Meiotic recombination in a commonly used laboratory mouse strain showed sensitivity to dietary changes.

Recombination within the germline is a tightly controlled process. But new research suggests that nutrition may introduce some variability into this crucial step in genetic transmission, which could have implications for the design of future genetics studies.

A study published in GENETICS found that genome-wide recombination rates in the spermatogenesis of adult male mice were sensitive to dietary changes, but that the observed changes in crossover frequency differed depending on genetic background.

“Our findings were very unexpected,” says senior author Elena de la Casa-Esperon. “It was hard to believe that a small change in diet could affect recombination, something we thought was very much a protected process.”

Unexpected Environmental Effects

Not many studies had previously examined the impact of environmental factors on recombination and most had studied maturing germline cells in utero or in juvenile animals. In contrast, de la Casa-Esperon and her team used adult males from genetically diverse inbred strains.

“We found that there was a gap in knowledge,” says de la Casa-Esperon. “There were two classic studies in flies and yeast that hinted at the possibility of nutritional effects on levels of recombination, but the possibility of diet effects has largely been neglected in other studies.”

The researchers used immunohistochemistry to evaluate crossover frequency after 24 days of feeding each group of mice one of three diets: a standard “maintenance” diet, a 50% “undernourishment” diet, or a nutrient-rich “breeding” diet.

In two of the three genetic strains studied, recombination rate stayed relatively constant following a switch to the restrictive “undernourishment” diet, indicating that tight control over reproduction is maintained even under stressful conditions. However, the nutrient-rich diet caused a reproducible increase in recombination levels in C57BL/6 mice, one of the most commonly used strains of laboratory mice.

“This wasn’t the kind of study where you find a dramatic effect that’s easy to see—the changes we observed were small, but with large numbers we could see that they were consistent,” says de la Casa-Esperon. “As a scientist looking at these types of phenomena, first you have to convince yourself, then you have to convince others.”

Cautionary Findings and Lingering Questions

Despite her initial surprise, de la Casa-Esperon is now convinced that other researchers need to realize that germline genetic information is more sensitive to environmental factors than previously thought, which could have transgenerational and evolutionary implications. “The take-home message is that there are many environmental variables that affect the traits we study, and we need to be vigilant about the potential effects of these variables on our results,” she says.

She also points out that dietary changes throughout lifetime may be a common occurrence in animal facilities; for instance, pregnant and nursing females are sometimes fed with nutrient-rich “breeding” chows while offspring are later switched to leaner “maintenance” diets. Hence, dietary regimes must be controlled and reported in recombination studies, as they can affect the results.

Next research steps include attempting to isolate the diet components that significantly contribute to changes in recombination levels. Phytoestrogens are promising candidates, since they are known to have epigenetic effects in germline—and these could, in turn, affect recombination. De la Casa-Esperon hopes other groups will study whether diet—and other environmental factors—can alter recombination rates in other species.

CITATION:

Diet effects on mouse meiotic recombination: a warning for recombination studies

Angela Belmonte-Tebar, Estefania San Martin Perez, Syonghyun Nam Cha, Ana Josefa Soler Valls, Nadia D Singh, Elena de la Casa-Esperon

GENETICS

2022:iyab190

https://doi.org/10.1093/genetics/iyab190

Scientists generated a karyotype, chromosome-scale genome assembly, and manual genome annotation for a common ctenophore.

Ctenophores—beautiful marine invertebrates also known as “comb jellies”—have long fascinated and perplexed biologists. Phylogeneticists believe that either ctenophores or sponges were the first organisms to branch off from the tree of life, making them the “sister clade” to all other animals. However, debate is ongoing about which group can claim this title.

A study published in G3: Genes|Genomes|Genetics provides researchers with new tools to address these questions. The paper describes the first-ever published ctenophore karyotype and the first full annotated genome assembly for Hormiphora californensis, a small, globular ctenophore that is found off the shores of California and elsewhere in the Pacific Ocean.

The chromosome-scale H. californensis genome will allow whole-genome comparisons that weren’t possible with previous assemblies, says lead author Darrin Schultz. “The applications go beyond studying ctenophores—we can compare this genome to that of any other species to look for differences and understand more about the history of evolution.”

