fig. 1B). Primer pairs were similarly evaluated for each HHV family member (data not shown) and selected for further analysis if they generated a single amplified band at low viral DNA concentrations in the presence of 100 ng of human genomic DNA. The second criterion for primer selection was low limits of detection. Primer pairs were evaluated for amplification of viral DNA dilutions. For example, EBVspecific primers were used to amplify dilutions of EBV DNA from ;230 000 copies to 230 copies (Fig. 1C). In this experiment, primer sets EA, EC, ED, and EE generated specific bands from as little as 230 EBV copies (lanes 5, 13, 17, and 21), while the other pairs did not. The third criterion for primer selection was lack of cross-reactivity with the other HHV family members. Each primer pair was used in PCR reactions containing each of the HHV targets (Fig. 1D). All primer pairs were found to amplify their specific viral target and did not cross-react with the other family members under the conditions used. ICS design and construction. With two optimal primer pairs identified for each viral target (Table 1), a single ICS was designed to incorporate all of these sequences. In the final design (Fig. 2) the distance between primer pairs is such that the PCR product derived from the ICS differs in size from the product derived from the viral target by at least

Clinical Chemistry 43, No. 10, 1997

1845

Fig. 1. PCR primer evaluation for specificity, limits of detection, and cross-reactivity. PCR reactions were amplified for 32 cycles and analyzed by electrophoresis and ethidium bromide staining. 123-bp ladder DNA was used as a size marker throughout (M). A, amplification with high target concentration. Primers sets (1A–1F) specific for HSV1 were evaluated in PCR reactions containing 0.5 ng of HSV1 DNA and either 1, 10, or 100 ng of total genomic DNA isolated from the K562 erythroleukemia cell line as indicated. B, amplification with low target concentration. Primer sets were evaluated as above except with 0.05 ng of HSV1 DNA. C, primer limits of detection. Tenfold serial dilutions of EBV DNA (from 5.0 ng to 5.0 pg) were amplified with EBV-specific primer sets (EA–EF). The estimated viral titer is indicated. D, primer cross-reactivity. Two pairs of primers specific for each viral target, HSV1, HSV2, VZV, EBV, CMV, HHV6, HHV7, and KSHV (panels a– h) were used in amplification reactions containing each viral target as template, HSV1 (50 ng), HSV2 (3.3 ng), VZV (0.5 ng), EBV (0.5 ng), CMV (10 ng), HHV6 (3.3 ng), HHV7 (300 ng), and KSHV (200 ng) (lanes 1, 2, V, E, C, 6, 7, and K). Primer pairs used are indicated in Table 1. Template DNAs are described in Materials and Methods. Routine reaction conditions were used except for CMV primer pair A in which the annealing temperature was 72 °C. The faint band observed with EBV primer pair A and the VZV template has not been observed in subsequent experiments.

20% so that they can be separated easily by gel electrophoresis (Table 1). However, the products are no more than 40% different in size, to minimize potential differences in amplification efficiency as a result of large differences in product size. A random sequence of 40 nucleotides is located between all primer pairs that can be used as an ICS-specific hybridization probe. The compiled sequence was analyzed to ensure that no secondary structures were inadvertently generated. The sequence was also compared with the GenBank data base to ensure that homologies with known sequences were not inadvertently generated. The synthetic insert region was synthe-

sized with overlapping oligonucleotides and limited PCR as described in Materials and Methods. Validation studies. To use an ICS to generate accurate quantitative data, one requirement has to be fulfilled—the ICS target and the viral target have to be amplified with the same efficiency. Amplification efficiency can be determined by amplifying identical samples for different numbers of PCR cycles [5]. This is based on the equation for exponential growth: Nc 5 Ni~1 1 f! c or logNc 5 logNi 1 c[log~1 1 f)#

1846


Table 1. HHVQ amplication primers characteristics and performance. Virus

HSV1 HSV2 VZV EBV CMV HHV6 HHV7 KSHV Parvo Control

Primer set

Target gene

A B A B A B A B A B A B A B A B A B A B C

TK3 TK3 pol pol Gene 71 Gene 71 EBER EBER gpB gpB IE BamHI frag.

