INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Human herpesviruses, herpes simplex virus type 1 (HSV-1) and type 2 (HSV-2), varicella-zoster virus (VZV), cytomegalovirus (CMV), and Epstein-Barr virus (EBV) are ubiquitous agents causing a wide range of acute central nervous system (CNS) infections in humans. Encephalitis caused by HSV-1 or VZV has been described extensively, and most patients with HSV-1 or VZV encephalitis lack cutaneous vesicles at the onset of neurological disease (8, 16). Both primary and recurrent herpesvirus infections may lead to CNS infection and disease. Among these viruses, HSV accounts for approximately 2 to 19% of all cases of encephalitis and 20 to 75% of all cases of necrotizing encephalitis (34, 36). Herpes simplex encephalitis has a yearly incidence of one to four per million people (37) and may be associated with significant morbidity and mortality without treatment (11). Neonatal HSV encephalitis is a devastating disease most commonly occurring as a consequence of perinatal transmission of HSV-2. Life-threatening herpesvirus infections of the brain can occur throughout life. In immunocompetent patients older than 3 months, meningitis, meningoencephalitis, and myelitis are often associated with HSV-2 (2). Dual or triple infections (coinfections) due to concomitant infection of the CNS by two or three herpesviruses determined by PCR have been reported (23, 36). A clinical diagnosis of herpes simplex encephalitis is unreliable, as none of the presenting symptoms (fever, focal seizures, hemiparesis, and altered level of consciousness) is pathognomonic of herpes simplex encephalitis.

Early diagnosis of HSV encephalitis is important to start adequate treatment and to exclude other diseases that have a similar clinical presentation. The convincing method for diagnosis of HSV encephalitis is the isolation of virus from brain tissue obtained by biopsy (25), but in 40 to 60% of the cases, brain biopsy is an unnecessary procedure, as these patients were finally proved to be negative for HSV (9). Conventional laboratory diagnosis of CNS infections caused by human herpesviruses has not been productive. HSV is rarely recovered in cell cultures from cerebrospinal fluid (CSF). A tentative diagnosis of HSV encephalitis can be performed by the demonstration of intrathecally produced anti-HSV antibodies as expressed in an increased ratio of HSV CSF and serum antibodies. By using immunoprecipitation, radioimmunoassay, enzyme-linked immunosorbent assay (ELISA), and immunoblotting, several authors suggested that an increased CSF-to-serum antibody ratio is a reliable measure for the diagnosis of HSV encephalitis (5, 10, 15, 20, 26, 31). By contrast, other authors hypothesized that prolonged antigen stimulation is present in the CNS after acute HSV encephalitis and that serological measurements combined with immunoglobulin and albumin determinations provide only a tentative, not definite, diagnosis (18).

The viral antigen detection in CSF samples using indirect ELISA was proved to be not efficient enough, since the sensitivity of this assay was below that of serological methods applied to CSF (3).

The use of PCR for the diagnosis of CNS disease has been well evaluated for HSV encephalitis (11, 24, 33), and PCR is now the method of choice for herpesvirus detection. It was shown that a clinically diagnosed viral infection of the CNS was 88 times more likely to occur in a patient with a positive PCR result than in one with a negative result (13).

Usually, a series of independent, targeted PCRs in which each PCR detects a single virus is performed. In the present report, we describe a multiplex PCR assay for single-tube amplification of HSV-1, HSV-2, VZV, CMV, and EBV. Failure in DNA extraction procedure used or the presence of DNA amplification inhibitors are revealed by the inclusion of -tubulin primers in the multiplex PCR assay.

In the present methods development study, we report the results of a preliminary study on a multiplex, one-round PCR assay for single-tube amplification that was performed to increase the sensitivity of this assay, compared to already published multiplex-nested PCRs (6, 30), and to evaluate the feasibility of this assay on a limited number of clinical samples, raising confidence in negative results.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Specimens. A total of 86 CSF from patients with clinical suspicion of HSV encephalitis, meningitis, or meningoencephalitis was included in this preliminary study. At least one CSF sample, and for some patients two consecutive CSF samples, were taken within the first 10 days following onset of neurological symptoms. None of the patients included in the study had a previous human immunodeficiency virus (HIV) infection. Twelve CSF samples collected from patients with unrelated diseases (10 had noninfectious CNS disease and 2 had autoimmune diseases) were used as negative controls. CSF samples were stored at 20°C until DNA extraction and multiplex PCR were performed.

