An Optimized Genotyping Workflow for Identifying Highly SCRaMbLEd Synthetic Yeasts

Synthetic Sc2.0 yeast strains contain hundreds to thousands of loxPsym recombination sites that allow restructuring of the Saccharomyces cerevisiae genome by SCRaMbLE. Thus, a highly diverse yeast population can arise from a single genotype. The selection of genetically diverse candidates with rearranged synthetic chromosomes for downstream analysis requires an efficient and straightforward workflow. Here we present loxTags, a set of qPCR primers for genotyping across loxPsym sites to detect not only deletions but also inversions and translocations after SCRaMbLE. To cope with the large number of amplicons, we generated qTagGer, a qPCR genotyping primer prediction tool. Using loxTag-based genotyping and long-read sequencing, we show that light-inducible Cre recombinase L-SCRaMbLE can efficiently generate diverse recombination events when applied to Sc2.0 strains containing a linear or a circular version of synthetic chromosome III.


S 2
Notably, no negative control can be performed because the primer pairs can bind in both the synthetic and wild type chromosome III.Gel scan shows individual bands of the loxTags with the expected sizes (cf.Table S5).Amplicons spanning 2 loxPsym are indicated in orange, the reason for this is the small LU sizes ranging from 52 bp to 191 bp. and is therefore a proxy for Cre activity.Y1510-Y1512 cells were grown in darkness for 6 h.Induced samples were cultured in medium containing 25 µM PCB and irradiated by a 5-min red light pulse, followed by 10-sec red light pulses every 5 min for 1 h, 2 h and 4 h, respectively.Non-induced samples were cultured in darkness for 1 h, 2 h and 4 h, respectively.After the indicated time, serial dilutions were performed and 20 µL of each sample was spotted on SC-Leu plates.Each experiment was performed in three independent biological replicates.One representative replica is shown.While no obvious difference in cell growth was observed for Y1510 as expected, the colony number decreased sharply after only one hour of red light treatment of Y1511 and Y1512.Quality scores for de novo linear synIII assemblies and (C) circular synIII de novo assemblies sorted by strain names, respectively.In general strains with a higher sequencing depth and N50 scores show a lower number of small-scale errors.The QV value for multiple strains reflects the difficulty to solve highly SCRaMbLEd synthetic chromosome sequences with existing de novo assembly tools.This highlights the need for dedicated tools to assemble highly SCRaMbLEd synthetic yeast strains.QVs are calculated based on small-scale errors (collapse and expansion) divided by the total base pairs of the assembly. 1Strains with no QV (NA) lack structural and small-scale error and most likely are assembled correctly.arm of chrVI that is not visible in pulsed-field gel electrophoresis (see Fig. S6), and none of the progeny shows this pattern.We conclude this may be an artifact of the relatively low sequencing coverage of the strain (14-fold).Furthermore, this strain does not have a whole genome duplication (see Fig. S13).(C) to

Figure S1 |
Figure S1 | LoxTag amplicon test with standard Taq polymerase.Initial test of the loxTags with standard Taq polymerase (OneTaq 2X Master Mix, NEB) and synIII genomic DNA as template.For maximal resolution and sensitivity the Typhoon RGB laser scanning system was used to identify unspecific binding.

Figure S2 |
Figure S2 | Cell survival assay after red light-activated L-SCRaMbLE.In order to determine an appropriate induction time, L-SCRaMbLE was applied on wild type BY4742 (Y1510), linear synIII (Y1511) and circular synIII (Y1512), each carrying pLH_Scr15 with the expression cassettes of the red light-regulated Cre recombinase.Reduction of colony numbers results from the loss of essential genes caused by L-SCRaMbLE

Figure S3 |
Figure S3 | LoxTag screening of L-SCRaMbLE yeast isolates obtained without selection pressure.Y1511 was induced via red light for 1 hour (10-s red light pulses applied every 5 min) and resulting cultures were plated on SC.Twelve colonies were randomly selected and analyzed by loxTag screening (initial methodology).Eight candidates were selected for long-read sequencing based on a high number of observed variations and clear differences between the selected candidates.These candidates are represented by their corresponding unique identifier (SLy241 to SLy248).The candidates with numeric identifiers are not further investigated.Orange loxPsym junctions indicate loxTags which span 2 loxPsym, essential LUs are indicated by underlined numbers.Red boxes indicate crossing of the threshold indicating the presence of amplicons.Gray boxes indicate the absence of the respective amplicon because the threshold has not been crossed.

Figure S4 | 6 Figure S5 | 8 Figure S7 |Figure S8 | 11 Figure S9 |
Figure S4 | Comparison of qPCR analyses using either loxTags or PCRTags.Eight SCRaMbLEd synIII strains SLy241 to SLy248 and the original synIII strain SLy066 as a reference were analyzed via qPCR with either loxTags or PCRTags following the optimized qPCR methodology.The PCR results of the reference strain SLy066 indicate that all loxTags generate an amplicon as expected.The results of the SCRaMbLEd strains show that loxTags indicate more potential SCRaMbLE events, in contrast to PCRTags as intended.Gray boxes indicate presence and white boxes absence of amplicon.LUs without PCRTags are indicated in red and loxTags spanning two LUs are indicated by larger sizes of boxes.

S 12 [ 16 Figure S10 |
Figure S10 | Normalized coverage plots for all SCRaMbLEd strains with aneuploidy or whole genome duplication in comparison to the parental strain.(A) Linear (SLy066) and (B) circular (SLy117) synIII parental strains.The circular synIII in SLy117 shows a generally reduced abundance in comparison to the other chromosomes and the linear synIII in SLy066.SLy117 shows a potential amplification of the right

(Figure S11 |
Figure S11 | Ploidy analysis of SCRaMbLE isolates and reference strains by flow cytometry.(A) DNA content of control strains and parental strains (SLy066 and SLy117) used for L-SCRaMbLE.The parental strains are haploid and have no aneuploidies based on our NGS sequencing results (data not shown).Control strains visualizing known ploidies and serving as standard for the flow cytometry based DNA content determination.Haploid (BY4741), diploid (BY4743) and tetraploid (YCy2990; Schindler & Cai unpublished).Parental strains show the same DNA content as BY4741 and are consequently haploid.(B) Flow cytometry analysis of SCRaMbLEd linear synIII strains SLy241-248.SLy242, SLy246 and SLy248 of the linear synIII derivatives show a diploid DNA content after SCRaMbLE and indicate a whole genome duplication (no aneuploidies are detected in the NGS data).All remaining strains have a haploid genotype except for SLy245 where an aneuploidy was detected by long-read sequencing (cf.Fig. 5).(C) Flow cytometry analysis of SCRaMbLEd circular synIII strains SLy249-259.None of the circular synIII derivativesindicate a whole genome duplication, all strains are haploid.However, a partial chromosome duplication (SLy251) and aneuploidies were detected by long-read sequencing (SLy255 and SLy257; cf.Fig.5).

Plasmids constructed and used in this study.
pLH_Scr15CEN/ARS plasmid with expression cassettes for PhyBNT-CreN and PIF3-NLS-CreC fusion proteins for red/far-red light-regulated Cre recombinase activity in Saccharomyces cerevisiae.

Oligonucleotides used in this study.
All loxTags (incl.synIII) and the roxTags are provided with the supporting data (Supporting Data S1).
Tab. S5 | synIII loxTags used in this study.*primerpairs spanning two loxPsym sites are indicated in orange S 29Tab.