Defining essential genes and identifying virulence factors of Porphyromonas gingivalis by massively-parallel sequencing of transposon libraries (Tn-seq)

Summary

Porphyromonas gingivalis is a keystone pathogen in the development and progression of periodontal disease. Obstacles to the development of saturated transposon libraries have previously limited transposon mutant-based screens as well as essential gene studies. We have developed a system for efficient transposon mutagenesis of P. gingivalis using a modified mariner transposon. Tn-seq is a technique that allows for quantitative assessment of individual mutants within a transposon mutant library by sequencing the transposon-genome junctions and then compiling mutant presence by mapping to a base genome. Using Tn-seq, it is possible to quickly define all the insertional mutants in a library and thus identify non-essential genes under the conditions in which the library was produced. Identification of fitness of individual mutants under specific conditions can be performed by exposing the library to selective pressures.

1. Introduction

Porphyromonas gingivalis, a Gram-negative, anaerobic, asaccharolytic, black-pigmenting bacterium, is a keystone pathogen in the development and progression of periodontal disease (1–3). While many important virulence factors of P. gingivalis have been identified, tools for high-throughput screening of genic and intergenic function have been limited until recently (4–7). Transposon mutagenesis, an important tool for creating inactivation mutants, utilizes a mobile genetic element to generate insertional mutations. The value of transposon mutagenesis is based on the ability of the transposable element to randomly insert into sites throughout the host genome, while giving only one insertion per strain. While transposon mutagenesis has been used to great effect in many bacterial species, early attempts to create transposon mutant libraries of P. gingivalis using transposons based on Tn4435 and Tn4400 were limited by preferential insertion into ‘hot-spots’, cryptic vector insertions, and a limited range of P. gingivalis strains that could be mutagenized (8–10). The Mariner transposon, first identified and described in Drosophila melanogaster, has been highly effective for in vitro bacterial mutagenesis in phylogenetically, physiologically, and metabolically diverse species(11–13). The only well documented constraint or preference for Mariner transposition is preferential insertion into ‘TA’ nucleotide sequences, which are abundant throughout genomes even in the face of GC-content skewing. A recent advance in analysis of bacterial transposon libraries has been the application of massively parallel sequencing technology to quantitatively identify the location of the transposon insertions in a library pool by sequencing of the transposon-genome junctions (14, 15). These strategies have been variously named Tn-seq, HITS, or INSeq (16–18). In this chapter, we describe modifications in the processes of transposon library creation and screening by Tn-seq that we developed to optimize the systems for use with P. gingivalis (19). Steps involved include library creation, testing of library complexity by nested semi-random PCR, preparation of the library for high-throughput screening, and bioinformatic analysis of outputs from Tn-seq.

2. Materials

2.1 Transposon Mutagenesis (16, 19)

1. Bacterial strains and plasmids

   • P. gingivalis
   • E. coli S17-1 λpir
   • Plasmids pSAM and pSAM_Bt
2. Media

   • P. gingivalis solid media:
     Blood agar plates (BAPHK)

     • ▪
       Trypticase soy agar
     • ▪
       Defibrinated sheep’s blood (5.0% vol/vol)
     • ▪
       Hemin (5.0 μg/ml)
     • ▪
       Vitamin K (0.5 μg/ml)
   • P. gingivalis liquid culture:
     Brain-Heart Infusion Broth (BHIHKSbcStgC)

     • ▪
       Brain-Heart infusion
     • ▪
       Yeast extract (1.0 mg/ml)
     • ▪
       Hemin (5.0 μg/ml)
     • ▪
       Vitamin K (0.5 μg/ml)
     • ▪
       Sodium bicarbonate (1.0 mg/ml)
     • ▪
       Sodium thioglycolate (0.25 mg/ml)
     • ▪
       Cysteine (0.5 mg/ml)
   • P. gingivalis antibiotics Gentamicin (25–50 μg/ml) and Erythromycin (2–10 μg/ml)
   • E. coli Luria broth base (LB) or Luria agar
   • E. coli antibiotic Carbenicillin (50 μg/ml)

