Nested Polymerase Chain Reaction With Sequence-Specific Primers Typing for HLA-A, -B, and -C Alleles: Detection of Microchimerism in DR-Matched Individuals

IT IS WIDELY ACCEPTED that donor leukocytes can survive within a recipient after transplantation1-4 or blood transfusion,5-8 yet the relevance of this microchimerism remains a contentious issue.4 9-13 Further controlled studies are required to achieve an improved understanding of the functional importance of microchimerism.

The majority of assays used to detect donor-derived material in recipient blood exploit sex mismatching between donor and recipient1,14-16 or the polymorphism of the HLA-DR region of the major histocompatibility complex (MHC).2,4,9,10,17 The most commonly used techniques based on HLA-DR polymorphism are the polymerase chain reaction with either sequence-specific primers (PCR-SSP) or sequence-specific oligonucleotides (PCR-SSO). These methods are commonly used to HLA type donors and recipients before transplantation,18,19 but may be applied to the detection of microchimerism. The sensitivity of detection of such techniques is variable and can range from 0.5% to 0.01%,3,20,21 but the sensitivity of PCR-SSP for HLA-DR alleles has been increased by the introduction of a preliminary PCR (nested PCR-SSP), amplifying exon 2 of HLA-DRB1.4,17,22,23 It has been demonstrated that the sensitivity of nested PCR-SSP typing for HLA-DR alleles is at least 100-fold higher than that of standard PCR-SSP typing.17,22Nevertheless, the use of nested PCR-SSP typing for HLA-DR alleles is not applicable when the donor and recipient are HLA-DR matched. In such cases, it would be useful to be able to increase the power of detection of the technique by applying the recently developed molecular typing methods for HLA class I alleles.19 24-26 

We describe here the development of a nested PCR-SSP typing method for the detection of HLA class I alleles. During the development of this method, we detected bands on the gel that were not indicative of donor or recipient HLA type, entirely consistent with our previous findings when developing a nested PCR-SSP typing system for HLA-DR alleles.22 These nonspecific bands do not arise from contamination, but result from the nonspecific amplification of recipient DNA with certain primer sets. The application of the technique to clinical samples is described, emphasizing the importance of rigorous controls for this type of assay system.

Anticoagulated peripheral blood was obtained from selected HLA-typed healthy volunteers for the sensitivity experiments. To validate the technique, peripheral blood samples were obtained from patients at the Oxford Transplant Centre (Oxford, UK) awaiting a renal transplant who had received planned transfusions of HLA-typed blood. Each patient received 50 mL of freshly isolated blood (<36 hours from donation) from 2 healthy donors. Blood samples were obtained at 2 time points before the transfusion and at defined time points after transfusion.

Genomic DNA was isolated from leukocytes obtained from anticoagulated blood using a salting out procedure,27 was precipitated with ethanol, and was resuspended in sterile water. The genomic DNA was quantitated and purity was assessed by spectroscopic absorbance at 260 and 280 nm.

Two hundred nanograms of DNA was initially amplified in a buffer containing 67 mmol/L Tris, pH 8.8; 16.6 mmol/L NH₄SO₄; 200 μmol/L of each dATP, dCTP, dGTP, and dTTP; 1.0 mmol/L MgCl₂; 0.5 μmol/L forward and reverse primer (Table 1); and 0.25 U BioTaq polymerase (Bioline, London, UK) for 30 cycles according to Cereb et al28 in a PTC200-96v thermal cycler (Genetic Research Instrumentation, Essex, UK). The resultant PCR product was diluted 1:500 in water before PCR-SSP typing. Extreme care was taken to avoid contamination at any stage.29 Filter tips were used when pipetting (BioExpress UK Ltd, London, UK), and the work was performed in a clean air cabinet (Heto-Holten UK Ltd, Camberley, UK).

Allele-specific primers (0.5 μmol/L) designed on the basis of published sequences19,30 (Table2) were used in multiple amplification reactions consisting of diluted first-round PCR product (1/500); 67 mmol/L Tris, pH 8.8; 16.6 mmol/L NH₄SO₄; 200 μmol/L of each dATP, dCTP, dGTP, and dTTP; 2.0 mmol/L MgCl₂; and 0.25 U BioTaq polymerase. PCR amplifications were performed in a PTC200-96v thermal cycler, and the products were subsequently analyzed by agarose gel electrophoresis using the conditions exactly as previously described.19 

Primer mixes 151 (A*0101-4N), 9 (A*2501-2, A*2601-11N, A*3401-2, A*6601-3), 155 (B*0801-3), 47 (B*1301-3), 86 (Cw*0102-3), and 106 (Cw*1502-6) (Table 2) were randomly selected to determine the sensitivity of nested PCR typing. The experiments were performed as previously described and repeated on three occasions.22Briefly, decreasing amounts of DNA from selected HLA-typed healthy volunteers were mixed with DNA of known HLA type, both at a starting concentration of 200 ng/μL, to give final concentrations of 1%, 0.1%, 0.01%, and 0.001% (vol/vol). The mixtures were then subjected to nested PCR-SSP typing. All products were subsequently analyzed by agarose gel electrophoresis.