KS330 KS330 VP2 NS1 b-Globin b-Globin RNAPII

5* primer

3* primer

Ref.

Viral product, bp

agcgtcttgtcattggcgaa same gtcccacctcagcgatctcgcct cgtcctggagtttgacagcg cgagtcagcctgacgatcta same cccgcctacacaccaactat tagggtgtaaaacaccgacc tacccctatcgcgtgtgttc same ttctccagatgtgccaggga gatccgacgcctacaaacac cgcatacaccaaccctactg same agccgaaaggattccaccat agcaacacccagctagcagt aatgtacaatcccttatacggatcc aatacactgtggttttatgggccg gaagagccaaggacaggtac same gcatcaaatacccagagacg

ttttctgctccaggcggact acttccgtggcttcttgctg cagcagcgagtcctgcacacaa same tttggagacctagcaagcttcgttc ccgagatggactgagtttgtctg agtctgggaagacaaccaca cagaaagcagagtctgggaa ataggaggcgccacgtattc cctcctataacgcggctgta ggtgctgagtgatcagtttc taccgacatccttgacatattac cacaaaagcgtcgctatcaa gactcattatggggatcgac acatggacagatcgtcaagc tccgtgttgtctacgtccag acctgtgctaaggcctgttaaggc ccattgctggttataaccacagg gtctccttaaacctgtcttg gcctatcagaaacccaagag ccacgtgaaacacaggcttg

13 13 14 14 15 15 16 16 —d —d 17 18 19 19 20 20 21 21 22e 22e 23

342 325 490 445 304 516 210 290 254 320 525 249 300 264 274 269 205 284 323 376 300

S calibrated product, bp

Product size difference, bp (%)a

LODb

CVc

456 416 627 564 415 618 285 397 357 437 686 330 392 352 375 370 269 356 437 477 417

114 (33) 91 (28) 137 (28) 119 (27) 111 (37) 102 (20) 75 (36) 107 (37) 103 (40) 117 (37) 161 (31) 81 (33) 92 (31) 88 (33) 101 (37) 101 (38) 64 (31) 72 (25) 114 (35) 101 (27) 117 (39)

5 5 5 5 5 5 5 5 5 25 5 5 5 5 5 5 50 5 ND ND ND

13.7 20.4 17.4 15.0 18.6 31.3 11.2 13.7 6.2 23.4 26.6 8.2 13.0 16.4 15.8 14.5 7.6 25.8 ND ND ND

a

The difference in size between the PCR products derived from the viral target and the ICS. The Limits of Detection as measured as the minimum number of ICS targets isolated from 1 mL of whole blood giving a detectable band in experiments similar to those described in Fig. 5, except with the primer pair indicated. c For 10 samples amplified for different numbers of PCR cycles in a V/S analysis similar to that described in Fig. 3, except that the indicated viral targets and primer pairs were used. d GenBank accession number A13758, 1994. e GenBank NID accession number S455025, 1994. ND 5 not done. b

where c in the number of amplification cycles, Ni is the initial amount of target, Nc is the amount of product generated after c amplification cycles, and f is the amplification efficiency. In a plot of logNc vs cycle number, the slope of the curve is log(1 1 f). When these curves are generated independently for the ICS product (Sc) and the viral product (Vc), parallel curves (i.e., equal slopes) indicate that each is amplified with the same efficiency (i.e., fs 5 fv). In other words, a graph of logVc/Sc vs c will generate a horizontal line (slope 5 0) if the efficiency of amplification is the same for the two targets.

We have performed this kind of analysis with primer pairs specific for each viral target and template mixtures of ICS and virus DNA. An example of this kind of analysis is presented in Fig. 3. With increasing numbers of amplification cycles, more PCR product is detected from both the ICS and viral targets. Quantification of the amount of product shows an exponential increase in product amount up to 31 amplification cycles under these conditions (Fig. 3A), after which the reaction begins to plateau. Analysis of logVc/Sc vs cycle number reveals a line that closely approximates horizontal for each primer pair (Fig. 3B).

