Molecular and Cellular Probes (1992) 6, 32 7-331
Rapid DNA fingerprinting to control for specimen errors in HIV testing by the polymerase chain reaction Sharon Cassol ; * James Rudnik, 2 Teresa Salas,2 Michael Montpetit 2 Richard T. Pon ; Cheikh Tidiane Sy, 2 Stanley Read,4 Carol Major' and Michael V . O'Shaughnessy' 'The Centre of Excellence for HIV/AIDS, University of British Columbia and St Paul's Hospital, 1081 Burrard Street, Vancouver, British Columbia V6Z 1Y6, Canada, 2the Retrovirology Laboratories, Health and Welfare Canada, Ottawa, Ontario, Canada, 'the Regional DNA Synthesis Laboratory, University of Calgary, Alberta, Canada, 'the Department of Pediatrics, Hospital for Sick Children, Toronto, Ontario, Canada, and 'the Ontario Ministry of Health, Toronto, Ontario, Canada (Received 23 March 1992, Accepted 24 April 1992)
Variable-number-tandem-repeats (VNTRs) are highly polymorphic and provide informative genetic markers for distinguishing between individuals . We have used PCR amplification of VNTR locus pMCT118 to identify mislabelled specimens submitted for HIV PCR testing . The method is rapid, can be applied to large numbers of samples and eliminates the need for radioactive probes . DNA samples (10 ng) are amplified for 25 cycles using fluorescence-labelled oligonucleotide primers (blue dye) . An aliquot of the PCR product is then combined with an internal lane size standard (labelled with a red dye), electrophoresed through a 2% agarose gel on an automated fluorescence DNA fragment analyser and the size and quantity of the fragments determined automatically relative to the internal standard . Fifteen alleles, ranging in size from 398 tp 709 by were readily identified in a random sampling of DNA from 63 unrelated HIV-infected patients . Fragment size was reproducible and corresponded to alleles containing from 16 to 35 repeats of a 16 by unit . VNTR genotyping will prove useful for resolving discordant results due to specimen mix-up and ensuring that the correct samples have been analyzed .
KEYWORDS : DNA fingerprinting, specimen identification .
INTRODUCTION It is well recognized that great care must be taken to avoid false-positives when amplifying nucleic acids by the polymerase chain reaction (PCR) . A major source of false-positive results is contamination or carry-over of material from one tube to another .''' Tight quality assurance, redundant testing and stringent screening algorithms are needed to eliminate errors due to contamination .` Another source of error that is common to all test procedures is the possibility of specimen mix-up
when collecting or analyzing large numbers of samples in a batch process . Such errors have been reported in HIV serological testing" and they are a continuing cause for concern, especially when testing low-prevalency populations ." In PCR efficacy studies, conducted in five laboratories with extensive experience in PCR technology, accidental specimen interchange during lysate preparation was the suspected cause of error in six of 13 misclassifications ." An understanding of test limitations and the factors
* Author to whom correspondence should be addressed .
0890-8508/92/040327 + 05 $08 .00/0
327
© 1992 Academic Press Limited
328
S . Cassol et al .
contributing to false-positive PCR results is needed for rational policy-making, especially if PCR is to be used in situations where other tests are non-informative .",' 3 To control for errors resulting from specimen mixup, we have incorporated a DNA fingerprinting step into our PCR testing algorithm . The technique involves amplification of a variable-number-tandemrepeat (VNTR) locus, pMCT118, 14,1 ' and direct detec-
tion of the length polymorphisms on an automated gene fragment analyzer." The method is rapid, can be applied to large numbers of samples and eliminates the need for HLA-DQa amplification to assess the quality and integrity of the DNA . We have used VNTR genotyping to confirm the identify and origin of cell lines and for the identification of mislabelled samples submitted for HIV PCR testing. VNTR typing of large numbers of serial samples will prove valuable for obtaining concrete data on the frequency of administrative and technical errors .
MATERIAL AND METHODS Sample collection and processing VNTR analysis was performed on crude cell lysates and on DNA extracted from dried blood spot (DBS) specimens . Cultured cells and peripheral blood mononuclear cells (PBMCs) isolated from clinical samples by standard ficoll-gradient centrifugation were washed twice in PBS and lysed by resuspending in lysis buffer (50 mm KCI, 10 mm Tris-HCl [pH 8 . 3], 2-5 mm MgCl2, 0 . 45% NP-40, 0 . 45% Tween-20 and
60 pg ml - ' Proteinase-K [Sigma]) to give a final concentration of 6x10' cells/ml . Lysates were then incubated at 56 ° C for 2 h and the Proteinase-K inactivated by heating (95 ° C, 15 min) . DBS samples were eluted into SDS lysis buffer, digested overnight with Proteinase-K at 56 ° C and the genomic DNA recovered by organic extraction" and precipitation onto silica .
