- Email: [email protected]
A simple and sensitive method for detecting adenovirus in serum and urine
Journal of Virological Methods 71 (1998) 51 – 56
A simple and sensitive method for detecting adenovirus in serum and urine Ying C. Henderson a, Ta-Jen Liu a,b, Gary L. Clayman a,b,* a
Department of Head and Neck Surgery, Box 69, The Uni6ersity of Texas M.D. Anderson Cancer Center, 1515 Holcombe Bl6d., Houston, TX 77030, USA b Department of Tumor Biology, The Uni6ersity of Texas M.D. Anderson Cancer Center, Houston, TX 77030, USA Accepted 30 October 1997
Abstract A method to detect shed virus in patients’ serum and urine was developed following Ad5CMV-p53 gene transfer via direct tumor injections. The procedure differs from those reported previously in that it first uses polyethylene glycol to precipitate adenoviral particles from patient serum or urine. Adenoviral DNA is then extracted following proteinase K digestion. Finally, polymerase chain reaction (PCR) amplification followed by Southern blot transfer are employed to enhance the limit of detection to less than ten viral particles. The detection limit in a 0.25-ml sample is five viral particles in serum and one viral particle in urine. This procedure is sensitive, reproducible, and can be completed in less than 2 days. © 1998 Elsevier Science B.V. All rights reserved. Keywords: Adenovirus; Serum; Urine; Polyethylene glycol
1. Introduction The adenoviral vector has emerged recently as a front runner in the field of gene therapy (Clinical protocols, 1996). Although adenoviral vectors
* Corresponding author. Tel.: + 1 713 7926920; fax: +1 713 7944662.
provide high transduction efficiencies in a wide range of host cells, their usefulness in human trials has yet to be determined, partly because of insufficient data to support their safety in humans. We describe a sensitive method to detect shed adenovirus in serum and urine from head and neck cancer patients following Ad5CMV-p53 (wild-type p53 in replication defective adenovirus) gene transfer.
0166-0934/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII S 0 1 6 6 - 0 9 3 4 ( 9 7 ) 0 0 1 8 9 - 4
52
Y.C. Henderson et al. / Journal of Virological Methods 71 (1998) 51–56
2. Materials and methods
2.1. Viral DNA preparation Ad5CMV-p53 adenoviral DNA (Zhang et al., 1993) was prepared by a method modified from Cunningham et al. (1995). Briefly, distilled water was added to 0.5-ml aliquots of serum or urine to make a final volume of 1 ml, followed by precipitation with 0.5 ml of 30% polyethylene glycol 8000 (PEG). Following the PEG precipitation, the pellet was resuspended in 0.3 ml of 0.5% NaDodSO4 containing proteinase K (60 mg) and incubated at 50°C for 2–16 h (Norder et al., 1990). Samples were extracted with phenol, and viral DNA precipitated with ethanol in the presence of 10 mg glycogen (Cunningham et al., 1995). Precipitated DNA was recovered by centrifugation at 14 000× g for 10 min at 4°C, resuspended in 0.3 ml of distilled water, and reprecipitated with ethanol. The DNA pellet was rinsed with 70% ethanol, vacuum-dried, and dissolved in 10 ml of distiled water. Only 5 ml of prepared DNA was used in a single polymerase chain reaction (PCR).
2.2. PCR amplification PCR was carried out in a 20-ml volume containing 2 mM MgCl2, 50 mM KCl, 0.1% Triton X-100, 200 mM each of deoxyribonucleotide triphosphates (dNTPs), 10 mM Tris – HCl (pH 9.0), 5 pM each of the primers, and 1.7 U of Taq DNA polymerase (Promega, Madison, WI). The reactions were carried out at 94°C for 30 s, 58°C for 30 s, and 72°C for 60 s for 35 cycles, followed by a 10-min extension at 72°C. PCR primers were selected from the sequence of Ad5CMV-p53 (Zhang et al., 1993) with the sense primer located at the 3% end of the p53 cDNA (5%-GCCTGTCCTGGGAGAGACCG-3%), and the antisense primer located in the adenovirus type 5 genome (5%-CCCTTAAGCCACGCCCACAC-3%). The PCR product (an 838-bp fragment) was separated on 1% agarose gel and transferred to a nylon membrane (Hybond-N+, Amersham, Arlington Heights, IL) by capillary absorption (Maniatis et al., 1989). The same PCR product was subcloned into pCR-Script vector (Stratagene, La Jolla, CA).