Going Over the Genome with a Fine-Toothed Comb

Schultz and his fellow researchers started with karyotyping, using a simple DNA staining protocol on ctenophore embryos to determine that H. californensis has 13 chromosomes. Then, after successfully isolating DNA samples suitable for long-read sequencing—a feat in and of itself—the team used PacBio Iso-Seq and Illumina RNA-seq technologies to sequence and annotate the full 110-Mb genome, manually determining at high resolution how genes were arranged on each chromosome.

“We spent hundreds of hours going through the entire genome by hand and making sure that every single gene annotation was correct,” says Schultz. “We did this because we saw evidence that previous ctenophore genome assemblies contained incorrect fusions or splits in genes. We wanted it to be right this time.”

As the researchers sifted through the genome, they found that H. californensis has an unusually high degree of heterozygosity—up to eight or nine percent in some areas of the genome. This presented bioinformatic challenges.

“The copies of the genome from the maternal side and the paternal side are so different from each other that it was like I had to assemble two separate genomes and try to come up with one assembly to represent both of those unions at the same time,” says Schultz.

As they got further into the annotation process, the team realized that two to three percent of the protein-coding genes occur completely within the bounds of another gene—not just one inside another, but multiple genes nested like Russian dolls. This architectural quirk has been observed in other species, especially eukaryotes with smaller genomes, but scientists are not yet sure whether the transposition processes that likely cause this “nesting” have any implications for the overall biology of an organism.

Comparative and Evolutionary Studies in Store

The new annotated ctenophore genome assembly will likely be used for comparative studies to address the phylogenetic “sister clade” debate between ctenophores and sponges in the coming years. Other applications of this study include addressing even broader questions about life on Earth. Some researchers are studying the genetic underpinnings of symbiosis using marine models like corals, ctenophores, and other cnidarians that have symbiotic algae. Others are interested in using comparative genomic studies to examine the evolution of specific traits that separate animals from all other organisms.

“This genome is just another step toward understanding how neurons and muscle cells—these things that made all of us uniquely animals—evolved over 600 million years ago,” says Schultz.

Schultz also notes that H. californensis may eventually be able to become a model species. It has a short life cycle, clear embryos ideal for developmental observation, and is more easily maintained in culture than many other related organisms. And now, it has a complete annotated genome available for study.

CITATION:

A chromosome-scale genome assembly and karyotype of the ctenophore Hormiphora californensis
Darrin T Schultz, Warren R Francis, Jakob D McBroome, Lynne M Christianson, Steven H D Haddock, Richard E Green
G3 Genes|Genomes|Genetics 2021: jkab302
https://doi.org/10.1093/g3journal/jkab302

Researchers used a forward genetic approach to identify genes that affect a social behavior in honey bees.

For more than 30 years, honey bee geneticist Robert E. Page, Jr. and his colleagues have sought the genes that influence a colony trait that only emerges from interactions between thousands of individual bees — a social phenotype.

Such traits are notoriously difficult to study. As hard as it is to disentangle a gene’s effect on behavior from environmental influences, these challenges are greatly multiplied by the complex interactions between genetically different individuals forming and altering their own social environments. As a result, mapping quantitative trait loci (QTLs) for social behavior long seemed out of reach. A recent paper published in GENETICS surveys the three decades of work that culminated in identifying genes that affect a complex, socially-regulated foraging behavior.

“Everything flowed from our observations of the behaviors that were associated with foraging behavior and division of labor and the resulting impact on the amount of pollen stored in the comb,” says Page, a researcher at Arizona State University. “Over time, we adapted new technologies to ask questions about those phenomena in different ways. By bringing together a host of different toolkits, we were able to build a story of the underlying genetic basis of a very complex social trait.”

Pollen hoarders

In a honey bee colony, some of the bees specialize in collecting pollen and hoarding this protein-rich food in wax cells near the “nursery” where eggs and larvae develop. Some of the bees eat the pollen and produce glandular secretions to feed the larvae; in turn, the larvae produce pheromones that stimulate foragers to collect pollen.

