This web page was produced as an assignment for an undergraduate course at Davidson College.
Summary: Total Synthesis of a Functional Designer Eukaryotic Chromosome (Annaluru et. al., 2014)
In this paper, Annaluru et. al. describe how they created a synthetic copy of chromosome III in the common model organism Saccharomyces cerevisiae (yeast). Their purpose was to produce a synthetic genome with loxPsym sites on either side of non-essential genes. The loxPsym sites are palindromic stretches of DNA which, when induced by the enzyme Cre, will recombine with each other such that the gene they flank is removed from the genome; thus, with a signal, the researchers could simulate evolution and induce the loss of non-essential genes. While bacterial and viral genomes have been synthesized in the past, this is a pioneering attempt to design a eukaryotic chromosome and find the minimal genes required for life by removing non-essentials. As their model for the synthetic chromosome, Annaluru et. al. chose yeast’s native chromosome III, which is 316,617 base pairs and is the location of the MAT gene. The MAT gene has two alleles, α and a. Yeast which are haploid and have the α allele will be α-mating type, and those with the a will be a-mating type; the diploids do not mate. This MAT locus proved useful in later testing the stability of the synthetic chromosome.
In order to create the synthetic chromosome (synIII), Annaluru et. al. first created building blocks (BB’s) by linking oligonucleotides, short DNA stretches of about 60-79 nucleotides each, with PCR (polymerase chain reaction—rapid heating and cooling cycles with a heat-stable DNA polymerase allows for quick DNA synthesis and copying). These ~750 bp BB’s were then merged by recombination to create 2-4 kb DNA minichunks which contained overlapping portions. These overlapping minichunks, having been transformed into yeast, were recombined into the native chromosome III along with genetic markers in order to distinguish the cells with successfully altered chromosomes.
Once synIII had been created, Annaluru et. al. tested their success by comparing the genome of synIII with native III. They used PCRTag analysis, where small primers are able to latch onto and identify desired sequences, to ensure that synIII had successfully been constructed and was in the right place. They found an absence of the native sequences in synIII yeast but all of the synIII tags, and vice versa in the WT yeast. Annaluru et. al. then considered the size of the chromosome, as synIII should be smaller than its native cousin due to removed sequences. They found that synIII was indeed smaller than native III, and that an intermediate step in the construction of synIII was slightly smaller than native III, which logically made sense. Sequencing of synIII revealed a few differences in the actual sequence from the synIII sequence that had been designed; however, many of these mutations were insignificant and explainable by simple chance and error in the recombination construction method. Finally, to ensure that synIII was not harmful to the yeast cell and did not result in growth defects, Annaluru et. al. plated WT, synIIIL (an intermediate construct) and synIII yeast on varying media and conditions. They found that the negative effects of synIII on fitness were very small, as the colonies were in general the same size and grew with the same success. Native III and synIIIL were perhaps slightly more fit, but the researchers indicate that they did not find it significant.
In order to know whether their work had been successful, the researchers also needed to test the stability of the chromosome, especially considering the presence of the loxPsym sites, the removal of many repetitive DNA sequences, and the removal of several tRNA genes. So in the absence of the enzyme Cre, which would have induced the recombination of loxPsym, they measured the loss of synIII over approximately 125 generation in 30 strains and found that it was not lost significantly more than the native chromosome. This was done by testing whether the site for MATα, located on synIII, was lost and thus allowed yeast to become a type a-mater due to the MATa allele on its native III chromosome. Annaluru et. al. then tested their loxPsym site-design by inducing expression of the enzyme Cre, causing the loxPsym sites to recombine (they refer to this process as SCRaMbLEing) in yeast that were synIII/III diploids. They found that this did indeed cause genes, such as the MAT locus, to be lost; this was noted by the increase of a-mater yeast, since the loxPsym site cut the MATα site from synIII, making the “diploid” yeast suddenly haploid for that locus with the “a” allele and able to mate.
Of the details of the paper, I enjoyed the figures the most—they were easy to understand, well organized, and well-colored. Some things were poorly explained in the captions—what “KANR,” “G418R,” “G418S” are in Fig. 2C, for example, or what the “*” indicates in Fig. 3B. Additionally, there were a few processes that were unclear from the text, such as how the loxPsym sites function, but I recognize that this paper is directed at an audience that would have a better understanding of the mechanism. The result I am most wary of is the determination that synIII had little to no effect on fitness. The researchers support this conclusion with photographs in Fig. 3C, but it is easy to choose the best photograph for the paper and it is unclear whether this is a truly typical result.
I thought this paper was fascinating, since we now have the ability, however limited, to create a synthetic chromosome. Granted, this was done in yeast with a very small chromosome, but it represents an incredible leap forward in synthetic biology. But with it come many questions for the future. Could we design plant chromosomes to yield the right amount of nutrients, or to be more resilient to weather and climate change? Could we engineer little bio-bugs to eat pollutants? Could we inject a small, designer chromosome into an egg or an embryo and let it make up for natural genomic defects? Whatever the use, creating stable synthetic chromosomes has fantastic potential. We should be wary, though, of overstepping ethical bounds in our zeal for synthetic biology; while intentions may be good, it might prove all too easy to create our own Brave New World.
