Barry Ganetzky. American Scientist. Volume 88, Issue 2. Mar/Apr 2000.
In 1956, Yuichiro Hiraizumi, a graduate student in the laboratory of James F. Crow at the University of Wisconsin, made a remarkable discovery that contradicted a basic tenet of genetics-the principle that each chromosome of a pair has an equal chance of being passed on to the next generation. Hiraizumi was carrying out genetic studies of natural populations of the fruit fly Drosophila melanogaster, with the nominal goal of investigating genes affecting viability.
In these experiments, Hiraizumi crossed white-eyed females with redeyed males. The females were from an inbred laboratory strain, and both members of the relevant chromosome pair carried certain mutations that produced the white eye color. The members of the corresponding chromosome pair in the males were dissimilar. One chromosome was from the laboratory strain carrying the mutations for white eyes. The other chromosome of the pair was from wild-caught flies and carried the genes for normal red eye color. These eye-color genes merely served as convenient genetic markers enabling Hiraizumi to trace the transmission of each of the two chromosomes from the males to the next generation.
According to the principles of inheritance, roughly half of the offspring from these matings should have been red-eyed, and the other half whiteeyed. That’s not what happened. Quite unexpectedly, a few of the mating pairs produced only red-eyed offspring. These crosses flagrantly violated genetic law: The chromosome carrying the red-eye genes (the one derived from nature) was transmitted preferentially to the offspring, whereas the other member of the chromosome pair, the one carrying the white-eye mutations, seemed not to be transmitted at all.
The name Segregation Distorter (SD) was given to chromosomes that display this unusual pattern of transmission, and geneticists now know that roughly 3 to 5 percent of nearly every natural population of Drosophila melanogaster harbors SD chromosomes.
But the very notion that such transmission distortions can take place is disturbing to anyone who considers questions of evolution and natural selection. In theory, evolution by natural selection is a rigorous process that favors the retention of genes that enhance the ability of organisms to survive and reproduce. Chromosomes and the genes they carry are supposed to be meted out equally into eggs and sperm through the specialized cell divisions called meiosis. Proper meiosis ensures competing genes equal representation in the gametes and thus guarantees that each gene is exposed equally to the forces of selection.
A particular gene that figured out a way to beat the system by ending up in the vast majority of functional gametes would have an enormous but unfair advantage over competing genes. Such cheating genes would tend to increase in a population even if they conferred no selective advantage-or, indeed, were harmful to the organisms in which they were present. In principle, a situation like this could lead to the extinction of a population.
The potent impact such genes could have on natural populations was first pointed out in 1957 in a theoretical paper by Laurence M. Sandler and Edward Novitski of Oak Ridge National Laboratory. They coined the phrase “meiotic drive” to refer to any alteration of meiosis that resulted in excess transmission of one genetic variant over its alternative. The definition has now been expanded to include any alteration in meiosis or the subsequent production of gametes that results in preferential transmission of a particular genetic variant.
Examples of meiotic drive have now been discovered in a wide variety of organisms, including fungi, higher plants, insects and mammals. The mechanisms by which these various meiotic-drive systems operate remain almost a complete mystery However, members of my laboratory have recently made substantial progress in unraveling one of them-the SD system-the very one discovered by Hiraizumi more than 40 years ago.
Hiraizumi wisely forgot about the original purpose of his experiments and focused instead on the analysis of the serendipitously discovered SD chromosomes. By chance, it happened that Sandler had made arrangements to join the Crow laboratory as a postdoctoral fellow even before SD was discovered. When it became evident that SD represented an actual instance of meiotic drive, which Sandler had considered from a theoretical standpoint only, he quickly teamed up with Hiraizumi to delve further into the phenomenon.
Over the next several years, Hiraizumi and Sandler provided the framework for the basic understanding of the phenomenon of segregation distortion, on which all subsequent investigations, including my own, have been built.
It did not take long for the two to establish that SD acted only in males; transmission from females was in accord with genetic law. The team also showed that distortion did not depend particularly on the laboratory strain used in the original crosses, since distortion could also be observed when a variety of other chromosomes were paired with SD. Furthermore, the two demonstrated that distortion did not happen as a resuit of increased mortality of embryos that inherited the non-SD chromosome. Rather, they concluded that distortion was related to a dysfunction in the sperm. That is, some of the sperm (those bearing the non-SD chromosome) in these distorting males failed to develop or function properly and did not participate in fertilization at all.