Fluorescence-labelled primers The primer sequences were as published by Kasai et al . :" 5'-GAAACT000CTCCAAACACTGCCCGCCG 3' for the forward and 5'-GTCTTGTTGGAGATGCACGTGCCCCTTGC-3' for the reverse primer . To permit the attachment of a fluorescent dye, an amino linker (MMT-C6-linker, Hakabel Scientific) was added to the 5'-end of each primer during the last monomer addition, using a standard synthesis cycle (Applied Biosystems 394 DNA synthesizer) and the Tr-ON/
Auto synthesis end configuration . After deprotection in ammonium hydroxide (50 ° C, 16 h) and lyophilization, the crude product was re-dissolved in 80% acetic acid, left for 1 h, lyophilized and then desalted by gel filtration . The dye-labelling was performed by dissolving 10 ODU of amino-labelled oligonucleotide in pH 9, 0 . 5 M NaHCO 3 /Na2 CO 3 buffer (10 pU and adding 3 pd FAM-NHS ester/DMSO solution (Applied Biosystems) . The dye reactions were left for 1 to 3 days and then purified by preparative gel electrophoresis on 24% acrylamide/7M urea gels . The FAMlabelled product was easily distinguished as an orange or green band, when viewed respectively, against either a white or black background under normal room lighting . The product was extracted from the gel, desalted and quantitated by u .v . spectroscopy .
PCR amplification and gel electrophoresis Samples were denatured (100 ° C, 5 min; 4 ° C, 5 min) and an aliquot equivalent to 1500 cells (or 10 ng of genomic DNA) was reheated to 80 ° C and added to 50 pl of pre-heated (80 ° C) PCR mix to give a final concentration of 50 mm KCI, 10 mm Tris-HCI (pH 8 . 3), 2 . 5 mm MgCl,, 0 . 25% Tween-20, 0. 25% Nonidet P-40, 0 . 2 m m each of dATP, dCTP, dGTP and dTTP, 5 pmol of each primer and 1 U of AmpliTag
TM (Cetus) . Ampli-
fication was performed for 25 cycles (Perkin-Elmer/ Cetus thermocycler) using a thermoprofile of : 95° C, 1 min (denaturation) ; 65 ° C, 1 min (annealing) and 70 ° C, 8 min (extension) and a final extension of 13 min at 70 ° C. Following PCR, 2 pl of each amplified sample was mixed with an equal volume of 2X loading buffer (50 mg ml - ' Ficoll, 8 . 5 mg ml - ' blue dextran and 1 .7 mg ml - ' dextran sulfate) and 8 fmol of an internal lane size standard, Genescan-1000 labelled with a red (ROX) fluorophore (Applied Biosystems) . One-half of each sample was then loaded into a 2% agarose gel on an Applied Biosystems GeneScanner TM 362 Fluor-
escent Fragment Analyzer and electrophoresed in Tris-borate buffer (pH 8 . 0) for 5 h at 100V using a 'well-to-read' distance of 4 cm . The FAM-labelled PCR products, detected by laser scanning, were computer-analyzed and the quantities and lengths of the PCR products established automatically relative to the internal ROX-standard .
RESULTS To obtain an estimate of the number of alleles that would be revealed in a population of HIV-infected individuals, a cross-section of PBMC and DBS samples
Rapid DNA fingerprinting submitted to the Retrovirology Laboratories at the Laboratory Centre for Disease Control was amplified at the pMCT118 locus . Fifteen alleles ranging in size from 398 to 709 by were readily identified in a random sampling of 63 unrelated patients with ARC/ AIDS and asymptomatic HIV infection . The size of the allele fragments was highly reproducible and corresponded to previously published pMCT118 length polymorphisms containing from 16 to 35 repeats of a 16 by un it. 14,16 Distribution of the allele frequencies is shown in Fig. 1 . The observed heterozygosity was 74 . 6% with the most common genotype consisting of the 432 and 524 by alleles with a combined frequency of 0 . 11 . Amplification of 12 human cell lines gave a heterozygosity of 81 . 8o/o with three cell lines (HL-60, CEM-13D and the closely related CEMSS line) expressing the 432 allele in combination with 524 . To assess the ability of VNTR to distinguish between individuals and identify errors, we retrospectively examined records and re-analyzed samples where mix-up was suspected, but not confirmed . One set of serial samples came from an at-risk homosexual man (patient D .C .) sampled at four time intervals before and during seroconversion . Serum samples were tested for HIV antibodies by ELISA and western blot and for HIV p24 antigen by antigen capture ; matched PBMC lysates were amplified for HIV-proviral DNA . Serological assays were completely negative for the first two time points (June 1985 and September 1985) . A third serum sample (December 1985) tested p24 antigen positive and by the fourth sampling (May
1986) the patient's serum showed full seroconversion to HIV antibody positivity, as determined by both ELISA and Western blotting . HIV DNA was detectable in the third and fourth lysates but, in addition, the earliest (June) lysate was also strongly and repeatedly PCR positive . The second lysate was PCR negative . Taken at face value, these findings could suggest that PCR was the first indicator of infection, becoming positive 6 months before the appearance of antigen and 11 months before seroconversion . However, as shown in Fig . 2, the earliest PCR positive was a false result due to specimen mix-up . The VNTR pattern of the first lysate (Fig. 2, lane 3) was distinctly different from the other three lysates (Fig . 2, lanes 6, 9 and 12), providing strong evidence that it came from a different individual . After exclusion of the mislabelled sample, PCR positivity was shown to be co-incident with the first appearance of p24 antigen, a maximum of 5 months before seroconversion . In two other instances, specimen interchange was suspected when the numbers on coded PCR tubes did not correspond with the sample list (two samples) and when serological results (provided by the sending laboratory) did not correlate with results of PCR testing for eight serial samples submitted on the same day . In the latter instance, specimen mix-up was confirmed when it was revealed that a serum sample had been accidentally discarded and the serum (but not the PCR) tubes re-numbered to reflect this loss . All 10 samples gave a unique VNTR pattern which could be used to resolve the discrepancies by comparative testing of repeat samples of known identity .