Viruses and controls. HSV-1 strain F, HSV-2 strain G, VZV strain Ellen, CMV strain AD169, and EBV strain P-3 (American Type Culture Collection, Manassas, Va.) were used as positive controls. All but EBV were propagated in tubes in MRC-5 cells (American Type Culture Collection) grown in minimal essential medium Dulbecco's modified enriched medium (MEM-D) medium supplemented with 2% fetal calf serum and examined daily until observation of typical cytopathic effect (CPE). When 70 to 80% CPE was observed, tubes were frozen at 70°C. EBV strain P-3 was subjected to DNA extraction directly from the stock solution.

DNA extraction. HSV-1, HSV-2, VZV, CMV, and EBV DNA was recovered from CSF according to the method described by Casas et al. (4), with minor modifications. Briefly, 100 µl of CSF was incubated for 10 min at room temperature with 400 µl of lysis buffer (4 M GuSCN, 0.5% N-lauroyl sarcosine, 1 mM dithiothreitol, 25 mM sodium citrate, and 40 µg of glycogen/tube). Then, 500 µl of cold isopropyl alcohol were added. The tubes were vortexed, allowed to stand for 15 min in ice, and then centrifuged for 10 min at 14,000 × g at 4°C. The isopropyl alcohol was removed, and the pellet was washed by addition of 1 ml of 70% ethanol. The samples were centrifuged again as above. The ethanol was removed, and the pellet was dried and dissolved in 25 µl of sterile, double-distilled water.

Sensitivity and specificity. The sensitivity of the PCR was determined using serial dilutions of titrated suspensions of either HSV-1 or HSV-2 and by testing serial dilutions of VZV-, CMV-, and EBV-specific PCR products quantitated by GelPro Analyzer version 3.0 software (Media Cybernetics, Silver Spring, Md.). All dilutions were prepared in pooled lymphocytic CSF and were shown to be negative for the DNA of all five herpesviruses by PCR. These CSF samples and DNA extract of uninfected MRC-5 cells were used as negative controls. A similar approach for the quantitation of PCR sensitivity has already been reported by Johnson et al. (14).

Primers. The primer pairs could enable identical amplification efficiencies for their target sequence in a multiplex reaction. Primers with nearly identical optimum annealing temperatures should work under fairly similar conditions if they anneal with single-copy sequences. Primer concentration is a critical parameter for successful multiplex PCR. If all primers in a reaction mixture anneal with equal efficiencies, they can be used at the same concentration (21). The PCR primers were synthesized with an Applied Biosystems oligonucleotide synthesizer, and their sequences are reported in Table 1.

View this table: [in this window] [in a new window]   TABLE 1.   Primer sequencesdigestion of PCR amplicons as predicted by DNA sequence analysis

DNA amplification-multiplex PCR. To optimize the multiplex PCR, a series of titrations of primer concentrations and deoxynucleotide triphosphate (dNTP levels were performed. Primer concentrations of 10, 25, 50, and 100 pmol from each primer pair were titrated simultaneously with dNTP (0.1, 0.2, and 0.3 mM concentrations of each of the dNTPs).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Optimization of multiplex PCR. The selected concentration for all 12 primers was 10 pmol, as all the primers in the reaction mixture annealed with equal efficiencies when the concentration of dNTPs was 0.8 mM.