Figures
(All figures below courtesy of Annaluru et. al., 2014)
Figure 1: SynIII design
- Panel A: The S. cerevisiae III chromosome and the synIII chromosome are both shown here. While the III and synIII chromosomes share a repeat as well as non-essential, uncharacterized, and dubious Open Reading Frames (ORFs), the synIII chromosome also has loxPsym sites before non-essential ORF YCR098C and after the repeat, and an added universal telomere cap.
- Panel B: As an example, this shows us how the insertion of a loxPsym site—a palindromic DNA sequence—is inserted into an untranslated region of the 3’ strand, right next to a non-essential gene (in this figure, non-essential ORF YCL055W).
- Panel C: This figure shows the simple change in an essential ORF (YCL004W) from TAG toTAA creates a synthetic stop codon in synIII.
Figure 2: SynIII construction
- Panel A: This shows the first step in building synIII—creating the small BB’s from oligonucleotides. These oligonucleotides (60-79 nt’s long) were linked by PCR into BB’s, which could then be ligated into a plasmid that was transformed.
- Panel B: Here we see a depiction of the BB’s from the plasmids and a shuttle vector being combined and cotransformed; from there, recombination occurs so that the BB’s (which contained overlapping pieces of DNA to allow for recombination) merge to form a minichunk of 2-4 kb. “RE” denotes a restriction enzyme cutting site, for either the rare XmaI or NotI, placed on either side of the minichunk.
- Panel C: This shows how the minichunks were used to replace parts of the natural yeast chromosome III, creating synIII. The minichunks, designed to overlap, were recombined such that they merged together, and then into the chromosome. In transformation 1, minichunks 1-4 recombined, with genetic marker LEU2 and a native yeast chromosome III sequence (Linker L1) attached to minichunk 4; Linker L1 allowed the merged minichunks to recombine with chromosome III, producing a hybrid identifiable by the genetic marker LEU2. In yeast transformation 2, similar steps occurred, this time merging minichunks 4-13, with Linker L2 (genetic marker URA3 and a native chromosome III sequence) attached to minichunk 13. Minichunk 4 and Linker L2 allowed for recombination into the hybrid product of transformation 1, producing a synIII section that has replaced parts of native chromosome III with minichunks 1-13 and identifiable by the presence of genetic marker URA3 and absence of LEU2.
Figure 3: Characterization and Testing of the synIII strain
- Panel A: This panel shows us that when the left arm of WT yeast chromosome III was tested, there were matches for every WT PCRTag but no matches for synIII PCRTags; the opposite result was obtained when the left arm of synIII was tested, with no WT matches and all synIII matches.
- Panel B: This panel is a gel which has labeled the bands of native III, synIII left arm, synIII, and native IX, VI, and I. In the far right gel lane, we can see that the synIII left arm is slightly smaller than the native III; synIII is even smaller, and is the same size as native VI. The purpose of the “*” on synIII left arm is unclear.
- Panel C: This panel shows us the phenotype and survival of WT, synIIIL, and synIII in serial dilutions on various media and over time. For those grown for 2 days on yeast extract peptone dextrose (YPD), the yeast grown at 25oC look roughly the same no matter their chromosome; synIIIL looks to be perhaps slightly more successful. Similar results hold true for 30oC and 37oC, though colonies are larger across the board as temperature rises. For the rest of the plates (yeast extract peptone glycerol ethanol for 3 days at 30oC, pH of 4.0 or 9.0, both for 2 days at 30oC, and 0.05% methyl methanosulfate for 3 days at 30oC), the general trend seems to be that strains with WT, synIIIL, and synIII are about equal in growth and survival; if anything, the synIII strain is slightly less viable, but synIIIL is comparable to WT.
Figure 4: Genomic stability of the synIII strain
- Panel A: Panel A depicts the methods used to test 30 strains of yeast containing the synIII chromosome for the ability to keep synIII over about 125 generations, a statistic measured by assaying cells for 58 distinct segments lacking essential genes; synIII proved remarkably stable with no deletions occurring.
- Panel B: This panel compares the average loss rate of synIII to the average loss rate of native III; there proved to be no significant difference between the two, indicating that synIII is about equally likely to be lost as the native chromosome is.
- Panel C: The left part of Panel C shows us how SCRaMbLEing a heterodiploid (synIII/III) that does not mate can cause a deletion in part of the synIII chromosome that contained a MATα site. The loss of the MATα site leaves these cells with only one copy of the gene, the MATa allele, and it can therefore now mate. The table on the right part of Panel C describes the frequency of mating types after SCRaMbLEing: for synIII/III strains, 45% became a-maters, while none were α-maters, since it was the MATα site that was lost from synIII in the process. For III/III strains, which were also heterozygous for both MAT alleles, there were no mating cells, since nothing was lost in the process and diploid cells do not mate.
References
Annaluru, N et. al. Total Synthesis of a Functional Designer Eukaryotic Chromosome. Science, 2014. 344:55-58.
Mutational Analysis Reveals the Origin and Therapy-Driven Evolution of Recurrent Glioma
Genomics Page
Biology Home Page
© Copyright 2014 Department of Biology, Davidson College, Davidson, NC 28035