Selection of Antibiotic-Resistant Bacterial Strains and Identification of Genomic Alterations by Whole-Genome Sequencing: Using Staphylococcus aureus and Oxacillin Resistance as an Example

BACKGROUND

High-throughput sequencing technologies are powerful tools for the study of bacterial genetics, and bacterial genome sequencing is now routinely used to investigate antimicrobial resistance (Loman and Pallen 2015; Baker et al. 2018; Boolchandani et al. 2019; Davey and Valdivia 2020). To understand mechanisms behind antimicrobial resistance, studies are performed in which strains are passaged at low concentrations of an antibiotic or compound of interest to obtain resistant strains. The mutations acquired by bacterial strains following such selection screens can then be identified by whole-genome sequencing in a matter of a few days, which contrasts the months it would typically take when using classical gene mutation mapping methods. Such selection processes, followed by a whole-genome-sequencing approach, can provide genetic insights into how bacteria become resistant to certain antimicrobials and, thus, this approach can be helpful in understanding the mechanism of action of newly identified drugs. Furthermore, whole-genome sequencing has additional applications in bacterial genetics studies. For instance, it is now also frequently used as a final step in a bacterial mutant strain construction process, to confirm specific mutations and exclude other genomic alterations.

Illumina short-read sequencing runs (75–150-bp reads) are frequently used for genome-resequencing projects, and multiple strains can be sequenced in a single run. For instance, for genome-resequencing projects of bacteria such as Staphylococcus aureus (which has an approximate genome size of 3 Mb), we routinely sequence 24–30 strains in a single run using an MiSeq instrument and a standard MiSeq V2 Reagent 300-cycle kit (150-bp paired-end sequencing kit), with which it is possible to sequence 5.1 Gb. When sequencing 24 S. aureus strains under ideal sample-preparation conditions, this should yield a genome coverage of 70-fold for each strain sequenced. Using the protocols described in this collection, we typically obtain 50–60-fold coverage when sequencing 24–30 S. aureus strains (see Protocol: Agar Plate-Based Method for the Selection of Antibiotic-Resistant Bacterial Strains [Zeden and Gründling 2023a], Protocol: Preparation of Staphylococcus aureus Genomic DNA Using Promega Nuclei Lysis and Protein Precipitation Solutions, Followed by Additional Cleanup and Quantification Steps [Zeden and Gründling 2023b], Protocol: Preparation of Staphylococcus aureus Genomic DNA Using a Chloroform Extraction and Ethanol Precipitation Method, Followed by Additional Cleanup and Quantification Steps [Zeden and Gründling 2023c], Protocol: Small-Scale Illumina Library Preparation Using the Illumina Nextera XT DNA Library Preparation Kit [Zeden and Gründling 2023d], and Protocol: Bacterial Whole-Genome-Resequencing Analysis: Basic Steps Using the CLC Genomics Workbench Software [Zeden and Gründling 2023e]).

An important factor for successful sequencing projects is the quality of the input DNA sample. For bacterial strains that can be grown as pure cultures in laboratory medium, it is usually quite simple to isolate enough good-quality genomic DNA (gDNA) for the subsequent library preparation and sequencing run. As part of this collection, we introduce two different methods, which have been optimized for the isolation of gDNA from the Gram-positive bacterium S. aureus (see Protocol: Preparation of Staphylococcus aureus Genomic DNA Using Promega Nuclei Lysis and Protein Precipitation Solutions, Followed by Additional Cleanup and Quantification Steps [Zeden and Gründling 2023b] and Protocol: Preparation of Staphylococcus aureus Genomic DNA Using a Chloroform Extraction and Ethanol Precipitation Method, Followed by Additional Cleanup and Quantification Steps [Zeden and Gründling 2023c]). Even though these methods are optimized for S. aureus, they can easily be adapted for use with other bacteria. Additional alternative methods for the isolation of gDNA from bacteria are also included in this collection and are described in Protocol: Preparing Bacterial Genomic DNA (Figueroa-Bossi et al. 2023).

In general, isolation of gDNA from Gram-positive bacteria is slightly more involved than isolation of gDNA from Gram-negative bacteria, because of differences in their cell wall structure. To efficiently lyse Gram-positive bacteria, the cell wall either needs to be mechanically disrupted using homogenizers or bead beaters, or the peptidoglycan layer needs to be enzymatically digested using enzymes such as lysozyme, mutanolysin, or, for S. aureus, lysostaphin. Besides gDNA, many bacterial strains often harbor one or more plasmids. Depending on plasmid size and the gDNA purification method used, some plasmid DNA will copurify with the gDNA and, hence, will also be sequenced as part of the run. As mutations that cause specific phenotypes could potentially be located on plasmids, this can be advantageous. However, in the case of plasmid-containing S. aureus strains, we have found that choosing the appropriate gDNA isolation method is essential to avoid the copurification of too much plasmid DNA. Large amounts of plasmid DNA in a sample can result in too many sequencing reads mapping onto the plasmid and insufficient sequence coverage of chromosomal DNA.