View larger version (138K): [in this window] [in a new window]   FIG. 1.   Specificity of the multiplex PCR to detect HSV-1, EBV, HSV-2, CMV, and VZV. Detection and typing of clinical isolates by multiplex PCR. Amplification was performed with 0.8 mM dNTPs. All primer pairs were used at the concentration of 10 pmol. Amplification conditions were as follows: 40 cycles of 30 s at 94°C, 40 s at 60°C, and 50 s at 72°C. The amplified products were subjected to a 3.5% agarose gel electrophoresis containing 1 µg of ethidium bromide/ml in Tris-borate-EDTA buffer. Lane M, molecular weight markers corresponding to a HinfI digest of pBR322 DNA (Minotech, Crete, Greece). Shown in lanes 1 to 5 are amplifications performed on 10 log PCRD50 of HSV-1/ml, 160 ag of EBV, 7 log PCRD50 of HSV-2/ml, 18 ag of CMV, and 14 ag of VZV in the presence of 1 µg DNA of pooled CSF samples. Lanes 1, 2, 3, 4, and 5, amplifications with primer pairs H1P32/H1M32, EP5/EM3, H2M40/H2P4, CP15/CM3, and VP22/VM20, respectively; lane 6, positive multiplex PCR amplification with the five sets of primers of the above five herpesviruses (product size marker used in the screening of the clinical specimens). Shown in lanes 7 to 11 are amplifications performed with -tubulin primers and the above set of five primers on clinical samples. Lane 7, negative CSF sample; lanes 8, 9, 10, and 11, CSF samples containing HSV-1, HSV-2, CMV, and VZV, respectively. The predicted size of each amplimer is shown on the right.

Sensitivity and specificity of multiplex PCR assay. The sensitivity of the multiplex PCR was determined using serial dilutions of infected tissue culture supernatants of predetermined 50% tissue culture infective doses (TCID₅₀s) for HSV-1 and HSV-2. The reciprocal of the highest dilution positive by PCR adjusted to concentration per milliliter after DNA extraction was defined as PCRD₅₀/ml. The sensitivity of the multiplex PCR for VZV, EBV, and CMV was determined by testing serial dilutions of quantitated PCR products. All dilutions were prepared in lymphocytic CSF shown to be negative for all five herpesviruses (HSV-1, HSV-2, VZV, CMV, and EBV) by PCR. Following agarose gel electrophoresis, the minimal amounts of HSV-1 and HSV-2 detectable were 10 and 7 log PCRD₅₀/ml, respectively. For VZV, EBV, and CMV, genomic DNA was detected at 14, 160, and 18 ag, respectively (Table 2). The corresponding sensitivity in DNA copies was 48, 66, and 840 genome copies for VZV, CMV and EBV respectively. The sensitivity was also assayed by 10-fold dilutions of active virus particles of HSV-1 (stain McIntyre), HSV-2 (strain G), VZV (strain Rod), CMV (strain AD169), and EBV (strain B95-8) counted by negative stain electron microscopy (Advanced Biotechnologies Inc.), prior to DNA extraction. The limiting sensitivities measured were 52, 213, 90, 260, and 600 virus particles for HSV-1, HSV-2, VZV, CMV, and EBV, respectively.

Potential presence of inhibiting factors in CSF. The diagnostic value of CSF PCR may be compromised by the presence of inhibitors, and PCR may potentially lead to false-negative results. Elevated CSF protein levels that are observed in acute encephalitis may contribute to this inhibitory effect (28). In the present assay, the inclusion of -tubulin primers permitted the detection of occasional false-negative results. Thus, failure in the DNA extraction procedure or the presence of DNA amplification inhibitors is easily detected (22).

Multiplex PCR analysis of CSF samples. The presence of the HSV-1, HSV-2, VZV, CMV, and EBV genomes was investigated by multiplex PCR in a total of 86 CSF specimens. These samples were sent to our laboratory for routine diagnostic purposes.