60 55
Number of repeats Fig . 1 .
bars) .
329
Distribution of pMCT118 alleles among 63 HIV-infected patients (open bars) and 12 cell lines (solid
S. Cassol et al.
330 350
a,
550
750
950
1150
1350
400
U
u, 200 U U1 d
o
0 ∎ Lane 6 : T oronto= 64 1 :50
d
a
m
400 200 0 ∎ Lane 9 : Toronto=65 1 :50
V
400 200 0 ∎ Lane 12 : Toronto=661 :50
V
Time of electrophoresis
Fig. 2 . Electrophoretogram view of PCR products generated at the pMCT118 VNTR . Lane 3 : amplification of sample incorrectly assigned to patient D .C. (June 1985); lanes 6, 9, 12 : amplification of sequential samples from patient D .C . (September 1985, December 1985 and May 1986, respectively) . The VNTR sizing and allele assignments of the three samples from patient D .C . are concurrent (432 and 464 bp) and distinctly different from those of the mislabelled sample (524 and 582 bp) .
DISCUSSION
PCR testing is becoming recognized as a major research tool for the direct detection of HIV in blood and body fluids . The main diagnostic benefit of PCR, however, will be to recognize infection in clinical situations where other tests fail (i .e . in early stage patients or the perinatal period) ." ," Since other markers of infection may not be possible, interpretation of PCR will require a thorough knowledge of its technical limitations and the introduction of improved procedures to reduce false-positive (and false-negative) results . This report describes the use of VNTR genotyping to control for administrative errors and reduce the risk of accidental (or deliberate)$ mislabelling . Comparison of VNTR profiles from initial and confirmatory test samples, would alert the diagnostician to sample processing errors and help ensure that the correct samples had been analysed . The method is rapid, simple and provides an attractive alternative to traditional HLA-DQa or (3-globulin amplification to measure the quality and suitability of samples for PCR studies . Failure to generate a PCR product with primers to human sequences indicates that the DNA is degraded or not present in adequate amounts or that PCR inhibitors are present .
Although false-positive HIV results can be eventually resolved by follow-up, it is often difficult to reassure the patient that he/she is indeed negative or to convince institutions that a sampling error has occurred .' The ability to unequivocally demonstrate these errors by VNTR typing would provide faster and more convincing evidence . Another important application of VNTR would be to resolve discrepancies in the literature. A few laboratories have reported the detection of HIV DNA by PCR in at-risk individuals as long as 16 to 35 months before overt antibody seroconversion ." ," However, the significance of positive PCR results in seronegative infections remains controversial, especially since more recent studies have found little evidence for a long latency period between exposure to the virus and seroconversion .20 23 The availability of VNTR typing now means that early PCR positives can be re-examined to identify (or rule out) specimen errors . Accurate definition of the time to seroconversion has important clinical, blood banking and public health implications . In summary, VNTR genotyping should prove valuable for sample identification, not only in the HIV setting, but also in other clinical situations where erroneous diagnosis can have serious consequences
Rapid DNA fingerprinting
for both the screening program and the individual . The method is non-radiometric, accurate, automated and directly applicable to other informative VNTR loci such as apoB (12 alleles), 24 SM7 (14 alleles)" and pYNZ22 (10 alleles) ." Multiplex PCR using 3 or 4 sets of primers labelled with different fluorescent dyes27 would be expected to increase the power of discrimination in terms of individual identity . However, these assays will have to be carefully optimized to avoid artifacts such as the generation of high molecular weight heteroduplexes when template or primer concentrations are too high, or when the annealing 24-27 temperature is too low .
