Imaging Minimal Bacteria at the Nanoscale: a Reliable and Versatile Process to Perform Single-Molecule Localization Microscopy in Mycoplasmas

Abstract

Mycoplasmas are the smallest free-living organisms. These bacteria are important models for both fundamental and synthetic biology, owing to their highly reduced genomes. They are also relevant in the medical and veterinary fields, as they are pathogenic to both humans and most livestock species. Mycoplasma cells have minute sizes, often in the 300- to 800-nm range. As these dimensions are close to the diffraction limit of visible light, fluorescence imaging in mycoplasmas is often poorly informative. Recently developed superresolution imaging techniques can break this diffraction limit, improving the imaging resolution by an order of magnitude and offering a new nanoscale vision of the organization of these bacteria. These techniques have, however, not been applied to mycoplasmas before. Here, we describe an efficient and reliable protocol to perform single-molecule localization microscopy (SMLM) imaging in mycoplasmas. We provide a polyvalent transposon-based system to express the photoconvertible fluorescent protein mEos3.2, enabling photo-activated localization microscopy (PALM) in most Mycoplasma species. We also describe the application of direct stochastic optical reconstruction microscopy (dSTORM). We showcase the potential of these techniques by studying the subcellular localization of two proteins of interest. Our work highlights the benefits of state-of-the-art microscopy techniques for mycoplasmology and provides an incentive to further the development of SMLM strategies to study these organisms in the future. IMPORTANCE Mycoplasmas are important models in biology, as well as highly problematic pathogens in the medical and veterinary fields. The very small sizes of these bacteria, well below a micron, limits the usefulness of traditional fluorescence imaging methods, as their resolution limit is similar to the dimensions of the cells. Here, to bypass this issue, we established a set of state-of-the-art superresolution microscopy techniques in a wide range of Mycoplasma species. We describe two strategies: PALM, based on the expression of a specific photoconvertible fluorescent protein, and dSTORM, based on fluorophore-coupled antibody labeling. With these methods, we successfully performed single-molecule imaging of proteins of interest at the surface of the cells and in the cytoplasm, at lateral resolutions well below 50 nm. Our work paves the way toward a better understanding of mycoplasma biology through imaging of subcellular structures at the nanometer scale.

Keywords: Mycoplasma; PALM; dSTORM; single-molecule localization microscopy; superresolution microscopy.

FIG 1

Assessing the functionality of the photoconvertible fluorescent protein mEos3.2 in multiple Mycoplasma species. (A) Map of the plasmid pMT85-PSynMyco-mEos3.2. The main genetic components of the plasmid are indicated. IR, inverted repeat; tnpA, transposase; aacA-aphD, gentamicin resistance. (B) Distribution of the Mycoplasma species used in this study. A phylogenetic tree of representative Mollicutes species was inferred using the maximum-likelihood method from the concatenated multiple alignments of 79 proteins encoded by genes present at one copy in each genome (adapted from Grosjean et al. 2014 [99]). The main phylogenetic groups are indicated by gray boxes. The six species transformed with pMT85-PSynMyco-mEos3.2 are identified by red arrowheads. (C) Sample images of mycoplasma cells expressing mEos3.2 in their cytoplasm. For each of the six species, cells transformed with the plasmid pMT85-PSynMyco-mEos3.2 were imaged using PALM. A representative subset of the field of view is given (from left to right, phase-contrast image, superresolved reconstruction at 40-nm pixel size, and Tesseler segmentation of the localizations). Scale bars = 1 μm. (D) Quantification of the PALM signal intensity. For each of the six species, both wild-type (WT) and pMT85-PSynMyco-mEos3.2-transformed cells (mEos) were imaged by PALM, and the data collected from a single representative field of view (512 by 512 pixels; pixel size = 0.16 μm) were analyzed. The dot plot presents the number of detections measured in each object segmented by Tesseler (equivalent to a cell), on top of which a boxplot showing the median, interquartile range, minimum, and maximum values is overlaid. A statistical test (Mann-Whitney) was performed to compare the two conditions, WT and mEos. ***, P < 1.10⁻¹⁰.