Despite the complexity of factors involved — thousands of individual bees, larvae, their interactions, and their environment — the total amount of pollen stored in the colony is a regulated trait. In a previous study, ASU reseacher Professor Jennifer established this by adding and removing pollen from honey bee colonies and observing the changes in foraging behavior. Each colony had a set level of stored pollen that the bees collectively sought to maintain; when researchers added pollen, foraging decreased until excess pollen had been consumed, and when researchers removed pollen, foraging increased until pollen again reached the colony’s standard level.

Using selective breeding, Page’s team generated strains of honey bees that substantially differ in amount of pollen stored within just three generations. “People have tried for decades to breed bees that store more honey, but honey storage is a very sloppily regulated trait,” says Page. “As long as nectar is available, bees will bring it back and stick it anywhere they can find space in the hive. In contrast, pollen storage offers excellent consistency of measurement.”

Phenotypic and genotypic analysis

Using a wide range of methods, Page and his colleagues studied the phenotypes and genotypes of these strains for 42 generations of selection. Phenotypic mapping revealed that bees in colonies that store more pollen are likely to have more ovarioles (the egg-producing structures in the insect ovary) and to be more sensitive to sugar than those from the colonies with less pollen. One research collaborator, Ying Wang, went so far as to painstakingly graft ovarioles from worker bees into recipients from a colony that produced less, which in turn affected the recipients’ behavior.

When they began to map genetic trait determination, the researchers expected that individual behavior would be more selectable than complex social traits playing out across large groups. Instead, they were surprised to find that genotype explained roughly 41 percent of colony variance in stored pollen but only two percent of individual variance in pollen collection.

Digging down to the genetic level, Page and his colleagues performed QTL mapping for the social phenotype of pollen hoarding along with individual foraging behavior, physiology, and anatomical traits. From the gene lists for each QTL they identified candidate genes of interest based on the phenotypic architecture and assessed them using expression assays and gene knockdown. Ultimately, they identified three genes of special interest, all of which have some association with ovary size. Future studies will involve examining the effects of these genes in developing larvae more closely.

A vertical approach

Page attributes his success in part to his tenacity in pursuing a “vertical approach”—interrogating a single question at every level of influence, starting with social interaction and working down to examine individual behavior, morphological differences, physiology, developmental processes, and genetic variation.

Pursuing this vertical approach over so many years involved the work of countless expert apiculturists, laboratory technicians, students, and fellow researchers alongside Page. “Our success came from combining the right people with the right tools as they became available,” he says. “Everyone who came through my lab made it better and brought something new.”

CITATION:
Societies to genes: can we get there from here?
Robert E Page, Jr.
GENETICS 2021; iyab104
https://doi.org/10.1093/genetics/iyab104

Researchers identified dozens of quantitative trait loci controlling important traits in Cannabis sativa.

In 2014, United States federal law changed to allow scientific research on Cannabis sativa in states with regulated hemp programs. This legal shift opened the door to research that had previously been slow and difficult due to regulatory hurdles and funding challenges.

A new study published in GENETICS capitalized on this new opportunity and identified 69 quantitative trait loci (QTLs) that are responsible for variation in key agronomic and biochemical traits in C. sativa. This research is a step towards understanding the genetic control of complex traits in hemp and will inform future investigations into the overall evolution and function of complex traits across multiple species.

Hemp is grown for a wide range of commercial uses, including in building materials, textiles, and composite plastics, food and drink, animal feed, and pharmaceutical cannabinoid products, says study leader John McKay of Colorado State University. “It’s important to understand the genes controlling this plant as a crop, and it’s also interesting from a fundamental evolutionary standpoint.”

Identifying QTLs in a non-model species

“Scientists have previously made progress in identifying the genetic basis of complex traits in model and crop species—our team wanted to ask those questions in C. sativa because it’s an understudied species,” says McKay.

Today’s genome sequencing and assembly technologies are largely species-agnostic. As a result, scientists can now take bioinformatic and statistical genetic tools that were initially developed and tested in fruit flies and Arabidopsis and increasingly apply them to non-model species.

McKay and his team developed an F₂ hemp population by crossing two phenotypically distinct varieties—a tall, late-flowering cultivar bred for fiber production and a shorter, early maturing hemp that was bred for both fiber and grain crops. They then used whole genome sequencing to map QTLs associated with traits of interest, such as grain yield and stem biomass, along with 17 biochemical traits. These included levels of cannabidiol and terpenes, which have medical uses, and THC, which is the psychoactive component of marijuana. THC is strictly regulated in the US, and THC levels in industrial hemp must be below 0.3% to remain legal crops.

