University of Massachusetts, Worcester

A biologist despairs over the difficulty of demonstrating heritability of chromatin states.

Chromatin, the packaged bundles of protein and DNA that make up eukaryotic genomes, is widely believed to be a carrier of 'epigenetic' inheritance — that is, heritable information not encoded by DNA. In multicellular organisms, the chromatin of mother and daughter cells is generally of similar shape, exposing similar regions of DNA for expression. And chromatin regulators often seem to be required for epigenetic states to be inherited. But there is a problem. It is possible that some other information carrier is inherited, and then directs chromatin regulators to re-establish a functional state.

One purported example of chromatin inheritance comes from yeast, which seem to 'remember' prior growth conditions. Galactose-naive yeast induce genes for Gal enzymes slowly; those whose recent ancestors experienced galactose induce them much faster. Because this 'memory' requires certain chromatin regulators, it has been suggested that it provided evidence for a heritable chromatin state.

Zacharioudakis et al. investigate this idea using heterokaryons, fused pairs of yeast cells that have mixed cytoplasmic contents but maintain separate nuclei. By inducing memory in one yeast and seeing speedy GAL1 expression in the other, they show that memory of galactose is transferable through cytoplasm (I. Zacharioudakis et al. Curr. Biol. 17, 2041–2046; 2007). Thus, the chromatin state around the GAL1 gene cannot be the heritable factor, and the authors further identify the probable inheritance factor as a soluble enzyme.

These results demonstrate the difficulty of proving that any example of epigenetic inheritance is due to inheritance of chromatin state per se. One wonders whether any chromatin state will ever be proved to be heritable, given the difficulty of proving the absence of another information carrier.