The sequence of the insert was confirmed by sequencing and the gel-purified insert used as a probe (probe A) to detect PCR products after labeling with [32P]dCTP using the Megaprime DNA labeling system (Amersham). The specific activity of probe A was approximately 109 dpm/mg.
2.3. Southern blot analysis Southern blot analysis was used to verify the specificity of PCR product. The membrane was prehybridized for 15 min at 65°C in Rapid-hyb buffer (Amersham) and hybridized in the same buffer containing 32P-labelled probe A for 1–2 h. The membrane was washed in 0.1× SSC and 0.1% NaDodSO4 at room temperature twice, and followed by two washes at 65°C (15 min per wash). The washed membrane was exposed to X-ray film at − 70°C for 1–16 h with intensifying screen.
2.4. Titer of the 6irus The titer of CsCl-purified viral particles was determined by plaque assay (Graham and Van Der Eb, 1973). An OD of one at 260 nm is equal to 1012 adenovirus particles/ml. All oligonucleotides were synthesized by Genosys Biotechnologies (Woodlands, TX).
3. Results
3.1. Testing different methods of 6iral DNA preparation Detection of cytomegalovirus and hepatitis B viral DNA in serum by the PCR has been reported (Norder et al., 1990; Ishigaki et al., 1991; Brytting et al., 1992; Cunningham et al., 1995; Tokimatsu et al., 1995). According to those reports, 10–400 ml of serum was used to extract viral DNA and about 10–25% of the extracted DNA was used in PCR. To establish a method for adenovirus DNA detection and to increase the sensitivity to detect adenoviral DNA, which is limited by the small amount of serum used in these assays, several existing methods were tested
Y.C. Henderson et al. / Journal of Virological Methods 71 (1998) 51–56
for cytomegaloviral DNA preparation. Adenovirus (104 Ad5CMV-p53 particles) were spiked into 0.5 or 1 ml of serum. Two PEG precipitations were carried out to ensure that all of the virus particles had been precipitated. Pellets after the first and second PEG precipitations were treated with NaDodSO4 alone or NaDodSO4 containing proteinase K. The samples were extracted with phenol and precipitated with ethanol. We found that proteinase K was required for detection of viral DNA. NaDodSO4 alone was able to release adenoviral DNA, inhibitors were present and no specific viral DNA was detected after PCR (data not shown, Y. Henderson). Precipitation with PEG once was sufficient to concentrate all of the virus. No viral DNA was detected in the supernatant following the first PEG precipitation. Viral DNA could be detected in 0.5 ml of serum but not in 1 ml of serum. However, if 1:10 dilution was made from the 1-ml serum DNA sample, a viral DNA PCR product could be detected in Southern Blot analysis. This observation is due most likely to the presence of PCR inhibitor(s) present in the 1-ml serum sample. The PCR inhibitor(s) could also be removed by repeated extraction with phenol or phenol/chloroform. Repeated extractions, however, result in heavy loss of DNA, thereby providing poor detection sensitivities of viral DNA (data not shown, Y. Henderson). Therefore, serum volumes not exceeding 0.5 ml were used for further analysis.
53
ern blot analysis, including the serum-only sample lacking added virus (lane 3). This band was not detected in ethidium bromide-stained agarose gels.
3.3. Comparison of the optimized 6iral DNA preparation with commercially a6ailable kits The optimized method of adenoviral DNA detection was compared with those of commercially available kits. The QIAamp blood kit (Qiagen) and QIAamp HCV kit were used to prepare viral DNA from serum and urine, respectively. A total of 100 viral particles could be detected from 0.25 ml of serum when viral DNA was prepared using the QIAamp blood kit (Fig. 2A). The detection
3.2. Determining the sensiti6ity of the assay After the method was optimized for viral DNA preparation, we proceeded to test the sensitivity of the assay. Serial dilutions of the Ad5CMV-p53 virus were made and known amounts of virus were spiked into serum or urine samples from a healthy donor. As shown in Fig. 1, DNA (a 838-bp fragment by PCR) corresponding to only five to ten viral particles could be detected in 0.25 ml of serum (Fig. 1A, lanes 2 and 4) and down to one viral particle could be detected in 0.25 ml of urine (Fig. 1 B, lane 3) by Southern blot analysis. A faster migrating band (about 600 bp in size) observed in Fig. 1A below the specific PCR product was visible in all serum samples in South-
Fig. 1. Autoradiography of the adenovirus detected in serum (A) and urine (B). The DNA was extracted from serum or urine using the method described in the text, resolved by electrophoresis on a 1% agarose gel, Southern blotted, and hybridized with the 838-bp probe A specific for detecting the p53 cDNA and adenovirus boundary. Amounts of DNA corresponding to 50, 5, 10, or 1 viral particles in serum were used in lanes 1, 2, 4, and 5, respectively. Arrow indicates the position of the specific PCR product. Lane 3 was serum only as a negative control. (B) The DNA extracted from urine containing 100, ten, or one viral particles was used in PCR (lanes 1, 2, and 3, respectively). Lane 4 represents urine containing no viral particles and was used as a negative control.