3. GasPak™ EZ Anaerobe Pouch System (BD Biosciences) (http://www.bdbiosciences.com/)

2.2 Nested Semi-Random PCR

1. DNeasy (Qiagen) (http://www.qiagen.com/)

2. Nuclease-free water (Ambion) (Invitrogen) (http://www.lifetechnologies.com/)

3. GoTaq DNA Polymerase (Promega) (http://www.promega.com/)

2.3 Library Preparation for Tn-seq (18–20)

1. 2 mL round-bottom microfuge tubes

2. Branson 450 sonifier

3. High intensity cup horn

4. Circulating water bath

5. TdT enzyme (Promega) (http://www.promega.com/)

6. 5× TdT reaction buffer (Promega) (http://www.promega.com/)

7. dideoxy CTP (Affymetrix) (www.affymetrix.com/)

8. dCTP

9. dNTP mix

10. Easy-A DNA polymerase mix (Agilent) (www.agilent.com/)

11. 10× Easy-A DNA polymerase reaction buffer (Agilent) (www.agilent.com/)

12. Performa gel filtration cartridge (Edge Biosystems) (http://www.edgebio.com/)

13. Thermocycler (BioRad) (http://www.bio-rad.com/)

14. Nanodrop/spectrophotometer (Thermo Scientific) (http://www.thermofisher.com/)

15. ChemiDoc XRS+ imaging system (BioRad) (http://www.bio-rad.com/)

16. Illumina Genome Analyzer II (Illumina) (http://www.illumina.com/)

17. ABI 3130XL DNA sequencer (Invitrogen) (http://www.lifetechnologies.com/)

3. Methods

Media and culture conditions (19)

Wild type (WT) Porphyromonas gingivalis strain ATCC 33277 [^Note1] was grown and maintained at 37°C under anaerobic conditions using the GasPak™ EZ Anaerobe Pouch System [^Note5]. Blood agar plates (BAPHK) containing trypticase soy agar supplemented with defibrinated sheep’s blood (5% vol/vol), hemin (5 μg/ml), and menadione (0.5 μg/ml) as well as brain-heart infusion broth (BHIHKS_bcS_tgC) containing brain-heart infusion, yeast extract (1 mg/ml), hemin (5 μg/ml), and menadione (0.5 μg/ml), sodium bicarbonate (1 mg/ml), sodium thioglycolate (0.25 mg/ml), and cysteine (0.5 mg/ml) were used for solid and liquid culture of P. gingivalis, respectively [^Note3]. Gentamicin (25–50 μg/ml) and erythromycin (2–10 μg/ml) were used when appropriate for prevention of contamination as well as isolation and maintenance of P. gingivalis mutants [^Note4].

Escherichia coli strain S17-1 λpir/pSAM_Bt was grown at 37°C under aerobic conditions in Luria broth base (LB) and Luria agar [^Note2]. Carbenicillin (50 μg/ml) was added for plasmid maintenance and to prevent contamination.

Transposon mutagenesis

WT P. gingivalis (strain ATCC 33277) was inoculated into brain-heart infusion broth (BHIHKS_bcS_tgC) without antibiotics. Broth cultures were grown to optical densities (OD₆₀₀) between 0.50 and 1.00. Escherichia coli strain S17-1 λpir containing the pSAM_Bt plasmid was grown to optical densities OD 0.50–1.00. Broth cultures were set up such that between a 5:1 and 10:1 ratio of P. gingivalis (recipient) to E. coli (donor) was achieved [^Notes9/10]. Although P. gingivalis is categorized as an obligate anaerobe it is able to survive without significant CFU loss (less than a log₁₀) for up to 6 hours under aerobic conditions when incubated alone on BAPHK at 37°C [^Note6].

The E. coli donor strain carrying the Mariner transposon on a suicide plasmid vector was conjugated with WT P. gingivalis using a bi-parental procedure where the E. coli donor strain and P. gingivalis recipient strain are cultured together on an agar plate to allow for plasmid transfer [^Note6]. Conjugation was carried out aerobically at 37°C for 5 hr [^Note6]. As P. gingivalis is naturally resistant to gentamicin, this antibiotic was used for selection against the donor E. coli following the conjugation [^Note4]. The transposon contains an erythromycin resistance gene (ermG) used to select for P. gingivalis transposon insertion mutants [^Note4].

