Comparison between CD4 and CD8 Lymphocytes as Controls for Methylation-Based Clonal Assay.

In somatic cells from normal female subjects, random inactivation of one of two X chromosomes occurs at an early stage of embryogenesis. Therefore, half of normal females possess an inactivated maternal X chromosome and half possess an inactivated paternal X chromosome. The human androgen receptor gene (HUMARA) contains highly polymorphic CAG trinucleotide repeats and several methyl-sensitive restriction enzyme (me-RE) sites. This polymorphism enables discrimination between paternal and maternal alleles; methylation status can easily be determined using me-RE because unmethylated (activated) sequences are sensitive to me-RE cutting while methylated (inactivated) sequences are resistant. Therefore, in clonal populations, one of the two HUMARA genes should be completely lost.

HUMARA-based assays have been applied to evaluate the clonal nature of various hematologic diseases. However, the main drawback of this method is the biased X chromosome inactivation, which is extremely skewed even in normal females. Consequently, HUMARA methylation patterns must be compared with adequate controls and clonal evaluation is impossible when both target and control samples show extremely skewed patterns. In most previous reports, granulocyte lineage clonality is assessed using T-lymphocytes or whole mononuclear cells as controls. To evaluate applicability of CD4 and CD8 cells as controls for clonal analysis, we examined X chromosome inactivation status in 22 healthy heterozygous females. Peripheral CD4⁺ and CD8⁺ lymphocytes were separated using the MACS system and DNA was extracted from both samples for PCR-based HUMARA assay. Although the 2 HUMARA peaks were almost identical in the HpaII pre-digested samples, there was skewing in some samples from elderly subjects after HpaII digestion: only 3 females in the CD4 group showed biased X inactivation, while 7 females in the CD8 group showed biased inactivation. These results indicate that CD8⁺ lymphocytes tend to more frequently show age-dependent skewing when compared with CD4⁺ lymphocytes.

Because it has been reported that the frequency of peripheral naive T lymphocytes is reduced in advanced age, we investigated the correlation between CD28 expression and age in 25 healthy controls (median age of 67), as the majority of CD28-positive cells were naive T cells. The proportion of the naive T cell population that was composed of CD8⁺ lymphocytes decreased markedly with age when compared with that of CD4⁺ cells.

Subsequently, we investigated age-related T cell receptor repertoire alterations. To evaluate the complexity of the TCR repertoire, the TCR Vβ repertoires of CD4⁺ and CD8⁺ lymphocytes were assessed by CDR3 size distribution analysis. The mean complexity score for control CD8⁺ lymphocytes (72.0) was significantly lower than that for CD4⁺ ymphocytes (126.5, P=0.011). In conclusion, frequent skewing of X chromosome inactivation was paralleled by severe reductions in naive T lymphocytes and T cell repertoire in CD8⁺ lymphocytes. Therefore, CD4⁺ lymphocytes might be better suited as controls for methylation-based clonal assay than CD8⁺ lymphocytes.