54
Y.C. Henderson et al. / Journal of Virological Methods 71 (1998) 51–56
carrier DNA). The majority of viral DNA was found in the flow-through fractions (Fig. 2B). At least a 10-fold detection difference was also observed when our method was compared with that of the QIAamp HCV kit, when viral DNA was prepared from urine (data not shown, Y. Henderson).
4. Discussion
Fig. 2. Detection of the adenoviral DNA prepared by Qiagen kit. (A) Serum (0.2 ml) spiked with 100 –105 Ad5CMV-p53 particles was prepared according to manufacture protocol of the QIAamp blood kit (Qiagen). All samples were passed through a spin column once. Serum only was used as a negative control. (B) Serum spiked with 104 viral particles was passed continuously through three independent spin columns before the DNA was eluted. The flow through from the first column was passed through a second column and flow through from second column was passed through a third column. DNA eluted from the first spin column (eluate), second column (1st flow through), and third column (2nd flow through) was used in PCR. The specific viral PCR product (indicated by arrow) was detected by Southern blot analysis using 32P-probe A.
sensitivity provided by the QIAamp blood kit was substantially less than the PEG precipitation method described herein by at least 10-fold; we detected ten particles or less in 0.25 ml of serum (Fig. 1A). The most likely explanation for the lower sensitivity of the QIAamp blood kit was due to the poor binding affinity of viral DNA to the Qiagen® spin column (even in the presence of
An easy-to-follow method was developed for the detection of viral DNA from human samples. As few as five viral particles can be detected in 0.25 ml of serum. The same method can be applied to analyze viral DNA in urine with a detection limit of one particle in 0.25 ml of urine. While developing this method, it was found that precipitation with polyethylene glycol was crucial, because centrifugation of the serum for 30 min at 14 000×g without PEG precipitation would only pellet 40% of viral particles, most of the viral particles remaining in the serum (data not shown, Y. Henderson). The PEG precipitation removed efficiently all viral particles from the serum. Since: (a) no viral DNA was found in the supernatant after PEG precipitation as tested by Southern blot analysis; (b) inhibitor was not considered as a factor here since no viral DNA product was detected in PCR from diluted samples; and (c) the intensity of specific PCR product after PEG precipitation from serum was the same as the intensity of specific PCR product obtained from viral DNA prepared from known concentrations of viral particles; we assume that all viral particles were precipitated by the first PEG precipitation. Several rounds of phenol extraction and ethanol precipitation were needed to obtain pure DNA for use in PCR. Viral DNA was not detected from 1 ml serum samples when 5 ml of DNA was used in the PCR, although proteinase K was present in the procedure. In separate experiments, when the identical 1 ml serum sample was extracted twice with phenol or when 1 ml of DNA prepared with one phenol extraction was used in PCR instead of 5 ml, viral DNA was detected after Southern blot (data not shown, Y. Henderson). This observation suggests that precipitated
Y.C. Henderson et al. / Journal of Virological Methods 71 (1998) 51–56
serologic products may inhibit the PCR amplification and identification process. Analysis was therefore limited to 0.5-ml serum aliquots. The sensitivity of our method for the detection of adenoviral DNA is greatly enhanced by the combination of PCR and Southern blot analysis, as compared with the use of PCR alone (Brytting et al., 1992; Cunningham et al., 1995). The sensitivity is also important when a limited sample volume is required. The detection limit of the new method is dependent on the real amount of viral particles being used. To ensure an accurate number of viral particles tested in the sensitivity assays, the titer of the Ad5CMV-p53 virus was determined by plaque assay and converted to viral particles as described in Section 2. The viral particle number was also determined by PCR and Southern blot analysis. DNA was prepared from nine urine samples spiked with one viral