1. Inoculate WT P. gingivalis from frozen stock onto BAPHK containing gentamicin (25 μg/ml) [^Note8]

2. Incubate the BAPHK in GasPak at 37°C for 5–7 days [^Note5]

3. Inoculate a single colony of the WT P. gingivalis from BAPHK into BHIHKS_bcS_tgC without antibiotics and incubate anaerobically in GasPak at 37°C overnight

4. Back-dilute the P. gingivalis overnight into BHIHKS_bcS_tgC without antibiotics in a 10:1 ratio [^Note9]

5. Grow anaerobically in GasPak at 37°C to OD₆₀₀ between 0.50 and 1.00

6. At the same time as step 3, plate E. coli strain S17-1 λpir containing the pSAM_Bt plasmid onto LB agar (carbenicillin 50 μg/ml); grow aerobically at 37°C

7. Inoculate LB broth (carbenicillin 50 μg/ml) with a single E. coli colony from the agar plate; grow to OD₆₀₀ 0.50 – 1.00 in 25 ml LB (carbenicillin 50 μg/ml); grow on a shaker, aerobically at 37°C

8. Remove fluid from the P. gingivalis and E. coli cultures in a 5:1 to 10:1 ratio of P. gingivalis (recipient) to E. coli (donor) and place into separate centrifuge tubes

9. Pellet each broth culture and resuspend in 2 ml BHIHKS_bcS_tgC

10. Plate four 500 μL puddles onto BAPHK without antibiotics [^Note6]

11. Incubate for 5–6 hours aerobically at 37°C

12. Resuspend conjugation reactions in a total of 25 mL PBS [^Note7]

13. Plate 100–400 μL aliquots onto BAPHK supplemented with gentamicin and erythromycin [^Note7]

14. Incubate anaerobically in Gaspaks at 37°C for at least 4 days, but up to 21 if desired [^Note11]

    [To determine transposition efficiency or for individual stocking]

15. Select individual colonies and transfer onto BAPHK containing erythromycin 5 μg/ml and gentamicin 25 μg/ml

16. Test individual P. gingivalis colonies using PCR to detect the presence of the transposon (ermG) as well as vector backbone components (bla and himar1C9a transposase) [Table 1] [Fig.1]

    [For pooling mutant colonies instead of maintaining stocks of individual mutant strains]

17. Scrape and resuspend mutant strain colonies from BAPHK into BHIHKS_bcS_tgC [^Note12]

Table 1.

Primers necessary for checking for parts of the mutagenesis vector in individual mutant strains, nested semi-random PCR, and Tn-seq. The ermG and bla primers are for the genes of pGERM (EF155418.1) and the himar1c9a primers are for the gene of pBADC9 (PMCID: PMC18050).

Primer	Sequence	Function
ermG +	TAGGTGCAGGGAAAGGTCAT	Vector
ermG −	CCATTTTTGCTGGCTTTCTT	Vector
bla +	TTGCCGGGAAGCTAGAGTAA	Vector
bla −	GCTATGTGGCGGTATTAT	Vector
himar1c9a +	GACGGAAAAACTCGGGTGTA	Vector
himar1c9a −	TTCAAGCGTGGTGAAATGAG	Vector
SAMseq1	ACGTACTCATGGTTCATCCCGATA	Semi-random
SAMseq2	GCGTATCGGTCTGTATATCAGCAA	Semi-random
SAMseq3	TCTATTCTCATCTTTCTGAGTCCAC	Semi-random
ARB1	GGCCACGCGTGCACTAGTACN10TACNG	Semi-random
ARB2	GGCCACGCGTGCACTAGTAC	Semi-random
Olj376	GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGGGGGGGGGGGGGGGG	Illumina
pSAM1	CCTGACGGATGGCCTTTTTGCGTTTCTACC	Illumina
pSAM2	AATGATACGGCGACCACCGAGATCTACACTCTTTGACCGGGGACTTATCATCCAACCTGTTA	Illumina
pSAM3	ACACTCTTTGACCGGGGACTTATCATCCAACCTGTTA	Illumina

Figure 1.

