To determine the efficiency of reprogramming in cell fusion experiments, the most informative are those that result in heterokaryons. Efficiency can be assessed as the proportion of selected heterokaryons (1–2% of total fusions attempted) that express pluripotency genes, such as Oct4. Transcription of such genes can be detected in 70% of the heterokaryons (mouse ES and human fibroblasts) within one day [9], although the level of this expression is likely to be low (i.e. approximately 1% of the expression of these genes in ES cells) [18]. When one donor cell is highly differentiated, a lower proportion of heterokaryons activate some of the genes that are not expressed in the starting somatic cells [19]. For example, the proportion of heterokaryons that are induced to express the human muscle gene 5.1H11 6 days following fusion with mouse muscle cells is 95% for human lung fibroblasts, 60% for human keratinocytes and 25% for human hepatocytes (Figure 2d) [19]. We conclude that, in heterokaryons, as in nuclear transfers, nuclei from the most specialised cells are much more resistant to reprogramming than those of less specialised cells.