particle each. Seven out of nine detected a positive PCR product following Southern blot analysis (data not shown, Y. Henderson). This result follows a Poisson distribution in which 77% of tested samples were found to be positive. This result indicates further that our detection limit for urine sample is one particle in 0.25 ml of urine. The faster migrating band observed in Fig. 1A was observed on Southern blot analysis only. We do not know the nature of this band. Several possibilities for the presence of the band on the Southern blot can be considered. The band might be associated with serum as a non-specific byproduct of the PCR and hybridization since it was detected in serum samples containing no virus (Fig. 1A, lane 3). The possibility of cross-contamination of virus during preparation of the samples cannot be excluded completely. However, every effort was made during the gene transfer phase I trial to prevent cross-contamination of specimens using careful handling and biological safety cabinets throughout all sample manipulations. The faster migrating band was detected in some but not all patient samples tested by Southern blot analysis. Attempts to isolate and further characterize the band were not successful. A technical difficulty was encountered in isolating this band due to its absence from the ethidium bromidestained agarose gel, despite the fact that the inten-
55
sity of the band was as strong as the specific viral product on the Southern blot analysis (Fig. 1A). In conclusion, this procedure provides a sensitive and reproducible method for the detection of adenovirus particles present in serum and urine samples in less than 2 days. Because the antisense primer for PCR is part of the adenovirus, a sense primer for the gene of interest is the only requirement to adapt this method for use with other adenoviral vectors containing genes other than p53. This method has been utilized successfully in an ongoing gene transfer trial to detect Ad5CMVp53 virus in patients’ serum and urine. Although virus may also be harbored in circulating cellular components, these analyses are critical to our understanding of the systemic biodistribution of adenoviral vectors. Acknowledgements This work was supported in part by a NIH R29 grant to Dr Gary Clayman, The University of Texas M.D. Anderson Cancer Center Core Grant CA-16672, and a sponsored research agreement with Introgen Therapeutics. We thank Mary Wang for preparing the Ad5CMV-p53 virus, Lore Feldman for editing the manuscript, Dr Dennis A. Johnston for a helpful discussion of relevant statistical issues, Dr J. Michael Hudson for critical reading of the manuscript, and Lillian K. Meander for editorial assistance. References Brytting, M., Xu, W., Wahren, B., Sundqvist, V.-A., 1992. Cytomegalovirus DNA detection in sera from patients with active cytomegalovirus infections. J. Clin. Microbiol. 30, 1937 – 1941. Clinical protocols, 1996. Cancer Gene Ther. 3, 58 – 68. Cunningham, R., Harris, A., Frankton, A., Irving, W., 1995. Detection of cytomegalovirus using PCR in serum from renal transplant recipients. J. Clin. Pathol. 48, 575 – 577. Graham, F.L., Van Der Eb, A.J., 1973. A new technique for the assay of infectivity of human adenovirus 5 DNA. Virology 52, 456 – 467. Ishigaki, S., Takeda, M., Kura, T., Ban, N., Saitoh, T., Sakamaki, S., Watanabe, N., Kohgo, Y., Niitsu, Y., 1991. Cytomegalovirus DNA in the sera of patients with cytomegalovirus pneumonia. Br. J. Haematol. 79, 198 – 204.
Y.C. Henderson et al. / Journal of Virological Methods 71 (1998) 51–56
56
Maniatis, T., Fritsch, E.F., Sambrook, J., 1989. Molecular Cloning: A Laboratory Manual, 2nd. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Norder, H., Hammas, B., Magnius, L.O., 1990. Typing of hepatitis B virus genomies by a simplified polymerase chain reaction. J. Med. Virol. 31, 215–221.
.
Tokimatsu, I., Tashiro, T., Nasu, M., 1995. Early diagnosis and monitoring of human cytomegalovirus pneumonia in patients with adult T-cell leukemia by DNA amplification in serum. Chest 107, 1024 – 1027. Zhang, W.-W., Fang, X., Branch, C.D., Mazur, W., French, B.A., Roth, J.A., 1993. Generation and identification of recombinant adenovirus by liposome-mediated transfection and PCR analysis. Bio/Techniques 15, 868 – 872.