Most QTLs they identified clustered into one of four genomic regions, suggesting that much of the difference between the two varieties is due to a small number of genes that have large pleiotropic effects. Two candidate genes emerged that may underlie some of these clusters: the homolog of an Arabidopsis transcription factor gene called TINY may be associated with a cluster of agronomic traits, and the gene for olivetol synthase appears to underlie variation in a cluster of biochemical traits, consistent with the enzyme’s role in cannabinoid synthesis. The researchers functionally validated the olivetol synthase candidate by expressing the two hemp alleles in yeast. They found that the allele from the low-cannabinoid cultivar produced less olivetol in the yeast expression system, supporting the hypothesis that allelic variation at this gene plays a role in the observed phenotypic variation.

The researchers also observed epistatic interactions between some of the QTL clusters, further complicating attempts to elucidate any one trait’s exact genetic underpinnings.

“This study definitely adds to the broader conversation about complex traits,” says McKay. “For example, everyone agrees epistasis exists, but breeders and geneticists like to argue about whether it’s important to include in prediction models. Documenting additional cases like this in which epistasis contributes to variation adds to our understanding of the basis of complex traits.”

Traits are complicated but still predictable

The results of this latest study contradict a paper from 2003 that concluded that variation in cannabinoid production is controlled by a single genetic locus. The team from Colorado State University identified at least four loci controlling variation in these chemotypes.

“The field of genetics has always been a friendly place to hypothesize that something—anything—is polygenic,” laughs McKay. “Finding multiple loci controlling a single biochemical trait wasn’t surprising to me, because the abundance of any molecule can be influenced not only by the pathway that makes that molecule but also the ones that influence the cells and machinery that contribute to the process.”

However, despite overturning assumptions of one-to-one genotype-phenotype interactions, McKay emphasizes that he still views the hemp traits in the study as predictable. Future grants would allow research groups like his to dive deeper into the adaptive value of cannabinoids in hemp plants and create more precise genetic manipulations of key traits of interest. Researchers are eager to see policy informed by scientific understanding of the factors that predictably affect cannabinoid content and other traits in hemp crops.

CITATION:
Quantitative Trait Loci Controlling Agronomic and Biochemical Traits in Cannabis sativa
Patrick Woods, Brian J. Campbell, Timothy J. Nicodemus, Edgar B. Cahoon, Jack L. Mullen, and John K. McKay
GENETICS 2021; iyab099
https://doi.org/10.1093/genetics/iyab099

Multiple divergent retrocopies of the well-characterized Ku70 gene were identified in humans and other primates.

The last 63 million years of primate evolution have been strongly shaped by genetic retrotransposition; thousands of genes and proteins with new functions have evolved from retrocopies scattered throughout the genome. These retrocopies arise when retrotransposons reverse transcribe a cellular mRNA and insert the resulting cDNA copy back into the genome.

A recent study published in G3: Genes|Genomes|Genetics characterized a new family of five retrocopies in the human genome. These retrocopies—which researchers named NUKUs—are derived from KU70, a highly conserved gene that encodes a ubiquitous protein involved in DNA double-strand break repair.

“Many people think retrocopies are junk DNA, but there are many examples where they have functional relevance in humans and other organisms,” says lead author Paul Rowley of the University of Idaho. Rowley and his collaborators found that NUKUs are among those retrocopies that may have a function that differs from that of the original parent gene.

The long journey to identifying and characterizing NUKUs

More than 10 years ago, scientists studying Ku70 performed a BLAST search for the gene in the human genome. They noticed that additional genes came up in the search results, often truncated or missing introns.

“It raised some eyebrows, because very few other genes in the DNA double-strand break repair pathway have this many retrocopies,” says Rowley. Since that initial observation, numerous scientists have contributed to the ongoing process of characterizing these NUKU retrocopies.