REFERENCES 1 . Kellog, D . E . & Kwok, S . (1990) . Detection of human immunodeficiency virus . In PCR protocols, a guide to methods and applications . (Innis M . A ., Gelfand, D . H ., Sninsky, J . J ., White, T . J ., eds), pp . 337-347 . New York : Academic Press, Inc . 2 . Clewley, J . P. (1989) . The polymerase chain reaction, a review of the practical limitations for human immunodeficiency virus diagnosis . Journal of Virological Methods 25, 179-88 . 3 . Kwok, S . & Higuchi, R. (1989) . Avoiding false positives with PCR . Nature (London) 339, 237-8 . 4 . Kitchin, P . A ., Szotyori, Z., Fromhold, C ., & Almon, N . (1990) . Avoidance of false positive. Nature (London) 344, 201 . 5 . Cimino, G . D ., Metchette, K . C ., Tessman, J . E ., Hearst, J . E . & Isaacs, S . T. (1991) . Post-amplification sterilization technique: a method to control PCR 'carryover for diagnostics . Nucleic Acids Research 19, 99-107 . 6 . Perkin-Elmer/Cetus . (1991) . GeneAmp PCR carry-over prevention kit. Amplifications 6, 12-13 . 7 . Wages, J . M ., Hamdallah, M., Calabro, M . A . et al. (1991) . Clinical performances of a polymerase chain reaction testing algorithm for diagnosis of HIV-1 infection in peripheral blood mononuclear cells . Journal of Medical Virology 33, 58-63 . 8 . Gill, M . J ., Rachlis, A . & Anand, C . (1991) . Fives cases of erroneously diagnosed HIV infection . Canadian Medical Association journal 145, 1593-5 . 9 . Holmberg, S . D ., Byers, R . H ., Schable, C . A ., Lifson, A . R ., Seage, G . R . & Schoenbaum, E . E . (1988) . Errors in reporting seropositivity for infection with human immunodeficiency virus (HIV) . Annales of Internal Medicine 108, 679-80. 10 . Meyer, K . B. & Pauker, S . G . (1987) . Screening for HIV : Can we afford the false-positive rate . New England Journal of Medicine 317, 238-41 . 11 . Sheppard, H . W ., Ascher, M . S ., Busch, P . P . et al. (1991) . A multicenter proficiency trial of gene amplification (PCR) for the detection of HIV-1 . Journal of Acquired Immune Deficiency Syndromes 4, 277-83 . 12 . Rogers, M . F., Ou, C-Y ., Rayfield, P. A . et al. (1989) . Use of the polymerase chain reaction for early detection of the proviral sequences of human immunodeficiency virus in infants born to seropositive mothers . New England Journal of Medicine 320, 1649-54 .
33 1
13 . Cassol, S. A ., Lapointe, N ., Salas, T . et al. (1992) . Diagnosis of vertical HIV-1 transmission using the polymerase chain reaction and dried blood spot specimens . Journal of Acquired Immune Deficiency Syndromes 5,
113-19 . 14 . Kasai, K ., Nakamura, Y ., & White, R . (1990) . Amplification of a variable number of tandem repeats (VNTR) locus (pMCT118) by the polymerase chain reaction (PCR) and its application to forensic science . Journal of Forensic Science 35, 1196-200. 15 . Nakamura, Y ., Carlson, M ., Krapcho, K., & White, R . (1988) . Isolation and mapping of a polymorphic DNA sequence, pMCT118, on Chromosome 1p (D1S58) . Nucleic Acids Research 16, 9364. 16 . Robertson, J ., Ziegle, J ., Kronick, M., Madden, D ., & Budowle, B . (1991) . Genetic typing using automated electrophoresis and fluorescence detection . In DNA Fingerpringting Approaches and Applications . (Burke, T ., Dolf, A., Jeffreys, J . and Wolff, R . Eds), pp . 391-8 . 17 . Cassol, S . A ., Salas, T ., Arella, M . et al . (1991). Use of dried blood spot specimens in the detection of human immunodeficiency virus type I by the polymerase chain reaction . Journal of Clinical Microbiology 29, 667-71 . 18 . Imagawa, D . T ., Lee, M . H ., Wolinsky, S . M. et al. (1989) . Human immunodeficiency virus type I infection in homosexual men who remain seronegative for prolonged periods . New England Journal of Medicine 320, 1458-62 . 19 . Ranki, A ., Valle, S . L., Krohn, M . et al. (1988). Long latency precedes overt seroconversion in sexually transmitted active wives of infected hemophilic men . American Journal of Medicine 85, 27-32 . 20. Horsburgh, C . R ., Jason, J ., Longini, I . M . et al . (1989) . Duration of human immunodeficiency virus infection before detection of antibody . Lancet i, 637-40. 21 . Jackson, J . B ., Kwok, S . Y ., Hopsicker, J . S . et al. (1989) . Absence of HIV-1 infection in antibody-negative sexual partners of HIV-1 infected hemophiliacs . Transfusion 29, 265-7. 22 . Nielsen, C ., Teglbjaerg, L . S ., Pedersen, C ., Lundrgren, J . D ., Nielsen, C. M. & Vestergaard, B . F . (1991) . Prevalence of HIV infection in seronegative high-risk, individuals examined by virus isolation and PCR . Journal of Acquired Immune Deficiency Syndromes 4, 1107-11 . 23 . Jason, J., Ou, C-Y ., Moore, J . L ., Lawrence, D . N ., Schochetman, G . & Evatt, B . L. (1989) . Prevalence of human immunodeficiency virus type I in hemophilic men and their sex partners . Journal of Infectious Diseases 160, 789-94. 24. Boerwinkle, E ., Siong, W ., Fourest, E . & Chan, L . (1989) . Rapid typing of tandemly repeated hypervariable loci by the polymerase chain reaction : application to the apolipoprotein B 3' hypervariable region . Proceedings of the National Academy of Science, USA 86, 212-16 . 25 . Harris, P . C ., Thomas, S ., Ratcliffe, P . J ., Breuning, NI . H ., Coto, E . & Lopez-Larrea, C . (1991) . Rapid genetic analysis of families with polycystic kidney disease 1 by means of a microsatellite marker . Lancet 338, 1484-7 . 26 . Horn, G . T ., Richards, B . & Klinger, K . W . (1989) . Amplification of a highly polymorphic VNTR segment by the polymerase chain reaction . Nucleic Acids Research 17, 2140 . 27 . Jeffreys, A . J., Wilson, V ., Neumann, R . & Keyte, J . (1988) . Amplification of human minisatellites by the polymerase chain reaction . Nucleic Acids Research 16, 1095371 .