Fig. 2. HHVQ-1 calibrator design and construction: structure of the synthetic HHVQ-1 insert. Primer sequences specific for HSV1 (1A and 1B), HSV2 (2A and 2B), VZV (VA and VB), EBV (EA and EB), CMV (CA and CB), HHV6 (6A and 6B), HHV7 (7A and 7B), KSHV (KA and KB), b-globin (bGA and bGB), and RNA polII (R) were positioned to give products of appropriate sizes. A 40-bp random sequence (AGGCGACGTATCGATCCTGACGGGTCTCGCAGATCAGTAA) was included between the primer blocks as a hybridization probe. An extra 10 bp that contain unique restriction enzyme sites to facilitate cloning was included at each end.


This analysis was performed for each primer pair targeting each virus. In every case the variability in V/S over the range of cycle numbers analyzed was ,27% (see Table 1). The highest variability was observed as the reactions began to enter the plateau phase of the reaction at higher cycle numbers (data not shown). In our experience, the most important variable that impacts amplification efficiency appears to be the sequence of the oligonucleotide primers; the sequence and length of the intervening DNA seem to have little impact on amplification efficiency, at least under the conditions used here. Because two primer pairs have been identified for each viral target and included in the HHVQ-1 ICS, accurate viral quantitation should be similar regardless of which primer pair is used. Mixtures of ICS and appropriate virus were amplified with each of the specific primer pairs and analyzed by fluorescence imaging (Fig. 3C). Although the absolute amount of product generated by each primer pair will differ depending on the efficiency of amplification for that pair, the V/S ratio should be the same. For example, amplification of HHV7 with primer set B (lane 15) generates 4.5 times the amount of product generated with primer set A (lane 14), and yet the V/S ratio differs by a factor of ,0.4. Of all of the viral targets this is the largest V/S difference observed between the two sets of primers. The superiority of using an ICS to perform quantitation over simple PCR is illustrated in precision testing. Intraassay precision was measured by preparing a master amplification mixture and aliquoting this into 10 separate tubes before amplification. Gel analysis of PCR products (Fig. 4, lanes 2–11) shows that although each tube contained an identical mixture, the amount of product generated from both the ICS and viral targets varied considerably. The CV for the viral product was 19%. However, whenever the viral product increased or decreased, the same change was observed in the ICS product. This is reflected by a much smaller CV (6.3%) for the V/S ratio. Interassay precision testing was performed by setting up identical PCR reaction mixtures and amplification for 10 consecutive days, after which all samples were applied to a single gel (Fig. 4, lanes 12–21). The interassay CV for the V/S ratios was 27%. This experiment also illustrates another important point concerning the superiority of this technology over simple PCR. In this experiment the samples loaded in lanes 14 and 17 did not exhibit an amplified product from the viral target. This could result from either the absence of a viral target in the mixture or inefficient amplification of a viral target that was indeed present because of the presence of an inhibitor. However, these samples also did not give an ICS product, which indicates that they were not amplified efficiently. In a clinical setting, the presence of an ICS band would be extremely useful in ruling out false-negative results. Finally, each primer pair was evaluated to determine the limits of detection. ICS DNA was either diluted in