A simple and sensitive method for detecting adenovirus in serum and urine Ying C. Henderson a, Ta-Jen Liu a,b, Gary L. Clayman a,b,* a
Department of Head and Neck Surgery, Box 69, The Uni6ersity of Texas M.D. Anderson Cancer Center, 1515 Holcombe Bl6d., Houston, TX 77030, USA b Department of Tumor Biology, The Uni6ersity of Texas M.D. Anderson Cancer Center, Houston, TX 77030, USA Accepted 30 October 1997
Abstract A method to detect shed virus in patients’ serum and urine was developed following Ad5CMV-p53 gene transfer via direct tumor injections. The procedure differs from those reported previously in that it first uses polyethylene glycol to precipitate adenoviral particles from patient serum or urine. Adenoviral DNA is then extracted following proteinase K digestion. Finally, polymerase chain reaction (PCR) amplification followed by Southern blot transfer are employed to enhance the limit of detection to less than ten viral particles. The detection limit in a 0.25-ml sample is five viral particles in serum and one viral particle in urine. This procedure is sensitive, reproducible, and can be completed in less than 2 days. © 1998 Elsevier Science B.V. All rights reserved. Keywords: Adenovirus; Serum; Urine; Polyethylene glycol
1. Introduction The adenoviral vector has emerged recently as a front runner in the field of gene therapy (Clinical protocols, 1996). Although adenoviral vectors
* Corresponding author. Tel.: + 1 713 7926920; fax: +1 713 7944662.
provide high transduction efficiencies in a wide range of host cells, their usefulness in human trials has yet to be determined, partly because of insufficient data to support their safety in humans. We describe a sensitive method to detect shed adenovirus in serum and urine from head and neck cancer patients following Ad5CMV-p53 (wild-type p53 in replication defective adenovirus) gene transfer.
0166-0934/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII S 0 1 6 6 - 0 9 3 4 ( 9 7 ) 0 0 1 8 9 - 4
52
Y.C. Henderson et al. / Journal of Virological Methods 71 (1998) 51–56
2. Materials and methods
2.1. Viral DNA preparation Ad5CMV-p53 adenoviral DNA (Zhang et al., 1993) was prepared by a method modified from Cunningham et al. (1995). Briefly, distilled water was added to 0.5-ml aliquots of serum or urine to make a final volume of 1 ml, followed by precipitation with 0.5 ml of 30% polyethylene glycol 8000 (PEG). Following the PEG precipitation, the pellet was resuspended in 0.3 ml of 0.5% NaDodSO4 containing proteinase K (60 mg) and incubated at 50°C for 2–16 h (Norder et al., 1990). Samples were extracted with phenol, and viral DNA precipitated with ethanol in the presence of 10 mg glycogen (Cunningham et al., 1995). Precipitated DNA was recovered by centrifugation at 14 000× g for 10 min at 4°C, resuspended in 0.3 ml of distilled water, and reprecipitated with ethanol. The DNA pellet was rinsed with 70% ethanol, vacuum-dried, and dissolved in 10 ml of distiled water. Only 5 ml of prepared DNA was used in a single polymerase chain reaction (PCR).
2.2. PCR amplification PCR was carried out in a 20-ml volume containing 2 mM MgCl2, 50 mM KCl, 0.1% Triton X-100, 200 mM each of deoxyribonucleotide triphosphates (dNTPs), 10 mM Tris – HCl (pH 9.0), 5 pM each of the primers, and 1.7 U of Taq DNA polymerase (Promega, Madison, WI). The reactions were carried out at 94°C for 30 s, 58°C for 30 s, and 72°C for 60 s for 35 cycles, followed by a 10-min extension at 72°C. PCR primers were selected from the sequence of Ad5CMV-p53 (Zhang et al., 1993) with the sense primer located at the 3% end of the p53 cDNA (5%-GCCTGTCCTGGGAGAGACCG-3%), and the antisense primer located in the adenovirus type 5 genome (5%-CCCTTAAGCCACGCCCACAC-3%). The PCR product (an 838-bp fragment) was separated on 1% agarose gel and transferred to a nylon membrane (Hybond-N+, Amersham, Arlington Heights, IL) by capillary absorption (Maniatis et al., 1989). The same PCR product was subcloned into pCR-Script vector (Stratagene, La Jolla, CA).
