Main

We have already reported that HLA allele matching is important in unrelated bone marrow transplantation (BMT).1 However, there are some patients who developed severe GVHD even in the case of HLA allele-matched transplantation. One of the reasons for GVHD development following HLA allele-matched transplantation is minor antigen mismatching. Some minor antigens have been identified from the peripheral blood of patients who developed GVHD after BMT. Only HA-12,3 was reported to be a minor antigen which statistically increases the probability of acute GVHD development. Minor antigens are MHC-binding peptides derived from polymorphic molecules. Therefore, there is a possibility that all polymorphic molecules become suppliers of minor antigens if the polymorphic region matches a peptide-binding motif of a certain HLA molecule.

It is well known that mitochondrial DNA (mtDNA) is highly polymorphic. mtDNA encodes subunit proteins, which are components of enzymes for the electron transfer pathway. A large number of substitutions have been identified not only in the non-coding region but also in the protein-coding region (http://www.gen.emory.edu/mitomap.html). In the murine system, minor antigens derived from mitochondrial polymorphisms are presented by H2-M3, class Ib molecule,4 and class Ia molecule.5,6 In the human system, it is not clear whether mitochondrial proteins function as minor antigens or not. If the polymorphisms of mitochondrial proteins are recognized as minor antigens, a large number of minor antigens will be supplied from the mitochondria. In this study, the association between mitochondorial polymorphism matching and clinical outcomes of HLA-allele-matched BMT was determined.

Materials and methods

Samples

HLA-A, -B, -DRB1 allele-matched pairs (340 pairs) were selected from cases of unrelated BMT performed by JMDP between February 1994 and January 1998. mtDNA with genomic DNA was prepared using a DNA Blood kit (Qiagen, Hilden, Germany). HLA-A, -B, -C and -DRB1 alleles were identified by the methods described previously.7 Briefly, PCR-SSO (sequence-specific oligonucleotide probes) (for HLA-A, -B, -C typing), PCR-SSP (sequence-specific primers) (for HLA-A26, Cw*08, 15 alleles), PCR-SSCP (single strand conformation polymorphism) (for HLA-B7, 15, 22, 37, 44, 59 alleles) and PCR-MPH (microtiter plate hybridization) (for HLA-DRB1 typing) were used.

Analysis of mtDNA polymorphism

Frequencies of mtDNA polymorphism were determined in DNA samples from 80 healthy Japanese donors by the PCR-SSCP method. Forty-five fragments of the entire 11-kb coding region of mtDNA were separately amplified using primer sets designed to amplify 300- to 400-bp fragments with 40–50 bp overlapping and whose positions are shown in Table 1. PCR amplification was performed for 30 cycles, each of which consisted of 20 s at 94°C, 20 s at 60°C, and 30 s at 72°C. After heat-denaturation, PCR products were analyzed using 10% polyacrylamide minigel. Electrophoresis was performed at 20 mA/gel for 2 h at 10°C. SSCP bands were visualized by silver staining (Daiichi Kagaku, Japan).

Table 1 PCR-SSCP analysis of mtDNA
Full size table

To determine nonsynonymous substitutions, the nucleotide sequences of fragments, which exhibit a different SSCP pattern, were determined directly.

Statistical analysis

Cumulative disease-free survival rates and cumulative incidence rates of acute GVHD (grade III–IV) of the recipients were estimated using the Kaplan–Meier method, and log-rank statistics was used to analyze differences.

Results

Polymorphism of mtDNA

Using samples from 80 normal Japanese donors, frequencies of polymorphisms were determined by the PCR-SSCP method (Table 1). Most of the PCR products showed multiple SSCP patterns. Based on the combination of SSCP patterns of 45 fragments covering the entire protein-coding regions, the 80 individuals can be divided into 59 types of SSCP-pattern combinations. Fifty individuals exhibited unique combinations. The most frequent combination was shared by eight individuals.

Polymorphisms with a relatively high frequency were found in ATP synthase 8 (MTATP8), NADH dehydrogenase 3 (MTND3) and cytochrome b (MTCYB). The nucleotide sequences of these fragments were determined (Table 2). Although polymorphisms in MTCYB at positions 15043 and 15301 caused synonymous substitutions, polymorphisms in MTATP8 and MTND3 caused amino acid substitutions. To detect the nonsynonymous polymorphisms in BMT samples, specific primers were designed (Table 3). The polymorphisms of HLA allele-matched BMT pairs were analyzed by the PCR-SSCP and the PCR-SSP (sequence-specific primer) methods (Table 2). The frequencies of polymorphisms of mtDNA of donors and recipients were about the same as that of the normal individuals.