Illumina short-read sequencing technology is used as part of the methods discussed in this collection. There are several different DNA fragmentation and library preparation methods that can be used. In the workflow introduced in this collection, gDNA is broken into fragments of ∼200–1000 bp in a so-called tagmentation reaction, during which a transposon cleaves double-stranded DNA and, at the same time, “tags” the DNA ends with two different types of universal sequences. The DNA fragments are subsequently amplified in a limited-cycle PCR using primers with sequences complementary to the added tags and, at the same time, each sample is tagged with specific barcode sequences that are also introduced as part of the primer sequences. This allows multiple samples to be pooled and sequenced as part of a single run. For the tagmentation reaction, it is essential that the correct amount of DNA is added to each reaction, as quantities of DNA that are too large or too small will affect the tagmentation process, resulting in fragments that are either too large or too small. We introduce additional gDNA purification and gDNA quantification steps for consistent tagmentation reaction results as part of the accompanying gDNA-purification protocols (see Protocol: Preparation of Staphylococcus aureus Genomic DNA Using Promega Nuclei Lysis and Protein Precipitation Solutions, Followed by Additional Cleanup and Quantification Steps [Zeden and Gründling 2023b] and Protocol: Preparation of Staphylococcus aureus Genomic DNA Using a Chloroform Extraction and Ethanol Precipitation Method, Followed by Additional Cleanup and Quantification Steps [Zeden and Gründling 2023c]).

Short-read sequences such as Illumina reads are well-suited for genome-resequencing projects when complete and annotated reference genomes of the bacterial strain used in an experiment are already available. Even if the genome of the strain used in a study is already available, however, it is always advisable, in any sequencing project, to resequence the wild-type strain used by a specific research group, as bacteria might have acquired additional mutations during propagation and long-term storage in a laboratory. By aligning the Illumina reads of the “laboratory wild-type strain” to a published reference genome, any background mutations present in the laboratory strain can be identified. Additionally, if the reference genome of the exact strain used in a study is not available and only the genome sequence of a related strain is available, the Illumina reads of the “laboratory wild-type strain” can be used to generate a new reference genome. This can be done as described in Protocol: Bacterial Whole-Genome-Resequencing Analysis: Basic Steps Using the CLC Genomics Workbench Software (Zeden and Gründling 2023e). Short-read sequences can also be used for de novo genome assembly, as, for instance, described in Protocol: De Novo Genome Sequencing, Annotation, and Taxonomy of Unknown Bacteria (Camilli 2023). In this case, Illumina reads are often combined with long-read sequencing technologies. Short-read sequences can also be used for high-throughput mutant screening, transposon-sequencing experiments, as described in Protocol: High-Throughput Mutant Screening via Transposon-Sequencing (Bourgeois and Camilli 2023).

In this collection, we introduce methods for the selection of antibiotic-resistant bacterial strains on agar plates and the subsequent identification of genomic alterations by whole-genome sequencing. We will use as a specific example the methicillin-sensitive S. aureus strain RN4220, and introduce a procedure for the selection of strains with increased oxacillin antibiotic resistance. The accompanying protocols can be easily adapted and applied to different bacterial species and other antibiotics. This approach can also be used with new antimicrobial compounds that have an unknown mode of action. In this case, the selection of resistant mutants and mapping of chromosomal mutations can aid in the identification of the mode of action of the compound used.

The procedure is carried out by performing the following protocols: Protocol: Agar Plate-Based Method for the Selection of Antibiotic-Resistant Bacterial Strains (Zeden and Gründling 2023a), Protocol: Preparation of Staphylococcus aureus Genomic DNA Using Promega Nuclei Lysis and Protein Precipitation Solutions, Followed by Additional Cleanup and Quantification Steps (Zeden and Gründling 2023b) (or by performing an alternative genomic DNA-isolation method better suitable for strains containing plasmids, Protocol: Preparation of Staphylococcus aureus Genomic DNA Using a Chloroform Extraction and Ethanol Precipitation Method, Followed by Additional Cleanup and Quantification Steps (Zeden and Gründling 2023c), Protocol: Small-Scale Illumina Library Preparation Using the Illumina Nextera XT DNA Library Preparation Kit (Zeden and Gründling 2023d), and Protocol: Bacterial Whole-Genome-Resequencing Analysis: Basic Steps Using the CLC Genomics Workbench Software (Zeden and Gründling 2023e).