The sequence of the insert was confirmed by sequencing and the gel-purified insert used as a probe (probe A) to detect PCR products after labeling with [32P]dCTP using the Megaprime DNA labeling system (Amersham). The specific activity of probe A was approximately 109 dpm/mg.
2.3. Southern blot analysis Southern blot analysis was used to verify the specificity of PCR product. The membrane was prehybridized for 15 min at 65°C in Rapid-hyb buffer (Amersham) and hybridized in the same buffer containing 32P-labelled probe A for 1–2 h. The membrane was washed in 0.1× SSC and 0.1% NaDodSO4 at room temperature twice, and followed by two washes at 65°C (15 min per wash). The washed membrane was exposed to X-ray film at − 70°C for 1–16 h with intensifying screen.
2.4. Titer of the 6irus The titer of CsCl-purified viral particles was determined by plaque assay (Graham and Van Der Eb, 1973). An OD of one at 260 nm is equal to 1012 adenovirus particles/ml. All oligonucleotides were synthesized by Genosys Biotechnologies (Woodlands, TX).
3. Results
3.1. Testing different methods of 6iral DNA preparation Detection of cytomegalovirus and hepatitis B viral DNA in serum by the PCR has been reported (Norder et al., 1990; Ishigaki et al., 1991; Brytting et al., 1992; Cunningham et al., 1995; Tokimatsu et al., 1995). According to those reports, 10–400 ml of serum was used to extract viral DNA and about 10–25% of the extracted DNA was used in PCR. To establish a method for adenovirus DNA detection and to increase the sensitivity to detect adenoviral DNA, which is limited by the small amount of serum used in these assays, several existing methods were tested
Y.C. Henderson et al. / Journal of Virological Methods 71 (1998) 51–56
for cytomegaloviral DNA preparation. Adenovirus (104 Ad5CMV-p53 particles) were spiked into 0.5 or 1 ml of serum. Two PEG precipitations were carried out to ensure that all of the virus particles had been precipitated. Pellets after the first and second PEG precipitations were treated with NaDodSO4 alone or NaDodSO4 containing proteinase K. The samples were extracted with phenol and precipitated with ethanol. We found that proteinase K was required for detection of viral DNA. NaDodSO4 alone was able to release adenoviral DNA, inhibitors were present and no specific viral DNA was detected after PCR (data not shown, Y. Henderson). Precipitation with PEG once was sufficient to concentrate all of the virus. No viral DNA was detected in the supernatant following the first PEG precipitation. Viral DNA could be detected in 0.5 ml of serum but not in 1 ml of serum. However, if 1:10 dilution was made from the 1-ml serum DNA sample, a viral DNA PCR product could be detected in Southern Blot analysis. This observation is due most likely to the presence of PCR inhibitor(s) present in the 1-ml serum sample. The PCR inhibitor(s) could also be removed by repeated extraction with phenol or phenol/chloroform. Repeated extractions, however, result in heavy loss of DNA, thereby providing poor detection sensitivities of viral DNA (data not shown, Y. Henderson). Therefore, serum volumes not exceeding 0.5 ml were used for further analysis.
53
ern blot analysis, including the serum-only sample lacking added virus (lane 3). This band was not detected in ethidium bromide-stained agarose gels.
3.3. Comparison of the optimized 6iral DNA preparation with commercially a6ailable kits The optimized method of adenoviral DNA detection was compared with those of commercially available kits. The QIAamp blood kit (Qiagen) and QIAamp HCV kit were used to prepare viral DNA from serum and urine, respectively. A total of 100 viral particles could be detected from 0.25 ml of serum when viral DNA was prepared using the QIAamp blood kit (Fig. 2A). The detection
3.2. Determining the sensiti6ity of the assay After the method was optimized for viral DNA preparation, we proceeded to test the sensitivity of the assay. Serial dilutions of the Ad5CMV-p53 virus were made and known amounts of virus were spiked into serum or urine samples from a healthy donor. As shown in Fig. 1, DNA (a 838-bp fragment by PCR) corresponding to only five to ten viral particles could be detected in 0.25 ml of serum (Fig. 1A, lanes 2 and 4) and down to one viral particle could be detected in 0.25 ml of urine (Fig. 1 B, lane 3) by Southern blot analysis. A faster migrating band (about 600 bp in size) observed in Fig. 1A below the specific PCR product was visible in all serum samples in South-
Fig. 1. Autoradiography of the adenovirus detected in serum (A) and urine (B). The DNA was extracted from serum or urine using the method described in the text, resolved by electrophoresis on a 1% agarose gel, Southern blotted, and hybridized with the 838-bp probe A specific for detecting the p53 cDNA and adenovirus boundary. Amounts of DNA corresponding to 50, 5, 10, or 1 viral particles in serum were used in lanes 1, 2, 4, and 5, respectively. Arrow indicates the position of the specific PCR product. Lane 3 was serum only as a negative control. (B) The DNA extracted from urine containing 100, ten, or one viral particles was used in PCR (lanes 1, 2, and 3, respectively). Lane 4 represents urine containing no viral particles and was used as a negative control.