Table 2 Polymorphisms of mtDNA with relatively high frequencies
Full size table
Table 3 Primers for detection of nonsynonymous substitutions
Full size table

Effects of mtDNA polymorphisms on clinical outcomes of BMT

The effects of MTATP8 and MTND3 polymorphism mismatching on 340 recipients transplanted with HLA-A, -B, -DRB1 allele-matched bone marrow were analyzed. The donor/recipient gender, donor parity, and the stage of malignancy and extent of total body irradiation were not different between the matched group and the mismatched group (data not shown). Table 4 shows the effect of mtDNA mismatching on cumulative disease-free survival and the incidence rate of GVHD of more than grade III. No significant effect was observed. The peptides derived from MTATP8, MTATP8-L: TMITPMLLT, MTATP8-F: TMITPMFLT, and from MTND3, MTND3-A: WLQKGLDWA, MTND3-T: WLQKGLDWT, have peptide-binding motifs for HLA-A*0201.8 The allele frequency of A*0201 is about 10% in the Japanese population.7 The effects of mtDNA mismatching on HLA-allele matched transplantation having the HLA-A*0201 were analyzed; however, no mismatching effects were observed. Although transplantations having frequent alleles, A*1101 (10%), A*2402 (48%), A*3303 (10%), B*4403 (10%), B*5201 (22%), were also analyzed, the mtDNA mismatches did not affect the outcomes. About 25% of the HLA-A, -B, DRB1 allele-matched pairs have incompatibility in HLA-C alleles. Again, effect of mtDNA mismatching was observed in analysis using HLA-A, -B, -C, -DRB1 allele-matched pairs (data not shown).

Table 4 Effect of mtDNA polymorphism compatability on incidence of acute GVHD and on disease-free survival
Full size table

Discussion

A large number of polymorphisms of mtDNA have been reported. However, the frequencies and associations among the polymorphisms are not well known. In this study, polymorphisms of mtDNA in the entire protein-coding region were analyzed by the PCR-SSCP method. The 80 individuals were divided into 59 types of SSCP-pattern combinations. However, the SSCP method cannot distinguish all polymorphisms. Determining the polymorphisms of the entire protein-coding region of mtDNA at the nucleotide sequence level may reveal that all individuals have unique combinations of substitutions.

Nonsynonymous polymorphisms of mtDNA with a relatively high frequency were identified and the effect of the mismatching on the outcome of unrelated bone marrow transplantation was studied. All nonsynonymous polymorphic molecules may become minor antigens. In this study, however, polymorphisms with a relatively high frequency were selected for statistical analysis. Substitutions at nucleotide position 8414 in MTATP8 and at 10398, 10400 in MTND3 have already been reported. In this study, the frequencies of these polymorphisms in the Japanese population were determined.

It has been reported that HLA-A, -B1 and -DRB16 allele mismatching are strong risk factors for acute GVHD and mortality in unrelated BMT. Therefore, to clarify the effects of mtDNA polymorphisms on outcomes of BMT, HLA-A, -B, -DRB1 allele-matched transplantation were selected for the analysis. The peptides derived from polymorphisms of mtDNA determined in this study are presumed to bind to HLA-A*0201.8 However, the mismatching did not affect the cumulative disease-free survival and the incidence rate of GVHD. Effect of HLA-C allele mismatching on BMT outcome is open to discussion. No effect of mtDNA polymorphisms was observed in HLA-A, -B, -C, -DRB1 allele-matched transplantation.

There are several explanations for the polymorphic molecules having peptide-binding motifs not functioning as minor antigens. (1) Peptides having binding motifs may not have been generated. The site of digestion by proteasome is reported to be dependent on the amino acid sequence.9 (2) Although it is not clear how many minor antigens affect clinical outcomes of BMT, the number is presumed not to be small. Therefore, the matching effect of one or two minor antigens may not be detected in statistical analysis. (3) Some polymorphic molecules including those found in this study generate both binding peptides with and without substitutions. In fact, the difference between the original peptide- and substituted peptide-HLA complex is only one amino acid. The number of T cells that can recognize only a single amino acid difference may not be sufficient to develop GVHD. Some minor antigens have been identified.10,11,12,13,14 However, most of the counterpart peptides are not known and most reports did not mention them. Counterpart peptides HA-110 and H-Y12 derived from the SMCY gene have been known. Goulmy et al2 investigated whether mismatching of minor histocompatibility antigens, HA-1, -2, -3, -4 and -5, contributes to acute GVHD development in recipients of genotypically HLA-matched BMT. These minor antigens were identified in recipients who developed acute GVHD. However, they reported that only HA-1 antigen mismatching was significantly correlated with GVHD development. Mismatching of the other antigens did not increase the incidence of acute GVHD development. HA-1 is a dimorphic antigen, HA-1H and HA-1R, differing in only one amino acid. The HA-1H peptide binds to the restriction molecule HLA-A2, but HA-1R does not.10 They estimated the effect of H-Y antigens by comparing donor's and recipient's gender combinations, and concluded that the mismatching of H-Y antigens has no effect on the incidence of acute GVHD development. We have obtained the same results (data not shown). The counterpart of H-Y antigen is a peptide derived from SMCX gene. Both peptides bind HLA-B7 molecules.12

Our results cannot exclude the possibility that a polymorphic peptide is recognized by T cells as a minor antigen leading to GVHD development. However, taking our results and other reports into consideration, peptides derived from polymorphic molecules hardly develop GVHD if the counterparts also bind to the same restriction molecules. In the research of minor antigens, not only the peptide but also its counterpart should be identified and considered.