Identifying the origin of single corneal cells by DNA fingerprinting: part II-- application to limbal allografting

Successful limbal allotransplantation allows regression of limbal stem cell deficiency features. Transplant survival is presumed if clinical improvement occurs. However, positive proof of surviving transplanted stem cells remains difficult. This follow-up study attempted to prove donor cell survival 5 years after limbal stem cell allograft in one woman with aniridia.

Impression cytology and single-cell DNA fingerprinting were used to investigate a previously studied patient. Corneal epithelial cells were harvested from five sites and isolated by micromanipulation. Polymerase chain reaction and short tandem repeat profiling were used to obtain forensic standard "DNA fingerprints" from single cells. (The technique is described in the preceding article, Part I.) Blood samples yielded host and donor DNA for comparison. Negative controls were performed for impression cytology and polymerase chain reaction. Simultaneous micro-scrape samples were also taken.

Impression cytology samples permitted informative DNA fingerprints from all corneal sites and represented 76% (23/30) of tested cells. Fifty percent (15/30) of the fingerprints were "specific" but 83% (19/23) matched the host DNA fingerprint. The remaining 17% (4/23) represented contamination from various sources. Specific fingerprints were obtained in 55% (10/18) of the cells from micro-scrape samples. All samples giving sufficient information matched the host DNA fingerprint. All tested blood samples gave specific fingerprints. None of the sampled corneal cells gave a donor DNA fingerprint.

In a single patient, no detectable long-term donor cell survival exists at 5 years. Positive identification would have provided unequivocal proof of donor cell survival. This technique gives useful information even if contamination occurs.