Through experiments, the researchers demonstrated that NUKUs are transcribed in human tissue. To determine whether these retrogene transcripts were capable of producing proteins, they mined existing Ribo-seq (ribosome profiling) datasets to look for unique NUKU transcripts that were easily distinguishable from Ku70 and other transcripts.

“The day when we confirmed the ribosome association of the transcripts—that was a good day,” says Rowley. “We had already spent a lot of time showing that NUKUs are transcribed, but without ribosome association, any case for full protein translation is pretty much dead in the water.”

Rowley emphasizes the collaborative nature of this work. “Being supported through the Institute for Modeling Collaboration and Innovation enabled me as an empiricist to bring expertise in molecular and computational modeling on board, which really strengthened the paper and laid better groundwork for future studies,” he says.

Hypothesizing functionality

The researchers have not yet determined the functionality of NUKU genes, but they know enough to conclude that they aren’t exact functional duplicates of Ku70.

“You would imagine that a retrocopy would function as a bona fide copy of its parental gene, but if these proteins are doing something in the cell, they certainly seem to have lost their canonical function through truncation and specific mutations,” says Rowley.

The Ku70p protein interacts with Ku80p, the other half of the eukaryotic Ku heterodimer. Through computational modeling, collaborator Jagdish Patel demonstrated that the mutations in the NUKU retrogenes would disrupt the NUKU proteins’ ability to bind Ku80p. Without the ability to form a heterodimer, it is unlikely that NUKUs are involved in DNA double-strand break repair. Their expression patterns suggest divergence too: while Ku70 is expressed fairly ubiquitously, NUKU transcripts show greater tissue specificity. Together, the study’s results suggest that NUKUs evolved rapidly during primate speciation and that the genes developed novel functions separate from Ku70.

Future studies will aim to identify possible functions of this familiar-yet-distinct gene family. These studies might also provide broader evolutionary insights by revealing factors driving the selection of specific mutations in NUKUs, such as regulation or interactions with pathogens or other host proteins.

CITATION

NUKU, a family of primate retrocopies derived from KU70

Paul A Rowley, Aisha Ellahi, Kyudong Han, Jagdish Suresh Patel, James T Van Leuven, Sara L Sawyer

G3 Genes|Genomes|Genetics 2021; jkab163

https://doi.org/10.1093/g3journal/jkab163

Over 90 percent of the genome of this pathogen threatening Australian ecosystems is made of transposable elements.

Myrtle rust (Austropuccinia psidii) is a pathogenic fungus that has spread rapidly across the globe with devastating effects on local vegetation, including agricultural crops. It is extremely versatile, with over 480 known host species in the widespread myrtle (Myrtaceae) family. Researchers have prioritized studying myrtle rust in order to protect biodiversity, but these studies have been hindered by the lack of a complete reference genome—until now.

A study published in G3: Genes|Genomes|Genetics describes the assembly of the myrtle rust’s unusual genome sequence and reveals features that scientists hope will help them combat this destructive invasive species.

A. psidii has been particularly damaging in Australia, where Myrtaceae make up a large proportion of the native flora.

“Our dominant species in Australia are in the family Myrtaceae, so it was very alarming when we found we had this new disease that could potentially have a devastating impact on our ecosystems, as well as our forestry and other industries,” says Peri Tobias, PhD, one of the project’s principal investigators.

To sequence the A. psidii genome, a group of Australian researchers joined forces with scientists from New Zealand, where myrtle rust was first detected in 2017 and is now found across a wide area of the country.

A gigabase-sized undertaking

The team initially struggled to assemble the genome because early size predictions from previous studies were too small by nearly an order of magnitude. To determine the correct depth of coverage needed, Tobias and her team used flow cytometry and short-read analysis tests to estimate the fungus’s nucleus size and predict the size of the genome. The fungal nuclei predominate in a di-haploid state, and the researchers predicted that the haploid genome would be just over 1 Gb (gigabase), much larger than any previously assembled rust fungus genomes.

Building on this information, the researchers performed more long-read sequencing and mapped out scaffold grouping predictions based on the presence of telomeres. Assembling the A. psidii genome used a staggering 359,878 CPU hours, with the computations running continuously for over four months on the University of Sydney High Performance Computing Cluster. BUSCO (Benchmarking Universal Single-Copy Orthologs) confirmed that the assembled genome, which matched the prediction of ~1 Gb, was high-quality and over 94.7% complete.