Most important, Hiraizumi and Sandler showed that distortion was not a property of the SD chromosome itself, but was caused by at least two discrete genes with distinct roles. In later studies, Daniel L. Hartl at the University of Minnesota expanded and clarified the definition of the two key components of the SD chromosomes.
One of these elements is called Sd, a particular gene carried on SD chromosomes. Sd was defined as the gene primarily responsible for causing distortion. The other element is called Responder (Rsp) and was defined as the apparent target of distortion. At least three distinct variants of Rsp have been found in nature and in laboratory strains: Rsp^sup i^ (Responder insensitive) is found on all SD chromosomes, as well as on some non-SD chromosomes; Rsps (Responder sensitive) is found on chromosomes that are sensitive to the action of Sd; and Rsp^sup ss^ (Responder supersensitive), is particularly sensitive to distortion.
Following the initial observations by Hiraizumi and Sandler, studies by several other investigators helped to fill in pieces of the puzzle. James W. Peacock and John Erickson, working at the University of Oregon, demonstrated that meiosis itself proceeds normally in distorting males-chromosomes segregate and are apportioned normally into the immature sperm cells called spermatids. Careful measurements by Hartl and Hiraizumi and by Benedetto Nicoletti and Gianni Trippa at the University of Rome confirmed sperm dysfunction as the ultimate basis of distortion.
Distorting males produce only half as many progeny as normal males even though the embryos produced with their sperm do not experience a greater rate of mortality. The decrease in progeny therefore implied that, compared with normal males, distorting males generate only half as many functional sperm, the vast majority of which contain the SD chromosome and not the homologue. The other half of the sperm-those that received the non-SD chromosome-are eliminated or are otherwise rendered unable to participate in fertilization.
Taken together, these results all indicate that SD manifests its effects during the time that the immature-but seemingly normal-spermatids mature into fully functional sperm. What goes wrong between formation of spermatids by meiosis and their subsequent maturation to sperm? The answer was provided by electron-microscope studies of sperm maturation by Kiyoteru T. Tokuyasu. and his colleagues at the University of California, San Diego.
In the course of normal maturation, just before sperm acquire their characteristic elongated cell bodies and their tails, the chromosomes become extremely condensed within the highly compacted sperm nucleus. However, Tokuyasu and his colleagues found that in distorting males the chromosomes in precisely half of the spermatids fail to condense and instead remain dispersed. These spermatids are unable to form mature, viable sperm.
The general interpretation that emerges from these results is that in distorting males, the SD chromosome produces a deleterious effect on its partner chromosome. Spermatids that receive this partner chromosome then fail to mature properly, whereas spermatids that receive the SD chromosome develop normally and carry the SD chromosome into the next generation.
From this perspective, SD chromosomes have managed not merely to cheat the system, but to do it in a diabolical manner worthy perhaps of Shakespeare-they eliminate the competition by fratricide.
In spite of having this general understanding of the basis of distortion, biologists would like to know how, on a molecular level, these events transpire. We would like to know how the specific genes involved-Sd and Rspinteract. And to understand that, we need to know what products these genes encode. Only when we have answered these questions can we fully unravel the mystery of SD. These are the questions we are now trying to answer in my laboratory.
Our recent results are the outcome of studies that I began more than 20 years ago, when I was a graduate student in Sandler’s laboratory at the University of Washington. I set out to generate chromosomes from which either the Sd or the Rsp gene was deleted by exposing the chromosomes to x rays. The resulting small deletions would be useful in precisely pinpointing the chromosome location of these genes and in characterizing their functional properties.
I showed that when the Sd gene was deleted from an SD chromosome, the deleted chromosome was no longer able to distort a sensitive partner chromosome; both chromosomes were then transmitted to offspring in normal ratios. This result demonstrated that the Sd mutation caused some new function to be acquired. (This is in contrast with most genetic mutations, which cause some normal function to be lost.) When the gene producing this novel activity was completely eliminated by a deletion, an otherwise intact SD chromosome lost all ability to cause distortion.
Furthermore, when Rsp^sup s^ was deleted from the partner chromosome, it was no longer subject to distortion by SD. Instead, it was transmitted normally, as though it carried the Rsp^sup i^ gene. These results supported the idea that as a consequence of some action of Sd, the chromosomes in a spermatid nucleus that receive the Rsp^sup s^ gene fail to condense properly during sperm maturation. A chromosome that entirely lacks Rsp is immune to the effects of Sd. Interestingly, Rsp does not appear to have any essential function of its own. Even flies that are missing Rsp from both chromosomes are viable and fertile.