View larger version (156K): [in this window] [in a new window]   FIG. 2.   Restriction enzyme digestion pattern for HSV-1, EBV, HSV-2, CMV, and VZV by HpaII (lanes 1 to 5) and Tru9I (lanes 6 to 10). Lane M1, molecular weight marker, corresponding to a HinfI digest of pBR322 DNA (Minotech, Crete, Greece); lane M2, molecular weight marker, 50-bp DNA ladder (MBI, Vilnius, Lithuania). Fragments smaller than 30 bp are not visible on the ethidium bromide-stained agarose gel. Lane 1, HSV-1 digest (58- and 62-bp fragments; the 27-bp fragment is not visible); lane 2, EBV, no digestion; lane 3, HSV-2 digest (61-, 82-, and 84-bp fragments); lane 4, CMV digest (235-bp fragment; the 21-bp fragment is not visible); lane 5, VZV, no digestion; lane 6, HSV-1, no digestion; lane 7, EBV digest (49- and 133-bp fragments); lane 8, HSV-2, no digestion; lane 9, CMV digest (60- and 171-bp fragments; the 25-bp fragment is not visible); lane 10, VZV digest (75-, 78-, and 122-bp fragments).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The detection of CNS infections by PCR has provided a new tool of laboratory diagnosis compared to conventional techniques of cell culture and serology. Target genes of organisms responsible for CNS infections are generally present early in the acute phase of the disease; in contrast, CSF samples from patients without an infectious etiology do not contain these nucleic acid sequences (1). PCR was evaluated in a large series of patients with biopsy-proven herpes simplex encephalitis, and 98% sensitivity was achieved using a single-round PCR protocol (19). Thus, the relevance of PCR for the early diagnosis of herpes simplex encephalitis was established when a therapeutic decision is urgent.

The present multiplex PCR assay was developed to allow simultaneous detection and typing of HSV-1, HSV-2, VZV, CMV, and EBV DNA within the same tube. Particular emphasis was given to how the method was optimized and carried out. In addition, as the detection of these viruses is performed with a single nucleic acid extraction and amplification conditions, it is convenient and cost-effective for use on a routine basis. The limited clinical data in this study are used to show that the assay may be implemented as a routine diagnostic test. The results of this preliminary study should prompt a more exhaustive analysis of the clinical value of the present multiplex PCR assay.

A sensitive assay was established by careful choice of primers and adjustment of dNTP and primer concentrations. The rather high number of 40 cycles was chosen to drive PCR into the linear phase of amplification, when the amount of primers or enzyme became limiting. In each PCR amplification, 2.5 U of cloned Pfu DNA polymerase were used. The advantage was that it provided adequate enzymatic activity when one DNA amplification was less efficient than another in a simultaneous reaction. The annealing temperature was chosen to be as high as possible (60°C), taking care not to reduce the sensitivity of the assay.

The sensitivity of detection of each of the viral nucleic acids was equivalent by each single primer pair and by multiplex primer pair reactions. The multiplex PCR has a high molecular sensitivity for all five herpesviruses detected in the present assay. For HSV-1 and HSV-2, log TCID₅₀/ml values were 8 and 10, respectively, and log PCRD₅₀/ml values were 5.5 and 7, respectively. Thus, the present multiplex PCR assay has a higher molecular sensitivity than those reported by Cassinoti et al. (6) (0.003 and 0.3 TCID₅₀ for HSV-1 and HSV-2, respectively), while the sensitivity is 30-fold greater than those reported by Read and Kurtz (30) for HSV-1 (TCID₅₀/PCRD₅₀ of 0.3 instead of 0.01; Table 2). A potential difficulty in making such a comparison is the lack of well-validated standard materials for such comparisons. TCID₅₀ is a semiquantitative estimate of the infectivity. By titration of this material in PCR (PCRD₅₀), a detection limit of 0.01 and 0.03 TCID₅₀/PCRD₅₀ for HSV-1 and HSV-2, respectively, was achieved. The problem with such an approach is that the infectivity of a virus preparation may not be an accurate indication of the number of viral genomes in a preparation (i.e., there may be significant numbers of non-infectious particles). To properly compare the assays, the same batch of material would need to be tested in all three assays. For VZV, CMV, and EBV, the sensitivity of the assay was of 14, 18, and 160 ag, or 48, 66, and 840 genome copies, respectively. In addition, the present assay has the considerable advantage that it is a single-round multiplex PCR, while Cassinoti et al. (6) and Read and Kurtz (30) performed two-round PCRs of 35 and 15 cycles for the first round and of 33 cycles for each round, respectively, exposing the assay to potential carryover contamination.

