A fungal gene reinforces Mendel's laws by counteracting genetic cheating

Cheating is in. Everybody is doing it: high-school students, businessmen, and tour de France racers. Every so often, individuals break or bend collective rules to gain an advantage in a competitive situation. There are also rules in the competition of genes and organisms. In sexually reproducing organisms, one central law is Mendel’s equal transmission of the two versions of a heterozygous gene pair to the next generation. For any given species, it is paramount that this rule be scrupulously respected. Only then can the Darwinian game of selection be properly played. Genetic mechanisms that allow a gene to act selfish and get more than its fair share in the next generation are collectively designated meiotic drive. Just as cheaters threaten the equilibrium of human societies, genetic cheaters can have deleterious effects on the fitness of organisms. Indeed, meiotic drive allows genes to prosper regardless of their advantage or disadvantage to the fitness of the species. Meiotic drive systems have been described in insects, mammals, fungi, and plants. In PNAS, Hammond et al. identify a fungal gene that prevents genetic cheating by resisting meiotic drive (1) and thus provide unique insights into the mechanism of meiotic drive in fungi.

Mendelian genetics are based on the postulate that during meiosis, for any given heterozygous gene (or chromosome) pair, equal amounts of functional gametes of each type are produced. But there are exceptional cases in which gametes of a particular type are produced in excess (2), a process termed meiotic drive. Linked mutations can hitchhike with the meiotic drive gene and spread in populations, even at a fitness cost. Meiotic drive elements can reside on autosomes or sex chromosomes. In the latter case, segregation distortion skews the sex ratio and this type of drive can potentially lead species to extinction. The mechanism of meiotic drive is different in males and females. In male meiosis, the meiotic drive gene (or chromosome) gains an advantage over the sensitive allele, because the sperm carrying the sensitive allele ends up nonfunctional. Two of the most studied meiotic drive elements, the SD segregation distortion locus in Drosophila and the t-haplotype in mice, operate that way. The SD locus of Drosophila is one of the very few meiotic drive systems for which a mechanistic model for drive could be proposed (3, 4). This system comprises the Sd gene responsible for the drive and a responder element, composed of noncoding DNA repeats and termed Rsp that is the target of Sd. Drive-sensitive and -insensitive alleles of Rsp differ by the number of repeat units, with many repeats in Rsp^s alleles and very few in Rspⁱ. The Sd driver gene encodes a truncated RanGAP protein. It is proposed that this truncated RanGAP disrupts the nuclear trafficking of small RNAs controlling RNAi-mediated heterochromatin formation at the Rsp locus. The Rsp^s nuclei, due to their large repeat array, would be particularly sensitive to this defect and would therefore fail to complete the chromatin compaction process required for proper maturation into functional gametes. In females, meiosis is asymmetric: only one of the four meiotic products is entitled to a genetic future in the egg. The three others are doomed to evolutionary death as polar bodies. Here for the driving gene (or chromosome), the foul play consists in gaining a transmission advantage to access the egg. It is proposed that centromeres can behave as female drive elements and that this centromere drive is the cause of the rapid evolution and complexity of centromeres (5, 6). In this view, centromeres compete for optimal microtubule attachment to gain access to the egg cell. This competition causes multiple cycles of runaway expansion of centromere satellite repeats and consecutive drive suppression by rapid evolution of kinetochore proteins. Apparently, male meiosis drive systems in flies and female “centromere-drive” systems involve what could be loosely defined as modifications in chromosome structure (chromatin condensation and centromere/kinetochore structure). Another example of meiotic drive operating both in males and females is GC-biased gene conversion (7). During meiosis, at heterozygous sites, a heteroduplex can form during recombination. Such mismatches are corrected but repair is biased and favors GC over AT alleles. This GC-biased gene conversion is thought to explain the correlation between GC content and recombination rates across genomes. This short enumeration should emphasize that meiotic drive, because it violates Mendelian rules and interferes with Darwinian selection, can have a profound impact on the fitness and evolution of sexually reproducing species. Not only does it allow for spreading of deleterious mutations but also it appears to shape the very structure and function of chromosomes and of the segregation machinery. To go back to an anthropomorphic analogy, just as cheaters impact human affairs by inducing multiplications of countermeasures to cheating, genetic cheaters alter the structure of the genetic machinery of the organisms hosting them.

