Context-dependent neocentromere activity in synthetic yeast chromosome VIII

Abstract

Pioneering advances in genome engineering, and specifically in genome writing, have revolutionized the field of synthetic biology, propelling us toward the creation of synthetic genomes. The Sc2.0 project aims to build the first fully synthetic eukaryotic organism by assembling the genome of Saccharomyces cerevisiae. With the completion of synthetic chromosome VIII (synVIII) described here, this goal is within reach. In addition to writing the yeast genome, we sought to manipulate an essential functional element: the point centromere. By relocating the native centromere sequence to various positions along chromosome VIII, we discovered that the minimal 118-bp CEN8 sequence is insufficient for conferring chromosomal stability at ectopic locations. Expanding the transplanted sequence to include a small segment (∼500 bp) of the CDEIII-proximal pericentromere improved chromosome stability, demonstrating that minimal centromeres display context-dependent functionality.

Keywords: CRISPR; Saccharomyces cerevisiae; aneuploidy; centromere; chromosomal stability; genome engineering; genome rearrangements; pericentromere; synVIII.

Graphical abstract

Figure 1

Design and assembly of synVIII (A) SynVIII hierarchical assembly workflow. Building blocks, assembled from overlapping ∼70 bp oligonucleotides as part of Sc2.0’s Build-A-Genome course, were used to generate single minichunks ranging from ∼2 to 4 kb in size. Homologous recombination between overlapping minichunks resulted in ∼10-kb chunks that were ligated together after restriction enzyme digestion to form megachunks A–N. oligos, oligonucleotides; RE, restriction enzyme; bp, base pairs; kb, kilobase pairs. (B) Schematic representation of several Sc2.0 design features within megachunk D. An essential gene (red) does not contain a loxPsym site in the 3′-end of the UTR. (C) WGS analysis revealed a duplication in megachunk G. (D) The duplication was repaired using a two-step CRISPR-Cas9 approach. Paired red and black arrows indicate the approximate binding locations of primers that bind two distinct locations in strains with the duplication. PCR results shown are representative of five technical replicates. WT, wild type. (E) Removal of duplicate sequences restored the read depth profile. (C, E) Read depth was calculated for non-overlapping 500-bp windows and normalized to the median depth across the chromosome. Arrows represent decreased sequencing depth at the CUP1 locus, which is present in single copy in synVIII but present in multiple copies in wild-type yeast and the S288C reference.

Figure 2

Debugging and characterization of synVIII (A—C) 10-fold serial dilution spot assays of synVIII strains, with BY4741 serving as a wild-type control. (A) Minichunk C1.13 contains a 1-bp deletion in the gene TDA3, which leads to a fitness defect on YPD medium at 22°C and 25°C in strains that contain the minichunk and yeast_chr08_9_4, which contains all of megachunk C. (B) The fitness defects observed on YPD medium at 22°C for synVIII are similar to defects observed in a tda3 knockout that is otherwise genetically identical to BY4741. (C) The final version of synVIII, yeast_chr08_9_11, which includes the repaired version of TDA3, grows comparably well with wild-type yeast under eight different growth conditions. YPD, yeast extract peptone dextrose; YPG, yeast extract peptone glycerol; SC, synthetic complete. (D) Volcano plot illustrating gene expression differences between synVIII and BY4741. Upregulated genes on chromosome VIII are in red and downregulated genes on chromosome VIII are in blue. FLO5 is a “repeat-smashed” gene shown in teal (these genes were pervasively recoded using GeneDesign’s RepeatSmasher^, and thus do not provide accurate RNA-seq data), CUP1 and associated genes are in yellow, and DEGs on other chromosomes are in purple. The fold change cutoff is 4, with a p value cutoff of 0.01. (E) Schematic illustration of the CUP1 locus with arrows representing primers used for diagnostic PCR. The red arrow indicates a primer that binds in two distinct locations due to the presence of homologous sequence. (F) Diagnostic colony PCR demonstrates that YHR054C is deleted. Four independent colonies were tested for each strain. (G) A diploid homozygous for synVIII creates viable spores after tetrad dissection.