Quality controls are of crucial importance to monitor the potential occurrence of false-positive as well as false-negative PCR results. Potential contaminations leading to false-positive results were monitored by submitting negative controls to the whole PCR process, including DNA extraction. Occasional negative PCR results were observed for patients with proven herpes simplex encephalitis (19, 27) raise an important problem; the competence of each sample for PCR amplification. It is well known that CSF may contain inhibitors that can partially or completely block DNA polymerase activity and cause false negative results. Due to the clinical and therapeutic implications of a false-negative PCR result, identification of inhibited PCRs is a priority. Therefore, when PCR is employed for diagnostic purposes, it is imperative to adopt adequate controls for assessing sample suitability for PCR. In the present multiplex PCR assay, occasional false-negative results for herpesvirus DNA amplification could lead to unidentifiable false-negative results if -tubulin primers were not included in the assay. Thus, failure of DNA extraction or failure to remove any inhibitors of DNA amplification may be avoided by the present assay. In routine practice, failure of DNA extraction or the presence of inhibitors was subsequently assessed by testing a distinct aliquot of each of the six CSF samples diluted 1/5 in distilled water prior to DNA extraction.

In the present study, we investigated a PCR-based DNA amplification technique for detecting herpesvirus DNA in the CSF samples of patients manifesting symptoms of viral encephalitis. The CSF samples evaluated in our study were from patient populations with CNS disease not associated with HIV infection. Coinfection was not detected. Detection of more than one virus from any clinical specimen is uncommon. Moreover, documented reports of viral coinfections in CNS diseases in immunocompetent patients determined by PCR have also been uncommon (29). The diagnostic procedure which was elaborated during the present study allowed the rapid detection of herpesvirus DNA in patients with the symptoms of encephalitis or meningitis. All patients were subsequently treated with acyclovir for 14 days (30 mg/kg of body weight/day, given intravenously). Four of the nine herpes virus-positive patients (Table 3) were clinically diagnosed with meningitis (detection of VZV DNA in CSF), and they had a full recovery. Three patients presented the clinical symptoms of encephalitis (detection of HSV-1 DNA in CSF); two of them recovered and the third one had neurological sequelae during a 6-month follow-up. The two neonates (detection of HSV-2 or CMV DNA in respective CSF samples) recovered with mild to moderate sequelae. Conventional virologic laboratory diagnosis (cell culture) of these CNS infections was not productive. All CSF samples were inoculated into MRC-5 and Vero cells and were followed for at least 8 days. Early CSF samples showed mild to moderate pleocytosis ranging from 20 to 400 cells/mm³ and consisted mainly of mononuclear cells. Protein content was either normal (<0.5g/liter), or increased up to 2 g/liter. In ELISAs, all nine patients showed serological evidence of the corresponding CNS infection by one of the VZV, HSV-1, HSV-2, or CMV viruses (high titers of anti-immunoglobulin G [IgG] antibodies to VZV, or HSV-1 and detection of anti-IgM antibodies to HSV-2, or CMV). The multiplex PCR assay described appears to be reliable for use with clinical samples. It has been shown to be appropriate for the early and type-specific detection of HSV-1, HSV-2, VZV, CMV, and EBV genomes in patients with suspected CNS infection. This assay has many advantages over other types of laboratory tests including single target PCR or multiplex nested PCR as has already been discussed. It detects simultaneously five different herpesviruses and sample suitability for PCR in a single amplification round of 40 cycles. There are clear advantages to a method that involves only one reaction: less sample material, reagents, and time are required.

In conclusion, the multiplex PCR assay presented in this study can provide an early, rapid, reliable noninvasive diagnostic tool allowing antiviral therapy to be initiated on the grounds of a specific viral diagnosis. The clinical features of neurological disease due to non-polio enteroviruses can overlap those caused by herpesviruses. There may be unnecessary hospitalization and anti-herpetic therapy may, therefore, be initiated for patients with mild enterovirus infection (32). However, additional studies are required in order to evaluate the present assay with a greater number of CSF specimens in the clinical context by testing it against well-defined specimens that have been thoroughly characterized by other methods and, finally, in order to assess its clinical significance and its utility in monitoring the efficacy of treatment with antiviral drugs. Such a study should be the subject of a subsequent submission.