Additional studies allowed me to determine exactly where on the chromosome these genes lie. Rsp turns out to be very close to the center of the chromosome, near a structure called the centromere, which is important for chromosome movement during cell divisions-during, for example, meiosis. This particular chromosomal region contains mostly heterochromatin, highly repetitive, simple DNA sequences that generally do not code for protein. Nevertheless, heterochromatin constitutes about one-third of the total length of the chromosome. The precise function of heterochromatin is still unknown, although it is thought that this region has some structural role and may be involved in meiosis.
The location of Rsp in heterochromatin was consistent with its genetic behavior as some kind of target for the action of Sd, rather than as a typical gene encoding a protein product.
This result was confirmed by a molecular analysis by Chung-I Wu at the University of Rochester and Terrence W Lyttle at the University of Hawaii. They successfully cloned and sequenced Rsp and showed that it does not code for a protein. Instead, Rsp is composed of a simple DNA sequence, containing 120 nucleotide bases, repeated over and over again for its entire length. Furthermore, they found that the sensitivity of Rsp to the action of Sd is a direct consequence of the number of times this sequence is repeated. The insensitive variant of Rsp, Rsp^sup i^, contains fewer than 50 copies of this sequence. The sensitive variant, Rsp^sup s^, contains several hundred copies, and the supersensitive variant, Rsp^sup ss^, contains about 1,000. The sensitive variants are so large that Sergio Pimpinelli and Patrizio Dimitri at the University of Rome have shown that these DNA segments can be seen under the microscope as a discrete blocks of heterochromatin.
At the same time that other laboratories were learning about Responder, people in my lab were trying to understand the other half of the problem. We were trying to make sense of Sd-the key gene required for distortion. This was difficult, since we had no idea what the gene looked like and didn’t know whether we would recognize it if we did in fact come across it.
On the basis of its genetic properties, we anticipated that the mutant geneSd-would differ from its normal counterpart-Sd+-by more than just a single nucleotide base. We believed the difference to be more substantial, and therefore that it could be readily discerned by standard molecular biological methods. Using cytogenetic techniques, John B. Brittnacher extended my original deletional analysis and identified a chromosome segment of roughly 200,000 bases in which Sd was located. Patricia Powers cloned this entire region as a series of small overlapping fragments. She then compared each fragment from the SD chromosome, which carries the mutant gene, with the corresponding fragment from normal chromosomes.
The comparison revealed only a single difference. A particular fragment was about 6,500 nucleotide bases long on the normal chromosome but was almost twice as long-12,000 bases-on the SD chromosome. Further analysis revealed the reason for this size difference. A segment of DNA is duplicated on the mutant chromosome.
Since this was the only detectable difference between the normal and the distorting chromosome, we concluded that it contained at least part of the Sd gene. But we needed to know whether it contained all of Sd.
To find out, Janna, McLean and Cynthia Merrill injected the DNA fragment we believed to contain the distorter gene into Drosophila embryos containing only normal chromosomes. The inserted genes can become incorporated in the DNA of some of these embryos. We predicted that if we actually had the Sd gene, that embryos receiving and integrating the inserted DNA would acquire the ability to produce offspring that could cause distortion. This in fact is what happened.
Now that we knew that our inserted DNA fragment contained a gene or genes capable of inducing full distorting ability, we hoped we could identify the Sd gene itself and determine its function.
From her analysis, Merrill determined that the normal fragment of 6,500 bases actually contains two overlapping genes. One of these is the Drosophila counterpart of a mammalian gene encoding heparan-sulfate-2-sulfotransferase (HS2ST). The second gene encoded the Drosophila counterpart of a protein known in yeast and mammals as RanGAP. RanGAP has recently been shown to be an essential component of a complex system that transports proteins and RNA molecules into and out of the cell’s nucleus.
We discovered that both the HS2ST and the RanGAP genes are represented twice on the SD chromosome, as opposed to just once on the normal counterpart. Both genes appear to be normal on the right hand portion of the duplication, as does the HS2ST gene on the left. But the RanGAP gene on the left is not normal; rather it encodes a mutant RanGAP that lacks the last 234 amino acids.
Because this truncated RanGAP protein was the only substantially altered protein encoded in the SD fragment, we concluded that this was the one responsible for the distorting activity. If that were true, we expected to be able to find the truncated protein in the testes of distorting males. Leyla Bayraktaroglu and Ayumi Kusano demonstrated that the truncated protein is indeed found in the testes of distorting males, as is the normal-sized protein. In contrast, normal males produce only the normal-size protein.
