A new platform for ultra-high density Staphylococcus aureus transposon libraries

Background Staphylococcus aureus readily develops resistance to antibiotics and achieving effective therapies to overcome resistance requires in-depth understanding of S. aureus biology. High throughput, parallel-sequencing methods for analyzing transposon mutant libraries have the potential to revolutionize studies of S. aureus, but the genetic tools to take advantage of the power of next generation sequencing have not been fully developed. Results Here we report a phage-based transposition system to make ultra-high density transposon libraries for genome-wide analysis of mutant fitness in any Φ11-transducible S. aureus strain. The high efficiency of the delivery system has made it possible to multiplex transposon cassettes containing different regulatory elements in order to make libraries in which genes are over- or under-expressed as well as deleted. By incorporating transposon-specific barcodes into the cassettes, we can evaluate how null mutations and changes in gene expression levels affect fitness in a single sequencing data set. Demonstrating the power of the system, we have prepared a library containing more than 690,000 unique insertions. Because one unique feature of the phage-based approach is that temperature-sensitive mutants are retained, we have carried out a genome-wide study of S. aureus genes involved in withstanding temperature stress. We find that many genes previously identified as essential are temperature sensitive and also identify a number of genes that, when disrupted, confer a growth advantage at elevated temperatures. Conclusions The platform described here reliably provides mutant collections of unparalleled genotypic diversity and will enable a wide range of functional genomic studies in S. aureus. Electronic supplementary material The online version of this article (doi:10.1186/s12864-015-1361-3) contains supplementary material, which is available to authorized users.