Unique genomic features

The massive size of the A. psidii genome can be attributed almost entirely to repetitive regions that make up over 91% of the genome. Most of these regions are novel transposable elements that are unique to this species rather than being shared with other fungi.

In addition to the sheer number of transposable elements, the researchers were also surprised by the arrangements of repetitive regions within the genome. Rather than clustering together, most genes are quite spread out and interspersed with long non-coding regions; this suggests that some of the transposable elements may insert preferentially into intergenic regions rather than forming long stretches of repetitive non-coding DNA.

“We were quite gobsmacked by how much repetitive DNA was in there due to these transposable elements,” says Tobias. “It was very interesting to us because it seems to indicate high evolution potential in this fungus.”

Combating disease spread

The researchers believe this high evolution potential may have contributed to the fungus’s ability to infect such a wide range of hosts. To explore this hypothesis, Tobias and her colleagues are performing comparative studies between different strains of this pathogen. Future research will also focus on investigating the effector genes that manipulate host plants. Ultimately, understanding host-pathogen interactions will help the researchers uncover the characteristics that can make potential hosts resistant.

“When we’re looking at host-pathogen interactions, we’re always trying to find the mechanism by which the pathogen infects the host,” says Tobias. “Knowledge is power. Staying one step ahead of the pathogen, stacking resistance genes in our crops, working with the attributes that we know enable resistance—that’s how we’ll start to address these large problems.”

CITATION

Austropuccinia psidii, causing myrtle rust, has a gigabase-sized genome shaped by transposable elements

Peri A Tobias, Benjamin Schwessinger, Cecilia H Deng, Chen Wu, Chongmei Dong, Jana Sperschneider, Ashley Jones, Zhenyan Lou, Peng Zhang, Karanjeet Sandhu, Grant R Smith, Josquin Tibbits, David Chagné, Robert F Park

G3 Genes|Genomes|Genetics, Volume 11, Issue 3, March 2021, jkaa015, https://doi.org/10.1093/g3journal/jkaa015

Yeast circumvent Dollo’s Law and reacquire traits lost tens of millions of years ago from distant relatives.

Dollo’s Law, first proposed in the 19^th century, holds that evolutionary losses are irreversible — if a species discards a complex trait, it will never reacquire exactly that trait.

However, a study published in Genetics highlighted one way that unicellular fungi can break Dollo’s Law: horizontal gene transfer. Although budding yeasts do not normally mate with distant relatives, data indicate that they can occasionally exchange genetic material and thereby regain lost genes or even gene clusters, despite functioning without them for millions of years.

Odd species out

“Laws in biology are almost meant to be violated. It’s a science of exceptions,” says Chris Todd Hittinger, one of the senior authors. Hittinger leads a laboratory at the University of Wisconsin-Madison, where he and his team study yeast metabolism, biodiversity, evolution, and ecology. This recent study emerged from data observations collected as part of the Y1000+ Project, a National Science Foundation-funded collaboration between Hittinger’s lab, Antonis Rokas’ lab, and others worldwide. The initiative is aimed at cataloging and analyzing genomic and metabolic data for all known yeast species from the subphylum Saccharomycotina in order to understand their evolution and ecological functions.

First author Max Hasse, who was an undergraduate researcher in Hittinger’s lab at the time, was the first to notice something odd in the Y1000+ Project data: some species that feed on the sugar galactose are deeply embedded in clades in which the other species generally cannot use galactose.

“His instincts were absolutely right,” says Hittinger. “For a junior in college to have that kind of scientific insight and know where to follow things really bodes well for his career.”

The researchers investigated the genomic data further and found that in these out-of-place species, the structures of the GAL gene clusters (needed for metabolizing galactose) were strikingly similar to those of distantly related budding yeasts, specifically species in the CUG-Ser1 clade. Further investigation led them to conclude that by far the most likely explanation was that CUG-Ser1 yeast served as common donors of the GAL gene cluster to at least four other species via horizontal gene transfer.

