methylMapR—an R package to generate the functional prokaryotic methylome

TL;DR

methylMapR is an R package that helps analyze prokaryotic methylomes using PacBio sequencing data for three bacterial species.

Contribution

The novel contribution is the development of methylMapR for functional methylome analysis in prokaryotes.

Findings

methylMapR enables functional methylome analysis from PacBio sequencing data.

The package was applied to Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa.

Comparative analyses revealed insights into prokaryotic methylation patterns.

Abstract

Here, we present an R package called methylMapR to capture the functional methylome from long read sequencing of prokaryotic genomes using the PacBio sequencing platform. We then describe its utility by comparative analyses of the functional methylomes in three bacterial species—Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa.

Figures

A flowchart showing inputs and outputs generated by functions in the methylMapR R package.

Tables

TABLE 1. Methylation/transcription summary metrics estimated from the methylMapR pipeline in three prokaryotic species—E. coli, K. pneumoniae, and P. aeruginosa, by methylation typea

Prokaryote	IPD ratio at methylation target bases and all other bases	Number and proportion of methylation types	Methylation motif/TFBS crowdedness (10 bp window, both strands) Avg and Max	Number and proportion of promoter region-associated motifs	TFBS/methyl interaction types – ratio of promoting to repressive
E. coli	m4C: 1.0657m5C: 1.1643m6A: 1.1945modified base: 1.0656non-target base: 1.0679	m4C: 3215, 0.0236m5C: 25187, 0.1848m6A: 40168, 0.2947modified base: 67721, 0.4969	m4C: 1.4325, 4m5C: 1.4406, 6m6A: 1.4075, 4modified base: 1.4324, 6	m4C: 27, 0.0260m5C: 212, 0.2040m6A: 325, 0.3128modified base: 475, 0.4572	m4C: 62:445, 0.1393m5C: 784:6381, 0.1229m6A: 1592:11919, 0.1336modified base: 941:7380, 0.1275
K. pneumoniae	m4C: 1.0137m6A: 1.1001modified base: 1.0398non-target base: 0.9735	m4C: 3464, 0.0297m6A: 77724,0.6666modified base: 35415, 0.3037	m4C: 2.0370, 6m6A: 2.030791, 6modified base: 2.059907, 6	m4C: 50, 0.0302m6A: 1041,0.6286modified base: 565, 0.3412	m4C: 21:93, 0.2258m6A: 705:4622, 0.1525modified base: 184:1275, 0.1443
P. aeruginosa	m4C: 1.0523m6A: 1.0018modified base: 1.0403non-target base: 1.0143	m4C: 3126, 0.0398m6A: 21121, 0.2692modified base: 54216, 0.6910	m4C: 1.0000, 1m6A: 1.0364, 2modified base: 1.0159, 2	m4C: 19, 0.0294m6A: 178, 0.2755modified base: 449, 0.6950	m4C: 0:0, 0m6A: 2:3, 0.6667modified base: 4:5, 0.8000

Full text

ANNOUNCEMENT

DNA methylation has been studied extensively in transcription regulation and gene expression across prokaryotes (1–4). A shared aspect of prokaryotic DNA methylation is that DNA methyltransferase (DNMT) requires the methyl donor molecule, S-adenosyl methionine, and covalently transfers the heritable epigenetic methyl tag to a base within its target motif (5). During binary fission, each daughter chromosome is hemi-methylated with the daughter strand retaining methylation patterns from its parent chromosome. The nascent strand remains unmethylated, needing to be remethylated by DNMT (6). Active gene processes like transcription can block or modulate the methylation reaction during re-methylation events (7). Conversely, methylation has been shown to both promote and repress transcription. Some technologies like PacBio Sequencing, henceforth PBS, and its associated software, kineticsTools, can monitor polymerase kinetics by tracking the time of incorporation of each new base on the nascent strand and estimate the interpulse duration ratio (IPD), the likelihood that a methyl functional group is present at that base (8, 9). A high IPD ratio is indicative of a slower incorporation rate by DNA polymerase, which can be caused by the presence of a methyl group (10, 11). Here, we present methylMapR, a tool that leverages the IPD ratio data from PBS and combines it with other functional genomic data to describe the functional prokaryotic methylome.

methylMapR implements several mapping functions that calculate new metrics and/or add columns to the output data frames, including assigning transcription factor methylation interaction type, finding promoter-associated motif sites, and calculating the number of TFBSs near each motif (Fig. 1). We utilized and compared functional methylomes from publicly available NCBI repositories for three well-studied prokaryotes, Escherichia coli, specifically the laboratory-born K12 strain, Klebsiella pneumoniae, and Pseudomonas aeruginosa. We acknowledge that none of these data comes from a single source or experiment(s). Our analyses here were therefore designed to help compare the power of methylMapR across multiple genera.

A flowchart showing inputs and outputs generated by functions in the methylMapR R package.

A comparison of the output from methylMapR and summary metrics of the three functional methylomes regarding transcription is presented in Table 1. The global IPD ratio observed across methylomes suggested that E. coli had the highest average IPD, for all known methylation types (m4C: 1.0657, m5C: 1.1643, and m6A: 1.1945). Global IPD ratios in both K. pneumoniae and P. aeruginosa were lower, exhibiting 1.1001 and 1.0018 IPD ratio at target adenines with a m6A motif, respectively. Comparing the types and proportions of motifs that were methylated, we observe one main difference between the isolates*—E. coli* has an extra methyl transferase/motif type (m5C) that the other two prokaryotes do not. Additionally, we uncovered differences in proportions of detected methylation motifs across all species. K pneumoniae shows over 60% of its methylation motifs are from the m6A tag. The m4C tag, though present in all species, accounts for less than 5% of all motifs in all three methylomes. All methylomes also show interactions with transcription factor binding sites, and the repressive methylation type was the most common interaction predicted. None of the methylomes had a methylation promotion rate above 20% at all methylated motif sites.

The methylMapR R package is available at https://github.com/therealtotodile/methylMapR/ and was tested on R 4.4.2 GUI 1.81 Big Sur ARM build (8462). To install methylMapR, source code, example data sets and code outputs, users can download the .gz to run on MacOSX, Windows, or Linux. From RStudio, users can go to Tools->Install packages, and under Install From , select the methylMapR Package Archive File (.tar.gz), then click install. R code to replicate analyses described in this manuscript is available via FigShare DOI: 10.6084/m9.figshare.2858581710.6084/m9.figshare.28585817.