54
Y.C. Henderson et al. / Journal of Virological Methods 71 (1998) 51–56
carrier DNA). The majority of viral DNA was found in the flow-through fractions (Fig. 2B). At least a 10-fold detection difference was also observed when our method was compared with that of the QIAamp HCV kit, when viral DNA was prepared from urine (data not shown, Y. Henderson).
4. Discussion
Fig. 2. Detection of the adenoviral DNA prepared by Qiagen kit. (A) Serum (0.2 ml) spiked with 100 –105 Ad5CMV-p53 particles was prepared according to manufacture protocol of the QIAamp blood kit (Qiagen). All samples were passed through a spin column once. Serum only was used as a negative control. (B) Serum spiked with 104 viral particles was passed continuously through three independent spin columns before the DNA was eluted. The flow through from the first column was passed through a second column and flow through from second column was passed through a third column. DNA eluted from the first spin column (eluate), second column (1st flow through), and third column (2nd flow through) was used in PCR. The specific viral PCR product (indicated by arrow) was detected by Southern blot analysis using 32P-probe A.
sensitivity provided by the QIAamp blood kit was substantially less than the PEG precipitation method described herein by at least 10-fold; we detected ten particles or less in 0.25 ml of serum (Fig. 1A). The most likely explanation for the lower sensitivity of the QIAamp blood kit was due to the poor binding affinity of viral DNA to the Qiagen® spin column (even in the presence of
An easy-to-follow method was developed for the detection of viral DNA from human samples. As few as five viral particles can be detected in 0.25 ml of serum. The same method can be applied to analyze viral DNA in urine with a detection limit of one particle in 0.25 ml of urine. While developing this method, it was found that precipitation with polyethylene glycol was crucial, because centrifugation of the serum for 30 min at 14 000×g without PEG precipitation would only pellet 40% of viral particles, most of the viral particles remaining in the serum (data not shown, Y. Henderson). The PEG precipitation removed efficiently all viral particles from the serum. Since: (a) no viral DNA was found in the supernatant after PEG precipitation as tested by Southern blot analysis; (b) inhibitor was not considered as a factor here since no viral DNA product was detected in PCR from diluted samples; and (c) the intensity of specific PCR product after PEG precipitation from serum was the same as the intensity of specific PCR product obtained from viral DNA prepared from known concentrations of viral particles; we assume that all viral particles were precipitated by the first PEG precipitation. Several rounds of phenol extraction and ethanol precipitation were needed to obtain pure DNA for use in PCR. Viral DNA was not detected from 1 ml serum samples when 5 ml of DNA was used in the PCR, although proteinase K was present in the procedure. In separate experiments, when the identical 1 ml serum sample was extracted twice with phenol or when 1 ml of DNA prepared with one phenol extraction was used in PCR instead of 5 ml, viral DNA was detected after Southern blot (data not shown, Y. Henderson). This observation suggests that precipitated
Y.C. Henderson et al. / Journal of Virological Methods 71 (1998) 51–56
serologic products may inhibit the PCR amplification and identification process. Analysis was therefore limited to 0.5-ml serum aliquots. The sensitivity of our method for the detection of adenoviral DNA is greatly enhanced by the combination of PCR and Southern blot analysis, as compared with the use of PCR alone (Brytting et al., 1992; Cunningham et al., 1995). The sensitivity is also important when a limited sample volume is required. The detection limit of the new method is dependent on the real amount of viral particles being used. To ensure an accurate number of viral particles tested in the sensitivity assays, the titer of the Ad5CMV-p53 virus was determined by plaque assay and converted to viral particles as described in Section 2. The viral particle number was also determined by PCR and Southern blot analysis. DNA was prepared from nine urine samples spiked with one viral