NEW YORK (GenomeWeb) – An ancient maize cob found in Mexico near the center of corn domestication is more similar to modern maize than to its wild forebears, according to a new sequencing study.
A team led by University of Copenhagen researchers sequenced the genome of a 5,310-year-old maize cob that had been uncovered during an archaeological dig in the Tehuacan Valley of Mexico. As they reported this week in Current Biology, when the researchers compared the genome sequence from this ancient sample to a reference panel of modern maize landraces and teosinte grass genomes, they found that it was more similar genetically to the modern landraces. However, it differed from modern maize at some key sites, the researchers added.
"Around 9,000 years ago in modern-day Mexico, people started collecting and consuming teosinte, a wild grass," Nathan Wales, a postdoc at Copenhagen and the Natural History Museum of Denmark, said in a statement. "Over the course of several thousand years, human-driven selection caused major physical changes, turning the unproductive plant into modern maize, commonly known as corn. Maize as we know it looks so different from its wild ancestor that a couple of decades ago scientists had not reached a consensus regarding the true ancestor of maize."
Wales and his colleagues sequenced the ancient maize sample, dubbed Tehuacan162, to an average 1.7x depth of coverage, or to 6x coverage for the regions accessible via short reads. They bolstered their whole-genome shotgun sequence data with sequencing data from a library enriched for a set of 348 loci.
As Zea mays ssp. parviglumis is the progenitor of modern maize along with introgression from Z. mays ssp. mexicana, the researchers used the phylogenetic tool TreeMix to establish the ancient sample's relatedness to a reference panel containing those along with modern maize landrace and teosinte genomes. This placed Tehuacan162 as an outgroup to modern landraces.
Using D statistics, the researchers tested the relationship between Tehuacan162 and teosinte and modern landraces, finding that the ancient sample is more closely related to parviglumis thanto mexicana, as expected, though also finding that it is more closely related to the modern landraces than it is to parviglumis. This, the researchers said, indicates that Tehuacan162 should belong to the same phylogenetic clade as maize.
Wales and his colleagues further used D statistics to compare Tehuacan162 to all 264 landraces in the reference panel. This gave them a picture of which derived alleles Tehuacan162 shared with which landrace. They also noted that they could not reject the hypothesis that Tehuacan162 was an outgroup to any given pair of landraces. Together, this and other data indicated to the researchers that Tehuacan162 represents an ancient form of maize that is closely related to the ancestor of maize, but is distinct from parviglumis.
While Tehuacan162 is closely related to modern maize, the researchers found that it shares certain gene variants that were selected for by humans during domestication, while lacking others.
For instance, when the researchers compared the Tehuacan162 sequence to the genomes included on the reference panel and previously published ancient genomes, they found that Tehuacan162 harbored the derived version of the gene that determines whether corn kernels are encased or exposed. The derived version indicates Tehuacan162 has exposed, edible kernels, the researchers noted.
However, the researchers also found that Tehuacan162 contained the ancestral version of the ear-shattering gene, which governs seed dispersal, even though analogous versions of that gene in other domesticated cereals changed early in the domestication process.
"Wild plants naturally release their seeds at the appropriate time, but humans have modified domesticated cereals so they retain their seeds so they can be easily collected from fields," Wales added. "The finding that the ancient maize cob has the ancestral version of the gene is unexpected and encourages further research."