Rapid DNA fingerprinting to control for specimen errors in HIV testing by the polymerase chain reaction Sharon Cassol ; * James Rudnik, 2 Teresa Salas,2 Michael Montpetit 2 Richard T. Pon ; Cheikh Tidiane Sy, 2 Stanley Read,4 Carol Major' and Michael V . O'Shaughnessy' 'The Centre of Excellence for HIV/AIDS, University of British Columbia and St Paul's Hospital, 1081 Burrard Street, Vancouver, British Columbia V6Z 1Y6, Canada, 2the Retrovirology Laboratories, Health and Welfare Canada, Ottawa, Ontario, Canada, 'the Regional DNA Synthesis Laboratory, University of Calgary, Alberta, Canada, 'the Department of Pediatrics, Hospital for Sick Children, Toronto, Ontario, Canada, and 'the Ontario Ministry of Health, Toronto, Ontario, Canada (Received 23 March 1992, Accepted 24 April 1992)
Variable-number-tandem-repeats (VNTRs) are highly polymorphic and provide informative genetic markers for distinguishing between individuals . We have used PCR amplification of VNTR locus pMCT118 to identify mislabelled specimens submitted for HIV PCR testing . The method is rapid, can be applied to large numbers of samples and eliminates the need for radioactive probes . DNA samples (10 ng) are amplified for 25 cycles using fluorescence-labelled oligonucleotide primers (blue dye) . An aliquot of the PCR product is then combined with an internal lane size standard (labelled with a red dye), electrophoresed through a 2% agarose gel on an automated fluorescence DNA fragment analyser and the size and quantity of the fragments determined automatically relative to the internal standard . Fifteen alleles, ranging in size from 398 tp 709 by were readily identified in a random sampling of DNA from 63 unrelated HIV-infected patients . Fragment size was reproducible and corresponded to alleles containing from 16 to 35 repeats of a 16 by unit . VNTR genotyping will prove useful for resolving discordant results due to specimen mix-up and ensuring that the correct samples have been analyzed .
KEYWORDS : DNA fingerprinting, specimen identification .
INTRODUCTION It is well recognized that great care must be taken to avoid false-positives when amplifying nucleic acids by the polymerase chain reaction (PCR) . A major source of false-positive results is contamination or carry-over of material from one tube to another .''' Tight quality assurance, redundant testing and stringent screening algorithms are needed to eliminate errors due to contamination .` Another source of error that is common to all test procedures is the possibility of specimen mix-up
when collecting or analyzing large numbers of samples in a batch process . Such errors have been reported in HIV serological testing" and they are a continuing cause for concern, especially when testing low-prevalency populations ." In PCR efficacy studies, conducted in five laboratories with extensive experience in PCR technology, accidental specimen interchange during lysate preparation was the suspected cause of error in six of 13 misclassifications ." An understanding of test limitations and the factors
* Author to whom correspondence should be addressed .
0890-8508/92/040327 + 05 $08 .00/0
327
© 1992 Academic Press Limited
328
S . Cassol et al .
contributing to false-positive PCR results is needed for rational policy-making, especially if PCR is to be used in situations where other tests are non-informative .",' 3 To control for errors resulting from specimen mixup, we have incorporated a DNA fingerprinting step into our PCR testing algorithm . The technique involves amplification of a variable-number-tandemrepeat (VNTR) locus, pMCT118, 14,1 ' and direct detec-
tion of the length polymorphisms on an automated gene fragment analyzer." The method is rapid, can be applied to large numbers of samples and eliminates the need for HLA-DQa amplification to assess the quality and integrity of the DNA . We have used VNTR genotyping to confirm the identify and origin of cell lines and for the identification of mislabelled samples submitted for HIV PCR testing. VNTR typing of large numbers of serial samples will prove valuable for obtaining concrete data on the frequency of administrative and technical errors .
MATERIAL AND METHODS Sample collection and processing VNTR analysis was performed on crude cell lysates and on DNA extracted from dried blood spot (DBS) specimens . Cultured cells and peripheral blood mononuclear cells (PBMCs) isolated from clinical samples by standard ficoll-gradient centrifugation were washed twice in PBS and lysed by resuspending in lysis buffer (50 mm KCI, 10 mm Tris-HCl [pH 8 . 3], 2-5 mm MgCl2, 0 . 45% NP-40, 0 . 45% Tween-20 and
60 pg ml - ' Proteinase-K [Sigma]) to give a final concentration of 6x10' cells/ml . Lysates were then incubated at 56 ° C for 2 h and the Proteinase-K inactivated by heating (95 ° C, 15 min) . DBS samples were eluted into SDS lysis buffer, digested overnight with Proteinase-K at 56 ° C and the genomic DNA recovered by organic extraction" and precipitation onto silica .