FIG 2

PALM imaging of an F-type ATPase in Mycoplasma mycoides subsp. capri. M. mycoides subsp. capri mEos3.2-0575 cells, expressing a mEos-fused variant of the β-subunit of the ATPase F_1-like domain, were imaged by PALM. In this M. mycoides subsp. capri mutant, the fluorescent fusion protein is expressed from the native genomic locus and replaces the wild-type variant. The data presented here correspond to a single representative field of view (512 by 512 pixels; pixel size = 160 nm). (A) Sample images of M. mycoides subsp. capri mEos3.2-0575 cells. For each field of view, the images correspond to epifluorescence (diffraction limited) (top), superresolved reconstruction (40-nm pixel) (middle), and Tesseler segmentation (bottom). Scale bar = 1 μm. (B) Tesseler clustering of the fluorescence signal. The number of clusters per Tesseler-segmented object was computed. The bar graphs display the distribution of the numbers of clusters per object. (C) Object and cluster sizes. The dot plot presents the area (in nm²) of each object and cluster segmented by Tesseler, to which a boxplot showing the median, interquartile range, minimum, and maximum values is overlaid. The median value of each data set is indicated. (D) Evaluation of the PALM imaging pointing accuracy. Inset, example of the tracks computed using PALMTracer from which the MSD₀ and pointing accuracy values are derived. The bar graphs display the distribution of the pointing accuracies derived from each track. The median value of the data set is indicated.

FIG 3

dSTORM imaging of an antibody-specific protease in Mycoplasma mycoides subsp. capri. M. mycoides subsp. capri 0582-HA cells expressing an HA tag-fused variant of the serine protease MIP₈₂ were immunolabeled and imaged by dSTORM. The tagged fusion protein is expressed from the native genomic locus and replaces the wild-type variant. The data presented here correspond to a single representative field of view (512 by 512 pixels; pixel size = 160 nm). (A) Sample images of M. mycoides subsp. capri 0582-HA cells. For each field of view, the images correspond to epifluorescence (diffraction limited) (left), superresolved reconstruction (40 nm pixel) (middle), and Tesseler segmentation (right). Scale bar = 1 μm. (B) Tesseler clustering of the fluorescence signal. For each field of view, the number of clusters per Tesseler-segmented object was computed. The bar graphs display the distribution of the numbers of clusters per object. (C) Object and cluster sizes. The dot plot presents the area (in nm²) of each object and cluster segmented by Tesseler, to which a boxplot showing the median, interquartile range, minimum, and maximum values is overlaid. The median value of each data set is indicated.

FIG 4

PALM/dSTORM two-color imaging of Mycoplasma mycoides subsp. capri. Sample images of M. mycoides subsp. capri 0582-HA pMT85-PSynMyco-mEos3.2 cells, expressing both an HA tag-fused variant of the serine protease MIP₀₅₈₂ and the fluorescent protein mEos3.2. The tagged fusion protein is expressed from the native genomic locus and replaces the wild-type variant. mEos3.2 is expressed from a transposon inserted at a random site in the bacterial chromosome. For each field of view, the images correspond to a reconstructed PALM image (40-nm pixel) (left), a reconstructed dSTORM image (40-nm pixel) (middle), and an overlay of the reconstructed PALM and dSTORM images (right). Scale bar = 1 μm. All the images were sampled from the same coverslip and field of view.

FIG 5

Evaluation of promoter strength and cell size through PALM imaging in M. mycoides subsp. capri. (A) Comparison of promoter strength by PALM imaging. M. mycoides subsp. capri cells, either wild type (WT) or transformed with plasmid pMT85-P438-mEos3.2 (promoter P438), pMT85-PSynMyco-mEos3.2 (promoter PSynMyco), or pMT85-PSpi-mEos3.2 (promoter PSpi), were imaged by PALM. For each strain, the data collected from a single representative field of view (512 by 512 pixels; pixel size = 0.16 μm) were analyzed. The dot plot presents the number of detections measured in each object segmented by Tesseler (equivalent to a cell), on top of which a boxplot showing the median, interquartile range, minimum, and maximum values is overlaid. Statistics tests (Mann-Whitney) were performed to compare the four strains (*, P < 0.05; ***, P < 1.10⁻¹⁰). (B) Deriving cell size data from PALM images. M. mycoides subsp. capri cells transformed with plasmid pMT85-PSpi-mEos3.2 were imaged by PALM. The data collected from a single representative field of view (512 by 512 pixels; pixel size = 0.16 μm) were analyzed. The dot plot presents the dimensions (in nm) of the major axis and the minor axis of each object segmented by Tesseler (Fig. S7), on top of which a boxplot showing the median, interquartile range, minimum, and maximum values is overlaid.