transcriptional terminator of the erythromycin resistance gene by inverse PCR (primers Tm167-170) of a stock plasmid, 5'-phosphorylation with T4 polynucleotide kinase, and self-ligation.
Finally, a construct containing dual outward facing promoters (pTM244) was made in two steps by both ligating the P pen promoter element and removing the transcriptional terminator as described.
The Φ11-FRT phage variant, which does not express a functional integrase gene and lacks the attP site, was created by sequentially exchanging each of the attL/attR attachment sites of the Φ11 prophage in HG003 with the FRT site specific recombination sequence from Saccharomyces cerevesiae [2,3]. The allelic exchange vectors pTM204attLint and pTM204attR were constructed by 3-piece overlap assembly of 1-kb DNA regions flanking the respective att site with EcoRI/HindIII digested pKFC using the In-Fusion Kit from Clontech. The FRT DNA sequence was introduced during PCR into the 5'-tail of the P2-P3 primers (Table S3). Allelic exchange at both sites in strain TM222 was confirmed by PCR using flanking primers. The FLP expressing vector pTM195 was next constructed by ligating the FLP recombinase gene, amplified with primers Tm79-Tm80 from pCP20, to the pLI50-Ppen E. coli-S. aureus shuttle vector digested with BamHI/AscI. The att-defective prophage of TM222 was induced by electroporation of pTM195, which constitutively expresses the FLP recombinase, and supplementation of outgrowth media with 1 µg/mL of mitomycin C. After 3 hours of growth at 30°C, cells were removed by centrifugation and the serial dilutions of supernatant were added to an RN4220 top agar overlay. Discrete plaques were isolated, and the expected attP-int::FRT locus was confirmed by sequencing. To selectively excise the Φ11 prophage in HG003, TM222 was likewise utilized except an aliquot of the pTM195 electroporation cell suspension was directly plated on TSB agar (Cm 10 µg/mL) without prior mitomycin C treatment. Colonies containing plasmid were then subcultured overnight in 10 mL of TSB to allow for Φ11 prophage excision and plasmid loss, before being plated to single colonies. Colonies were replica plated to check for Cm sensitivity, indicating loss of pTM195, and sensitivity to Φ11 infection. The Φ11::FRT locus of one colony with the expected phenotype, TM226, was confirmed by DNA sequencing.
Gene deletion and replacement with the kan R gene were done as previously described with few modifications [4]. Briefly, primers lyrA1-6 or mprF1-6 were used to amplify 1000bp upstream and 1000 bp downstream of the target gene to be deleted as well as the kan R fragments. Equimolar concentrations of the three fragments (-1000, kan R , and +1000) were spliced by PCR, digested with BamHI-HF and SalI-HF (NEB) and ligated into the pKFC plasmid digested with the same restriction enzymes. RN4220 was transformed with each plasmid and deletion/replacement of the target gene with kan R was performed by single crossover integration of the plasmid at 42°C and curing of the plasmid at 30°C. Deletion/replacements were verified by PCR and genomic DNA sequencing and subsequently transduced into S. aureus HG003 using phage Φ85.
Preparation and sequencing of transposon library Unless otherwise described, reactions are mixed and incubated in 1.5 mL Eppendorf LoBind tubes. DNA was stored at -20°C overnight, and at -80°C long term. At least 10 μg of high molecular weight genomic DNA was purified using the GES protocol [5]. 10 μg of genomic DNA was digested with 50-100 U NotI in a 600 μl reaction in NEB Buffer #3 supplemented with BSA. The reaction was vortexed gently, mixed by inversion, spun down, and incubated at 37°C for seven hours (mixing and spinning down once halfway through). NotI was inactivated at 70°C for 20 minutes, and cooled to room temperature (5 minutes).
200μl of 4x precipitation buffer was added to the 600μl digest reaction and vortexed and inverted (not pipetted) to mix. This reaction was incubated in an ice water bath in a 4°C cold room for 12-16 hours. After incubation, the reaction was spun down in a tabletop centrifuge at maximum speed for 20 minutes at 4°C. Then, the supernatant was removed, and the precipitated DNA was washed once with cold 1x Precipitation buffer, spinning down the DNA at maximum speed for 10 minutes at 4°C. To further purify the DNA, pellets were washed twice with room temperature 70% ethanol, spinning down the DNA at maximum speed for 5 minutes at room temperature between washes. The final DNA pellet was dried and resuspended in 50-100μl standard elution buffer (10mM Tris, pH 8.5) by pipetting. seconds, and 70°C for 30 seconds; and 12°C for storage. CES buffer consists of 2.7M betaine, 6.7mM DTT, 6.7% DMSO, and 55μg/ml BSA [8]. Beginning at the end of 9 cycles, a tube from each category of reaction (undigested with experimental primers, digested with experimental primers, undigested with control primers, digested with control primers) was removed from the thermocycler every three cycles and quenched with DNA gel loading buffer supplemented with 0.1% SDS. Samples were stored at -20°C until all cycles were completed. The final tubes were removed at the end of the 27 th cycle. These reactions were run on a 1% agarose gel in TAE buffer. In a sample deemed sufficiently digested, a six cycle product detection threshold difference between the undigested and digested samples when using the experimental primers (785 bp) in comparison to the control primers (1298 bp) was required ( Figure S3).
Once we confirmed that the DNA had been sufficiently digested, the biotinylated adaptors were ligated. 50 μM of annealed adaptors were prepared by mixing 15 μl of 100 μM TM214 and 15 μl of 100 μM TM215 with 1.5 μl 1M NaCl. This reaction was incubated in a thermocycler at 95°C for 5 minutes followed by cooling to 4°C at a rate of 0.1°C/second. Each Another size-selective precipitation was performed to remove un-ligated adaptors from genomic DNA. 50 μl of the same 4x PEG solution described above was added to the ligation reaction, and the tube mixed by vortexing, spun down, and incubated at 4°C or in the cold room for 12-16 hours. After incubation, the DNA is washed and dried in the same manner as described above after the NotI digest. MmeI-digested DNA was ligated to streptavidin dynabeads via the biotinylated adaptor.
This required three buffers. The 2x BandW Buffer consists of 2 M NaCl, 10 mM Tris, and 1 mM EDTA pH 7.5 with concentrated HCl [9]. LoTE Buffer consists of 3 mM Tris, 0.2 mM EDTA pH 7.5 with concentrated HCl [9]. LoTE+Tween Buffer is the same as LoTE buffer, but supplemented with 0.05% Tween 20. 200 μl of 2x BandW buffer was added to each sample. 32 μl/sample of Dynabeads® M-280 Streptavidin beads was added to a LoBind tube and placed in the magnetic particle collector (MPC). Then, the supernatant was removed, and the beads were washed three times with 1 mL of 1x BandW buffer. Finally, the sample was resuspended in 32 μl/sample of 1x BandW buffer, and 32 μl of beads were added to each diluted MmeI digest sample. These were incubated at room temperature for one hour, resuspending by tapping and inversion every 10-15 minutes LIB_AdaptB_5_long, LIB_AdaptT_6_long, LIB_AdaptB_6_long) annealed to each other (T to B) as described above. We would like to note that neither the barcodes in these adaptors nor the transposon construct-specific barcodes use an error-correcting barcode sequence [10,11].
Without error correcting barcodes, it is possible that sequences could be mis-assigned to the wrong sample or wrong transposon construct due to errors in sequencing. However, because we only used sequences with a high quality score in our analysis, we assume that the fraction of      has a dramatic growth defect at high temperatures, while ΔlyrA grows more rapidly than wild type. Table S1. Gene Essentiality in HG003

ADDITIONAL TABLES
Please refer to Additional File 2