Fluorescence-labelled primers The primer sequences were as published by Kasai et al . :" 5'-GAAACT000CTCCAAACACTGCCCGCCG 3' for the forward and 5'-GTCTTGTTGGAGATGCACGTGCCCCTTGC-3' for the reverse primer . To permit the attachment of a fluorescent dye, an amino linker (MMT-C6-linker, Hakabel Scientific) was added to the 5'-end of each primer during the last monomer addition, using a standard synthesis cycle (Applied Biosystems 394 DNA synthesizer) and the Tr-ON/
Auto synthesis end configuration . After deprotection in ammonium hydroxide (50 ° C, 16 h) and lyophilization, the crude product was re-dissolved in 80% acetic acid, left for 1 h, lyophilized and then desalted by gel filtration . The dye-labelling was performed by dissolving 10 ODU of amino-labelled oligonucleotide in pH 9, 0 . 5 M NaHCO 3 /Na2 CO 3 buffer (10 pU and adding 3 pd FAM-NHS ester/DMSO solution (Applied Biosystems) . The dye reactions were left for 1 to 3 days and then purified by preparative gel electrophoresis on 24% acrylamide/7M urea gels . The FAMlabelled product was easily distinguished as an orange or green band, when viewed respectively, against either a white or black background under normal room lighting . The product was extracted from the gel, desalted and quantitated by u .v . spectroscopy .
PCR amplification and gel electrophoresis Samples were denatured (100 ° C, 5 min; 4 ° C, 5 min) and an aliquot equivalent to 1500 cells (or 10 ng of genomic DNA) was reheated to 80 ° C and added to 50 pl of pre-heated (80 ° C) PCR mix to give a final concentration of 50 mm KCI, 10 mm Tris-HCI (pH 8 . 3), 2 . 5 mm MgCl,, 0 . 25% Tween-20, 0. 25% Nonidet P-40, 0 . 2 m m each of dATP, dCTP, dGTP and dTTP, 5 pmol of each primer and 1 U of AmpliTag
TM (Cetus) . Ampli-
fication was performed for 25 cycles (Perkin-Elmer/ Cetus thermocycler) using a thermoprofile of : 95° C, 1 min (denaturation) ; 65 ° C, 1 min (annealing) and 70 ° C, 8 min (extension) and a final extension of 13 min at 70 ° C. Following PCR, 2 pl of each amplified sample was mixed with an equal volume of 2X loading buffer (50 mg ml - ' Ficoll, 8 . 5 mg ml - ' blue dextran and 1 .7 mg ml - ' dextran sulfate) and 8 fmol of an internal lane size standard, Genescan-1000 labelled with a red (ROX) fluorophore (Applied Biosystems) . One-half of each sample was then loaded into a 2% agarose gel on an Applied Biosystems GeneScanner TM 362 Fluor-
escent Fragment Analyzer and electrophoresed in Tris-borate buffer (pH 8 . 0) for 5 h at 100V using a 'well-to-read' distance of 4 cm . The FAM-labelled PCR products, detected by laser scanning, were computer-analyzed and the quantities and lengths of the PCR products established automatically relative to the internal ROX-standard .
RESULTS To obtain an estimate of the number of alleles that would be revealed in a population of HIV-infected individuals, a cross-section of PBMC and DBS samples
Rapid DNA fingerprinting submitted to the Retrovirology Laboratories at the Laboratory Centre for Disease Control was amplified at the pMCT118 locus . Fifteen alleles ranging in size from 398 to 709 by were readily identified in a random sampling of 63 unrelated patients with ARC/ AIDS and asymptomatic HIV infection . The size of the allele fragments was highly reproducible and corresponded to previously published pMCT118 length polymorphisms containing from 16 to 35 repeats of a 16 by un it. 14,16 Distribution of the allele frequencies is shown in Fig. 1 . The observed heterozygosity was 74 . 6% with the most common genotype consisting of the 432 and 524 by alleles with a combined frequency of 0 . 11 . Amplification of 12 human cell lines gave a heterozygosity of 81 . 8o/o with three cell lines (HL-60, CEM-13D and the closely related CEMSS line) expressing the 432 allele in combination with 524 . To assess the ability of VNTR to distinguish between individuals and identify errors, we retrospectively examined records and re-analyzed samples where mix-up was suspected, but not confirmed . One set of serial samples came from an at-risk homosexual man (patient D .C .) sampled at four time intervals before and during seroconversion . Serum samples were tested for HIV antibodies by ELISA and western blot and for HIV p24 antigen by antigen capture ; matched PBMC lysates were amplified for HIV-proviral DNA . Serological assays were completely negative for the first two time points (June 1985 and September 1985) . A third serum sample (December 1985) tested p24 antigen positive and by the fourth sampling (May
1986) the patient's serum showed full seroconversion to HIV antibody positivity, as determined by both ELISA and Western blotting . HIV DNA was detectable in the third and fourth lysates but, in addition, the earliest (June) lysate was also strongly and repeatedly PCR positive . The second lysate was PCR negative . Taken at face value, these findings could suggest that PCR was the first indicator of infection, becoming positive 6 months before the appearance of antigen and 11 months before seroconversion . However, as shown in Fig . 2, the earliest PCR positive was a false result due to specimen mix-up . The VNTR pattern of the first lysate (Fig. 2, lane 3) was distinctly different from the other three lysates (Fig . 2, lanes 6, 9 and 12), providing strong evidence that it came from a different individual . After exclusion of the mislabelled sample, PCR positivity was shown to be co-incident with the first appearance of p24 antigen, a maximum of 5 months before seroconversion . In two other instances, specimen interchange was suspected when the numbers on coded PCR tubes did not correspond with the sample list (two samples) and when serological results (provided by the sending laboratory) did not correlate with results of PCR testing for eight serial samples submitted on the same day . In the latter instance, specimen mix-up was confirmed when it was revealed that a serum sample had been accidentally discarded and the serum (but not the PCR) tubes re-numbered to reflect this loss . All 10 samples gave a unique VNTR pattern which could be used to resolve the discrepancies by comparative testing of repeat samples of known identity .