1847

Fig. 3. HHVQ-1 ICS and viral targets are amplified with the same efficiency. A, PCR reactions were set up with EBV primer sets A or B and aliquoted into 10 separate tubes, each containing 250 molecules of HHVQ-1 ICS DNA and 5.4 pg of EBV DNA. One tube was removed from the thermocycler after different numbers of amplification cycles from 25 through 34 and analyzed by gel electrophoresis. After the gels were stained with SyberGreenII (Molecular Probes) they were analyzed on a Fluorimager (Molecular Dynamics). Each band was quantitated with ImageQuant software, and the pixel values were plotted against amplification cycle number in a semilog plot. Products derived from viral target (closed symbols) and ICS target (open symbols) for primer pairs A (circles) and B (squares) are indicated. B, the ratio of the pixel values derived from the viral target (V) and the ICS target (S) is indicated for primer set A (closed triangles) or B (open triangles). C, PCR reactions contained each primer set (A in even- and B in odd-numbered lanes) specific for each virus. Each reaction contained 500 molecules of HHVQ-1 ICS and 10 pg of EBV (lanes 2 and 3), 50 pg of CMV (lanes 4 and 5), 20 pg of VZV (lanes 6 and 7), 80 pg of HHV6 (lanes 8 and 9), 60 pg of HSV2 (lanes 10 and 11), 50 pg of HSV1 (lanes 12 and 13), 30 pg of HHV7 (lanes 14 and 15), 200 pg of KSHV (lanes 16 and 17), or 100 pg of parvovirus (lanes 18 and 19) DNA and was analyzed after amplification for 29 cycles and gel electrophoresis as described above. The V/S ratio for each sample is indicated over the respective lane.

1848


Fig. 4. Inter- and intraassay precision. A PCR mixture was set up and divided into 10 tubes such that each contained 250 molecules of HHVQ-1 ICS DNA and 10 pg of EBV DNA and amplified for 29 cycles (lanes 2–11). Similar single PCR reactions were set up and amplified on each of 10 consecutive days except that 5.0 pg of EBV DNA was used (lanes 12–21). After electrophoresis, gels were stained with SyberGreenII and analyzed by fluorescence imaging. Samples from lanes 14 and 17 were excluded from the CV analysis because no product was observed from either the viral or ICS targets.

water and used for amplification or diluted in whole blood and total DNA was isolated before amplification. Analysis of the PCR products for an EBV primer pair is depicted in Fig. 5. This set of primers was able to detect as few as 5 target molecules, whether diluted in water or isolated in whole blood. The limits of detection for all primer pairs are indicated in Table 1. EBV viral burden analysis in posttransplant lymphoproliferative disorder (PTLD). To demonstrate the utility of quantitative PCR in a clinical setting, we examined DNA isolated from whole blood of 10 pediatric transplant patients that had been diagnosed with PTLD for EBV viral burden by ICS PCR. This was compared with EBV viral burden from 10 nontransplant patients (Fig. 6). In four of the nontransplant patients, low EBV titers above the limits of detection (300 viral targets/mL of blood) could be detected, ranging from 330 to 1950 viral targets/mL of blood. These EBV targets presumably represent latently

infected circulating cells. In contrast, EBV viral burden in PTLD patients ranged from 18 000 to 7 300 000 viral targets/mL of blood. The mean values for the two groups, ,640 and 1 300 000 viral targets/mL of blood, for the reference group and the PTLD group, respectively, were significantly different by Student’s t test (P ,0.05). These results are consistent with previously published studies of viral burden analysis in PTLD patients by viral culture, immunofluorescence, and simple and semiquantitative PCR methods [3, 8 –12]. The increase in EBV viral burden in PTLD patients appears to be the extreme end of the continuum of EBV activation associated with immunosuppression. Thus, in organ transplant patients without PTLD, the average EBV viral burden was found to be 36 000 viral targets/mL of blood, 36-fold lower than that seen in PTLD but 56-fold higher than the nontransplant population (Fig. 6). The difference in mean viral titers between the transplant and the PTLD patients would not be considered significant (P 5 0.052). This is probably related to the large variability in absolute viral titers in both populations coupled with small sample size. In addition, overlap of the absolute values in these two groups suggests that increases in viral burden may be a better predictor of the onset of disease rather than absolute viral burden. In summary, we report the development and validation of a quantitative PCR method to measure the viral burden of all eight HHV family members in patients’ samples. The method involves the use of an ICS that is coamplified with

Fig. 6. EBV viral burden analysis in PTLD. Fig. 5. Limits of detection of viral DNA isolation from whole blood. HHVQ-1 ICS DNA was diluted in distilled deionized H2O such that 1 mL used for amplification contained the number of molecules indicated, from 5000 to 5 (lanes 2–5). In parallel, HHVQ-1 DNA, from 106–103 molecules, was added to 200 mL of whole blood, and DNA was isolated with a QIAamp Blood Kit (Qiagen) such that 1 mL used for amplification contained the number of molecules indicated (lanes 7–10). PCR reactions were amplified for 34 cycles with EBV-specific primer set A.

