Artificial chromosomes are essentially miniature versions of real chromosomes that can replicate alongside their natural counterparts in host cells. They have the potential to be “incredibly useful for genome engineering, especially in cases where you want to put in a very large piece of DNA that, let’s say, [encodes] a whole cascade of enzymes involved in a particular pathway,” says chromatin researcher Gary Karpen of Lawrence Berkeley National Laboratory. However, this potential isn’t always realized because there’s a stubborn hurdle hindering artificial chromosome construction: creating centromeres.
Like natural chromosomes, artificial ones need centromeres to attach to mitotic spindles and separate sister chromatids during cell divisions. Centromeres are defined by the presence of a specialized histone called CENP-A that’s critical for connecting to spindles. But how CENP-A is initially recruited is not entirely clear. Centromeres tend to be buried deep in a jungle of repetitive DNA, known as α-satellite DNA in humans, and those repeats often contain binding sites for CENP-B, a protein thought to contribute to CENP-A loading. Researchers therefore generally include α-satellite DNA with CENP-B binding sites in sequences they wish to convert to human artificial chromosomes (HACs), but even with these seemingly appropriate sequences, centromere formation is hit-and-miss.
Chromosome biologist Ben Black of the University of Pennsylvania and colleagues have now taken a more direct approach, essentially forcibly recruiting CENP-A into their chosen piece of cloned DNA. They first incorporate repeats of a 27-base-pair sequence called LacO in the part of the overall DNA molecule where they want a centromere to form. They then transfect the LacO-containing DNA into human cells engineered to express a fusion protein composed of a LacO binding domain and a protein called HJURP—an epigenetic factor responsible for incorporating CENP-A into chromatin. After binding to the LacO repeats, the fusion protein integrates CENP-A into and along the surrounding DNA, forming a centromere to support the stable replication of the HAC.
Using this method, the team was able to create functional centromeres in previously recalcitrant α-satellite–containing regions of cloned DNA. The researchers were even able to generate HACs using a region of chromosome 4 that entirely lacked α-satellite repeats. If other DNA sequences are similarly amenable to α-satellite–free HAC production, this could eliminate the added difficulties of handling highly repetitive DNA.
“I think the idea of preloading the DNA with CENP-A is a really good one,” says Karpen, who was not involved in the study. “It’s an important breakthrough. . . . [and] opens up new opportunities for creating artificial chromosomes.” (Cell, 178:P624–39.E19, 2019).
|Human artificial chromosome formation approach||Components||Rationale||Stability of HAC Clones|
|Sequence-based||40–200 kb of α-satellite repeats with a high frequency of the CENP-B binding sequence||α-satellite DNA is found at all human centromeres. CENP-B can facilitate CENP-A nucleosome assembly.||Varies considerably. At the low end, a HAC clone can be considered stable if the HAC is present in more than 20 percent of cells.|
|Epigenetic||A 10 kb array of LacO repeats that bind HJURP fusion proteins and approximately 200 kb of surrounding DNA||CENP-A is indispensable for centromere formation and function. Forced CENP-A seeding can generate self-propagating centromeric chromatin on plasmids in fruit fly cells.||The majority of clones have the HAC present in 80 to 100 percent of cells.|