Figure 3

Centromere transplantation via SSICT results in persistent aneuploidy regardless of strain background or position (A) Schematic illustration of ectopic centromere positions on chromosome VIII and the outcomes of successful SSICT at each position. SV, structural variant. (B) Schematic representation of the SSICT method, which uses two to three sgRNAs targeting native and ectopic positions to simultaneously delete nCEN8 and integrate eCEN8 via homologous recombination. (C) Read depth profiles analyzed from WGS data indicate that chrVIII is present in two copies after successful SSICT for two representative strains. Results are similar for left arm telocentric transplants in the wild-type background and proximal transplants in the synVIII background; see Table S5. (D) Three different methods were used to attempt to destabilize or lose the aneuploid chromosome of the left arm telocentric eCEN8 strain ySLL260. Copy number was calculated by dividing the median depth of each chromosome by the median depth of the genome. (E) Schematic illustration of fusion chromosome IX-III-I generated in Luo et al. before and after SSICT. Different colors represent each chromosome: chrI (red), chrIII (gold), and chrIX (teal). (F) Read depth profiles analyzed from WGS data indicate that chrIX-III-I in the fusion strain and chrIX in the wild-type background are present in two to three copies. (C and F) Relative depth was calculated by determining the median read depth, calculated from reads per position, across each chromosome relative to the genome average. Reads were randomly downsampled, and plots were modified for presentation purposes.

Figure 4

SSICT results in structural variation and hybrid centromere contexts (A) Two representative insets of Hi-C contact maps for strain ySLL223, obtained after a metacentric transplantation attempt. In the left-hand map, the expected structure of the metacentrically located centromere, designed using the S288C reference sequences of chromosomes VIII (light blue; modified to include the ectopic CEN8 sequence) and IX (gray), are schematized; in the right-hand map, the corrected chrVIII reference sequence in which the approximate position of the inverted sequence, as revealed by whole-genome sequencing, is indicated by double arrows. The full Hi-C map is included as Figure S7. Schematics of chrVIII and IX are shown on the x and y axes; arrowheads indicate left and right end of the intervening inverted sequence between the native and ectopic CEN8 positions. The Hi-C maps were generated from 5-kb bins; violet to white color scale represented by the right-hand panel reflects high to low contact frequency (log10). (B) 3D average representations of the right hand Hi-C map in (A). chrVIII is shown in light blue, and the inferred native and ectopic CEN8 positions are colored in yellow and orange, respectively. The remaining chromosomes and centromeres are shown in light gray and black, respectively. See Video S1. (C) Schematic representation of wild-type chrVIII including the native context of CEN8 and the context of the metacentric and right arm telocentric ectopic positions. (D) Schematic illustration of the hybrid centromere context accounting for the ∼100-kb inversion in strain ySLL223. (E) Schematic illustration of the hybrid centromere context generated for ySLL224, which was characterized after a right arm telocentric transplantation attempt. Inverted duplication of the left arm forms an isochromosome, a structure confirmed by WGS. (F) Read depth profile for ySLL224. An absence of reads at the right telomere indicates a deletion, whereas the left arm has been duplicated. Relative depth was calculated by determining the median read depth, calculated from reads per position, across chrVIII relative to the genome average. Reads were randomly downsampled, and the plot was modified for presentation purposes. Arrowhead represents increased depth at the CUP1 locus (inset), which is known to be present in multiple copies in wild-type yeast. (G) Diagnostic colony PCR provides further evidence for structure of the isochromosome. Two independent colonies of ySLL224 were compared with the wild-type strain BY4741. Red arrows represent primers that bind to CDEIII of CEN8. WT, wild-type.

Figure 5

Pericentromeric sequences improve chromosome stability (A) An ade2 ade3 derivative of the wild-type strain BY4741 was transformed with CEN8-containing minichromosomes, selected and maintained on SC-Ura, and then plated to SC medium with low adenine concentration. Red or partially red colony color shown here on representative SC plates indicates that the minichromosome is maintained, even in the absence of selection. The original construct contains CEN8 and 5 kb total of pericentromeric sequence, the minimal construct contains only the 118-bp CEN8 sequence, and the remaining constructs contain 500 or 1.5 kb of pericentromeric sequence. The red colony area was quantified using ImageJ. Averages and standard deviation were calculated from a total of five biological replicates per construct. Raw data are available in Table S7. (B) Repeating SSICT at the left arm telocentric location with 500 bp of CDEIII-flanking pericentromeric sequence decreased the incidence of chrVIII aneuploidy. Copy number was calculated by dividing the median depth of each chromosome by the median depth of the genome.