EBV viral burden was measured in blood from 10 pediatric transplant patients (six liver and four renal) diagnosed with PTLD, from 10 nontransplant patients undergoing routine blood evaluation with normal blood cell counts (reference), and from 14 consecutive pediatric transplant patients. Each PCR reaction contained 50 molecules (for PTLD samples) or 5 molecules (for reference and transplant samples) of HHVQ-1 ICS and dilutions of DNA isolated from whole blood with the IsoQuick Nucleic Acid Extraction Kit (ORCA Research) and was amplified for 34 cycles. Dilutions used for quantitation were chosen where the V/S ratio was closest to 1. The bars indicate the mean values for each group.


the viral target. This allows for the quantification of viral genomes in absolute terms (e.g. viral targets/mL of blood) and can be used to help rule out false-negative results. The technique is rapid and simple and can easily be used in a routine molecular diagnostics laboratory. With this technique, it will now be possible to determine whether changes in EBV viral titer can be used to predict which patients are at risk of developing PTLD and to closely follow their responses to antiviral therapy. Similar investigations concerning the relationship between viral burden of the other herpesviruses and other clinical problems eventually will be possible with this technology.

We thank V. Fleming for the isolation of patient DNA, K. Krisher for the HSV1 culture supernatant, L. Arvanitakis and E. Cesarman for the BC-3 cell line, B. Clinchy for the Jijoye cell line, and C. Cabradilla, K. Reagan, B. Shuman, N. Stoller, J. Chamberlain, L. Picker, and members of the Scheuermann laboratory for helpful discussion. This work was supported in part by a grant from the Texas Higher Education Coordinating Board.

References 1. Couriel D, Canosa J, Engler H, Collins A, Dunbar C, Barrett AJ. Early reactivation of cytomegalovirus and high risk of interstitial pneumonitis following T-depleted BMT for adults with hematological malignancies. Bone Marrow Transplant 1996;18:347–53 2. Craig FE, Gulley MG, Banks PM. Posttransplant lymphoproliferative disorder. Am J Clin Pathol 1993;99:265–76. 3. Riddler SA, Breinig MC, McKnight JLC. Increased levels of circulating Epstein–Barr virus (EBV)-infected lymphocytes and decreased EBV nuclear antigen antibody responses are associated with the development of posttransplant lymphoproliferative disease in solid-organ transplant recipients. Blood 1994;84:972– 84. 4. Wang AM, Doyle MV, Mark DF. Quantitation of mRNA by the polymerase chain reaction. Proc Natl Acad Sci U S A 1989;86: 9717–21. 5. Scheuermann RH, Bauer SR. Polymerase chain reaction-based mRNA quantification using an internal standard: analysis of oncogene expression. Methods Enzymol 1993;218:446 –73. 6. Arvanitakis L, Mesri EA, Nador RG, Said JW, Asch AS, Knowles DM, Cesarman E. Establishment and characterization of a primary effusion (body cavity-based) lymphoma cell line (BC-3) harboring Kaposi sarcoma-associated herpesvirus (KSHV/HHV-8) in the absence of Epstein–Barr virus. Blood 1996;88:2648 –54. 7. Lin J-C, Lin S-C, De BK, Chan W-P, Evatt BL. Precision of genotyping of Epstein–Barr virus by polymerase chain reaction using three gene loci (EBNA-2, EBNA-3C, and EBER): predominance of Type A virus associated with Hodgkin’s disease. Blood 1993;81:3372– 81.