particle each. Seven out of nine detected a positive PCR product following Southern blot analysis (data not shown, Y. Henderson). This result follows a Poisson distribution in which 77% of tested samples were found to be positive. This result indicates further that our detection limit for urine sample is one particle in 0.25 ml of urine. The faster migrating band observed in Fig. 1A was observed on Southern blot analysis only. We do not know the nature of this band. Several possibilities for the presence of the band on the Southern blot can be considered. The band might be associated with serum as a non-specific byproduct of the PCR and hybridization since it was detected in serum samples containing no virus (Fig. 1A, lane 3). The possibility of cross-contamination of virus during preparation of the samples cannot be excluded completely. However, every effort was made during the gene transfer phase I trial to prevent cross-contamination of specimens using careful handling and biological safety cabinets throughout all sample manipulations. The faster migrating band was detected in some but not all patient samples tested by Southern blot analysis. Attempts to isolate and further characterize the band were not successful. A technical difficulty was encountered in isolating this band due to its absence from the ethidium bromidestained agarose gel, despite the fact that the inten-
55
sity of the band was as strong as the specific viral product on the Southern blot analysis (Fig. 1A). In conclusion, this procedure provides a sensitive and reproducible method for the detection of adenovirus particles present in serum and urine samples in less than 2 days. Because the antisense primer for PCR is part of the adenovirus, a sense primer for the gene of interest is the only requirement to adapt this method for use with other adenoviral vectors containing genes other than p53. This method has been utilized successfully in an ongoing gene transfer trial to detect Ad5CMVp53 virus in patients’ serum and urine. Although virus may also be harbored in circulating cellular components, these analyses are critical to our understanding of the systemic biodistribution of adenoviral vectors. Acknowledgements This work was supported in part by a NIH R29 grant to Dr Gary Clayman, The University of Texas M.D. Anderson Cancer Center Core Grant CA-16672, and a sponsored research agreement with Introgen Therapeutics. We thank Mary Wang for preparing the Ad5CMV-p53 virus, Lore Feldman for editing the manuscript, Dr Dennis A. Johnston for a helpful discussion of relevant statistical issues, Dr J. Michael Hudson for critical reading of the manuscript, and Lillian K. Meander for editorial assistance. References Brytting, M., Xu, W., Wahren, B., Sundqvist, V.-A., 1992. Cytomegalovirus DNA detection in sera from patients with active cytomegalovirus infections. J. Clin. Microbiol. 30, 1937 – 1941. Clinical protocols, 1996. Cancer Gene Ther. 3, 58 – 68. Cunningham, R., Harris, A., Frankton, A., Irving, W., 1995. Detection of cytomegalovirus using PCR in serum from renal transplant recipients. J. Clin. Pathol. 48, 575 – 577. Graham, F.L., Van Der Eb, A.J., 1973. A new technique for the assay of infectivity of human adenovirus 5 DNA. Virology 52, 456 – 467. Ishigaki, S., Takeda, M., Kura, T., Ban, N., Saitoh, T., Sakamaki, S., Watanabe, N., Kohgo, Y., Niitsu, Y., 1991. Cytomegalovirus DNA in the sera of patients with cytomegalovirus pneumonia. Br. J. Haematol. 79, 198 – 204.
Y.C. Henderson et al. / Journal of Virological Methods 71 (1998) 51–56
56
Maniatis, T., Fritsch, E.F., Sambrook, J., 1989. Molecular Cloning: A Laboratory Manual, 2nd. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Norder, H., Hammas, B., Magnius, L.O., 1990. Typing of hepatitis B virus genomies by a simplified polymerase chain reaction. J. Med. Virol. 31, 215–221.
.
Tokimatsu, I., Tashiro, T., Nasu, M., 1995. Early diagnosis and monitoring of human cytomegalovirus pneumonia in patients with adult T-cell leukemia by DNA amplification in serum. Chest 107, 1024 – 1027. Zhang, W.-W., Fang, X., Branch, C.D., Mazur, W., French, B.A., Roth, J.A., 1993. Generation and identification of recombinant adenovirus by liposome-mediated transfection and PCR analysis. Bio/Techniques 15, 868 – 872.