60 55
Number of repeats Fig . 1 .
bars) .
329
Distribution of pMCT118 alleles among 63 HIV-infected patients (open bars) and 12 cell lines (solid
S. Cassol et al.
330 350
a,
550
750
950
1150
1350
400
U
u, 200 U U1 d
o
0 ∎ Lane 6 : T oronto= 64 1 :50
d
a
m
400 200 0 ∎ Lane 9 : Toronto=65 1 :50
V
400 200 0 ∎ Lane 12 : Toronto=661 :50
V
Time of electrophoresis
Fig. 2 . Electrophoretogram view of PCR products generated at the pMCT118 VNTR . Lane 3 : amplification of sample incorrectly assigned to patient D .C. (June 1985); lanes 6, 9, 12 : amplification of sequential samples from patient D .C . (September 1985, December 1985 and May 1986, respectively) . The VNTR sizing and allele assignments of the three samples from patient D .C . are concurrent (432 and 464 bp) and distinctly different from those of the mislabelled sample (524 and 582 bp) .
DISCUSSION
PCR testing is becoming recognized as a major research tool for the direct detection of HIV in blood and body fluids . The main diagnostic benefit of PCR, however, will be to recognize infection in clinical situations where other tests fail (i .e . in early stage patients or the perinatal period) ." ," Since other markers of infection may not be possible, interpretation of PCR will require a thorough knowledge of its technical limitations and the introduction of improved procedures to reduce false-positive (and false-negative) results . This report describes the use of VNTR genotyping to control for administrative errors and reduce the risk of accidental (or deliberate)$ mislabelling . Comparison of VNTR profiles from initial and confirmatory test samples, would alert the diagnostician to sample processing errors and help ensure that the correct samples had been analysed . The method is rapid, simple and provides an attractive alternative to traditional HLA-DQa or (3-globulin amplification to measure the quality and suitability of samples for PCR studies . Failure to generate a PCR product with primers to human sequences indicates that the DNA is degraded or not present in adequate amounts or that PCR inhibitors are present .
Although false-positive HIV results can be eventually resolved by follow-up, it is often difficult to reassure the patient that he/she is indeed negative or to convince institutions that a sampling error has occurred .' The ability to unequivocally demonstrate these errors by VNTR typing would provide faster and more convincing evidence . Another important application of VNTR would be to resolve discrepancies in the literature. A few laboratories have reported the detection of HIV DNA by PCR in at-risk individuals as long as 16 to 35 months before overt antibody seroconversion ." ," However, the significance of positive PCR results in seronegative infections remains controversial, especially since more recent studies have found little evidence for a long latency period between exposure to the virus and seroconversion .20 23 The availability of VNTR typing now means that early PCR positives can be re-examined to identify (or rule out) specimen errors . Accurate definition of the time to seroconversion has important clinical, blood banking and public health implications . In summary, VNTR genotyping should prove valuable for sample identification, not only in the HIV setting, but also in other clinical situations where erroneous diagnosis can have serious consequences
Rapid DNA fingerprinting
for both the screening program and the individual . The method is non-radiometric, accurate, automated and directly applicable to other informative VNTR loci such as apoB (12 alleles), 24 SM7 (14 alleles)" and pYNZ22 (10 alleles) ." Multiplex PCR using 3 or 4 sets of primers labelled with different fluorescent dyes27 would be expected to increase the power of discrimination in terms of individual identity . However, these assays will have to be carefully optimized to avoid artifacts such as the generation of high molecular weight heteroduplexes when template or primer concentrations are too high, or when the annealing 24-27 temperature is too low .
REFERENCES 1 . Kellog, D . E . & Kwok, S . (1990) . Detection of human immunodeficiency virus . In PCR protocols, a guide to methods and applications . (Innis M . A ., Gelfand, D . H ., Sninsky, J . J ., White, T . J ., eds), pp . 337-347 . New York : Academic Press, Inc . 2 . Clewley, J . P. (1989) . The polymerase chain reaction, a review of the practical limitations for human immunodeficiency virus diagnosis . Journal of Virological Methods 25, 179-88 . 3 . Kwok, S . & Higuchi, R. (1989) . Avoiding false positives with PCR . Nature (London) 339, 237-8 . 4 . Kitchin, P . A ., Szotyori, Z., Fromhold, C ., & Almon, N . (1990) . Avoidance of false positive. Nature (London) 344, 201 . 5 . Cimino, G . D ., Metchette, K . C ., Tessman, J . E ., Hearst, J . E . & Isaacs, S . T. (1991) . Post-amplification sterilization technique: a method to control PCR 'carryover for diagnostics . Nucleic Acids Research 19, 99-107 . 6 . Perkin-Elmer/Cetus . (1991) . GeneAmp PCR carry-over prevention kit. Amplifications 6, 12-13 . 7 . Wages, J . M ., Hamdallah, M., Calabro, M . A . et al. (1991) . Clinical performances of a polymerase chain reaction testing algorithm for diagnosis of HIV-1 infection in peripheral blood mononuclear cells . Journal of Medical Virology 33, 58-63 . 8 . Gill, M . J ., Rachlis, A . & Anand, C . (1991) . Fives cases of erroneously diagnosed HIV infection . Canadian Medical Association journal 145, 1593-5 . 9 . Holmberg, S . D ., Byers, R . H ., Schable, C . A ., Lifson, A . R ., Seage, G . R . & Schoenbaum, E . E . (1988) . Errors in reporting seropositivity for infection with human immunodeficiency virus (HIV) . Annales of Internal Medicine 108, 679-80. 10 . Meyer, K . B. & Pauker, S . G . (1987) . Screening for HIV : Can we afford the false-positive rate . New England Journal of Medicine 317, 238-41 . 11 . Sheppard, H . W ., Ascher, M . S ., Busch, P . P . et al. (1991) . A multicenter proficiency trial of gene amplification (PCR) for the detection of HIV-1 . Journal of Acquired Immune Deficiency Syndromes 4, 277-83 . 12 . Rogers, M . F., Ou, C-Y ., Rayfield, P. A . et al. (1989) . Use of the polymerase chain reaction for early detection of the proviral sequences of human immunodeficiency virus in infants born to seropositive mothers . New England Journal of Medicine 320, 1649-54 .