Map of the pSAM_Bt vector generated using Plasmapper. The three ORFs on the vector are denoted in pink. Antibiotic markers denoted in orange. The Amp and BT_1311 promoter regions are denoted in green. Unique restriction enzyme cut sites are labeled. (21)

Nested Semi-random PCR/Sequencing

While defining the library by Tn-seq can be done directly from the transformants, due to the time and expense of performing the sequencing repeatedly, we typically test the success of individual mutagenesis experiments by nested semi-random PCR. This allows us to quickly determine the success of an individual mutagenesis experiment in generating diverse insertions prior to adding the mutants to the library pool. This method can also be used when a mutant strain has been isolated from the pool and knowledge of the transposon insertion location is desired. Amplicons from the PCR reactions can be sequenced individually if desired, however, unless a mixture of strains has been prepared only one of the bands visible on the agarose gel following PCR will be recognized by the sequencing primer. In cases where a complete library has been tested by applying a selective pressure, nested semi-random PCR can be used to determine the complexity of the remaining mutants prior to subjecting to massively-parallel sequencing. If the nested semi-random PCR reveals limited remaining diversity of the surviving mutants, it may be more cost effective to sequence individual colonies.

• 1Prepare genomic DNA from selected mutants/clones as well as wild-type progenitor strain; done by kit based (Qiagen) isolation.
• 2
  Prepare the following PCR Reaction Mixture

  • 12.5 μl GoTaq Green Master Mix
  • 9.5 μl Nuclease-free dH₂O
  • 0.5 μl Primer ARB1 (GGCCACGCGTCGACTGTACN¹⁰….TACNG) [Table 1]
  • 0.5 μl Primer SAMSeq1 (ACGTACTCATGGTTCATCCCGATA) [Table 1]
  • 2.0 μl DNA template from genomic DNA
• 3
  First round of ‘random’ PCR:

  1. 95°C for 2 min
  2. 95°C for 1 min
  3. 30°C for 1 min
  4. 72°C for 1 min
  5. Repeat steps 2–4 six cycles
  6. 95°C for 1 min
  7. 55°C for 1 min
  8. 72°C for 1 min
  9. Repeat steps 6–8 for 30 cycles
  10. 72°C for 5 min
  11. 4°C HOLD

[Do not electrophorese products on agarose gel at this step; you will either not see anything or will have a ladder of DNA. You can choose to or not to PCR-purify the products from reaction 1 prior to reaction 2 at this point, but this is not necessary.]

• 4
  Second round of ‘random’ PCR: Prepare PCR reaction as above substituting ARB1 and SAMSeq1 for ARB2 and SAMSeq2

  • Primers [ARB2 and SAMSeq2]

    • (pSAMseq2 = GCGTATCGGTCTGTATATCAGCAA) [Table 1]
    • (ARB2 = GGCCACGCGTCGACTGTACN) [Table 1]

PCR Reaction 2

1. 95°C for 2 min

2. 95°C for 1 min

3. 55°C for 1 min

4. 72°C for 1 min

5. Repeat steps 2–4 for 40 cycles

6. 72°C for 5 min

7. 4°C HOLD

• Test 5 μl of PCR reaction 2 product on agarose gel [multiple bands expected] [Fig.2]

• PCR purify the remaining reaction round 2 PCR products using kit/kit protocol (Qiagen DNeasy)

• Send PCR products for sequencing using SAMSeq3 primer [Table 1]

• Use NCBI BLASTn [^Note13] program to search sequence (minus the transposon sequence) and map for genomic location to identify the transposon-chromosomal junction

Figure 2.