The researchers were somewhat surprised to find horizontal gene transfer in the yeast galactose utilization pathway because Saccharomyces cerevisiae — the best-known model yeast species — has a galactose utilization pathway that is encoded in multiple places throughout the genome, making transfers more difficult. These genes are tightly controlled by dedicated regulatory genes scattered across the genome. However, in CUG-Ser1 yeast, the GALgenes are clustered together and more loosely regulated by conserved factors.

Broader evolutionary takeaways

Although trait reacquisition via horizontal gene transfer is much more common in bacteria and unicellular eukaryotes in direct contact with their environment, this study underscores some broad evolutionary principles that hold true even in multicellular eukaryotes.

“Sometimes in the popular press, evolution gets portrayed as being all about chance,” says Hittinger. “There’s an element of that, but there’s also an element of predictability.” In the case of yeasts, evolutionary trait reconstruction modeling revealed that allowing trait reacquisition in some cases is more evolutionarily likely than only allowing for trait loss across all scenarios. 

However, says Hittinger, “it is still true to a point that once you lose something it’s very hard to regain it.” In the paper, the authors point out that the GAL genes of the CUG-Ser1 clade represent something of a “best-case scenario” for trait reacquisition. This is true for several reasons. As well as the genes being closely collocated and regulated by conserved factors, the phenotype offers a clear competitive advantage in galactose-rich environments, making it more likely that the trait will persist in the species once reacquired.

Even among budding yeasts, trait reacquisition is probably rare in the absence of these factors. The species that reacquired galactose utilization stood out in this study precisely because the vast majority of closely related species had not reacquired the trait, instead continuing down apparently loss-permanent evolutionary pathways.

Next Steps

As the Y1000+ Project continues, “there are a lot of exciting directions,” says Hittinger. In the immediate future, the team intends to follow up on other possible cases of horizontal gene transfer between yeast species, tying the known examples to other losses of metabolic pathways and reacquisitions. “There are lots of great candidates out there.”

CITATION:

Repeated horizontal gene transfer of GALactose metabolism genes violates Dollo’s law of irreversible loss

Max A B Haase, Jacek Kominek, Dana A Opulente, Xing-Xing Shen, Abigail L LaBella, Xiaofan Zhou, Jeremy DeVirgilio, Amanda Beth Hulfachor, Cletus P Kurtzman, Antonis Rokas, Chris Todd Hittinger

GENETICS 2021 217: iyaa012. 

https://doi.org/10.1093/genetics/iyaa012

AP-2 transcription factors, which control sleep in flies and worms, are confirmed to do the same in mammals.

Humans are not alone in their deep need for sleep. Almost all animals, even tiny nematode worms and fruit flies, suffer when deprived of their Z’s, but little is known about how sleep is controlled. New work published in GENETICS advances our understanding of this mysterious physiological state by pinpointing a key gene family that affects sleep architecture.

Two research groups independently conducted parallel studies on neural crest-derived AP-2 transcription factors in mice. Their work demonstrates that the associated genes play a conserved role in mammalian sleep, though there is some evolutionary divergence and added complexity in mammals compared to invertebrates.

From Worms to Mice

Henrik Bringmann leads the Max Planck Research Group on Sleeping and Waking. In 2013, his laboratory screened C. elegans worms for mutations affecting sleep. AP-2 deletion mutants showed no detectable sleep, suggesting these transcription factors are important regulators of the process. In 2016, Bringmann led another study showing that AP-2 transcription factors are also needed for sleep in Drosophila. 

“Sleep appears to have evolved at least 600 million years ago and has been conserved since then, which suggests that many fundamental principles of sleep regulation are conserved,” said Bringmann. “This said, sleep is more complex in humans than in worms or flies, so we need mammalian models to understand additional levels of complexity.”

The researchers’ next step was to establish whether a similar sleep-regulating role is played by related genes in mammals, the TFAP2 genes. In their most recent study, the team generated mice that were heterozygous knockouts for either Tfap2a or Tfap2b and compared each mutant to wild-type littermates. In addition to analyzing sleep duration and brain waves, they examined symptoms of sleep deprivation such as memory loss and stress resistance.

