The unique kind of human artificial chromosome: Bypassing the requirement for repetitive centromere DNA

Abstract

Centromeres are essential components of all eukaryotic chromosomes, including artificial/synthetic ones built in the laboratory. In humans, centromeres are typically located on repetitive α-satellite DNA, and these sequences are the “major ingredient” in first-generation human artificial chromosomes (HACs). Repetitive centromeric sequences present a major challenge for the design of synthetic mammalian chromosomes because they are difficult to synthesize, assemble, and characterize. Additionally, in most eukaryotes, centromeres are defined epigenetically. Here, we review the role of the genetic and epigenetic contributions to establishing centromere identity, highlighting recent work to hijack the epigenetic machinery to initiate centromere identity on a new generation of HACs built without α-satellite DNA. We also discuss the opportunities and challenges in developing useful unique sequence-based HACs.

Introduction

Since the first de novo synthesis of a prokaryotic genome [1], the prospect of building eukaryotic genomes has enticed synthetic biologists. If we could build even a single chromosome, it would unlock numerous possibilities, including engineering new gene circuits or genes of any sort. If we could build an entire eukaryotic genome, one would be presented with the ultimate synthetic biology toolbox, which could lead to transformative advances such as transplantable human organs harvested from livestock animals and virus-resistant cell lines for faithful antibody generation [2]. Work in budding yeast reported within the last six years has broken ground by generating the first entirely synthetic yeast chromosomes, combining the de novo synthesis approaches used in the prokaryotic system with recombination-mediated assembly of small chromosome building blocks [[3], [4], [5], [6]]. Since budding yeast chromosomes are tiny compared to their mammalian counterparts (the entire haploid budding yeast genome is 12 Mbp, while the “puny” human chromosome 21 is ~48 Mbp, and the entire haploid human genome is estimated to be ~3100 Mbp), one might envision that the substantial “scale up” will be the greatest challenge to translating success from yeast to mammalian genomes. That could prove true, but advances in DNA synthesis/assembly methodologies and instrumentation are likely to continue shortening this particular gap. However, other major challenges include the inclusion of essential mammalian chromosomal elements, the delivery of large artificial chromosomes into mammalian cells, and the efficiency in isolating cells that carry a functional artificial chromosome.

Over the last twenty years, the human artificial chromosome (HAC) field has contributed a wealth of information to circumvent some of these potential issues by defining the minimal components required for HAC propagation and inheritance through cell divisions alongside their natural chromosome counterparts. While these efforts have been reviewed previously [7,8] and elsewhere in this issue, we briefly mention the current possible “parts list” for HACs: telomeres, an origin of replication, and a centromere (Fig. 1). Telomeres are genetically encoded repeats that cap the ends of natural chromosomes in mammals and budding yeast, and the addition of telomere sequences to the ends of a linear HAC molecule is sufficient to protect HAC DNA ends [9]. However, telomeres are only needed if there actually are DNA ends, so one can also generate circular HACs that obviate this requirement [10,11](Fig. 1). While strict sequence-dependent origins of replication are required in budding yeast (the classic autonomous replicating sequence; ARS [12]), origins fire from a variety of sequences in mammals. Some differences in efficiency exist depending on sequence [13], but it remains unclear if a lack of efficient origin firing ever compromises HAC formation or stability, as long as sufficient lengths of mammalian genomic DNA is present to represent possible origin firing locations. The final requirement, the centromere, poses a greater challenge. This review will discuss efforts to use HACs as a tool to resolve these problems. In particular, HACs made from a non-repetitive DNA template can alleviate many of the challenges caused by the repetitive DNA typically found at human centromeres.