Nested semi-random PCR for sequencing/identification of individual mutants. PCR from individual colonies was carried out using primers specific to the transposon and random primers for the P. gingivalis genome. Agarose gel following second round of nested PCR shown above. Lanes 1 and 2 contain DNA ladder markers, lanes 3 through 5 are negative controls of WT P. gingivalis strain W83, template only and primer only, and individual mutants in the remaining lanes 6–15. Using SAMSeq3 primer for sequencing one will only receive a single product, even though there are multiple bands, unless the prepared gDNA contains two mutant strains. Of note, insertions within the same gene can give different banding patters, as similar banding patterns can be coincidental as well, so extrapolation of insertion location from PCR gel bands is not advisable.

Construction and sequencing of libraries

After confirming successful transposon mutagenesis, mutant library pools (or post-selection pools if experiments beyond the BAPHK mutagenesis have been performed) must be compiled and prepared for sequencing. Preparation for sequencing will include the elements necessary for bioinformatic analyses downstream of sequencing. Detailed methods specific for creation of DNA libraries for use with Tn-seq have been previously well described(18–20); below is a brief outline. For determining the in vitro blood agar based essential genes of P. gingivalis, a single-use mutant pool aliquot was processed using Tn-seq. First, genomic DNA from a mutant pool aliquot was sheared. Deoxycytidine triphosphate tails (C-tails) were then added to the sheared gDNA through PCR. Non-gDNA was removed during a gel filtration cleanup. Transposon containing fragments were amplified in a PCR reaction and adapters specific for sequencing and library identification were added. The Tn-seq-prepared mutant gDNA was run in a sequencing reaction using an Illumina sequencing platform.

We run between 8 and 12 samples per lane of the Illumina sequencer. Each sample is tagged with a unique barcode as part of the preparation for sequencing. Several dozen barcodes are available, which have been described by van Opijnen et al. and Gordon et al., as well as those available through Illumina and New England Biolabs. At the time of our initial publication in 2013, 8–12 samples in a lane resulted in an average of 5–10,000,000 reads per sample which was sufficient depth to provide reliable analysis for our purposies. Since then, Illumina reads per lane have increased dramatically, which would allow the inclusion of more bar codes per lane. However, this must be adjusted for the depth of sequencing required for the specific indication.

1. Prepare the mutant pool with a genomic DNA extraction kit/kit protocol (Qiagen DNeasy)

2. Elute genomic DNA with 100 μL elution buffer into a microfuge tube

3. Place eluent into a 2 mL round-bottom microfuge tube

4. Shear for 2 min in a Branson sonicator to generate fragments between 200–600 bp

5. Add poly C-tails to 1 μg of sheared DNA in a 20 μl reaction using TdT as per the manufacturer’s instructions (1hr incubation at 37°C)

6. Heat-inactivate at 75°C for 20 min

7. Remove ddCTP and other small molecules using a Performa gel filtration cartridge

8. Amplify transposon-containing fragments by PCR (primers pSAM1 and Olj376)

9. Remove PCR primers and other small molecules using a Performa gel filtration cartridge

10. Amplify the exact Tn-gDNA junction and add sequences needed for Illumina sequencing and indexing (primers: pSAM2 and indexing/barcoding)

11. Pool sequencing preparations if running more than one experiment per lane and run single-end sequencing reaction on Illumina GAII using pSAM3 and Illumina indexing primers

Data analysis

All read mapping and primary data analysis can be carried out on the Galaxy server in combination with Microsoft Office software (22). In-depth descriptions of specific scripts have been previously reported (and are either available on the open-access Galaxy server or can be generated from previous publications) (18–20). Sequencing output should be generated/received as either FASTQ or FASTA files. If received as FASTA, or potentially other formats, conversion to FASTQ prior to other downstream analyses will be required (and can be found on the Galaxy server under ‘NGS: QC and Manipulation’). Visualizations of Tn-seq data GenomeView allows for comparison of insertions and insertion frequency under multiple conditions (tracts) at once, can be customized manually, but also allows for overlay of NCBI (and other) documentation (23) [Fig.3].