To obtain decisive proof that we had the right gene, we once again injected a DNA fragment into normal embryos to create distorting flies. This time, Merrill inserted only the left half of the fragment-the one that contains the truncated RanGAP gene. (Since this gene overlaps with the HS2ST gene, it is impossible to insert RanGAP alone. Both genes were inserted together, but HS2ST was disabled and rendered nonfunctional. Only the truncated RanGAP protein could be produced from this fragment.)
The results from this experiment were unequivocal. Flies that received the engineered DNA fragment acquired the ability to cause distortion with the same strength as flies carrying a native SD chromosome. Therefore, we concluded that the truncated RanGAP is indeed the functional Sd product.
As this was the first time that the molecular defect in any meiotic-drive system had been identified, it represented a major step forward. Now that we know that Sd encodes an altered RanGAP protein and that RanGAP is important for molecular trafficking across the nuclear membrane, we can start to speculate on possible scenarios for segregation distortion.
We know that chromosome condensation is necessary for spermatids to mature into viable sperm and that this step fails in distorting males when a sensitive Responder is present in spermatid nuclei. Various sperm-specific chromosomal proteins are required to bring about chromosome condensation. The messenger RNAs encoding these proteins must be exported from the nucleus to the cytoplasm where the proteins are manufactured. After their synthesis, the proteins must then be imported into the nucleus. Thus, we can readily imagine how a perturbation in nuclear transport could result in failed chromosome condensation by affecting the production or nuclear abundance of these proteins. Because this extreme compaction of chromosomes happens exclusively during sperm development, this probably explains why sperm but not eggs are affected by Sd.
Our current studies are aimed at trying to unravel how the truncated RanGAP interferes with nuclear transport. Kusano, in my laboratory, has demonstrated that the mutant RanGAP still retains its normal biochemical activity. However, for various reasons we suspect that the truncated RanGAP could be mislocalized within the cell. This is important because the mechanism of nuclear transport is critically dependent on the normal cytoplasmic localization of RanGAP activity. Nuclear transport would be disrupted if RanGAP were within the nucleus. Merrill has obtained some tantalizing preliminary evidence indicating that the truncated RanGAP is indeed mislocalized to the nucleus during some stages of sperm development. These results need to be confirmed, and additional experiments are required to determine whether this is what is ultimately responsible for the failed chromosome condensation that occurs in dysfunctional sperm. Experiments to address these issues are under way
What is more difficult to understand is why in a distorting male, only those spermatids that receive a Rsps- or Rsp”bearing chromosome are affected by Sd. Why do the Rsp^sup i^-bearing spermatids still develop normally? One possibility is that, for some reason, the truncated RanGAP is mislocalized to the nucleus only in spermatids containing Rsp^sup s^ or Rsp^sup ss^.
Another possibility is that the spermatids containing Rsp^sup s^- or Rsp^sup ss^-bearing chromosomes might be more susceptible to a defect in nuclear transport because they contain many more copies of a particular DNA sequence than do Rsp^sup i^-containing spermatids. These sequences could preferentially bind to the proteins that facilitate chromosome condensation. If the amount of these proteins is limited owing to a defect in nuclear transport, there may not be enough to condense the rest of the chromosomes inside the nucleus.
Of course, much more work is needed to test these ideas. Segregation distortion has been a puzzle for more than 40 years, and it is probably too much to expect that it will fully give up its remaining secrets any time soon. Nevertheless, for the first time we have been able to identify the underlying molecular defect in a meiotic-drive system.
This advance has offered us an entirely new perspective on segregation distortion, enabling us to frame specific questions and pointing us in the direction of further investigations. It is exciting and satisfying that we have been able to establish a link between SD and nuclear transport, a process of fundamental biological importance and currently one of the most vigorously studied areas of cell biology. Undoubtedly, efforts to elucidate the mechanisms of distortion at the cellular level will benefit from the studies of nuclear transport in other experimental systems. Conversely, future studies of SD should not only resolve the remaining questions about its mechanism, but may also offer novel insights into the important process of nuclear transport as well.
The author would like to thank the members of his laboratory, past and present, for their important contributions to the work reviewed here and in particular, Cynthia Merrill, for her exceptional efforts, key discoveries and helpful comments on the manuscript. This work was supported by the National Science Foundation.