33 1
13 . Cassol, S. A ., Lapointe, N ., Salas, T . et al. (1992) . Diagnosis of vertical HIV-1 transmission using the polymerase chain reaction and dried blood spot specimens . Journal of Acquired Immune Deficiency Syndromes 5,
113-19 . 14 . Kasai, K ., Nakamura, Y ., & White, R . (1990) . Amplification of a variable number of tandem repeats (VNTR) locus (pMCT118) by the polymerase chain reaction (PCR) and its application to forensic science . Journal of Forensic Science 35, 1196-200. 15 . Nakamura, Y ., Carlson, M ., Krapcho, K., & White, R . (1988) . Isolation and mapping of a polymorphic DNA sequence, pMCT118, on Chromosome 1p (D1S58) . Nucleic Acids Research 16, 9364. 16 . Robertson, J ., Ziegle, J ., Kronick, M., Madden, D ., & Budowle, B . (1991) . Genetic typing using automated electrophoresis and fluorescence detection . In DNA Fingerpringting Approaches and Applications . (Burke, T ., Dolf, A., Jeffreys, J . and Wolff, R . Eds), pp . 391-8 . 17 . Cassol, S . A ., Salas, T ., Arella, M . et al . (1991). Use of dried blood spot specimens in the detection of human immunodeficiency virus type I by the polymerase chain reaction . Journal of Clinical Microbiology 29, 667-71 . 18 . Imagawa, D . T ., Lee, M . H ., Wolinsky, S . M. et al. (1989) . Human immunodeficiency virus type I infection in homosexual men who remain seronegative for prolonged periods . New England Journal of Medicine 320, 1458-62 . 19 . Ranki, A ., Valle, S . L., Krohn, M . et al. (1988). Long latency precedes overt seroconversion in sexually transmitted active wives of infected hemophilic men . American Journal of Medicine 85, 27-32 . 20. Horsburgh, C . R ., Jason, J ., Longini, I . M . et al . (1989) . Duration of human immunodeficiency virus infection before detection of antibody . Lancet i, 637-40. 21 . Jackson, J . B ., Kwok, S . Y ., Hopsicker, J . S . et al. (1989) . Absence of HIV-1 infection in antibody-negative sexual partners of HIV-1 infected hemophiliacs . Transfusion 29, 265-7. 22 . Nielsen, C ., Teglbjaerg, L . S ., Pedersen, C ., Lundrgren, J . D ., Nielsen, C. M. & Vestergaard, B . F . (1991) . Prevalence of HIV infection in seronegative high-risk, individuals examined by virus isolation and PCR . Journal of Acquired Immune Deficiency Syndromes 4, 1107-11 . 23 . Jason, J., Ou, C-Y ., Moore, J . L ., Lawrence, D . N ., Schochetman, G . & Evatt, B . L. (1989) . Prevalence of human immunodeficiency virus type I in hemophilic men and their sex partners . Journal of Infectious Diseases 160, 789-94. 24. Boerwinkle, E ., Siong, W ., Fourest, E . & Chan, L . (1989) . Rapid typing of tandemly repeated hypervariable loci by the polymerase chain reaction : application to the apolipoprotein B 3' hypervariable region . Proceedings of the National Academy of Science, USA 86, 212-16 . 25 . Harris, P . C ., Thomas, S ., Ratcliffe, P . J ., Breuning, NI . H ., Coto, E . & Lopez-Larrea, C . (1991) . Rapid genetic analysis of families with polycystic kidney disease 1 by means of a microsatellite marker . Lancet 338, 1484-7 . 26 . Horn, G . T ., Richards, B . & Klinger, K . W . (1989) . Amplification of a highly polymorphic VNTR segment by the polymerase chain reaction . Nucleic Acids Research 17, 2140 . 27 . Jeffreys, A . J., Wilson, V ., Neumann, R . & Keyte, J . (1988) . Amplification of human minisatellites by the polymerase chain reaction . Nucleic Acids Research 16, 1095371 .