1. Import FASTQ sequencing files (Galaxy→Get Data →Upload File)

2. Remove C-tails using adapter clipping script (Galaxy→NGS: QC and Manipulation →Clip)

3. Filter out low quality sequencing reads (Galaxy→NGS: QC and Manipulation→Filter By Quality)

4. Align clipped/filtered reads to reference genome (Galaxy→NGS: Mapping →Bowtie for Illumina)

5. Use bowtie output (SAM file) to tabulate insertion sites and instances an insertion site was sequenced

6. Generate Excel spreadsheet file indicating position in the genome, locus (if within a gene) that an insertion maps to, DNA strand associated with the insertion as well as the frequency of reads for each insertion site

7. Use Excel output to estimate complexity of a given transposon mutant library as well as to compare specific mutant strains in input and output samples

8. Convert Excel output file to BED file for visualization using GenomeView; alternatively, SAM file from step 5 can be converted to a BAM file (Galaxy→NGS: SAM Tools→Convert SAM-to-BAM) and then convert BAM file into a BED file (Galaxy→BED Tools→Convert BAM-to-BED)

   [If visualization of insertions/reads is desired]

9. Download NCBI GenBank/FASTA files of reference genome for GenomeView visualizations

10. Use BED files from step 8 as input for GenomeView visualizations

    [If attempting to determine or compare essentiality/conditional essentiality of a gene or intergenic region]

11. Retrieve nucleotide sequences of selected CDS using BROP [Note¹³]

12. Utilize BLASTn function in DEG (24) to determine putative essentially matches

Figure 3.

Using GenomeView for visualization of Tn-seq data. BED files (for Tn-seq data) and GenBank files (for NCBI) were uploaded to generate tracts, in which the insertion locus becomes an arrow and the numbers inside the arrow are the reads for that given insertions under the experimental condition. Top tract in blue are genes of the P. gingivalis ATCC 33277 genome from NCBI. Subsequent tracts are the BAPHK-based library followed by BHIHKS_bcS_tgC- based condition. Under these conditions, insertions into an ISPg3 element and tnaA stay relatively constant (top), while insertions into ragA/ragB –operon show a BAPHK-essential gene, one with reduced fitness under BHIHKS_bcS_tgC condition relative to BAPHK, as well as an unaffected gene (bottom).

Applications

There are many potential applications for transposon mutagenesis of a species followed by insertion site sequencing of the resultant mutant populations (14, 15, 25). Although for some species there have already been comprehensive, non-transposon mutagenesis-based essential gene studies (26–28), one can begin by attempting to define an in vitro essential gene list if no comprehensive clean deletion collection or saturated mutant library have been created in the species of interest. Beyond defining a set of genes that cannot be mutagenized under an in vitro condition, one can use a Tn-based mutant library to query genes and/or intergenic regions for relative ‘fitness’ under an in vitro or in vivo condition (16, 17, 29). In vitro conditions can include variations in growth medium nutrient compositions (carbohydrates, peptides, vitamins and minerals) [Fig.3], environmental stresses (temperature, salinity, oxygen levels and light) or host-derived immune stresses (complement components, antimicrobial peptides and phagocytosis). In vivo conditions can include colonization and survival within host-based niches as well as transit from host to host or reservoir to host.

Given that there had been no previous saturated mutant library or comprehensive deletion library created for P. gingivalis, we decided to use our transposon mutant library to define the in vitro (BAPHK-based) putatively essential genes of P. gingivalis (19). The genes identified during our study as essential have been deposited and curated in the DEG (24). An in silico bioinformatic analysis was performed using the P. gingivalis essential gene data as validation for the program GepTop, which is a tool for predicting gene essentiality in sequenced genomes (30).

Abbreviations

Pg

Porphyromonas gingivalis

PCR

Polymerase Chain Reaction

BHIHKS_bcS_tgC

supplemented Brain-Heart Infusion

BAPHK

supplemented Blood Agar Plate

BROP

Bioinformatics Resource Oral Pathogens

BED

Browser Extensible Data format

SAM

Sequence Alignment/Map format

BAM

Binary Alignment/Map format

BLAST

Basic Local Alignment Search Tool

WT

Wild-Type

DEG

Database of Essential Genes