Main

Sgaramella and Zinder2 recently asserted that sheep are “highly inbred”, and they queried whether the microsatellite evidence that we originally provided1 was sufficiently robust to exclude the possibility that Dolly was derived from embryonic or fetal cells from a different animal. They also suggested that, because the ewe was pregnant, Dolly might have been derived from a stray fetal cell contaminating the mammary tissue rather than from a mammary cell itself.

This hypothesis seems improbable as there are few fetal cells in the circulation of pregnant women (from 1 in 105 to 1 in 109 maternal cells3) and there is a much less intimate relationship between maternal and fetal circulation in sheep than in humans4. We think it highly unlikely that the partially purified mammary cells could have been overgrown by contaminating fetal cells during their relatively short time in culture5.

Sheep populations are usually outbred rather than inbred and the five microsatellites used in the previous analysis (data from four of which were published1) were chosen because they are polymorphic in sheep6. However, in the absence of information about allele frequencies in the specific population from which the 6-year-old ewe was taken, we could not estimate how useful these markers are in discriminating between a fetal or maternal origin for Dolly. Thus we carried out the more detailed analysis described here.

The cells used to produce Dolly were prepared from mammary tissue taken from a 6-year-old ewe as part of a separate collaboration between the Hannah Research Institute in Scotland and PPL Therapeutics. A sample of the tissue that had been stored, frozen, at the Hannah Research Institute since the ewe was killed in 1995, and DNA from the original cell populations, was provided by C. Wilde for this analysis. We calculated the frequency of alleles in the Finn Dorset flock at the Hannah Research Institute from data from 44 individuals drawn from two separate generations but with only one representative from each full-sib family. Blood samples were delivered to Rosgen, a specialist genotyping company set up by the Roslin Institute and accredited to the quality standard ISO9002. DNA was extracted using standard protocols.

Microsatellite amplification was done using three of the primer sets used previously and seven proprietary markers from Perkin Elmer, and the products of DNA amplification by the polymerase chain reaction were analysed using an ABI 377 Prism Sequencer. The markers from Perkin Elmer were originally developed for parentage testing in cattle but had also been shown to be polymorphic in sheep. The presence or absence of each allele was determined in DNA extracted from blood from Dolly and the other Finn Dorset animals, from cells from the original mammary tissue, and from cells at passage four (Dolly was derived from cells at passage three).

The data on allele frequencies are summarized in Table 1. All but one of the ten microsatellite markers were polymorphic, with from four to nine alleles present per marker. The alleles present in DNA from Dolly were identical to those in the original mammary tissue, in the cell population prepared from that tissue and in the cells cultured to passage four. We estimated the probability that another sheep from the same population would have the same genotype as the 6-year-old ewe to be between 1.9×10−12 and 2.7×10−10 (95% confidence interval). We conclude that it is extraordinarily unlikely that Dolly was derived from a different Finn Dorset animal and, therefore, reject the hypothesis that “imagined and unimagined experimental error”2 occurred.

Table 1 Microsatellite analysis of Dolly

If Dolly were derived from a fetal cell, she would have derived half of her alleles from the sire of the fetus and half from the 6-year-old ewe. We calculated the chance of a fetal cell having the same genotype as the 6-year-old ewe to be between 1.1×10−6 and 9.2×10−6 (95% confidence interval). We conclude that Dolly was derived from a mammary cell of the 6-year-old donor ewe.