1849

8. Telenti A, Marshall WF, Smith TF. Detection of Epstein–Barr virus by polymerase chain reaction. J Clin Microbiol 1990;28:2187–90. 9. Savoie A, Perpete C, Carpentier L, Joncas J, Alfieri C. Direct correlation between the load of Epstein–Barr virus-infected lymphocytes in the peripheral blood of pediatric transplant patients and risk of lymphoproliferative disease. Blood 1994;83:2715–22. 10. Crompton CH, Cheung RK, Donjon C, Miyazaki I, Feinmesser R, Hebert D, Dosch H-M. Epstein–Barr virus surveillance after renal transplantation. Transplantation 1994;57:1182–9. 11. Martinez OM, Villanueva JC, Lawrence-Miyasaki L, Quinn MB, Cox K, Krams SM. Viral and immunologic aspects of Epstein–Barr virus infection in pediatric liver transplant recipients. Transplantation 1995;59:519 –24. 12. Kenagy DN, Schlesinger Y, Weck K, Ritter JH, Gaudreault-Keener MM, Storch GA. Epstein–Barr virus DNA in peripheral blood leukocytes of patients with posttransplant lymphoproliferative disease. Transplantation 1995;60:547–54. 13. Kit S, Kit M, Zavi H, Trkula D, Otsuka H. Nucleotide sequence of the herpes simplex virus type 2 (HSV-2) thymidine kinase gene and predicted amino acid sequence of thymidine kinase polypeptide and its comparison with the HSV-1 thymidine kinase gene. Biochim Biophys Acta 1983;741:158 –70. 14. Tsurumi T, Maeno K, Nishiyama Y. Nucleotide sequence of the DNA polymerase gene of herpes simplex virus Type 2 and comparison with the Type 1 counterpart. Gene 1987;52:129 –37. 15. Davison AJ, Scott JE. The complete DNA sequence of varicellazoster virus. J Gen Virol 1986;67:1759 – 816. 16. Rosa MD, Gottlieb E, Lerner MR, Steitz JA. Striking similarities are exhibited by two small Epstein–Barr virus-encoded ribonucleic acids and the adenovirus-associated ribonucleic acids vai and vaii. Mol Cell Biol 1981;1:785–96. 17. Yamamoto T, Mukai T, Kondo K, Yamanishi K. Variation of DNA sequence in immediate-early gene of human herpesvirus 6 and variant identification by PCR. J Clin Microbiol 1994;32:473–76. 18. Aubin J-T, Collandre H, Candotti D, Ingrand D, Rouzioux C, Burgard M, et al. Several groups among human herpesvirus 6 strains can be distinguished by Southern blotting and polymerase chain reaction. J Clin Microbiol 1991;29:367–72. 19. Okuno T, Oishi H, Hayashi K, Nonogaki M, Tanaka K, Yamanishi K. Human herpesviruses 6 and 7 in cervixes of pregnant women. J Clin Microbiol 1995;33:1968 –70. 20. Chang Y, Cesarman E, Pessin MS, Lee F, Culpepper J, Knowles DM, Moore PS. Identification of herpesvirus-like DNA sequences in AIDS-associated Kaposi’s sarcoma. Science 1994;266: 1865–9. 21. Shade RO, Blundell MC, Cotmore SF, Tattersall P, Astell CR. Nucleotide sequence and genome organization of human parvovirus B19 isolated from the serum of a child during aplastic crisis. J Virol 1986;58:921–36. 22. Liu J, Johnson RM, Traweek ST. Rearrangement of the bcl-2 gene in follicular lymphoma: detection by PCR in both fresh and fixed tissue samples. Diagn Mol Pathol 1993;2:241–7. 23. Wintzerith M, Acker J, Vicaire S, Vigneron M, Kedinger C. Complete sequence of the human RNA polymerase II largest subunit. Nucleic Acids Res 1992;20:910.