Detection of maternal DNA in placental/umbilical cord blood by locus-specific amplification of the noninherited maternal HLA gene.

A critical issue regarding the broader utilization of placental/ umbilical cord blood (PCB) in unrelated bone marrow restoration is the possibility of contamination with maternal lymphocytes capable of immunological reactivity against the eventual recipient. On transplantation, such maternal cells might lead to graft-versus-host disease (GVHD) even if the intended donor's neonatal lymphocytes were unresponsive. We measured the proportion of PCB samples that were contaminated with maternal cells. Placental-maternal sample pairs were selected so that the mother was heterozygous for the DR53 haplotype, whereas the placental sample was DR53-negative. The PCB samples were investigated for the presence of the noninherited maternal gene DRB4, exclusive to the DR53 haplotypes. Locus-specific polymerase chain reaction amplification with DRB4 sequence-specific primers was followed by either gel electrophoresis or blotting and hybridization to an internal sequence DRB4 probe. Polymerase chain reaction products from DNA mixtures containing as low as 0.5 ng of a DRB4-positive DNA control in 1.0 microgram of a DRB4-negative DNA sample (1:2 x 10(3) dilution) showed a visible DRB4 band in agarose gels stained with ethidium bromide. Locus-specific hybridization increased the detection sensitivity to 1:10(5) (0.01 ng of the DRB4-positive DNA control). Control mixtures of known amounts of DRB4-positive and -negative DNA were included in all experiments. Comparison of the thickness of DRB4 bands after electrophoresis and the intensity of the DRB4-specific hybridization signals to the concentration controls allowed a rough estimation of the amount of maternal DNA in the placental blood specimens. A total of 213 PCB samples were tested. By gel electrophoresis, DRB4-specific bands were observed to be as strong or stronger in 23 (10.8%) samples as those in the 1:2 x 10(3) control, and 153 (17.8%) samples were negative in this test. The remaining 37 (17.3%) samples disclosed weaker DRB4 bands, suggesting the presence of maternal genetic material. By hybridization, 81 (38%) samples were positive and 132 were negative for the noninherited maternal gene. Review of the clinical characteristics of the mothers (demographics and labor and delivery information), the newborns (birth weight, sex, and gestational age), and PCB collections (placental weight, white blood cell count, and collected volume) failed to show any significant differences between the units testing positive or negative for the noninherited maternal gene. Thus, transplantable PCB units carry a high probability of having maternal DNA in detectable amounts. Whether this DNA comes from potentially graft-versus-host disease-inducing maternal lymphocytes or whether the putatively transplacentally-acquired maternal cells are immunologically dysfunctional, as in most infants with severe combined immunodeficiency disease, remains to be shown.