Since the AP-2 transcription factors promote sleep in invertebrates, researchers initially expected that the equivalent genes would also promote sleep in mice. They were therefore surprised to find that while losing Tfap2bfunction reduced both sleep quality and quantity, Tfap2a mutants slept for the typical amount of time and the quality of their sleep was actually higher than wild-type. This result suggests that the function of AP-2 transcription factors has diverged over the course of evolution, perhaps to allow sleep quality to be fine-tuned in either direction.

Working Backwards from Human Disorders

Meanwhile, another research group also published an article in the same issue of GENETICS examining the effects of two specific TFAP2B mutations in mice. Yu Hayashi of Kyoto University  and University of Tsukuba said his lab was inspired to pursue this avenue of research by a paper in PNAS. That study described multiple human families with mutations in TFAP2B that are associated with a rare disorder known as Char syndrome. Two of these families exhibited symptoms of disordered sleep, including sleepwalking and extremely shortened sleep.

“We were astonished to read about a family that slept just two to three hours per night and had no signs of fatigue. We thought that maybe analyzing this gene can help us understand what sleep is for in the first place,” said Hayashi. “We even thought that maybe this mutation might somehow substitute for the function of sleep. Could there be a way to help humans need less sleep?”

To answer these questions, Hayashi’s team set out to replicate those families’ specific TFAP2B point mutations in mouse models. They measured both how long the mice stayed awake and the duration of different sleep phases compared to wild-type mice and heterozygous knockouts. The results showed that TFAP2B helps determine the amount of nonrapid eye movement sleep (NREMS). However, the effects of the point mutations in mice did not match the symptoms observed in humans—the mice did not sleepwalk or show dramatically shortened sleep.

In the mice that carried the same mutation as the human family with short sleep, female mice showed fragmented sleep, while male mice did not. This result was unforeseen because there was no reported gender difference in the human family. “It was surprising for me to see the gender difference,” said lead author Ayaka Nakai, a graduate student in the Hayashi lab. It is possible that future research may uncover sex differences in sleep regulation.

In terms of generating a model of fatigue-free short sleep, the results did not match their initial hopes—sleep was reduced overall in heterozygous mutant mice, but it was generally fragmented rather than shortened. However, the results clearly established TFAP2B’s important role in sleep architecture and laid the groundwork for learning more about how sleep works.

Next Steps

Ultimately, said Bringmann, “it was satisfying to hear that both approaches converged on the same conclusions regarding TFAP2B’s role in sleep.” Together, these two papers establish that AP-2 transcription factors contribute to sleep control in mammals—just as they do in flies and worms.

The gene’s evolutionary conservation is a key lesson from the study, said Nakai.

For Nakai and Yu, the next research step is creating knockdown mice that have Tfap2b downregulated only in the nervous system. This will allow them to observe the neural effect of a homozygous loss-of-function mutation, which is developmentally lethal if the knockout is genome-wide. Studying a homozygous knockdown may offer clearer insights into how the gene affects neuron specification and activity.

Next, Bringmann is interested in examining other mammalian AP-2 paralogs, but he says invertebrates will continue to be important for identifying other genes that contribute to sleep regulation. “Going back and forth between different models will be the future of sleep research for the next years, and this will be facilitated by looking at homologous genes and conserved principles.”

CITATION:

Functional Divergence of Mammalian TFAP2a and TFAP2b Transcription Factors for Bidirectional Sleep Control

Yang Hu, Alejandra Korovaichuk,  Mariana Astiz, Henning Schroeder, Rezaul Islam, Jon Barrenetxea, Andre Fischer, Henrik Oster and Henrik Bringmann

GENETICS 2020 216: 735-752. 

https://doi.org/10.1534/genetics.120.303533

Sleep Architecture in Mice Is Shaped by the Transcription Factor AP-2β

Ayaka Nakai, Tomoyuki Fujiyama, Nanae Nagata, Mitsuaki Kashiwagi, Aya Ikkyu, Marina Takagi, Chika Tatsuzawa, Kaeko Tanaka, Miyo Kakizaki, Mika Kanuka, Taizo Kawano, Seiya Mizuno, Fumihiro Sugiyama, Satoru Takahashi, Hiromasa Funato, Takeshi Sakurai, Masashi Yanagisawa and Yu Hayashi

GENETICS 2020 216:753-764

https://doi.org/10.1534/genetics.120.303435