In fungi, meiosis is zygotic as opposed to gametic. Instead of producing gametes (as in animals), meiosis yields progeny. Meiosis occurs in the zygote immediately after karyogamy and leads to the formation of haploid spores. Spores germinate and divide mitotically to produce haploid individuals (which are genetically identical to the gametes they will eventually produce). Meiotic drive thus takes a different form in these organisms. Fungal meiotic drive elements are detected as spore killers (8). In Neurospora crassa, a sexual cross leads to the formation of spore sacs (asci), each containing eight haploid spores. In a cross between a killer and a sensitive strain, all progeny harboring the sensitive nuclei die; only the killer nuclei survive (Fig. 1). Some call this process a fratricide. Several such spore-killer genes have been described, among them, spore-killer 2 (Sk-2). The Sk-2^k allele kills all meiotic products bearing the sensitive Sk-2^s allele (9). Strikingly, sensitive Sk-2^s nuclei escape killing when they are enclosed (by a genetic trick) in the same spore as Sk-2^k killer nuclei. Sk-2^k maps to a region that can be defined only as a 30-map-unit interval because recombination is blocked between spore killers and sensitive strains in that region. This large recombination block region around the spore killer constitutes a major hurdle for the genetic dissection of the locus. A second spore killer termed Sk-3 behaves similarly to Sk-2 and maps to the same recombination block region. Sk-2 and Sk-3 kill each other; in a Sk-2^k × Sk-3^k cross, 99.9% of the spores are inviable. Some natural strains are nonkiller but yet are resistant to killing. Some resist Sk-2^k or Sk-3^k and some resist both. Strains resistant to Sk-2^k bear a resistance gene termed r(Sk-2) that maps to the recombination block region. Importantly, recombination is not blocked between sensitive and resistant strains in that region, allowing for straightforward mapping of the r(Sk-2) gene. Hammond et al. take advantage of this convenient backdoor to gain entrance into the molecular structure of this spore-killer system. The authors carried out mapping of r(Sk-2) and pinpoint the gene to an ∼10-kb region containing five candidate ORFs, one of which was found to confer resistance to Sk-2^k. This gene (rsk) encodes a lineage-specific 486-aa protein with no recognizable motifs. When this gene is deleted in a Sk-2^k background, spore formation is totally eliminated (the killer kills itself), indicating that the killer uses the RSK protein for self-resistance. Rsk also ensures resistance to Sk-3^k. Rsk alleles from various wild isolates that confer resistance to Sk-2 are closely related in sequence but distinct from alleles that confer resistance to Sk-3 or sensitive alleles. The level of divergence between the different types of alleles is substantial with as low

Fig. 1.

Meiotic drive caused by the Sk-2 spore-killer gene in Neurospora. In Neurospora, the meiotic progeny obtained after a sexual cross form bunches of eight-spored asci (spore sacs). Asci are eight-spored due to a postmeiotic mitotic division that makes eight haploid spores of the four meiotic products. When a Sk2^k killer strain is crossed with a sensitive Sk-2^s strain, all spores that do not contain the Sk2^k killer gene fail to mature and die, yielding asci with four unviable white Sk-2^s spores and four normal black Sk-2^k spores. Homozygous crosses between killer strains or sensitive strains yield normal asci.

Hammond et al. identify a fungal gene that prevents genetic cheating by resisting meiotic drive.

as ∼60% identity between alleles, which could suggest that rsk alleles are rapidly evolving. Meiotic drive elements are often understood as two-component systems with a driver product and a responder locus that is the target of the driver. For Sk-2, the authors propose an alternate plausible model resembling a poison-antidote system: The (yet to be identified) Sk-2^k killer factor would disrupt ascospore formation unless it is shielded by the RSK resistance factor. When ascospores become delimitated, the sensitive nuclei are no longer protected by RSK and further ascospore development is impeded. This study also makes clear that the process targeted by the killer factor is specific to the meiotic program as coexistence of Sk-2^k and sensitive nuclei in somatic cells has no adverse effect, even in the absence of the shielding RSK factor.

There is another fungal meiotic drive system that has been studied at the molecular level, the het-s gene of Podospora anserina (10). Het-s encodes a prion. Driving occurs when the product of the het-s gene is in its prion conformation and leads to specific abortion of the spore carrying the alternate het-S allele. Apparently, the mechanism of spore killing differs in these two fungal systems. The product of the driving gene (the prion form of HET-s) is not toxic per se but turns HET-S into a toxic protein. Thus, there is no requirement for self-resistance to prevent the driver from killing itself as seen in Sk-2.

In addition to the question of the killing mechanism per se, this study on Neurospora raises many questions on the evolutionary dynamics of this spore-killing system in natural populations. Is there a cost to be a killer? What is the impact of the resistant strains? How efficient are they in preventing spreading of the killer?

The approach Hammond et al. took to identify components of the Neurospora spore-killer systems led them to identify the antidote before the poison. However, the authors have laid a foot on the Land of Nod (where Cain was exiled after his misdeed) and are now in a position to track down the culprit of the Neurospora spore fratricide. Much is to be gained from the comparative study of meiotic drive systems operating in gametic and zygotic meiosis. Although these systems are probably mechanistically very different, they can enlighten us on the common consequences of the existence of Mendelian outlaws.