The sensitivity of detection of 6 of the HLA class I primer mixes (151, 9, 155, 47, 86, and 106; see Table 2) was determined by performing mixing experiments, and the results are shown in Fig 1. Each primer mix was capable of reproducibly detecting DNA to a level of 0.001% (equivalent to a dilution of DNA of 1:100,000), demonstrating that the sensitivity of this technique is comparable to that of other nested PCR-SSP techniques.17 22 

Samples were investigated from a patient who had received planned HLA-typed blood transfusions mismatched for HLA class I alleles. Samples obtained before and 2 days after blood transfusion were compared (Fig 2). All recipient alleles were detected in both samples (arrowed) and, in the posttransfusion samples, additional bands corresponding to the HLA alleles of both blood transfusion donors were evident (asterisked). However, nonspecific products also appeared in certain lanes on the gels, a finding consistent with our results of nested PCR-SSP typing for HLA-DR.22 The bands were not compatible with the known HLA type of the donors or the recipient. The pattern of bands generated by the nested PCR-SSP typing of the pretransfusion sample remains consistent in the posttransfusion samples and, thus, by inference, is recipient HLA-type dependent.

To control for the presence of the nonspecific products resulting from mispriming, a specificity control in the form of a pretransfusion sample is required. Rigorous analysis of this sample establishes a baseline for the analysis of posttransfusion samples. Therefore, nested PCR-SSP typing is routinely performed 5 times on 2 samples obtained before transfusion and on samples after transfusion. The results from one such analysis of the HLA-A locus are shown in Fig 3.

The recipient HLA type (A*0301, A*2402) is present in all repeats of the pretransfusion and posttransfusion samples (PM 4 and 5; dark shading). Potentially informative reactions are those reactions that do not yield a band in the pretransfusion samples (in this case, primer mixes 151, 9, 12, 152, 18, 184, and 154). The blood donors were typed as HLA-A*0101 (donor 1) and HLA-A*2501, A*3002 (donor 2), and these alleles are recognized by primer mixes 151 (A*0101), 9 (A*2501), 10 (A*2501), and 18 (A*3002). All of these primer mixes gave rise to a band when the posttransfusion sample was analyzed, but in the case of primer mix 10, a band was also evident in both pretransfusion samples. The reaction with primer mix 10 is therefore termed a noninformative reaction. Furthermore, there were other primer mixes that gave rise to bands in the pretransfusion samples that were also present in posttransfusion samples (PM 3, 6, 13, 14, 15, 153, and 24). The presence of these nonspecific products is dependent on the recipient HLA type.

In this report, we have described a method for the detection of HLA class I donor alleles after blood transfusion using the nested PCR-SSP technology previously developed for HLA-DR alleles.17,22,31The technique is finely tuned to achieve maximum sensitivity while retaining specificity; for example, increased sensitivity can be achieved by increasing the amount of DNA in the first-round amplification, but such a modification will result in a decrease in specificity. The sensitivity of the technique has been demonstrated using mixing experiments and DNA can be detected to a level of at least 0.001% (equivalent to a dilution of 1:100,000), comparable to the techniques developed for the detection of HLA-DR alleles.17 22 

The high level of sensitivity achieved with the technique leads to a concomitant decrease in specificity. When the nested PCR-SSP typing was applied to the detection of donor-type microchimerism in a patient receiving an HLA-mismatched blood transfusion, nonspecific products appeared that did not correspond to the donor or recipient HLA type. This finding is entirely consistent with our previous results obtained while developing a similar system for the detection of HLA-DR alleles. In the previous study, we sequenced several of these nonspecific products and determined that they did not result from contamination but rather from mispriming events that led to amplification of recipient-derived HLA alleles or associated pseudogenes.22It is therefore possible that some of the nonspecific products seen in the HLA class I nested PCR-SSP typing system result from the amplification of HLA class I-associated pseudogenes, of which 6 have been currently identified.32 33 Therefore, to use nested PCR-SSP typing to analyze clinical samples, a means of carefully controlling the system is required.

In the system we have devised, a pretransfusion blood sample is analyzed to act as a baseline and specificity control. Multiple analysis of this sample with nested PCR-SSP typing allows a rigorous baseline to be established for all subsequent analyses (Figs 2 and 3). If such a sample is not included in the analysis, then amplification appearing in posttransfusion samples may be wrongly assumed to be indicative of donor alleles. For example, in Fig 2, primer mix 10, which amplifies the donor allele HLA-A*2501, yields a band in the posttransfusion sample suggesting the presence of microchimerism. However, this band is also detected in the pretransfusion sample; therefore, the amplification in the posttransfusion samples is noninformative. The requirement for a pretransfusion/transplant sample has also been shown by Elwood et al,4 who, by using nested PCR-SSP typing for HLA-DR alleles, demonstrated that false-positive results for microchimerism would have been obtained in 17% of patients had a recipient pretransplant DNA sample not been included in the analysis.

The inclusion of HLA-A, -B, and -C alleles in the nested PCR-SSP system significantly increases the power of this type of analysis where donors and recipients are frequently matched for HLA-DR alleles. The technique is flexible and widely applicable to the detection of microchimerism after blood transfusion or solid organ transplantation and may provide the means for understanding the true relevance of microchimerism.