Sometimes, a proposed idea is so contrary to current scientific thinking that even though there is data to support it, most scientists remain skeptical. They accept it only after a very simple experiment is done.

In January 2000, Howard Hughes Medical Institute researchers Liming Li and Susan Lindquist at the University of Chicago performed such a set of experiments, providing the most definitive evidence yet that prions exist in yeast. [1]

For years the accepted dogma has been that all of the heritable information in an organism is contained in DNA - or in RNA, in the case of a few specialized viruses - but the data now show that some proteins can carry hereditary information as well.

And following Li and Lindquist's work seems to be a small avalanche of new data about fungal prions, including the identification of several new, putative prions and information about their regulation.

In 1994, Reed Wickner at the National Institutes of Health proposed that the two yeast phenotypes [URE3] and [PSI⁺] were prions, based on four genetic criteria. First, prions have a non-Mendelian mode of inheritance. Second, the phenotype can be cured reversibly. Third, prion formation is induced by overexpression of the normal protein. Finally, the phenotype of the prion resembles that of a mutation in the gene encoding the normal protein.

While Wickner says that "no one objected to this hypothesis," skeptics in the scientific community doubted the very existence of prions. After all, we normally think of DNA, and its constituent genes, as controlling the traits to be passed on to the next generation. But prions challenge this dogma in a manner that some apparently find unsettling.

In theory, the way prions work is very simple. A prion protein has two conformations, which can be thought of as an active conformation, frequently soluble within the cell, and an inactive form, which is protease resistant and found in insoluble aggregates. That a protein can have active and inactive states is not unusual in the least. What is surprising about prions is that once one protein in a cell takes on its inactive conformation, it influences the rest of the molecules of the same species to adopt a similar conformation and to aggregate together as an insoluble mass. This means that the conformation of one protein influences the rest of the proteins in the cell. Significantly, this influence is continued even after the cell undergoes division and new protein is made, or if the cytoplasm of two cells is mixed, as in mating.

Therefore, prions confer a heritable phenotype that is based on a change in protein conformation alone and has no underlying alteration in the DNA.

The Definitive Data

The most talked-about prion is the mammalian PrP protein, responsible for spongiform encephalitis - more commonly known as mad-cow disease in cattle, scrapie in sheep, and Creutzfeldt-Jakob disease in humans. But this example has remained a matter of debate among scientists because of the complexity of the experimental systems involved.

Hence the beauty of Li and Lindquist's experiments lies in their very simplicity.

In the first segment of their experiments, Li and Lindquist show that prion behavior is transferred to a distinct protein simply by fusing a prion-determining domain to it. They engineered a fusion protein containing the prion-determining domain (NM) of the Sup35p protein, from which the [PSI⁺] prion is formed, and a rat glucocorticoid receptor (GR). When the NMGR fusion protein was expressed at low levels in yeast, it activated a beta-galactosidase reporter gene, turning the cells blue. But when NMGR was overexpressed, it behaved like a prion, forming nonfunctional aggregates; the reporter gene was not activated and the cells remained white. When blue NMGR cells were mated to white NMGR cells, all the progeny were white, indicating that the prion behavior was transmitted to all the progeny in a non-Mendelian fashion.

Thus far, the experiments are interesting but not a definitive case for the existence of prions.

The really definitive experiment relies on a few standard yeast genetic tricks and demonstrates that the prion-like behavior continues even after the DNA plasmid encoding the fusion protein is removed from the cell, and therefore has to be the result of the protein itself.

Specifically, the researchers introduced two variants of the fusion protein, NMGR⁵²⁶ and NMGR, into two different cell lines of yeast. When white cells expressing a prion-like form of NMGR⁵²⁶ were mated with blue cells expressing a functional NMGR protein, all of the progeny were white - as in the experiment above. But, remarkably, when these daughter cells were allowed to lose the plasmid carrying the NMGR⁵²⁶ gene, they remained white. Therefore, with no change to the DNA of the NMGR cells, a stable, heritable change in phenotype has occurred due solely to the protein components of the cell. What better proof could one want of a protein-only form of inheritance?

And the Race Is On

That is only the beginning of what has happened in the field of yeast prions this year.

Several groups have reported the identification of new, putative prions. Two were identified based on sequence similarities to the prion-inducing domains of Sup35 and Ure2p, the protein responsible for the [URE3] prion. The best characterized of these newly identified prions is Rnq1, which stands for rich in asparagine(N) and glutamine(Q), in reference to the unusually high percentage of these amino acids in the prion forming domains. [2] Researchers have also identified putative prions in other yeast species by looking for homologues of the N domain of Sup35. These will be useful in exploring species barriers in the transmission of prions - an issue critical for understanding disease transmission in mammals. [3]

Another exciting development in the field is the identification of several regulators of prion formation. Wickner's group found that a protein that negatively regulates the normal Ure2p is required for the formation of [URE3]. [4] Similarly, Susan Liebman's group at the University of Illinois at Chicago showed that a recently identified prion, [PIN⁺], is required for the induction of [PSI⁺], though not for its maintenance. [5] Although these data don't explain the formation or regulation of prions, it is a significant step to understanding the process.

Why all this flurry now? According to Wickner, it isn't a sudden change in the field, but instead the fruition of years of effort. Liebman sees it a little differently. She thinks that now that there is clear evidence that prions exist in yeast, people are actively looking for them. "It used to be that people would find these bizarre things, briefly characterize them, and then stop working on them. Now they're taking out their bizarre things and looking at them again."

The Implications of Prions

The existence of prions is not only changing how biologists view issues of heredity; it is also critical for evolution itself. The phenotypes generated by prions are only semi-permanent or metastable. At some low frequency, the prions in a cell spontaneously switch from the off state to the on state, or vice versa. (Wickner estimates that the change occurs in approximately 1 in 10⁶ cells for [PSI⁺] and [URE3].) This is different from the way DNA works to control a cell. When changes occur in the DNA, they are generally permanent.

Therefore, by combining the effects of prions and DNA mutations, an organism has both permanent and semi-permanent ways to adapt to an environment over an evolutionary time scale.

Are prions likely to be widely found in biology? According to Lindquist, Wickner, and Liebman they are - but nucleic acids will still be the main agent of heredity. However, Lindquist speculates that since we now know how to look for them, we will find them all over the place.